Yingqi Xiong
403 S. New Ave., Monterey Park, CA 91755
(847) 868-5857 • yxb936@ucla.edu
https://www.linkedin.com/in/yingqi-xiong-8224a856/
I am a Ph.D student working on machine learning and optimization. I am seeking for 2018 summer internship opportunities
in machine learning, software engineering and data science. My research is mainly focusing on electricity load forecasting,
electric vehicle user behavior prediction and optimal distributed energy resources management & transaction. In addition
to my research, I have also contributed to the infrastructure construction for UCLA microgrid system, such as developing
charging station controller firmware, building control center server and database, etc. I have successfully delivered programs
and control algorithms written in Python, C++ and C to, and provided continuous customer support and updates for more
than 300 active users.
EDUCATION
○
University of California, Los Angeles (UCLA), Los Angeles, CA Expected December 2018
Doctor of Philosophy in Mechanical Engineering (GPA: 3.75/4.0)
○
Northwestern University, Evanston, IL June 2014
Master of Science in Mechanical Engineering (GPA: 4.0/4.0)
○
Northeastern University, China June 2012
Bachelor of Engineering in Mechanical Engineering and Automation (GPA: 3.82/4.0)
RESEARCH EXPERIENCE
University of California, Los Angeles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
○ Solar Power Generation Forecasting via Long-Short-Term-Memory Networks (LSTM) 2017
• Designed a symmetric Auto-Encoder (AE) as feature extractor to process numerical weather prediction data
• Developed an AE-LSTM network by connecting with split pre-trained AE then fine-tuning weights of LSTM
• Outperformed conventional ARIMA technique and other DNNs with lower RMSE and Bias
○ Hybrid Deep Belief Network (DBN) - Support Vector Regression (SVR) for Load Forecasting 2017
• Generated outputs from stacked Restricted Boltzmann Machine with different back propagation epochs
• Channeled all DBN outputs into Radial Basis Function (RBF) kernel SVR and trained with expected electricity load
• Provided high score accurate load forecasting as guidance for implementing energy policy
○ Electric Vehicle User Behavior Classification using Multilayer Perceptron (MLP) 2016
• Applied PCA to reduce the dimensions of over 40GB electric vehicle user charging record data
• Created MLP network in TensorFlow with sigmoid activation function to classify user behavior
• Fine-tuned MLP hyper-parameters using exhaustive grid search with cross validation to obtain an 85% accuracy
○ Distributed stochastic Algorithm for Optimal Energy Resources Management in Microgrid 2015
• Designed decentralized algorithm using first-order optimal condition and Alternating Direction Method of Multipliers
• Built stochastic user model describing uncertainty by Sample Average Approximation and Monte-Carlo simulation
• Integrated solar panel and electric vehicle into UCLA microgrid achieved 35% power quality improvement
○ Microgrid Energy Management Algorithm Complexity Reduction and Real-Time Implementation 2015
• Developed Augmented Lagrangian optimization solver in C++ which outperformed Gradient Decent in speed
• Merged algorithm with dynamic programming for real-time implementation in UCLA and Santa Monica
• Reduced algorithm time complexity from O(n2
) to O(nlogn)
1/2
Northwestern University. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
○ Laser Shock Peening Metal Surface Enhancement and Optimization 2014
• Developed laser shock peening platform control system and firmware for fully automatic operation
• Optimized laser induced plasma parameters to increase metal surface fatigue resistance and hardness by 20%
○ Force Control in Double Sides Incremental Forming (DSIF) via Statistical Modeling and Finite Element 2013
• Designed a force-feedback based tool control strategy for DSIF machine
• Combined statistical model with finite element analysis to predict possible fracture location
PROFESSIONAL EXPERIENCE
○
Smart Grid Energy Research Center Los Angeles, CA
Software Development Engineer July 2015 - Sept 2017
• Built an http server using Python Twisted framework to process requests from user and control center
• Conducted Test Driven Development (TDD) to develop firmware in C and Python for electric vehicle charging station
• Deployed distributed optimal charging control program in 217 stations reducing total energy cost by 40%
○
Smart Grid Energy Research Center Los Angeles, CA
Electrical Engineer July 2014 - June 2015
• Designed pilot signal and Ground Fault Interrupter (GFI) circuit for electric vehicle charging station
• Lead a team of 10 people to manufacture charging stations for UCLA, LADWP, SCE and the city of Santa Monica
• Established communication network and conducted embedded system on-site debugging for charging stations
SPECIALTIES and SKILLS
○ Specialties: Machine Learning, Deep Learning, Optimization, Big Data Pipeline & Analytics, Finite Element Analysis
○ Programming Languages: Python, C, C++, Java, SQL, Hadoop, Bash
○ Machine Learning: pandas, sklearn, TensorFlow, PyTorch.
