TY - DATA T1 - Code accompanying the paper on aircraft path planning in continuous environments with deep reinforcement learning PY - 2023/06/15 AU - Jan Groot UR - DO - 10.4121/b68d7aa9-235b-4f8b-b8c8-a977eaceba50.v1 KW - Soft Actor Critic KW - BlueSky KW - Deep Reinforcement Learning KW - Path Planning N2 - <p>This code is used for generating the results shown in the paper on aircraft path planning in continuous environments with deep reinforcement learning.</p><p></p><p>When the paper is published it will be referenced here.</p><p></p><p>The code is structured in 3 folders, all using the same population data, obtained from <a href="https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/grids" target="_blank">Eurostats</a>. The discrete_environment folder contains all of the code related to the discretization, Dijkstra solutions and postprocessing of the Dijkstra output. The continuous_environment folder contains a fork of the <a href="https://github.com/TUDelft-CNS-ATM/bluesky" target="_blank">BlueSky Open Air Traffic Simulator</a> repository, with all of the plugins related to training the Deep Reinforcement Learning algorithm, and evaluating of the paths in the continuous environment. Finally the policy_plotter folder contains all of the tools for generating the visuals presented in the paper.</p><p></p><p>Before running or going through the code make sure to read the ReadMe file.</p><p></p><p>The software is also available on github: https://github.com/jangroter//PathplanningDRL</p> ER -