ENS Rennes - Optimization

Optimization is a solver for P2P energetic problem.

---

Description

This project is a solver to minimize P2P energetic problem based on a European dataset releases in this journal.

This project read all CSV file, execute script on the collected data, then display the result in a graphstream windows.

There are some arguments you can use to launch this programm :

• -debug false will print on terminal some debug logs if set to true (default false).
• -gui false will display data on graphstream windows if set to true (defautl false).
• -generator false will display on graphstream generators if set to true (default false).
• -date 2013-12-01 00:00:00 a date to load on the CSV loads file.
• -algo scipy which algo to use, scipy is based on scipy optimize, when admm is a custom implementation.
• -worker 5 number of threads used to compute the specified algo on dataset when date is set to all (default 2).

Date need to be formated as follow : YYYY-MM-DD HH:00:00 with HH between 00 and 23.

---

Setup

This project is compiled under java 8 in order to deploy it more easily. Make sure to install Java 8 and set java.exe to your PATH variables.

Download from this permalink the network_edges.csv and load_signal.csv that are too heavy for git and place these files under src/main/resources/.

Comment the first line of these two files with a #.

---

Docker

To create a local docker image, make sure to execute gradlew clean jar first to generate the jar file, which is located under build/libs/.

Then run :

docker build --no-cache . -t optimization[:tag]

You can run this image with custom jvm arguments through env var.

docker run -d --name optimization --mount type=bind,source="/c/sqlite",target=/optimize/sqlite optimization[:tag] -debug false -date all -worker 5