基于粒子群优化RBF神经网络的水轮发电机组振动故障诊断  

Hydraulic generating vibration faults diagnosis by RBF neural network based on particle swarm optimization

作　　者：贾嵘[1] 陈晓芸[1] 李辉[1] 席文飞[2] 

机构地区：[1]西安理工大学电力工程系,陕西西安710048 [2]内蒙古电力科学研究院,内蒙古呼和浩特010020

出　　处：《西北农林科技大学学报（自然科学版）》2009年第6期229-234,共6页Journal of Northwest A&F University(Natural Science Edition)

基　　金：陕西省科技厅2007年工业攻关计划项目(2007K05-15)

摘　　要：【目的】针对单一径向基(RBF)神经网络在水轮发电机组振动故障诊断中泛化能力不足的缺点,提出基于粒子群(PSO)算法优化的RBF神经网络。【方法】利用PSO算法操作简单、容易实现等特点及其深刻的智能背景,对RBF神经网络的参数(中心和宽度)、连接权重进行优化,并用经PSO算法优化的RBF神经网络对水轮发电机组振动故障进行仿真诊断。【结果】仿真诊断结果表明,PSO算法优化的RBF神经网络具有较好的分类效果,较RBF诊断模型精度高、收敛快。【结论】PSO算法优化的RBF神经网络,适用于水轮发电机组振动故障诊断,其诊断精度较高,具有推广应用价值。【Objective】For the system of vibration faults diagnosis of hydraulic generating,the deficiency of generalization ability using single BP Network is analyzed and a Radial Basis Function（RBF） Neural Network algorithm based on Particle Swarm Optimization（PSO） is presented.The system of vibration faults diagnosis of hydraulic generating is simulated.【Method】Being easy to realize,simple to operate with profound intelligence background,the parameters and connection weight are optimized by the algorithm and vibration faults diagnosis of hydraulic generating is simulated by the optimized RBF Neural Network.【Result】The diagnostic results of the instance show that it has better classifying results,higher precision,faster convergence than that of BP diagnosis model.【Conclusion】The optimized RBF Neural Network is suitable for fault diagnosis of hydraulic generating.The method has good diagnosis accuracy and popularization value.

关 键 词：水轮机 振动故障诊断 粒子群算法 神经网络 

分 类 号：TM312[电气工程—电机]