Hydraulic gate control and monitoring system based on machine vision

Abstract

The present invention relates to the field of machine vision technology, and specifically to a water conservancy gate control and monitoring system based on machine vision, the system comprising a visual data acquisition module, a visual feature extraction module, a time series analysis module, a prediction and decision module, a fault diagnosis module and an execution control module. In the present invention, by directly extracting the visual features of water level and flow rate from an image using a convolutional neural network, the system not only improves the speed of data analysis, but also enhances the accuracy of the data and the predictability of the operation. The digitalized hydrological characteristic data reveals the dynamic characteristics of hydrological changes through time series analysis, enhances the automation of the system and optimizes the decision-making process. The application of the Bayesian network model makes the prediction of the hydrological condition more accurate, provides data support for gate operation, thereby ensuring the efficiency and safety of water resources management, and the automatic identification and risk assessment of structural defects reduce potential safety risks.

Description

Technical Field

[0001] The invention relates to the technical field of machine vision, and in particular to a water conservancy gate control and monitoring system based on machine vision. Background Art

[0002] Machine vision is a technology that uses digital processing and intelligent algorithms to analyze images to imitate human visual judgment. It integrates technologies and methods from multiple disciplines such as optics, mechanics, electronics, and computer science. It mainly includes four basic steps: image capture, image processing, image analysis, and image understanding. In industrial automation, machine vision is widely used in quality control, automatic inspection, and robot navigation.

[0003] Among them, the machine vision gate operation management system refers to the use of machine vision technology to control and manage the opening and closing operations of water gates. Video or image data is collected by cameras installed in the gate area, and then image processing algorithms are used to analyze the current status of the gate and the surrounding environment. The system can automatically detect water level changes, identify obstructions, and adjust the opening and closing of the gate according to preset rules to ensure the efficiency and safety of water flow management, including flood control, irrigation management and regulation of urban water systems, effectively reducing the need and errors of manual operation.

[0004] Existing technologies are delayed in responding to sudden hydrological events and lack effective prediction capabilities, which affects the timeliness and accuracy of gate operations. For example, during floods, delayed gate adjustments can lead to insufficient flood control and increased risk of damage. In addition, insufficient monitoring of the health of gate structures is also a major weakness. Problems such as cracks or corrosion cannot be detected in a timely manner, delaying maintenance opportunities, leading to equipment failures or serious accidents, thereby increasing maintenance costs and safety hazards. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a water conservancy gate control and monitoring system based on machine vision.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a water conservancy gate control and monitoring system based on machine vision comprises:

[0007] The visual data acquisition module is based on a camera, which regularly collects images of the gate and its surrounding environment, performs time synchronization processing on continuous image frames, and adds a timestamp to each frame of the image according to the time information of the image frame to obtain a timestamp image set;

[0008] The visual feature extraction module optimizes image contrast and clarity based on the timestamp image set, extracts visual features of water level and flow velocity using a convolutional neural network, and generates a hydrological feature data set;

[0009] The time series analysis module converts the hydrological characteristic data set into digital time series data, unifies the data format, analyzes the change trend of the hydrological series data, and generates time variation characteristics;

[0010] The prediction and decision-making module analyzes the time variation characteristics through the Bayesian network model based on the time variation characteristics, predicts the hydrological conditions in the future time period, formulates the gate operation process, and generates operation control data;

[0011] The fault diagnosis module identifies structural defects in the image data frames, including cracks and corrosion, compares the defect characteristics with historical maintenance data, evaluates the potential risks and maintenance needs of the gate, and generates structural health assessment results;

[0012] The execution control module calculates and sets the optimal gate operation parameters according to the operation control data and the structural health assessment results, automatically executes the gate opening or closing action, and monitors the gate operation status in real time to obtain a control execution log.

[0013] As a further solution of the present invention, the timestamp image set includes time identification, spatial coordinates, and environmental characteristic indicators; the hydrological characteristic data set specifically includes water level index, flow velocity, and clarity level; the time variation characteristics include periodic change information, trend fluctuation information, and continuous dependency information; the operation control data specifically includes control details, emergency response plan, and operation optimization information; the control execution log includes execution detail records, status update records, and performance adjustment feedback records.

[0014] As a further solution of the present invention, the visual data acquisition module includes an image capture submodule, a time synchronization submodule, and a time marking submodule;

[0015] The image capture submodule is based on a camera, which regularly collects images of the gate and its surrounding environment, and simultaneously checks whether the image quality matches the current environment to obtain a complete image data set;

[0016] The time synchronization submodule performs time synchronization processing of image frames through an internal clock based on the complete image data set, verifies the consistency of the image frames in time, and obtains a set of synchronized frames;

[0017] The time stamp submodule embeds a verified timestamp into the synchronous frame set, and uses the system clock to mark the acquisition time for each frame of the image, thereby generating a timestamp image set.

