Planetary hybrid electric buses usually have a great jerk while the drive mode switched. PID-based controller could not effectively ensure vehicle ride comfort during the mode switching process. Model predictive control(MPC) ,which can obtain optimal control sequences by online rolling optimization, is used to solve this issue, and a dynamic coordination control method based on MPC is proposed to start the engine. Based on the vehicle dynamic equations and vehicle historical data, the data-driven engine model and the closed-loop simulation model are set up in the Matlab/Simulink. Taking the start-up process of the engine as a constrained multi-objective optimization problem, the model predictive controller is designed according to the system state space equation and optimization problem. In the process of switching from the pure electric mode to the hybrid mode, the model predictive controller is compared with the traditional PID-based control method and without control. The simulation results demonstrate that under the premise of ensuring the dynamic performance of the vehicle, compared with the PID control and without control, during the mode switching process, the dynamic coordinated control method based on MPC can not only achieve the normal start of the engine, but also greatly reduce the peak jerk and also make the vehicle follow the target speed well.