基于快速自适应超螺旋算法的高速列车最优黏着控制

Optimal adhesion control of high-speed train based on fast adaptive super-twisting algorithm

为解决轨面状态变化时列车牵引/制动力得不到有效发挥的问题,考虑轮轨间黏着状态具有较强的非线性和时变性,设计一种快速自适应超螺旋(FAST)滑模控制器。根据轮轨黏着机制和列车运行原理建立列车模型,对不同轨面下黏着特性曲线进行研究。使用全维状态观测器对负载转矩进行观测,并计算获取当前轨面黏着系数。在此基础上,利用滑模极值算法动态搜索当前轨面最佳蠕滑速度。采用快速自适应超螺旋算法设计的滑模控制器控制牵引电机转矩,使列车蠕滑速度稳定跟踪于当前轨面最佳蠕滑速度处。利用MATLAB软件对FAST滑模控制器设计的列车运行控制系统进行仿真,将其与采用超螺旋滑模控制器和标准滑模控制器的仿真结果进行对比。研究结果表明,FAST算法将系统稳定时间缩短至5 s,收敛速度加快。不同轨面搜索到的蠕滑速度稳定在最优阈值1.4~1.5 m/s之间,轨面变化时,转矩变化趋势与轨面变化趋势一致,数值分别稳定在67200和32720左右。在模型中加入干扰后,系统能快速稳定,转矩相对误差处于±0.05%,控制器有较强鲁棒性。由此可知,快速自适应超螺旋算法实现了对不同轨面下的列车最佳蠕滑速度的快速跟踪,充分利用了当前轨面的最大黏着系数,极大发挥了列车牵引/制动力,实现了高速列车的最优黏着控制。

In order to solve the problem that the train traction and braking force are not effectively exerted when the rail surface state changes,a fast adaptive super-twisting(FAST)sliding mode controller was designed considering the strong nonlinearity and time-varying nature of the wheel-rail adhesion state.The train model was established according to the wheel-rail adhesion mechanism and the train operation principle,and the load torque was observed using a full-dimensional state observer and the current rail adhesion coefficient was calculated.On this basis,the sliding mode extreme value algorithm was used to dynamically search for the optimal creep slip speed of the current rail surface.The traction motor torque was controlled by the sliding mode controller designed with the fast adaptive super-twisting algorithm,which enabled the train creep speed to be tracked steadily at the optimal creeping speed of the current rail surface.The MATLAB software was used to simulate the train operation control system designed by the FAST sliding mode controller,and the simulation results were compared with those of the super-twisting sliding mode controller and the standard sliding mode controller.The results show that the FAST algorithm reduces the system stabilization time to 5 seconds and the convergence speed is faster.The creep-slip speed searched at different rail surfaces is stable between the optimal threshold of 1.4~1.5 m/s.The trend of torque change is consistent with the trend of rail surface change when the rail surface changes,and the values are stable around 67200 and 32720 respectively.After adding disturbance to the model,the system can be stabilized quickly and the relative error of torque is less than±0.05%,and the controller has strong robustness.It can be seen that the fast adaptive super-twisting algorithm achieves the fast tracking of the best creep speed of the train under different rail surfaces,and makes full use of the maximum adhesion coefficient of the current rail surface.It greatly exerts the train traction and braking force,and realizes the optimal adhesion control of the high-speed train.