一种改进人工蜂群的分数阶PID控制器优化算法

An Improved Artificial Colony Optimized Algorithm for Fractional-Order PID Controller

摘要:

针对分数阶PID控制器参数整定过程参数多复杂性大,传统靠经验试凑的方法不易实现且优化效果差的问题,提出了一种改进的人工蜂群算法,实现分数阶PID控制器参数的整定;该算法通过改进人工蜂群算法中搜索方程,并引入一个淘汰机制,对分数阶控制器参数进行群智能搜索,将搜索到的参数送至分数阶PID控制器中反复迭代,以带有权值的误差绝对值积分指标(AIE)作为人工蜂群寻优的目标函数,最后得出控制器;本文以非线性系统为被控对象,经过实例仿真,验证了该算法实现的控制器比传统整数阶控制器和未改进的人工蜂群算法实现的分数阶控制器的动态性能和稳态性能都有所提高,在超调、上升时间、振荡性方面都优于未改进算法。

Abstract:

It is complicated to set parameters of fractional order PID controller, the optimization results from traditional method are not only difficult to achieve but poor, so this paper put forward an improved artificial colony algorithm to complete the fractional order PID controller’s parameter setting. First, improving the searching equation and adding an elimination mechanism in artificial colony algorithm to achieve parameters’ intelligent searching. Next those parameters will be sent to the fractional order PID controller and then repeat forementioned process. Meanwhile, using the integral of absolute error index (AIE) with weights as the objective function. Finally the controller can be concluded. This paper used a nonlinear system as a controlled object, by simulation and analysis, proved the controller’ performance of improved algorithm is more better in dynamic and steady-state than the controller of traditional integer order and the controller without improved artificial colony algorithm.In addition, this new algorithm we proposed is better than not improved algorithm in the aspect of maximum overshoot, rise time and oscillation.