Intelligent analysis method and system for sluice hydraulic section model test result

By building a physical model of the sluice section and using data acquisition equipment, combined with principal component analysis and machine learning, the problems of unstable data and unintelligent processing in the sluice hydrocarbon section model test are solved, and intelligent evaluation of the sluice performance and Analysis improves the scientificity and security of design and operation.

CN118965295AInactive Publication Date: 2024-11-15东莞市运河治理中心
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411139073.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-20
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing technology has problems such as unstable data, unintelligent processing and easy-to-miss features in the test of sluice hydrocarbon section models, which leads to the inability to comprehensively evaluate the performance of sluice, making it difficult to improve the scientificity and effectiveness of design and operation.

Method used

The physical model of the sluice section is used to record key data in real time using data acquisition equipment, extract key features through principal component analysis, build a sluice operation model, and use machine learning to conduct data-driven evaluation, and monitor and optimize the sluice status in real time.

Benefits of technology

It realizes intelligent evaluation and analysis of sluice performance, extracts key features, improves data reliability and processing efficiency, and enhances the scientificity and safety of sluice design and operation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965295A_ABST
    Figure CN118965295A_ABST
Patent Text Reader

Abstract

The invention provides an intelligent analysis method and system for a water gate hydraulic section model test result, and the method comprises the steps: building a water gate section physical model, and recording key data in a test process in real time through data collection equipment; performing cleaning and time sequence formatting on the collected original key data to obtain a first data set; constructing a sluice operation model according to the extracted key features, and parameterizing related variables such as flow and pressure by using the key features; evaluating flood discharge capacity and flow state connection conditions under different opening degrees; flow coefficients are calculated and subjected to statistical analysis, and test values of the flow coefficients under different operation conditions are obtained; and associating the extracted key features with sluice performance indexes, and training a data-driven evaluation model based on machine learning. The system comprises a data acquisition module, a data analysis module and a result output module. According to the method, systematic evaluation and optimization of the hydraulic section performance of the sluice are realized, and the adaptive capacity of the sluice in a complex environment is enhanced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an intelligent analysis method and system for test results of a sluice hydraulic section model. Background Art

[0002] The hydraulic cross-section model of a sluice is a scaled model created through a physical model or computer simulation to study the behavior of water flow in and around the sluice. The model is usually scaled according to a certain ratio (such as 1:10 or 1:25) based on the design parameters and structural characteristics of the actual sluice, aiming to simulate and analyze the effect of water flow on the sluice structure, flow pattern changes, and flood discharge capacity of the sluice. The purposes of conducting model test analysis of sluice hydraulic sections include: evaluating the design of the sluice, through which the structural design of the sluice can be verified to ensure that it has sufficient stability and safety under different flow patterns and flow conditions; optimizing energy dissipation measures, testing the rationality of the design of different energy dissipators (such as energy dissipation pools, energy dissipation retaining walls, etc.), and proposing effective energy dissipation and anti-scouring measures to prevent scouring and erosion in the downstream; studying flood discharge capacity, analyzing the flood discharge capacity of the sluice at different openings, obtaining the flow coefficient, and evaluating its performance under extreme water level conditions to ensure that the sluice operates safely and reliably during floods; ensuring safe operation, by understanding the impact of water flow on the sluice, exploring safe and efficient operation methods, evaluating various states during the operation of the sluice, and reducing the risk of failures; improving engineering reliability, through intelligent analysis of test data, improving the scientific nature of sluice performance evaluation, and improving the reliability and scientific decision-making ability of engineering projects. However, existing technical data show that physical model tests are usually limited by environmental factors (such as flow rate, temperature, etc.), which may lead to unstable or inconsistent experimental results; traditional test data processing and analysis methods mostly rely on manual calculations, which are not intelligent enough and have difficulty processing large amounts of data, resulting in potential features being ignored; traditional test systems may not be able to monitor water flow changes in real time, and may not respond to emergencies in a timely manner, making it difficult to make effective adjustments and interventions.

[0003] Prior art 1, application number: CN202010941875.7 discloses a visualization processing method for the test results of hydraulic and river engineering models. The sand-laying elevation of the hydraulic or river engineering dynamic bed model is determined according to the elevation of the prototype riverbed. After the riverbed in the dynamic bed model area is sand-laid, the measurement origin is defined and the topographic measurement points are arranged in a grid shape. Then, the riverbed elevation of each measurement point is measured using topographic measurement instruments and equipment. The water flow mathematical model terrain processing method is used to realize the visualization, convenient processing and statistical analysis of the test results of hydraulic and river engineering models. Although it solves the technical problems that the conventional hydraulic and river engineering model scouring and silting test results processing methods are complex, the visibility of the scouring and silting changes in the riverbed plane is poor, and the terrain longitudinal and cross-sectional diagrams are usually limited by the position of the measuring section, and the terrain profile diagrams at any position cannot be drawn, so the statistics of scouring and silting are difficult and the accuracy is poor. However, due to the limitations of riverbed environmental factors (such as flow, temperature, etc.), the experimental results may be unstable or inconsistent.

[0004] Prior art 2, application number: CN202310999978.2 discloses a high-precision hydraulic model test method and system, which constructs a digital model through the demand information of the water conservancy project and the structural information of the water conservancy and hydropower buildings; determines the scale of the digital model according to the principle of model similarity, obtains the detailed structure and overall outline size of the water conservancy project building, and after obtaining the determined detailed structure and overall outline size, selects the test site using the overall size of the determined digital model, and solidifies the model base platform; uses 3D printing equipment to print the digital model to obtain the digital model corresponding to the digital model; sets the experimental conditions required for the hydraulic model test for the digital model, obtains the experimental data, and analyzes the experimental data to obtain the optimized hydraulic design and scheduling operation plan. The system includes modules corresponding to the method steps. Although it has the characteristics of high model accuracy, high production efficiency, and recyclable materials; but relying too much on digital models leads to high requirements for sample data accuracy and increases the processing time of sample data.

[0005] Prior art three, application number: CN202311446380.7 discloses a wave model test method for mushroom-head drainage engineering. Aiming at the existing problem that the judgment can only be made by relying on the personal experience of technicians and is prone to errors, the following scheme is proposed, which includes the following steps: through wave mathematical model test, wave mathematical model calculation is performed on the drainage outlet area, and the design wave elements corresponding to each design water level are provided to provide a basis for engineering design. Before the drainage head structure is placed, wave simulation is performed, and then the drainage head structure and part of the drainage tunnel structure are simulated in the laboratory pool according to the specific location, elevation and permeability characteristics of the actual project. Through the wave local overall physical model test, the pressure of the drainage head structure under the combined action of waves and drainage and the stability of the bottom protection stone around the drainage outlet under different working conditions are determined, providing a basis for designing the structural strength of the drainage head structure. However, the experimental data processing and analysis methods mostly rely on manual calculations, which are not intelligent enough and difficult to process large amounts of data.

[0006] At present, the existing technologies 1, 2 and 3 cannot fully evaluate the performance of the sluice gate, key features are easily missed, and the scientificity and effectiveness of the sluice gate in the design, maintenance and operation process cannot be effectively improved. Therefore, the present invention provides an intelligent analysis method and system for the test results of the hydraulic section model of the sluice gate, which can extract the key features in the test data of the sluice section model and realize the intelligent evaluation and analysis of the sluice gate reconstruction and expansion project. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides an intelligent analysis method and system for the test results of a sluice hydraulic section model, comprising the following steps:

[0008] Build a physical model of the sluice section and use data acquisition equipment to record key data in real time during the test; clean the collected original key data, remove noise and outliers, and format the original key data according to time series to obtain the first data set;

[0009] Using principal component analysis, key features are extracted from the first data set to identify parameters that have a significant impact on the performance of the sluice gate. Based on the extracted key features, a sluice gate operation model is constructed, and the key features are used to parameterize flow and pressure related variables.

[0010] The extracted key features are associated with the sluice performance indicators, and a data-driven evaluation model based on machine learning is trained. The sluice performance indicators include flood discharge capacity and flow connection at different openings, sluice operation mode and flow coefficient; the evaluation model calculation results are compared with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the evaluation model calculation results and suggestions are provided.

[0011] Optionally, the process of obtaining the first data set includes the following steps:

[0012] Install various sensors related to data acquisition equipment at key locations of the physical model of the sluice section, including water level sensors, flow meters and pressure sensors;

[0013] Receive key data collected by each data collection device and use low-pass filter to remove high-frequency noise, calculate the deviation of each data point from the mean value of key data, and regard key data exceeding the preset threshold as outliers, and remove the key data corresponding to the outliers; organize the cleaned key data into a time series format containing timestamps and corresponding values ​​of each key parameter;

[0014] The processed key data is converted into a data frame using the pandas library to generate the first data set.

[0015] Optionally, the process of installing various sensors associated with the data acquisition device includes the following steps:

[0016] Identify key locations, including the entrance, exit, and different levels within the sluice structure; analyze historical monitoring data, fluid dynamics simulation results, and previous test results for each key location to determine the relative importance of each location in the overall sluice performance, and determine the weight coefficient setting criteria based on water flow characteristics, impact on the overall sluice performance, and historical data performance;

[0017] A weighted scoring method is used to assign a weight coefficient to each key position, and the weight coefficient is between 1 and 10. According to the weight coefficient of the key position, the type and quantity of the installed data sampling equipment are obtained;

[0018] After the data collection equipment is deployed, it is initialized, and regular data collection and monitoring are carried out. The data is combined with the weight coefficient to analyze the performance of each key location.

[0019] Optionally, key data includes water level, flow rate and pressure information.

[0020] Optionally, the process of building a sluice operation model includes the following steps:

[0021] Determine the input variables and output variables required in the sluice gate operation model, extract key features from the first data set, combine the key features into a multidimensional feature set by multiplication, and form new key features;

[0022] Develop mathematical relationships between input variables, pressures, and new key features;

[0023] The sluice operation model is fitted using the training data divided from the first data set, and the parameters of the sluice operation model are estimated using the least squares method; after fitting, the performance of the sluice operation model is evaluated through cross-validation; and the flow and pressure are calculated in real time based on the new input data.

