| The water supply network is a huge investment in urban infrastructure.Optimizing the design of the water supply network will not only help to ensure the reliability of the water supply,but also reduce t he project cost,which has important research significance.Since the optimal design of the water supply network is a complex discrete combination problem,the intelligent optimization algorithm at this stage has unique advantages in solving this type of p roblem.Existing research shows that the difference algorithm in intelligent optimization algorithms(such as genetic algorithm,particle swarm optimization,ant colony algorithm,etc.)is suitable for solving water supply network optimization problems.Ho wever,the difference algorithm has the problems of low computational efficiency and premature maturity when dealing with large-scale water supply network optimization problems.To solve this problem,researchers have improved the difference algorithm from different aspects,such as improving its mutation or crossover operations.Different from previous research,this paper proposes a differential evolution algorithm based on optimal selection strategy,which improves the efficiency of differential calculat ion by improving the selection strategy of the differential algorithm and obtains a higher-precision solution.The research content and achievements of this article include:1.Based on the existing water supply network optimization design modeling method,a combination method of MATLAB calling EPANET software is proposed.Firstly,it analyzes the hydraulic calculation theory and mathematical model of the optimal design of the water supply network;secondly,it summarizes the EPANET software hydraulic calcu lation principle and modeling process;finally,discusses how the EPANETH software and MATLAB software establish the calling relationship through the EPANETH toolkit.2.Aiming at the problem that the traditional differential algorithm selection strategy is not conducive to quickly reducing population diversity,a differential algorithm based on optimal selection strategy is proposed.First,the principle of the differential algorithm for the optimal selection strategy is explained;then,taking the three w ater supply network of New York Tunnels,Hanoi,and Balerma Network as an example,the feasibility of the differential algorithm for optimal selection strategy is analyzed.It is concluded that the optimal selection strategy difference algorithm and the tr aditional differential algorithm have the same parameter combination,and the calculation efficiency is increased by 43.60%,46.70%,and 72.86% respectively.Finally,the parameter sensitivity of the optimal selection strategy difference algorithm is analy zed,and the experimental results show that Compared with other optimization algorithms,the proposed algorithm uses a smaller amount of calculation to find the new optimal solution of the BN case(1.921337 € M),which shows the superiority of the proposed algorithm in terms of computational efficiency and search ability.Specifically,it saves 91.26%,34.86%,and 57.04% of computational power than the NSGA-II(1.9215 € M),NLP-DE(1.923 € M),and HD-DDS(1.941 € M)algorithms,respectively.3.In order to solve the problem that the optimization of large-scale water supply network needs to check the parameter combination repeatedly,a new progressive parameter checking method is proposed.Firstly,after a few optimization operations in the parameter combinat ion,the sensitivity analysis of the parameter combination is carried out according to the solution quality;secondly,the interval containing the best parameter combination is estimated;finally,multiple optimization operations are carried out in the int erval to determine the best parameter combination.The results show that the known optimal solution of BN case is found by using this checking method,which can minimize the cost and maximize the performance of the algorithm within the allowed calculation amount,and is helpful to analyze the sensitivity of parameter combination. |
PDF Full Text Request