Data processing method and system for hydraulic engineering

Abstract

The invention discloses a data processing method and system for a water conservancy project, and belongs to the technical field of data processing, and the method specifically comprises the steps: collecting water conservancy project multi-source data, carrying out the preprocessing and dynamic correction, building a nonlinear hydrodynamic model according to the local water flow characteristics, and calculating the state distribution of water flow. The first water flow state distribution is obtained, the water flow state is predicted through machine learning, the second water flow state distribution is obtained, the final water flow state distribution is obtained in combination with the first water flow state distribution, and whether abnormity occurs or not is analyzed; by combining various data preprocessing technologies, data quality is effectively improved, water flow state information is effectively analyzed for tasks such as accurate prediction of the water flow state and flood early warning, and a decision maker is helped to timely obtain water flow change information and make a quick decision.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and specifically relates to a data processing method and system for water conservancy projects. Background Art

[0002] In the design, construction, and management processes of water conservancy projects, a large amount of water flow data, meteorological data, geographical data, etc. are involved. These data usually have the characteristics of large scale, complexity, and diversity. For example, variables such as water flow status, precipitation, temperature, and flow velocity are crucial for water resource management, flood warning, irrigation scheduling, etc. With the rapid development of information technology and the Internet of Things, more and more sensors, remote sensing technologies, and monitoring devices are applied to water conservancy projects, generating a large amount of real-time and historical data.

[0003] However, in the face of massive and complex data, traditional data processing methods often struggle to cope and have the following problems: 1. Data noise and incompleteness: The data in water conservancy projects may be affected by factors such as environmental changes and equipment failures, resulting in data noise or missing; 2. Data inconsistency: Data from different sources, at different times, and in different formats are inconsistent; 3. The efficiency and real-time nature of data processing; 4. Complex data relationships.

[0004] For example, the Chinese patent with the authorization announcement number CN117520752B discloses a water conservancy project information management method based on big data, including: based on the water level change amount corresponding to the initial water level data, dividing the water level data line graph into regions, confirming the high-frequency water level region and the low-frequency water level region, and then calculating the wavelet decomposition layers corresponding to each region in the high-frequency water level region and the low-frequency water level region according to the wavelet layer score factors corresponding to the high-frequency water level region and the low-frequency water level region. Finally, based on the wavelet decomposition layers and a preset denoising algorithm, the initial water level data in each region of the high-frequency water level region and the low-frequency water level region are respectively denoised to confirm the final water level data after denoising. By dividing the water level data line graph into regions and denoising the initial water level data in each region according to the wavelet decomposition layers of each region, compared with the traditional denoising method, the accuracy of denoising is improved, and thus the denoising cost is reduced.

[0005] The defect of the above patent is that it does not consider the complex relationships between data and it is difficult to use the data for tasks such as prediction. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the present invention proposes a data processing method and system for water conservancy projects, aiming to improve the accuracy, efficiency, and real-time nature of data processing in water conservancy projects through means such as multi-dimensional fusion, machine learning, and prediction fusion.

[0007] To achieve the above object, the present invention provides the following technical solutions: A data processing method for water conservancy projects, comprising the following specific steps: Collect multi-source data of water conservancy projects, and perform preprocessing and dynamic calibration; Establish a non-linear hydrodynamics model for local water flow characteristics, calculate the state distribution of the water flow, and obtain the first water flow state distribution; Use machine learning to predict the water flow state, obtain the second water flow state distribution, combine it with the first water flow state distribution, obtain the final water flow state distribution, and analyze whether an anomaly occurs.

[0008] Specifically, the dynamic calibration includes: Calculate the calibration factor using the calibration formula, and the specific formula is: , where K represents the calibration factor, represents the i-th preprocessed data, represents the weight factor corresponding to the i-th preprocessed data, represents the data amplification factor, represents the calibration coefficient, q represents the data compression coefficient, represents an integer.

