摘要
液压系统故障诊断中采用BP神经网络进行故障的模式识别,存在着收敛率较低的问题。结合粒子群算法和BP算法各自的优势,提出了一种基于改进的PSO-BP液压系统故障诊断方法。对标准粒子群算法的惯性权重和学习因子进行改进,再对BP神经网络的权值和阈值进行优化,达到改善BP网络性能的目标。仿真结果表明该方法提高了BP网络的收敛率,减小了诊断误差。
Hydraulic system fault diagnosis uses BP neural network for failure pattern recognition, but it exists the problem of low convergence rate. Combining the respective advantages of particle swarm algorithm and BP algorithm,a new modified PSO-BP hydraulic system fault diagnosis methods was proposed. Firstly, the inertia weight and learning factor of the standard particle swarm algorithm was improved, then BP neural network weights and threshold is optimized by modified PSO algorithm. BP network performance was ameliorated. The simulation results show that this method improves the convergence rate of the BP network, and it could diagnose the failure of engineering machinery hydraulic system.
出处
《液压气动与密封》
2012年第11期13-14,18,共3页
Hydraulics Pneumatics & Seals
关键词
粒子群算法
BP神经网络
液压系统
故障诊断
particle swarm optimization algorithm
BP neural network
hydraulic system
fault diagnosis