基于改进MPC的重载列车虚拟编组协同运行控制方法研究OA

To optimize tracking intervals in virtual coupling operations of heavy-haul railways,a model predictive control(MPC)-based cooperative control method was developed.A multi-mass-point dynamic model and a"soft wall"collision dynamic spacing control model were established for heavy-haul trains,with subsequent model discretization.A MPC optimization algorithm suitable for virtual coupling operation control of heavy-haul railways was then designed and combined with a Kalman filter for state estimation.This algorithm utilized a multi-objective function based on speed tracking error and train interval error to generate optimal traction/braking force sequences.A dual virtual-coupled train closed-loop simulation system was constructed using railway line parameters of a railway group company.Dynamic coupling characteristics were analyzed across dynamic coupling,cooperative operation,and decoupling scenarios.Results demonstrate that during dynamic coupling,the following train's speed tracking error converges to 0.4 km/h with train interval error maintaining near safety thresholds.In the cooperative operation phase,train interval errors do not exceed 20 meters,while during decoupling,dynamic safety constraints are satisfied.The MPC effectiveness in virtual coupling coordination is verified.This study provides references for achieving reduced tracking intervals in virtual coupling for heavy-haul trains.