CAJ | 학술논문

基于故障轴承的特征提取,提出一种基于小渡包与径向基RBF神经网络相结合的故障诊断方法,克服了以往常用诊断方法中的小波BP神经网络网络收敛慢、训练时间长、而且常常陷入局部极小点的缺点.采用小波滤波技术对采集到的滚动轴承振动信号进行滤波处理,利用小波包分解获得滚动轴承振动信号的特征向量作为故障样本对RBF网络进行训练,进行了详细的故障诊断试验研究.实验结果表明训练好的RBF网络能够很好地诊断出轴承故障类型,故本方法在旋转机械故障诊断方面具有良好的应用价值.
기우고장축승적특정제취,제출일충기우소도포여경향기RBF신경망락상결합적고장진단방법,극복료이왕상용진단방법중적소파BP신경망락망락수렴만、훈련시간장、이차상상함입국부겁소점적결점.채용소파려파기술대채집도적곤동축승진동신호진행려파처리,이용소파포분해획득곤동축승진동신호적특정향량작위고장양본대RBF망락진행훈련,진행료상세적고장진단시험연구.실험결과표명훈련호적RBF망락능구흔호지진단출축승고장류형,고본방법재선전궤계고장진단방면구유량호적응용개치.
A fault dingnesis method combined wavelet packet transform with RBF neural network is presented bearings faults based On feature extracting of fault bearing. It overcomes BP neural network' s slow convergence, long hours of training, and falling into the local minimum point. Using the new method, the signal is processed first through the wavelet denoising, Then, three-layer wavelet packet is adopted to decompose the denoised signal of rolling beatings, and the wavelet packet energy eigenvector is obtained which can be used, as fault samples to train RBF neural network. The experimental results show that the trained RBF neural network can diagnose the kind of roiling bearing faults, the method has fair prospects of application for the rotary machine fault diagnosis.