有约束随动系统控制策略研究

提出了一种新的有约束的复杂随动系统--小碗摆球系统,并针对该系统利用达朗伯静力学的方法进行了建模。在讨论了系统能控和能观性的基础上分别采用状态反馈的极点配置法和基于遗传算法的LQR最优控制2种方法进行了实际系统控制效果的实验对比,通过对比可知极点配置的状态反馈控制器具有更好的鲁棒性和瞬态特性,而遗传算法优化的LQR控制具有更好的稳态特性,以及更短的调节时间。同时在参数选择方面相比于极点配置试特征值的方法,遗传算法优化LQR控制控制器更有针对性,便于实际的应用操作。

Research on a New Constrained Servo System Control Strategy

A new constrained complex follow-up system-cup and ball pendulum system was proposed. The new system was modeled by the method of lagrange dynamics. controllability and observability of the system were studied firstly. Two control methods of state feedback controller with pole assignment and LQR optimal control based on genetic algorithm were used to achieve the experiment comparison in this paper. Through the comparative analysis, state feedback controller with pole assignment has better robustness and transient characteristics. And LQR optimal control based on genetic algorithm has better steady-state characteristics and shorter tuning time. Meanwhile, compared with state feedback controller with pole assignment LQR optimal control based on genetic algorithm is more targeted in terms of parameter selection, therefore it is convenient for practical application.