Evolutionary and Swarm Based Optimization of Fit k-Nearest Neighbor Classifier for Classification of Power Quality Disturbances

Abstract

In the electric network, integration of renewable energy sources and switching in large industrial loads and non-linear loads result in Power Quality Disturbances (PQDs). PQDs arise malfunctioning the sensitive equipment integrated with power grid; therefore, it is necessary to overcome the effects of poor power quality (PQ) in electric power supply. It is possible to monitor PQ with the advancement of signal processing and artificial intelligence automatically. In this paper, techniques of automatic detection and classification of PQDs are proposed using discrete wavelet transform–multi-resolution analysis (DWT–MRA) and evolutionary and swarm intelligence algorithms in renewable integrated power grid. Fit k-nearest neighbor (KNN) classifier along with DWT–MRA technique is used to classify the PQDs and it is optimized using evolutionary and swarm intelligence algorithms like, genetic algorithms, particle swarm optimization, and grey wolf optimization. Performance of algorithm is compared using accuracy of classification and confusion matrix. Further robustness of classifier is tested with Simulink model of IEEE bus test system that indicated the renewable integrated power grid environment.

Keywords:

• power quality disturbances
• discrete wavelet transform
• binary grey wolf optimization
• renewable integrated power grid

AUTHOR CONTRIBUTIONS

Dazi Li: supervision, data curation, investigation, formal analysis; Irfan Ali Channa: conceptualization, data curation, formal analysis, investigation, methodology, resources, validation, writing—original draft; Xun Chen: formal analysis, data curation; Li Song: investigation, formal analysis.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported in part by the National Natural Science Foundation of China under Grant 62273026.

Notes on contributors

Dazi Li

Dazi Li received the Ph.D. degree in engineering from the Department of Electrical and Electronic Systems, Kyushu University, Fukuoka, Japan, in 2004. She is currently a Full Professor and the Vice Dean of the College of Information Science and Technology, Beijing University of Chemical Technology, Beijing, China. Her research interests include machine learning and artificial intelligence, advanced process control, complex system modeling and optimization, and fractional calculus system. Prof. Li is currently an Associate Editor of ISA Transactions.

Irfan Ali Channa

Irfan Ali Channa Received his Bachelor and Master degrees from the Electrical Engineering Department, Quaid e Awam University of Engineering Science and Technology, Nawabshah Pakistan. Currently, he is PhD student in the Institute of Automation, Beijing University of Chemical Technology, Beijing China. His interests include Power quality, signal processing, machine learning, deep learning, big data analysis, Renewable energy.

Xun Chen

Xun Chen Received Bachelor and Master degrees from Institute of Automation, Beijing University of Chemical Technology, Beijing China, Currently he is doing PhD in Beijing University of Chemical Technology Beijing China. His research interests include: Machine learning, PID controller, ADRC and optimization.

Li Song

Li Song Received PhD degree from Institute of Automation, Beijing University of Chemical Technology, Beijing China, Currently she is doing PostDoc in Zhejiang University, Hangzhou China. Her research interests include: Reinforcement learning, Proximal optimization, Q-learning and inverse Reinforcement learning.