This project implements various algorithms for solving the Job Shop Scheduling Problem (JSSP), including Q-Learning and Actor-Critic methods. It provides both a C++ library and Python bindings for easy integration into different environments.
- Job Shop Environment simulation
- Q-Learning algorithm implementation
- Actor-Critic algorithm implementation
- Taillard problem instance generator
- Visualization of scheduling results
- Python bindings for easy integration
| Algorithm | Implemented | Notes |
|---|---|---|
| Q-Learning | ✅ | |
| SARSA | ❌ | |
| DQN (Deep Q-Network) | ❌ | |
| Double DQN | ❌ | |
| Dueling DQN | ❌ | |
| Actor-Critic | ✅ | |
| A2C (Advantage Actor-Critic) | ❌ | |
| A3C (Asynchronous Advantage Actor-Critic) | ❌ | |
| PPO (Proximal Policy Optimization) | ❌ | |
| TRPO (Trust Region Policy Optimization) | ❌ | |
| SAC (Soft Actor-Critic) | ❌ | |
| DDPG (Deep Deterministic Policy Gradient) | ❌ | |
| TD3 (Twin Delayed DDPG) | ❌ | |
| MADDPG (Multi-Agent DDPG) | ❌ | |
| QMIX (Q-Mixing) | ❌ | |
| VDN (Value Decomposition Networks) | ❌ | |
| Genetic Algorithms | ❌ | |
| Particle Swarm Optimization | ❌ | |
| Simulated Annealing | ❌ | |
| Ant Colony Optimization | ❌ |
| Feature | Implemented | Notes |
|---|---|---|
| Jobs | ✅ | |
| Operations | ✅ | |
| Actions | ✅ | |
| Bindings | ✅ | |
| Taillard | ✅ | |
| Manual Environment Creator | ❌ | |
| Tool Changing | ❌ | |
| Tool Properties | ❌ | |
| Machine Breakdowns | ❌ | |
| Job Priority | ❌ | |
| Setup Times | ❌ | |
| Due Dates | ❌ |
This project uses vcpkg to manage C++ dependencies. The dependencies are specified in the vcpkg.json manifest file in the root directory of the project.
Before building the project, ensure you have the following installed:
sudo apt-get install build-essential pkg-config cmake git curl zip unzip tar autoconf autoconf-archive libtoolsudo apt-get install libxmu-dev libxi-dev libgl-dev libglu1-mesa-devsudo apt-get install libxrandr-dev libxinerama-dev libxcursor-dev libxi-devsudo apt-get install python3 python3-devsudo apt-get install linux-tools-common linux-tools-generic
sudo /usr/lib/linux-tools-6.8.0-36/perfTo install vcpkg:
git clone https://github.com/Microsoft/vcpkg.git
cd vcpkg
./bootstrap-vcpkg.shSet the VCPKG_ROOT environment variable to the vcpkg installation directory:
export VCPKG_ROOT=/path/to/vcpkgThis project uses CMakePresets.json to manage build configurations. There are two main configurations: debug and release.
-
Clone the repository:
git clone https://github.com/cair/job-shop-simulator.git cd job-shop-simulator -
To configure and build the project in debug mode:
cmake --preset=debug cmake --build --preset=debug
-
To configure and build the project in release mode:
cmake --preset=release cmake --build --preset=release
To install the Python package, first ensure that the VCPKG_ROOT environment variable is set, then run:
pip install .This will automatically use CMakePresets.json to build the project and install the Python package.
To use the C++ library in your project, include the necessary headers and link against the library:
#include <jobshop/job_shop_environment.h>
#include <jobshop/job_shop_qlearning.h>
// Your code hereAfter installing the Python package, you can use it in your Python code:
import jobshop
# Create a job shop environment
env = jobshop.JobShopEnvironment(jobs)
# Create a Q-Learning agent
agent = jobshop.JobShopQLearning(env, alpha=0.1, gamma=0.9, epsilon=0.3)
# Train the agent
agent.train(num_episodes=1000)
# Print the best schedule
agent.printBestSchedule()src/: Contains the C++ source filesinclude/: Contains the header filesbindings/: Contains the Python bindingsCMakeLists.txt: The main CMake configuration fileCMakePresets.json: Defines the build presetsvcpkg.json: Specifies the project dependencies for vcpkg
This project is licensed under the MIT License. See the LICENSE file for details.
Contributions are welcome! Please feel free to submit a Pull Request.
If you have any questions or feedback, please open an issue on the GitHub repository.