基于KPCA与粒子群优化SVM的扫路车驱动电机故障诊断

Fault diagnosis of driving motor of road sweeper based on KPCA and particle swarm optimization SVM

针对环卫车辆驱动电机故障特征不明显且存在大量干扰因素等问题,为了提高故障诊断准确率,提出一种基于核主元分析和粒子群优化支持向量机(SVM)的故障诊断方法。该方法利用在电机故障状态下的振动信号构建时频域混合特征集,通过核主元分析,对特征集内的特征量进行降维处理;通过选择主元特征和利用粒子群算法,优化SVM的主要参数,将得到的特征量输入到优化后的SVM中进行计算,并与未进行核主元分析的SVM进行对比分析。计算结果表明:该方法能够显著提高扫路车驱动电机的故障诊断准确率。

In order to improve the accuracy of fault diagnosis,a fault diagnosis method based on kernel principal component analysis and particle swarm optimization support vector machine(SVM)is proposed for the problems that the fault characteristics of the drive motor of sanitation vehicles are not obvious and there are a large number of interference factors.In this method,the vibration signal in the motor fault state is used to construct the time-frequency domain hybrid feature set,and the dimension of the feature quantity in the feature set is reduced by kernel principal component analysis.By selecting the principal component features,the particle swarm optimization algorithm is used to optimize the main parameters of support vector machine.The obtained feature quantities are input into the optimized support vector machine for calculation,and compared with the support vector machine without kernel principal component analysis.The calculation results show that this method can significantly improve the fault diagnosis accuracy of the driving motor of the sweeper.