基于MPSO-SVM的不确定条形区域故障诊断研究

Research on Fault Diagnosis of Uncertain Strip Area Based on MPSO-SVM

针对一类线性不确定性系统,基于条形区域下,研究了执行器连续增益故障可靠控制的问题。对于极点数据难以获取的缺陷,给出状态观测器的核心算法,从而实现系统状态的实时采集。同时,为解决支持向量机在极点配置中选取参数困难的问题,利用改进的粒子群算法(MPSO)优化支持向量机的结构,进而获得更合理的核宽度系数和惩罚系数。与网格搜寻法(Grid search-SVM)和粒子群算法(PSO)相比,该方法可获得较高的分类精度,仿真实例进一步证明了设计可靠控制器的必要性。

To a class of linear uncertain systems,in terms of the strip region,the problem of actuator continuous gain fault and reliable control is studied.In this paper,in order to solve the problem that pole data is difficult to obtain,the core algorithm of state observer is proposed to realize the real-time acquisition of system state.At the same time,in order to solve the problem that it is difficult to select the parameters of support vector machine in pole assignment,an improved particle swarm optimization(MPSO)algorithm is proposed to optimize the structure of support vector machine(SVM),so as to obtain more reasonable kernel width coefficient and penalty coefficient.Compared with grid search SVM and particle swarm optimization(PSO),this method has higher classification accuracy,and the simulation example further proves the necessity of reliable controller design.