查看文章

We has presented a methodology for the application of genetic algorithm technique to pipeline network optimization. The GA codes the pipe sizes available for selection as binary strings. We have used a simple three-operator genetic algorithm namely reproduction, crossover, and mutation. Results represented to show that the genetic algorithm techniques are very effective in finding near-optimal or optimal solutions for a case study network in relatively few evaluations. The GA technique generates a whole class of alternatives solutions close to optimum. One of these alternative solutions may actually be preferred to the optimum solution based on non quantifiable measures. This is a major benefit of genetic algorithm method. The genetic algorithms technique is infancy, and further development should provide improvements in these search method for practical problem.