Prediction method for dissolved oxygen concentration of river section, medium and product

Abstract

The invention discloses a method for predicting the dissolved oxygen concentration of a river section, a medium and a product in the technical field of urban water environment prediction, and the method comprises the steps: obtaining basic data of the river section with the dissolved oxygen concentration to be measured, inputting a dissolved oxygen concentration prediction model, outputting a dissolved oxygen concentration prediction value, and determining the dissolved oxygen concentration prediction model. Comprising the following steps: acquiring historical basic data of a river section, respectively inputting a random forest prediction model, a gradient boosting tree prediction model, an XGBoost prediction model, a BP neural network prediction model and a support vector machine prediction model which are trained in advance, and outputting a comparison scatter diagram of a dissolved oxygen concentration prediction value and a dissolved oxygen concentration true value of each model; and determining a dissolved oxygen concentration prediction model based on the evaluation indexes. According to the method, the problems of low efficiency, insufficient coverage, high cost, insufficient prediction accuracy and poor model adaptability in monitoring and predicting the concentration of the dissolved oxygen in the water body in the prior art are solved.

Description

Technical Field

[0001] The invention relates to a method, medium and product for predicting dissolved oxygen concentration in a river section, belonging to the technical field of urban water environment prediction. Background Art

[0002] With the improvement of people's living standards, the quality of urban water environment has also received much attention. On the one hand, with the advancement of black and odorous water treatment, many urban water bodies have basically eliminated the black and odorous water on sunny days, but this is accompanied by prominent phenomena such as the river returning to black and poor quality during the flood season and abnormal fluctuations in water quality; on the other hand, in recent years, global warming has become an environmental issue that cannot be ignored, and has had a profound impact on natural ecosystems, especially in freshwater ecosystems. With the increase in temperature, the dissolved oxygen concentration in water bodies generally shows a downward trend, posing a serious threat to aquatic organisms and the entire ecosystem. To improve the national water environment quality, it provides strong support for accurately predicting the dissolved oxygen concentration in river sections.

[0003] At present, in the research field of dissolved oxygen concentration in water bodies, although a variety of monitoring methods have been developed, there are still several significant deficiencies and challenges. Manual monitoring is inefficient and has limited coverage; fixed-site monitoring is difficult to meet the needs of large-scale synchronous dynamics; although remote sensing monitoring has advantages, satellite remote sensing is costly and has low resolution, and ground remote sensing is limited in large-scale continuous monitoring. Furthermore, the data currently used in the prediction of dissolved oxygen concentration often have insufficient coverage, so the prediction is not accurate enough. At the same time, a single model has limitations and inadaptability in predicting dissolved oxygen concentration in different regions and different river sections. We need to start from the comprehensiveness and accuracy of monitoring indicators, the scientificity and systematicness of evaluation methods, and the comprehensive application of data, and strengthen relevant research and practical work to better protect and manage the water environment.

[0004] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgement or any form of suggestion that the information constitutes the prior art already known to ordinary technicians in this field. Summary of the invention

[0005] The technical problem to be solved by the present invention is that the existing technology has the problems of low efficiency, insufficient coverage, high cost, insufficient prediction accuracy and poor model adaptability in monitoring and predicting dissolved oxygen concentration in water bodies.

[0006] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solutions:

[0007] In a first aspect, the present invention provides a method for predicting dissolved oxygen concentration in a river section, characterized by comprising:

[0008] Obtaining basic data of the river section where the dissolved oxygen concentration is to be measured, the basic data including water quality data, hydrological data, section land type data, associated pump station data and meteorological data;

[0009] Inputting the basic data of the river section of the to-be-measured dissolved oxygen concentration into a dissolved oxygen concentration prediction model for prediction, and outputting a predicted value of dissolved oxygen concentration;

[0010] Wherein, the method for determining the dissolved oxygen concentration prediction model includes:

[0011] Obtain historical basic data of the river section, input the pre-trained random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model respectively, and output a scatter plot comparing the predicted value of dissolved oxygen concentration of each model with the actual value of dissolved oxygen concentration;

[0012] Based on the evaluation indicators, the prediction accuracy and fitting effect of each prediction model were analyzed according to the comparative scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of the dissolved oxygen concentration, and the dissolved oxygen concentration prediction model was determined.

