A cyclic well optimization design method, computer equipment and medium

Abstract

The application discloses a circulating well optimization design method, computer equipment and medium, and relates to the technical field of water pollution remediation, and the method comprises the following steps: obtaining hydrogeological parameters of a target site; a plurality of decision variable arrays are randomly generated; the decision variable array comprises a circulating well structure parameter and a circulating well operation parameter; a plurality of target arrays are respectively input into a trained circulating well operation effect representation parameter prediction model to obtain a plurality of circulating well operation effect representation parameters; and the genetic algorithm is adopted based on the circulating well operation effect representation parameters to optimize individuals to obtain an optimal solution of the decision variable array. The application can simultaneously optimize a plurality of circulating well parameters, and the circulating well parameters are obtained through the trained circulating well operation effect representation parameter prediction model, which is simpler, does not need to model the target site circulating well, and reduces the time cost.

Description

Technical Field

[0001] This invention relates to the field of water pollution remediation technology, and in particular to a method for optimizing the design of circulating wells, computer equipment, and media. Background Technology

[0002] Due to the impact of human activities such as urbanization, industrialization, and agricultural development, groundwater pollution has become a major environmental problem. Among existing groundwater remediation technologies, circulating well technology, as an in-situ remediation technique, has advantages such as minimal disturbance to the underground environment, low energy consumption, and low remediation costs. It can also be coupled with other remediation technologies such as multiphase extraction, chemical oxidation, and electrochemical processes, and has been used in practice abroad for many years. Circulating well technology pumps groundwater into the well through a screen pipe, while simultaneously injecting remediated clean water into the aquifer through another screen pipe. This promotes three-dimensional circulation of groundwater around the circulating well, creating horizontal and vertical groundwater flow fields within the aquifer, improving the removal efficiency of pollutants extracted with the water flow. At the same time, the groundwater circulation process increases the oxygen content of the underground environment, which is conducive to the aerobic degradation of organic pollutants.

[0003] Currently, there are no mature actual site remediation cases of circulating well technology. The main limiting factors are: (1) Circulating well technology is subject to strict hydrogeological conditions. The unreasonable design of circulating well process parameters will cause groundwater to "short-flow" near the circulating well, that is, the groundwater only circulates near the well, the radius of influence is small, resulting in a longer circulation cycle and increased remediation costs. (2) The three-dimensional flow field formed by a single well of a traditional double-screen circulating well has a remediation blind zone for pollutants accumulated at the top and bottom of the aquifer, especially for non-aqueous phase liquids (NAPLs) that are insoluble in water, which can easily lead to low remediation efficiency and poor remediation effect. (3) If the capture capacity of a single well is insufficient to capture all the pollution plume, it will lead to the expansion of the pollution plume range. Based on this, in order to further improve the application scope of circulating well technology, a circulating well group remediation system using multiple circulating wells can be considered. For example, multiple circulating wells can be arranged parallel or perpendicular to the direction of groundwater flow. By adjusting the hydrodynamic control parameters of the circulating well group, including the pumping and injection flow rate and pumping and injection mode, pollutants enriched in the remediation blind area can be effectively captured.

[0004] After clarifying the hydrogeological conditions of the site, optimizing the structural parameters and key process parameters of the circulating well is a prerequisite for the rational application of circulating well technology. Compared with the design of a single circulating well, the process design parameters of a multi-circulation well also need to consider issues such as well spacing, making the situation more complex. Traditional circulating well optimization design methods include physical testing and numerical simulation to simulate the circulating well structure under specific hydrogeological conditions, determine whether its operation meets the requirements, and if not, continuously adjust the circulating well structural parameters for optimization. The above-mentioned traditional circulating well design methods have the following disadvantages: (1) The simulation time is long. Each time the parameters are optimized, the circulating well structure needs to be remodeled and simulated again, resulting in a large overall time cost; (2) The optimization process has limitations. Many parameters need to be adjusted, but it is difficult to optimize multiple parameters at the same time. If only one parameter is optimized while other parameters are fixed, the optimization result is prone to getting stuck in a local optimum. Summary of the Invention

