基于PSO-WNN的液压泵故障诊断  

Hydraulic Pump Fault Diagnosis Based on PSO-WNN

作　　者：钱宝存 赵龙章[1] 何如意[1] 高士翔[1] 

机构地区：[1]南京工业大学自动化与电气工程学院,江苏南京211816

出　　处：《液压与气动》2014年第7期34-37,48,共5页Chinese Hydraulics & Pneumatics

基　　金：浙江省重大科技专项优先主题重大项目(2008C13024-1)

摘　　要：提出一种多理论融合的新型故障诊断方法并实验验证。选择在函数逼近、学习速率等方面有优势的径向基(RBF)网络,同时采用小波包分解对振动信号降噪并提取频带能量作为RBF的输入。引入粒子群算法(PSO)对RBF神经网络的参数进行优化。结果表明PSO优化的小波RBF神经网络能较为准确地对液压泵故障做出识别与监测。A new fault diagnosis method is proposed that multiple theories cooperate. The radial basis function （RBF） neural networks which take advantages in function approximation and learning rate are chosen. In order to improve the ability of the hydraulic pump fault diagnosis, the wavelet packet decomposition is used to reduce the noise of the vibration signal, and the band energy is taken as the input of RBF. The particle swarm optimization （PSO） is introduced to optimize the weights and the threshold value of the RBF neural network. Results indicate that optimized wavelet RBF neural network by PSO can identify and monitor the fault of the pump accurately.

关 键 词：液压泵 PSO优化 小波包分解 RBF神经网络 

分 类 号：TH137[机械工程—机械制造及自动化] TP13[自动化与计算机技术—控制理论与控制工程]