A DMPC Based Lane Change Combination Decision and Control Method for Autonomous Vehicles

In order to improve the application of model predictive control (MPC) method in the trajectory planning of high-speed autonomous vehicles and the effective stability control in tracking. Firstly, a dynamic scene switching model predictive planning control (DMPC) is proposed that combines local motion planning and longitudinal control. A longitudinal speed control model, considering surrounding vehicles, is established for longitudinal control and local lateral trajectory planning. Secondly, a vehicle lateral dynamics model, accounting for tire slip, is established to track the local lateral reference trajectory planned by the longitudinal controller. Thirdly, the Big M method is employed to combine decision commands with system state constraints. Under the command of decision module, the algorithm can automatically activate or deactivate the relevant constraints, ensuring the accurate execution of local motion commands without the need for a separate trajectory planning module. Through simulation verification in different driving scenarios, it can be seen that the proposed method can automatically execute the corresponding maneuver commands under complex driving conditions. Compared with traditional lane-centering controllers, the obtained results in smoother lane changing trajectories with smaller heading angle changes, while not introducing additional computational burden. This significantly enhances both driving comfort and safety.