[0013] Furthermore, the water quality data includes dissolved oxygen concentration values ​​of water quality indicators of river sections;

[0014] The hydrological data include the average water temperature and average flow of the river section;

[0015] The cross-sectional land use type data include the proportion of built-up areas;

[0016] The associated pump station data include the dissolved oxygen concentration value and discharge flow rate of the water quality index of the forebay of the associated pump station at each section under rainfall;

[0017] The meteorological data includes rainfall data.

[0018] Furthermore, the method for acquiring basic data of the river section includes:

[0019] Count the water quality index concentration values ​​at the beginning of each rainfall, including the dissolved oxygen concentration value, and the minimum dissolved oxygen concentration value of the water quality index of the river section within 24 hours from the beginning to the end of the rainfall, to obtain the water quality data;

[0020] The average water temperature and average flow rate of the river section within 24 hours from the beginning to the end of each rainfall event are counted to obtain the hydrological data;

[0021] Count the proportion of built-up areas in riverbank sections to obtain land use type data of the sections;

[0022] Count the rainfall type of each rainfall event and the average water quality concentration of the forebay of the associated pumping station under the corresponding rainfall type, including the dissolved oxygen concentration value and the discharge flow, to obtain the data of the associated pumping station;

[0023] 24 hours is taken as the minimum interval between rainfall events, and the accumulated rainfall events greater than or equal to 3 mm are regarded as rainfall events, and the rainfall amount, rainfall duration, maximum rainfall intensity, average rainfall intensity, and previous drought time of each rainfall event are counted to obtain the meteorological data;

[0024] The water quality data, the hydrological data, the section land type data, the associated pump station data and the meteorological data are integrated into the basic data of the river section in units of rainfall events.

[0025] Furthermore, the historical basic data of the river section is obtained, and the pre-trained random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model are input respectively, and the comparison scatter plot of the predicted value of dissolved oxygen concentration of each model and the actual value of dissolved oxygen concentration is output, including:

[0026] Dividing the historical basic data of the river section into a training set and a test set;

[0027] Use the training set to train the random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model respectively, and use the grid search method to find the optimal parameter configuration of each prediction model;

[0028] According to the prediction results of the test set and the model, a scatter plot comparing the predicted values ​​of dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration is generated.

[0029] Furthermore, the method of finding the optimal parameter configuration of each prediction model by a grid search method includes:

[0030] For random forest prediction models: find the optimal parameter configuration for the number of estimators, the maximum depth of the decision tree, and the minimum number of samples required for a decision tree node to split;

[0031] For the gradient boosted tree prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of the decision tree;

[0032] For XGBoost prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of decision tree;

[0033] For the BP neural network prediction model: find the optimal parameter configuration for the number of hidden layers, the number of neurons in each layer of the neural network, the type of activation function, the type of optimization algorithm, and the number of training rounds;

[0034] For the support vector machine prediction model: find the optimal parameter configuration of the penalty coefficient, kernel function, and hyperparameters that control the influence range of the kernel function.

[0035] Furthermore, based on the evaluation index, according to the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of the dissolved oxygen concentration, the accuracy and fitting effect of the prediction of each prediction model are analyzed to determine the dissolved oxygen concentration prediction model, including:

[0036] According to the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration, the determination coefficient R of each prediction model is calculated. 2 Value and Akaike Information Criterion AIC value;

[0037] According to the determination coefficient R of each prediction model 2 The prediction accuracy, fit and complexity of each prediction model were evaluated by using the Akaike information criterion (AIC) value, and the optimal prediction model was selected from each prediction model as the prediction model for dissolved oxygen concentration.

[0038] Among them, the closer the determination coefficient R² value is to 1, the higher the fit of the prediction model is, and the smaller the Akaike information criterion AIC value is, the more balanced the prediction model is between fitting data and maintaining complexity.

[0039] Furthermore, the determination coefficient R of each prediction model is 2 The value calculation expression is:

[0040] (1);

[0041] In the formula, Indicates The true value of dissolved oxygen concentration of training samples, Indicates The predicted value of dissolved oxygen concentration of training samples, Indicates the average dissolved oxygen concentration.

