机械振动信号的改进数学形态滤波分析

ANALYSIS OF MECHANICAL VIBRATION SIGNAL USING IMPROVED MATHEMATICAL MORPHOLOGICAL FILTER

为克服传统形态滤波器结构元素不能自适应调整长度的缺点,采用量子理论对数学形态学结构元素长度进行优化,实现了结构元素长度随振动信号变化自动修正。采用量子比特对振动信号进行分析,并将其拓展为振动信号的多量子比特系统,进而在量子系统内对振动信号的波形进行测量,最终结合量化指标对结构元素进行动态优化。将经过长度修正的结构元素用于故障轴承的振动信号分析,准确提取出异常脉冲信号。

For overcoming the defect that the length of conventional structuring element in mathematical morphology can’t change adaptively,quantum theory is utilized to optimize its length,which finally realizes the goal that the length of structuring element can adjust with mechanical vibration signal automatically. The quantum bit is employed to analyze vibration signal,and furthermore it is extended to multiple quantum bits system for vibration signal analysis. The waveform of vibration signal is measured in the framework of quantum system,thus the structuring element is modified by measured index. Applying the modified structuring element to vibration signal sampled from fault bearing,the abnormal impulse signal is extracted correctly.