[0009] Specifically, the step of establishing a non-linear hydrodynamics model for local water flow characteristics, calculating the state distribution of the water flow, and obtaining the first water flow state distribution includes: Establish a non-linear hydrodynamics model for local water flow characteristics, and the specific formula is: ; where, represents the fluid density, represents the fluid velocity vector, represents the fluid pressure, reflecting the internal mechanical state of the fluid, represents the dynamic viscosity of the fluid, represents the gradient, represents the partial derivative function, represents the turbulence correction term, represents the boundary effect correction term; According to the actual structure of the water conservancy project, determine the kinematic boundary conditions. At the solid boundary (, the no-slip boundary condition is adopted, that is, the flow velocity is zero. At the free surface boundary, consider the effects of surface tension and atmospheric pressure, and combine kinematics to calculate the dynamic boundary conditions. The specific formula for the kinematic boundary conditions is: , where, represents the water surface elevation, represents the velocity component in the x direction in the two-dimensional plane, Represents the flow velocity component in the y - direction within the two - dimensional plane; The finite volume method is used to numerically solve the non - linear hydrodynamic model to obtain the state distribution of the water flow The calculated water flow state distribution is compared and verified with the actual measurement data, the error index is calculated, and according to the comparison and verification results, the parameters of the non - linear hydrodynamic model are adjusted and optimized, and the calculation is carried out again to finally obtain the first water flow state distribution.

[0010] Specifically, using machine learning to predict the water flow state to obtain the second water flow state distribution, combining with the first water flow state distribution to obtain the final water flow state distribution, and analyzing whether an anomaly occurs, including: Select and train a machine learning model to predict the future water flow state to obtain the second water flow state distribution; Use a fusion strategy to fuse the first water conservancy state distribution and the second water flow state distribution to obtain the final water flow state distribution; According to the final water flow state distribution, analyze whether an anomaly occurs.

[0011] Specifically, the selecting and training of the machine learning model to predict the future water flow state to obtain the second water flow state distribution includes: Collect historical water flow state data to construct a time - series data set; Based on the features extracted from the non - linear hydrodynamic model, further extract water flow features, including the degree of water flow urgency, Reynolds number, and time - series features; Select a machine learning model according to the complexity of the water flow state of the water conservancy project and the data characteristics; Divide the historical data into a training set and a test set according to a preset ratio. During the training process, the early stopping method is used to prevent overfitting. When the objective function no longer decreases within consecutive iteration cycles, stop the training and save the model parameters with the best current performance to obtain the trained machine learning model; Input the real - time multi - source data of the water conservancy project into the trained machine learning model, and output the predicted water flow state, that is, the second water flow state distribution.

[0012] Specifically, the using of the fusion strategy to fuse the first water conservancy state distribution and the second water flow state distribution to obtain the final water flow state distribution includes: For the non - linear hydrodynamic model, calculate the credibility according to the accuracy of the input parameters of the non - linear hydrodynamic model and the calculation credibility of the boundary conditions; For the machine learning model, calculate the credibility based on the training effect and generalization ability of the model; Set the fusion weights according to the credibility of the non - linear hydrodynamic model and the machine learning model, and obtain the final water flow state distribution through weighted fusion according to the fusion weights.

[0013] Specifically, analyzing whether an anomaly occurs based on the final water flow state distribution includes: Calculate the statistical features of the final water flow state distribution, including: mean and standard deviation; Set the threshold interval of the water flow state distribution. When the statistical features exceed the threshold interval of the water flow state distribution, it is determined as an anomaly; Record the time, location and type of the anomaly occurrence, and formulate corresponding treatment measures according to the anomaly cause.

