Optimization Of PMA Partition Of Urban Water Supply Network By Graph Theory And Evaluation System

The leakage rate of majority of urban water supply network is high, which caused a great waste of water and energy, and it need to carry out leakage control over the whole country in the water supply network. The foundation of the leakage control of water supply network is the partition of the network, the partition form of water supply pipe network directly determine the efficiency of leakage control. Pressure control is the most economic method of leakage control in water supply network. In this paper, the water supply network pressure management area(PMA) is pressure control based, this form of partition belong to medium-sized district metering area(DMA), but it is different with DMA target partition, PMA partition is mainly relate to pressure control, and taking into account the district metering.Because the actual pipe network cannot be experimented. Therefore, this study first established leakage of hydraulic simulation model based on pressure driven node demand(PDD). PDD modle is built based on traditional hydraulic modle.The node flow in traditional hydraulic modle is divided into two parts, the user requirement and the leakage flow. The user requirement varies with the time,and the leakage flow is along with the change of node pressure. PDD modle can simulate the leakage of water supply network, and provide a basis for water supply network PMA partition and evaluation.In order to guide the partition decision, the partition evaluation system,which include the pre evalution and the partitioned evaluation, is established.The pre evaluation is mainly used to determine the feasibility of partition and leakage levels, and the partitioned evaluation is used to evaluate the optimization of the partition scheme.This study consider PMA partition method of graph theory and evaluation system. First of all,it use adaptive AP clustering algorithm along with economic calculation to form the initial partition of the water supply network. Then the paper applies Dijkstra algorithm to calculate the shortest path of each cluster center point to the source of the water, to determine the position of entrance of each partition. Then it establish a partition boundary optimization model, simulated annealing algorithm is used to solve the model. Finally it conbine part of the partition to form the final partition scheme.The method of water network partition is applied in Y City water supply network, and it eventually find a way to meet the hydraulic conditions of the partitioning scheme.The program satisfy the evaluation criteria and it can be used to guide the actual water supply network partition.