Real-Time Comprehensive Assessment of IGJO-CNN Power Quality Based on Combined Empowerment Approach
J F Wang, Yongzhi Min · 2024
An IGJO-CNN (Improved Golden Jackal Optimization- Convolutional Neural Network) real-time hierarchical power quality assessment method based on the combined assignment method is proposed. A comprehensive power quality assessment system is constructed; a weighting matrix for power quality assessment is constructed by combining the fuzzy hierarchical analysis method, CRITIC method, and state weighting; the Improved Golden Jackal Optimization (IGJO) algorithm is used to realize the optimization of the hyper-parameters in the convolutional neural network (CNN) model; and the entropy weighting method is used to synthesize real-time grades of power quality to obtain the comprehensive assessment grade of power quality, which is used as a criterion for the comprehensive assessment grade. The overall state evaluation of power quality based on the comprehensive evaluation level and the analysis of the causes of level change based on the real-time evaluation level are obtained. The example analysis shows that the proposed method can obtain accurate and comprehensive power quality grading results.