[0024] Optionally, after constructing the sluice operation model, computational fluid dynamics software is used to simulate the water flow behavior and evaluate the flood discharge capacity and flow connection at different openings.

[0025] Optionally, particle swarm optimization is used to optimize the sluice operation mode, calculate the flow coefficient and perform statistical analysis to obtain experimental values ​​of the flow coefficient under different operating conditions.

[0026] Optionally, the process of training a data-driven machine learning-based evaluation model consists of the following steps:

[0027] According to the time series, the value of each key feature is matched with the value of the corresponding sluice performance indicator; all time series data are combined into a data frame, with the horizontal axis being the time point and the vertical axis being the key features and performance indicators; the data frame is randomly divided into a training set and a test set;

[0028] The model is trained using the training set, with the key features in the training set as input independent variables and the sluice performance indicators as target output variables, and the random forest algorithm is used for training;

[0029] After training, the accuracy of the model is evaluated through the test set, and the regression error indicators mean square error and root mean square error are used to evaluate the model; if the model effect does not meet the standard, grid search is used to tune the hyperparameters.

[0030] Optionally, the process of matching the value of each key feature with the value of the corresponding sluice performance indicator comprises the following steps:

[0031] Unify the time stamps of key features and corresponding sluice performance indicators into a unified format;

[0032] Generate a time series by the interval between the minimum and maximum timestamps, creating a list of all possible time points;

[0033] For each time point, the key features and sluice performance indicators are outer joined to form a complete data frame; if a time point is missing in a certain data set, the missing value is filled using interpolation.

[0034] The present invention provides an intelligent analysis system for test results of a sluice hydraulic section model, comprising:

[0035] The data acquisition module is responsible for building a physical model of the sluice section and using data acquisition equipment to record key data in the test process in real time. The key data includes water level, flow rate and pressure information; cleaning the collected original key data, removing noise and outliers, and formatting the original key data according to the time series to obtain the first data set;

[0036] The data analysis module is responsible for extracting key features from the first data set using principal component analysis and identifying parameters that have a significant impact on the performance of the sluice gate. Based on the extracted key features, a sluice gate operation model is constructed, and the key features are used to parameterize flow and pressure-related variables. Computational fluid dynamics software is used to simulate water flow behavior and evaluate the flood discharge capacity and flow pattern connection under different openings. Particle swarm optimization is used to optimize the sluice gate operation mode, calculate the flow coefficient and perform statistical analysis to obtain the experimental values ​​of the flow coefficient under different operating conditions.

[0037] The result output module is responsible for associating the extracted key features with the sluice performance indicators and training the data-driven evaluation model based on machine learning. The sluice performance indicators include the flood discharge capacity and flow connection under different openings, the sluice operation mode and the flow coefficient; comparing the calculation results of the evaluation model with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the calculation results of the evaluation model and suggestions are provided.

[0038] The present invention provides accurate raw data for subsequent analysis by building a physical model of the sluice section and using data acquisition equipment to monitor key parameters such as water level, flow rate and pressure in real time; eliminates noise and outliers during the cleaning process to ensure the reliability and validity of the data; formats the data into a time series format suitable for analysis so that subsequent analysis can use standardized and consistent data sets; lays a solid data foundation for downstream analysis to ensure that the data on which the model relies has sufficient accuracy and integrity; improves data quality through data cleaning and processing to avoid data problems affecting subsequent analysis results. Principal component analysis is used to extract key features that affect sluice performance to streamline the data set and enhance the interpretability and predictive power of the model; a sluice operation model is constructed based on the extracted features to parameterize variables such as flow and pressure to reflect the complexity of water flow behavior; computational fluid dynamics (CFD) software is used to perform flow simulation, evaluate the flood discharge capacity and flow state of different openings, and optimize the sluice operation mode in combination with the particle swarm optimization algorithm to obtain statistical data on the flow coefficient. Scientific decision support: Through in-depth analysis of flow behavior, the operating conditions of the sluice are optimized, providing scientific basis for engineers, enhancing the scientificity and rationality of sluice design and operation; Improving efficiency and safety: The simulation results help identify the optimal flow pattern and opening setting, reduce the risk of scouring, and improve the overall efficiency and operational safety of the sluice. Model training and evaluation: Combine the extracted key features with the sluice performance indicators (such as flood discharge capacity and flow coefficient), use machine learning methods to train data-driven evaluation models, so that they can make predictions under given conditions; compare the model calculation results with actual test data, verify the accuracy of the model, and adjust the model parameters to achieve continuous improvement; use the verified evaluation model to monitor and evaluate the sluice status in real time and provide scientific operation suggestions. Intelligent management and decision-making: Establish a decision support system through machine learning to make the management of the sluice more intelligent, thereby reducing human intervention and improving management efficiency; Improve accuracy and response speed: Improve the accuracy of predictions through model verification and adjustment, ensure timely response to changes and potential risks in actual operation, and enhance the safety of sluice operation.

[0039] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0040] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0042] Figure 1 It is a flow chart of the intelligent analysis method of the test results of the hydraulic section model of the sluice in Example 1 of the present invention;

[0043] Figure 2 FIG. 2 is a diagram of the process of obtaining the first data set in Embodiment 2 of the present invention;

[0044] Figure 3 A process diagram of installing various sensors related to a data acquisition device in Embodiment 3 of the present invention;

[0045] Figure 4 A process diagram of constructing a sluice operation model in Embodiment 4 of the present invention;

[0046] Figure 5 A process diagram of extracting key features from a first data set in Embodiment 5 of the present invention;

[0047] Figure 6 This is a process diagram for fitting the sluice operation model in Example 6 of the present invention;

[0048] Figure 7 This is a process diagram of simulating water flow behavior using computational fluid dynamics software in Example 7 of the present invention;

[0049] Figure 8 A process diagram of training a data-driven evaluation model based on machine learning in Example 8 of the present invention;

[0050] Fig. 9 A process diagram of matching the value of each key feature with the value of the corresponding sluice performance index in Embodiment 9 of the present invention;

[0051] Fig.10 It is a block diagram of the intelligent analysis system for the test results of the sluice hydraulic section model in Example 10 of the present invention. DETAILED DESCRIPTION

[0052] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0053] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. The singular forms of "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0054] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0055] Example 1: Figure 1 As shown, an embodiment of the present invention provides an intelligent analysis method for the test results of a sluice hydraulic section model, comprising the following steps:

[0056] S100: Build a physical model of the sluice section, use data acquisition equipment to record key data in the test process in real time, the key data includes water level, flow rate, pressure and other information; clean the collected original key data, remove noise and outliers, format the original key data according to time series, and obtain the first data set;

[0057] S200: Use principal component analysis to extract key features from the first data set and identify parameters that significantly affect the performance of the sluice gate. Based on the extracted key features, build a sluice gate operation model and use the key features to parameterize related variables such as flow and pressure. Use computational fluid dynamics software to simulate water flow behavior and evaluate the flood discharge capacity and flow pattern connection under different openings. Use particle swarm optimization to optimize the sluice gate operation mode, calculate the flow coefficient and perform statistical analysis to obtain the experimental values ​​of the flow coefficient under different operating conditions.

[0058] S300: Associate the extracted key features with the sluice performance indicators and train a data-driven evaluation model based on machine learning. The sluice performance indicators include the flood discharge capacity and flow connection at different openings, the sluice operation mode and the flow coefficient; compare the evaluation model calculation results with the actual test data to verify the accuracy of the evaluation model. According to the verification results, adjust the evaluation model parameters and feedback them to the evaluation model to achieve continuous improvement; evaluate the sluice status in real time and provide suggestions based on the evaluation model calculation results.

[0059] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, a physical model of the sluice section is first constructed, and the key data of the test process is recorded in real time using data acquisition equipment, and the key data include information such as water level, flow rate and pressure; the collected original key data are cleaned, noise and outliers are removed, and the original key data are formatted according to time series to obtain a first data set; secondly, principal component analysis is used to extract key features from the first data set to identify parameters that have a significant impact on the performance of the sluice; based on the extracted key features, a sluice operation model is constructed, and the key features are used to parameterize related variables such as flow and pressure; computational fluid dynamics software is used to simulate water flow behavior and evaluate different The flood discharge capacity and flow pattern connection under different openings; particle swarm optimization is used to optimize the sluice operation mode, the flow coefficient is calculated and statistically analyzed, and the test values ​​of the flow coefficient under different operating conditions are obtained; finally, the extracted key features are associated with the sluice performance indicators, and a data-driven evaluation model based on machine learning is trained. The sluice performance indicators include flood discharge capacity and flow pattern connection under different openings, sluice operation mode and flow coefficient; the evaluation model calculation results are compared with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the evaluation model calculation results and suggestions are provided. Step S100 of the above scheme provides accurate raw data for subsequent analysis by building a physical model of the sluice section and using data acquisition equipment to monitor key parameters such as water level, flow rate and pressure in real time; removes noise and outliers during the cleaning process to ensure the reliability and validity of the data; formats the data into a time series format suitable for analysis so that subsequent analysis can use standardized and consistent data sets; lays a solid data foundation for downstream analysis and ensures that the data on which the model relies has sufficient accuracy and integrity; improves data quality through data cleaning and processing to avoid data problems affecting subsequent analysis results. Step S200 uses principal component analysis to extract key features that affect sluice performance, so as to streamline the data set and enhance the interpretability and predictive ability of the model; constructs a sluice operation model based on the extracted features to parameterize variables such as flow and pressure to reflect the complexity of water flow behavior; applies computational fluid dynamics (CFD) software to perform flow simulation, evaluates the flood discharge capacity and flow state of different openings, and optimizes the sluice operation mode in combination with the particle swarm optimization algorithm to obtain statistical data of the flow coefficient. Scientific decision support: Through in-depth analysis of flow behavior and optimization of the operating conditions of the sluice, scientific basis is provided for engineers to enhance the scientificity and rationality of sluice design and operation; Improve efficiency and safety: simulation results help identify the optimal flow pattern and opening setting, reduce the risk of scour, and improve the overall efficiency and operational safety of the sluice.Step S300: Model training and evaluation. Combine the extracted key features with the performance indicators of the sluice gate (such as flood discharge capacity and flow coefficient), use machine learning methods to train the data-driven evaluation model so that it can make predictions under given conditions; compare the model calculation results with the actual test data, verify the accuracy of the model, and adjust the model parameters to achieve continuous improvement; use the verified evaluation model to monitor and evaluate the sluice gate status in real time and provide scientific operation suggestions. Intelligent management and decision-making: Establish a decision support system through machine learning to make the management of the sluice gate more intelligent, thereby reducing human intervention and improving management efficiency; Improve accuracy and response speed: Improve the accuracy of predictions through model verification and adjustment, ensure timely response to changes and potential risks in actual operations, and enhance the safety of sluice gate operation.

