An Energy-Efficient Driving Method for Connected and Automated Vehicles Based on Reinforcement Learning

The development of connected and automated vehicles (CAV) technology not only helps to reduce traffic accidents improve efficiency, but also has significant potential for energy saving emission reduction. Using the dynamic flow information around vehicle optimize trajectory is conducive improving efficiency vehicle. Therefore, an energy-efficient driving method CAVs based on reinforcement learning proposed in this paper. Firstly, a set prediction models long short-term memory (LSTM) neural networks are developed, which integrate intention lane change time accuracy surrounding trajectories. Secondly, model built Proximity Policy Optimization (PPO) learning. takes current states predicted trajectories as input information, outputs energy-saving control variables while taking into account various constraints, such safety, comfort, travel efficiency. Finally, tested by simulation NGSIM dataset, results show that can save consumption 9–22%.