Study On Site Selection Of Water Quality Monitoring Points In Water Supply Network Based On Multiobjective Optimization

The water supply network is one of the major national infrastructures and an important carrier for transporting water and ensuring water quality.However,the water supply network has a complex structure and is widely distributed,making it difficult to achieve all-round and time-sensitive monitoring of the network considering the monitoring costs and subsequent economic benefits.Therefore,the limited number of monitoring points to quickly monitor the pollution events is very critical,and how to select a representative network node as a water supply network water quality monitoring points is a key challenge.This study selected a quasi-dynamic hydraulic water quality model,the construction of "node importance index highest" "cover the largest amount of water" "the average monitoring time is the shortest" multi-objective optimization model,in MATLAB software platform using the improved Harris Eagle algorithm(INSHHO)for its solution,for the water supply network water quality monitoring point location to provide a solution.The main research work is as follows:(1)based on EPANET2.2 pipe network model,in UCINET software platform based on the graph theory to build a pipe network topology model.Innovative network analysis method applied to the water quality monitoring point of the water supply network site selection study.The network index of pipe network nodes and hydraulic water quality condition indexes combined to build a node importance index model,and the use of hierarchical analysis and entropy method to assign weight to the combination of indicators.The results show that the points with high importance index are mainly distributed in the key position of the structure and poor hydraulic and water quality conditions.The node importance index provides a scientific reference basis for the location of water quality monitoring points.(2)with the help of MATLAB software platform,in 24 working conditions,the use of quasi-dynamic hydraulic water quality model to obtain the simulation data,the construction of the corresponding time step node water quality ratio matrix;and the preparation of Dijkstra(Dijkstra)algorithm program,the hydraulic data into any node transmission time matrix.The results show that the large amount of basic data processed by the working conditions can more accurately reflect the actual operation of the pipe network.(3)The "water coverage" and "average monitoring time" are selected as the monitoring objectives of endogenous and exogenous pollution for cooperative tracking,and the principle and expression method of the optimization objectives are analyzed.At the same time,on the basis of the two-objective optimization model and the innovative addition of "node importance index maximum" objective,forming a three-objective optimization model,to provide decision makers with a third dimensional decisionmaking basis,to improve the comprehensiveness and reliability of the layout of water quality monitoring site selection.(4)The Harris Hawk algorithm(HHO)is improved.Using HHO algorithm,nondominated ranking multi-objective genetic algorithm(NSGA-II),multi-objective particle swarm optimization algorithm(MOPSO)and the improved INSHHO algorithm to solve the algorithm case two-objective optimization model,it can be found that the INSHHO algorithm obtains the most number of optimal solutions,and the objective function value is optimal year-on-year.Meanwhile,in the three-objective optimization model,the IGD(Inverted Generational Distance)of the INSHHO algorithm is the smallest and the HV(Hypervolume)of the hypercube is the largest compared with other algorithms,which proves the excellent performance of the INSHHO algorithm.In addition,the validation of the pipe network case also shows that the three-objective optimization model solution results are more representative.(5)The actual water supply network data in area A was imported into EPANET2.2 platform for modeling,and the final solution was obtained after calibration and model solving using the above method.the number of monitoring points in each of the two areas in area A was set to 6,and the coverage rate was greater than 75%,the average monitoring time was around 0.5 hours,and the node importance index was also in the top.In addition,also on the pipe network for pollution source intrusion simulation,verification results show that the scheme can meet the coverage of water and monitoring time requirements,and the location of water quality monitoring points are more typical,the monitoring benefits of the pipe network is good.