Multi-Agent Reinforcement Learning-Based UAV Pathfinding for Obstacle Avoidance in Stochastic Environment
This repository contains the code for the implementation of CTPDE, CTFDE, and CTFDE-MPC for UAV obstacle avoidance in the stochastic environment. Our experiments are conducted in simulations and real-robot systems. The experiment video is available at https://www.bilibili.com/video/BV1gw41197hV/?vd\_source=9de61aecdd9fb684e546d032ef7fe7bf
How to use the code: Training: To start training run the main.py file. Evaluation: The trained model for testing can be found inside the directory. Run the test.py file with the path to the trained model. You can also run the test.m file to visulize the obstacle avoidance process.
Dependencies: The code is written in Python. We recommend using Python 3.7. The required packages can be found in the file environment.yml.