城轨列车自动驾驶广义预测控制器的算法研究  

Study of ATO control algorithm in urban rail transit based on generalized predictive control

作　　者：赵文天[1] 郜春海[2] 

机构地区：[1]北京交通大学轨道交通控制与安全国家重点实验室,北京100044 [2]北京交通大学轨道交通运行控制系统国家工程研究中心,北京100044

出　　处：《计算机工程与应用》2015年第21期259-264,共6页Computer Engineering and Applications

基　　金：城市轨道交通北京实验室资助;城市轨道交通自动化与控制北京市重点实验室资助;城市轨道交通CBTC系统北京市高等学校工程研究中心资助;国家高技术研究发展计划(863)(No.2011AA110502)

摘　　要：近年来,轨道交通发展非常迅速,其中ATO设备负责列车的牵引和制动,这直接影响城轨运行的安全和效率。已有的城轨ATO控制方法由于控制算法过于复杂,或者需要迭代求解且计算时间不确定等,因而仅具有理论意义,难以在工程上实现。广义预测控制算法就是一种简单高效、易于工程实现的控制算法,但是其设计之初并没有考虑ATO控制的特殊需求。通过改进广义预测控制算法,在优化目标中增加了超调抑制项,重新推导预测输入表达式,并增加输入饱和约束,设计了采用二次曲线的平滑的牵引和制动目标曲线,从而保障乘客乘车的安全性和舒适性。通过仿真,证明算法的易用性、安全性和有效性。In recent years, rail transportation is developing very rapidly, of which ATO device, directly affecting the opera-tion 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.

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 

分 类 号：U231.6[交通运输工程—道路与铁道工程]