[0018] As a further solution of the present invention, the visual feature extraction module includes an image enhancement submodule, a feature recognition submodule, and a data set construction submodule;

[0019] The image enhancement submodule adjusts the brightness and contrast settings of the image based on the timestamp image set to optimize the image usability and obtain an optimized post-image set;

[0020] The feature recognition submodule uses a convolutional neural network to extract key visual features, including water level and flow velocity dynamics, based on the optimized post-image set, and captures key visual information affecting hydrological analysis to form a key morphological feature set;

[0021] The data set construction submodule processes the key morphological feature set, compiles and archives it according to a predetermined data format, and integrates the information to obtain a hydrological feature data set.

[0022] As a further solution of the present invention, the time series analysis module includes a data conversion submodule, a trend analysis submodule, and a characteristic generation submodule;

[0023] The data conversion submodule performs digital processing based on the hydrological feature data set to convert the visual feature data into a digital time series to obtain a unified digital sequence;

[0024] The trend analysis submodule performs trend analysis on the unified digital sequence, calculates statistical indicators of the data, including mean value and standard deviation, identifies the change trend and periodicity of key variables, and obtains the deviation dynamic analysis results;

[0025] The characteristic generation submodule refines key data characteristics according to the offset dynamic analysis results, extracts the variation characteristics of the time series, analyzes the behavior pattern of the hydrological data, and generates time variation characteristics.

[0026] As a further solution of the present invention, the prediction and decision module includes a model application submodule, a prediction analysis submodule, and a decision support submodule;

[0027] The model application submodule uses the Bayesian network model based on the time variation characteristics, combines historical data and current analysis to make probability predictions, predicts the hydrological conditions in the future time period, and generates a hydrological condition prediction model;

[0028] The prediction and analysis submodule analyzes the hydrological conditions in the future time period based on the hydrological condition prediction model, identifies potential risks and change trends, and obtains risk trend analysis results;

[0029] The decision support submodule formulates the gate operation process according to the risk trend analysis results, and verifies whether the operation process matches the current analysis data and prediction results to form operation control data.

[0030] As a further solution of the present invention, the hydrological conditions in the future time period are analyzed to identify potential risks and changing trends, using the formula:

[0031]

[0032] Get risk trend analysis results , where Representative time Hydrological parameters The value of Representative time Hydrological parameters The value of represents the adjustment factor, Represents the total number of hydrological parameters.

[0033] As a further solution of the present invention, the fault diagnosis module includes a defect detection submodule, a data comparison submodule, and a risk assessment submodule;

[0034] The defect detection submodule identifies structural defects in the image data frame, analyzes crack and corrosion features in the image, locates structural defects, and obtains structural defect location records;

[0035] The data comparison submodule compares the current image data with the records in the historical maintenance data based on the structural defect location records, evaluates the defect change trend and severity, calculates the risk level, and generates defect change trend data;

[0036] The risk assessment submodule evaluates the health status and potential maintenance needs of the structure based on the defect change trend data, analyzes potential safety hazards, and obtains structural health assessment results.

[0037] As a further solution of the present invention, the crack and corrosion features in the image are analyzed to locate the structural defects, using the formula:

[0038]

[0039] Calculate the concentration of defect areas , where Representative The gray value of a pixel, represents the total number of pixels in the analysis area, Represents the total number of pixels within the defect area.

[0040] As a further solution of the present invention, the execution control module includes a parameter calculation submodule, an operation execution submodule, and a state monitoring submodule;

[0041] The parameter calculation submodule calculates the operation parameters based on the operation control data and the structural health assessment results, adjusts the key parameters of the gate operation, and generates the gate operation optimization parameters;

[0042] The operation execution submodule automatically controls the opening or closing action of the gate according to the gate operation optimization parameters, verifies whether the operation is correctly executed through real-time feedback information, and obtains the operation feedback record;

[0043] The status monitoring submodule monitors the gate status and environmental changes in real time according to the operation feedback records, collects key operation data, performs data analysis to verify the operation effect, and generates a control execution log.

