Improved Dual Coding Genetic Algorithm for the Optimal Design of Looped Pipe Network Based on Cycle

An optimization model for looped pipe network is established, in which the least investment is taken as objective function, and the improved genetic algorithm is used to solve the problem. Based on the characteristics of looped pipe network optimization, the dual coding genetic algorithm combining binary coding with integer coding is adopted to optimize the pipe network layout and pipe diameter simultaneously, and the restriction of pipe diameter is satisfied automatically. Aiming at the binary coding, in the operations of initial solutions creating, crossing and mutation, cycle-based method is adopted, the infeasible solutions are avoided or the probability of producing infeasible solutions is reduced, and the fitness function is improved, so the optimum performance of the algorithm is improved. In the example the investment cost is reduced 11.65%, and it shows that the algorithm is efficient.