基于遗传算法的WANN的汽车发动机故障诊断分析  

The analysis of fault diagnosis for automotive engine based on genetic algorithm,wavelet transform and neural Network

作　　者：余丙荣[1] 周明龙[2] 李玲纯[2] 

机构地区：[1]安徽机电职业技术学院,安徽芜湖241000 [2]安徽工程大学,安徽芜湖241000

出　　处：《安徽工程大学学报》2011年第1期55-57,共3页Journal of Anhui Polytechnic University

摘　　要：利用基于遗传算法的全局优化能力,小波分析具有数据压缩和特征提取的特性,神经网络具有非线性映射和学习推理的优点.结合三者的特点,提出了一种基于遗传算法、小波与神经网络的汽车发动机故障诊断方法,应用汽车发动机的故障数据作为实例验证,GA-WANN模型诊断速度快,鲁棒性好,故障诊断正确率高.By means of genetic algorithms-based global optimization,wavelet analysis prosesses the features of data compression and feature extraction and neural networks have the nonlinearity mapping and the advantages of learning reasoning,combining the characteristics of the three,a method of afault diagnosis for automotive engine based on genetic algorithm,wavelet transform and neural network is presented,the practical application to fault diagnosis for gasoline engine verifies that the model enjoys the features of being fast,good robustness and accurate diagnosing.

关 键 词：小波变换 人工神经网络 遗传算法 故障诊断 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]