[0044] Compared with the prior art, the advantages and positive effects of the present invention are:

[0045] In the present invention, by using convolutional neural networks to directly extract visual features of water level and flow rate from images, the system not only improves the speed of data analysis, but also enhances the accuracy of data and the predictability of operations. The digitized hydrological characteristic data reveals the dynamic characteristics of hydrological changes through time series analysis, enhances the degree of automation of the system and optimizes the decision-making process. The application of the Bayesian network model makes the prediction of hydrological conditions more accurate and provides data support for gate operations, thereby ensuring the efficiency and safety of water resources management. The automatic identification and risk assessment of structural defects reduce potential safety risks, while real-time monitoring and automatic control effectively reduce the need for manual intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a system flow chart of the present invention;

[0047] Figure 2 It is a schematic diagram of the system framework of the present invention;

[0048] Figure 3 This is a flow chart of the visual data acquisition module of the present invention;

[0049] Figure 4 It is a flow chart of the visual feature extraction module of the present invention;

[0050] Figure 5 It is a flow chart of the time series analysis module of the present invention;

[0051] Figure 6 is a flow chart of the prediction and decision-making module of the present invention;

[0052] Figure 7 It is a flow chart of the fault diagnosis module of the present invention;

[0053] Figure 8 It is a flow chart of the execution control module of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0056] Embodiment 1

[0057] See also Figure 1 The present invention provides a technical solution: a water conservancy gate control and monitoring system based on machine vision includes:

[0058] The visual data acquisition module is based on a camera, which regularly collects images of the gate and its surrounding environment, performs time synchronization processing on continuous image frames, and adds a timestamp to each frame of the image according to the time information of the image frame to obtain a timestamp image set;

[0059] The visual feature extraction module optimizes the image contrast and clarity based on the timestamped image set, uses a convolutional neural network to extract the visual features of water level and flow velocity, and generates a hydrological feature dataset;

[0060] The time series analysis module converts the hydrological characteristic data set into digital time series data, unifies the data format, analyzes the changing trend of the hydrological series data, and generates time variation characteristics;

[0061] The prediction and decision-making module is based on the time variation characteristics. It analyzes the time variation characteristics through the Bayesian network model, predicts the hydrological conditions in the future time period, formulates the gate operation process, and generates operation control data;

[0062] The fault diagnosis module identifies structural defects in the image data frames, including cracks and corrosion, compares the defect characteristics with historical maintenance data, evaluates the potential risks and maintenance needs of the gate, and generates structural health assessment results;

[0063] The execution control module calculates and sets the optimal gate operation parameters based on the operation control data and structural health assessment results, automatically executes the gate opening or closing action, and monitors the gate's operating status in real time to ensure that the operation execution matches the predicted hydrological conditions and actual conditions, and obtains the control execution log.

[0064] The timestamp image set includes time identification, spatial coordinates, and environmental characteristic indicators. The hydrological characteristic data set specifically includes water level index, flow velocity, and clarity level. The time variation characteristics include periodic change information, trend fluctuation information, and continuous dependency information. The operation and control data specifically includes control details, emergency response plans, and operation optimization guidelines. The control execution log includes execution detail records, status update records, and performance adjustment feedback records.

[0065] See also Figure 2 and Figure 3 ,The visual data acquisition module includes an image capturing submodule, a time synchronization submodule, and a time marking submodule;

[0066] The image capture submodule is based on a camera, which regularly collects images of the gate and its surrounding environment, and simultaneously checks whether the image quality matches the current environment to obtain a complete image data set;

[0067] The camera regularly collects images of the gate and its surrounding environment. By setting the collection interval, the system automatically activates the camera to capture images. The collected images are then transmitted for preliminary quality analysis, including detection of image clarity, lighting conditions and interference. Images that do not meet the quality standards will be re-collected. Once the image meets the preset quality standards, the system will store the image data in the image library for further processing and analysis to form a complete image data set.

[0068] The time synchronization submodule performs time synchronization processing of image frames based on the complete image data set through the internal clock, verifies the temporal consistency of image frames, and obtains a set of synchronized frames;

[0069] The complete image data set is time-synchronized, and the internal clock is used to correct and synchronize the image frames. First, the metadata of each frame of the image is extracted from the data set, including the shooting timestamp. The timestamp is compared with the internal clock to identify the time deviation. By calculating the time deviation, the timestamp of the image frame is adjusted to make all image frames consistent in time. The processing process involves comparing timestamps, calculating deviations, and re-marking. The image frames are reorganized into a temporally coherent set to provide a time basis for subsequent analysis and application.

[0070] The time stamp submodule embeds the verified timestamp in the synchronous frame set, and uses the system clock to mark the acquisition time for each frame of the image to generate a timestamp image set;

[0071] Timestamps are embedded in synchronized sets of image frames. The specific process includes obtaining the current time of the system clock and embedding this timestamp into the metadata of each frame of the image. The timestamp embedding process uses a standard time formatting method to ensure the consistency and accuracy of each timestamp. After each frame of the image is assigned an exact acquisition time mark, the generated timestamp image set provides time information for subsequent processing, analysis, and archiving.

