基于电流时频特征的不对中故障诊断研究

摘要

电动机电流信号分析广泛应用于电机本身的监测和故障诊断,但该技术与转动系统的研究却比较少.针对转子系统的轴系不对中故障,提出了基于经验模态分解(EMD)和遗传算法支持向量机(GA-SVM)的不对中故障诊断方法.首先通过EMD方法将电流信号分解成若干个本征模函数(IMF);然后计算各IMF分量的能量特征和峭度值;最后从包含有故障信息的IMF分量的能量特征和峭度值作为输入建立支持向量机(SVM)判断轴系故障类型.实验表明,该方法可以有效地实现对于转子系统不对中故障类型和故障程度的诊断,且相对于只依靠能量特征的诊断方法,该方法对于不对中故障的诊断正确率有了明显的提高.%Motor current signal analysis has been an effective way to monitor and diagnose electrical machines. However,little research work has been reported in using this technique for rotor systems. In order to diagnose the misalignment fault of rotor system,a method was proposed which included the empirical model decomposition(EMD) and genetic algorithm optimization support vector machine(GA-SVM). First,the EMD was used to decompose the current signal into several IMFs.Then,the energy characteristics and kurtosis of each IMF component were calculated. Finally,the energy features and kurtosis of the IMFs containing the fault information were input to the GA-SVM for fault classification and recognition. The experimental results show that this method can effectively diagnose the misalignment fault type and fault degree of rotor system.The method can improve the correct rate of the fault diagnosis, compared with the method that only depends on the EMD energy characteristic.