摘要
发动机作为一种非线性对象,导致传统诊断方法的结果存在极强的不确定性,同时为了提升汽车发动机的使用寿命,降低发动机维修人员工作强度,提出一种基于人机一体化的汽车发动机故障诊断方法。通过小波窗口分析发动机形状的变化,得到发动机振动信号的细节与概貌,构建小波系数矩阵对振动信号降噪,并遵照曲轴转角分割时域能量,获取振动信号特征,依靠支持向量机算法组建汽车发动机特征分类器,最后利用人工蜂群方法计算权重函数优化故障分类器,使用该分类器实现对汽车发动机故障的诊断。实验证明,所提方法在汽车发动机故障诊断中,有着诊断精确度高与诊断范围大的优点。
As a non-linear object,the results of traditional diagnosis methods for engines are highly uncertain.At the same time,in order to improve the service life of automobile engine and reduce the work intensity of engine maintenance personnel,a fault diagnosis method of automobile engine based on human-computer integration is proposed.Through the wavelet window analysis of the engine shape changes,the details and overview of the engine vibration signal are obtained,the wavelet coefficient matrix is constructed to de noise the vibration signal,and the time domain energy is segmented according to the crankshaft angle to obtain the vibration signal features.The support vector machine algorithm is used to build the vehicle engine feature classifier.Finally,the artificial bee colony method is used to calculate the weight function to optimize the fault classification.The classifier is used to realize the fault diagnosis of automobile engine.Experiments show that the proposed method has the advantages of high diagnosis accuracy and wide diagnosis range in the automotive engine fault diagnosis.
作者
黄元银
黄雪
HUANG Yuan-yin;HUANG Xue(Repair Branch,Hangzhou Xiaoshan Public Transportation Co.,Ltd.,Hangzhou 311201)
出处
《环境技术》
2021年第2期183-187,193,共6页
Environmental Technology
关键词
人机一体化
汽车发动机
故障诊断
小波分析
故障分类器
man machine integration
automobile engine
fault diagnosis
wavelet analysis
fault classifier