论文部分内容阅读
针对一类参数大范围变化的不确定系统,提出一种基于分类转换策略的神经滑模控制方法.按小偏差原理对系统模型进行划分,利用结合主成分分析的最小二乘支持向量机进行分类训练,并分别设计基于径向基函数神经网络在线调整切换项增益的滑模控制器,在线时利用分类器按系统数据自动选择相应的控制器.同时,引入结合混沌机制的量子粒子群算法,并将其用于控制器近似最佳切换函数的构造.仿真结果表明,系统具有良好的跟踪性能和较强的鲁棒性,有效地降低了抖振.
Aiming at the uncertain system with a wide range of parameters, this paper proposes a neural network sliding mode control method based on categorical transformation strategy. According to the principle of small deviation, the system model is divided and classified by least square support vector machine combined with principal component analysis Training, and separately design sliding mode controller based on radial basis function neural network to adjust the gain of switching item online, and automatically select the corresponding controller according to the system data by using classifier while online.Meanwhile, the quantum particle swarm optimization algorithm with chaos mechanism is introduced, Which is used to construct the approximate optimal switching function of the controller.The simulation results show that the system has good tracking performance and strong robustness and effectively reduces the chattering.