在柴油机故障诊断中,对实测数据进行局域波分解,得到信号在各个瞬时频率上的基本模式分量,通过求解各模式分量的关联维数,对故障进行诊断。研究结果表明:故障状态对应的模式分量关联维数发生较大变化,而其他模式分量变化很小,证实了该方法的有效性和工程实用性。
In the fault diagnosis of diesel engine, intrinsic mode function of data in each instantaneous frequency was gained by using local-wave method. Throught calculating correlation dimensions of each mode function, the faults were diagnosed. The results show that the correlation dimensions of mode function corresponding to fault state are in great changes while others are in little changes. The validity and practicality of this method are confirmed.