To improve the highway congestion and low road transportation efficiency, heavy duty vehicle platooning driving in formation can improve the expressway in this paper. Firstly, a heavy duty vehicle longitudinal dynamic model and an inverse longitudinal dynamic model is established, an inverse engine model based on centroid moment of inertia balance is proposed for heavy duty vehicle platooning. Secondly, the distance between the front and rear vehicles, speed, acceleration and other relevant parameters in the heavy duty vehicle platooning are used as the input state of the controller, the vehicle acceleration required to follow the vehicle is determined by the model predictive control algorithm. At the same time, the hierarchical control system is established by the model predictive control algorithm and optimize it. Both simulation verification and intelligent vehicle experiment results show that the optimized platooning controller has a better performance than the general controller. The stability and effectiveness of the platooning are improved by this control method.