[0060] In summary, this embodiment aims to achieve a systematic evaluation and optimization of the hydraulic section performance of the sluice through the above three steps, and enhance the adaptability of the sluice in complex environments. The specific significance is reflected in: improving the safety of the project: through intelligent analysis, timely discovering and solving potential problems to ensure the safety of the sluice under extreme basin conditions; promoting scientific decision-making: making scientific decisions based on the results of data analysis, optimizing the operation mode of the sluice, and improving its adaptability to environmental changes; saving resources and costs: improving the efficiency of sluice operation and maintenance and reducing human errors, reducing maintenance and operating costs, and maximizing resource utilization; promoting technological progress: integrating advanced data analysis and optimization algorithms, which will help promote technological innovation and development in the field of hydraulic engineering and improve the overall intelligence level of water conservancy projects; achieving comprehensive, accurate, and real-time evaluation and optimization of sluices, and enhancing their role in practical applications.

[0061] Example 2: Figure 2 As shown, based on Example 1, the process of obtaining the first data set provided in this embodiment of the present invention includes the following steps:

[0062] S101: Install various sensors related to data acquisition equipment at key locations of the physical model of the sluice section, including water level sensors, flow meters, and pressure sensors;

[0063] S102: Receive the key data collected by each data collection device and use a low-pass filter to remove high-frequency noise, calculate the deviation of each data point from the mean value of the key data, and regard the key data exceeding the preset threshold as an outlier, and remove the key data corresponding to the outlier; organize the cleaned key data into a time series format including a timestamp and the corresponding values ​​of each key parameter;

[0064] S103: Use the pandas library to convert the processed key data into a data frame to generate a first data set.

[0065] The working principle and beneficial effects of the above technical solution are as follows: First, various sensors related to data acquisition equipment are installed at key positions of the physical model of the sluice section, including water level sensors, flow meters, and pressure sensors; secondly, the key data collected by each data acquisition device are received, and a low-pass filter is used to remove high-frequency noise, and the deviation of each data point from the mean of the key data is calculated, and the key data exceeding the preset threshold is regarded as an outlier, and the key data corresponding to the outlier is removed; the cleaned key data is sorted into a time series format containing a timestamp and the corresponding values ​​of each key parameter; finally, the processed key data is converted into a data frame using the pandas library to generate a first data set. Step S101 of the above solution installs sensors and obtains data diversity: by installing water level sensors, flow meters, and pressure sensors at key positions of the sluice section model, important parameters related to water flow behavior and sluice performance can be obtained in real time; sensors with reasonable layout can reflect different states and characteristics of water flow, ensuring the representativeness and comprehensiveness of the data; so that data changes can be recorded in real time during the test, which is helpful for subsequent analysis. Data accuracy and validity: High-quality sensor feedback can improve the accuracy and validity of data, laying a good foundation for subsequent analysis; Comprehensive understanding of sluice performance: The collection of multiple key parameters can provide necessary information for the comprehensive analysis of sluice performance and help understand how different parameters affect each other during operation. Step S102 Data cleaning, using a low-pass filter to remove high-frequency noise, can improve the quality of the signal and avoid erroneous analysis caused by noise; By calculating the deviation of each data point from the mean of the key data, judge and remove outliers that exceed the preset threshold to ensure the consistency and authenticity of the data; Integrate the cleaned data into a time series format of timestamps and key parameters to make it suitable for subsequent modeling and analysis; Data neatness: The cleaned data is neater, which improves the integration of the data and helps to improve the accuracy of the subsequent analysis model; Reliable basic data: Provides a set of verified data to ensure that the data foundation for subsequent analysis and model building is credible and available. Step S103: Data frame generation, using the Pandas library to build a data frame: The processed key data is converted into a data frame through the Pandas library, making data processing more efficient and intuitive; convenient for analysis and processing: the data frame format can effectively utilize various data processing and analysis tools of Pandas to quickly implement data screening, cleaning, aggregation and visualization, etc. Improve data analysis efficiency: the data frame is highly structured, making subsequent analysis, modeling and visualization simpler and more efficient; convenient for data sharing and recording: the generated data frame is not only convenient for subsequent analysis, but also convenient for recording and sharing with other researchers, providing a good data foundation for research in the field of hydraulic engineering.

[0066] In summary, the acquisition process of the entire first data set in this embodiment, through the above steps, gradually captures the original, unprocessed data by installing sensors, cleans the data to ensure the quality and validity of the data, and finally converts it into a high-quality data frame for analysis and modeling. This not only ensures the accuracy and completeness of the data, but also provides strong support for subsequent analysis and subsequent optimization of sluice performance.

[0067] Example 3: Figure 3 As shown, based on Example 2, the process of installing various sensors related to the data acquisition device provided in the embodiment of the present invention includes the following steps:

[0068] S1011: Identify key locations, including the entrance, exit and different levels within the sluice structure; analyze historical monitoring data, fluid dynamics simulation results and previous test results for each key location to determine the relative importance of each location in the overall sluice performance, and determine the weight coefficient setting criteria based on water flow characteristics, impact on the overall sluice performance and historical data performance;

[0069] Among them, water flow characteristics include the degree of influence of flow rate, flow velocity and water level changes; the impact on the overall performance of the sluice: for example, the impact on the safety and operating efficiency of the sluice; historical data performance: the performance collected in past monitoring, such as failure frequency, maintenance records, etc.;

[0070] S1012: Use a weighted scoring method to assign a weight coefficient to each key position, where the weight coefficient is between 1 and 10. According to the weight coefficient of the key position, the type and quantity of installed data sampling equipment are obtained;

[0071] Among them, the high-weighted position: the entrance area of ​​the sluice gate; sensor configuration, flow meter; quantity: 3 flow meters; type: electromagnetic flow meters arranged at different depths, which can monitor the water flow speed and direction in real time and capture the hierarchical changes of the water flow;

[0072] Water level sensor, quantity: 2 water level sensors; location: one is installed on the water surface to monitor the inlet water level, and the other is arranged at a certain depth under the water to obtain the water level changes before and after entering the sluice;

[0073] Pressure sensor, quantity: 2 pressure sensors, location: installed in the upstream and downstream of the water flow to compare the pressure difference between input and output, and help analyze the water flow behavior and the stress of the sluice gate;

[0074] Medium weighted location: overflow port on the side of a sluice gate; Sensor configuration, flow meter, quantity: 1 flow meter type: Select an ultrasonic flow meter that can be mounted in a side position and measure the overflow rate. ;

[0075] Water level sensor: 1 water level sensor; Location: Installed at the water level position of the overflow to monitor the changes in water level;

[0076] Low weight location: backflow area of ​​the sluice gate, sensor configuration: water level sensor; quantity: 1 water level sensor; type: choose a simple piezoelectric water level sensor, installed in the backflow area to monitor occasional water level changes;

[0077] S1013: After the data collection equipment is deployed, it is initialized, and regular data collection and monitoring are performed. The data is combined with the weight coefficient to analyze the performance of each key location.

[0078] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the key positions are first determined, including the entrance, the exit and different levels in the sluice structure; historical monitoring data, fluid dynamics simulation results and results of previous tests are analyzed for each key position to determine the relative importance of each position in the overall sluice performance, and the weight coefficient setting standard is determined according to the water flow characteristics, the impact on the overall performance of the sluice and the historical data performance; secondly, a weighted scoring method is used to assign a weight coefficient to each key position, and the weight coefficient is between 1 and 10. According to the weight coefficient of the key position, the type and quantity of the installed data sampling equipment are obtained; finally, after the data acquisition equipment position is arranged, it is initialized, and regular data acquisition and monitoring are carried out, and the data is combined with the weight coefficient to analyze the performance of each key position. Step S1011 of the above scheme determines the key positions and weight coefficient settings. By analyzing the different levels of the entrance, exit and inside the sluice gate, the most important monitoring positions are identified, which can ensure the core area of ​​tracking water flow characteristics and sluice performance; combining historical monitoring data, fluid dynamics simulation and previous test results to ensure comprehensive and scientific data decision-making; comprehensive analysis of water flow characteristics, overall performance impact and historical performance can improve the accuracy of the importance assessment of each position. Scientific decision-making basis: provide data-driven decision support to enhance the understanding of sluice design and operation; optimize resource allocation: clarify which positions need more attention and data collection, which helps to achieve efficient utilization in equipment resource allocation and avoid unnecessary complexity and cost waste. Step S1012 uses the weighted scoring method to allocate weight coefficients. Through the weighted scoring method, a weight coefficient (1-10 points) is allocated to each key position based on the relative importance of the position in the system, so that each position can take into account its importance in the monitoring strategy; according to the weight coefficient, the type and number of data sampling equipment required to be deployed at each position are reasonably determined to ensure the pertinence and effectiveness of information collection. Rational configuration: Ensure that high-weighted positions are equipped with multiple types of high-precision sensors to capture comprehensive data, while medium and low-weighted positions limit the number of sensors according to their importance to optimize resource utilization; Enhance monitoring effect: Improve the data quality and efficiency of the entire monitoring system, making the collected data more valuable for reference. Step S1013 Initialization and data collection after the sensor is deployed. After the sensor is deployed, all sensors are initialized to ensure that they are operating normally, confirm the device connection and functional status, and avoid technical failures in data collection; Carry out regular data collection, combine the weight coefficients of different positions, analyze the performance of each key position, and reflect the water flow conditions in a timely manner.Real-time monitoring and evaluation can track the operating status of the sluice gate over a long period of time, ensuring that monitoring data under different conditions are available in a dynamic environment; effective decision support combines analytical data with weight coefficients to help formulate scientific operation and maintenance plans, response strategies and subsequent technical improvements; real-time monitoring of safety risk management ensures that potential abnormal situations can be discovered in a timely manner, improving the safety and reliability of the sluice gate and reducing the risk of failure.