[0005] The purpose of this invention is to provide a circulating well optimization design method, computer equipment, and medium that can simultaneously optimize multiple circulating well parameters and obtain circulating well parameters through a trained circulating well operation performance characterization parameter prediction model. This method is simpler, eliminates the need to model circulating wells at the target site, and reduces time costs.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A circulating well optimization design method, the method comprising:

[0008] Obtain the hydrogeological parameters of the target site;

[0009] Multiple decision variable arrays are randomly generated; the decision variable arrays include circulating well structural parameters and circulating well operating parameters; the circulating well structural parameters include the distance between the top of the injection screen of each circulating well and the top plate of the aquifer, the distance between the top of the pumping screen of each circulating well and the top plate of the aquifer, the length of the injection screen of each circulating well, and the length of the pumping screen of each circulating well; the circulating well operating parameters include the pumping and injection rate of each circulating well and the well spacing between adjacent circulating wells;

[0010] Multiple target arrays are input into a pre-trained circulating well performance characterization parameter prediction model to obtain multiple circulating well performance characterization parameters. Each target array includes the hydrogeological parameters and a decision variable array. The circulating well performance characterization parameters include circulation efficiency and pollutant removal efficiency. The pre-trained circulating well performance characterization parameter prediction model is a model trained using sample target arrays as input and sample circulating well performance characterization parameters as labels. The sample target arrays include sample hydrogeological parameters, sample circulating well structural parameters, and sample circulating well operational parameters.

[0011] Based on the performance characteristics of the circulating wells, a genetic algorithm is used to optimize the individuals to obtain the optimal solution of the decision variable array; each individual is a decision variable array; the optimal solution of the decision variables is used for the repair of the circulating wells in the target site.

[0012] Optionally, the genetic algorithm is the NSGA-II algorithm.

[0013] Optionally, the optimization of individuals based on the circulation well operation performance characterization parameters using a genetic algorithm to obtain the optimal solution for the decision variable array specifically includes:

[0014] Initialize the population parameters; the population parameters include the population size N, the maximum number of iterations, the crossover probability, and the mutation probability;

[0015] The offspring population Q generated in generation t t With parent population P t Merging to form population R t The population R t The size is 2N; if it is the first generation population, then the first generation population is taken as the population R. t ;

[0016] For the population R t Perform a non-dominated sort to obtain the non-dominated set;

[0017] Calculate the crowding degree of each individual in the non-dominated set, and obtain the new parent population P based on the crowding degree. t+1 The congestion level is determined by the parameters characterizing the operating performance of the circulating well.

[0018] For the parent population P t+1 A new offspring population Q is obtained by performing crossover and mutation operations. t+1 ;

[0019] Determine whether the current iteration number has reached the maximum iteration number. If yes, take the value of the decision variable corresponding to the current iteration number as the optimal solution for the decision variable; otherwise, proceed to step "take the offspring population Q generated in generation t".t With parent population P t Merging to form population R t "until the maximum number of iterations is reached."

[0020] Optionally, before inputting the multiple target arrays into the trained circulating well operation performance characterization parameter prediction model, the method further includes: training the circulating well operation performance characterization parameter prediction model, specifically including:

[0021] Obtain the dataset; the dataset includes several sample target arrays and sample loop well performance characterization parameters corresponding to each sample target array;

[0022] The dataset was used to train the prediction model of the characterization parameters of the circulation well operation effect, and the trained prediction model of the characterization parameters of the circulation well operation effect was obtained.

[0023] Optionally, obtaining the dataset specifically includes:

[0024] Establish groundwater flow simulation models and solute transport simulation models;

[0025] Under several sample hydrogeological parameters, numerical simulations are performed based on the groundwater flow simulation model and the solute transport simulation model to obtain the sample circulation well structural parameters, sample circulation well operation parameters, and sample circulation well operation effect characterization parameters corresponding to each sample hydrogeological parameter; the sample hydrogeological parameters, the sample circulation well structural parameters, and the sample circulation well operation parameters constitute the sample target array.

