基于遗传算法优化BP神经网络的提升机制动系统故障诊断  

Fault Diagnosis of Hoist Braking System Based on BP Neural Network Optimized by Genetic Algorithm

作　　者：刘锦荣[1] 王绍进[1] 任芳[1] 杨兆建[1] 

机构地区：[1]太原理工大学机械工程学院,太原030024

出　　处：《煤矿机械》2011年第5期246-248,共3页Coal Mine Machinery

基　　金：山西省留学基金资助项目(200419)

摘　　要：针对提升机故障诊断系统的复杂性,利用浮点数编码的遗传算法优化BP神经网络的初始权值和阈值,再按BP算法沿负梯度方向进行网络学习直至收敛,构建起遗传算法优化BP神经网络的诊断方法。将此诊断方法应用于2JTP-1.2型提升机液压制动系统,诊断结果表明,该方法用于提升机液压制动系统常见故障诊断可行。For the complexity of hoist fault diagnosis system,initial weights and thresholds of neural network have been optimized by using floating-point coded genetic algorithm,the method of network learning and converging was analyzed using BP with negative gradient searching,and the neural network diagnostic method of genetic algorithm optimization was built.The diagnosis method has been applied to 2JTP-1.2 hoist hydraulic braking system,and the results show that the method used in the common fault diagnosis of hoist hydraulic system is feasible.

关 键 词：遗传算法 神经网络 提升机 故障诊断 

分 类 号：TD444[矿业工程—矿山机电]