论文部分内容阅读
提出了一种基于特征波形稀疏匹配的滚动轴承故障模式识别方法.该方法通过自行设计的搜索算法从信号中提取多段特征波形,并对其进行学习优化,以优化后的特征波形作为基原子模型生成原子库及模式匹配库.将待识别信号在模式匹配库上进行一阶匹配分析,实现轴承故障的模式识别.对正常轴承、滚动体故障、内圈故障和外圈故障信号进行实验,验证了方法的有效性和鲁棒性.
A method of pattern recognition of rolling bearing based on sparse matching of feature waveform is proposed.The multi-segment feature waveform is extracted from the signal by self-designed search algorithm and optimized by learning, and the optimized feature waveform is generated as a basic atomic model Atomic Library and Pattern Matching Library First-order matching analysis of the signal to be recognized on the pattern matching library to realize the pattern recognition of bearing faults.Experiments on normal bearing, rolling element failure, inner ring fault and outer ring fault signal are carried out to verify The validity and robustness of the method.