This repository contains code with algorithms for encoding system architecture selection choices: discrete choices (e.g. between different material types or fuel types) that influence each other. It is useful in the context of system architecture optimization, where some model of the architecture design space is created and automatically searched (optimized) to find the best architecture instance.
For more background on architecture optimization and modeling architecture design spaces, see: System Architecture Design Space Exploration: An Approach to Modeling and Optimization
An encoder is instantiated and queried for a given encoding scenario, specified by:
- A list of nodes, representing elements of the architecture (design space)
- A list of choice nodes mapping to mutually-exclusive option nodes
- An influence matrix
- A choice constraint map
The influence matrix is an n_nodes x n_nodes matrix, specifying for each node their status and influence:
- On the diagonal, the initial node status is defined
- On the off-diagonal, the influence of node
ion nodejis defined
| Status code | Diagonal | Off-diagonal |
|---|---|---|
| 0 | Not selected | No influence |
| 1 | Selected/included | Node i selects node j |
| 2 | Removed | Node i removes node j |
| 3 | Invalid architecture | |
| 4 | Choice made |
The choice constraint map (optional) defines for each choice-option pair which other choice-option pairs are removed. This is used for modeling choice linking and other constraints.
Clone the repository and:
pip install -e .
The project is coordinated by: Jasper Bussemaker (jasper.bussemaker at dlr.de)
If you find a bug or have a feature request, please file an issue using the Github issue tracker. If you require support for using SelectionChoiceEncoding or want to collaborate, feel free to contact me.
Contributions are appreciated:
- Fork the repository
- Add your contributions to the fork
- Update/add documentation
- Add tests and make sure they pass (tests are run using
pytest)
- Read and sign the Contributor License Agreement (CLA) , and send it to the project coordinator
- Issue a pull request