DESIGN OPTIMIZATION OF GAS TRANSMISSION SYSTEM WITH DIFFERENTIAL EVOLUTION ALGORITHM

Abstrak

Gas distribution through a pipeline network is a highly complex process and requires significant financial investment. This network system consists of a source, pipe, the compressor and sink (consumer). The source is the node where the producer has the gas pressure that will be distributed, the pipe is used to connect the producer and the consumer. Between the pipes there is a compressor which functions to increase the pressure. This network system was created at a significant cost, so it is necessary to search for minimal costs, but consumer demand is still met. This research discusses the search for an optimal gas network with minimum costs. This minimum cost depends on several parameters i.e. the length and diameter of pipe, also the pressure on the compressors entry and exit points. There are many optimization methods, but one of the simple and easy to implement methods is the Differential Evolution Algorithm, so this method is used to determine the optimal solution to this problem. Researchers used seven DE variants based on mutation strategies, namely DE/rand/1, DE/best/1, DE/rand/2, DE/best/2, DE/current-to-best/1, DE/current-to- rand/1, and DE/rand-to-best/1. The seven variants have never been used in gas distribution networks by previous researchers. Therefore, the seven variants were compared, and the minimum solution was determined. The results show that the DE/best/2 variant is the variant that produces the minimum total costs compared to the other variants. DE/best/2 achieved the lowest annual operating cost at USD 13.99 million.