Abstract:This study proposes autonomous eco-driving strategies for a traffic environment with limited information available based on three popular Reinforcement Learning (RL) algorithms for continuous actions, i.e., Deep Deterministic Policy Gradient (DDPG), Proximal Policy Optimization (PPO) and Soft Actor–Critic (SAC) approaches, to address a serious challenge in the literature. The challenge is, despite the potential of connected and automated vehicles (CAVs) to diminish traffic interruptions caused by traffic signals on urban streets, for a long period of time, the accurate and immediate exchange of information via vehicle-to-everything (V2X) communication remains a problem because of the low penetration rate of CAVs and communication quality issues. The available information is assumed to only include signal phase and timing (SPaT) information and traffic states of the leading vehicle, respectively. To optimize the overall driving performance, a trade-off among safety, efficiency, energy, and ride comfort is considered in the reward function. Moreover, a hybrid policy is proposed to take advantage of RL and an analytical car-following model. The strategies enable CAVs to safely, efficiently and comfortably traverse signalized intersections while all the other vehicles are unconnected human-driven vehicles (HVs). Trajectories from the pNEUMA dataset are used to train and test the proposed models. The performance of the proposed models is compared to the naturalistic driving data, the Intelligent Driver Model (IDM) and an eco-driving method based on rules and optimization (Trigo). Testing results show that DDPG and SAC with the hybrid policy (HybridSAC) have the best overall performance, i.e., better than human drivers in all aspects; better safety, energy and comfort than the Trigo model; similar performance to the IDM but better energy efficiency. The temporal and spatial generalization capabilities of the RL-based methods are also tested, among which HybridSAC has the best performance as a whole.

Hybrid Car-Following Strategy based on Deep Deterministic Policy Gradient and Cooperative Adaptive Cruise Control

Hybrid Car-Following Strategy Based on Deep Deterministic Policy Gradient and Cooperative Adaptive Cruise Control

Adaptive Kalman-based hybrid car following strategy using TD3 and CACC

Path Following for Autonomous Ground Vehicle Using DDPG Algorithm: A Reinforcement Learning Approach

Multi-Objective Optimization of Vehicle-Following Control for Connected Electric Vehicles Based on Deep Deterministic Policy Gradient

Altruistic cooperative adaptive cruise control of mixed traffic platoon based on deep reinforcement learning

Model Predictive Adaptive Cruise Control of Intelligent Electric Vehicles Based on Deep Reinforcement Learning Algorithm FWOR Driver Characteristics

A Decision-Making Strategy for Car Following Based on Naturalist Driving Data via Deep Reinforcement Learning

Coordinated Decision Control of Lane-Change and Car-Following for Intelligent Vehicle Based on Time Series Prediction and Deep Reinforcement Learning

Car-following stability improvement of cooperative adaptive cruise control based on distributed model predictive control

Improved Deep Reinforcement Learning for Car-Following Decision-Making

A Deep Reinforcement Learning Based Hierarchical Eco-Driving Strategy for Connected and Automated HEVs

Deep Reinforcement Learning Car-Following Control Based on Multivehicle Motion Prediction

A New Hybrid Model Predictive Controller Design for Adaptive Cruise of Autonomous Electric Vehicles

Human-Like Autonomous Car-Following Model with Deep Reinforcement Learning

Hybrid car following control for CAVs: Integrating linear feedback and deep reinforcement learning to stabilize mixed traffic

A Novel Vehicle Platoon Following Controller Based on Deep Deterministic Policy Gradient Algorithms

Deep Reinforcement Learning Car-Following Model Considering Longitudinal and Lateral Control

Eco-driving strategies using reinforcement learning for mixed traffic in the vicinity of signalized intersections

A Real-Time Optimal Car-Following Power Management Strategy for Hybrid Electric Vehicles with ACC Systems

Personalized Adaptive Cruise Control with Deep Reinforcement Learning