基于置信规则库的汽车视镜系统声音信号故障诊断

A fault diagnosis of sound signals method for automotive mirror system based on belief rule base

考虑到声音信号非接触、获取方便的优势,提出了一种基于置信规则库(Belief Rule Base, BRB)的汽车视镜系统声音信号故障诊断方法。首先,采用多尺度散布熵(Multiscale Dispersion Entropy, MDE)来提取视镜系统声音信号特征;然后,融合提取的特征及专家知识建立BRB故障诊断模型;之后,采用协方差矩阵适应进化策略优化算法(Projection Covariance Matrix Adaptive Evolutionary Strategy, P-CMA-ES)对BRB中专家给定的初始参数进行优化,提高模型精度;最后,利用某型汽车视镜系统耐久试验过程的声音信号监测数据验证了所提方法的有效性和准确性。

Considering the advantages of non-contact and convenient acquisition of sound signals,this paper proposes a belief rule base(BRB)-based fault diagnosis method for the sound signals of automotive mirror system.Firstly,multiscale dispersion entropy(MDE)is used to extract the features of the sound signals;then,the extracted features and the expert's empirical knowledge are fused to establish a belief rule base fault diagnosis model;finally,the projection covariance matrix adaptive evolutionary strategy(P-CMA-ES)is used to optimize the initial parameters given by the experts in the BRB to improve the accuracy of the model.Finally,the effectiveness and accuracy of the proposed method are verified using the sound signal monitoring data using the endurance test of a certain type of automobile sight glass system.