This model serves as the baseline for the Aerial Wildfire Suppression environment, trained and tested on task 2 with difficulty 2 using the Proximal Policy Optimization (PPO) algorithm.
Environment: Aerial Wildfire Suppression
Task: 2
Difficulty: 2
Algorithm: PPO
Episode Length: 3000
Training max_steps: 1800000
Testing max_steps: 180000
Train & Test Scripts
Download the Environment
Evaluation results
- Crash Count on hivex-aerial-wildfire-suppressionself-reported0.12222222536802292 +/- 0.16381577701779051
- Extinguishing Trees on hivex-aerial-wildfire-suppressionself-reported6.599999992549419 +/- 15.242522731448657
- Extinguishing Trees Reward on hivex-aerial-wildfire-suppressionself-reported32.9999993622303 +/- 76.21261195341214
- Fire Out on hivex-aerial-wildfire-suppressionself-reported0.3138888914138079 +/- 0.4019425711973155
- Fire too Close to City on hivex-aerial-wildfire-suppressionself-reported0.6666666671633721 +/- 0.44261318597616683
- Preparing Trees on hivex-aerial-wildfire-suppressionself-reported747.7777755737304 +/- 635.3383235803965
- Preparing Trees Reward on hivex-aerial-wildfire-suppressionself-reported3738.888851928711 +/- 3176.691613050029
- Water Drop on hivex-aerial-wildfire-suppressionself-reported23.388888955116272 +/- 12.81474416895113
- Water Pickup on hivex-aerial-wildfire-suppressionself-reported22.874999976158144 +/- 12.792184644801207
- Cumulative Reward on hivex-aerial-wildfire-suppressionself-reported3947.1150146484374 +/- 2234.072313108481