Underground water pollution source identification system based on multi-objective optimization algorithm

Abstract

The invention provides an underground water pollution source identification system based on a multi-objective optimization algorithm. According to the system, the pollution source identification precision and efficiency are remarkably improved, and firstly, pollution source data in underground water are collected in real time through a distributed sensor network; then, predicting the dynamic change of the pollution source by using a deep learning model, and optimizing a plurality of targets of pollution source position, concentration change, recognition efficiency, resource consumption and environmental influence in combination with a multi-target optimization algorithm; and finally, the system combines historical data and real-time data to carry out environmental influence evaluation on the pollution source and generates a targeted pollution source treatment strategy. Through the multi-objective optimization method, the underground water pollution source identification precision, efficiency, resource utilization efficiency and environmental friendliness are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of groundwater pollution source identification, and specifically to a groundwater pollution source identification system based on a multi-objective optimization algorithm. Background Technique

[0002] With the increasing requirements for environmental protection, the identification of groundwater pollution sources has become one of the key issues in the field of environmental monitoring. Traditional groundwater pollution source identification methods mostly rely on static sampling and manual analysis. This method is not only time-consuming and laborious, but also often unable to predict the diffusion and change trend of pollution sources in a timely and effective manner due to the dynamic changes of pollution sources. Therefore, how to improve the accuracy and real-time performance of pollution source identification has become the research focus in this field.

[0003] In recent years, with the development of sensor technology and data acquisition systems, distributed sensor networks have been widely used in groundwater pollution monitoring. These sensors can collect relevant data of pollution sources in real time, such as pollutant concentration, location and environmental factors. However, how to efficiently process and analyze a large amount of real-time data, especially when facing multiple objective optimization problems, is still a challenge.

[0004] Multi-objective optimization algorithms have become effective tools for solving complex environmental problems because they can find a balance among multiple conflicting objectives. In groundwater pollution source identification, applying multi-objective optimization algorithms can consider multiple aspects such as the accuracy of pollution source location, the change of pollution source concentration, calculation efficiency, resource consumption and environmental impact, so as to provide a more comprehensive and efficient solution for pollution source identification. Summary of the Invention

[0005] In order to solve the technical problems mentioned in the current background technique, the present invention proposes a groundwater pollution source identification system based on a multi-objective optimization algorithm.

[0006] For this reason, the technical solution adopted by the present invention is as follows:

[0007] M1. Data acquisition and processing module, including a data acquisition unit and a data processing unit. The data acquisition unit collects pollution source data in groundwater in real time through distributed sensors. The pollution source data includes the location, concentration and environmental factors of the pollution source, and collects the acquisition time and resource consumption of the distributed sensors to form sensor status data. The data processing unit performs cleaning, denoising and normalization processing on the pollution source data, and outputs the sensor status data and the processed pollution source data to the multi-objective optimization module;

[0008] M2. The multi-objective optimization module predicts the change in the source pollution location through the first deep learning model and the change in the source pollution concentration through the second deep learning model based on the processed source pollution data. Combining the sensor status data, it performs source pollution target optimization through the evolutionary algorithm (NSGA-III). The optimization objectives include the accuracy of source pollution location movement, the accuracy of source pollution concentration change assessment, the identification efficiency, the minimization of resource costs, and the minimization of environmental impacts, and finally outputs the multi-objective optimization results.

[0009] M3. The result feedback and evaluation module receives the optimization results output by the multi-objective optimization module, conducts an environmental impact assessment of the source pollution, and generates a source pollution control strategy based on the environmental impact assessment.

[0010] Further, the data acquisition unit collects the source pollution data in groundwater by arranging monitoring points through distributed sensors. The collected source pollution data in groundwater is represented as a time series data set X t , X t ={x1, x2, …, x n} where t is the time at moment t, and x n is the nth source pollution data.

