基于PSO的工程机械故障智能诊断及应用

打开文本图片集

中图分类号：TP311;TQ018 文献标识码：A 文章编号：1001-5922（2025）12-0241-04

Abstract：Inorder to improve the accuracyand eficiency of engineering machinery fault diagnosis，this paper proposes an inteligent diagnosis model（PSO-RBF） based on particle swarm optimization algorithm（PSO）to optimize radial basis function network（RBF）.The study takes elevator vibration signals asexperimental objects，uses multi-scale permutation entropy and local mean decomposition methods to extract signal features，and constructs featurevectors to characterize fault states；Then，the PSO algorithm is used todynamicallyoptimize the hidden layer parametersof theRBFnetwork，solving theproblemof traditionalRBFnetworksbeingsensitivetoinitial parameters and prone tofalling into localoptima，whichcaneffectively improve theaccuracyoffaultclassification.The results showed that the PSO-RBF model achieved an average diagnostic accuracy of 96.5% in mild，moderate，and severe fault states，witha training time ofonly12.3 seconds，outperforming traditional RBFandBP networks in performance.In addition，the inteligent monitoring system designed based on this model canachievereal-time warning and remote management offaults，providing eficient solutions forthe intellgent operationand maintenance of construction machinery.

Key words：engineering machinery;inteligent fault diagnosis;particle swarm optimization;radial basis function network

随着现代工业的快速发展，工程机械在建筑、制造、交通等领域的作用日益凸显，其运行状态的稳定性直接关系到生产效率和安全性。（剩余5467字）