This paper explores the application of artificial neural networks for online identification of a multimachine power system. A recurrent neural network has been proposed as the identifier of the two area, four machine system which is a benchmark system for studying electromechanical oscillations in multimachine power systems. This neural identifier is trained using the static Backpropagation algorithm. The emphasis of the paper is on investigating the performance of the variants of the Backpropagation algorithm in training the neural identifier. The paper also compares the performances of the neural identifiers trained using variants of the Backpropagation algorithm over a wide range of operating conditions. The simulation results establish a satisfactory performance of the trained neural identifiers in identification of the test power system.
Neural network based identification of multimachine power systemcsandit
In recent years, the golden codes have proven to exhibit a superior performance in a wireless
MIMO (Multiple Input Multiple Output) scenario than any other code. However, a serious
limitation associated with it is its increased decoding complexity. This paper attempts to resolve
this challenge through suitable modification of golden code such that a less complex sphere
decoder could be used without much compromising the error rates. In this paper, a minimum
polynomial equation is introduced to obtain a reduced golden ratio (RGR) number for golden
code which demands only for a low complexity decoding procedure. One of the attractive
approaches used in this paper is that the effective channel matrix has been exploited to perform
a single symbol wise decoding instead of grouped symbols using a sphere decoder with tree
search algorithm. It has been observed that the low decoding complexity of O (q1.5) is obtained
against conventional method of O (q2.5). Simulation analysis envisages that in addition to
reduced decoding, improved error rates is also obtained.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
International Journal of Computational Engineering Research(IJCER) is an intentional online Journal in English monthly publishing journal. This Journal publish original research work that contributes significantly to further the scientific knowledge in engineering and Technology
ARTIFICIAL NEURAL NETWORK BASED POWER SYSTEM STABILIZER ON A SINGLE MACHINE ...EEIJ journal
In this paper the use of artificial neural network in power system stability is studied. A predictive
controller based on two neural networks is designed and tested on a single machine infinite bus system
which is used to replace conventional power system stabilizers. They have been used for decades in
power system to dampen small amplitude low frequency oscillation in power systems. The increases in
size and complexity of power systems have cast a shadow on efficiency of conventional method. New
control strategies have been proposed in many researches. Artificial Neural Networks have been studied
in many publications but lack of assurance of their functionality has hindered the practical usage of them
in utilities. The proposed control structure is modelled using a novel data exchange established between
MATLAB and DIgSILENT power factory. The result of simulation proves the efficiency of the proposed
structure.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
Neural network based identification of multimachine power systemcsandit
In recent years, the golden codes have proven to exhibit a superior performance in a wireless
MIMO (Multiple Input Multiple Output) scenario than any other code. However, a serious
limitation associated with it is its increased decoding complexity. This paper attempts to resolve
this challenge through suitable modification of golden code such that a less complex sphere
decoder could be used without much compromising the error rates. In this paper, a minimum
polynomial equation is introduced to obtain a reduced golden ratio (RGR) number for golden
code which demands only for a low complexity decoding procedure. One of the attractive
approaches used in this paper is that the effective channel matrix has been exploited to perform
a single symbol wise decoding instead of grouped symbols using a sphere decoder with tree
search algorithm. It has been observed that the low decoding complexity of O (q1.5) is obtained
against conventional method of O (q2.5). Simulation analysis envisages that in addition to
reduced decoding, improved error rates is also obtained.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
International Journal of Computational Engineering Research(IJCER) is an intentional online Journal in English monthly publishing journal. This Journal publish original research work that contributes significantly to further the scientific knowledge in engineering and Technology
ARTIFICIAL NEURAL NETWORK BASED POWER SYSTEM STABILIZER ON A SINGLE MACHINE ...EEIJ journal
In this paper the use of artificial neural network in power system stability is studied. A predictive
controller based on two neural networks is designed and tested on a single machine infinite bus system
which is used to replace conventional power system stabilizers. They have been used for decades in
power system to dampen small amplitude low frequency oscillation in power systems. The increases in
size and complexity of power systems have cast a shadow on efficiency of conventional method. New
control strategies have been proposed in many researches. Artificial Neural Networks have been studied
in many publications but lack of assurance of their functionality has hindered the practical usage of them
in utilities. The proposed control structure is modelled using a novel data exchange established between
MATLAB and DIgSILENT power factory. The result of simulation proves the efficiency of the proposed
structure.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Performance assessment of an optimization strategy proposed for power systemsTELKOMNIKA JOURNAL
In the present article, the selection process of the topology of an artificial neural network (ANN) as well as its configuration are exposed. The ANN was adapted to work with the Newton Raphson (NR) method for the calculation of power flow and voltage optimization in the PQ nodes of a 10-node power system represented by the IEEE 1250 standard system. The purpose is to assess and compare its results with the ones obtained by implementing ant colony and genetic algorithms in the optimization of the same system. As a result, it is stated that the voltages in all system nodes surpass 0,99 p.u., thus representing a 20% increase in the optimal scenario, where the algorithm took 30 seconds, of which 9 seconds were used in the training and validation processes of the ANN.
HEURISTIC BASED OPTIMAL PMU ROUTING IN KPTCL POWER GRIDIAEME Publication
Power system monitoring is an important process in an efficient smart grid. The control centers used in smart grid requires restructuring. State measurements rather than state estimationare pre-requisite for the modern control center. The Phasor Measurement Unit (PMU) measures the synchronized voltage and current parameters. Placement of minimum number of PMUs in a bus system such that the wholes system becomes observable is considered as Optimal PMU Placement (OPP) problem. In this paper, Hybrid Distance Optimization (HDO) algorithm is proposed to reduce the number of PMUs for complete observability along with the minimum length of fiber optic cable required for interconnecting the PMU nodes
Short Term Load Forecasting: One Week (With & Without Weekend) Using Artifici...IJLT EMAS
This paper present for analysis of short term load forecasting: one week (with & without weekend) using ANN techniques for SLDC of Gujarat. In this paper short term electric load forecasting using neural network; based on historical load demand, The Levenberg-Marquardt optimization technique which has one of the best learning rates was used as a back propagation algorithm for the Multilayer Feed Forward ANN model using MATLAB.12 ANN tool box. Design a model for one week (with & w/o weekend) load pattern for STLF using the neural network have been input variables are (Min., Avg., & Max. load demands for previous week, Min., Avg., & Max. temperature for previous week & Min., Avg., & Max. humidity for previous week). And Nov-12 to Apr-13 (6 Months) historical load data from the SLDC, Gujarat are used for training, testing and showing the good performance. Using this ANN model computing the mean absolute error between the exact and predicted values, we were able to obtain an absolute mean error within specified limit and regression value close to one. This represents a high degree of accuracy.
As Wireless Sensor Networks are penetrating into the industrial domain, many research opportunities are emerging. One such essential and challenging application is that of node localization. A feed-forward neural network based methodology is adopted in this paper. The Received Signal Strength Indicator (RSSI) values of the anchor node beacons are used. The number of anchor nodes and their configurations has an impact on the accuracy of the localization system, which is also addressed in this paper. Five different training algorithms are evaluated to find the training algorithm that gives the best result. The multi-layer Perceptron (MLP) neural network model was trained using Matlab. In order to evaluate the performance of the proposed method in real time, the model obtained was then implemented on the Arduino microcontroller. With four anchor nodes, an average 2D localization error of 0.2953 m has been achieved with a 12-12-2 neural network structure. The proposed method can also be implemented on any other embedded microcontroller system.
Mobile Network Coverage Determination at 900MHz for Abuja Rural Areas using A...ijtsrd
This study proposes Artificial Neural Network ANN based field strength prediction models for the rural areas of Abuja, the federal capital territory of Nigeria. The ANN based models were created on bases of the Generalized Regression Neural network GRNN and the Multi Layer Perceptron Neural Network MLP NN . These networks were created, trained and tested for field strength prediction using received power data recorded at 900MHz from multiple Base Transceiver Stations BTSs distributed across the rural areas. Results indicate that the GRNN and MLP NN based models with Root Mean Squared Error RMSE values of 4.78dBm and 5.56dBm respectively, offer significant improvement over the empirical Hata Okumura counterpart, which overestimates the signal strength by an RMSE value of 20.17dBm. Deme C. Abraham ""Mobile Network Coverage Determination at 900MHz for Abuja Rural Areas using Artificial Neural Networks"" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-2 , February 2020,
URL: https://www.ijtsrd.com/papers/ijtsrd30228.pdf
Paper Url : https://www.ijtsrd.com/computer-science/artificial-intelligence/30228/mobile-network-coverage-determination-at-900mhz-for-abuja-rural-areas-using-artificial-neural-networks/deme-c-abraham
Implementation of Feed Forward Neural Network for Classification by Education...ijsrd.com
in the last few years, the electronic devices production field has witness a great revolution by having the new birth of the extraordinary FPGA (Field Programmable Gate Array) family platforms. These platforms are the optimum and best choice for the modern digital systems now a day. The parallel structure of a neural network makes it potentially fast for the computation of certain tasks. The same feature makes a neural network well suited for implementation in VLSI technology. In this paper a hardware design of an artificial neural network on Field Programmable Gate Arrays (FPGA) is presented. Digital system architecture is designed to realize a feed forward multilayer neural network. The designed architecture is described using Very High Speed Integrated Circuits Hardware Description Language (VHDL).General Terms-Network.
