蚁群算法在PID控制中的应用及其参数影响  

Application and parameter influence of ACO in PID control

作　　者：孙铁成[1] 张思敏[1] 李超波[1] 

机构地区：[1]中国科学院微电子研究所微电子器件与集成技术重点实验室,北京100029

出　　处：《现代电子技术》2015年第20期20-25,共6页Modern Electronics Technique

摘　　要：鉴于传统PID参数整定方法的不足,提出了一种采用蚁群算法优化选取PID控制参数的方法。通过建立数学模型将PID控制参数选择问题抽象成路径选择问题,从而将蚁群算法成功的应用于PID参数优选,并对寻优过程进行了仿真。将结果与常用的临界比例度法整定的结果进行了比较,发现基于蚁群算法的PID参数优选方案可使系统超调量大幅减小,并明显缩短系统调节时间,具有良好的应用前景。此外,讨论了蚁群算法中的关键参数对算法性能的影响,对比了不同参数下算法的收敛速度和求解质量。In view of the deficiencies of traditional PID parameters tuning method,a new method to optimize and select PID control parameters by means of ACO（ant colony algorithm）is proposed,in which the selection problem of PID control parameters is abstracted into the routing selection problem by building a mathematical model,thus ACO is applied successfully to PID parametric optimization and the optimizing process is simulated. It is found by comparing with tuned results of common critical proportioning method that the PID parameter optimization scheme based on ACO can reduce system overshoot significantly and shorten tuning time of system obviously,and has a great application prospect. In addition,the influence of key parameter in ACO on algorithm performance is discussed,and convergence velocity and solution quality of the algorithm with different parameters are compared.

关 键 词：PID控制 蚁群算法 信息素 参数优选 

分 类 号：TN911-34[电子电信—通信与信息系统] TP301.6[电子电信—信息与通信工程]