○ Industry Software: Matlab, Abaqus, SolidWorks, Powerworld, RSCAD
PUBLICATIONS
○ Real-Time Bi-directional Electric Vehicle Charging Control with Distribution Grid Implementation
Yingqi Xiong, Behnam Khaki, Peter Chu, Rajit Gadh
IEEE Transmission & Distribution Conference and Exposition, Denver, Colorado, 16-19 April 2018
○ Distributed Optimal Vehicle Grid Integration Strategy with User Behavior Prediction
Yingqi Xiong, Bin Wang, Peter Chu, Rajit Gadh
IEEE PES General Meeting, Chicago, Illinois, 16-20 July 2017
○ Extension of IEC61850 with smart EV charging
Yingqi Xiong, Bin Wang, Peter Chu, Rajit Gadh
IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia), Melbourne, Australia, 2016, pp. 294–299
○ IoT based manufacturing system with a focus on energy efficiency
Zhiyuan Cao; Yu-Wei Chung; Yingqi Xiong; Chi-Cheng Chu; Rajit Gadh
IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia), Melbourne, Australia, 2016, pp. 545-552
CERTIFICATIONS
○ Predictive Analytics and Optimization for Electric Power Distribution
Electric Power Research Institute (EPRI), 2017
○ Machine Learning by Stanford University
Coursera, 2016
2/2

YingqiCV

  • 1.
    Yingqi Xiong 403 S.New Ave., Monterey Park, CA 91755 (847) 868-5857 • yxb936@ucla.edu https://www.linkedin.com/in/yingqi-xiong-8224a856/ I am a Ph.D student working on machine learning and optimization. I am seeking for 2018 summer internship opportunities in machine learning, software engineering and data science. My research is mainly focusing on electricity load forecasting, electric vehicle user behavior prediction and optimal distributed energy resources management & transaction. In addition to my research, I have also contributed to the infrastructure construction for UCLA microgrid system, such as developing charging station controller firmware, building control center server and database, etc. I have successfully delivered programs and control algorithms written in Python, C++ and C to, and provided continuous customer support and updates for more than 300 active users. EDUCATION ○ University of California, Los Angeles (UCLA), Los Angeles, CA Expected December 2018 Doctor of Philosophy in Mechanical Engineering (GPA: 3.75/4.0) ○ Northwestern University, Evanston, IL June 2014 Master of Science in Mechanical Engineering (GPA: 4.0/4.0) ○ Northeastern University, China June 2012 Bachelor of Engineering in Mechanical Engineering and Automation (GPA: 3.82/4.0) RESEARCH EXPERIENCE University of California, Los Angeles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ○ Solar Power Generation Forecasting via Long-Short-Term-Memory Networks (LSTM) 2017 • Designed a symmetric Auto-Encoder (AE) as feature extractor to process numerical weather prediction data • Developed an AE-LSTM network by connecting with split pre-trained AE then fine-tuning weights of LSTM • Outperformed conventional ARIMA technique and other DNNs with lower RMSE and Bias ○ Hybrid Deep Belief Network (DBN) - Support Vector Regression (SVR) for Load Forecasting 2017 • Generated outputs from stacked Restricted Boltzmann Machine with different back propagation epochs • Channeled all DBN outputs into Radial Basis Function (RBF) kernel SVR and trained with expected electricity load • Provided high score accurate load forecasting as guidance for implementing energy policy ○ Electric Vehicle User Behavior Classification using Multilayer Perceptron (MLP) 2016 • Applied PCA to reduce the dimensions of over 40GB electric vehicle user charging record data • Created MLP network in TensorFlow with sigmoid activation function to classify user behavior • Fine-tuned MLP hyper-parameters using exhaustive grid search with cross validation to obtain an 85% accuracy ○ Distributed stochastic Algorithm for Optimal Energy Resources Management in Microgrid 2015 • Designed decentralized algorithm using first-order optimal condition and Alternating Direction Method of Multipliers • Built stochastic user model describing uncertainty by Sample Average Approximation and Monte-Carlo simulation • Integrated solar panel and electric vehicle into UCLA microgrid achieved 35% power quality improvement ○ Microgrid Energy Management Algorithm Complexity Reduction and Real-Time Implementation 2015 • Developed Augmented Lagrangian optimization solver in C++ which outperformed Gradient Decent in speed • Merged algorithm with dynamic programming for real-time implementation in UCLA and Santa Monica • Reduced algorithm time complexity from O(n2 ) to O(nlogn) 1/2
  • 2.