[0026] Optionally, before training the prediction model for the characterization parameters of the circulating well operation using the dataset, the method further includes:

[0027] The data in the dataset is normalized.

[0028] Optionally, the prediction model for the characterization parameters of the circulating well operation effect includes a first convolutional layer, a second convolutional layer, a first maximum pooling layer, a third convolutional layer, a second maximum pooling layer, a fourth convolutional layer, a third maximum pooling layer, a flattening layer, a first fully connected layer, a second fully connected layer, and a third fully connected layer connected in sequence.

[0029] Optionally, the hydrogeological parameters include aquifer thickness, hydraulic gradient, horizontal permeability coefficient, horizontal permeability coefficient / vertical permeability coefficient, specific yield, elastic release coefficient, and porosity.

[0030] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the above-described circulating well optimization design method.

[0031] The present invention also provides a computer-readable storage medium storing a computer program adapted to be loaded by a processor and executed by the above-described circulating well optimization design method.

[0032] According to specific embodiments provided by the present invention, the following technical effects are disclosed: The present invention provides a circulating well optimization design method, computer equipment, and medium. The method includes: acquiring hydrogeological parameters of the target site; randomly generating multiple decision variable arrays; the decision variable arrays include circulating well structural parameters and circulating well operating parameters; inputting the multiple target arrays into a trained circulating well operating effect characterization parameter prediction model to obtain multiple circulating well operating effect characterization parameters; each target array includes hydrogeological parameters and a decision variable array; optimizing individuals based on the circulating well operating effect characterization parameters using a genetic algorithm to obtain the optimal solution of the decision variable array; each individual is a decision variable array; the optimal solution of the decision variables is used for the repair of circulating wells at the target site. The present invention can simultaneously optimize multiple circulating well parameters (circulating well structural parameters and circulating well operating parameters), and obtains circulating well parameters through a trained circulating well operating effect characterization parameter prediction model, which is simpler, eliminates the need to model circulating wells at the target site, and reduces time costs. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a schematic diagram of the circulation well optimization design method provided in an embodiment of the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the specific implementation process of the circulating well optimization design method provided in this embodiment of the invention;

[0036] Figure 3 This is a schematic diagram of the predictive model structure for the characterization parameters of circulating well operation provided in an embodiment of the present invention;

[0037] Figure 4 This is a schematic diagram of the NSGA-II algorithm's method for optimizing the decision variable array, provided in an embodiment of the present invention.

[0038] Figure 5 This is a schematic diagram of the structure of a computer device provided by the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] The purpose of this invention is to provide a circulating well optimization design method, computer equipment, and medium. Through a genetic algorithm, multiple circulating well parameters (circulating well structural parameters and circulating well operating parameters) can be optimized simultaneously. Furthermore, the circulating well parameters can be obtained through a trained circulating well operating effect characterization parameter prediction model, which is simpler and does not require modeling of the circulating well at the target site, thus reducing time costs.

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] like Figure 1 and Figure 2 As shown, the present invention provides a circulating well optimization design method, the method comprising:

[0043] S1: Obtain the hydrogeological parameters of the target site.

[0044] S2: Randomly generate multiple decision variable arrays; the decision variable arrays include circulating well structural parameters and circulating well operating parameters; the circulating well structural parameters include the distance between the top of the injection screen of each circulating well and the top plate of the aquifer, the distance between the top of the pumping screen of each circulating well and the top plate of the aquifer, the length of the injection screen of each circulating well, and the length of the pumping screen of each circulating well; the circulating well operating parameters include the pumping and injection volume of each circulating well and the well spacing between adjacent circulating wells.

[0045] S3: Input multiple target arrays into the trained circulating well operation performance characterization parameter prediction model to obtain multiple circulating well operation performance characterization parameters; each target array includes the hydrogeological parameters and a decision variable array; the circulating well operation performance characterization parameters include circulation efficiency and pollutant removal efficiency; the trained circulating well operation performance characterization parameter prediction model is a model trained with sample target arrays as input and sample circulating well operation performance characterization parameters as labels; the sample target array includes sample hydrogeological parameters, sample circulating well structural parameters, and sample circulating well operation parameters.