[0011] Further, the data processing unit processes the source pollution, including data cleaning, denoising processing, and normalization processing. The data cleaning removes invalid and abnormal data information, and the abnormal data information includes values that do not conform to physical and chemical laws and abnormal values caused by sensor failures.

[0012] The denoising processing is performed through the Kalman filtering algorithm, and the formula is:

[0013]

[0014] where is the data after denoising; K t is the Kalman gain; H t is the observation matrix;

[0015] The denoised data is normalized through z-score standardization, and the processed source pollution data Z t is output.

[0016] Further, the steps of the source pollution target optimization include data collection, population generation, objective function calculation, non-dominated sorting, crossover and mutation operations, and weight adjustment.

[0017] The data collection includes sensor status data and processed pollution source data obtained from the data acquisition and processing module, as well as pollution source location change data predicted by the first deep learning model and pollution source concentration change data predicted by the second deep learning model.

[0018] Based on the data collection, the population generation generates an initial population P0, and each individual in the population is a potential pollution source identification solution.

[0019] Furthermore, the objective function calculation calculates the corresponding identification solutions for each optimization objective. The identification solutions are generated from the sensor status data and the processed pollution source data. Each identification solution represents a solution to an optimization objective. Through the objective function calculation, the quality of each identification solution is evaluated.

[0020] The identification solution for the accuracy of pollution source location movement is calculated based on the difference from the actual pollution source location. The formula is:

[0021]

[0022] where f1(x) is the objective function of the identification solution for the accuracy of pollution source location movement; x i (t) is the actual pollution source location; is the predicted pollution source location; m is the number of samples; is the distance between the actual location and the predicted location;

[0023] The identification solution for the accuracy of pollution source concentration change assessment is calculated based on the difference between the predicted pollution source concentration and the actual concentration. The formula is:

[0024]

[0025] where f2(x) is the objective function of the identification solution for the accuracy of pollution source concentration change assessment; C i (t) is the actual pollution source concentration; is the predicted pollution source concentration;

[0026] The identification solution for the identification efficiency is based on the acquisition time of the distributed sensors, and the optimization objective is to minimize the time. The formula is:

[0027] f3(x) = T(x)

[0028] where f3(x) is the objective function of the identification solution for the identification efficiency; T(x) is the acquisition time;

[0029] The identification solution for minimizing the resource cost is to minimize the consumption of acquisition resources. The formula is:

[0030] f4(x) = C total(x)

[0031] Among them, f4(x) is the identification solution objective function for minimizing resource costs; C total (x) is the total resource cost;

[0032] The identification solution for minimizing environmental impact is to minimize the impact of environmental factors, and the formula is:

[0033]

[0034] Among them, f5(x) is the identification solution objective function for minimizing environmental impact; I i is the impact assessment value of the environmental factor of the i-th pollution source; m is the number of pollution sources.

[0035] Furthermore, the non-dominated sorting sorts all solutions according to the calculation results of the identification solution objective functions of the respective optimization objectives to obtain the Pareto front solutions.

[0036] The crossover and mutation operations are used to generate new identification solutions. By simulating the gene recombination process in the genetic algorithm, the information of the parent solutions (the previous identification solutions) is combined to generate new offspring solutions (the next identification solutions). In the optimization of the pollution source objectives, the parent solutions are an identification solution. In the crossover operation, the advantageous information in the parent solutions is combined to form a new identification solution, that is, the offspring solution.

[0037] The weight adjustment adjusts the weights of the objectives according to the new identification solutions. According to the weight adjustment of the objectives, the objective function calculations are repeated until the iteration conditions are met, and the multi-objective optimization results of the pollution source are output, that is, the balance of the accuracy of the pollution source position movement, the accuracy of the assessment of the pollution source concentration change, the identification efficiency, the minimization of resource costs, and the minimization of environmental impact.

