Research On Traceability Method Of River Point Source Water Pollution

Since the beginning of the 21st century,with the continuous development of industrial modernization and urban intelligence in China,various major emergencies have seriously endangered public safety,especially those related to surface water pollution.The research on water pollution traceability currently focuses mostly on groundwater and the water supply pipe network,with little attention paid to accidents involving surface water pollution in rivers and lakes.However,current water pollution traceability solutions continue to encounter problems like inaccurate traceability location,limited research scenarios,and inadequate universality of traceability models.Therefore,when sudden river pollution accidents occur,it is particularly important to quickly and accurately trace the source of the accident.At the same time,traceability work can also assist relevant departments in timely initiating response measures to prevent further spread of pollution,providing theoretical basis and data support for emergency management work.This thesis mainly focuses on the problem of sudden source water pollution in rivers,and uses COMSOL software to build a simulation case of sudden source water pollution events on the riverbank.The focus is on studying related issues and technologies such as single point source pollution,multi-point source pollution,and traceability methods.And the applicability of relevant methods was analyzed through comparative experiments.The main work and innovative points of this thesis are as follows:1.The thesis analyzes the sources and causes of general pollutants,basic traceability principles,and classification of traceability issues.On this basis,pollutant migration and diffusion models for single point sources and multiple point sources were established,and the solution process was provided.The related principles and research steps of Bayesian Markov Monte Carlo framework in the probability statistics method in the sudden water pollution problem are sorted out in detail,and its core idea is to use Bayesian theorem to transform the pollution source identification problem into a sampling problem of posterior probability density function with unknown parameters.A brief overview of common optimization methods,a detailed explanation of the principle and process steps of differential evolution algorithm,and an analysis of the main factors affecting algorithm performance.2.Traceability issues for single source water pollution accidents.In order to reduce the impact of low initial point selection and acceptance rate of the MCMC(Markov chain Monte Carlo)algorithm,a two-stage MH(Metropolis Hastings)algorithm based on equidistant random sampling method is proposed for tracing single point source pollution accidents in rivers.The use of equidistant random sampling method to adjust the selection of initial points in the algorithm improves the rationality of the selection method;In response to the low acceptance rate of the MH algorithm,a two-stage processing idea was introduced.By constructing an approximate distribution of the objective function and performing segmented calculations,the overall acceptance rate of the algorithm was improved.The simulation case study shows that the acceptance rate of the proposed algorithm is about twice that of the MH algorithm,and it exhibits high traceability accuracy and stability in the solution results of unknown parameters.3.Traceability issues for complex multi-point source water pollution accidents.In order to improve the traceability accuracy of unknown parameters from uncontaminated sources,a multi-point source water pollution accident traceability algorithm based on adaptive differential evolution algorithm is proposed.Firstly,based on the diffusion law of pollutants,a multi-point source pollutant diffusion model is constructed,and the traceability problem is transformed into an optimization problem using the difference between the theoretical concentration value and the observed concentration value.Then,by adaptively adjusting the mutation factor F and crossover probability Cr in the algorithm,the algorithm can converge to the vicinity of the optimal solution faster.Finally,combined with the constructed simulation example,the applicability of the fixed stagger differential evolution algorithm and the adaptive differential evolution algorithm in the problem of multi-point source water pollution was compared and analyzed.The research has shown that the proposed adaptive differential evolution algorithm tends to solve unknown parameters near the true value after approximately 600 iterations.The experimental results show that the proposed algorithm can accurately and efficiently achieve pollutant traceability,and has certain reference value.