[0046] S4: Based on the performance characteristics parameters of the circulating wells, a genetic algorithm is used to optimize the individuals to obtain the optimal solution of the decision variable array; each individual is a decision variable array; the optimal solution of the decision variables is used for the repair of the circulating wells in the target site.

[0047] In this embodiment, the hydrogeological parameters include aquifer thickness (M), hydraulic gradient (I), and horizontal permeability coefficient (K). H ), horizontal permeability coefficient / vertical permeability coefficient (i.e., the ratio K of horizontal permeability coefficient to vertical permeability coefficient) H / K V ), water yield (μ), elastic water release coefficient (ss), and porosity (n).

[0048] The structural parameters of the circulating well (i=1, 2 in this embodiment) include the distance (d) between the top of the injection screen of the i-th circulating well and the top plate of the aquifer. i in The distance (d) between the top of the pumping screen of the i-th circulating well and the top plate of the aquifer. i out The length L of the injection screen in the i-th circulating well i in and the length L of the pumping screen of the i-th circulating well. i out .

[0049] The operating parameters of the circulating well include the pumping and injection rate (Q) of the i-th circulating well. i ) and the well spacing (d) between adjacent circulating wells.

[0050] Circulation well operation performance parameters include circulation efficiency ( ) and pollutant removal efficiency ( ).

[0051] In this embodiment, the prediction model for the characterization parameters of the circulating well operation effect can employ a convolutional neural network. The prediction model includes, in sequence, a first convolutional layer, a second convolutional layer, a first max-pooling layer, a third convolutional layer, a second max-pooling layer, a fourth convolutional layer, a third max-pooling layer, a flattened layer, a first fully connected layer, a second fully connected layer, and a third fully connected layer. The structure of the prediction model for the characterization parameters of the circulating well operation effect is as follows: Figure 3 As shown. Zero-padding is performed in the same mode, the activation function is ReLU, and the three pooling windows are all (2, 1). The number of neurons in the three fully connected layers are 16, 16, and 2, respectively. The activation functions of the first and second fully connected layers are both ReLU, and the activation function of the third fully connected layer is a linear function. The optimizer is in Adam mode.

[0052] Before inputting the multiple target arrays into the trained circulating well operation performance characterization parameter prediction model, this embodiment further includes: training the circulating well operation performance characterization parameter prediction model, specifically including:

[0053] Obtain the dataset (simulation dataset); the dataset includes several sample target arrays and the sample loop well operation performance characterization parameters corresponding to each sample target array.

[0054] The dataset was used to train the prediction model of the characterization parameters of the circulation well operation effect, and the trained prediction model of the characterization parameters of the circulation well operation effect was obtained.

[0055] Specifically, obtaining the dataset includes:

[0056] Establish groundwater flow simulation models and solute transport simulation models.

[0057] Under several sample hydrogeological parameters, numerical simulations are performed based on the groundwater flow simulation model and the solute transport simulation model to obtain the sample circulation well structural parameters, sample circulation well operation parameters, and sample circulation well operation effect characterization parameters corresponding to each sample hydrogeological parameter; the sample hydrogeological parameters, the sample circulation well structural parameters, and the sample circulation well operation parameters constitute the sample target array.

[0058] After obtaining the dataset through simulation, it is necessary to preprocess the data to avoid discrepancies between the predicted results and expectations caused by low-quality data.

[0059] First, data cleaning is performed. This is because when generating input data, there may be situations where the input data design is unreasonable, resulting in the inability to generate output data or generating unreasonable output data. Such situations are considered as missing data, and the data needs to be found and deleted in the data cleaning step.

[0060] Secondly, before training the prediction model for the characterization parameters of the circulating well operation using the aforementioned dataset, the following steps are also included:

[0061] The data in the dataset is normalized.

[0062] Specifically, the data in the dataset needs to be standardized because differences in magnitude can lead to attributes with larger magnitudes dominating. This embodiment uses a normalization method to process the data. The specific process is as follows: for each attribute, let minA and maxA be the maximum and minimum values ​​of attribute A, respectively. The original value x in A is normalized and mapped to a value x' in the interval [0, 1]. The specific formula is as follows:

[0063] .