[0079] In summary, this embodiment has established a scientific and systematic sluice monitoring solution and achieved data-driven decision-making: ensuring that scientific data can be used as the basis in all decision-making processes to improve management and operational efficiency. It has achieved improved monitoring accuracy and scope: by reasonably configuring sensors, sufficient data can be obtained at different weight positions to enhance the understanding and control of the entire sluice system. It has achieved resource optimization: while effectively utilizing resources, it has reduced the risk of technology implementation and waste of resources, ensuring the economy of system operation. It has achieved continuous improvement capabilities: providing data support for the subsequent monitoring and maintenance of the sluice, making the operation and maintenance process more forward-looking and predictable, thereby continuously improving the overall performance and safety of the sluice.

[0080] Example 4: Figure 4 As shown, based on Example 1, the process of constructing a sluice operation model provided in the embodiment of the present invention includes the following steps:

[0081] S201: Determine the input variables and output variables required in the sluice operation model, extract key features from the first data set, combine the key features into a multidimensional feature set by multiplication, and form new key features;

[0082] Among them, the input variables include flow rate, opening, pressure and characteristic parameters, and the output variables include estimated pressure and flow rate;

[0083] S202: Establishing a mathematical relationship between input variables, pressure and new key features, and forming a mathematical equation according to the type of the sluice operation model;

[0084] y=β0+β1x1+β2x2+…+β k x k ;

[0085] Where y is the output (such as flow or pressure), x is the input feature, and β is the model parameter;

[0086] S203: Fitting the sluice operation model using the training data divided from the first data set, and estimating the parameters of the sluice operation model using the least squares method; after fitting, evaluating the performance of the sluice operation model through cross-validation; and calculating the flow and pressure in real time according to the new input data.

[0087] The working principle and beneficial effects of the above technical solution are as follows: This embodiment first determines the input variables and output variables required in the sluice operation model, extracts key features from the first data set, and combines the key features into a multidimensional feature set by multiplication to form new key features; wherein, the input variables include flow, opening, pressure and characteristic parameters, etc., and the output variables include estimated pressure and flow, etc.; secondly, a mathematical relationship between the input variables, pressure and the new key features is established, and a mathematical equation is formed according to the type of the sluice operation model; finally, the sluice operation model is fitted using the training data divided from the first data set, and the parameters of the sluice operation model are estimated using the least squares method; after fitting, the performance of the sluice operation model is evaluated by cross-validation; and the flow and pressure are calculated in real time according to the new input data. Step S201 of the above scheme extracts and combines key features. By determining the input and output variables required in the sluice operation model, the scope and objectives of the study are clarified, ensuring the systematic and targeted nature of subsequent analysis and modeling; extracting key features from the first data set can reduce the input dimension of the model and focus on the parameters that are important for the performance of the sluice, thereby improving the efficiency of modeling; combining key features by multiplication and other methods to form a new multidimensional feature set can capture the nonlinear relationship and interaction between input variables, which helps improve the prediction ability of the model. Improve the accuracy of the model: through effective feature extraction and combination, high-quality input is provided to the model, which enhances the learning ability of the model; optimize the modeling process: ensure that the model captures important physical characteristics when processing inputs, reduces the influence of redundant variables, and improves the flexibility of the model by combining features, which helps reveal potential water flow behavior characteristics. Step S202 establishes a mathematical relationship. By establishing a mathematical relationship between input variables and output variables (such as a linear regression equation), a clear theoretical framework is provided for the model, and the causal relationship between variables is clarified; the model parameterization is formalized so that specific values ​​can be obtained through data fitting, and the β parameter is used to represent the degree of influence of each variable on the output, which also provides interpretability for subsequent analysis. Quantitative analysis: Establishing a clear mathematical relationship can quantitatively analyze the real system and promote the understanding and prediction of the operation characteristics of the sluice gate; Model reference: The formed mathematical equation provides a basis for future prediction and optimization, and the output variable can be quickly calculated through the equation. Step S203 Model fitting and verification By using the training data divided from the first data set, the sluice gate operation model is fitted using the least squares method and other techniques to obtain appropriate model parameters (β values); the model is evaluated by means such as cross-validation to ensure that the model has good generalization ability on unseen data and avoid overfitting; by constructing a good model, the flow and pressure can be calculated at one time according to the new input data to quickly obtain the prediction results.Improve reliability: Through the systematic model fitting and verification process, the established model is ensured to have high accuracy and reliability, providing a solid foundation for practical application; Practical application: The ability to calculate flow and pressure in real time enables the model to be applied in actual sluice operations for dynamic monitoring and regulation, thereby improving the management efficiency and safety of the sluice.

[0088] In summary, in the entire process of constructing the sluice operation model, steps S201, S202 and S203 of this embodiment all provide different degrees of guarantee for the effectiveness and reliability of the final model; through reasonable feature selection and combination, the establishment of clear mathematical relationships, and the systematic fitting and verification process, the model can accurately reflect the flow behavior and realize the intelligent and efficient management of the sluice, thus achieving the important goal of optimizing water resources management and improving the ability of the sluice to cope with various changes in conditions.

[0089] Example 5: Figure 5 As shown, based on Example 4, the process of extracting key features from the first data set provided in the embodiment of the present invention includes the following steps:

[0090] S2011: Obtain all relevant variables in the first data set, including flow rate, opening set pressure, etc.; check whether there are missing values ​​in the data set, and use the mean to fill in the missing values; standardize the variables, subtract the mean and divide by the standard deviation, so that each variable has a mean of 0 and a standard deviation of 1;

[0091] S2012: Calculate the covariance matrix between the variables in the first data set after standardization. For n variables, the covariance matrix is ​​an n×n symmetric matrix, in which each element represents the covariance between the corresponding variables; perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors. The eigenvalues ​​represent the importance of each principal component, and the eigenvectors represent the directions of the principal components; select the principal components according to the size of the eigenvalues, and select the first k principal components with larger eigenvalues;

[0092] S2013: Calculate the proportion of the eigenvalues ​​corresponding to each principal component to the total eigenvalues, as well as the cumulative variance explanation rate, and select the principal components with a cumulative variance explanation rate of more than 90%; Based on the selected principal components, use the eigenvectors to reduce the dimension of the original data to a new space to obtain a new data set in which the variables are the principal components; Based on the results of the principal component analysis, extract the key features that affect the performance of the sluice.

[0093] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, all relevant variables in the first data set are first obtained, including flow rate, opening set pressure, etc.; whether there are missing values ​​in the data set is checked, and the missing values ​​are processed by filling the mean; the variables are standardized, the mean is subtracted and divided by the standard deviation, so that each variable has a mean of 0 and a standard deviation of 1; secondly, the covariance matrix between the variables in the first data set after standardization is calculated. For n variables, the covariance matrix is ​​an n×n symmetric matrix, in which each element represents the covariance between the corresponding variables; the covariance matrix is ​​analyzed by eigenvalues. Solve, get eigenvalues ​​and eigenvectors, eigenvalues ​​represent the importance of each principal component, and eigenvectors represent the direction of the principal component; select the principal component according to the size of the eigenvalue, and select the principal components with the largest eigenvalues ​​in the first k; finally, calculate the proportion of the eigenvalues ​​corresponding to each principal component to the total eigenvalues, as well as the cumulative variance explanation rate, and select the principal component with a cumulative variance explanation rate of more than 90%; according to the selected principal component, use the eigenvector to reduce the dimension of the original data to a new space to obtain a new data set, in which the variables are the principal components; according to the results of the principal component analysis, extract the key features that affect the performance of the sluice. Step S2011 of the above scheme is data preparation and standardization, extracting all relevant variables (such as flow, opening, pressure, etc.) in the first data set to lay the foundation for subsequent analysis; handle missing values ​​by filling in the mean to maintain the integrity of the data and reduce the deviation and error caused by missing values; standardize the variables to ensure that all variables are on the same scale (mean is 0, standard deviation is 1) to avoid different dimensions and ranges affecting the results of subsequent analysis. Improve data quality: Missing value processing and standardization make the data set more rigorous when analyzing, reducing errors caused by inconsistent data; Balance variable importance: Standardization ensures that all features contribute equally to principal component analysis, so that the analysis results will not be dominated by the large numerical range of some variables, making the analysis results more reliable and reasonable. Step S2012 calculates the covariance matrix and eigenvalue decomposition. By calculating the covariance matrix between the variables of the standardized data set, the linear relationship between the variables and the degree of mutual influence are measured; the covariance matrix is ​​decomposed by eigenvalue to obtain the importance (eigenvalue) and direction (eigenvector) of each principal component; select the first k principal components with larger eigenvalues ​​according to the size of the eigenvalue, and retain the principal component with the largest contribution to the data variability. Understand variable relationships: Through the covariance matrix, you can better understand the correlation between variables in the data and its impact on the overall structure; Data dimensionality reduction: By selecting eigenvalues, the most important information is extracted, and noise and redundant features are filtered out, thereby effectively reducing the dimension of the data, which can improve the calculation efficiency and model performance for subsequent analysis.Step S2013 variance explanation and dimensionality reduction, calculate the proportion of the total variance accounted for by each principal component, and the cumulative variance explanation rate, evaluate the effectiveness and information retention of the selected principal components; select the principal components that can explain more than 90% of the total variance based on the variance explanation rate, ensuring that the new data set can reflect most of the information of the original data; map the original data to the principal component space through the eigenvector, and create a new data set, in which the variables are the selected principal components. Information retention: Selecting principal components with a cumulative variance explanation rate higher than 90% can ensure that important information is retained and increase the accuracy of the model; Effectiveness improvement: The new data set obtained after dimensionality reduction can simplify subsequent analysis and modeling work, reduce computational costs and complexity, avoid overfitting, and increase the generalization ability of the model; Identify key features: The principal components extracted through this process help identify key features that have a significant impact on the performance of the sluice, providing an important basis for subsequent modeling and decision-making.

