基于粒子群算法的时变滑模解耦控制方法

Decoupled Sliding-mode Controller with Time-varying Sliding Surfaces Using PSO Algorithm

研究六阶非线性耦合系统滑模解耦算法收敛性和时变滑模算法比例系数选取问题,提出了一种基于粒子群算法的时变滑模解耦控制方法;算法首先根据六阶非线性系统的耦合特性,对六阶非线性系统进行了两次解耦;然后,用粒子群算法对时变滑模的比例系数进行了优化;最后利用二级倒立摆系统验证了文中所提算法并把所提算法与现存的滑模解耦方法作了对比,结果表明文中所提算法不但解决了现有滑模解耦算法不能使二级倒立摆的小车稳定在平衡位置的问题,而且收敛速度更快,稳定性更好。

This paper presents decoupled sliding--mode controller using time--varying sliding surfaces based on PSO. The twice decoupied method provides a way to achieve asymptotic stability compared with the existing decoupled control methods for six--order nonlinear system. Then we use PSO to select the scaling gain of the time--varying sliding surfaces..The results show that the proposed algorithm can not only make the car stable in the equilibrium position of the double inverted pendulum that the existing sliding mode decoupling algorithm failed to solve, and the convergence faster, better stability.