关键词: 列车自动驾驶系统（ATO）, 城市轨道交通, 广义预测控制器, 超调抑制, 输入约束, 平滑目标曲线

Abstract: In recent years, rail transportation is developing very rapidly, of which ATO device, directly affecting the operation of urban rail safety and efficiency, is responsible for the train traction and braking. ATO control algorithms during recent years are too complex or requiring iterative solutions and computation time uncertain. So they only have theoretical significance and are difficult to realize in engineering. Generalized predictive control algorithm is simple, efficient and easy to implement, but its initial design doesn’t consider the special needs of ATO control. By adding overshoot-reduction item in the optimization target to re-derive predictive input expression, and input saturation constraints, and using smooth quadratic curve in designing traction and braking target curve, passengers’ safety and comfort are ensured. Through simulation, the algorithm proposed in this paper is proven to be ease of use, safe and efficient.

Key words: Automatic Train Operation（ATO）, urban rail transit, Generalized Predictive Control（GPC）, overshoot reduction, input constraints, smooth target curve