基于优化RBF网络的抽油机故障诊断方法研究

Research on Fault Diagnosis Method of Pumping Unit Based on Optimized RBF Network

本文介绍了RBF神经网络结构和原理,针对RBF神经网络在抽油机故障诊断中核函数参数的局限性,使用K-Means算法优化RBF网络中心参数,使用动态参数的PSO算法优化RNF网络权值和宽度参数,建立PSO-RBF神经网络。最后将PSO-RBF神经网络与RBF神经网络应用于抽油机故障诊断,证明了优化后的PSO-RBF神经网络在计算速度和诊断准确率上更加优秀。

This paper introduces the structure and principle of RBF neural network.In order to limit the kernel function parameters of RBF neural network in pumping unit fault diagnosis,K-Means algorithm is used to optimize RBF network center parameters,and dynamic parameter PSO algorithm is used to optimize RNF network.The weight and width parameters are used to establish the PSO-RBF neural network.Finally,the PSO-RBF neural network and RBF neural network are applied to the pumping unit fault diagnosis,which proves that the optimized PSO-RBF neural network is superior in calculation speed and diagnostic accuracy.