基于提升小波变换与EEMD的神经网络齿轮故障诊断方法

发布时间：2018-01-09 09:38

本文关键词：基于提升小波变换与EEMD的神经网络齿轮故障诊断方法　出处：《组合机床与自动化加工技术》2017年09期 　论文类型：期刊论文

【摘要】：针对单一的信号处理诊断方法难以实现齿轮故障准确诊断的局限性,文章将提升小波变换、集成经验模态分解(ensemble empirical mode decomposition,EEMD)与相关系数相结合,提出一种新的信号消噪方法,并在此方法的基础上,分别利用BP、Elman和RBF神经网络完成了齿轮故障诊断。首先采用提升小波变换对故障信号进行初步消噪,然后对其作EEMD分解,得到一组固有模态函数(intrinsic mode function,IMF)分量;然后计算各分量的相关系数,剔除相关性较小的伪分量后进行重构,完成二次消噪;最后计算剩余分量的能量特征,并将其作为神经网络的输入向量,进而完成齿轮断齿、裂纹和磨损状况下的故障诊断。仿真分析和应用实例表明:基于提升小波变换与EEMD分解并结合相关系数筛选的消噪方法,比仅用提升小波方法消噪的效果更好。三种神经网络均成功辨别出了齿轮的故障类型,但不同方法各有优劣之处;就诊断效率和准确性而言,BP神经网络的诊断效果最好。
[Abstract]:Aiming at the limitation that it is difficult to realize the accurate diagnosis of gear fault by single signal processing diagnosis method, this paper will raise wavelet transform. The integrated empirical mode decomposition (EMD) is combined with the correlation coefficient. A new signal denoising method is proposed, and based on this method, BP is used separately. Elman and RBF neural network complete the gear fault diagnosis. Firstly, the lifting wavelet transform is used to eliminate the noise of the fault signal, and then it is decomposed by EEMD. A set of intrinsic mode function IMF components are obtained. Then the correlation coefficients of each component are calculated and the pseudo-components with less correlation are eliminated and reconstructed to complete the quadratic de-noising. Finally, the energy characteristic of the residual component is calculated and used as the input vector of the neural network to complete the gear tooth breaking. Fault diagnosis under the condition of crack and Wear. Simulation analysis and application examples show that the noise reduction method is based on lifting wavelet transform and EEMD decomposition combined with correlation coefficient screening. The three neural networks have successfully identified the fault types of gears, but different methods have their own advantages and disadvantages. In terms of diagnostic efficiency and accuracy, BP neural network has the best diagnostic effect.
【作者单位】：宁德师范学院物理与电气工程系;上海大学机电工程与自动化学院;
【基金】：福建省自然科学基金项目(2015J01643) 福建中青年教师教育科研项目(JA15545) 宁德市科技计划项目(20150034) 福建省高校杰出青年科研人才培育计划资助(闽教科[2015]54号)
【分类号】：TH132.41
【正文快照】： 海200072)0引言齿轮是机械设备中最重要的连接和传动部件,及时发现齿轮在传动过程中的故障,能经济合理地安排设备的维护与维修时间,避免发生事故[1]。目前的齿轮故障诊断方法多建立于采集振动信号的基础上,尚无较为系统的诊断方法,因此本文对齿轮故障诊断方法的探索具有一定的

【相似文献】