[0042] Furthermore, the calculation expression of the Akaike Information Criterion AIC value of each prediction model is:

[0043] (2);

[0044] In the formula, represents the number of training samples, represents the logarithmic function, represents the likelihood function.

[0045] In a second aspect, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, wherein the computer program / instruction, when executed by a processor, implements the steps of the method described in the first aspect.

[0046] In a third aspect, the present invention provides a computer program product, comprising a computer program / instruction, characterized in that when the computer program / instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. The present invention predicts dissolved oxygen concentration by using five different prediction models, namely random forest, gradient boosting tree, XGBoost, BP neural network, and support vector machine. Each algorithm can capture different features in the data with its unique advantages, thereby significantly improving the accuracy of the prediction. By comparing the output results of the five prediction models and selecting the best performing model as the final prediction model for dissolved oxygen concentration, the present invention effectively avoids the deviations or limitations that may exist in the prediction of a single algorithm. This multi-algorithm fusion selection mechanism greatly enhances the overall robustness, allowing the prediction model to more comprehensively utilize the information in the data and maintain stable prediction performance when facing complex and changeable water environment data.

[0049] 2. The present invention does not rely on a single data source, but integrates a wide range of data sources, including water quality data, hydrological data, section land use type data, associated pump station data and meteorological data. The comprehensive use of these data can more comprehensively reflect the actual situation of the river section. At the same time, the present invention uses a specific combination of prediction models to determine the optimal dissolved oxygen concentration prediction model. This method of integrating data sources further improves the accuracy and practicality of the prediction. By comprehensively considering multiple factors, the present invention provides more comprehensive and accurate prediction support for the protection and management of the water environment.

[0050] 3. The present invention calculates the determination coefficient R² and the Akaike information criterion AIC value by comparing the comparison scatter plot of the predicted value of dissolved oxygen concentration output by each prediction model with the actual value. The balance between the model's fit and complexity is measured. Based on these evaluation indicators, the prediction accuracy, fit and complexity of each model are comprehensively evaluated at the same time, and finally the optimal prediction model is selected as the dissolved oxygen concentration prediction model. It not only ensures that the selected model has a high degree of prediction accuracy, but also achieves the best fit of the data while maintaining the complexity, providing reliable technical support for the monitoring and management of the water environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1FIG. 1 is a flow chart of a method for predicting dissolved oxygen concentration in a river section provided by an embodiment of the present invention;

[0052] Figure 2 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the random forest prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0053] Figure 3 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the gradient boosting tree prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0054] Figure 4 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the XGBoost prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0055] Figure 5 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the BP neural network prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0056] Figure 6 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the support vector machine model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration. DETAILED DESCRIPTION

[0057] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0058] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.

[0059] Example 1

[0060] like Figure 1 As shown, this embodiment introduces a method for predicting dissolved oxygen concentration in a river section, including:

[0061] Step 1: Obtain basic data of the river section to be studied.

[0062] This example predicts the dissolved oxygen concentration of a river section based on the monitoring data of multiple sections of a river in Nanjing. The hydrological monitoring data and water quality monitoring data of the section from 2021 to 2023 are collected, the rainfall data of the meteorological station closest to the section from 2021 to 2023 are collected, and the water quality data of the pump station associated with the section from 2021 to 2023 are collected. The collected data are processed and organized into a total of 261 data sets, some of which are shown in Table 1 and Table 2:

[0063]

[0064]

[0065] Obtain historical basic data of the river section under study, including water quality data, hydrological data, section land use type data, associated pumping station data and meteorological data.

[0066] The water quality data include the dissolved oxygen concentration value of the water quality indicator of the river section;

[0067] The hydrological data include the average water temperature and average flow of the river section;

[0068] The cross-sectional land use type data include the proportion of built-up areas;

[0069] The associated pump station data include the dissolved oxygen concentration value and discharge flow rate of the water quality index of the forebay of the associated pump station at each section under rainfall;

[0070] The meteorological data includes rainfall data.

