Optimal Design Of Water Supply Systems Via Improved Adaptive Particle Swarm Optimization

Water supply network is very important for the water supply systems, and plays an important role in the protection of rapid economic development and people’s daily life. As the rapid progress of urbanization, the existing water supply network can’t meet people’s growing demand for the water, so the reconstruction of water supply system has become an urgent problem. Because of the high cost, for the water supply system, the reasonable reconstruction and optimization design can not only save a lot of project investment, but also can increase water supply system reliability. The accomplishment of this paper puts forward a network optimization models and designs a new improved Particle Swarm Optimization algorithm, being used to optimize the water supply network reconstruction. The primary content and innovations of this research work are as follows:(1) The improved adaptive design for particle swarm optimization algorithmIn this paper, aiming at the shortcomings of the particle swarm optimization algorithm to solve combinatorial optimization problems, from the point of view of mathematical, the concept of similarity is put forward. To assess population distribution information, and adjust adaptively the parameters to balance the global and local searching ability of the algorithm, a novel improved particle swarm optimization algorithm is proposed. By optimizing the benchmark test functions, the results show that the improved adaptive particle swarm optimization algorithm can jump out of local optimal and be able to find the global optimal value with the faster convergence speed. The improved algorithm can avoid trapping in local optimum and accelerate the convergence speed, but algorithm local search ability is not strong. Therefore, to make full use of population information, the concept of convergence factor is proposed, the novel improved algorithm can evaluate population information, adjust adaptively parameters, and is to join the local search mechanism. Hanoi network and New York is optimized, the results show that the improved algorithm can find the global optimal value with the smaller computational cost, effectively solve this kind of large scale combinatorial optimization problem.(2) The establishment and optimization of the water supply network modelWater supply network model is very important for the water supply network reconstruction. The water demand prediction can provide reasonable basis for the establishment of the optimization model. In this paper, on the basis of the analysis of the traditional forecast methods, the predictive model of neural network is proposed to predict the water demand according to the historical data, the results of the model validation show that neural network model can improve prediction accuracy even with less data. On the basis of water demand prediction data, through the analysis of the economic evaluation index of the optimization model, the optimization model of water supply network is established, considering water supply system reliability and hydraulic balance relationship. Combined with improved particle swarm algorithm, the water supply network optimization model was designed for the actual engineering. This proposed a method for the optimization of the water supply network. Through calibration the optimization model, the results show that the model accords with the engineering practice, and can be used to design the water supply network reconstruction optimization.(3) The design of the optimization platform and practical engineering optimizationWater supply network reconstruction is a complex project, it is necessary to establish a unified, standardized platform. In this paper, according to the engineering practice, a water supply network optimization platform is designed. By an engineering example, the optimization model and improved algorithm is verified scientifically and reasonably on the optimization platform. By the traditional design method and the improved particle swarm algorithm, the water system of a university in Beijing is designed, the results show that the improved particle swarm algorithm gains better scheme and save project investment, on the basis of meeting the actual project requirements for water quantity, water quality and water pressure. The research has a certain theoretical significance on the improvement of particle swarm algorithm.