In response to the problems of unclear boundaries,insufficient division criteria,and premature convergence of training algorithms,an algorithm model for dividing and identifying the operational states of AC contactors based on weighted grey target and Levy Flight Particle Swarm Optimization (LPSO) algorithm is proposed.Firstly,the threshold denoising and feature selection are performed on the raw data.Secondly, by using the Analytic Hierarchy Process (AHP) to weight and improve the grey target algorithm,the grey target value of the entire life cycle of the AC contactor is obtained as the data basis for dividing the operational state into three states: good, general, and high-risk.Finally, the BP neural network optimized by LPSO (LPSO-BP) algorithm is used for classification prediction,and compared with other algorithms to verify that the proposed algorithm has better stability and accuracy.
