遗传神经网络在齿轮故障诊断中的应用  

Application of Genetic Neural Network to Gear Fault Diagnosis

作　　者：刘景艳[1] 李玉东[1] 杨晓邦[1] 

机构地区：[1]河南理工大学电气工程与自动化学院,河南焦作454000

出　　处：《郑州大学学报（工学版）》2012年第3期36-39,共4页Journal of Zhengzhou University（Engineering Science）

基　　金：中国煤炭工业协会科技计划资助项目(MTKJ08-343)

摘　　要：由于齿轮故障征兆与故障之间具有非线性和耦合性等特点,采用BP神经网络对齿轮进行故障诊断存在着收敛速度慢和可靠性差等缺点,提出了一种基于遗传算法的BP神经网络齿轮故障诊断方法,即在利用BP神经网络对齿轮进行故障诊断的基础上,利用遗传算法对神经网络的权值和阈值进行修正,得到全局的最优值.仿真结果表明,该诊断策略具有故障诊断能力强和诊断效率高的特点,改善了齿轮故障诊断的精度和速度.Because gear faults have characteristics of being nonlinear and coupling between fault symptoms and fault, BP neural network gear fault diagnosis has slow convergence speed and poor reliability. The BP neural network gear fault diagnosis method based on genetic algorithm is put forward. Namely, the BP neural network is used in gear fault diagnosis, and genetic algorithm is applied to optimize the weights and thresholds of the network. So the global optimal value is obtained. The simulation results show that the diagnosis strategy has the characteristic of strong diagnosis ability and high diagnosis efficiency and improves the gear fault diagnosis precision and speed.

关 键 词：齿轮 故障诊断 神经网络 遗传算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]