[0071] The method for acquiring basic data of the river section includes:

[0072] Count the water quality index concentration values ​​at the beginning of each rainfall, including the dissolved oxygen concentration value, and the minimum dissolved oxygen concentration value of the water quality index of the river section within 24 hours from the beginning to the end of the rainfall, to obtain the water quality data;

[0073] The average water temperature and average flow rate of the river section within 24 hours from the beginning to the end of each rainfall event are counted to obtain the hydrological data;

[0074] Count the proportion of built-up areas in riverbank sections to obtain land use type data of the sections;

[0075] Count the rainfall type of each rainfall event and the average water quality concentration of the forebay of the associated pumping station under the corresponding rainfall type, including the dissolved oxygen concentration value and the discharge flow, to obtain the data of the associated pumping station;

[0076] 24 hours is taken as the minimum interval between rainfall events, and the accumulated rainfall events greater than or equal to 3 mm are regarded as rainfall events, and the rainfall amount, rainfall duration, maximum rainfall intensity, average rainfall intensity, and previous drought time of each rainfall event are counted to obtain the meteorological data;

[0077] The water quality data, the hydrological data, the section land type data, the associated pump station data and the meteorological data are integrated into the basic data of the river section in units of rainfall events.

[0078] Step 2: Process the data and create a dataset.

[0079] The historical basic data of the river section are divided into a training set and a test set according to a preset ratio of 4:1.

[0080] Step 3: Train the prediction model.

[0081] Use the training set to train the random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model respectively, and use the grid search method to find the optimal parameter configuration of each prediction model;

[0082] According to the prediction results of the test set and the model, a scatter plot comparing the predicted values ​​of dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration is generated.

[0083] in, Figure 2 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the random forest prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0084] Figure 3 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the gradient boosting tree prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0085] Figure 4 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the XGBoost prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0086] Figure 5 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the BP neural network prediction model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration;

[0087] Figure 6 Shown is a scatter plot comparing the predicted value of dissolved oxygen concentration output by the support vector machine model provided by an embodiment of the present invention and the actual value of dissolved oxygen concentration.

[0088] The method of finding the optimal parameter configuration of each prediction model by using a grid search method includes:

[0089] For random forest prediction models: find the optimal parameter configuration for the number of estimators, the maximum depth of the decision tree, and the minimum number of samples required for a decision tree node to split;

[0090] For the gradient boosted tree prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of the decision tree;

[0091] For XGBoost prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of decision tree;

[0092] For the BP neural network prediction model: find the optimal parameter configuration for the number of hidden layers, the number of neurons in each layer of the neural network, the type of activation function, the type of optimization algorithm, and the number of training rounds;

[0093] For the support vector machine prediction model: find the optimal parameter configuration of the penalty coefficient, kernel function, and hyperparameters that control the influence range of the kernel function.

[0094] Step 4: Determine the dissolved oxygen concentration prediction model.

[0095] Based on the evaluation indicators, according to the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of the dissolved oxygen concentration, the prediction accuracy and fitting effect of each prediction model are analyzed to determine the dissolved oxygen concentration prediction model, including:

[0096] According to the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration, the determination coefficient R of each prediction model is calculated. 2 Value and Akaike Information Criterion AIC value;

[0097] According to the determination coefficient R of each prediction model 2 The prediction accuracy, fit and complexity of each prediction model were evaluated by using the Akaike information criterion (AIC) value, and the optimal prediction model was selected from each prediction model as the prediction model for dissolved oxygen concentration.

[0098] Among them, the closer the determination coefficient R² value is to 1, the higher the fit of the prediction model is, and the smaller the Akaike information criterion AIC value is, the more balanced the prediction model is between fitting data and maintaining complexity.

[0099] The determination coefficient R of each prediction model 2 The value calculation expression is:

[0100] (1);

[0101] In the formula, Indicates The true value of dissolved oxygen concentration of training samples, Indicates The predicted value of dissolved oxygen concentration of training samples, Indicates the average dissolved oxygen concentration.

[0102] The calculation expression of the Akaike Information Criterion AIC value of each prediction model is:

[0103] (2);

[0104] In the formula, represents the number of training samples, represents the logarithmic function, represents the likelihood function.

[0105] The Akaike Information Criterion ACI value and the determination coefficient R of the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual dissolved oxygen concentration obtained by calculation in this embodiment 2 The values ​​are shown in Table 3:

[0106]

[0107] The specific parameter configuration of the dissolved oxygen concentration prediction model is shown in Table 4:

[0108]

[0109] Step 5: Input the basic data of the river section where the dissolved oxygen concentration is to be measured into the dissolved oxygen concentration prediction model for prediction, and output the predicted value of dissolved oxygen concentration.

