Spatial Solution Method Of Residual Chlorine Model For Water Supply Network Based On Cellular Automata

The safety of residential drinking water as an important part of water supply safety has been raised to a national strategic level,With the promulgation and implementation of the "Hygienic Standards for Drinking Water"(GB 5749-2022),there are higher accuracy and adaptability requirements for residual chlorine prediction models in water supply networks,and most of the current residual chlorine model operating efficiency and prediction accuracy,especially for the low and medium flow rate pipe network end pipe section residual chlorine accuracy does not meet the existing water quality monitoring needs.Therefore,this paper uses cellular automata to establish Eulerian method-cellular automata residual chlorine model and Lagrangian method-cellular automata residual chlorine model based on two residual chlorine model spatial resolution methods,and study the model optimization method,parameter sensitivity,numerical accuracy and adaptability conditions.Firstly,In the process of solving the residual chlorine model of water supply network by Euler method,the graph based extended cellular automata algorithm is introduced,and an activation/negative activation asynchronous method is used to optimize the model calculation steps,and a Euler method method cellular automata residual chlorine prediction model is established.and the performance of the model was analyzed by using the water supply network,and two types of residual chlorine inputs,constant input and impulse input,were used to study the model adaptability.The results show that the MSE values of this Eulerian model and Epanet2.0 in two sets of simulations are 0.04 and 0.07 respectively,which verifies the prediction accuracy of the new model.The time required to reach the maximum number of nodes before and after the introduction of the asynchronous method is 1250 s and 750 s,and the maximum acceleration ratios of both in two sets of simulations are 3.9 and 4.2 and 3.2 and 2.9 respectively,which verifies that the asynchronous method has a certain improvement on the model The maximum speedup ratios in the two sets of simulations are 3.9,4.2 and 3.2,2.9,respectively,which verifies that the asynchronous method has a certain improvement effect on the computational efficiency of the model.Then,the analytical model of residual chlorine by cellular automaton and Lagrangian method is combined with the first-order reaction kinetic equation based on chlorine to establish the ordinary differential equation,and the Lagrangian method-cellular automaton residual chlorine model is established by solving the equation with the numerical method of Longacurta.The flow velocity and random displacement are used to represent the advection and radial diffusion processes in the model,respectively,and the CPU running time is reduced based on a model optimization method.The model is analyzed using a pipe network,and the results show that the optimization method reduces the CPU running time to 0.11 s.The new model produces lower MSE than the Epanet 2.0 simulation by selecting appropriate The model simultaneously considers fluid advection,dispersion,and the main water reaction and wall water reaction,and is almost independent of numerical dispersion.The model is almost independent of numerical dispersion and produces smooth residual chlorine concentration simulation curves,and the simulation effect under medium radial diffusion conditions is better than that under low radial diffusion conditions,while the residual chlorine decay coefficient does not affect the simulation accuracy of the model.Finally,the performance of the two Lagrangian models including Euler method,dispersion model and advection model is compared by using the water supply network in North Malyn Raion District.The results show that the first two models have good running speed,and the Lagrangian Cellular automaton residual chlorine model considering the effect of pipeline dispersion has better simulation effect than the other two methods under the condition of low residual chlorine concentration.And in response to the problem of low residual chlorine prediction values in some nodes and time periods of the pipeline network,the 24-hour residual chlorine concentration prediction curve of nodes and the proportion of water sources at each node are analyzed,and suggestions for optimizing the operation status are given.