A study on multi-agent reinforcement learning for autonomous distribution vehicles

A self-driving automobile, also known as an autonomous vehicle, can observe its surroundings and maneuver without the assistance of a driver thanks to software algorithms completely automated driving system. As result, car may react environment in way that is comparable human driver. Road regulations limitations are required guarantee security effectiveness delivery services, well stop accidents damage brought on by technology failures. This paper formulates Autonomous Delivery Vehicles optimization problem proposes Multi-Agent Reinforcement Learning approach makes use shortest-path data have been obtained analytically. method allows for training multiple agents work together improve delivery. Using this strategy, vehicles' efficiency be markedly enhanced.