Deep Deterministic Policy Gradiant for TORCS
Overview
- Date: 06/10/21
- Team: Rafail Islam, Siming Liu
- Environment: Ubuntu, TORCS, Tensorflow, Python
- Goal: Deep reinforcement learning algorithms for single-agent and multi-agent environments.
- Keywords: Deep reinforcement learning, multiagent system, DDPG, TORCS
In recent years, reinforcement learning algorithms have been used in the field of multi-agent systems to help the agents with interactions and cooperation on a variety of tasks. Controlling multiple agents simultaneously is extremely challenging as the complexity increases drastically with the number of agents in the system. In this study, we propose a novel semi-centralized deep reinforcement learning algorithm for mixed cooperative and competitive multi-agent environments. Specifically, we design a robust DenseNet-style actor-critic structured deep neural network for controlling multiple agents based on the combination of local observation and abstracted global information to compete with opponent agents. We extract common knowledge through influence maps considering both enemy and friendly agents for unit positioning and decision-making in combat.
Resources
- Linux Environment - Ubuntu
- TORCS Reinforcement Learning Environment Installation Guide for gym_torcs in Ubuntu (.docx)
- TORCS instruction
- TensorFlow 2
- Reinforcement Learning
Open research questions
- How to learn to race in TORCS?
- How to learn from other champion drivers?
- How to drive with other cars (Multi-Agent Systems)?
- How to drive like a human driver?