Improving Speed Trajectory Optimization for Electric Vehicles with Pre-Calculated Energy Maps and Dynamic Profile Discretization

Ongoing research and development in the field of electric vehicles (EVs) have resulted in a continuous expansion of their range. Additionally, advancements in vehicle connectivity have created new opportunities for intelligent driving assistance and energy optimization, particularly through the use of cloud data. However, the integration of eco-driving assistance with numerical optimization of speed trajectories remains challenging due to the high computational demands of these methods. To address this challenge and make such a system feasible for integration into vehicle systems, the computational effort required for an optimized driving trajectory must be minimized. This paper presents a method to accelerate speed trajectory optimization using pre-calculated energy and time consumption maps. For this purpose, a dynamic discretization of the anticipated driving profile is applied. Initial results show a substantial reduction in computation time, varying with different scenarios. Furthermore, trajectory optimizations have led to energy consumption reductions between 5% and 30% in simulation trials. The structure of this paper is as follows: First, the vehicle model, implemented in Matlab/Simulink, and the longitudinal dynamics, with a focus on key components, are explained. Next, the numerical optimization implementation using Dynamic Programming, the Ford-Bellman Algorithm, and the necessary discretization of the driving profile to be optimized are described. This is followed by an explanation of the hybrid energy consumption calculation. Then, the functional implementation of the algorithm in Matlab pseudocode is presented. Various optimized driving trajectories for a predefined route are then generated with the optimizer, and the simulation results are analyzed, including the computational effort required for optimization. Finally, the results are summarized, and an outlook on the optimizer’s potential future applications is provided.