[0072] See also Figure 2 and Figure 4 ,The visual feature extraction module includes an image enhancement submodule, a feature recognition submodule, and a data set construction submodule;

[0073] The image enhancement submodule adjusts the brightness and contrast settings of the image based on the timestamp image set, optimizes the image usability, and obtains an optimized post-image set;

[0074] Receive a set of images marked with a timestamp, and adjust the brightness and contrast for each image. Detect the initial brightness and contrast levels of each image. For each image, dynamically adjust the brightness and contrast parameters according to its content characteristics and environmental background. Optimize the parameter settings based on the histogram data of the image, and display the pixel distribution of each brightness level in the image. By analyzing the data, determine the optimal values ​​of brightness and contrast to ensure the visual balance and clarity of the image. After the adjustment is completed, re-encode the image and save the optimized image in the post-image set.

[0075] The feature recognition submodule uses a convolutional neural network to extract key visual features, including water level and flow velocity dynamics, based on the optimized post-image set, and captures key visual information that affects hydrological analysis to form a key morphological feature set;

[0076] After processing the set of images optimized for brightness and contrast, a convolutional neural network (CNN) is used to analyze the images to extract key visual features such as water levels and flow velocity. It starts with loading the trained CNN model, which is optimized for specific hydrological-related visual features. The image is passed through the network layers, and each layer extracts different levels of features, such as edges, shapes, and motion information. The deep layers of the network recognize more complex patterns, such as dynamic changes in flow velocity. After layer-by-layer processing, the key visual information of the image is output. The information is encoded and stored as a set of key morphological features, providing the data required for accurate hydrological dynamic analysis.

[0077] The dataset construction submodule processes the key morphological feature set, compiles and archives it according to the predetermined data format, and integrates the information to obtain the hydrological feature dataset;

[0078] The key morphological feature set extracted from the convolutional neural network is processed. First, the feature data format is standardized to ensure data consistency and compatibility. The processing flow includes normalization and encoding of feature data to ensure balanced expression of each feature in the set. Subsequently, the standardized data is sorted and archived to form an ordered data structure. During the archiving process, the system classifies and labels the features according to predetermined classification rules so that each hydrological feature is properly recorded and indexed. After completing the steps, a hydrological feature dataset is formed.

[0079] See also Figure 2 and Figure 5 ,The time series analysis module includes a data conversion submodule, a trend analysis submodule, and a feature generation submodule;

[0080] The data conversion submodule performs digital processing based on the hydrological feature data set, converting the visual feature data into a digital time series to obtain a unified digital sequence;

[0081] Call the digitization processing function to convert the visual feature data into a digital format. The digitization process includes setting an independent data encoding format for each visual feature such as water level and flow rate. The encoding process takes into account the characteristics of the visual data, such as the resolution and color depth of the image. After digitization, each visual feature generates corresponding time series data, which is converted into a standardized digital sequence through an algorithm. This sequence unifies the data representation to facilitate subsequent analysis and processing. Each conversion step ensures that the data retains its spatiotemporal characteristics and maintains the continuity and accuracy of the data. The generated unified digital sequence contains all necessary time markers and feature values.

[0082] The trend analysis submodule performs trend analysis on the unified digital sequence, calculates the statistical indicators of the data, including the mean and standard deviation, identifies the change trend and periodicity of key variables, and obtains the deviation dynamic analysis results;

[0083] Receive a unified digital sequence and perform statistical analysis on it. First, calculate the basic statistical indicators of the sequence, such as the mean and standard deviation. The calculation relies on the functions in the statistical software package. Then, use time series analysis methods to identify trends and periodicity in the data, including the application of moving average and exponential smoothing techniques. The analysis work emphasizes identifying the changing trends of key variables such as water level and flow rate. The trend analysis results help predict future hydrological behavior. The generated offset dynamic analysis results will show the changes of key variables over time and provide a basis for further decision-making and research.

[0084] The feature generation submodule refines key data features based on the results of the dynamic analysis of the offset, extracts the variation characteristics of the time series, and analyzes the behavior patterns of the hydrological data to generate time variation characteristics;

[0085] Based on the results of the dynamic analysis of the offset, the extraction of data features is deepened, which involves further mining of time series data. The variation characteristics of the data are identified and extracted through statistical and machine learning algorithms. The algorithms include random forests and support vector machines, which are used to analyze the behavioral patterns and anomalies of hydrological data. The feature extraction process not only focuses on the statistical properties of the data, but also includes the behavioral trends of the data. The generated set of time variation characteristics records the changing trends of hydrological characteristics such as flow rate and water level.