NEURAL NETWORK BASED IDENTIFICATION OF MULTIMACHINE POWER SYSTEMcscpconf
This paper demonstrates an effective application of artificial neural networks for online identification of a multimachine power system. The paper presents a recurrent neural network as the identifier of the benchmark two area, four machine system. This neural identifier is trained using the static Backpropagation algorithm. The trained neural identifier is then tested using datasets generated by simulating the system under consideration at different operating
points and a different loading condition. The test results clearly establish a satisfactory performance of the trained neural identifier in identification of the power system considered.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
A novel efficient adaptive-neuro fuzzy inference system control based smart ...IJECEIAES
A novel adaptive-neuro fuzzy inference system (ANFIS) control algorithm-based smart grid to solve power quality issues is investigated in this paper. To improve the steady-state and transient response of the solar-wind and grid integrated system proposed ANFIS controller works very well. Fuzzy maximum power point tracking (MPPT) algorithm-based DC-DC converters are utilized to extract maximum power from solar. A permanent magnet synchronous generator (PMSG) is employed to get maximum power from wind. To maximize both power generations, back-to-back voltage source converters (VSC) are operated with an intelligent ANFIS controller. Optimal power converters are adopted this proposed methodology and improved the overall performance of the system to an acceptable limit. The simulation results are obtained for a different mode of smart grid and non-linear fault conditions and the proven proposed control algorithm works well.
ARTIFICIAL NEURAL NETWORK BASED POWER SYSTEM STABILIZER ON A SINGLE MACHINE ...EEIJ journal
In this paper the use of artificial neural network in power system stability is studied. A predictive
controller based on two neural networks is designed and tested on a single machine infinite bus system
which is used to replace conventional power system stabilizers. They have been used for decades in
power system to dampen small amplitude low frequency oscillation in power systems. The increases in
size and complexity of power systems have cast a shadow on efficiency of conventional method. New
control strategies have been proposed in many researches. Artificial Neural Networks have been studied
in many publications but lack of assurance of their functionality has hindered the practical usage of them
in utilities. The proposed control structure is modelled using a novel data exchange established between
MATLAB and DIgSILENT power factory. The result of simulation proves the efficiency of the proposed
structure.
Power system transient stability margin estimation using artificial neural ne...elelijjournal
This paper presents a methodology for estimating the normalized transient stability margin by using the multilayered perceptron (MLP) neural network. The complex relationship between the input variables and output variables is established by using the neural networks. The nonlinear mapping relation between the normalized transient stability margin and the operating conditions of the power system is established by using the MLP neural network. To obtain the training set of the neural network the potential energy boundary surface (PEBS) method along with time domain simulation method is used. The proposed method is applied on IEEE 9 bus system and the results shows that the proposed method provides fast and accurate tool to assess online transient stability.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Performance assessment of an optimization strategy proposed for power systemsTELKOMNIKA JOURNAL
In the present article, the selection process of the topology of an artificial neural network (ANN) as well as its configuration are exposed. The ANN was adapted to work with the Newton Raphson (NR) method for the calculation of power flow and voltage optimization in the PQ nodes of a 10-node power system represented by the IEEE 1250 standard system. The purpose is to assess and compare its results with the ones obtained by implementing ant colony and genetic algorithms in the optimization of the same system. As a result, it is stated that the voltages in all system nodes surpass 0,99 p.u., thus representing a 20% increase in the optimal scenario, where the algorithm took 30 seconds, of which 9 seconds were used in the training and validation processes of the ANN.
HEURISTIC BASED OPTIMAL PMU ROUTING IN KPTCL POWER GRIDIAEME Publication
Power system monitoring is an important process in an efficient smart grid. The control centers used in smart grid requires restructuring. State measurements rather than state estimationare pre-requisite for the modern control center. The Phasor Measurement Unit (PMU) measures the synchronized voltage and current parameters. Placement of minimum number of PMUs in a bus system such that the wholes system becomes observable is considered as Optimal PMU Placement (OPP) problem. In this paper, Hybrid Distance Optimization (HDO) algorithm is proposed to reduce the number of PMUs for complete observability along with the minimum length of fiber optic cable required for interconnecting the PMU nodes
Short Term Load Forecasting: One Week (With & Without Weekend) Using Artifici...IJLT EMAS
This paper present for analysis of short term load forecasting: one week (with & without weekend) using ANN techniques for SLDC of Gujarat. In this paper short term electric load forecasting using neural network; based on historical load demand, The Levenberg-Marquardt optimization technique which has one of the best learning rates was used as a back propagation algorithm for the Multilayer Feed Forward ANN model using MATLAB.12 ANN tool box. Design a model for one week (with & w/o weekend) load pattern for STLF using the neural network have been input variables are (Min., Avg., & Max. load demands for previous week, Min., Avg., & Max. temperature for previous week & Min., Avg., & Max. humidity for previous week). And Nov-12 to Apr-13 (6 Months) historical load data from the SLDC, Gujarat are used for training, testing and showing the good performance. Using this ANN model computing the mean absolute error between the exact and predicted values, we were able to obtain an absolute mean error within specified limit and regression value close to one. This represents a high degree of accuracy.
As Wireless Sensor Networks are penetrating into the industrial domain, many research opportunities are emerging. One such essential and challenging application is that of node localization. A feed-forward neural network based methodology is adopted in this paper. The Received Signal Strength Indicator (RSSI) values of the anchor node beacons are used. The number of anchor nodes and their configurations has an impact on the accuracy of the localization system, which is also addressed in this paper. Five different training algorithms are evaluated to find the training algorithm that gives the best result. The multi-layer Perceptron (MLP) neural network model was trained using Matlab. In order to evaluate the performance of the proposed method in real time, the model obtained was then implemented on the Arduino microcontroller. With four anchor nodes, an average 2D localization error of 0.2953 m has been achieved with a 12-12-2 neural network structure. The proposed method can also be implemented on any other embedded microcontroller system.
Mobile Network Coverage Determination at 900MHz for Abuja Rural Areas using A...ijtsrd
This study proposes Artificial Neural Network ANN based field strength prediction models for the rural areas of Abuja, the federal capital territory of Nigeria. The ANN based models were created on bases of the Generalized Regression Neural network GRNN and the Multi Layer Perceptron Neural Network MLP NN . These networks were created, trained and tested for field strength prediction using received power data recorded at 900MHz from multiple Base Transceiver Stations BTSs distributed across the rural areas. Results indicate that the GRNN and MLP NN based models with Root Mean Squared Error RMSE values of 4.78dBm and 5.56dBm respectively, offer significant improvement over the empirical Hata Okumura counterpart, which overestimates the signal strength by an RMSE value of 20.17dBm. Deme C. Abraham ""Mobile Network Coverage Determination at 900MHz for Abuja Rural Areas using Artificial Neural Networks"" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-2 , February 2020,
URL: https://www.ijtsrd.com/papers/ijtsrd30228.pdf
Paper Url : https://www.ijtsrd.com/computer-science/artificial-intelligence/30228/mobile-network-coverage-determination-at-900mhz-for-abuja-rural-areas-using-artificial-neural-networks/deme-c-abraham
Implementation of Feed Forward Neural Network for Classification by Education...ijsrd.com
in the last few years, the electronic devices production field has witness a great revolution by having the new birth of the extraordinary FPGA (Field Programmable Gate Array) family platforms. These platforms are the optimum and best choice for the modern digital systems now a day. The parallel structure of a neural network makes it potentially fast for the computation of certain tasks. The same feature makes a neural network well suited for implementation in VLSI technology. In this paper a hardware design of an artificial neural network on Field Programmable Gate Arrays (FPGA) is presented. Digital system architecture is designed to realize a feed forward multilayer neural network. The designed architecture is described using Very High Speed Integrated Circuits Hardware Description Language (VHDL).General Terms-Network.
