基于遗传算法的线性多输入系统极点配置设计方法的研究

A Method Study of Pole Placement based on Genetic Algorithm for MI System

在确定的闭环极点下,多输入线性时不变系统的状态反馈矩阵的解不唯一。针对无法解决的优化问题,提出了一种新的基于遗传算法的设计方法,即将传统的直接计算法和遗传算法相结合,找到合适的优化参数,从而找到最优阵。该算法不需要复杂的计算,物力实现容易。对于具有普遍性的对象进行MATLAB仿真,试验结果表明新算法得到的反馈矩阵优于其它各种方法,从而具有一定的理论意义和实践意义。

The solution of the state feedback matrix K is not unique when poles of closed-loop are defined. Aiming at the problem that common design cannot get the optimized design of K, this paper puts forward a new method based on genetic algorithms, combining conventional direct algorithms with genetic algorithms, which can find the favorite optimized parameters and then find the optimized K. Both the new algorithms and the application are easier than any other algorithms. The MATLAB results tests based on popular problem： the feedback matrix K of the new design is better than other＇s, which is important in control theories and engineered practice.