[0014] A data processing system for water conservancy projects, used to implement the data processing method for water conservancy projects described above, includes: a data processing module, a first state calculation module and an analysis module; The data processing module is used to collect multi - source data of water conservancy projects and perform pre - processing and dynamic correction; The first state calculation module is used to establish a non - linear hydrodynamic model for local water flow characteristics, calculate the state distribution of water flow, and obtain the first water flow state distribution; The analysis module is used to predict the water flow state using machine learning to obtain the second water flow state distribution, combine it with the first water flow state distribution to obtain the final water flow state distribution, and analyze whether an anomaly occurs.

[0015] Specifically, the analysis module includes: a second state calculation unit and a fusion analysis unit; The second state calculation unit is used to predict the second water flow state distribution using a machine learning model; The fusion analysis unit is used to fuse the first water flow state distribution and the second water flow state distribution using a fusion strategy and perform anomaly analysis.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention proposes a data processing method for water conservancy projects. By combining various data pre - processing technologies, the quality of data is effectively improved. For tasks such as accurate prediction of water flow state and flood warning, the accuracy of data is crucial.

[0017] 2. The present invention proposes a data processing method for water conservancy projects. Considering the complex relationships of water conservancy project data, it realizes real - time data processing and instant feedback, helping decision - makers obtain water flow change information in a timely manner and make quick decisions.

[0018] 3. The present invention provides a data processing method for water conservancy projects. Through physical models and machine learning models, the water flow state can be effectively analyzed, enhancing the decision-making support ability and adaptive ability of water conservancy projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flowchart of a data processing method for water conservancy projects provided by the present invention; Figure 2 It is an architecture diagram of a data processing system for water conservancy projects provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] The following will describe the present application in detail with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any form. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made. These all fall within the protection scope of the present application.

[0021] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0022] It should be noted that if there is no conflict, the various features in the embodiments of the present application can be combined with each other, and all are within the protection scope of the present application. In addition, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order from the module division in the device or the flowchart. In addition, the terms "first", "second", "third", etc. used in the present application do not limit the data and execution order, but only distinguish the same items or similar items with basically the same functions and effects. Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used in this specification in the description of the present application are only for the purpose of describing specific embodiments and are not used to limit the present application. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0023]

[0024] Embodiment 1 Please refer to Figure 1 , an embodiment provided by the present invention: A data processing method for water conservancy projects includes the following specific steps: Step S1: Collect multi-source data of water conservancy projects, and perform preprocessing and dynamic calibration; The multi-source data of the water conservancy project includes: sensor data, remote sensing images, meteorological data, and historical project data; the sensor data includes: water flow data, using a water flow sensor to measure the water velocity, flow rate, and direction of the water body. Usually, an ultrasonic flowmeter or an electromagnetic flowmeter is used, which can maintain high accuracy under different flow velocity conditions; pressure data, using a pressure sensor to monitor the water pressure of the water body and the water pressure inside the dam or pipeline, facilitating real-time understanding of the physical load borne by the water body; temperature data, using a temperature sensor to monitor the change of water temperature. The change of water temperature has an impact on both the water density and the flow pattern; rainfall, using a rainfall sensor to monitor the rainfall intensity and cumulative rainfall, which directly affects water body recharge and flood risk assessment; structural stress data, using a structural stress sensor to monitor the stress and vibration of the structure, preventing safety problems caused by structural fatigue or abnormal vibration; The preprocessing includes: filtering, using algorithms such as low-pass filters, median filters, or Kalman filters to perform noise reduction processing on the original data. For example, for water flow data, Kalman filtering can be used to smooth short-term fluctuations and remove random noise; signal smoothing: smoothing the data to reduce the impact of sudden noise, but at the same time retaining the key dynamic change characteristics in the actual signal; data normalization: in order to eliminate the differences in the measurement ranges of different sensors, perform normalization processing on the data of each sensor before data fusion, reducing the differences between different data magnitudes, and facilitating subsequent multi-data fusion and model training; missing value processing: interpolation method, when data is missing during the acquisition process, methods such as linear interpolation and spline interpolation can be used to supplement the missing data to ensure data continuity; statistical completion, for long-term missing data, the mean or median of historical data can be used for completion, and at the same time, an outlier detection mechanism is considered to prevent interference caused by incorrect filling to subsequent analysis; In the data acquisition and preprocessing stage, not only is it required that the sensor layout is reasonable and the acquisition equipment is accurate, but also the original data needs to be finely processed through various preprocessing techniques (denoising, normalization, missing value completion, and outlier detection) to lay a solid foundation for subsequent data fusion, model establishment, and intelligent prediction. Only by ensuring the data quality can accurate and efficient monitoring and early warning in water conservancy projects be achieved; The dynamic calibration in Step S1 includes: Calculate the calibration factor using the calibration formula. The specific formula is: , where K represents the calibration factor, represents the i-th preprocessed data, represents the weight factor corresponding to the i-th preprocessed data, represents the data amplification factor, represents the calibration coefficient, and q represents the data compression coefficient. represents an integer to prevent division by zero.