[0064] Where x' is the normalized value; x is the value before normalization; min(x) is the minimum value in the data; and max(x) is the maximum value in the data.

[0065] The preprocessed hydrogeological parameters, circulating well structural parameters, and circulating well operating parameters of the samples, along with their corresponding circulating well operating effect characterization parameters, are used as a set of data in the dataset.

[0066] The causes of circulating well operational performance are complex, and relying on simulation models to obtain performance data is too time-consuming. Therefore, a convolutional neural network is constructed to obtain the influencing factors of circulating well operational performance and the response relationship between performance data, thereby establishing an optimized design model for circulating wells. The prediction network is trained using a preprocessed simulation dataset to obtain a high-accuracy prediction model for circulating well operational performance parameters. 80% of the simulation dataset is used as the training set to train the model, and 20% is used as the test set to test the network's actual learning ability.

[0067] The performance and accuracy of the prediction network are measured using a loss function and an evaluation function, respectively. The loss function represents the difference between the true and predicted values. During training, the convolutional neural network continuously updates its model parameters by calculating the loss function; therefore, a smaller loss function is better. The evaluation function represents the degree of matching between the true and predicted values. The convolutional neural network ultimately uses the evaluation function to represent the network's prediction performance. The evaluation function is similar to the loss function, except that its result is not used during training. In this embodiment, the mean squared error (MSE) is used as the loss function, defined as follows:

[0068] .

[0069] In the formula, n represents the number of samples. Represents the true value. This represents the predicted value.

[0070] Through the above process, a well-trained predictive model for the characterization parameters of circulating well operation can be obtained.

[0071] Then, site tests are conducted, and the actual hydrogeological parameters of the target site are determined through site tests and hydrogeological survey data.

[0072] Based on the hydrogeological parameters obtained from the site test, the structural parameters and operational parameters of the circulating well are used as the decision variable array, and the operational effect parameters of the circulating well are used as the optimization objective. The aforementioned prediction network serves as the intrinsic link between the optimization objective and the decision variables, acting as the objective function. A genetic algorithm is then used to perform optimization calculations. In this embodiment, the genetic algorithm is the NSGA-II algorithm.

[0073] The genetic algorithm is used to optimize the individual parameters based on the circulating well operation performance characteristics to obtain the optimal solution for the decision variable array, specifically including:

[0074] Initialize the population parameters; the population parameters include the population size N, the maximum number of iterations, the crossover probability, and the mutation probability.

[0075] The offspring population Q generated in generation t t With parent population P t Merging to form population R t The population R t The size is 2N; if it is the first generation population, then the first generation population is taken as the population R. t .

[0076] For the population R t Perform a non-dominated sort to obtain the non-dominated set.

[0077] Calculate the crowding degree of each individual in the non-dominated set, and obtain the new parent population P based on the crowding degree. t+1 The crowding degree is used to measure the density of each solution in the non-dominated set. The purpose is to disperse the obtained Pareto optimal solutions as much as possible, thereby maintaining the diversity of individuals in the population.

[0078] For the parent population P t+1 A new offspring population Q is obtained by performing crossover and mutation operations. t+1 .

[0079] Determine whether the current iteration number has reached the maximum iteration number. If yes, take the value of the decision variable corresponding to the current iteration number as the optimal solution for the decision variable; otherwise, proceed to step "take the offspring population Q generated in generation t". t With parent population P tMerging to form population R t "until the maximum number of iterations is reached."

[0080] Specifically: S501: Set the parameters of the NSGA-II algorithm, including initializing the population size N=100, the maximum number of iterations 200, the crossover probability 0.8, and the mutation probability 0.2 to obtain the parent population P. t , t=1.

[0081] S502: Perform non-dominated sorting on individuals to obtain non-dominated sets, calculate the crowding degree for each individual in each non-dominated set, and then obtain the offspring population Q through selection, crossover, and mutation. t .

[0082] S503: Transform the offspring population Q t With parent population P t Merge to form a population R of size 2N. t The population is sorted non-dominated according to the elite strategy to generate a non-dominated set and the crowding degree is calculated, resulting in a new offspring population Q. t+1 .

