README.md

Optimizing the topology of a gas transmission network

Table of Contents

• Introduction
• Implementation

Introduction

This project implements a mathematical formulations of gas network topology optimization, which is solved by using IBM® Decision Optimization (DO) for Data Science Experience (DSX) Local.

The project is composed by the following assets:

• Notebooks:
  • Gas_Network_Optimization
• DO Model:
  • Gas_Network_Optimization
• Datasets:
  • A collection of CSV files representing the data of the gas network (derived from gaslib.zib.de), plus two different transportation scenarios that specify certain gas quantities

Implementation

The implementation is based on Jupyter notebooks with Python 2.7, and DO Models for DSX. The main project assets are described in the following.

Notebook Gas_Network_Optimization

This notebook implements a mathematical model for the gas network topology optimization problem. A simplification is used which leads to a mixed-integer linear program that is modeled and solved with IBM® DO CPLEX® Modeling for Python: docplex. The addressed formulation is described and discussed within the notebook.

DO Model Gas_Network_Optimization

This DO Model handles the optimization approach implemented in the notebook Gas_Network_Optimization. In particular, several DO Scenarios are created and the optimization is performed for each scenario. This DO Model contains also a dashboard that reports a summary chart comparing cost KPIs and an overview of new pipelines which have been selected by the optimization model.