Gearshift Controller Optimization for Electric Vehicles with Reinforcement Learning

This article proposes an optimization method for electric vehicle dual-clutch transmission gearshift controllers. A feedforward-feedback control model is established, and the feedback controller is an RBFNN controller optimized by the PILCO algorithm. The control effect of the proposed method is compared with the PID feedback controller. Applying the learned controller to various slope and load circumstances demonstrates its optimization effectiveness with good robustness. The PILCO algorithm requires only a few experiments for optimization, making it a useful tool for engineers and technicians to increase the effectiveness of their Research and Development for various gearshift controllers.