中文摘要：

本文研究多优化目标和多约束条件下的动车组列车自动驾驶ATO控制问题。通过分段线性化列车运行阻力,引入列车运行状态整数变量,建立了动车组混合整数列车运行模型。提出基于混合系统模型预测控制HMPC的列车自动驾驶策略,并应用输入分块化技术和显式模型预测控制,降低算法的计算量,以提高硬件实现的可行性。前者通过固定一定时间段内的控制量,对输入序列进行分块化,降低控制器的自由度;后者通过离线设计和在线综合的方法减少算法的在线计算时间。最后利用Matlab仿真软件环境下的MPT 3.0扩展工具箱对所提控制策略进行数值仿真。结果表明：该控制器能在列车运行安全约束下,合理分配动车组各车厢的牵引力和制动力,保障列车准时及节能高效地运行,同时所提出的改进算法能有效降低计算量。

英文摘要：

This paper addressed an automatic train operation （ATO） problem of the electric multiple unit （EMU） train under consideration of the multiple optimal objectives and constraints. First, after piecewise lin- earization of the train running resistance, a hybrid integer EMU train running model that includes a running status integer variable was derived. Then an ATO algorithm based on the hybrid model predictive control （HMPC） was proposed, which utilized move-blocking strategy and explicit model predictive control to simplify the on-line computation for the sake of controller implementation in practice. Specifically, the former reduced the degrees of freedom of the controller by fixing the system input to keep it constant over several time-steps, and the latter decreased the online computation time by combining the off-line design with the on-line synthesis method. Finally, the numerical simulation experiments were conducted to verify the effectiveness of the pro- posed algorithm by using Matlab-based MPT 3. 0 toolbox. Results showed that the controller reasonably dis- tributed the traction force and braking force of each car of the EMU train under the train running safety con- straints and ensured the punctual and energy efficient operation of the train. At the same time, the proposed al- gorithm effectively reduced the computation of the algorithm.