NEURAL NETWORK BASED IDENTIFICATION OF MULTIMACHINE POWER SYSTEMcscpconf
This paper demonstrates an effective application of artificial neural networks for online identification of a multimachine power system. The paper presents a recurrent neural network as the identifier of the benchmark two area, four machine system. This neural identifier is trained using the static Backpropagation algorithm. The trained neural identifier is then tested using datasets generated by simulating the system under consideration at different operating
points and a different loading condition. The test results clearly establish a satisfactory performance of the trained neural identifier in identification of the power system considered.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
A novel efficient adaptive-neuro fuzzy inference system control based smart ...IJECEIAES
A novel adaptive-neuro fuzzy inference system (ANFIS) control algorithm-based smart grid to solve power quality issues is investigated in this paper. To improve the steady-state and transient response of the solar-wind and grid integrated system proposed ANFIS controller works very well. Fuzzy maximum power point tracking (MPPT) algorithm-based DC-DC converters are utilized to extract maximum power from solar. A permanent magnet synchronous generator (PMSG) is employed to get maximum power from wind. To maximize both power generations, back-to-back voltage source converters (VSC) are operated with an intelligent ANFIS controller. Optimal power converters are adopted this proposed methodology and improved the overall performance of the system to an acceptable limit. The simulation results are obtained for a different mode of smart grid and non-linear fault conditions and the proven proposed control algorithm works well.
ARTIFICIAL NEURAL NETWORK BASED POWER SYSTEM STABILIZER ON A SINGLE MACHINE ...EEIJ journal
In this paper the use of artificial neural network in power system stability is studied. A predictive
controller based on two neural networks is designed and tested on a single machine infinite bus system
which is used to replace conventional power system stabilizers. They have been used for decades in
power system to dampen small amplitude low frequency oscillation in power systems. The increases in
size and complexity of power systems have cast a shadow on efficiency of conventional method. New
control strategies have been proposed in many researches. Artificial Neural Networks have been studied
in many publications but lack of assurance of their functionality has hindered the practical usage of them
in utilities. The proposed control structure is modelled using a novel data exchange established between
MATLAB and DIgSILENT power factory. The result of simulation proves the efficiency of the proposed
structure.
Power system transient stability margin estimation using artificial neural ne...elelijjournal
This paper presents a methodology for estimating the normalized transient stability margin by using the multilayered perceptron (MLP) neural network. The complex relationship between the input variables and output variables is established by using the neural networks. The nonlinear mapping relation between the normalized transient stability margin and the operating conditions of the power system is established by using the MLP neural network. To obtain the training set of the neural network the potential energy boundary surface (PEBS) method along with time domain simulation method is used. The proposed method is applied on IEEE 9 bus system and the results shows that the proposed method provides fast and accurate tool to assess online transient stability.
Neural Network-Based Stabilizer for the Improvement of Power System Dynamic P...TELKOMNIKA JOURNAL
This paper develops an adaptive control coordination scheme for power system stabilizers (PSSs)
to improve the oscillation damping and dynamic performance of interconnected multimachine power
system. The scheme was based on the use of a neural network which identifies online the optimal
controller parameters. The inputs to the neural network include the active- and reactive- power of the
synchronous generators which represent the power loading on the system, and elements of the reduced
nodal impedance matrix for representing the power system configuration. The outputs of the neural
network were the parameters of the PSSs which lead to optimal oscillation damping for the prevailing
system configuration and operating condition. For a representative power system, the neural network has
been trained and tested for a wide range of credible operating conditions and contingencies. Both
eigenvalue calculations and time-domain simulations were used in the testing and verification of the
performance of the neural network-based stabilizer.
Transient Stability Assessment and Enhancement in Power SystemIJMER
Power system is subjected to sudden changes in load levels. Stability is an important concept
which determines the stable operation of power system. For the improvement of transient stability the
general methods adopted are fast acting exciters, circuit breakers and reduction in system transfer
reactance. The modern trend is to employ FACTS devices in the existing system for effective utilization
of existing transmission resources. The critical clearing time is a measure to assess transient instability.
Using PSAT, the critical clearing time (CCT) corresponding to various faults are calculated. The most
critical faults were identified using this calculation. The CCT for the critical faults were found to change
with change in operating point. The CCT values are predicted using Artificial Neural Network (ANN) to
study the training effects of ANN. TCSC is selected as the FACTS device for transient stability
enhancement. Particle Swarm Optimization method is used to find the optimal position of TCSC using
the objective function real power loss minimization. The result shows that the technique effectively
increases the transient stability of the system
New artificial neural network design for Chua chaotic system prediction usin...IJECEIAES
This study aims to design a new architecture of the artificial neural networks (ANNs) using the Xilinx system generator (XSG) and its hardware co-simulation equivalent model using field programmable gate array (FPGA) to predict the behavior of Chua’s chaotic system and use it in hiding information. The work proposed consists of two main sections. In the first section, MATLAB R2016a was used to build a 3×4×3 feed forward neural network (FFNN). The training results demonstrate that FFNN training in the Bayesian regulation algorithm is sufficiently accurate to directly implement. The second section demonstrates the hardware implementation of the network with the XSG on the Xilinx artix7 xc7a100t-1csg324 chip. Finally, the message was first encrypted using a dynamic Chua system and then decrypted using ANN’s chaotic dynamics. ANN models were developed to implement hardware in the FPGA system using the IEEE 754 Single precision floating-point format. The ANN design method illustrated can be extended to other chaotic systems in general.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Calculating voltage magnitudes and voltage phase angles of real electrical ne...IJECEIAES
In the field of electrical network, it is necessary, under different conditions, to learn about the behavior of the system. Power Flow Analysis is the tool per excellent that allow as to make a deep study and define all quantities of each bus of the system. To determine power flow analysis there is a lot of methods, we have either numerical or intelligent techniques. Lately, researchers always work on finding intelligent methods that allow them to solve their complex problems. The goal of this article is to compare two intelligent methods that are capable of predicting quantities; artificial neural network and adaptive neuro-fuzzy inference system using real electrical networks. To do that we used few significant discrepancies. These methods are characterized by giving results in real time. To make this comparison successful, we implemented these two methods, to predict the voltage magnitudes and the voltage phase angles, on two Moroccan electrical networks. The results of the comparison show that the method of adaptive neuro-fuzzy inference system have more advantages than the method of artificial neural network.
An algorithm for fault node recovery of wireless sensor networkeSAT Publishing House
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
This paper proposes an improvement of the direct power control (DPC) scheme of a grid connected three phase voltage source inverter based on artificial neural networks (ANN) and fuzzy logic (FL) techniques for the renewable energy applications. This advanced control strategy is based on two intelligent operations, the first one is the replacement of the conventional switching table of a three phase voltage source inverter (VSI) by a selector based on artificial neural networks approach, and the second one is the replacement of the hysteresis comparators by fuzzy logic controllers for the instantaneous active and reactive power errors. These operations enable to reduce the power ripples, the harmonic disturbances and increase the response time period of the system. Finally, the simulation results were obtained by Matlab/Simulink environment, under a unity power factor (UPF). These results verify the transient performances, the validity and the efficiency of the proposed DPC scheme.
International Journal of Computational Engineering Research(IJCER)ijceronline
International Journal of Computational Engineering Research(IJCER) is an intentional online Journal in English monthly publishing journal. This Journal publish original research work that contributes significantly to further the scientific knowledge in engineering and Technology.
AN OPTIMIZED WEIGHT BASED CLUSTERING ALGORITHM IN HETEROGENEOUS WIRELESS SENS...cscpconf
The last few years have seen an increased interest in the potential use of wireless sensor networks (WSNs) in various fields like disastermanagementbattle field surveillance, and border security surveillance. In such applications, a large number of sensor nodes are deployed, which are often unattended and work autonomously. The process of dividing the network into interconnected substructures is called clustering and the interconnected substructures are called clusters. The cluster head (CH) of each cluster act as a coordinator within the substructure. Each CH acts as a temporary base station within its zone or cluster. It also communicates with other CHs. Clustering is a key technique used to extend the lifetime of a sensor network by reducing energy consumption. It can also increase network scalability. Researchers in all fields of wireless sensor network believe that nodes are homogeneous, but
some nodes may be of different characteristics to prolong the lifetime of a WSN and its reliability. We have proposed an algorithm for better cluster head selection based on weights for different parameter that influence on energy consumption which includes distance from base station as a new parameter to reduce number of transmissions and reduce energy consumption by sensor nodes. Finally proposed algorithm compared with the WCA, IWCA algorithm in terms of number of clusters and energy consumption.
