摘要
为了辨识双重不确定性系统未知参数,跟踪系统目标输出值,本文对未知参数随机系统分别进行了对偶控制和最小二乘(RLS)辨识控制.提出了双重不确定随机系统对偶控制和最小二乘辨识控制策略.通过双重不确定随机系统仿真,比较分析了对偶控制和最小二乘辨识控制策略的性能.结果表明:该对偶控制策略用于双重不确定性随机系统参数辨识时实时性和稳定性优于最小二乘辨识控制;随机系统在运行过程中未知参数变化时,对偶控制策略仍可有效跟踪系统目标输出值,且对偶控制跟踪误差和切换之前保持一致,而RLS辨识控制不能有效跟踪系统目标输出值,跟踪失效.
In order to identify unknown parameters and to track system target output value in dual uncertainty system, dual control and recursive least squares （RLS） identification control are presented for the stochastic system. The control strategy of double dual control and the recursive least squares identification controland is put forward. Through dual uncertain stochastic system simulation, the dual control and least squares identification performance of the control strategy are comparatively analyzed. The results show： The dual control strategy is used for dual uncertainty parameter identification of stochastic system,its realtime performance and stability are better than the least squares identification control; When the unknown parameters of stochastic system change in the running process, dual control strategy can still be effective track system for the target output value, and the tracking error is consistent with the value before switching;RLS identification control cannot effectively track system output value, and fails to track.
出处
《西安工业大学学报》
CAS
2016年第10期843-847,共5页
Journal of Xi’an Technological University
基金
国家自然科学基金重点项目(61533014)
国家自然科学基金(61273127
61304204)
陕西省国际科技合作重点项目(2015KW-024)
陕西省教育厅专项科学研究计划项目(16JK1364)
关键词
双重不确定系统
对偶控制
辨识控制
实时性
dual uncertainties system
dual control
identification control
real-time
作者简介
杨恒占（1976-），男，西安工业大学讲师，主要研究方向为随机系统、最优控制．E-mail：yanghengzhan@xatu．edu．cn
