This model serves as the baseline for the Drone-Based Reforestation environment, trained and tested on task 3 with difficulty 4 using the Proximal Policy Optimization (PPO) algorithm.
Environment: Drone-Based Reforestation
Task: 3
Difficulty: 4
Algorithm: PPO
Episode Length: 2000
Training max_steps: 1200000
Testing max_steps: 300000
Train & Test Scripts
Download the Environment
Evaluation results
- Cumulative Distance Reward on hivex-drone-based-reforestationself-reported1.3088074851036071 +/- 0.2150446017537585
- Cumulative Distance Until Tree Drop on hivex-drone-based-reforestationself-reported48.39169616699219 +/- 5.7290635352219415
- Cumulative Distance to Existing Trees on hivex-drone-based-reforestationself-reported64.92751655578613 +/- 4.971308759365728
- Cumulative Normalized Distance Until Tree Drop on hivex-drone-based-reforestationself-reported0.13088074818253517 +/- 0.021504460692074904
- Cumulative Tree Drop Reward on hivex-drone-based-reforestationself-reported3.975467157363892 +/- 0.5825542394332213
- Out of Energy Count on hivex-drone-based-reforestationself-reported0.04072725491598248 +/- 0.02455264640086014
- Recharge Energy Count on hivex-drone-based-reforestationself-reported10.695373344421387 +/- 0.5710744663818678
- Tree Drop Count on hivex-drone-based-reforestationself-reported0.9469023418426513 +/- 0.033324031370099066
- Cumulative Reward on hivex-drone-based-reforestationself-reported101.1289744567871 +/- 3.8212373566735938