ANFIS used as a Maximum Power Point Tracking Algorithmfor a Photovoltaic Syst...IJECEIAES
Photovoltaic (PV) modules play an important role in modern distribution networks; however, from the beginning, PV modules have mostly been used in order to produce clean, green energy and to make a profit. Working effectively during the day, PV systems tend to achieve a maximum power point accomplished by inverters with built-in Maximum Power Point Tracking (MPPT) algorithms. This paper presents an Adaptive Neuro-Fuzzy Inference System (ANFIS), as a method for predicting an MPP based on data on solar exposure and the surrounding temperature. The advantages of the proposed method are a fast response, non-invasive sampling, total harmonic distortion reduction, more efficient usage of PV modules and a simple training of the ANFIS algorithm. To demonstrate the effectiveness and accuracy of the ANFIS in relation to the MPPT algorithm, a practical sample case of 10 kW PV system and its measurements are used as a model for simulation. Modelling and simulations are performed using all available components provided by technical data. The results obtained from the simulations point to the more efficient usage of the ANFIS model proposed as an MPPT algorithm for PV modules in comparison to other existing methods.
Computational Approaches for Monitoring Voltage Stability in Power NetworksAM Publications
Voltage collapse and major blackouts have been repeatedly encountered in large power networks. The prime reason for this, is failure of systems ability to maintain synchronization. Under such condition system fails to maintain steady voltage at all buses and rapid growth in system collapse is assured. The basic work in this paper is to find the system stability by analyzing Jacobean determinant and Reactive Power in per unit. For this purpose we use Ward-Hale 6 Bus System with reactive power in per unit and Jacobean determinant. We compute critical value of reactive power by artificial neural network and Extrapolation through MS-Excel, where the system becomes unstable. We provide a computational framework which helps to analyze system stability limit.
High - Performance using Neural Networks in Direct Torque Control for Asynchr...IJECEIAES
This article investigates solution for the biggest problem of the Direct Torque Control on the asynchronous machine to have the high dynamic performance with very simple hysteresis control scheme. The Conventional Direct Torque Control (CDTC) suffers from some drawbacks such as high current, flux and torque ripple, as well as flux control at very low speed. In this paper, we propose an intelligent approach to improve the direct torque control of induction machine which is an artificial neural networks control. The principle, the numerical procedure and the performances of this method are presented. Simulations results show that the proposed ANN-DTC strategy effectively reduces the torque and flux ripples at low switching frequency, compared with Fuzzy Logic DTC and The Conventional DTC.
In this paper a new combination Radial Basis Function Neural Network and p-q Power Theory (RBFNN-PQ) proposed to control shunt active power filters (SAPF). The recommended system has better specifications in comparison with other control methods. In the proposed combination an RBF neural network is employed to extract compensation reference current when supply voltages are distorted and/or unbalance sinusoidal. In order to make the employed model much simpler and tighter an adaptive algorithm for RBF network is proposed. The proposed RBFNN filtering algorithm is based on efficient training methods called hybrid learning method.The method requires a small size network, very robust, and the proposed algorithms are very effective. Extensive simulations are carried out with PI as well as RBFNN controller for p-q control strategies by considering different voltage conditions and adequate results were presented.
Similar to COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN NEURAL NETWORK BASED IDENTIFICATION OF POWER SYSTEM (20)
Encryption in Microsoft 365 - ExpertsLive Netherlands 2024Albert Hoitingh
In this session I delve into the encryption technology used in Microsoft 365 and Microsoft Purview. Including the concepts of Customer Key and Double Key Encryption.
Pushing the limits of ePRTC: 100ns holdover for 100 daysAdtran
At WSTS 2024, Alon Stern explored the topic of parametric holdover and explained how recent research findings can be implemented in real-world PNT networks to achieve 100 nanoseconds of accuracy for up to 100 days.
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...UiPathCommunity
💥 Speed, accuracy, and scaling – discover the superpowers of GenAI in action with UiPath Document Understanding and Communications Mining™:
See how to accelerate model training and optimize model performance with active learning
Learn about the latest enhancements to out-of-the-box document processing – with little to no training required
Get an exclusive demo of the new family of UiPath LLMs – GenAI models specialized for processing different types of documents and messages
This is a hands-on session specifically designed for automation developers and AI enthusiasts seeking to enhance their knowledge in leveraging the latest intelligent document processing capabilities offered by UiPath.
Speakers:
👨🏫 Andras Palfi, Senior Product Manager, UiPath
👩🏫 Lenka Dulovicova, Product Program Manager, UiPath
Climate Impact of Software Testing at Nordic Testing DaysKari Kakkonen
My slides at Nordic Testing Days 6.6.2024
Climate impact / sustainability of software testing discussed on the talk. ICT and testing must carry their part of global responsibility to help with the climat warming. We can minimize the carbon footprint but we can also have a carbon handprint, a positive impact on the climate. Quality characteristics can be added with sustainability, and then measured continuously. Test environments can be used less, and in smaller scale and on demand. Test techniques can be used in optimizing or minimizing number of tests. Test automation can be used to speed up testing.
Enhancing Performance with Globus and the Science DMZGlobus
ESnet has led the way in helping national facilities—and many other institutions in the research community—configure Science DMZs and troubleshoot network issues to maximize data transfer performance. In this talk we will present a summary of approaches and tips for getting the most out of your network infrastructure using Globus Connect Server.
Alt. GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using ...James Anderson
Effective Application Security in Software Delivery lifecycle using Deployment Firewall and DBOM
The modern software delivery process (or the CI/CD process) includes many tools, distributed teams, open-source code, and cloud platforms. Constant focus on speed to release software to market, along with the traditional slow and manual security checks has caused gaps in continuous security as an important piece in the software supply chain. Today organizations feel more susceptible to external and internal cyber threats due to the vast attack surface in their applications supply chain and the lack of end-to-end governance and risk management.
The software team must secure its software delivery process to avoid vulnerability and security breaches. This needs to be achieved with existing tool chains and without extensive rework of the delivery processes. This talk will present strategies and techniques for providing visibility into the true risk of the existing vulnerabilities, preventing the introduction of security issues in the software, resolving vulnerabilities in production environments quickly, and capturing the deployment bill of materials (DBOM).
Speakers:
Bob Boule
Robert Boule is a technology enthusiast with PASSION for technology and making things work along with a knack for helping others understand how things work. He comes with around 20 years of solution engineering experience in application security, software continuous delivery, and SaaS platforms. He is known for his dynamic presentations in CI/CD and application security integrated in software delivery lifecycle.
Gopinath Rebala
Gopinath Rebala is the CTO of OpsMx, where he has overall responsibility for the machine learning and data processing architectures for Secure Software Delivery. Gopi also has a strong connection with our customers, leading design and architecture for strategic implementations. Gopi is a frequent speaker and well-known leader in continuous delivery and integrating security into software delivery.
Epistemic Interaction - tuning interfaces to provide information for AI supportAlan Dix
Paper presented at SYNERGY workshop at AVI 2024, Genoa, Italy. 3rd June 2024
https://alandix.com/academic/papers/synergy2024-epistemic/
As machine learning integrates deeper into human-computer interactions, the concept of epistemic interaction emerges, aiming to refine these interactions to enhance system adaptability. This approach encourages minor, intentional adjustments in user behaviour to enrich the data available for system learning. This paper introduces epistemic interaction within the context of human-system communication, illustrating how deliberate interaction design can improve system understanding and adaptation. Through concrete examples, we demonstrate the potential of epistemic interaction to significantly advance human-computer interaction by leveraging intuitive human communication strategies to inform system design and functionality, offering a novel pathway for enriching user-system engagements.
Observability Concepts EVERY Developer Should Know -- DeveloperWeek Europe.pdfPaige Cruz
Monitoring and observability aren’t traditionally found in software curriculums and many of us cobble this knowledge together from whatever vendor or ecosystem we were first introduced to and whatever is a part of your current company’s observability stack.
While the dev and ops silo continues to crumble….many organizations still relegate monitoring & observability as the purview of ops, infra and SRE teams. This is a mistake - achieving a highly observable system requires collaboration up and down the stack.