[0025] Step S2: For the local flow characteristics, establish a non - linear hydrodynamic model, calculate the state distribution of the water flow, and obtain the first water flow state distribution. The specific steps of Step S2 are as follows: Step S201: For the local flow characteristics, establish a non - linear hydrodynamic model. The specific formula is: ; where, represents the fluid density, which usually depends on water temperature and salt content. represents the fluid velocity vector, which describes the motion velocity of the fluid in each direction. represents the fluid pressure, which reflects the internal mechanical state of the fluid. represents the dynamic viscosity of the fluid, which reflects the internal friction characteristics of the fluid. represents the gradient. represents the partial derivative function. represents the turbulence correction term, which is used to describe the energy dissipation and momentum transfer deviation caused by local turbulence. represents the boundary effect correction term, which considers the additional resistance and reaction force generated when the water body contacts the solid boundary (such as dams, riverbeds), so that the flow state in the boundary layer can more accurately reflect the actual situation. In this embodiment, the non - linear convection term describes the non - linear acceleration effect generated by the fluid's own motion. This term is particularly crucial in high - flow - velocity and complex flow fields. The diffusion term represents the momentum diffusion effect caused by molecular viscosity, which smooths the velocity gradient in the flow and helps describe the energy dissipation process. The turbulence correction term can, to a certain extent, compensate for the energy and momentum perturbations caused by turbulence, making the model more sensitive to local turbulence effects. In actual water conservancy projects, the interaction between water flow and solid boundaries is very complex. By introducing the boundary effect correction term, the flow resistance and recirculation phenomenon in the boundary layer can be better simulated, thereby improving the accuracy of the entire model. The advantages of the formula are as follows: By introducing the turbulence correction and boundary effect correction terms, the model is closer to the actual working conditions when describing complex flow phenomena. Especially in the cases of high Reynolds numbers and obvious boundary layers, it can provide more accurate flow field predictions. The modified model is not only applicable to the uniform flow under ideal conditions but also can handle the non - uniform, non - linear, and turbulent complexity problems common in water conservancy projects, improving the applicability and robustness of the model in various actual scenarios.

[0026] Step S202: Determine the accurate kinematic boundary conditions according to the actual structure of the water conservancy project. At the solid boundaries (such as the dam body, pipe wall, etc.), the no-slip boundary condition is adopted, that is, the flow velocity is zero. At the free water surface boundary, considering the effects of surface tension and atmospheric pressure, the dynamic boundary conditions are combined with kinematic calculations. The specific formula for the kinematic boundary conditions is as follows: , where, represents the water surface elevation, represents the velocity component in the x-direction in the two-dimensional plane, represents the velocity component in the y-direction in the two-dimensional plane; Step S203: Numerically solve the non-linear hydrodynamic model using the finite volume method to obtain the state distribution of the water flow; Specifically, the computational domain is divided into a series of non-overlapping control volumes, the conservation equations within the control volumes are integrated, the partial differential equations are transformed into algebraic equations, and through the Gauss divergence theorem, the surface integral is transformed into a line integral, and further discretized to obtain a system of linear equations; Step S204: Compare and verify the calculated water flow state distribution with the actual measurement data, calculate the error index, adjust and optimize the parameters of the non-linear hydrodynamic model according to the comparison and verification results, and perform calculations again to finally obtain the first water flow state distribution.

