中文摘要：

针对高速列车运行过程的模型不确定性问题,提出了高速列车的自适应子空间预测控制方法.首先,由列车观测数据集得到高速列车初始子空间预报模型,并利用快速LQ分解方法实现列车观测数据集的在线滑动窗口更新,从而获得列车自适应预报模型;然后,提出了融合当前时刻与过去时刻的加权误差信息的列车模型自适应切换策略,进而设计了高速列车自适应子空间预测控制器;最后,进行了高速列车运行的数值仿真实验,结果表明提出的方法具有较好的列车跟踪性能.

英文摘要：

To explore the uncertainty of the model for a high-speed running train, an adaptive predictive control method for high speed trains was proposed. Firstly, the subspace prediction initial model of high-speed train was obtained from the observation data, and the adaptive prediction model of high-speed trains was presented by sliding window formulations of the fast LQ decomposition for online update. Secondly, an adaptive model for switching strategies was proposed after considering the predictive errors of the high-speed trains, and the adaptive subspace predictive controller for high-speed trains was designed. Finally, the control simulation of the high-speed trains was implemented. The results show the tracking performance is enhanced by the proposed method.