I, a former op, would like to extend an invitation to all application developers to join the observability party will share these foundational concepts to build on:
Why You Should Replace Windows 11 with Nitrux Linux 3.5.0 for enhanced perfor...SOFTTECHHUB
The choice of an operating system plays a pivotal role in shaping our computing experience. For decades, Microsoft's Windows has dominated the market, offering a familiar and widely adopted platform for personal and professional use. However, as technological advancements continue to push the boundaries of innovation, alternative operating systems have emerged, challenging the status quo and offering users a fresh perspective on computing.
One such alternative that has garnered significant attention and acclaim is Nitrux Linux 3.5.0, a sleek, powerful, and user-friendly Linux distribution that promises to redefine the way we interact with our devices. With its focus on performance, security, and customization, Nitrux Linux presents a compelling case for those seeking to break free from the constraints of proprietary software and embrace the freedom and flexibility of open-source computing.
Securing your Kubernetes cluster_ a step-by-step guide to success !KatiaHIMEUR1
Today, after several years of existence, an extremely active community and an ultra-dynamic ecosystem, Kubernetes has established itself as the de facto standard in container orchestration. Thanks to a wide range of managed services, it has never been so easy to set up a ready-to-use Kubernetes cluster.
However, this ease of use means that the subject of security in Kubernetes is often left for later, or even neglected. This exposes companies to significant risks.
In this talk, I'll show you step-by-step how to secure your Kubernetes cluster for greater peace of mind and reliability.
LF Energy Webinar: Electrical Grid Modelling and Simulation Through PowSyBl -...DanBrown980551
Do you want to learn how to model and simulate an electrical network from scratch in under an hour?
Then welcome to this PowSyBl workshop, hosted by Rte, the French Transmission System Operator (TSO)!
During the webinar, you will discover the PowSyBl ecosystem as well as handle and study an electrical network through an interactive Python notebook.
PowSyBl is an open source project hosted by LF Energy, which offers a comprehensive set of features for electrical grid modelling and simulation. Among other advanced features, PowSyBl provides:
- A fully editable and extendable library for grid component modelling;
- Visualization tools to display your network;
- Grid simulation tools, such as power flows, security analyses (with or without remedial actions) and sensitivity analyses;
The framework is mostly written in Java, with a Python binding so that Python developers can access PowSyBl functionalities as well.
What you will learn during the webinar:
- For beginners: discover PowSyBl's functionalities through a quick general presentation and the notebook, without needing any expert coding skills;
- For advanced developers: master the skills to efficiently apply PowSyBl functionalities to your real-world scenarios.
Free Complete Python - A step towards Data Science
COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN NEURAL NETWORK BASED IDENTIFICATION OF POWER SYSTEM
1. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
DOI : 10.5121/ijcsit.2013.5407 93
COMPARATIVE STUDY OF BACKPROPAGATION
ALGORITHMS IN NEURAL NETWORK BASED
IDENTIFICATION OF POWER SYSTEM
Sheela Tiwari1
, Ram Naresh2
, Rameshwar Jha3
1
Department of Instrumentation and Control Engineering, Dr. B R Ambedkar National
Institute of Technology, Jalandhar, Punjab, India
tiwaris@nitj.ac.in
2
Department of Electrical Engineering, National Institute of Technology, Hamirpur,
Himachal Pradesh, India
rnareshnith@gmail.com
3
IET Bhaddal Technical Campus, Ropar, Punjab, India
rjharjha@yahoo.com
ABSTRACT
This paper explores the application of artificial neural networks for online identification of a multimachine
power system. A recurrent neural network has been proposed as the identifier of the two area, four machine
system which is a benchmark system for studying electromechanical oscillations in multimachine power
systems. This neural identifier is trained using the static Backpropagation algorithm. The emphasis of the
paper is on investigating the performance of the variants of the Backpropagation algorithm in training the
neural identifier. The paper also compares the performances of the neural identifiers trained using variants
of the Backpropagation algorithm over a wide range of operating conditions. The simulation results
establish a satisfactory performance of the trained neural identifiers in identification of the test power
system.
KEYWORDS
System Identification, Recurrent Neural Networks, Static Backpropagation (BP)
1. INTRODUCTION
With a continuously increasing demand for electric power worldwide, there is an impending need
of augmenting the power carrying capacity of the existing power grid. The existing power grid
requires to be “smartened up” to improve the reliability, security and efficiency of the electric
power system. Continuous monitoring and intelligent control of the grid activities is the key to a
smart grid. The conventional controllers require acceptable approximate mathematical models of
the power systems and the involved uncertainties but with increasing complexities in the
contemporary power grid, it is becoming tedious and time consuming to generate such models.
Artificial neural networks (ANNs) have been known to have the capability to learn the complex
approximate relationships between the inputs and the outputs of the system and are not restricted
by the size and complexity of the system [1]. The ANNs learn these approximate relationships on
the basis of actual inputs and outputs. Therefore, they are generally more accurate as compared to
the relationships based on assumptions. This imparts immense potential to the ANNs for use in
identification and control of the modern power systems. ANNs have been proposed for detection
of power system harmonics [2-4], fault section estimation [5-6], fault diagnostics of power plant
2. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
94
[7] and in protection strategies [8-9]. Applications of ANNs in reactive power transfer allocation
[10] and ATC estimation [11] have been reported. Stability issues like damping of oscillations
[12-14], prediction of loadability margins [15] and voltage contingency screening [16] have been
successfully addressed by ANN based solutions. ANNs have the potential of application for real-
time control [17]. Accurate system identification is of immense importance in power system
operation and control. The capabilities of the ANNs make them a suitable choice for this
application. Multilayer feedforward artificial neural networks using Backpropagation algorithm
for training have been proposed for successful online model identification of synchronous
generator [18] and a UPFC equipped single machine infinite bus system [19]. A neural network
based estimation unit has been proposed to estimate in real time, the parameters for an interfacing
scheme for grid-connected inverters and simultaneously estimating the grid voltage [20]. The
authors have employed neural network for system identification for predictive control of a
multimachine power system operating under widely varying operating conditions and subjected to
transient conditions [21].
The work undertaken proposes to use a recurrent neural network for online identification of a
multimachine power system. Since the Backpropagation algorithm has been successfully
employed to train system identifiers as already reported in literature [18-19], this work aims to
investigate the training performance of some of the variants of the Backpropagation algorithm in
training the proposed neural identifier. The testing performances of the neural identifiers trained
using variants of the Backpropagation algorithm are also compared to establish the accuracy of
the differently trained identifiers.
This paper is organized as: Section 2 presents a brief overview of the recurrent neural networks
including the types of architecture for training such networks. Section 3 describes the system
under consideration and outlines the objectives of this work. Section 4 presents the architecture of
the proposed neural identifier. Section 5 discusses the training of the proposed identifier and
presents a brief overview of the variants of the Backpropagation algorithm used for training the
proposed neural identifier in this work. The simulation results of training and testing the proposed
identifier and the discussions are presented in section 6 followed by the conclusions in section 7.
2. RECURRENT NEURAL NETWORKS
Neural networks are broadly classified as static networks and dynamic networks. In case of static
networks, the output of the network is dependent only on the current input to the network and is
calculated directly from the input passing through the feedforward connections. There are no
delay elements and no feedback elements present in the static networks. In dynamic networks, the
output of the network depends on the current input as well as on current and/or previous inputs
and outputs i.e. states of the network. These networks therefore, are said to possess memory and
can be trained to learn sequential or time varying patterns [22], making them more powerful than
the static networks. Dynamic networks are further divided into two categories: those having only
feedforward connections with delays at the input layer only/or distributed throughout the network
and those having feedback or recurrent connections.
The recurrent neural networks have the capability to predict the future values based on the values
at the preceding instants. The nonlinear autoregressive network with exogenous (independent)
inputs i.e. NARX, is a recurrent dynamic network defined by
3. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
95
Where y(k) and u(k) are the outputs and inputs at the kth instant and ny and nu are the number of
time steps for which the current output is regressed on the output and input respectively. A
diagram showing the implementation of the NARX model using a feedforward neural network to
approximate the function f in (1) is given in Figure 1.
Two different architectures have been proposed to train a NARX network [22]. First is the
parallel architecture as shown in Figure 2, where the output of the neural network is fed back to
the input of the feedforward neural network as part of the standard NARX architecture. In
contrast, in the series-parallel architecture as shown in Figure 3, the true output of the plant (not
the output of the identifier) is fed to the neural network model as it is available during training.