[0038] Furthermore, the environmental impact assessment calculates the diffusion path of pollutants and the influence range of the pollution source based on the predicted movement of the pollution source position and the change in concentration according to the multi-objective optimization results. The formula is:

[0039]

[0040] Among them, is the pollutant concentration change rate; D is the diffusion coefficient; v is the groundwater flow velocity; is the gradient of the concentration; According to the diffusion path and the multi-objective optimization results, the impact of the pollution source on the groundwater quality is evaluated.

[0041] Through the environmental impact assessment, pollution source treatment strategies and plans are generated, including physical remediation measures, chemical remediation measures, biological remediation measures, and source control measures.

[0042] Compared with the prior art, the advantages of the present invention are as follows:

[0043] 1. Multi-objective optimization ability: The present invention adopts an advanced multi-objective optimization algorithm, which can find the optimal balance among multiple objectives. By comprehensively considering various factors such as the accuracy of the pollution source location, the accuracy of the pollution source concentration prediction, the calculation efficiency, the resource consumption, and the environmental impact, the present invention ensures that the best compromise can be achieved between accuracy and efficiency during the pollution source identification process, thereby improving the overall performance of the system.

[0044] 2. Efficient resource utilization and low cost: Through the dynamic weight adjustment of the optimization objectives, the present invention not only improves the accuracy of pollution source identification but also optimizes the use of system resources, minimizing the consumption of computing resources and operating costs to the greatest extent. Especially in resource-constrained environments, the present invention can achieve low-cost and high-performance pollution source identification and avoid the problems brought by high resource consumption and long-time calculation in traditional methods.

[0045] 3. Minimization of environmental impact: The present invention designs an objective function for minimizing environmental impact, optimizes the frequency of data collection and the efficiency of algorithm operation, reduces the interference to the groundwater environment, and ensures that the interference to the environment during the pollution source treatment process is minimized. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is the flow chart of the groundwater pollution source identification system of the present invention;

[0048] Figure 2 It is the flow chart of the data collection and processing module of the present invention;

[0049] Figure 3 It is the flow chart of the multi-objective optimization module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] To achieve the above objectives, the present invention is implemented through the following technical solutions. The present invention provides a groundwater pollution source identification system based on a multi-objective optimization algorithm. Please refer to Figures 1-3 , the system includes:

[0051] M1. Data acquisition and processing module, including a data acquisition unit and a data processing unit. The data acquisition unit collects groundwater pollution source data in real time through distributed sensors. The pollution source data includes the location, concentration, and environmental factors of the pollution source, and collects the acquisition time and resource consumption of the distributed sensors to form sensor status data. The data processing unit cleans, denoises, and normalizes the pollution source data, and outputs the sensor status data and the processed pollution source data to the multi-objective optimization module.

[0052] The data acquisition unit arranges multiple monitoring points through distributed sensors to collect groundwater pollution source data in real time. The sensors collect data according to a preset time interval, and the collected pollution source data is represented as a time series data set X. t , X t ={x1,x2,…,x n}, where t is the time at moment t, and x n is the nth pollution source data.

[0053] The data processing unit processes the pollution source data, including cleaning, denoising, and normalization. The purpose of data cleaning is to remove invalid and abnormal data information. Abnormal data includes values that do not conform to physical and chemical laws and abnormal values caused by sensor failures.

[0054] The purpose of denoising is to eliminate the noise generated by the sensors during the measurement process. Denoising is performed through the Kalman filter algorithm, and the formula is:

[0055]

[0056] where is the data after denoising; K t is the Kalman gain; H t is the observation matrix.

[0057] The denoised data is normalized through z-score standardization, and the processed pollution source data Z t .

[0058] M2. Multi-objective optimization module. Based on the processed pollution source data, the change of the pollution source location is predicted through the first deep learning model, and the change of the pollution source concentration is predicted through the second deep learning model; combined with the sensor status data, the pollution source target optimization is carried out through the evolutionary algorithm (NSGA-III). The optimization objectives include the accuracy of the pollution source location movement, the accuracy of the pollution source concentration change assessment, the identification efficiency, the minimization of resource costs, and the minimization of environmental impacts. Finally, the multi-objective optimization result is output.

