基于电机电流信号的齿轮泵故障识别方法

Fault Identification Method of Gear Pump Based on Motor Current Signal

针对机械类信号在齿轮泵故障识别与诊断中存在的信号获取成本高、信噪比低、故障特征不易获取等问题,提出一种基于电机电流信号的液压齿轮泵故障识别方法。分析通过驱动电机电流信号对齿轮泵故障进行识别的可行性,对所采用的VMD方法的参数进行了优化,结合齿轮泵运行工况对IMF分量的相关性进行分析,并重构了电流信号,依据其排列熵和均方根值所构造的特征样本并融合KFCM聚类算法,对齿轮泵进行故障识别与诊断。并通过机电液试验台对不同故障类型的齿轮泵进行试验,试验结果表明:所提电机电流信号分析与特征提取方法可准确而有效识别齿轮泵故障。

In the fault identification and diagnosis of gear pump,mechanical signals have such problems as high signal acquisition cost,low signal-to-noise ratio and difficulty in obtaining fault characteristics.To solve these problems,a fault identification method for hydraulic gear pump was presented based on motor current signal.The feasibility of identifying gear pump fault by driving motor current signal was analyzed.The parameters of the VMD method were optimized.Then,the correlation of IMF components was analyzed and the current signal was reconstructed according to the operating conditions of the gear pump.Based on the characteristic samples constructed by the entropy and root mean square,KFCM clustering algorithm was fused to identify and diagnose gear pump faults.At last,the experiments for gear pumps with different fault types were carried out on the electromechanical hydraulic test bench.The experimental results show that the current signal analysis and feature extraction method can be used to identify gear pump faults accurately and effectively.