This architecture has two advantages [22]. The first is a more accurate value presented as input to
the neural network. The second advantage is the absence of a feedback loop in the network
thereby enabling the use of static backpropagation for training instead of the computationally
expensive dynamic backpropagation required for the parallel architecture. Also, assuming the
output error tends to a small value asymptotically so that , the series parallel
model may be replaced by a parallel model without serious consequences if required.
Figure 1. Implementation of NARX model
Figure 2. Parallel architecture of training a NARX model
4. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
96
Figure 3. Series-Parallel architecture of training a NARX model
3. SYSTEM DESCRIPTION
The two area, four machine system with active power flowing from Area 1 to Area 2 [23],
proposed to study the electromechanical oscillations of a multimachine power system is taken as
the test system for the work undertaken. In spite of the small size of the system, its behaviour
mimics the behaviour of a large power system in actual operation. Each area comprises of two
900 MVA machines and the two areas are connected by a 220 kV double circuit line of length
220 km. The load voltage profile is improved by installing additional 187 MVAr capacitors in
each area. The system under study is equipped with PSS and has a UPFC installed between bus
11 and bus 12 with bus 11 common to the shunt and series converters and the other side of the
series converter connected to bus 12 as shown in Figure 4.
Figure 4. 2-Area system equipped with UPFC
The effective utilization of the UPFC in the system requires implementation of various control
schemes, many of which require the system to be identified. In this work, the active power at bus
12, corresponding to a specific value of the quadrature component of the series voltage
5. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
97
injected by the UPFC is to be predicted under various operating conditions. This next step value
of is predicted using the values of and at some preceding instants.
Objectives
i. To design a neural identifier to predict the next step value of on the basis of the
values of and at preceding time instants.
ii. To investigate the performances of the variants of the Backpropagation algorithm for
training the neural identifier.
iii. To investigate the performances of the neural identifiers trained using the variants of the
Backpropagation algorithm.
iv. To propose the most suitable of the considered variants of the Backpropagation
algorithm for online application for the system under consideration.
4. ARCHITECTURE OF THE NEURAL IDENTIFIER
A neural identifier that predicts the next step value of on the basis of the values of and
at four preceding instants is proposed. As the objective clearly requires a dynamic neural
network, the NARX model is used. A two layer neural network with sigmoidal hidden layer
neurons and linear output layer neurons can identify any system with any degree of accuracy,
subject to the availability of sufficient number of hidden neurons [24]. Therefore, the NARX
model is implemented using the two layer feedforward neural network. Since the true value of the
output for the preceding instants are available, the series parallel architecture is used. As the
values of and at four preceding instants are to be used, the total number of inputs to the
neural network is eight as shown in Figure 5. The number of sigmoidal neurons in the hidden
layer has been fixed at thirteen using trial and error approach and the output layer has one linear
neuron.
Figure 5. Architecture of the proposed neural identifier
6. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
98
The system under consideration is modelled and simulated using MATLAB / SIMULINK to
generate data for training and testing the proposed neural identifier. The operation of the system is
simulated by applying restricted within the range +0.1 pu and -0.1 pu (restricting the
quadrature component of the series injected voltage to 10% of the nominal line-to-ground
voltage) and sampling the input, and output, at the rate of 32 samples per second. The
neural network is presented with the following inputs
And
for predicting For linear input neurons, the output of the input neurons is same as
the input given by
The output of the hidden layer (layer 1), consisting of 13 sigmoidal neurons is given by
Where, is the weight matrix between input and hidden layers and is the bias to
the hidden layer neurons. Similarly, the output of the output layer (layer 2), comprising of one
linear neuron is given by
Where, is the weight matrix between hidden and output and layers and is the bias to
the output layer neurons.
5. TRAINING OF THE NEURAL IDENTIFIER
Identification requires setting up a suitably parameterized identification model and adjustment of
these parameters of the model to optimize a performance function based on the error between the
outputs from the plant and the identification model [22]. It is assumed that the weight matrices of
the neural network proposed as the identifier exists, for which, both plant and the identifier have
the same output for any specified inputs, for the same initial conditions [22].
The system under consideration is simulated at different operating conditions for a wide range of
the steady state active power flow level in the tie-line flowing between the two areas to generate
data for training. The training data set consists of 1288 data points spread over a wide range of
operation. This training dataset is employed in training the proposed neural identifier offline
through simulation to make it learn the forward dynamics of the plant. During training the
weights and biases of the network are iteratively adjusted to minimize the network performance
function. The performance function used for the neural identifier under consideration is the mean
square error, mse, given by
7. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
99
Where, N is the size of the training dataset, and are the target and predicted value of the
output of the neural network when the input is presented and is the error (difference
between the target and predicted value) for the input. The performance index V in (7) is a
function of weights and biases, and can be given by
The performance of the neural network can be improved by modifying till the desired level of
the performance index, is achieved. This is achieved by minimizing with respect to
and the gradient required for this is given by
where, is the Jacobian matrix given by
and is the error for all the inputs. The gradient in (9) is determined using backpropagation,
which involves performing computations backward through the network. This gradient is then
used by different algorithms to update the weights of the network. These algorithms differ in the
way they use the gradient to update the weights of the network and are known as the variants of
the Backpropagation algorithm.
This work compares the performance of the basic implementation of the Backpropagation
algorithm i.e. Gradient descent algorithm with other variants in order to investigate the potentials
of these algorithms for online applications in power system identification. A brief overview of the
different algorithms considered in this work is given under:
i) Gradient Descent algorithm (GD): The network weights and biases, is modified in a
direction that reduces the performance function in (8) most rapidly i.e. the negative of the
gradient of the performance function [25]. The updated weights and biases in this algorithm are
given by
Where, is the vector of the current weights and biases, is the current gradient of the
performance function and is the learning rate.
8. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
100
ii) Scaled Conjugate Gradient Descent algorithm (SCGD): The gradient descent algorithm
updates the weights and biases along the steepest descent direction but is usually associated with
poor convergence rate as compared to the Conjugate Gradient Descent algorithms, which
generally result in faster convergence [26]. In the Conjugate Gradient Descent algorithms, a
search is made along the conjugate gradient direction to determine the step size that minimizes
the performance function along that line. This time consuming line search is required during all
the iterations of the weight update. However, the Scaled Conjugate Gradient Descent algorithm
does not require the computationally expensive line search and at the same time has the advantage
of the Conjugate Gradient Descent algorithms [26]. The step size in the conjugate direction in this
case is determined using the Levenberg-Marquardt approach. The algorithm starts in the direction
of the steepest descent given by the negative of the gradient as
The updated weights and biases are then given by
Where, is the step size determined by the Levenberg-Marquardt algorithm [27]. The next
search direction that is conjugate to the previous search directions is determined by combining the
new steepest descent direction with the previous search direction and is given by
The value of is as given in [26], by
Where is given by
iii) Levenberg-Marquardt algorithm (LM): Since the performance index in (8) is sum of
squares of non linear function, the numerical optimization techniques for non linear least squares
can be used to minimize this cost function. The Levenberg-Marquardt algorithm, which is an
approximation to the Newton’s method is said to be more efficient in comparison to other
methods for convergence of the Backpropagation algorithm for training a moderate-sized
feedforward neural network [27]. As the cost function is a sum of squares of non linear function,
the Hessian matrix required for updating the weights and biases need not be calculated and can be
approximated as
The updated weights and biases are given by
9. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
101
Where, μ is a scalar and I is the identity matrix.
iv) Automated Bayesian Regularization (BR): Regularization as a mean of improving network
generalization is used within the Levenberg-Marquardt algorithm. Regularization involves
modification in the performance function. The performance function for this is the sum of the
squares of the errors and it is modified to include a term that consists of the sum of squares of the
network weights and biases. The modified performance function is given by
Where SSE and SSW are given by
Where, is the total number of weights and biases, in the network. The performance index
in (19) forces the weights and biases to be small, which produces a smoother network response
and avoids over fitting. The values of α and β are determined using Bayesian Regularization in an
automated manner [28-29].
6. SIMULATION RESULTS AND DISCUSSION
6.1. TRAINING
The neural network proposed in section 4 was trained using the training set and the training
algorithms described in section 5. A Pentium (R) Dual-Core CPU T4400 @2.20 GHz was used to
train the proposed neural identifier. Table 1 summarizes the results of training the proposed
network using the four training algorithms discussed in the earlier section. Each entry in the table
represents 50 different trials, with random initial weights taken for each trial to rule out the
weight sensitivity of the performance of the different training algorithms. The network was
trained in each case till the value of the performance index in (8) was 0.0001 or less.