[0059] The goal of this module is to optimize multiple conflicting objective functions simultaneously, including the accuracy of the movement of the pollution source location, the accuracy of the assessment of the change in the pollution source concentration, the identification efficiency, the minimization of resource costs, and the minimization of environmental impacts. To solve multiple objectives simultaneously, this module adopts the NSGA-III evolutionary algorithm and dynamically adjusts the weights of the objective functions to achieve a balance between different objectives, ensuring that the system not only considers accuracy but also efficiency, cost, and environmental impact during the pollution source identification process. The specific technical steps include data collection, population generation, objective function calculation, non-dominated sorting, crossover and mutation operations, and weight adjustment:

[0060] 1) Data collection includes the sensor status data and processed pollution source data obtained from the data acquisition and processing module, and the pollution source location change data predicted by the first deep learning model and the pollution source concentration change data predicted by the second deep learning model;

[0061] 2) Population generation generates an initial population P0 based on the data collection, and each individual in the population is a potential pollution source identification solution;

[0062] 3) Objective function calculation calculates the corresponding identification solutions for each optimization objective. The identification solutions are generated from the sensor status data and processed pollution source data. Each identification solution represents a solution for an optimization objective. Through objective function calculation, the advantages and disadvantages of each identification solution are evaluated.

[0063] The identification solution for the accuracy of the movement of the pollution source location is calculated by the difference from the actual pollution source location, and the formula is:

[0064]

[0065] where f1(x) is the objective function of the identification solution for the accuracy of the movement of the pollution source location; x i (t) is the actual pollution source location; is the predicted pollution source location; n is the number of samples; is the distance between the actual location and the predicted location;

[0066] The identification solution for the accuracy of the assessment of the change in the pollution source concentration is calculated by the difference between the predicted pollution source concentration and the actual concentration, and the formula is:

[0067]

[0068] where f2(x) is the objective function of the identification solution for the accuracy of the assessment of the change in the pollution source concentration; C i (t) is the actual pollution source concentration; is the predicted pollution source concentration;

[0069] The recognition solution for recognition efficiency is optimized based on the acquisition time of distributed sensors, with the optimization objective being to minimize time. The formula is:

[0070] f3(x) = T(x)

[0071] Among them, f3(x) is the recognition solution objective function for recognition efficiency; T(x) is the acquisition time;

[0072] The recognition solution for minimizing resource cost is to minimize the consumption of acquisition resources. The formula is:

[0073] f4(x) = C total (x)

[0074] Among them, f4(x) is the recognition solution objective function for minimizing resource cost; C total (x) is the total resource cost; the objective of this formula is to minimize resource consumption, that is, to reduce resource consumption during operation, which is crucial for low-cost and high-efficiency operation in actual deployment, especially important in resource-constrained environments;

[0075] The recognition solution for minimizing environmental impact is to minimize the impact of environmental factors. The formula is:

[0076]

[0077] Among them, f5(x) is the recognition solution objective function for minimizing environmental impact; I i is the impact assessment value of the environmental factor of the i-th pollution source; m is the number of pollution sources;

[0078] By calculating the objective function, the calculation results of the objective functions of each recognition solution, f1(x), f2(x), f3(x), f4(x), f5(x), are output;

[0079] 4) According to the calculation results of the objective functions of each recognition solution, use the non-dominated sorting method to sort all solutions, and find the Pareto front solutions. Each solution is sorted according to the relative superiority between multiple objectives, representing the best balance among all objectives;

[0080] 5) Generate new solutions through crossover and mutation operations. The crossover operation simulates the gene recombination process in genetic algorithms, combines the information of two parent solutions to generate a new offspring solution. Each crossover operation randomly selects two parent solutions and generates a new solution by exchanging information. The purpose of crossover is to utilize the information complementarity between parent solutions to produce potential high-quality solutions. In the pollution source objective optimization, the parent solution is a recognition solution. In the crossover operation, the advantageous information in the parent solutions is combined to form a new individual, that is, the offspring solution.