[0094] In summary, each step of this embodiment, from data preparation to feature extraction, is serving to extract parameters that have a significant impact on the performance of the sluice gate and improve the accuracy and effectiveness of the model; through strict data processing, reasonable statistical analysis and effective dimensionality reduction practices, it ultimately provides solid support for the optimization of sluice gate operation.

[0095] Example 6: Figure 6 As shown, based on Example 4, the process of fitting the sluice operation model provided in the embodiment of the present invention includes the following steps:

[0096] S2031: Divide the first data set into a training set and a test set, and extract input variables and corresponding output variables from the first data set;

[0097] S2032: Select mean square error (MSE) as the loss function to measure the difference between the model prediction value and the true value, initialize parameter β, and calculate the prediction output under the current parameters. Calculate the current loss; update the parameters by gradient descent, repeat until the stopping criteria is met and the loss function converges;

[0098]

[0099] Among them, y i is the actual output, is the output predicted by the model, and n is the number of samples;

[0100]

[0101] Among them, β j is the learning rate, is the gradient of the loss function with respect to the parameters;

[0102] S2033: Use K-fold cross validation to evaluate the performance of the sluice operation model on different data sets. Divide the training set into K parts, use K-1 parts for training in turn, and use the remaining 1 part for validation; calculate the mean square error of the evaluation indicator for each round, and calculate the average performance indicator.

[0103] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the first data set is first divided into a training set and a test set, and input variables and corresponding output variables are extracted from the first data set; secondly, the mean square error MSE is selected as the loss function to measure the difference between the model prediction value and the true value, the parameter β is initialized, and the prediction output under the current parameter is calculated. Calculate the current loss; update the parameters by gradient descent, repeat until the stopping criteria is met and the loss function converges;

[0104]

[0105] Among them, y i is the actual output, is the output predicted by the model, and n is the number of samples;

[0106]

[0107] Among them, β j is the learning rate, is the gradient of the loss function with respect to the parameters;

[0108] Finally, K-fold cross validation is used to evaluate the performance of the sluice operation model on different data sets. The training set is divided into K parts, and K-1 parts are used for training in turn, and the remaining 1 part is used for verification; the mean square error of the evaluation index for each round is calculated, and the average performance index is calculated. Step S2031 of the above scheme is data set division and feature extraction. The first data set is divided into a training set and a test set. Usually, about 70%-80% of the data is used for model training, and the rest is used for model performance evaluation. This ensures the generalization ability of the model on unseen data; extract input variables (such as flow, opening, pressure, etc.) and corresponding output variables (such as predicted pressure and flow) from the first data set to provide necessary data support for subsequent modeling. Model verification: dividing the data set allows the performance of the model on unseen data to be evaluated after model training, helping to discover overfitting or deficiencies of the model; extracting useful input and output variables ensures that the training data of the model is closely related to the actual situation of the problem, thereby improving the effectiveness of the model. Step S2032 Loss function and parameter optimization Using mean square error (MSE) as the loss function can quantify the gap between the model prediction value and the true value, and provide a clear goal for model optimization; parameter initialization provides a starting point for model training preparation, and setting a reasonable initial value helps to converge quickly; by calculating the predicted output and loss under the current parameters, and using the gradient descent method to gradually update the parameters until the stop condition is met. Optimization accuracy: By minimizing the loss function, the fitting accuracy of the model can be effectively improved to ensure that the model can better map the relationship between the input variables and the output variables; the gradient descent method is used to enable the parameter update process to gradually approach the optimal solution, thereby improving the performance and stability of the model. Step S2033 Cross-validation evaluation further divides the training set into K parts, uses K-1 parts for training in turn, and the remaining 1 part is used as the validation set; calculates the mean square error (MSE) in each round of cross-validation, and obtains the average performance index for the evaluation results of K rounds. Cross-validation can effectively reduce the evaluation bias caused by specific data division, making the results of model evaluation more representative and stable. By observing the performance of the model on the validation set, we can better judge whether the model is overfitting or underfitting, and make parameter adjustments or model selection accordingly. The results of K-fold cross-validation provide a practical basis for subsequent model selection, parameter adjustment and optimization.

[0109] In summary, this embodiment provides a systematic methodology for the effective construction of the sluice operation model from data preparation to model optimization to performance evaluation. Each step has its important technical effect, ensuring that the solution can adapt to the complexity of the real environment and improve the generalization ability and reliability of the model; it provides a solid foundation for the prediction of sluice performance in practical applications and helps to improve the management and scheduling efficiency of water conservancy projects.

[0110] Example 7: Figure 7 As shown, based on Example 1, the process of simulating water flow behavior using computational fluid dynamics software provided in the embodiment of the present invention includes the following steps:

[0111] S204: Based on the actual structure of the sluice, use CAD software to establish a three-dimensional geometric model of the sluice, including water fluid channels, gates and related facilities; divide the computational grids of the fluid region and the solid region in the software; set the viscosity and density of water and the steady state or transient state of the flow; select the turbulence model k-ε to process the turbulence characteristics in the flow;

[0112] S205: setting the inlet, outlet and side wall boundary conditions, inputting specific flow and water level conditions, setting the gate opening and flood discharge mode; the solver performs calculations, confirms the time step and iteration number of the solution, starts the calculation, performs numerical simulation of the flow field, and monitors the convergence of the calculation;

[0113] S206: Output the result data, analyze the flow velocity, pressure distribution and water level changes, analyze the flow pattern connection and flood discharge capacity through the visualized flow velocity cloud map; evaluate the flow pattern changes under different openings, and record the flood discharge capacity and water flow behavior under each state.

[0114] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, first, according to the actual structure of the sluice, a three-dimensional geometric model of the sluice is established using CAD software, including water fluid channels, gates and related facilities; the calculation grid of the fluid area and the solid area is divided in the software; the viscosity and density of the water and the steady state or transient state of the flow are set; the turbulence model k-ε is selected to process the turbulent characteristics in the flow; secondly, the inlet, outlet and side wall boundary conditions are set, specific flow and water level conditions are input, and the gate opening and flood discharge mode are set; the solver performs calculations, confirms the time step and number of iterations for the solution, starts the calculation, performs numerical simulation of the flow field, and monitors the convergence of the calculation; finally, the result data is output, the flow velocity, pressure distribution and water level changes are analyzed, and the flow state connection and flood discharge capacity are analyzed through a visualized flow velocity cloud map; the flow state changes under different openings are evaluated, and the flood discharge capacity and water flow behavior under each state are recorded. In step S204 of the above scheme, the geometric model is created and the fluid area is meshed. The CAD software is used to create an accurate three-dimensional model based on the actual structure of the sluice, which can carefully reflect the spatial layout of the water fluid channel, gates and related facilities; the geometric model is divided into computational grids of the fluid area and the solid area to ensure that the computational area can fully capture the details and characteristics of the water flow, providing a basis for subsequent simulations. Improve accuracy: high-quality geometric models and meshing can effectively improve the accuracy of the calculation results, so that the simulation of flow behavior can more realistically reflect the actual situation; enhance flow analysis capabilities: through accurate model and fluid property (such as viscosity and density) settings, it lays the foundation for subsequent turbulence and flow pattern analysis, so that fluid characteristics can be more clearly presented. Step S205: Boundary condition setting and numerical solution. The boundary condition setting accurately sets the inlet, outlet and side wall boundary conditions, as well as the specific values ​​of flow and water level in the simulation, to ensure that the calculation process meets the actual operating conditions and uses the correct mode (such as gate opening and flood discharge method); Solver configuration and calculation: select the appropriate time step and number of iterations, start the calculation through the solver, perform numerical simulation of the flow field, and monitor the convergence of the calculation process in real time. Ensure the accuracy of the solution: reasonable boundary conditions and effective solution strategies ensure that the calculation can proceed smoothly, and at the same time, the simulation results can truly reproduce the actual water flow behavior; flow field dynamic capture: through the simulation of the flow field, more detailed fluid dynamic changes can be obtained, providing a comprehensive data basis for subsequent data analysis.Step S206 Result output and flow pattern analysis will output the simulation calculation results (such as flow velocity, pressure distribution, water level changes, etc.) in the form of data for subsequent analysis; Visualization of flow velocity cloud map: Data is presented as flow velocity cloud map and other graphics through visualization software, which can intuitively analyze the connection of flow pattern and flood discharge capacity; Flow pattern change evaluation: Record and analyze the flow pattern changes in the basin under different openings, and evaluate the characteristics of flood discharge capacity and water flow behavior; Flow characteristic evaluation: By analyzing the flow characteristics, evaluate whether the design and structure of the sluice meet the expected needs, and have an in-depth understanding of the management and control capabilities of the water flow; Basis for optimized design: Analysis of result data provides a scientific basis for future design improvements, which can help designers optimize sluices and related facilities to improve their flood discharge capacity and safety; Practical application value: Analysis of simulation data provides support for decision-making of engineering projects, ensures the safety and effectiveness of sluices, and provides a basis for environmental protection and rational use of water resources.

[0115] In summary, this embodiment can effectively simulate water flow behavior, conduct in-depth analysis and evaluation of sluice design, realize the transformation from theoretical design to practical application, and provide important support for ensuring the safe and efficient operation of water conservancy projects. Through CAD simulation, potential problems can be identified in the early stages of design, thereby effectively reducing risks and costs in the later stages of construction.

[0116] Example 8: Figure 8 As shown, based on Example 1, the process of training a data-driven evaluation model based on machine learning provided in this embodiment of the present invention includes the following steps:

[0117] S301: According to the time series, the value of each key feature is matched with the value of the corresponding sluice performance index; all the time series data are combined into a data frame, with the horizontal axis being the time point and the vertical axis being the key feature and the performance index; the data frame is randomly divided into a training set and a test set;

[0118] S302: using the training set to perform model training, taking the key features in the training set as input independent variables, and taking the sluice performance index as the target output variable, and using the random forest algorithm for training;

[0119] S303: After training, the accuracy of the model is evaluated through the test set, and the regression error indicators mean square error and root mean square error are used to evaluate the model; if the model effect does not meet the standard, grid search is used to tune the hyperparameters.

