This paper presents an implementation of a MultiAgent based Energy Management System for a micro grid with JADE (Java Agent Development Framework). The MAS is applied for a micro grid consisting of different distributed energy sources such as solar PV system, wind power system, diesel generator system, storage system, and critical and noncritical loads. Different agents are developed on JADE framework and they are given responsibilities of relevant DES’s (Distributed Energy Source) and loads. A runtime environment for Agents are created and a dynamic simulation model developed through JADE considering the intermittent qualities of renewable energy sources. The case studies presented on this paper are modeled on JADE platform. Developing MAS in JADE runtime environment using AOP (Agent-Oriented Programming) helps agents to operate with all their autonomous, rational, reactive, and proactive qualities. The use of MAS concept in micro grids improves its efficiency in various aspects. The MAS based micro grid management system implemented in JADE platform and can be used to carry out various simulations to study about agent behaviors in different environments and different system objectives. The main purpose of the paper is to prove the possibility of using multiagent concept in micro grid energy management systems. For more information : https://www.researchgate.net/publication/339981129_Developing_Multi-Agent_Based_Micro-Grid_Management_System_in_JADE

1. 1
Multi–Agent Controlled Energy
Management System For Micro-Grids
Progress Review 1 Start : 10 / 12 / 2018
Presented by : Vimukkthi Vithanage
Department of Electrical Engineering
University of Moratuwa
Supervised by : Prof. K.T.M.U Hemapala
Eng. Anura Wijayapala

2. 2
Introduction
Micro-Grid
With
Energy Management System
Using
Multi-Agent Concept

3. 3
Introduction
-No significant change in centrally controlled structure of
Power Electrical Grids for last century.
- Centralized structure and distributed structure

4. 4
Introduction
The Future of Global Energy / 3D’s
● Decarbonize
● Decentralize
● Democratize

5. 5
What is a microgrid ?
A Microgrid is a scaled down version of the
centralized power system that generates,
distributes, and regulates the flow of
electricity.

6. 6
What is Microgrid?
● Small-scale power supply network.
● Local power generation for local loads.
● Highly flexible and efficient.
● Connected to both the local generating
units and the utility.
● Excess power can be sold to the utility grid.
● Size may vary.

7. 7
Advantages of microgrids ?
● Improvements in local energy distributions
● Increasing power reliability
● Financial benefits for both consumer and utility
● Makes utility grid more resilience
● Environment friendly green concept.

8. 8
Drawbacks of current microgrid systems ?
● Voltage, frequency and power quality must be
considered and controlled to acceptable
standards.
● High installation cost
● Standards are not yet available for addressing
operation and protection issues.
● Resynchronization with the utility grid is
difficult.
● Microgrid protection

9. 9
Objectives
Main objective
- To design and develop multi-agent controlled micro grid
with multiple agents for distributed energy sources
which interact each other and work as an integrated
system.
Specific objectives
- To propose control algorithms, communication
protocols and data collection patterns
- To develop an energy management system based on
MAS.

10. 10
Methodology
- Conduct a detailed research and study on current
exiting designs and concepts
- Exploring components, tools related to the system
design requirements.
- Define the control and network requirements for the
system
- Designing and implementing the agents and develop
internal wireless communication.
- Validate the proposed system in a test-bed

11. 11
What is an Agent ?
- A software designed in a computer system
- An agent can perform,
* Autonomous
* Logical
* Predefined
actions to achieve design objectives of a system.
- An agent has four basic characteristics
* Controlling
* Learning
* Optimizing
* Reasoning

12. 12
What is Multi Agent Concept ?
An MAS is a software based computer system
that is situated in some environment, and that is
capable of taking autonomous action in this
environment in order to meet its design
objectives.
Characteristics of a multi agent system
• Autonomy
• Reactivity
• Proactiveness
• Social ability

13. 13
A Multi-Agent System

14. 14
Design and Implementation of Multi-Agent System
Server
Agent
Load
Agent
Load
Agent
Load
Agent
DES
Agent
DES
Agent

15. 15
Proposed System Architecture

16. Operation of Agents
DES Agent
Load Agent
Server Agent
Main Grid Circuit Breaker and
Measurements
DER Circuit
Breaker and
Measurements
Load Circuit
Breaker and
Measurements

17. 17
Design and Implementation of Multi-
Agent System
Responsibilities of a Server agent
● Accept incoming registration
● Respond to registrations
● Monitor system voltage
● Update subscribers
● Interpret main grid state
● Store information

18. 18
Design and Implementation of Multi-
Agent System
Agent Specifications and Responsibilities DES
agent
● Register with the server agent
● Communicate with other user or DER agents
● Sense and react to the external measurements
● Update the user on current progress to Server
agent

19. 19
proposal
received ?
Request
message
connect?
confirm
status ?
Send disconnect command to
breaker
Send status message to control
agent
Send accept proposal with
available power
Start
Yes
No
Yes
No
Send connect command to
breaker
End
No
Yes
Operation of DES Agent
Measurements

20. Micro-grid Grid Outage Management using Multi Agent Systems
- Use Multi-Agent concept for Energy management in
Microgrids
- Dynamic model considering the intermittent nature of DES
and Loads
- Simulation of the model is developed
Research Gap
- Dynamic model developed only for solar pv source and
loads.
- Centralized structure
- No communication between Agents
[1]Leo Raju, Antony Amalraj Morais, Ramyaa Rathnakumar, P Soundaryaa, L D Thavam, Micro-grid
Grid Outage Management using Multi Agent Systems,Energy Procedia,2017
Literature Review

