| The water distribution networks of urban in China invaded by the exogenous sudden pollution incidents are increasing in recent years. As the pipe network itself has pushed the migration flow, inertial flow, hydraulic transients, topology and complex characteristics. Large-scale urban of water supply pipe network, once caused by the occurrence of exogenous sudden pollution accidents caused by human factors such as the deliberate destruction,the contaminants in the pipe network easily diffused to form a global danger. However, the emergency treatment of this kind of exogenous sudden pollution accidents invasion, the most important thing is to accurately and quickly determine the location of sources of invasion, invasion time and intrusion concentration. To solve the problem of pollution locating under the exogenous sudden pollution incidents in supply pipe network, the article as water supply pipe network of the region of the South G city for the study, expand the water supply network to track anti-burst sources to build the model and solve the model research methods. According to the attribute data, the call data and the business data of water supply network of the town, using EPANET software to establish the model of water hydraulic and quality, based on the forward modeling and the inverse modeling of water quality, building simulation-optimization backtracking model and proposing an optimal simulation method based on the particle swarm optimization (PSO) combined with ant colony optimization (ACO). It provides an effective method for locating and identifying source of pollution of water supply network.Firstly, based on the database of the static and dynamic of the water supply pipe network in the town, the software platform EPANET 2.0 was used to set up hydraulic model, man the variation of flow and velocity under different time and space of the network was analyzed to closing to the real operational states.Secondly, based on the optimization location of water supply network with water quality monitoring of exogenous under sudden pollution accident, according to the static and dynamic of hydraulic model, adaptive particle swarm K-medodis clustering algorithm was used to arrange the location of sudden pollution water quality ten monitoring stations with 89.8% coverage rate. The test about heavy metals non-attenuation contaminants invading the network was done by the software platform EPANET 2.0. Forward modeling of water quality made a simulation of exogenous sudden pollution accident invading the real network to analyze the variation of flow and velocity under different time and space.Thirdly, the mathematical model of the city for the simulation and optimization of the water supply pipe network was built,the Toolkit Dynamic Link Library (DDL) of the software EPANET 2.0 was invoked on the MATLAB software platform to get the data of water supply network water quality monitoring stations under the simulation of sudden pollution accident. According to the data which was get before, in order to get the mathematical optimization expression of the difference between real values with analogue value.In the end, based on PSO-ACO fusion algorithm, the solution model of pollution source simulation and optimization back tracking model is proposed. In view of the defects and shortcomings of the PSO algorithm to solve the model, fusion ACO algorithm pheromone mechanism, and puts forward the ability of PSO-ACO fusion algorithm of search threshold with the advantages of full called particle swarm optimization (PSO) algorithm to improve the solving the pollution source trace back model. This train of thought on the platform of MATLAB software based on the source of pollution of the water supply network positioning system program code and call the EPANET software Toolkit dynamic link library, and then used to solve the mathematical model. Associated with the same parameters, the fusion of PSO and PSO-ACO algorithm was used to solve the model. The results show that, compared with the PSO algorithm, PSO-ACO fusion algorithm not only greatly improve the model accuracy, and can effectively shorten the solving time model. In addition, the effect of PSO-ACO fusion optimization algorithm of two parameters search threshold and population size respectively to solve the model results accuracy and computational efficiency is analyzed. The results show that, reasonable setting these two parameters, the accuracy and computational efficiency of the model are higher. |
PDF Full Text Request