基于蚁群算法的智能人工腿最优PID控制器设计

来源期刊：中南大学学报(自然科学版)2004年第1期

论文作者：谭冠政 李文斌

文章页码：91 - 96

关键词：蚁群系统;蚁群算法; PID控制器;智能仿生人工腿

Key words：ant system; ant algorithm; PID controller; intelligent bionic artificial leg

摘    要：以蚁群系统和蚁群算法为基础,提出了一种新的具有不完全微分的最优PID控制器的设计方法。该控制器以系统单位阶跃响应的超调量σ、上升时间t_r以及调整时间t_s为性能指标,针对给定的控制对象,利用所建立的蚁群算法搜索出一组最优PID参数K^*_p,T^*_i及T^*_d,作为实时控制中PID控制器的参数。该控制器被用于控制智能仿生人工腿中的执行电机。计算机仿真结果表明,与传统的PID控制器相比,这种基于蚁群算法的最优PID控制器具有良好的动态和稳态性能,可用于控制多种不同的对象和过程。

Abstract: A new designing method for the optimal PID controller with incomplete derivation is proposed based on ant system and ant algorithm. In this method, for a given system, by taking the overshootσ, rise timetr, and setting timetsof the system′s unit step response as the controller′s performance indexes and by means of the ant algorithm developed by the authors, a group of optimal PID parametersK*_p,T*_I and T*_d are obtained, which are used as the parameters of the PID controller in re-al-time control. This kind of controller has been used to control the motor of the intelligent bionic artificial leg designed by the authors. The results of computer simulation show that this controller is better than the conventional PID controllers in the dynamic behavior of system response. This kind of PID controller possesses dynamic controlling and robust performance, and can be used to control many different kinds of objects and processes.