[0083] S504: Determine if the maximum number of iterations has been reached. If it is, proceed to S505; otherwise, return to S502, set t = t + 1, and perform iterative calculations.

[0084] S505: Output the optimal solution for the decision variables, and the algorithm terminates.

[0085] The above calculations yield the optimal values ​​of the decision variable array (circulation well structural parameters and circulation well operating parameters), at which point the circulation efficiency ( ) and pollutant removal efficiency ( () Reach its maximum.

[0086] Specifically, the dominance relation in S502 is as follows: For two solutions x and y, if for any i = 1, 2, ..., n, we have... And exist If x dominates y, then x is said to dominate y.

[0087] S503 Congestion Level ( The calculation of the function (i-1) involves sorting individuals within a single frontier according to the values ​​of a function in ascending order, then using the maximum function value as `max` and the minimum function value as `min`. The distance between each individual's nearest neighbor, i.e., the difference between the function values ​​of `i-1` and `i+1`, is calculated using `max` and `min` to normalize the values. The specific formula is as follows:

[0088] .

[0089] In the formula, and Let be the function values ​​of the i-th solution followed by the i-th solution and the i-th solution preceding the i-th solution on the m-function, respectively. and These represent the maximum and minimum values ​​of the m-function, respectively. The m-function is the prediction principle formula of the prediction model for the parameters characterizing the operation effect of circulating wells. Let be the parameters representing the operating effect of the circulating well corresponding to the (i+1)th group of decision variables (the (i+1)th individual). The parameters representing the operating effect of the circulating well are the decision variables of the (i-1)th group (the (i-1)th individual).

[0090] This invention not only solves the problems of long simulation time, need for manual intervention, and easy getting trapped in local optima in traditional circulating well design methods, but also optimizes the design of circulating wells by comprehensively considering the structural parameters of circulating wells and the factors affecting the operating effect of circulating wells under different hydrogeological conditions, thereby improving the operating effect of circulating wells and providing ideas for the layout of multiple wells.

[0091] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the above-described circulating well optimization design method.

[0092] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer device provided in this application. For example... Figure 5 As shown, computer device 1000 may include: a processor 1001, a network interface 1004, and a memory 1005. Furthermore, computer device 1000 may also include: a user interface 1003, and at least one communication bus 1002. The communication bus 1002 is used to implement communication between these components. The user interface 1003 may include a display screen and a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as at least one disk storage device. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 5 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a device control application program.

[0093] exist Figure 5In the computer device 1000 shown, the network interface 1004 provides network communication functions; the user interface 1003 is mainly used to provide an input interface for users; and the processor 1001 can be used to call the device control application stored in the memory 1005 to implement the circulating well optimization design method described in the above embodiments, which will not be described in detail here.

[0094] The present invention also provides a computer-readable storage medium storing a computer program that is adapted to be loaded by a processor and executed by the circulating well optimization design method described in the above embodiments, which will not be described in detail here.

[0095] The above program can be deployed and executed on a single computer device, or deployed and executed on multiple computer devices located in one location, or executed on multiple computer devices distributed across multiple locations and interconnected through a communication network. Multiple computer devices distributed across multiple locations and interconnected through a communication network can form a blockchain network.

[0096] The aforementioned computer-readable storage medium can be an internal storage unit of the computer device, such as a hard drive or memory. It can also be an external storage device, such as a plug-in hard drive, smart media card (SMC), secure digital card (SD), or flash card. Furthermore, the computer-readable storage medium can include both internal and external storage units. This computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. It can also be used to temporarily store data that has been output or will be output.