Table 1 Statistical comparison of different training algorithms
S. No. Training Algorithm
Average
Time (s)
Ratio
Maximum
Time (s)
Minimum
Time (s)
Std Dev.
1. Gradient Descent Very slow in converging
2.
Scaled Conjugate
Gradient Descent
280.8473 47.865 911.5366 114.2788 180.7087
3. Levenberg Marquardt 5.8675 1 9.4187 2.9393 1.6995
4.
Bayesian
Regularization
7.6748 1.308 14.4054 3.1373 1.9246
10. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
102
The Gradient Descent algorithm was very slow in converging for the required value of the
performance index. The average time required for training the network using the Levenberg-
Marquardt algorithm was the least whereas, maximum time was required for training the network
using the Scaled Conjugate Gradient Descent algorithm. The training algorithm employing
Bayesian Regularization continuously modifies its performance function and hence, takes more
time as compared to the Levenberg-Marquardt algorithm but this time is still far less than the
Scaled Conjugate Gradient Descent method. From Table 1, it can be established that the
Levenberg-Marquardt algorithm is the fastest of all the training algorithms considered in this
work for training a neural network to identify the multimachine power system. Since the training
time required for different training algorithms have been compared, the conclusion drawn from
the results for the offline training may also be extended to online training. Therefore, it can be
assumed that similar trend of training time required by the different training algorithms will be
exhibited during online training of the proposed neural identifier for continuous updating of the
offline trained identifier.
6.2. TESTING
The neural networks trained using the training algorithms listed in Table1 were tested on the same
CPU. The test datasets consisted of data points not included in the training set. The system under
consideration was simulated at two such operating points for which no data point was included in
the training set. The operation of the multimachine power system under consideration at these two
operating points was simulated using MATLAB/SIMULINK for a period of 13 seconds each.
This period also included a 3-phase short circuit fault at point A at t=10 s for a duration of 200 ms
with the circuit breakers auto reclosing after 12 cycles. The data during these two simulations was
sampled at the rate of 32 samples per second to form two test sets: Test Set I and Test Set II,
corresponding to the two operating points. As the Gradient Descent algorithm was too slow to
converge for the desired value of the performance index, the neural networks that were trained
using the rest of the three training algorithms listed in Table 1 were tested using these two test
sets.
Test Set I: Test Set 1 consists of 417 data points generated at such an operating point at which the
active power flow in the system at steady state is 399 MW, which is within the range of the active
power flowing in the system at the steady state at the operating points considered for generating
the training set. The first four data points are used to predict the output at the next instant.
Therefore, the number of predicted outputs for this test set is 413. A three-phase short circuit fault
is simulated at t=10 s which corresponds to the sample point number 321 in the test set. The
actual output for the system after autoreclosure of the circuit breaker is available in the sample
point number 329 of the test set. The actual values of the output power and the values predicted
for the same using different neural networks during a part of the steady state and transient period
are shown in Figures 6 and 7 respectively.
11. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
103
Figure 6. Actual and predicted values of output power during steady state for test set I
Figure 6 clearly shows that the predictive quality of all the neural networks is satisfactory during
the steady state period. The effect of the 3-phase short circuit fault on the system is captured in
the sample point 322 of the actual output power as shown in Figure 7. However, the
Figure 7. Actual and predicted values of output power during transient period for test set I
circuit breakers operate and an improved system performance is reflected in subsequent samples.
As the neural networks have been trained to use the information at four preceding instants to
predict the next step output, the effect of decrease in the actual output in sample 322 is reflected
immediately in the values predicted by neural networks trained using SCG and LM methods of
training in sample 323. The output power predicted by the neural network trained using the BR
method also shows a downward trend in sample 323 but the minimum value of output power
predicted by this neural network is in sample 324. Figure 7 clearly shows that the values predicted
using the neural network trained using the LM method follow the actual values closely in the
transient period. The effect of autoreclosure of the circuit breakers on the power level is visible in
sample number 329 of the actual output power. The increased value of the actual output power in
sample 329 due to the autoreclosure of the circuit breakers is reflected in the values predicted by
all the three neural networks in the subsequent instant. The predictive quality of the three trained
12. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
104
neural networks is investigated by comparing the average absolute error in the predicted values.
Table 2 shows the average absolute error along with the maximum error and the minimum error
in the values predicted by the trained neural networks.
Table 2 Performance of trained neural networks on test set I
S. No.
Training
Method
Average Absolute
Error for entire set
Maximum
Error
Minimum
Error
1. SCG 0.0117 1.0689 1.4809e-5
2. LM 0.0103 0.9960 1.0254e-5
3. BR 0.0171 1.8511 7.0297e-6
The average absolute error is the least for the neural network trained using the LM method. The
maximum error for every neural network is reported in the transient period and is least for the
network trained using the LM method. The neural network trained using the BR method follows
the actual values satisfactorily but has slightly higher values of average absolute error and
maximum error associated with it as compared to the first two methods. However, the BR method
imparts generalization to the network and the neural network trained using BR reported the least
minimum error for this test set.
Test Set II: This test set also consists of 417 data points sampled during simulation of the system
at an operating point for which the active power flowing in the system at steady state is 156 MW,
which is outside the range of active powers flowing at steady state corresponding to the operating
points considered for generating the training set. For the same reasons as for Test Set I, the
number of predicted values in Test Set II is also 413. In this case too, a 3-phase short circuit fault
of duration 200 ms is simulated at t=10 s. Figures 8 and 9 show the actual and predicted output
values for a section of steady state and transient period of the system respectively, at the
operating point under consideration.
Figure 8. Actual and predicted values of output power during steady state for test set II
13. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
105
Figure 9. Actual and predicted values of output power during transient period for test set II
It is clear from Figure 8 that the actual power output values and the values predicted using the
different neural networks in steady state are in close proximity even at this operating point. The
decrease in the actual value for the active power immediately after the 3-phase short circuit fault
is captured in sample 321 of Figure 9. This information is used by all the neural networks for
making predictions for the next instant and all the neural networks successfully predict the
decrease in the active power with the minimum power value predicted for sample 322. Similarly,
the effect of autoreclosing of the circuit breakers is also predicted successfully by the three neural
networks. The average absolute error along with the minimum and maximum errors in the values
predicted by the trained neural networks are given in Table 3.
Table 3 clearly shows that the average absolute errors even at such an operating point for which
the steady state active power flowing in the system is outside the range of the active powers
flowing at operating points considered for generation of training data set are following the same
trend as reported in case of Test Set I.
Table 3 Performance of trained neural networks on test set II
S. No.
Training
Method
Average Absolute
Error for entire set
Maximum
Error
Minimum
Error
1. SCG 0.0092 0.7465 9.9872e-6
2. LM 0.0078 0.5507 1.2711e-5
3. BR 0.0105 0.6905 2.8250e-6
7. CONCLUSION
A neural network has been proposed to predict the next step value of the output power on the
basis of the values of the control input and output power at preceding time instants. The proposed
neural network is trained using different variants of the Backpropagation algorithm.
Investigations in to the training performance of the different algorithms establish that the
Levenberg-Marquardt algorithm is the fastest to converge. Comparison of the predictions made
by the different neural networks reveal that the neural network trained using the Levenberg-
Marquardt algorithm gives the most accurate predictions. The availability of fast computing
machines in current times and the accurate predictions reported in this work clearly establish the
scope for online application of neural networks for identification of multimachine power systems.
14. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
106
The fast convergence teamed with good predictive quality makes Levenberg-Marquardt algorithm
the most suitable choice of all the variants considered in this work for training a neural network
for this application.
REFERENCES
[1] Halpin, S.M. & Burch, R.F.,(1997) “Applicability of neural networks to industrial and commercial
power systems: a tutorial overview”, IEEE Trans. Industry Applications, Vol. 33, No. 5, pp1355-
1361.
[2] Hsiung, C.L., (2007) “Intelligent Neural Network-Based Fast Power System Harmonic Detection”,
IEEE Trans. Industrial Electronics, Vol. 54, No. 1, pp43-52.
[3] Mazumdar, J., Harley, R.G., Lambert, F.C., & Venayagamoorthy, G.K., (2007) “Neural Network
Based Method for Predicting Nonlinear Load Harmonics”, IEEE Trans. Power Electronics, Vol. 22,
No. 3, pp1036-1045.
[4] Mazumdar, J., Harley, R.G., Lambert, F.C., Venayagamoorthy, G.K., & Page, M.L., (2008)
“Intelligent Tool for Determining the True Harmonic Current Contribution of a Customer in a Power
Distribution Network”, IEEE Trans. Industry Applications, Vol. 44, No. 5, pp1477-1485.