[0120] The working principle and beneficial effects of the above technical solution are as follows: This embodiment first matches the value of each key feature with the value of the corresponding sluice performance indicator according to the time series; combines all time series data into a data frame, with the horizontal axis representing the time point and the vertical axis representing the key features and performance indicators; randomly divides the data frame into a training set and a test set; secondly, uses the training set to train the model, takes the key features in the training set as input independent variables, and takes the sluice performance indicators as target output variables, and uses the random forest algorithm for training; finally, after training, evaluates the accuracy of the model through the test set, and uses the regression error indicators mean square error and root mean square error to evaluate the model; if the model effect does not meet the standard, uses grid search to perform hyperparameter tuning. Step S301 of the above scheme: Data preparation and construction of data frame, by matching the key features with the data of the sluice performance indicators in the time series, ensuring that the features at each time point correspond to the corresponding performance results, and ensuring the basis of data consistency and relevance; organizing the matched data into a systematic and structured data frame (DataFrame), so that the data of each feature and indicator can be easily accessed and processed; the data frame uses the time point as the index, the key features and performance indicators as the column, which is convenient for subsequent large-scale data operations; the data is divided into training set and test set by random sampling, ensuring that the training set contains enough information for model training, and the test set realizes the generalization ability of the verification model. Basic data preparation: lays a good foundation for subsequent modeling. The quality and structure of the data directly affect the learning effect of the model; improve the accuracy and reliability of the model: by dividing the training set and the test set, the model is avoided from overfitting, and the performance of the model on unseen data can be effectively evaluated to ensure the effectiveness of the model in practical applications. Step S302: Model training uses the random forest algorithm as the basis for the evaluation model. Random forest is an integrated learning method with strong fitting ability and can effectively handle nonlinear relationships and high-dimensional data. Using the data of the training set, the model can find the complex relationship between key features and sluice performance indicators by continuously adjusting the weights and structures of each decision tree of the model. During the training process, a mapping relationship is established between the input data (independent variables) and the output results (dependent variables), so that the model can learn the functional relationship from features to performance. Build an effective model: Train the model through a suitable algorithm so that it can understand the relationship between input features and sluice performance indicators, thereby improving the accuracy of predictions. Take advantage of the integrated method: Random forest reduces the bias and variance of the model by combining the results of multiple decision trees, and improves stability and robustness.Step S303: Model evaluation and hyperparameter tuning. The prediction ability of the model is evaluated through the test set. The performance of the model is quantitatively evaluated using the two regression error indicators, mean square error (MSE) and root mean square error (RMSE). Through the calculation of the evaluation indicators, the model is found to be deficient in some aspects, such as large prediction errors or poor performance under specific conditions. If the performance of the model does not meet expectations, the model's hyperparameters are comprehensively searched through the grid search method to find the best parameter combination to improve the performance of the model. Verify the practicality of the model: The evaluation process ensures that the model not only performs well on the training set, but also can accurately predict in actual applications, enhancing its feasibility. Continuously improve the model: By tuning the parameters, the model can be further optimized to better adapt to the data, thereby improving the accuracy of the prediction. This is an important link to achieve continuous improvement and optimization of the model.

[0121] In summary, the entire process of training the data-driven evaluation model based on machine learning in this embodiment, through scientific steps, not only realizes data integration, model construction, evaluation and optimization, but also ensures that the model can effectively reflect the performance of the sluice in practical applications; it provides strong data support and decision-making basis for the intelligent management of the sluice, and helps to improve the safety and efficiency of water conservancy projects.

[0122] Example 9: Fig. 9 As shown, based on Example 8, the process of matching the value of each key feature with the value of the corresponding sluice performance indicator provided in the embodiment of the present invention includes the following steps:

[0123] S3011: Unify the time stamps of key features and corresponding sluice performance indicators into a unified format;

[0124] S3012: Generate a time series by the interval between the minimum timestamp and the maximum timestamp, and create a list containing all possible time points;

[0125] S3013: For each time point, the key features and sluice performance indicators are outer joined to form a complete data frame; if a time point is missing in a certain data set, choose to use interpolation to fill in the missing value.

[0126] The working principle and beneficial effects of the above technical solution are as follows: First, the key features and the timestamps of the corresponding sluice performance indicators are unified in format in this embodiment; secondly, a time series is generated by the interval between the minimum timestamp and the maximum timestamp to create a list containing all possible time points; finally, for each time point, the key features and the sluice performance indicators are outer joined and merged to form a complete data frame; if a time point is missing in a certain data set, the missing value is supplemented by interpolation. Step S3011 of the above scheme: the key features and the timestamps of the corresponding sluice performance indicators are unified in format, and the timestamps in different data sources are unified into the same format to eliminate the alignment problem caused by inconsistent formats; the timestamp is converted into a date and time object (such as a datetime object) through a function or operation to facilitate subsequent data processing and comparison. Improve data quality: after the format is unified, it can ensure that the time data between different data sources can be correctly matched in the subsequent merging operation, reducing the data loss caused by format problems; facilitate subsequent operations: create a unified data set based on time, lay the foundation for subsequent data alignment and analysis, and reduce the errors that may be encountered in subsequent data processing. Step S3012 generates a time series by the interval between the minimum timestamp and the maximum timestamp, generates a continuous time series by using the minimum and maximum timestamps, creates a time point list according to the set time interval (for example, every 5 minutes, every hour); forms a standardized time index for subsequent data merging, alignment and interpolation. Ensure data integrity: Generating a complete time series can ensure that all possible time points are covered in the data frame, and the values ​​of key features and performance indicators in these time periods can be compared to avoid analysis deviations caused by missing data; facilitate subsequent interpolation: provides a basic framework for subsequent interpolation processing, so that there can be data corresponding to each time point in the time series, which is convenient for generating a functionally complete data set. Step S3013 For each time point, the key features and sluice performance indicators are outer joined and merged, and the outer join operation is performed through the time series to merge all the key features and sluice performance indicators into a data frame to form a complete data set; interpolation processing: In the process of data merging, interpolation methods are used to process missing values ​​(for example, linear interpolation, time series interpolation, etc.) to fill in the missing data points in the time series. Form a structured data frame: The final data frame structures key information, which is convenient for subsequent machine learning model training and verification, so that input features and target variables can correspond one to one; Improve the reliability of the model: By processing missing data, the model is not affected by missing data during training, thereby improving the accuracy of the model's prediction ability; Reduce analysis bias: Ensure that all involved features and indicators can be observed at the same time point, which is convenient for time series analysis and model training, and reduces misleading results caused by missing data.

[0127] In summary, this embodiment can efficiently match key features with sluice performance indicators, providing high-quality data support for subsequent model training. Each step technically lays a solid foundation for the accuracy, completeness and subsequent analysis of the data. The significance of its implementation is to provide guarantees for the scientific nature of data analysis and the accuracy of model prediction, thereby improving the intelligence level and decision-making ability of the sluice management system.

[0128] Example 10: Fig.10 As shown, based on Embodiments 1 to 9, the intelligent analysis system for the test results of the sluice hydraulic section model provided by the embodiment of the present invention comprises:

[0129] The data acquisition module is responsible for building a physical model of the sluice section and using data acquisition equipment to record key data in the test process in real time. The key data includes information such as water level, flow rate and pressure. The collected original key data is cleaned to remove noise and outliers, and the original key data is formatted according to the time series to obtain the first data set.

[0130] The data analysis module is responsible for extracting key features from the first data set using principal component analysis and identifying parameters that have a significant impact on the performance of the sluice gate. Based on the extracted key features, a sluice gate operation model is constructed, and key features are used to parameterize related variables such as flow and pressure. Computational fluid dynamics software is used to simulate water flow behavior and evaluate the flood discharge capacity and flow pattern connection under different openings. Particle swarm optimization is used to optimize the sluice gate operation mode, calculate the flow coefficient and perform statistical analysis to obtain the experimental values ​​of the flow coefficient under different operating conditions.

[0131] The result output module is responsible for associating the extracted key features with the sluice performance indicators and training the data-driven evaluation model based on machine learning. The sluice performance indicators include the flood discharge capacity and flow connection under different openings, the sluice operation mode and the flow coefficient; comparing the calculation results of the evaluation model with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the calculation results of the evaluation model and suggestions are provided.

