基于感知模糊自适应蚁群算法的非线性PID控制

随着系统复杂度的提高和对象不确定性因素的增加,为克服线性PID动态性能和稳态性能差的缺陷,分析了非线性PID控制器各控制参数对误差的理想变化过程,构造非线性PID控制器。由于增益参数大量增加,传统参数优化方法不再适用,在分析蚁群算法的基础上,提出了基于感知自适应蚁群算法,并加入模糊自适应信息素更新机制,用于优化非线性PID控制器的设计方法。通过仿真实验将该控制器与基于蚁群算法的非线性PID控制器和基于蚁群算法、Z-N法的PID控制器进行对比,并对控制性能和收敛性能进行了分析,结果表明该算法有效克服了传统蚁群算法收敛速度较慢、容易陷入局部最优而停滞的缺陷,该控制器具有更好的动态性能和稳态性能。

Nonlinear PID Control of Based on Sensation Fuzzy Self-adaptive Ant Colony Optimization

With the system complexity and objects uncertainties increasing, for overcoming the defect of linear PID controller, a nonlinear PID controller was constructed by analyzing the ideal varying process of the individual tuning nonlinear PID controller concerning error. The traditional optimization was inapplicable due to the quantity of gain parameters increasing. On the basis of analyzing the control parameter optimization by ant colony optimization (ACO), the self-adaptive ant colony optimization based on sensation was used and the fuzzy self-adaptive update mechanism of pheromone was added. The controller was compared with nonlinear PID controller based on ACO, linear controller based on ACO and Z-N in simulation, their control performance and convergence performance were analyzed. The results show that the algorithm overcomes the defect of traditional ACO effectively which is slow in converging, possible to sink into local optimum and effected by initial value, and the controller has better dynamic performance and steady performance.