21. Distributed Optimization of Solar Micro-grid using Multi Agent
Reinforcement Learning
- Use Multi-Agent concept for Energy management in Solar PV
based micro grid
- The agent can reason, control,learn, optimize and coordinate
with other agents
Research Gap
- Model developed for only 2 Agents
- Less complexity of the considered problem
- Only proposing the concept - no model developed.
[2]Leo Raju, Sibi Sankar, R.S. Milton,Distributed Optimization of Solar Micro-grid Using Multi Agent
Reinforcement Learning,Procedia Computer Science,2015
Literature Review

22. Consideration of Smart Grid Operations by Multi-Agent
- An autonomous Smart Grid operation by using multi-agent
approach
- Proposed multi-agent system consists of five types of agent
and distributed communication arrangement
Research Gap
- no two way communication structure
[3]Yuji Ueda, Takeshi Nagata,Consideration of Smart Grid Operations by Multi-Agent,Energy
Procedia,2012
Literature Review

23. Multi-Agent-Based Cloud Architecture of Smart Grid
- A cloud architecture of smart grid is proposed based Multi-
Agent technology
- Can achieve dynamic, simple management and
expansibility, extremely low cost, flexible.
Research Gap
- Downtime, cloud computing systems are internet-based,
service outages are always an unfortunate possibility and
can occur for any reason.
- Low Security
[4]Xinkun Jin, Zijun He, Zongqi Liu,Multi-Agent-Based Cloud Architecture of Smart Grid,Energy
Procedia,2011
Literature Review

24. Fuzzy Q-Learning for multi-agent decentralized energy
management in microgrids
- Proposes a cooperative multi-agent system for managing the
energy of a stand-alone microgrid
- The multi-agent system learns to control the components of
the microgrid
- Distributed multi-agent system
Research Gap
- Dynamic model developed only for solar pv source and
loads.
[5]P. Kofinas, A.I. Dounis, G.A. Vouros,Fuzzy Q-Learning for multi-agent decentralized energy
management in microgrids,Applied Energy,2018
Literature Review

25. A multi-agent based energy management solution for
integrated buildings and microgrid system
- Multi-agent based energy management system is presented.
- Real-time monitoring and optimal control
- Cost-effective and comfort-aware energy management
solutions are presented.
Research Gap
- Centralized structure
- System was only mathematically formulated
- Communication channels are limited
[6]Amjad Anvari-Moghaddam, Ashkan Rahimi-Kian, Maryam S. Mirian, Josep M. Guerrero, A multi-
agent based energy management solution for integrated buildings and microgrid system,
Literature Review

26. Modelling and optimal management of distributed
microgrid using multi-agent systems
- A novel distributed energy management system using multi-
agent system is proposed.
- Different agents were installed in the MG to control and
supervise the system autonomously.
- The communication of intelligent agents
Research Gap
- Centralized structure with one way communication channels
[7]Muhammad Waseem Khan, Jie Wang, Linyun Xiong, Meiling Ma,Modelling and optimal
management of distributed microgrid using multi-agent systems,Sustainable Cities and Society,2018
Literature Review

27. 27
Summary of Literature Review
Research Decentralized
Structure
Two way
communication
Multiple
Interacting
Agents
Practically
implemented
[1] Leo Raju et al. , 2018 Yes No No No
[2] Sibi Sankar et al. , 2015 Yes Yes No Yes
[3] Yuji Ueda et al , 2012 Yes Yes Yes No
[4] Xinkun Jin et al .,2011 Yes No Yes No
[5] P. Kofinas et al., 2018 Yes Yes Yes No
[6] Amjad Anvari-Moghaddam
et al., 2018
No Yes Yes No
[7] Muhammad Waseem Khanet
al., 2018
No No Yes No

28. 28
Proposed Hardware
Controller module – Wifi based
ESP8266 development board
(NodeMCU)
● Processor: L106 32-bit RISC
microprocessor
● 32 KiB instruction RAM
● TCP/IP
● Open Source IoT platform
● Firmware – ESP8266 Wi-Fi SOC
● Hardware ESP 12

29. Months starting from
January Jan Feb Mar April May June July Aug Sep Oct Nov Dec
Literature Review
+
Publishing a Review
Paper
Design and Simulate
the design and error
correction
Publishing a
conference paper
based on simulation
Hardware
Implementation
Thesis and other
required document
preparation
Timeline
Prograss Review 1 Prograss Review 2

30. 30
References
[1]Leo Raju, Antony Amalraj Morais, Ramyaa Rathnakumar, P Soundaryaa, L
D Thavam, Micro-grid Grid Outage Management using Multi Agent
Systems,Energy Procedia,2017
[2]Leo Raju, Sibi Sankar, R.S. Milton,Distributed Optimization of Solar
Micro-grid Using Multi Agent Reinforcement Learning,Procedia Computer
Science,2015
[3]Yuji Ueda, Takeshi Nagata,Consideration of Smart Grid Operations by
Multi-Agent,Energy Procedia,2012
[4]Xinkun Jin, Zijun He, Zongqi Liu,Multi-Agent-Based Cloud Architecture of
Smart Grid,Energy Procedia,2011

31. 31
References
[5] P. Kofinas, A.I. Dounis, G.A. Vouros,Fuzzy Q-Learning for multi-agent
decentralized energy management in microgrids,Applied Energy,2018
[6] Amjad Anvari-Moghaddam, Ashkan Rahimi-Kian, Maryam S. Mirian,
Josep M. Guerrero, A multi-agent based energy management solution
for integrated buildings and microgrid system,Applied Energy,2017
[7] Muhammad Waseem Khan, Jie Wang, Linyun Xiong, Meiling
Ma,Modelling and optimal management of distributed microgrid using
multi-agent systems,Sustainable Cities and Society,2018

32. 32
Thank you !