Modeling of Nonlinear Load Electric Energy Measurement and Evaluation System Based on Artificial Intelligence Algorithm

Background: To improve the modeling efficiency of nonlinear load electric energy metering evaluation systems, a method based on an artificial intelligence algorithm was proposed. Methods: First, the artificial glowworm swarm optimization extreme learning machine, a potent tool that employs the artificial firefly algorithm for global optimization, was introduced. Then, the input weighting matrix, hidden layer offset matrix, extreme learning machine model, and hours of training error were determined. Moreover, during a certain time in a specific region of China, power load simulation using an experiment was employed to validate and evaluate the model. Results: The experimental results showed that the traditional back propagation (BP) neural network had the largest prediction relative error, the stability of BP neural network was poor, and the relative error time was large, which was related to the defect of the neural network. The prediction effect of the support vector machine (SVM) method was better than that of the BP neural network because SVM has a strict theoretical and mathematical basis; thus, its generalization ability was better than that of the BP neural network, and the algorithm showed global optimality. Conclusion: The chart analysis showed that the GSO-ELM algorithm performed better in terms of stability as well as test error. The modeling nonlinear load electrical energy measurement and evaluation system based on an artificial intelligence algorithm provides better results and is effective. The proposed algorithm outperforms the contemporary ones.