Study On Optimization Model Of Water Supply Network Based On Improved PSO-DE Hybrid Algorithm

With the increase and improvement of living standards of the urban population, urban water supply systems in the growing role of social development. The water distribution network system is an important part of the urban water supply systems, which generally invest 50% to 80% of the total investment of the entire water supply system. Pipe network design is reasonable not only relates to the water supply is safe, but also related to the construction and operating costs whether the economy. Therefore, to carry out water supply network optimization design is important.In a systematic review of domestic and foreign water supply network foundation to optimize the design of the progress in the study on the the traditional single objective network optimization considering only total network cost in converting value minimum problems, put forward to the total cost of the network reduced annual value of the smallest, the pipe network hydraulic reliability degree and reliable water quality is of the largest degree for multi-objective optimization model. This paper analyzes the see network optimization model and optimization algorithm and the Improved Particle Swarm- differential evolution(pso-de) hybrid algorithm of network of optimization design, and through the example analysis, the results show that the improved algorithm is reasonable and effective. The main conclusions are as follows:(1) Model analysis and discussion, aiming at the shortcomings of the grey GM(1,1) model is proposed to improve, through actual data of several commonly used forecasting model and improved grey GM(1,1) model was prediction effect contrast analysis. The results show that: to improve the Grey Prediction GM(1,1) model can better overcome GM(1,1) model of Gray's shortcomings, higher prediction accuracy of the model MAPE=0.159%, for actual urban water supply system mid long term load forecasting of several commonly used water.(2) The water supply network of single water source was analyzed, and the optimization model of multi objective network was established. Starting network was established to build the cost objective function is introduced water quality entropy concept, establish reliable water quality objective function, and pipeline construction costs and operating costs, and water pipe network system reliability objective functions simultaneously constructs a multi-objective optimization model.(3) According to the characteristics of particle swarm optimization(PSO) and differential evolution algorithm(DE), the two party is improved, and an improved particle swarm optimization algorithm is proposed. In the optimization design of circular pipe network, the method of quasi Newton method is used to calculate the hydraulic balance, which reduces the computational effort.(4) To has the representative of the water supply pipe network project to optimize the design, on the platform of matlab by results based on integer coding improved pso-de hybrid algorithm is used to solve the model, and the optimization with the traditional method, the results obtained were compared. The results show that compared with the traditional methods, the optimization model to obtain the design project with a total investment than traditional models and save costs of about 6.0%, pipe network system reliability degree increased 10.34%, the unit reliability cost reduced 14.81%.