A Genetic Algorithm Application for Gas Pipeline Network Optimization

During the last decade, demands for optimization of available gas distribution networks or design of new networks with minimum Total Annualized Cost (TAC) have been increased significantly. Due to complicated structure of gas networks their optimization requires an efficient and robust optimization algorithm. On the other hand, various types of software for simulation of gas networks intensively have been developed, but they can rarely optimize the network. It is particularly true when optimization of a network with a lot of loops has been investigated. In these problems, network solution strategy, selection of appropriate objective function, and utilizing an efficient and reliable optimization method are very challenging issues. In this paper “SharifGasNet” (SGN) is presented as a package whose capabilities are beyond steady-state network simulators. The main advantage of SGN over the other alternatives is the fact that it can optimize the gas pipeline network based on minimization of a more rational and yet realistic objective function. SGN obtains the optimum network based on Evolutionary Algorithm.

Keywords:

gas network , constrained optimization , genetic algorithm