基于遗传算法的差值形态滤波尺度参数优化方法研究  

Research on Measure Parameter Optimization Method of a Difference Morphological Filtering Based on Genetic Algorithm

作　　者：吕博[1] 张永祥[1] 柯维[1] 

机构地区：[1]海军工程大学,武汉430033

出　　处：《中国机械工程》2013年第3期355-359,共5页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(50979109)

摘　　要：针对使用差值形态滤波提取故障特征时结构元素难以确定的情况,对尺度参数实行优化从而寻求其最佳尺度。首先对滚动轴承信号进行降噪处理,之后对信号进行差值形态滤波提取故障信号,通过遗传算法寻找到了结构元素尺度的最优解。通过仿真和试验研究表明,该方法取得的效果优于非优化的处理效果和包络解调法。Scale parameters were optimized to search the best scale, which was due to the con- struction element was hard to be determined when fault characteristics were extracted using a differ- ence morphological filtering. The noise in antifriction bearing signals was reduced, the fault signals were extracted through the difference morphological filtering, and then the best results of construc- tion element scales were gotten through genetic algorithm. The experimental and simulation results show the new method presented herein is better than the method without optimization and the enve- lope demodulation method.

关 键 词：形态滤波 遗传算法 结构元素 故障诊断 

分 类 号：TH122[机械工程—机械设计及理论]