[0081] The mutation operation is a process of locally randomly perturbing the individual solution to explore other potential high-quality solutions in the solution space. By making small random changes to part of the solution, the mutation operation can prevent the algorithm from falling into a local optimal solution and enhance the diversity of solutions. In source identification, the mutation operation can generate new solutions by slightly adjusting the source location or concentration, which helps the system explore the optimization space from different perspectives.

[0082] Combining the crossover and mutation operations, the new generation population P1 (i.e., all new identification solutions) is output. It can not only find new possible solutions through the information recombination of the parent solutions but also explore a new solution space through local perturbations. Through this combination, the algorithm can effectively avoid falling into a local optimal solution and accelerate convergence to the global optimal solution.

[0083] 6) According to the identification solutions of the new generation population P1, adjust the weights of the objectives. According to the adjusted objective weights, repeat the calculation of the objective function until the iteration conditions are met, and obtain the final Pareto front solution set, that is, the optimized source identification solutions, which achieve an optimal balance among multiple objectives.

[0084] In the multi-objective optimization process, the solution space is explored and optimized through an evolutionary algorithm. The objective weights are dynamically adjusted during the optimization process, and finally, the multi-objective optimization results of the pollution sources are output. The optimization results include the accuracy of the movement of the pollution source location, the accuracy of the assessment of the change in the pollution source concentration, the identification efficiency, the balance of resource costs and environmental impacts, to ensure optimality in multiple aspects.

[0085] M3, the result feedback and evaluation module, receives the optimization results output by the multi-objective optimization module, conducts an environmental impact assessment of the pollution sources, and generates a pollution source treatment strategy based on the environmental impact assessment.

[0086] The main task of the result feedback and evaluation module is to receive the optimization results from the multi-objective optimization module and conduct an environmental impact assessment of the pollution sources.

[0087] The optimization results are the comprehensive results of the five optimization objectives of the multi-objective optimization module, reflecting the balanced optimization results of five objectives, namely, the accuracy of the change in the pollution source location, the accuracy of the assessment of the change in the pollution source concentration, the identification efficiency, the minimization of resource costs, and the minimization of environmental impacts, representing the optimized pollution source location and concentration and the overall optimization performance.

[0088] The environmental impact assessment includes the comprehensive impact of the multi-objective optimization results and the pollution source diffusion path.

[0089] In the comprehensive impact of the optimization results, the positions and concentrations of pollution sources at different time points can be accurately predicted. Based on this optimization result, the system can provide accurate pollution source diffusion paths and changes in pollutant concentrations. The optimized identification efficiency ensures that when the pollution source changes, the system can update the evaluation results in a timely manner, reduce delays, and quickly reflect the impact of pollution source diffusion on groundwater and the environment. The optimization result of minimizing resource costs ensures that during the pollution source assessment process, the system can efficiently utilize computing resources, reduce unnecessary storage and bandwidth consumption, and avoid delays in evaluation results or untimely processing caused by excessive resource consumption. The optimization result of minimizing environmental impact ensures that the pollution source assessment process minimizes interference with the environment, including optimizing the frequency of data collection, selecting remediation technologies, etc., to achieve minimal environmental interference, especially during the pollution source monitoring and treatment process.

[0090] The pollution source diffusion path is based on the predicted movement of the pollution source location and changes in concentration, and calculates the diffusion path of pollutants and the influence range of the pollution source. The formula is:

[0091]

[0092] Where, is the rate of change of pollutant concentration; D is the diffusion coefficient; v is the groundwater flow velocity; is the gradient of concentration; Based on the diffusion path and multi-objective optimization results, evaluate the impact of the pollution source on the groundwater quality.

