一种基于LS-SVM的道岔控制电路故障诊断

Fault Diagnosis for Railway Switch Control Circuit Based on ARPSO Least Squares Support Vector Machine

针对最小二乘支持向量机最优参数难以寻找的问题,提出了用ARPSO算法优化最小二乘支持向量机可调参数的方法,并将该方法应用于道岔控制电路的故障诊断中.ARPSO算法在保证种群多样性的同时,避免了基本PSO算法过早收敛的问题,能更高的提高算法效率.仿真证明ARPSO算法比基本PSO算法具有更高的收敛速度和效率,基于ARPSO最小二乘支持向量机的分类方法比最小二乘支持向量机分类方法具有更高的分类准确度.

For the selection difficulty of optimization parameters of least squares support vector machine,a method of Attractive and Repulsive Particle Swarm Optimizer algorithm is proposed.And this method is applied to the fault diagnosis of railway switch control circuit.ARPSO algorithm ensures the population diversity and avoids parameters convergence problem of the basic PSO algorithm at the same time,and then it improves the algorithm efficiency.The simulation result shows not only the convergence speed and efficiency of ARPSO algorithm is higher than basic PSO but also the classification accuracy of LS-SVM based on ARPSO algorithm is higher than LS-SVM.