动态矩阵控制中基于Kalman滤波的开环预测方法  

Open-loop prediction method in dynamic matrix control based on Kalman filter

作　　者：谢亚军[1] 丁宝苍[1] 陈桥[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049

出　　处：《控制与决策》2017年第3期419-426,共8页Control and Decision

基　　金：国家自然科学基金项目(61573269);国家863计划项目(2014AA041802);陕西省自然科学基金项目(2016JM6049)

摘　　要：用稳态Kalman滤波器来解释双层结构模型预测控制中的开环预测模块,得到的结果与非基于Kalman滤波导出的开环预测对应和等价.将该结果与文献中已有的Kalman滤波解释进行了详细对比,阐述了其中的不同点.针对双层结构预测控制的整体策略进行仿真,并针对积分输出的Kalman滤波进行仿真,验证了所得结论的有效性.The steady-state Kalman filter is utilized to interpret the open-loop prediction module in the double-layered model predictive control（MPC）.The results are consistent with, and equivalent to the open-loop prediction which is not deduced based on the Kalman filter.The results are compared with other Kalman filter interpretations in the literature,and the differences are stated.By simulating the results on the overall strategy of the double-layered MPC, and simulating the Kalman filter for the integral outputs, the effectiveness of the results is validated.

关 键 词：双层结构模型预测控制 开环预测 KALMAN滤波 积分输出 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]