[0093] According to the environmental impact assessment, generate a pollution source treatment strategy. Based on the pollution source diffusion path, changes in pollution source concentration, and environmental impact assessment results, determine the priority areas for pollution source treatment. The priority treatment areas are those with higher pollution concentrations, faster pollution source diffusion speeds, and greater environmental impacts.

[0094] Based on the multi-objective optimization results and environmental impact assessment of the pollution source, provide suitable pollution source treatment technologies and solutions, including physical remediation measures, chemical remediation measures, biological remediation measures, and source control measures.

[0095] The present invention proposes a groundwater pollution source identification system based on a multi-objective optimization algorithm, aiming to solve problems such as accuracy, efficiency, resource consumption, and environmental impact existing in the prior art during the pollution source identification process through the combination of real-time data collection, deep learning model prediction, and multi-objective optimization algorithm. The system collects relevant data of groundwater pollution sources in real time through a distributed sensor network, and uses an evolutionary algorithm (NSGA-III) to dynamically optimize multiple objectives such as pollution source location, concentration change, identification efficiency, resource cost, and environmental impact, thereby improving the overall performance of pollution source identification.

[0096] In summary, the advantages of the present invention are as follows: on the one hand, through the multi-objective optimization algorithm, the accuracy, efficiency, cost and environmental impact of pollution source identification are effectively balanced, significantly improving the identification accuracy and response speed; on the other hand, the present invention not only optimizes resource consumption, reduces the system operation cost, but also reduces the interference to the environment, ensuring more sustainability and environmental friendliness in the process of pollution source identification. These innovations make the present invention have important application prospects and practical value in the field of groundwater pollution source identification and treatment.

[0097] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A groundwater pollution source identification system based on a multi-objective optimization algorithm, characterized in that The system includes: M1, a data acquisition and processing module, including a data acquisition unit and a data processing unit. The data acquisition unit collects groundwater pollution source data in real time through distributed sensors. The pollution source data includes pollution source location, concentration, and environmental factors, and collects the acquisition time and resource consumption of the distributed sensors to form sensor status data. The data processing unit performs cleaning, denoising, and normalization processing on the pollution source data, and outputs the sensor status data and the processed pollution source data to a multi-objective optimization module; M2, a multi-objective optimization module, based on the processed pollution source data, predicts the change of pollution source location through a first deep learning model, and predicts the change of pollution source concentration through a second deep learning model; combining the sensor status data, performs pollution source target optimization through an evolutionary algorithm (NSGA-III). The optimization objectives include the accuracy of pollution source location movement, the accuracy of pollution source concentration change assessment, identification efficiency, minimization of resource cost, and minimization of environmental impact, and finally outputs a multi-objective optimization result; M3, a result feedback and evaluation module, receives the optimization result output by the multi-objective optimization module, conducts an environmental impact assessment of the pollution source, and generates a pollution source treatment strategy according to the environmental impact assessment.

2. The groundwater pollution source identification system based on the multi-objective optimization algorithm according to claim 1, wherein The data acquisition unit collects the data of pollution sources in groundwater by arranging monitoring points through distributed sensors, and the collected data of pollution sources in groundwater is represented as a time series data set X t , X t = {x1, x2, …, x n}, where t is the time at moment t, and x n is the nth pollution source data.

3. The groundwater pollution source identification system based on a multi-objective optimization algorithm according to claim 2, characterized in that The data processing unit processes the pollution source, including data cleaning, denoising processing, and normalization processing. The data cleaning removes invalid and abnormal data information. The abnormal data information includes values that do not conform to physical and chemical laws and abnormal values caused by sensor failures. The denoising processing is performed by the Kalman filtering algorithm, and the formula is: Among them, is the denoised data; K t is the Kalman gain; H t is the observation matrix; Normalize the denoised data through z-score normalization and output the processed pollution source data Z t .

