System design by constraint adaptation and differential evolution
IEEE Transactions on Evolutionary Computation
Hi-index | 0.00 |
A mathematical model of optimal scheduling scheme for natural gas pipeline network is established, which takes minimal annual operating cost of compressor stations as objective function after comprehensively considering the resources of gas field, operating parameters of compressor stations and work conditions of pipeline system. In the light of the characteristics of the objective function such as nonliner, more optimal variables and complicated constraint conditions, based on the thought of modern heuristic evolutionary-algorithm, this paper presented a new hybrid genetic algorithm, which is featured by global search, fast convergence and strong robustness. It combined the reproduction strategy of differential evolution algorithm with the crossover and mutation of genetic algorithm. With the dynamic calibration of fitness and the elitism strategy of the optimal individual, this algorithm can improve the ability of searching and avoid the premature convergence effectively. The case study of a certain pipeline network system with 11 nodes, 11 pipelines,2 compressor stations demonstrates the effectiveness and application of the established model and algorithm. The optimal scheduling scheme could be adapted to daily operation and future retrofit of gas pipeline network.