[0097] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0098] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for optimizing the design of circulating wells, characterized in that, The method includes: Obtain the hydrogeological parameters of the target site; Multiple decision variable arrays are randomly generated; the decision variable arrays include circulating well structural parameters and circulating well operating parameters; the circulating well structural parameters include the distance between the top of the injection screen of each circulating well and the top plate of the aquifer, the distance between the top of the pumping screen of each circulating well and the top plate of the aquifer, the length of the injection screen of each circulating well, and the length of the pumping screen of each circulating well; the circulating well operating parameters include the pumping and injection rate of each circulating well and the well spacing between adjacent circulating wells; Multiple target arrays are input into a pre-trained circulating well performance characterization parameter prediction model to obtain multiple circulating well performance characterization parameters. Each target array includes the hydrogeological parameters and a decision variable array. The circulating well performance characterization parameters include circulation efficiency and pollutant removal efficiency. The pre-trained circulating well performance characterization parameter prediction model is a model trained using sample target arrays as input and sample circulating well performance characterization parameters as labels. The sample target array includes sample hydrogeological parameters, sample circulating well structural parameters, and sample circulating well operating parameters. The circulating well performance characterization parameter prediction model includes a first convolutional layer, a second convolutional layer, a first max-pooling layer, a third convolutional layer, a second max-pooling layer, a fourth convolutional layer, a third max-pooling layer, a flattening layer, a first fully connected layer, a second fully connected layer, and a third fully connected layer, connected in sequence. Based on the aforementioned circulating well operation performance characterization parameters, a genetic algorithm is used to optimize individual wells to obtain the optimal solution for the decision variable array, specifically including: Initialize population parameters; the population parameters include the population size. N Maximum number of iterations, crossover probability, and mutation probability; The first t Offspring population produced by generation Q t with parental population P t Merging to form a population R t The population R t The size is 2 N If it is the first generation population, then the first generation population shall be regarded as the population mentioned above. R t ; For the population R t Perform a non-dominated sort to obtain the non-dominated set; Calculate the crowding degree of each individual in the non-dominated set, and obtain a new parent population based on the crowding degree. P t+1 The congestion level is determined by the parameters characterizing the operating performance of the circulating well. For the parent population P t+1 New offspring populations are obtained by performing crossover and mutation operations. Q t+1 ; Determine whether the current iteration number has reached the maximum iteration number. If yes, take the value of the decision variable corresponding to the current iteration number as the optimal solution for the decision variable; otherwise, proceed to step "to determine the offspring population generated in generation t". Q t with parental population P t Merging to form a population R t The iteration continues until the maximum number of iterations is reached; each individual is an array of decision variables; the optimal solution of the decision variables is used for the repair of the circulating well in the target site.

2. The circulating well optimization design method according to claim 1, characterized in that, The genetic algorithm used is the NSGA-II algorithm.

3. The circulating well optimization design method according to claim 1, characterized in that, Before inputting the multiple target arrays into the trained circulating well operation performance characterization parameter prediction model, the method further includes: training the circulating well operation performance characterization parameter prediction model, specifically including: Obtain the dataset; the dataset includes several sample target arrays and sample loop well performance characterization parameters corresponding to each sample target array; The dataset was used to train the prediction model of the characterization parameters of the circulation well operation effect, and the trained prediction model of the characterization parameters of the circulation well operation effect was obtained.

4. The circulating well optimization design method according to claim 3, characterized in that, The acquisition of the dataset specifically includes: Establish groundwater flow simulation models and solute transport simulation models; Under several sample hydrogeological parameters, numerical simulations are performed based on the groundwater flow simulation model and the solute transport simulation model to obtain the sample circulation well structural parameters, sample circulation well operation parameters, and sample circulation well operation effect characterization parameters corresponding to each sample hydrogeological parameter; the sample hydrogeological parameters, the sample circulation well structural parameters, and the sample circulation well operation parameters constitute the sample target array.

5. The circulating well optimization design method according to claim 3, characterized in that, Before training the prediction model for the characterization parameters of circulating well operation using the aforementioned dataset, the following steps are also included: The data in the dataset is normalized.

6. The circulating well optimization design method according to claim 1, characterized in that, The hydrogeological parameters include aquifer thickness, hydraulic gradient, horizontal permeability coefficient, horizontal permeability coefficient / vertical permeability coefficient, specific yield, elastic release coefficient, and porosity.

7. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method according to any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted to be loaded by a processor and executed as described in any one of claims 1-6.

Images