    Northwestern University. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ○ Laser Shock Peening Metal Surface Enhancement and Optimization 2014 • Developed laser shock peening platform control system and firmware for fully automatic operation • Optimized laser induced plasma parameters to increase metal surface fatigue resistance and hardness by 20% ○ Force Control in Double Sides Incremental Forming (DSIF) via Statistical Modeling and Finite Element 2013 • Designed a force-feedback based tool control strategy for DSIF machine • Combined statistical model with finite element analysis to predict possible fracture location PROFESSIONAL EXPERIENCE ○ Smart Grid Energy Research Center Los Angeles, CA Software Development Engineer July 2015 - Sept 2017 • Built an http server using Python Twisted framework to process requests from user and control center • Conducted Test Driven Development (TDD) to develop firmware in C and Python for electric vehicle charging station • Deployed distributed optimal charging control program in 217 stations reducing total energy cost by 40% ○ Smart Grid Energy Research Center Los Angeles, CA Electrical Engineer July 2014 - June 2015 • Designed pilot signal and Ground Fault Interrupter (GFI) circuit for electric vehicle charging station • Lead a team of 10 people to manufacture charging stations for UCLA, LADWP, SCE and the city of Santa Monica • Established communication network and conducted embedded system on-site debugging for charging stations SPECIALTIES and SKILLS ○ Specialties: Machine Learning, Deep Learning, Optimization, Big Data Pipeline & Analytics, Finite Element Analysis ○ Programming Languages: Python, C, C++, Java, SQL, Hadoop, Bash ○ Machine Learning: pandas, sklearn, TensorFlow, PyTorch. ○ Industry Software: Matlab, Abaqus, SolidWorks, Powerworld, RSCAD PUBLICATIONS ○ Real-Time Bi-directional Electric Vehicle Charging Control with Distribution Grid Implementation Yingqi Xiong, Behnam Khaki, Peter Chu, Rajit Gadh IEEE Transmission & Distribution Conference and Exposition, Denver, Colorado, 16-19 April 2018 ○ Distributed Optimal Vehicle Grid Integration Strategy with User Behavior Prediction Yingqi Xiong, Bin Wang, Peter Chu, Rajit Gadh IEEE PES General Meeting, Chicago, Illinois, 16-20 July 2017 ○ Extension of IEC61850 with smart EV charging Yingqi Xiong, Bin Wang, Peter Chu, Rajit Gadh IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia), Melbourne, Australia, 2016, pp. 294–299 ○ IoT based manufacturing system with a focus on energy efficiency Zhiyuan Cao; Yu-Wei Chung; Yingqi Xiong; Chi-Cheng Chu; Rajit Gadh IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia), Melbourne, Australia, 2016, pp. 545-552 CERTIFICATIONS ○ Predictive Analytics and Optimization for Electric Power Distribution Electric Power Research Institute (EPRI), 2017 ○ Machine Learning by Stanford University Coursera, 2016 2/2