Atomic Proximal Policy Optimization for Electric Robo-Taxi Dispatch and Charger Allocation
The Atomic Proximal Policy Optimization (Atomic-PPO) algorithm is a scalable deep reinforcement learning method designed to optimize electric robo-taxi dispatch by decomposing fleet decisions into sequential atomic actions for each vehicle. It outperforms benchmark policies and reveals that ridership-based charger placement and fast charging speed are more critical than simply increasing vehicle range for system performance.