4. The groundwater pollution source identification system based on a multi-objective optimization algorithm according to claim 3, wherein The steps of the pollution source target optimization include data collection, population generation, objective function calculation, non-dominated sorting, crossover and mutation operations, and weight adjustment. The data collection includes sensor status data and processed pollution source data obtained from the data acquisition and processing module, and pollution source location change data predicted by the first deep learning model and pollution source concentration change data predicted by the second deep learning model. The population generation generates an initial population P0 according to the data collection. Each individual in the population is a potential pollution source identification solution.

5. The groundwater pollution source identification system based on a multi-objective optimization algorithm according to claim 4, characterized in that, The objective function calculation is to calculate the corresponding identification solutions for each optimization objective. The identification solutions are generated from the sensor status data and the processed pollution source data. Each identification solution represents a solution for an optimization objective. Through the objective function calculation, the pros and cons of each identification solution are evaluated. The identification solution for the accuracy of pollution source location movement is calculated by the difference from the actual pollution source location, and the formula is: Among them, f1(x) is the recognition solution objective function for the accuracy of the pollution source location movement; x i (t) is the true pollution source location; is the predicted pollution source location; m is the number of samples; is the distance between the true location and the predicted location; The identification solution for the accuracy of pollution source concentration change assessment is calculated by the difference between the predicted pollution source concentration and the actual concentration, and the formula is: Among them, f2(x) is the identification solution objective function for evaluating the accuracy of the change in the pollution source concentration; C i (t) is the actual pollution source concentration; C i (t) is the predicted pollution source concentration; The identification solution for the identification efficiency is based on the acquisition time of the distributed sensors, and the optimization objective is to minimize the time, and the formula is: f3(x) = T(x) where f3(x) is the identification solution objective function for identification efficiency; T(x) is the acquisition time; The identified solution for minimizing resource costs is to minimize the consumption of acquisition resources, and the formula is: f4(x) = C total (x) Among them, f4(x) is the recognition solution objective function for minimizing resource costs; C total (x) is the total resource cost; The identified solution for minimizing environmental impacts is to minimize the impacts of environmental factors, and the formula is: Among them, f5(x) is the recognition solution objective function for minimizing environmental impacts; I i is the impact assessment value of the environmental factor of the i-th pollution source; m is the number of pollution sources.

6. A groundwater pollution source identification system based on a multi-objective optimization algorithm according to claim 5, characterized in that, The non-dominated sorting ranks all the identified solutions according to the calculation results of the objective functions of the identified solutions for each optimization objective to obtain the Pareto front solutions. The crossover and mutation operations are used to generate new identified solutions. By simulating the gene recombination process in the genetic algorithm, the information of the parent solutions (the previous identified solutions) is combined to generate new offspring solutions (the next identified solutions). In the optimization of the pollution source target, the parent solution is an identified solution. In the crossover operation, the advantageous information in the parent solutions is combined to form a new identified solution, that is, the offspring solution. The weight adjustment adjusts the weights of the objectives according to the new identified solutions. According to the weight adjustment of the objectives, the calculation of the objective functions is repeated until the iteration conditions are met, and the multi-objective optimization results of the pollution source are output, that is, the balance of the accuracy of the pollution source location movement, the accuracy of the assessment of the pollution source concentration change, the identification efficiency, the minimization of resource costs, and the minimization of environmental impacts.

7. A groundwater pollution source identification system based on a multi-objective optimization algorithm according to claim 6, characterized in that The environmental impact assessment calculates the diffusion path of pollutants and the influence range of the pollution source based on the predicted movement of the pollution source location and the change in concentration according to the multi-objective optimization results. The formula is: Among them, is the change rate of pollutant concentration; D is the diffusion coefficient; v is the groundwater flow velocity; is the concentration gradient; according to the diffusion path and the multi-objective optimization result, the impact of the pollution source on the groundwater quality is evaluated. Through the environmental impact assessment, pollution source control strategies and plans are generated, including physical remediation measures, chemical remediation measures, biological remediation measures, and source control measures.

Images