In this repository, we focus on utilizing Reinforcement Learning (RF) agents to solve the elevator problem. We explore different algorithms to train our agents, providing a rich environment for experimentation and learning.
- Red: Represents an outside call.
- Blue: Represents an inside call.
- Magenta: Used when both inside and outside calls are present.
Please note that the visualization does not depict individual humans, however, in the code, multiple humans can exist on the same level. Each human is assigned a specific weight, and the elevator has a defined MAX_weight that it can carry.
A foundational framework to generate various tests for training models with different parameters.
Interactive visualization built in Unity that communicates with the Python backend. Witness the training in real-time or dive deep into the results to understand agent behavior.
A naive approach that surprisingly provides commendable results. Use this as a starting point to understand the problem before diving into more complex algorithms.
Dive deeper with the Q-Table algorithm, exploring how a tabular approach can be used to optimize the agent's decisions in the elevator problem.
- git clone https://github.com/your_username/elevator_rf_agents.git
- cd elevator_solver
- pip3 install -r requirements.txt
- create the .env (see the .env.example)
- modify environment.py (/elevator_solver/core/utils/environment.py) to set up the expiriment
- python3 generate_cases.py
- python3 train.py
- python3 validate.py
- python3 server.py (start Falsk serveer to send action to Unity visualisation)
- start Unity project (/elevator_solver/unity/)
- Feel free to contribute to the project. Raise an issue or submit a pull request.
- Baseline Model
- Q-Table Algorithm
- Deep Q-Network Implementation
- Policy Gradient Methods
- Further Integration with Unity for Advanced Visualizations