遗传优化的LSSVM在故障诊断中的仿真研究

Simulation Research on Optimized Genetic LSSVM Used in Fault Diagnosis

研究机载电气故障诊断问题,采用支持向量方法。支持向量机是一种采用结构风险最小化原则的新型机器学习方法,具有出色的学习分类能力和推广能力。支持向量机的参数选择非常重要,决定故障诊断的精确度。针对最小二乘支持向量机的参数通常靠交叉试验来确定的情况,为了提高故障诊断的精度和效率,提出了一种模拟退火遗传算法和最小二乘支持向量机相结合的故障诊断方法,利用模拟退火遗传算法的全局搜索能力对最小二乘支持向量机的参数进行寻优,并以某型直升机机载电气盒的故障诊断为例对方法进行了仿真研究。实验结果表明,模拟退火遗传算法优化的最小二乘支持向量机取得了较好的故障诊断效果。

Support vector machine（SVM） has excellent learning,classification ability and generalization ability,which uses structural risk minimization.Parameters selection is very important and decides the fault diagnosis precision.Considering the fact that the parameters in least squares support vector machine（LSSVM） are usually decided by cross-validation,a new fault diagnosis method that combines simulated annealing-genetic algorithm with LSSVM is proposed in order to enhance accuracy and efficiency in fault diagnosis.This method searches the optimized parameters in LSSVM by taking advantage of the simulated annealing-genetic algorithm＇s powerful global searching ability.The research is provided using this method on the fault diagnosis of a certain type of helicopter＇s helicopter-electrical-box.The experimental results show that the proposed method achieves perfect accuracy and efficiency in fault diagnosis.