SIGNAL+DRAHT | Ausgabe 09/2024
Deep Reinforcement Learning for optimisation of train operation
This project focuses on optimising train operations using Deep Reinforcement Learning (DRL). The goal is to create an intelligent train control system that ensures safety by adhering to speed limits, enhances passenger comfort, maintains punctuality, achieves precise parking and improves energy efficiency. Two DRL algorithms, Double Deep Q-Network (DDQN) and Proximal Policy Optimisation (PPO), have been examined. PPO has outperformed DDQN, effectively balancing the objectives of safety, comfort, punctuality, parking precision and energy efficiency....