[0027] Step S3: Use machine learning to predict the water flow state, obtain the second water flow state distribution, combine it with the first water flow state distribution to obtain the final water flow state distribution, and analyze whether an anomaly occurs.

[0028] The specific steps of Step S3 are as follows: Step S301: Select and train a machine learning model to predict the future water flow state and obtain the second water flow state distribution; The specific steps of Step S301 are as follows: Step S3011: Collect historical water flow state data, integrate the flow velocity, pressure, water level, etc. data collected by sensors at different times and under different working conditions in the past, and construct a time series data set covering rich information; In this embodiment, these historical data reflect the water flow characteristics of the water conservancy project under various conditions such as normal operation, seasonal changes, and different flow rate regulations, and are cleaned and preprocessed to ensure the quality and integrity of the data; The quality of the data directly determines the accuracy of the prediction. Ensuring that the data is clear and without missing values helps to construct an effective model; by collecting multi-dimensional data, more information can be provided for the model and the accuracy of the prediction can be increased; Step S3012: Based on the extraction of features such as turbulent kinetic energy and vorticity in step S2 using a physical model, further expand the feature space, and calculate the degree of water flow rapidity, Reynolds number, time series features, and other features; In this embodiment, the degree of water flow rapidity is crucial for determining whether the water flow is in a critical state; the Reynolds number is used to measure whether the flow characteristics of the fluid are laminar or turbulent; perform difference processing on time series data to obtain the change rates of variables such as flow velocity and pressure, which helps to capture the dynamic trend of the water flow state. For example, the first-order difference can reflect the instantaneous increase or decrease of the flow velocity; adopt dimensionality reduction techniques such as principal component analysis (PCA) to screen out features with high contribution to the prediction of the water flow state, reduce data redundancy, and at the same time avoid overfitting problems and improve the training efficiency of the machine learning model; Feature engineering can remove irrelevant features and noise, thereby reducing interference in model training. By performing different transformations on the data, the model can be made more adaptable to the distribution characteristics of the data and improve the generalization ability; Step S3013: According to the complexity of the water flow state of the water conservancy project and the data characteristics, select a suitable machine learning model. For example, the long short-term memory network (LSTM) is very effective for capturing the continuous change law of the water flow state over time because it is good at dealing with long-term dependencies in time series data; the random forest model has strong anti-overfitting ability and can handle high-dimensional feature data, which has advantages in the classification prediction of water flow states under multiple working conditions; the gradient boosting decision tree (GBDT) constructs a strong classifier by iteratively training multiple weak classifiers and has good fitting ability for the non-linear changes of the water flow state; Step S3014: Divide the prepared historical data into a training set and a test set according to a preset ratio. During the training process, adopt the early stopping method to prevent overfitting. When the objective function no longer decreases within consecutive iteration cycles, stop the training and save the model parameters with the best current performance to obtain a trained machine learning model; In this embodiment, performance evaluation: Use the test set to evaluate the trained model and calculate various performance indicators. In addition to the root mean square error, the mean absolute error can also be calculated; Model tuning: According to the evaluation results, tune the model. If it is found that the model has underfitting phenomena (such as low R2, large MAE and RMSE), the complexity of the model can be increased. For example, for LSTM, increase the number of hidden layers or neurons; if there is an overfitting problem, adopt regularization techniques, such as adding an L2 regularization term in LSTM to constrain the weight matrix and prevent the model from overlearning the noise in the training data.

[0029] Step S3015: Input the real-time multi-source data of the water conservancy project into the trained machine learning model, and output the predicted water flow state, that is, the second water flow state distribution.

