| Urban Water supply network is an important component of the urban water distribution system. Water supply infrastructure, security, reliability, efficiency is the guarantee of social and economic development. It is an important part of the water supply system. Reasonable for pipe network hydraulic calculation and analysis, to achieve safe, reliable operation and management plays a crucial role. Through hydraulic station simulation on water supply network, one can control only a small amount of data points to reflect all points throughout the pipe network information, thus greatly save cost and enhance efficiency. Currently, research in this area there are many deficiencies, failed to give full play to the computer's function.Based on previous literature data and related engineering experience, a reasonable choice of model and analyze the constraints on the subject in-depth research. First, the standard genetic algorithm (SGA) was carried out simulations, and then gradually introduced to expand the sampling space, floating-point encoding, optimal preservation strategy, arithmetic crossover and dynamic mutation genetic algorithm technology for multi-angle standard improvements made good as expected. Finally, in theory, these improvements demonstrates the rationality and necessity of technology.To make up for lack of local search ability of genetic algorithm case, the paper introduced a simulated annealing algorithm (SA), simulated annealing SA algorithm as an improved GA module, the probability of a certain group of local search to further improve the GA performance, improve the search efficiency and accuracy.Programming software used in this paper has a good numerical simulation of MATLAB functions in MATLAB programming platform, the first simple example shows that the mining algorithm effectiveness and success of research results, and then verified, and are programmed to do a optimal design, finally, to a county for the actual simulation cases pipe network, and verify the feasibility and practicability of the improved method. |