针对旋转机械故障诊断中的不确定性问题,提出基于多传感器D—S(Dempster-sharer)证据理论和模糊数学相结合的信息融合算法;通过多传感器测出旋转机械振动位移和振动加速度,得出D—S证据理论中多传感器分别对旋转机械的信度函数分配值,使用改进的D—S证据算法得到融合后的信度函数分配值,由D—S合成规则确定故障类型,通过在多功能旋转机械平台上的试验得出改进后的证据理论明显提高了旋转机械故障诊断的精度。
In order to solve uncertain problem of rotation machine fault diagnosis, an information fusion algorithm based on multi-sensor combination of D-S evidential theory and fuzzy mathematics theory for rotation machine fault diagnosis is proposed. We can get displacement and acceleration of rotation machine through multi-sensor. Then the basic probability number will be obtained by using the algorithm. At last the type of fault will be ascertained over D-S objective judgment rule. The specific trial conveys that the new method was more accurate.