[0030] Step S302: Use a fusion strategy to fuse the first water state distribution and the second water flow state distribution to obtain the final water flow state distribution; The specific steps of step S302 are as follows: Step S3021: For the non - linear hydrodynamic model, consider factors such as the accuracy of model input parameters, the rationality of boundary condition setting, and the theoretical accuracy of the model itself to evaluate its credibility; Exemplarily, by analyzing the influence of sensor measurement errors on key parameters (such as fluid viscosity, density, etc.) in the hydrodynamic model, using error propagation theory to quantify the input uncertainty of the physical model. At the same time, compare the fitting degree between the simulation results of the physical model under different historical working conditions and the actual monitoring data, and use indicators such as root mean square error (RMSE), coefficient of determination (R²), etc. for quantitative evaluation to establish a credibility function based on these factors; Step S3022: For the machine - learning model (which obtains the second water flow state distribution), measure the credibility based on the training effect, generalization ability, and adaptability to new data of the model; On the one hand, during the training process, use techniques such as cross - validation to monitor the over - fitting or under - fitting situation of the model to ensure that the model has good generalization performance. On the other hand, as new water flow state data is continuously collected, real - time evaluate the deviation between the model prediction value and the actual value, and use similar statistical indicators to construct the credibility function of the machine - learning model; Step S3023: Set the fusion weights according to the credibility of the non - linear hydrodynamic model and the machine - learning model, and weighted - fuse according to the fusion weights to obtain the final water flow state distribution.

[0031] According to the above - mentioned credibility function, calculate the credibility weights of each model at the current moment, and the sum of the weights is 1. This fusion method ensures that in any working condition, the results of the model with higher credibility are preferentially adopted, while taking into account the supplementary information of the other model to improve the reliability of the final water flow state distribution.

[0032] Step S303: Analyze whether an anomaly occurs according to the final water flow state distribution.

[0033] The specific steps of step S303 are as follows: Step S3031: Calculate the statistical characteristics of the final water flow state distribution, such as mean, standard deviation, etc.; Step S3032: Set the threshold interval of the water flow state distribution, usually determined based on the statistical analysis of historical normal data. When the statistical characteristics exceed the threshold interval of the water flow state distribution, it is determined as an anomaly; In this embodiment, anomaly detection based on a machine learning classification model: Use the labeled historical water flow state data of anomalies and normals to train classification models such as support vector machines (SVMs) and decision trees. Take the current final water flow state distribution as the input, and the model outputs a category judgment of whether it belongs to an anomaly or normal. When training an SVM, find an optimal hyperplane to separate the two types of data and maximize the margin distance, so that the model can accurately distinguish between abnormal and normal states when facing new data. For a decision tree, construct decision branches according to the thresholds of different features and gradually judge the category of the water flow state; Deep learning anomaly detection: Adopt deep learning models such as autoencoders. First, perform unsupervised training on normal water flow state data to learn the latent feature representation of normal data. Then, in the prediction stage, input the current water flow state data into the autoencoder and calculate the reconstruction error. If the reconstruction error exceeds the preset threshold, it indicates that the current data deviates greatly from the normal mode and is determined to be an anomaly. The autoencoder can automatically capture the complex internal structure of the data and has good detection ability for some complex anomalies that are not easily identified by simple statistical features; Step S3033: Once an anomaly is detected, immediately record the time, location (positioned by sensor coordinate information), and type of anomaly (such as a sudden increase in flow velocity, abnormal pressure fluctuations, etc.). Develop corresponding handling measures according to the cause of the anomaly.