[0132] The working principle and beneficial effects of the above technical solution are as follows: the data acquisition module of this embodiment builds a physical model of the sluice section, and uses data acquisition equipment to record key data in the test process in real time, and the key data include information such as water level, flow rate and pressure; the collected original key data are cleaned, noise and outliers are removed, and the original key data are formatted according to time series to obtain a first data set; the data analysis module uses principal component analysis to extract key features from the first data set and identify parameters that have a significant impact on the performance of the sluice; based on the extracted key features, a sluice operation model is constructed, and the key features are used to parameterize related variables such as flow and pressure; computational fluid dynamics software is used to simulate water flow behavior and evaluate The flood discharge capacity and flow pattern connection under different openings are estimated; the particle swarm optimization is used to optimize the sluice operation mode, the flow coefficient is calculated and statistically analyzed, and the test values ​​of the flow coefficient under different operating conditions are obtained; the result output module associates the extracted key features with the sluice performance indicators, and trains the data-driven evaluation model based on machine learning. The sluice performance indicators include the flood discharge capacity and flow pattern connection under different openings, the sluice operation mode and the flow coefficient; the evaluation model calculation results are compared with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice state is evaluated in real time and suggestions are provided based on the evaluation model calculation results. The data acquisition module of the above scheme builds a physical model of the sluice to simulate the real water flow conditions to ensure the authenticity and validity of the test data; data acquisition equipment (such as flowmeters, water level sensors and pressure sensors) are used to comprehensively collect key data during the test, including water level, flow rate and pressure, etc., and the real-time performance ensures the timeliness of the data; the collected raw data is noise and outliers are eliminated and formatted into a standard time series data set, so that the subsequent analysis has a reliable basis. High-quality data source: Through accurate raw data collection and processing, the reliability and accuracy of the data are ensured, providing a solid foundation for subsequent data analysis; Real-time monitoring: Provides necessary data support for real-time evaluation of sluice performance, allowing rapid response to any situation that occurs during the test. The data analysis module extracts key features from the first data set, identifies significant parameters that affect sluice performance, simplifies data dimensions, and improves the efficiency of subsequent analysis; Based on the extracted key features, a mathematical model is established to parameterize important variables such as flow and pressure, so as to better understand their interrelationships and influences; CFD software is used to simulate water flow behavior, analyze the flood discharge capacity and flow pattern connection under different openings, and provide a deeper understanding of flow characteristics; By applying the PSO algorithm to optimize the sluice operation mode, the flow coefficient under different operating conditions is calculated and counted, and an effective operation strategy is provided. Optimization performance evaluation: By extracting key features and establishing models, the performance of the sluice under various operating conditions can be effectively evaluated to improve the accuracy of decision-making; Based on the simulation and optimization results, a scientific basis and suggestions are provided for sluice operation, thereby improving the efficient and safe operation of the sluice.The result output module associates the extracted key features with sluice performance indicators such as flood discharge capacity, flow pattern connection and flow coefficient to form a comprehensive evaluation system; by training a machine learning-based model, the sluice performance is evaluated in real time, creating a learnable system that can adapt to different operating conditions; the model calculation results are compared with the actual test data to verify the accuracy of the model, and the model parameters are dynamically adjusted according to the results to form a feedback loop for continuous improvement; combined with the evaluation calculation, the sluice status is monitored in real time, and corresponding operational suggestions are put forward to help managers respond to potential problems in a timely manner. Accurate performance evaluation: Through machine learning and data analysis models, high-precision sluice performance evaluation can be achieved, providing strong support for management and decision-making; dynamic adjustment and optimization can be achieved to improve the operational safety and efficiency of the sluice, reduce operational risks, and support the implementation of water resource management strategies.

[0133] In summary, this embodiment improves the level of intelligent management of sluice gates: provides data support and intelligent decision-making tools for the operation and management of sluice gates, and improves the efficiency and scientificity of overall management; realizes continuous improvement and feedback loops: forms a closed loop of continuous learning and improvement by dynamically adjusting the evaluation model to ensure the efficiency and safety of continuous operation of the sluice gates; provides visual results and reports: outputs the analysis results in the form of charts, reports, etc., to help managers quickly understand the performance of the sluice gates and make effective decisions; promotes technological progress and innovation: promotes technological progress in the field of hydraulic engineering and promotes wider application and research by continuously optimizing the intelligent analysis process in practice.

[0134] Embodiment 11: Based on Embodiment 10, the data acquisition module provided in the embodiment of the present invention comprises:

[0135] The equipment installation submodule is responsible for installing various sensors related to data acquisition equipment at key locations of the physical model of the sluice section, including water level sensors, flow meters, and pressure sensors;

[0136] The data cleaning submodule is responsible for receiving the key data collected by each data acquisition device, using a low-pass filter to remove high-frequency noise, calculating the deviation of each data point from the mean of the key data, and treating the key data exceeding the preset threshold as an outlier, and removing the key data corresponding to the outlier; organizing the cleaned key data into a time series format containing a timestamp and the corresponding values ​​of each key parameter;

[0137] The collection generation submodule is responsible for converting the processed key data into a data frame using the pandas library to generate the first data set.

[0138] The working principle and beneficial effects of the above technical solution are as follows: the equipment installation submodule of this embodiment installs various sensors related to the data acquisition equipment at the key positions of the physical model of the sluice section, including water level sensors, flow meters and pressure sensors, etc.; the data cleaning submodule receives the key data collected by each data acquisition device and uses a low-pass filter to remove high-frequency noise, calculates the deviation of each data point from the mean of the key data, and regards the key data exceeding the preset threshold as an outlier, and removes the key data corresponding to the outlier; organizes the cleaned key data into a time series format containing a timestamp and the corresponding values ​​of each key parameter; the set generation submodule uses the pandas library to convert the processed key data into a data frame to generate a first data set. The equipment installation submodule of the above scheme reasonably arranges water level sensors, flow meters, pressure sensors and other equipment according to the structure and function of the sluice to ensure that the required data can be collected at key positions; calibrates each sensor and sets the sampling frequency to ensure that environmental parameters can be collected accurately and in real time. Improve data accuracy: Installing sensors at key locations can maximize the capture of the performance and operating status of the sluice gate, ensuring that the acquired data is true and representative; Optimize data quality: Reasonable sensor selection and layout can effectively improve data diversity, ensuring that the data foundation for subsequent analysis and modeling is solid and rich. The data cleaning submodule uses a low-pass filter to remove high-frequency noise and reduce the impact of environmental interference on data quality; calculate the deviation of each data point from the mean of key data, determine whether the data is an outlier by the set threshold, and remove these abnormal data to ensure the neatness and reliability of the data set; organize the cleaned key data into a time series format containing timestamps and corresponding values ​​of each key parameter for subsequent processing and analysis. Ensure the validity of the data: After removing noise and outliers, the resulting data set is more accurate, which is convenient for scientific analysis and reduces the negative impact of data problems on model results; organize the data into a time series format to make it more operational and convenient for subsequent data processing and feature extraction. The collection generation submodule uses the pandas library to convert the cleaned and sorted data into a data frame (DataFrame) to form a structured data set for subsequent analysis and modeling; it generates the first data set, which includes timestamps and corresponding values ​​of key parameters, forming a preliminary basis for analysis. Provide flexible data processing capabilities: The data frame structure enables data to be flexibly screened, cleaned, transformed, and otherwise operated, facilitating subsequent analysis and modeling processes; it lays the foundation for subsequent analysis: the generated first data set is an important basis for subsequent feature engineering, model training, and performance evaluation, ensuring that the data processing process runs through the entire process and has good logic and consistency.

[0139] In summary, through the above-mentioned submodules, the entire data acquisition module of this embodiment can effectively collect data from the environment, clean up its noise and construct a structured data set. Each submodule jointly ensures the reliability and validity of the data, providing a solid foundation for subsequent data analysis, model construction and practical application; it plays an important supporting role in realizing intelligent monitoring and management of sluices.

[0140] Embodiment 12: Based on Embodiment 10, the data analysis module provided in the embodiment of the present invention includes:

[0141] The feature processing submodule is responsible for determining the input variables and output variables required in the sluice operation model, extracting key features from the first data set, and combining the key features into a multidimensional feature set through multiplication to form new key features;

[0142] Among them, the input variables include flow rate, opening, pressure and characteristic parameters, and the output variables include estimated pressure and flow rate;

[0143] The relationship determination submodule is responsible for establishing the mathematical relationship between the input variables, pressure and the new key features, and forming mathematical equations according to the type of sluice operation model;

[0144] y=β0+β1x1+β2x2+…+β k x k ;

[0145] Where y is the output (such as flow or pressure), x is the input feature, and β is the model parameter;

[0146] The model fitting submodule is responsible for fitting the sluice operation model using the training data divided from the first data set, estimating the parameters of the sluice operation model using the least squares method; after fitting, evaluating the performance of the sluice operation model through cross-validation; and calculating the flow and pressure in real time based on the new input data.

[0147] The working principle and beneficial effects of the above technical solution are as follows: the feature processing submodule of the present embodiment determines the input variables and output variables required in the sluice operation model, extracts key features from the first data set, and combines the key features into a multidimensional feature set by multiplication to form new key features; wherein the input variables include flow, opening, pressure and characteristic parameters, etc., and the output variables include estimated pressure and flow, etc.; the relationship determination submodule establishes a mathematical relationship between the input variables, pressure and the new key features, and forms a mathematical equation according to the type of the sluice operation model;

[0148] y=β0+β1x1+β2x2+…+β k x k ;

[0149] Where y is the output (such as flow or pressure), x is the input feature, and β is the model parameter;

[0150] The model fitting submodule uses the training data divided from the first data set to fit the sluice operation model, and estimates the parameters of the sluice operation model using the least squares method; after fitting, the performance of the sluice operation model is evaluated through cross-validation; and the flow and pressure are calculated in real time based on the new input data. The feature processing submodule of the above scheme specifies the input variables (such as flow, opening, pressure, etc.) and output variables (such as estimated pressure, flow, etc.) to be used; key features are extracted from the first data set, and they are combined into a multidimensional feature set through a product operation to generate new key features. This is crucial for capturing nonlinear relationships and complex interactions; the generated multidimensional feature sets can improve the performance of the model because they better describe the underlying structure and relationships of the data, especially in the case of complex systems such as sluices; by extracting and combining features, important information in the data set can be refined and redundant features can be removed, providing higher quality data support for subsequent modeling. The relationship determination submodule analyzes the relationship between input variables and output variables, and constructs corresponding mathematical equations (such as linear regression models) according to the type of sluice operation model. The mathematical equations make the model interpretable and can intuitively express the quantitative relationship and influence between input and output. The mathematical model provides a clear guideline for subsequent model development, helping researchers understand which features have the most significant impact on the output variables and providing a basis for model selection and optimization. The model fitting submodule uses the training data divided from the first data set and the least squares method to estimate the parameters in the sluice operation model, which is a key step in achieving model training and optimization. The performance of the model is evaluated through cross-validation, which helps to evaluate the generalization ability of the model on unseen data and avoid overfitting. Improve the accuracy of the model: Through fitting, the model can better adapt to known data, thereby achieving higher prediction accuracy in practical applications; real-time computing capability: After the model is fitted, the estimated parameters can be used to perform real-time calculations on new input data to quickly obtain important data such as flow and pressure. This is of great significance for decision support and system regulation.

[0151] In summary, the data analysis module of this embodiment realizes the development and optimization of the sluice operation model through the collaboration of multiple submodules such as feature processing, relationship determination and model fitting; each submodule improves the accuracy, stability and interpretability of the model, and provides a scientific basis for real-time monitoring and control of the sluice. The whole process not only provides a theoretical basis for subsequent decision support, but also promotes the effective management and utilization of water resources.

