Optimization Design of Natural Gas Pipeline Based on a Hybrid Intelligent Algorithm

Natural gas pipeline is an important link between natural gas exploitation and utilization. To coordinate the domestic and foreign natural gas resources and satisfy the needs of various regions’ economic development, China plans to continue accelerating the construction of natural gas pipeline network during the 13th five-year plan period. However, the problems of irrational pipeline construction and unexpected high cost of investment come out in the early stage. To solve these problems, this paper proposes a hybrid intelligent algorithm for optimization design. Firstly, the pipeline throughput is determined based on predicting the natural gas demand in the region with grayscale coupled neural network model, and then a mathematical model is established for natural gas pipeline system design, using particle swarm optimization (PSO) algorithm coupled with simulated annealing (SA) to solve this problem. A real gas pipeline in China is taken as an example to demonstrate the effectiveness of SA-PSO compared with traditional PSO algorithm.