摘要
以受控自回归滑动平均模型和受控自回归积分滑动平均模型为研究对象,根据估计理论,利用参数递推方法,构造了广义预测控制的一种状态空间实现,从而避免了解Diophantine方程,大大地减少了预估算法的计算量.为控制系统的性能分析提供了便利条件 文末,对状态空间实现的可控与可观性加以了证明.
Base on controlled autoregressive moving average model and controlled autoregressive integrated moving average model,according to estimation theory, using the method of parameters recurrence relations, a state space realization of generalized predictive control is structured and its observability and controllability is proved avoiding solving the equation of Diophantine, this algorithm can decrease the computing amount. And it provides a useful way for analysing the performance of the control system.
出处
《湖南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2003年第5期44-46,共3页
Journal of Hunan University:Natural Sciences
基金
湖南省自然科学基金资助项目(60272051)
关键词
广义预测控制
受控自回归滑动平均模型
受控自回归积分滑动平均模型
状态空间实现
generalied predictive control
controlled autoregressive moving average model
controlled autoregressive integrated moving average model
state space realization