[5] Cardoso, G., Rolim, J.G., & Zurn, H.H., (2004) “Application of neural-network modules to electric
power system fault section estimation”, IEEE Trans. Power Delivery, Vol. 19, No. 3, pp1034- 1041.
[6] Cardoso, G., Rolim, J.G., & Zurn, H.H., (2008) “Identifying the Primary Fault Section After
Contingencies in Bulk Power Systems”, IEEE Trans. Power Delivery, Vol. 23, No. 3, pp1335-1342.
[7] Liangyu Ma, Yongguang Ma, & Lee, K.Y., (2010) “An Intelligent Power Plant Fault Diagnostics for
Varying Degree of Severity and Loading Conditions”, IEEE Trans. Energy Conversion, Vol. 25, No.
2, pp546-554.
[8] Negnevitsky, M., & Pavlovsky, V., (2005) “Neural networks approach to online identification of
multiple failures of protection systems”, IEEE Trans. Power Delivery, Vol. 20, No. 2, pp588- 594.
[9] Thalassinakis, E.J., Dialynas, E.N., & Agoris, D., (2006) “Method Combining ANNs and Monte
Carlo Simulation for the Selection of the Load Shedding Protection Strategies in Autonomous Power
Systems”, IEEE Trans. Power Systems, Vol. 21, No. 4, pp1574-1582.
[10] Mustafa, M.W., Khalid, S.N., Shareef, H., & Khairuddin, A., (2008) “Reactive power transfer
allocation method with the application of artificial neural network”, IET Proc.- Genr. Transm.
Distrib., Vol. 2, No. 3, pp402-413.
[11] Pandey, S.N., Pandey, N.K., Tapaswi, S., & Srivastava, L., (2010) “Neural Network-Based Approach
for ATC Estimation Using Distributed Computing”, IEEE Trans. Power Systems, Vol. 25, No. 3,
pp1291-1300.
[12] Changaroon, B., Srivastava, S.C., & Thukaram, D., (2000) “A neural network based power system
stabilizer suitable for on-line training-a practical case study for EGAT system”, IEEE Trans. Energy
Conversion, Vol. 15, No. 1, pp103-109.
[13] Chao Lu, Si, J., & Xiaorong Xie., (2008) “Direct Heuristic Dynamic Programming for Damping
Oscillations in a Large Power System”, IEEE Trans. Systems, Man, and Cybernetics, Part B:
Cybernetics, Vol. 38, No. 4, pp1008-1013.
[14] Nguyen, T.T., & Gianto, R., “Neural networks for adaptive control coordination of PSSs and FACTS
devices in multimachine power system”, IET Proc.- Genr. Transm. Distrib., Vol. 2, No. 3, pp355-372.
[15] Gu, X., & Canizares, C.A., (2007) “Fast prediction of loadability margins using neural networks to
approximate security boundaries of power systems”, IET Proc.- Genr. Transm. Distrib., Vol. 1, No. 3,
pp466-475.
[16] Jain, T., Srivastava, L., & Singh, S.N., (2003) “Fast voltage contingency screening using radial basis
function neural network”, IEEE Trans. Power Systems, Vol. 18, No. 4, pp1359- 1366.
[17] Ray, S., & Venayagamoorthy, G.K., (2008) “Real-time implementation of a measurement-based
adaptive wide-area control system considering communication delays”, IET Proc.- Genr. Transm.
Distrib., Vol. 2, No. 1, pp62-70.
[18] Jung-Wook Park, Venayagamoorthy, G.K., & Harley, R.G., (2005) “MLP/RBF neural-networks-
based online global model identification of synchronous generator”, IEEE Trans. Industrial
Electronics, Vol. 52, No. 6, pp1685- 1695.
[19] Venayagamoorthy, G.K., & Kalyani, R.P., (2005) “Two separate continually online-trained
neurocontrollers for a unified power flow controller”, IEEE Trans. Industry Applications, Vol. 41,
No. 4, pp 906- 916.
15. International Journal of Computer Science & Information Technology (IJCSIT) Vol 5, No 4, August 2013
107
[20] Mohamed, Y.A.-R., & El-Saadany, E.F., (2008) “Adaptive Discrete-Time Grid-Voltage Sensorless
Interfacing Scheme for Grid-Connected DG-Inverters Based on Neural-Network Identification and
Deadbeat Current Regulation”, IEEE Trans. Power Electronics, Vol. 23, No. 1, pp308-321.
[21] Tiwari, S., Naresh, R., & Jha, R., (2011) “Neural network predictive control of UPFC for improving
transient stability performance of power system”, Appl Soft Comput, Vol. 11, No. 8, pp4581-4590.
[22] Narendra, K.S., & Parthasarathy, K., (1990) “Identification and control of dynamical systems using
neural networks”, IEEE Trans. Neural Networks, Vol. 1, No. 1, pp4-27.
[23] Kundur, P (1994) Power System Stability and Control, Tata McGraw-Hill, New Delhi.
[24] Hagan, M.T., & Demuth, H.B., (1999) “Neural networks for control”, American Control Conference.
[25] Hagan, M.T., Demuth, H.B., & Beale, M.H (1996) Neural Network Design, MA: PWS Publishing,
Boston.
[26] Moller, M.F., (1993) “A scaled conjugate gradient algorithm for fast supervised learning”, Neural
Networks, Vol. 6, pp525-533.
[27] Hagan, M.T., & Menhaj, M., (1994) “Training feed-forward networks with the Marquardt algorithm”,
IEEE Trans. Neural Networks, Vol. 5, No. 6, pp989-993.
[28] Foresee, F.D. & Hagan, M.T., (1997) “Gauss-Newton approximation to Bayesian regularization”,
International Joint Conference on Neural Networks.
[29] Mackay, D.J.C., (1992) “Bayesian interpolation”, Neural Computation, Vol. 4, No. 3, pp415-447.
Authors
Sheela Tiwari did her B.E. (Electrical Engineering) from Thapar Institute of
Engineering and Technology, Patiala, India in 1993 and M.Tech. in Instrumentation
and Control from Department of Electrical Engineering, Punjab Agricultural
University in 1996. She joined Regional Engineering College Jalandhar (now Dr. B.
R. Ambedkar National Institute of Technology, Jalandhar) as lecturer in the
Department of Instrumentation and Control Engineering in 1997 and is currently
working as Associate Professor in the same department. Her research interests are
artificial intelligence and its application in power systems.
Dr R. Naresh was born in Himachal Pradesh, India in 1965. He received BE in
Electrical Engineering from Thapar Institute of Engineering and Technology, Patiala,
India in 1987, ME in Power Systems from Punjab Engineering College, Chandigarh
in 1990 and Ph D from the University of Roorkee, Roorkee (now IIT Roorkee), India
in 1999. He joined Regional Engineering College, Hamirpur (now NIT Hamirpur) in
1989, worked as Assistant Professor in the Department of Electrical Engineering
from 2000 to 2007 and is working as Professor in the same department since August
2007. He has taught courses in basic electrical engineering, power systems, electrical machines and
artificial neural networks. He has published a number of research papers in national & international
journals. He has been providing consultancy services to electric power industry. His research interests are
artificial intelligence applications to power system optimization problems, evolutionary computation,
neural networks and fuzzy systems.
Dr. R. Jha was born in Bihar, India in 1945. He did B.Sc Engg. (Electrical Engg.) from
Bhagalpur Univ. Bhagalpur in 1965, M. Tech. in Control and Instrumentation Engg.
from IIT, New Delhi in 1970 and Ph.D in Electrical Engg. in 1980. He served as a
Lecturer in Electrical Engg. at Jorhat Engineering College, Jorhat (Assam) from 1965
till he joined as Asstt. Prof. at Panjab Engineering College, Chandigarh in 1972. He
joined as Professor of Electrical Engg. at REC Hamirpur (now NIT Hamirpur) in
1986 and was elevated to the post of Principal of REC Hamirpur in 1990 which he
continued till 1994. He joined the Department of Instrumentation & Control
Engineering at National Institute of Technology Jalandhar as Professor in September
1994 and attained superannuation in 2011. He is currently serving as Director-General, IET Technical
Campus, Bhaddal, Ropar, Punjab. Dr. Jha has published extensively in areas of non-linear control systems,
stability theory and its applications to electrical engineering problems. He has published more than 85-
research papers in both international and national journals of repute.