基于AGA-BP算法的智能故障诊断技术研究  

Research on the Technique of Fault Diagnosis Based on Adaptive Genetic Algorithm and BP Neural Network

作　　者：焦爱红[1] 袁力哲[2] 陈燕生[2] 

机构地区：[1]中国人民武装警察部队学院指挥系,河北廊坊065000 [2]炮兵指挥学院三系,河北廊坊065000

出　　处：《机床与液压》2011年第5期121-124,共4页Machine Tool & Hydraulics

摘　　要：针对传统BP神经网络的不足,提出基于自适应遗传算法的BP神经网络故障诊断算法。在迭代计算前期,采用自适应遗传算法对神经网络的权值和阈值进行全局优化;在迭代计算后期,利用改进的BP算法在近似最优解附近进行局部寻优。将该算法用于磨削烧伤的故障诊断之中,并将结果与基于改进BP网络的诊断结果进行比较,证明该方法的正确性和有效性。Fault diagnosis algorithm based on adaptive genetic algorithm and BP neural network （AGA-BP） was presented to avoid the defect of tradition BP neural networks. The adaptive genetic algorithm was used to optimize initial weights and thresholds of the BP neural network in earlier stage of iterative calculation, and the error back propagation algorithm with self study speed was used to improve the network problems of slow convergence speed in the later stage. The AGA-BP algorithm was used to diagnose grinding burn fault. The result was compared with that of the general network algorithm. It testifies the method is correct and valid.

关 键 词：故障诊断 自适应遗传算法 BP算法 

分 类 号：TH16[机械工程—机械制造及自动化] TP392[自动化与计算机技术—计算机应用技术]