[0034] In-depth analysis of the cause of the anomaly: Combine the current operating conditions of the water conservancy project (such as whether flood discharge, equipment maintenance, etc. are in progress), meteorological conditions (heavy rain, strong wind, etc.), and surrounding environmental factors (such as river channel siltation, landslide impact, etc.) to comprehensively judge the root cause of the anomaly. For example, if it is detected that the flow velocity near a certain sluice gate decreases abnormally and there are reports of a large amount of sediment deposition upstream at the same time, it may be due to the obstruction of the water flow channel caused by the sedimentation; Develop corresponding handling measures according to the cause of the anomaly: If the anomaly is caused by equipment failure, promptly notify the maintenance personnel for emergency repair; if it is caused by natural factors, such as a flood approaching, activate the emergency plan, adjust the operating parameters of the water conservancy facilities (such as increasing the flood discharge volume, closing some sluice gates, etc.) to ensure the safe and stable operation of the water conservancy project, and at the same time report the anomaly information to the management center in a timely manner for subsequent further analysis and decision-making; Comprehensively use the method of combining machine learning and physical models to accurately obtain the final water flow state distribution, and timely and effectively detect and handle anomalies, providing a solid guarantee for the intelligent management and safe operation of water conservancy projects.

[0035] Embodiment 2 Please refer to Figure 2 , another embodiment provided by the present invention: A data processing system for a water conservancy project, including: a data processing module, a first state calculation module, and an analysis module; The data processing module is used to collect multi-source data of water conservancy projects and perform preprocessing and dynamic calibration; The first state calculation module is used to establish a non-linear hydrodynamic model for local water flow characteristics, calculate the state distribution of the water flow, and obtain the first water flow state distribution; The analysis module is used to predict the water flow state using machine learning to obtain the second water flow state distribution, combine it with the first water flow state distribution to obtain the final water flow state distribution, and analyze whether an anomaly occurs.

[0036] The analysis module includes: a second state calculation unit and a fusion analysis unit; The second state calculation unit is used to predict the second water flow state distribution using a machine learning model; The fusion analysis unit is used to fuse the first water flow state distribution and the second water flow state distribution using a fusion strategy and perform anomaly analysis.

[0037] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the corresponding technical solutions in the prior art in terms of implementation principles are not described in detail to avoid excessive elaboration.

[0038] As described above in the specific embodiments, the purpose, technical solutions, and beneficial effects of the present invention have been further described in detail. It should be understood that the above is only the specific embodiments of the present invention and is not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data processing method for water conservancy projects, characterized in that, It includes the following specific steps: Collect multi-source data of water conservancy projects, and perform preprocessing and dynamic calibration; Establish a non-linear hydrodynamic model for local water flow characteristics, calculate the state distribution of the water flow, and obtain the first water flow state distribution; Use machine learning to predict the water flow state, obtain the second water flow state distribution, combine it with the first water flow state distribution, obtain the final water flow state distribution, and analyze whether an anomaly occurs.

2. The data processing method for a water conservancy project according to claim 1, wherein The dynamic calibration includes: Calculate the calibration factor using the calibration formula, and the specific formula is: ， where K represents a correction factor, represents the i-th preprocessed data, represents the weight factor corresponding to the i-th preprocessed data, represents a data amplification factor, represents a calibration coefficient, and q represents a data compression coefficient, represents an integer.

3. The data processing method for a water conservancy project according to claim 2, characterized in that, The step of establishing a non-linear hydrodynamic model for local water flow characteristics, calculating the state distribution of the water flow, and obtaining the first water flow state distribution includes: Establish a non-linear hydrodynamic model for local water flow characteristics, and the specific formula is: ； Among them, represents the fluid density, represents the fluid velocity vector, represents the fluid pressure, reflecting the internal mechanical state of the fluid, represents the dynamic viscosity of the fluid, represents the gradient, represents the partial derivative function, represents the turbulence correction term, represents the boundary effect correction term; According to the actual structure of the water conservancy project, determine the kinematic boundary conditions. At the solid boundary, the no-slip boundary condition is adopted, that is, the flow velocity is zero. At the free surface boundary, consider the effects of surface tension and atmospheric pressure, and combine kinematic calculations to obtain the dynamic boundary conditions. The specific formula for the kinematic boundary conditions is: ， Among them, represents the water surface elevation, represents the velocity component in the x-direction in the two-dimensional plane, represents the velocity component in the y-direction in the two-dimensional plane; Use the finite volume method to numerically solve the non-linear hydrodynamic model to obtain the state distribution of the water flow Compare and verify the calculated water flow state distribution with the actual measurement data, calculate the error index, adjust and optimize the parameters of the non-linear hydrodynamic model according to the comparison and verification results, and perform calculations again to finally obtain the first water flow state distribution.

