Application Of Artificial Fish Swarm Algorithm In Optimization Drainage Pipe Network

With the accelerated process of urbanization in China,people have paid more and more attention to the construction of drainage pipe network.Urban drainage pipe network construction is an important part of municipal facilities.Drainage pipe network construction has a very important impact on urban construction,so the study on the optimization of drainage pipe network is of great significance.Some achievements have been made in the research of drainage pipe network in China,but there is still a certain gap between the developed countries and the developed countries.In recent years,in addition to the traditional form algorithm,the genetic algorithm is used to optimize the calculation of drainage pipe network.Although the genetic algorithm has many advantages,it can not converge to the global optimal solution because of its lack of searching ability.Artificial fish swarm algorithm can quickly jump out of the local extreme,fast search characteristics,this paper attempts to use artificial fish swarm algorithm to optimize the drainage network.The artificial fish swarm algorithm can be applied to the continuous variable optimization problem,and the application of the discrete variable optimization problem is still limited,but the result is satisfactory in solving the TSP problem.This paper studies the characteristics and the optimization method of discrete variables,re encoding and decoding through the variable,and the concept of artificial fish swarm algorithm was modified and applied to the drainage pipe network optimization design of discrete variables.In this paper,the pipe diameter and depth are used as decision variables,and the relationship between flow rate,filling degree,design slope and other variables,such as the diameter and depth of.Simplify the model,reduce the difficulty of programming.Solve the minimum value of the water drainage pipe network investment cost function,with the MATLAB algorithm.Finally,a numerical example shows that the algorithm is simple and feasible.