[0110] The basic data set X_test and the predicted value of dissolved oxygen concentration of the river section to be tested are shown in Table 5:

[0111]

[0112] Example 2

[0113] Based on the same inventive concept as that of Example 1, this embodiment introduces a computer-readable storage medium on which computer instructions are stored, characterized in that the computer instructions implement the steps of the method of Example 1 above when executed by a processor.

[0114] Example 3

[0115] Based on the same inventive concept as other embodiments, the present invention further provides a computer program product, including computer instructions, characterized in that the computer instructions implement the steps of the method in the above-mentioned embodiment 1 when executed by a processor.

[0116] In summary, the present invention predicts dissolved oxygen concentration by using five different prediction models, namely random forest, gradient boosting tree, XGBoost, BP neural network, and support vector machine, at the same time. Each algorithm can capture different features in the data with its unique advantages, thereby significantly improving the accuracy of the prediction. By comparing the output results of the five prediction models and selecting the best performing model as the final prediction model for dissolved oxygen concentration, the present invention effectively avoids the deviations or limitations that may exist in the prediction of a single algorithm. This multi-algorithm fusion selection mechanism greatly enhances the overall robustness, so that the prediction model can more comprehensively utilize the information in the data and maintain stable prediction performance when facing complex and changeable water environment data.

[0117] The present invention does not rely on a single data source, but integrates a wide range of data sources, including water quality data, hydrological data, section land type data, associated pump station data and meteorological data. The comprehensive use of these data can more comprehensively reflect the actual situation of the river section. At the same time, the present invention uses a specific combination of prediction models to determine the optimal dissolved oxygen concentration prediction model. This method of integrating data sources further improves the accuracy and practicality of the prediction. By comprehensively considering multiple factors, the present invention provides more comprehensive and accurate prediction support for the protection and management of the water environment.

[0118] The present invention calculates the determination coefficient R² and the Akaike information criterion AIC value by comparing the comparison scatter plot of the predicted value of dissolved oxygen concentration output by each prediction model with the actual value. The balance between the model's fit and complexity is measured. Based on these evaluation indicators, the prediction accuracy, fit and complexity of each model are comprehensively evaluated at the same time, and finally the optimal prediction model is selected as the dissolved oxygen concentration prediction model. It not only ensures that the selected model has a high degree of prediction accuracy, but also achieves the best fit of the data while maintaining the complexity, providing reliable technical support for the monitoring and management of the water environment.

[0119] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0120] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0121] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0122] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0123] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A method for predicting dissolved oxygen concentration in a river section, characterized in that: include: Obtaining basic data of the river section where the dissolved oxygen concentration is to be measured, the basic data including water quality data, hydrological data, section land type data, associated pump station data and meteorological data; Inputting the basic data of the river section of the to-be-measured dissolved oxygen concentration into a dissolved oxygen concentration prediction model for prediction, and outputting a predicted value of dissolved oxygen concentration; Wherein, the method for determining the dissolved oxygen concentration prediction model includes: Obtain historical basic data of the river section, input the pre-trained random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model, and output a scatter plot comparing the predicted value of dissolved oxygen concentration of each model with the actual value of dissolved oxygen concentration; Based on the evaluation indicators, the prediction accuracy and fitting effect of each prediction model were analyzed according to the comparative scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of the dissolved oxygen concentration, and the dissolved oxygen concentration prediction model was determined.

2. The method for predicting dissolved oxygen concentration in a river section according to claim 1, characterized in that: The water quality data include the dissolved oxygen concentration value of the water quality indicator of the river section; The hydrological data include the average water temperature and average flow of the river section; The cross-sectional land use type data include the proportion of built-up areas; The associated pump station data include the dissolved oxygen concentration value and discharge flow rate of the water quality index of the forebay of the associated pump station at each section under rainfall; The meteorological data includes rainfall data.

