改进的阈值小波和PSO-SVM算法及应用

Algorithm and Application of Improved Threshold Wavelet and PSO-SVM

本文提出了基于改进阈值小波及改进支持向量机的低速重载滚动轴承故障识别方法.首先在实验台上测量三种不同工况的轴承信号,利用改进阈值小波对信号进行降噪处理,将降噪的信号利用小波包分解法得出各频带的特征能量值;其次利用粒子群算法来优化支持向量机的学习因子,提高支持向量机多分类器的识别性能;最后,将特征能量值导入优化后的多分类器,实现低速重载滚动轴承的智能识别.结果表明,结合后的方法有着良好的诊断效果.

This paper presents a method based on improved threshold wavelet and improved support vector machine which used in low - speed and high load bearing fault identification. Firstly, in the experiment equipment, three different kinds of bearing sample signals are collected, denoise these signals by improved thresh- old wavelet, then wavelet packet is used to decompose the denoised vibration signals. Secondly, particle swarm optimization is used to optimize the learning factors of the support vector machine to improve multi - classifier recognition performance. Finally, the characteristic energy values are imported to improved multi - classifier, in order to achieve low -speed rolling bearing heavy intelligent recognition.