[0152] Embodiment 13: Based on Embodiment 10, the result output module provided in the embodiment of the present invention comprises:

[0153] The numerical matching submodule is responsible for matching the value of each key feature with the value of the corresponding sluice performance indicator according to the time series; combining all time series data into a data frame, with the horizontal axis being the time point and the vertical axis being the key features and performance indicators; and randomly dividing the data frame into a training set and a test set;

[0154] The variable processing submodule is responsible for using the training set to train the model, taking the key features in the training set as input independent variables and the sluice performance indicators as target output variables, and using the random forest algorithm for training;

[0155] The error evaluation submodule is responsible for evaluating the accuracy of the model through the test set after training, and using the regression error indicators mean square error and root mean square error to evaluate the model; if the model effect does not meet the standard, grid search is used for hyperparameter tuning.

[0156] The working principle and beneficial effects of the above technical solution are as follows: the numerical matching submodule of this embodiment matches the value of each key feature with the value of the corresponding sluice performance indicator according to the time series; all time series data are combined into a data frame, with the horizontal axis being the time point and the vertical axis being the key features and performance indicators; the data frame is randomly divided into a training set and a test set; the variable processing submodule uses the training set to train the model, takes the key features in the training set as input independent variables, and takes the sluice performance indicators as target output variables, and uses the random forest algorithm for training; after training, the error evaluation submodule evaluates the accuracy of the model through the test set, and uses the regression error indicators mean square error and root mean square error to evaluate the model; if the model effect does not meet the standard, grid search is used for hyperparameter tuning. The numerical matching submodule of the above scheme matches the value of each key feature with the value of the corresponding sluice performance indicator according to the time series, involving the alignment of timestamps to ensure that all data are compared at the same time point; the matched data are combined into a data frame with time as the horizontal axis and key features and performance indicators as the vertical axis to facilitate subsequent analysis and processing; the data frame is randomly divided into training sets and test sets to ensure the representativeness of the data during the model training process. Ensure data consistency: The time series matching of data ensures the one-to-one correspondence between input features and target outputs, providing a consistent data foundation for model training; constructing a data frame makes the data more structured, which is convenient for intuitive analysis and the application of subsequent algorithms, and provides support for data visualization and understanding; dividing the data into training sets and test sets can evaluate the generalization ability of the model to a certain extent, that is, the performance of the model on unseen data. The variable processing submodule uses the key features of the training set as input independent variables and the sluice performance indicators as the target output variables, and applies the random forest algorithm to train and build the model; the random forest algorithm automatically evaluates the importance of features and can identify which input features have the greatest impact on the output variables during the training process. Improve prediction ability: As an integrated learning method, random forest can process high-dimensional data and improve the prediction accuracy of the model under multiple random sampling, reducing the risk of overfitting; the model has strong ability to handle nonlinear relationships, good adaptability, and can effectively deal with noise in the data and improve the stability of the model; through the feature importance evaluation of random forest, it can guide the subsequent feature selection and simplification, and improve the efficiency and effect of model training. The error evaluation submodule verifies the trained model through the test set, evaluates the performance of the model, and quantifies the accuracy of the model using indicators such as mean square error (MSE) and root mean square error (RMSE); if the performance of the model does not meet the set standards, use grid search hyperparameter tuning to optimize the parameter settings of the model.Comprehensively understand the model performance: Through error evaluation, we can clearly understand the prediction accuracy of the model and find potential improvement points so as to take appropriate optimization measures; Model improvement path: Hyperparameter tuning can effectively improve model performance, making the final output prediction results more reliable, thereby better supporting decision-making; Scientific reliability: Through quantitative error evaluation and tuning process, we ensure that the results provided for the final application scenario are scientific and reliable, thereby enhancing decision-making confidence.

[0157] In summary, the result output module of this embodiment realizes the training, evaluation and optimization of the sluice model through the cooperation of sub-modules such as numerical matching, variable processing and error evaluation, forming a complete set of model development and application processes. Each sub-module jointly realizes the monitoring and optimization of the model effect, providing strong data support and decision-making basis for sluice management. It not only ensures the accuracy and reliability of the model, but also lays a good foundation for subsequent intelligent decision-making and automated management.

[0158] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent analysis method for the test results of a sluice hydraulic section model, characterized in that: The following steps are involved: Build a physical model of the sluice section and use data acquisition equipment to record key data in real time during the test; clean the collected original key data, remove noise and outliers, and format the original key data according to time series to obtain the first data set; Using principal component analysis, key features are extracted from the first data set to identify parameters that have a significant impact on the performance of the sluice gate. Based on the extracted key features, a sluice gate operation model is constructed, and the key features are used to parameterize flow and pressure related variables. The extracted key features are associated with the sluice performance indicators, and a data-driven evaluation model based on machine learning is trained. The sluice performance indicators include flood discharge capacity and flow connection at different openings, sluice operation mode and flow coefficient; the evaluation model calculation results are compared with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the evaluation model calculation results and suggestions are provided.

2. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 1, characterized in that: The process of obtaining the first data set includes the following steps: Install various sensors related to data acquisition equipment at key locations of the physical model of the sluice section, including water level sensors, flow meters and pressure sensors; Receive key data collected by each data collection device and use a low-pass filter to remove high-frequency noise, calculate the deviation of each data point from the mean of the key data, and regard key data exceeding the preset threshold as outliers, and remove the key data corresponding to the outliers; Arrange the cleaned key data into a time series format containing timestamps and corresponding values ​​of each key parameter; The processed key data is converted into a data frame using the pandas library to generate the first data set.

3. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 2, characterized in that: The process of installing various sensors associated with data acquisition equipment includes the following steps: Identify key locations, including the entrance, exit, and different levels within the sluice structure; analyze historical monitoring data, fluid dynamics simulation results, and previous test results for each key location to determine the relative importance of each location in the overall sluice performance, and determine the weight coefficient setting criteria based on water flow characteristics, impact on the overall sluice performance, and historical data performance; A weighted scoring method is used to assign a weight coefficient to each key position, and the weight coefficient is between 1 and 10. According to the weight coefficient of the key position, the type and quantity of the installed data sampling equipment are obtained; After the data collection equipment is deployed, it is initialized, and regular data collection and monitoring are carried out. The data is combined with the weight coefficient to analyze the performance of each key location.

4. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 1, characterized in that: Key data includes water level, flow rate and pressure information.

5. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 1, characterized in that: The process of building a sluice operation model includes the following steps: Determine the input variables and output variables required in the sluice gate operation model, extract key features from the first data set, combine the key features into a multidimensional feature set by multiplication, and form new key features; Develop mathematical relationships between input variables, pressures, and new key features; Fitting the sluice operation model using the training data divided from the first data set, and estimating the sluice operation model parameters using the least squares method; After fitting, the performance of the sluice operation model is evaluated through cross-validation; flows and pressures are calculated in real time based on new input data.

6. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 1, characterized in that: After constructing the sluice operation model, computational fluid dynamics software was used to simulate the water flow behavior and evaluate the flood discharge capacity and flow connection under different opening degrees.

7. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 6, characterized in that: Particle swarm optimization is used to optimize the sluice operation mode, calculate the flow coefficient and perform statistical analysis to obtain the experimental values ​​of the flow coefficient under different operating conditions.

8. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 1, characterized in that: The process of training a data-driven evaluation model based on machine learning consists of the following steps: According to the time series, the value of each key feature is matched with the value of the corresponding sluice performance indicator; all time series data are combined into a data frame, with the horizontal axis being the time point and the vertical axis being the key features and performance indicators; the data frame is randomly divided into a training set and a test set; The model is trained using the training set, with the key features in the training set as input independent variables and the sluice performance indicators as target output variables, and the random forest algorithm is used for training; After training, the accuracy of the model is evaluated through the test set, and the regression error indicators mean square error and root mean square error are used to evaluate the model; if the model effect does not meet the standard, grid search is used to tune the hyperparameters.

9. The intelligent analysis method for the test results of the sluice hydraulic section model according to claim 8, characterized in that: The process of matching the value of each key feature with the value of the corresponding sluice performance indicator includes the following steps: Unify the time stamps of key features and corresponding sluice performance indicators into a unified format; Generate a time series by the interval between the minimum and maximum timestamps, creating a list of all possible time points; For each time point, the key features and sluice performance indicators are outer joined to form a complete data frame; if a time point is missing in a certain data set, the missing value is filled using interpolation.

10. An intelligent analysis system for the test results of a sluice hydraulic section model, characterized in that: Include: The data acquisition module is responsible for building a physical model of the sluice section and using data acquisition equipment to record key data in the test process in real time. The key data includes water level, flow rate and pressure information; cleaning the collected original key data, removing noise and outliers, and formatting the original key data according to the time series to obtain the first data set; The data analysis module is responsible for extracting key features from the first data set using principal component analysis and identifying parameters that have a significant impact on the performance of the sluice gate. Based on the extracted key features, a sluice gate operation model is constructed, and the key features are used to parameterize flow and pressure-related variables. Computational fluid dynamics software is used to simulate water flow behavior and evaluate the flood discharge capacity and flow pattern connection under different openings. Particle swarm optimization is used to optimize the sluice gate operation mode, calculate the flow coefficient and perform statistical analysis to obtain the experimental values ​​of the flow coefficient under different operating conditions. The result output module is responsible for associating the extracted key features with the sluice performance indicators and training the data-driven evaluation model based on machine learning. The sluice performance indicators include the flood discharge capacity and flow connection under different openings, the sluice operation mode and the flow coefficient; comparing the calculation results of the evaluation model with the actual test data to verify the accuracy of the evaluation model. According to the verification results, the evaluation model parameters are adjusted and fed back to the evaluation model to achieve continuous improvement; the sluice status is evaluated in real time based on the calculation results of the evaluation model and suggestions are provided.

Citation Information

Cited By

  • Smart target range data management method and system

    CN119206590A

  • Water gate intelligent control method based on water conservancy monitoring data

    CN119902571A

  • Scouring and silting change evaluation method based on seabed scouring mechanism of Yellow River underwater delta

    CN120030287A

  • Gas density measuring method based on gas densitometer

    CN120293768A