3. The method for predicting dissolved oxygen concentration in a river section according to claim 1, characterized in that: The method for acquiring basic data of the river section includes: Count the water quality index concentration values ​​at the beginning of each rainfall, including the dissolved oxygen concentration value, and the minimum dissolved oxygen concentration value of the water quality index of the river section within 24 hours from the beginning to the end of the rainfall, to obtain the water quality data; The average water temperature and average flow rate of the river section within 24 hours from the beginning to the end of each rainfall event are counted to obtain the hydrological data; Count the proportion of built-up areas in riverbank sections to obtain land use type data of the sections; Count the rainfall type of each rainfall event and the average water quality concentration of the forebay of the associated pumping station under the corresponding rainfall type, including the dissolved oxygen concentration value and the discharge flow, to obtain the data of the associated pumping station; 24 hours is taken as the minimum interval between rainfall events, and the accumulated rainfall events greater than or equal to 3 mm are regarded as rainfall events, and the rainfall amount, rainfall duration, maximum rainfall intensity, average rainfall intensity, and previous drought time of each rainfall event are counted to obtain the meteorological data; The water quality data, the hydrological data, the section land type data, the associated pump station data and the meteorological data are integrated into the basic data of the river section in units of rainfall events.

4. The method for predicting dissolved oxygen concentration in a river section according to claim 1, characterized in that: Obtain historical basic data of the river section, input the pre-trained random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model, and output a comparative scatter plot of the predicted value of dissolved oxygen concentration of each model and the actual value of dissolved oxygen concentration, including: Dividing the historical basic data of the river section into a training set and a test set; Use the training set to train the random forest prediction model, gradient boosting tree prediction model, XGBoost prediction model, BP neural network prediction model, and support vector machine prediction model respectively, and use the grid search method to find the optimal parameter configuration of each prediction model; According to the prediction results of the test set and the model, a scatter plot comparing the predicted values ​​of dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration is generated.

5. The method for predicting dissolved oxygen concentration in a river section according to claim 4, characterized in that: The method of finding the optimal parameter configuration of each prediction model by using a grid search method includes: For random forest prediction models: find the optimal parameter configuration for the number of estimators, the maximum depth of the decision tree, and the minimum number of samples required for a decision tree node to split; For the gradient boosted tree prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of the decision tree; For XGBoost prediction model: find the optimal parameter configuration for the number of evaluators, learning rate, and maximum depth of decision tree; For the BP neural network prediction model: find the optimal parameter configuration for the number of hidden layers, the number of neurons in each layer of the neural network, the type of activation function, the type of optimization algorithm, and the number of training rounds; For the support vector machine prediction model: find the optimal parameter configuration of the penalty coefficient, kernel function, and hyperparameters that control the influence range of the kernel function.

6. The method for predicting dissolved oxygen concentration in a river section according to claim 4, characterized in that: The evaluation index is based on the comparison scatter plot of the predicted dissolved oxygen concentration value output by each prediction model and the actual value of the dissolved oxygen concentration, and the prediction accuracy and fitting effect of each prediction model are analyzed to determine the dissolved oxygen concentration prediction model, including: According to the comparison scatter plot of the predicted dissolved oxygen concentration output by each prediction model and the actual value of dissolved oxygen concentration, the determination coefficient R of each prediction model is calculated. 2 Value and Akaike Information Criterion AIC value; According to the determination coefficient R of each prediction model 2 The prediction accuracy, fit and complexity of each prediction model were evaluated by using the Akaike information criterion (AIC) value, and the optimal prediction model was selected from each prediction model as the prediction model for dissolved oxygen concentration. Among them, the closer the determination coefficient R² value is to 1, the higher the fit of the prediction model is, and the smaller the Akaike information criterion AIC value is, the more balanced the prediction model is between fitting data and maintaining complexity.

7. The method for predicting dissolved oxygen concentration in a river section according to claim 6, characterized in that: The determination coefficient R of each prediction model 2 The value calculation expression is: （1）； In the formula, Indicates The true value of dissolved oxygen concentration of training samples, Indicates The predicted value of dissolved oxygen concentration of training samples, Indicates the average dissolved oxygen concentration.

8. The method for predicting dissolved oxygen concentration in a river section according to claim 6, characterized in that: The calculation expression of the Akaike Information Criterion AIC value of each prediction model is: （2）； In the formula, represents the number of training samples, represents the logarithmic function, represents the likelihood function.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images