动车组自动驾驶制动过程双自适应优化控制

Optimal Control with Dual Adaptation for the Braking Process of EMU Automatic Operation

在动车组安全可靠运行的研究中,依赖于列车自动驾驶(ATO)系统。针对现采用单一自适应控制方法的ATO控制器控制列车制动时,在制动开始和结束阶段会出现明显的控制偏差的问题,采用双自适应广义预测控制方法设计ATO控制器。双自适应广义预测控制方法具有实时自适应建模和自动调优功能,其为达到系统模型自适应的目的,采用可变遗忘因子递推最小二乘法(VFF-RLS)实时辨识动车组制动过程的模型参数,然后根据实时辨识的模型参数自适应建模和计算控制器调优参数,并设计了监督机制确保控制稳定。仿真结果表明,在列车制动的起始和结束阶段,采用双自适应控制方法的ATO控制器的控制误差明显小于采用单一自适应控制方法的ATO控制器,并可确保动车组制动过程安全可靠运行。

The safe and reliable operation of electric multiple unit （EMU） depends on the progressiveness of the automatic train operation （ATO） system. The control deviation of the single adaptive control method of ATO control- ler to control the train brake is obvious at the beginning and end stage of the brake. A dual adaptive generalized pre- dictive control method is used to design the ATO controller. The dual adaptive generalized predictive control method contains both model adaptation and auto - tuning capabilities. Firstly, the variable forgetting factor recursive least squares （VFF - RLS） algorithm is selected to identify instantaneously the model parameters in the braking process for the purpose of system model adaptation. Secondly, according to the model parameters, adaptive modeling is a- chieved and the tuning parameter is calculated, and the supervision mechanism is designed to ensure control stabili- zation. Finally, Simulation results show that the control error of the ATO controller with dual adaptive control method is significantly less than the single adaptive controller at the start and end stage of the train braking, and it can ensure the safe and reliable operation of the EMU braking process.