论文部分内容阅读
针对滚动轴承信号的非线性、非平稳性特点及诊断中冗余与噪音的干扰,引入了核主元分析法和BP神经网络相结合的方法对轴承的故障信号进行诊断,以提高轴承故障诊断的性能。通过5个传感器采集轴承不同状态的故障信号,利用小波包提取能量特征值,同时提取轴承的时-频域特征量组成原始特征空间,利用核主元分析方法对原始特征空间降维,提取主元特征量输入到BP神经网络中进行故障模式识别。试验结果表明,KPCA-BP网络模型的性能优于未筛选-BP网络,具有更好的诊断效果和抗干扰能力。
Aiming at the nonlinear and non-stationary characteristics of rolling bearing signals and the interference of redundancy and noise in diagnosis, the method of combining KPCA with BP neural network is introduced to diagnose the bearing fault signals to improve the bearing fault diagnosis performance. Five sensors are used to collect the fault signals of different states of the bearings. The wavelet packet is used to extract the energy eigenvalues. At the same time, the time-frequency domain features are extracted to form the original feature space. The principal component analysis (PCA) is used to reduce the dimension of the original feature space. Meta-feature input to the BP neural network for fault pattern recognition. The experimental results show that the performance of the KPCA-BP network model is better than the non-screening-BP network, which has better diagnostic results and anti-interference ability.