混合PSO在最小均方误差调节器中的应用  

Application of Hybrid PSO to Least Mean Square Error Regulator

作　　者：吴华丽[1] 吴进华[1] 柳爱利[1] 

机构地区：[1]海军航空工程学院控制工程系,烟台264001

出　　处：《计算机系统应用》2012年第4期212-215,共4页Computer Systems & Applications

基　　金：安徽省教育厅自然科学基金(2005KJ004ZD)

摘　　要：利用最小均方误差调节器来改善有限拍系统是一种常用的方法,而调节器参数的选择是非常重要的。本文在确定最小均方误差调节器的过程中设计基于模拟退火的粒子群算法对参数进行了优化,并对一具体系统进行了仿真。结果表明,混合粒子群算法比单独使用粒子群算法和模拟退火算法的效果要好,同相关文献确定参数方法相比,系统的超调量和调节时间都得到了明显的改善,验证了所提算法的有效性。The least mean square error regulator is adopted to improve performance of the dead beat control system as one method in common use, while selection of the regulator parameter is very important. This paper designs the particle swarm optimization algorithm based on simulated annealing to optimizing parameter in the course of defining the least mean square error regulator, which is used in simulating the concrete system. The results show that Hybrid PSO can get better optimization effect than SA or PSO single used, compared with the method of determining parameters in the relevant literature,while the overshoot and settling time of the system are improved obviously, which validates the effectiveness of the algorithra put forward.

关 键 词：最小均方误差 参数选择 粒子群算法 模拟退火 优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]