4. A data processing method for water conservancy projects according to claim 3, characterized in that, The step of using machine learning to predict the water flow state, obtaining the second water flow state distribution, combining it with the first water flow state distribution, obtaining the final water flow state distribution, and analyzing whether an anomaly occurs includes: Select and train a machine learning model to predict the future water flow state and obtain the second water flow state distribution; Use a fusion strategy to fuse the first water conservancy state distribution and the second water flow state distribution to obtain the final water flow state distribution; Analyze whether an anomaly occurs according to the final water flow state distribution.

5. A data processing method for water conservancy projects according to claim 4, characterized in that, The step of selecting and training a machine learning model to predict the future water flow state and obtain the second water flow state distribution includes: Collect historical water flow state data and construct a time series data set; Based on the features extracted from the non-linear hydrodynamic model, further extract water flow features, including the rapidity and slowness of the water flow, the Reynolds number, and time series features; Select a machine learning model according to the complexity of the water flow state of the water conservancy project and the data characteristics; Divide the historical data into a training set and a test set according to a preset ratio. During the training process, use the early stopping method to prevent overfitting. When the objective function no longer decreases within consecutive iteration cycles, stop the training and save the model parameters with the best current performance to obtain the trained machine learning model; Input the real-time multi-source data of the water conservancy project into the trained machine learning model, and output the predicted water flow state, that is, the second water flow state distribution.

6. The data processing method for a water conservancy project according to claim 4, characterized in that, The step of using a fusion strategy to fuse the first water conservancy state distribution and the second water flow state distribution to obtain the final water flow state distribution includes: For the non-linear hydrodynamic model, calculate the credibility of the input parameters and boundary conditions of the non-linear hydrodynamic model; For a machine learning model, calculate the credibility based on the training effect and generalization ability of the model; Set the fusion weight according to the credibility of the non - linear hydrodynamic model and the machine learning model, and obtain the final water flow state distribution through weighted fusion according to the fusion weight.

7. The data processing method for a water conservancy project according to claim 4, characterized in that According to the final water flow state distribution, analyze whether an anomaly occurs, including: Calculate the statistical features of the final water flow state distribution, including: mean and standard deviation; Set the threshold interval of the water flow state distribution. When the statistical features exceed the threshold interval of the water flow state distribution, it is determined as an anomaly; Record the time, location and type of the anomaly occurrence, and formulate corresponding treatment measures according to the cause of the anomaly.

8. A data processing system for water conservancy projects, which is used to implement a data processing method for water conservancy projects described in any one of claims 1-7, characterized in that, Including: A data processing module, a first state calculation module and an analysis module; The data processing module is used to collect multi - source data of the water conservancy project, and perform pre - processing and dynamic correction; The first state calculation module is used to establish a non - linear hydrodynamic model for local water flow characteristics, calculate the state distribution of the water flow, and obtain the first water flow state distribution; The analysis module is used to predict the water flow state using machine learning to obtain the second water flow state distribution, combine it with the first water flow state distribution to obtain the final water flow state distribution, and analyze whether an anomaly occurs.

9. The data processing system for a water conservancy project according to claim 8, wherein, The analysis module includes: a second state calculation unit and a fusion analysis unit; The second state calculation unit is used to predict the second water flow state distribution using a machine learning model; The fusion analysis unit is used to fuse the first water flow state distribution and the second water flow state distribution using a fusion strategy and perform anomaly analysis.

Images