Design And Application Of GIS-based Urban Water Supply Pipe Network Optimization

With the increasing size of the modern urban, the length and the complexity of thewater supply pipe network are also increasing. The construction, maintenance, operationcost of water supply pipe network is accounted for70%~80%of the whole pipe networksystem investment. And with the improving of the residents’ living standard and theefficiency of industrial production, the reliability of water supply system is put forwardstrict requirements. There is a close relationship between the setting of water supplypipe network radius and the improving of the economy and the reliability. Therefore,there is an important research significance of the optimal design of water supplynetwork.The main subject of the radius of the portfolio of the water supply networkoptimization. As the pipe network is large-scale, complex structure, inputpipenetwork-based data input is very complicated, so the GIS system and the pipe networkoptimization combined. Through the GIS system can quickly obtain basic data tomanage and modify data, and intuitive display optimization results, greatly shorten thedesign cycle.This paper aims to optimize the pipe diameter combination of the waterdistribution network (WDN). Inputting fundamental data of the WDN can becumbersome for its enormous network size and complex structure. Thus, the GISsystem and the WDN optimization are combined together. With the help of the GISsystem, it is made possible to obtain, manage and modify the fundamental data in aconvenient way. And optimization results are visually displayed. Which all contribute toshorten the design cycle greatly.The optimization model proposed in this paper covers the economical targets andthe reliability index. In the first place, the WDN reliability is calculated by a weightedaverage of nodal extra pressure. The effect of construction, maintenance, operation costand nodal extra pressure are all considered in the objective function. Therefore, thereliability of the WDN is improved while guaranteeing economical and operationalrequirements at the same time. Secondly, the adaptive genetic algorithm (GA) is chosen to encode design variablesand then through selection, crossover and mutation to produce each design variable withbetter fitness until the optimal solution is obtained. During this process, through elitiststrategy as well as adaptive crossover ratio and mutation ratio, the performance of thealgorithm is improved.Finally, with practical project cases, the GA optimization is compared with thetraditional design method (the economic velocity method). Which demonstrates that theGA is of practical significance in WDN optimization.