Application of Comprehensive Energy Efficiency Evaluation of Enterprise Based on SVM in Virtual Power Plant Operator System

In order to help the enterprises of the virtual power plant operator system understand their own comprehensive energy efficiency and improve the accuracy of energy efficiency evaluation, a comprehensive energy efficiency evaluation method was proposed based on support vector machine. The energy efficiency indicators are filtered from the dimensions of physics, economy, ecology, transmission, power quality, etc., and a comprehensive energy efficiency indicator evaluation system is constructed based on the selected indicators. In order to reduce the subjectivity of selecting the internal parameters of SVM, a combination of grid search, genetic algorithm and particle swarm optimization was taken by comparing to select the optimal penalty factors and kernel parameters. The final experimental results show that this method improves the accuracy and rationality of the comprehensive energy efficiency evaluation, and can provide data support for enterprises to carry out energy efficiency transformation.