[0086] See also Figure 2 and Figure 6 ,The prediction and decision making module includes the model application submodule, the prediction analysis submodule, and the decision support submodule;

[0087] The model application submodule uses the Bayesian network model based on the time variation characteristics, combines historical data and current analysis to make probabilistic predictions, predict the hydrological conditions in the future time period, and generate a hydrological condition prediction model;

[0088] The Bayesian network model is used to process data based on time-varying characteristics. First, a pre-trained Bayesian network is loaded, which has been trained and optimized for historical hydrological data. Then, the currently collected hydrological characteristic data is input into the model, including the latest observations of time-varying characteristics such as flow rate and water level. The Bayesian network combines real-time data with historical data sets and uses a probabilistic algorithm to calculate future hydrological conditions. The calculation process involves updating the network's conditional probability table and node probabilities. The prediction results output by the network will directly form a hydrological condition prediction model, which predicts the hydrological conditions in future time periods, including potential risks and hydrological change trends.

[0089] The prediction and analysis submodule analyzes the hydrological conditions in the future time period based on the hydrological condition prediction model, identifies potential risks and change trends, and obtains risk trend analysis results;

[0090] Analyze the hydrological conditions in the future time period, identify potential risks and change trends, and use the formula:

[0091]

[0092] Get risk trend analysis results , where Representative time Hydrological parameters The value of Representative time Hydrological parameters The value of represents the adjustment factor, represents the total number of hydrological parameters;

[0093] If the hydrological parameters to be analyzed include water level and flow velocity, then , if the current water level Meter, water level at the previous moment Meters, current flow rate m / s, flow rate at the previous moment m / s, adjustment factor The contribution of the balance parameter is calculated according to the formula:

[0094]

[0095]

[0096]

[0097]

[0098] The results show that the current hydrological conditions have changed compared with the previous moment, and the magnitude of this change is The value is given as a quantitative indicator for risk trend analysis. The value indicates that the risk has increased, but the magnitude is not large.

[0099] The decision support submodule formulates the gate operation process according to the risk trend analysis results, and verifies whether the operation process matches the current analysis data and prediction results to form operation control data;

[0100] Based on the results of the risk trend analysis, an appropriate gate operation process is formulated, including the design of each step of the operation process, such as the time, degree of opening and duration of the gate. Then, it is verified whether the operation process is suitable for the current hydrological conditions and predicted results. The simulation process takes into account different hydrological scenarios and response strategies to ensure that the operation process can effectively respond to various predicted hydrological conditions. After the verification is completed, the generated operation and control data includes all the necessary operation parameters and steps, and the data will be used for actual gate control.

[0101] See also Figure 2 and Figure 7 ,The fault diagnosis module includes a defect detection submodule, a data comparison submodule, and a risk assessment submodule;

[0102] The defect detection submodule identifies structural defects in the image data frame, analyzes crack and corrosion features in the image, locates structural defects, and obtains structural defect location records;

[0103] Analyze the crack and corrosion features in the image and locate the structural defects using the formula:

[0104]

[0105] Calculate the concentration of defect areas , where Representative The gray value of a pixel, represents the total number of pixels in the analysis area, Represents the total number of pixels within the defect area.

[0106] After processing the image, select the image area containing obvious cracks and corrosion for analysis, for example, if the analysis area Contains 1000 pixels, of which 200 pixels show obvious signs of cracks or corrosion. The average grayscale value of the defective pixels is assumed to be 150, and is calculated according to the formula:

[0107]

[0108] The result shows that the average concentration of defective pixels is high in the selected analysis area, and a large value indicates that there are significant structural defects in the area, and this information is recorded in the structural defect location record.

[0109] The data comparison submodule compares the current image data with the records in the historical maintenance data based on the structural defect location records, evaluates the defect change trend and severity, calculates the risk level, and generates defect change trend data;

[0110] Based on the structural defect location records, a comparative analysis with historical maintenance data is initiated. The process includes loading historical images and maintenance records stored in the database, using image comparison algorithms to compare areas at the same location in current and historical images, identifying defect change trends, such as crack extension or corrosion deepening, and evaluating the severity of each defect through comparison, and calculating the risk level of the defect. The calculation results are integrated into defect change trend data, which describes the development history and current status of each defect, providing decision makers with key information to formulate maintenance or repair strategies.

[0111] The risk assessment submodule evaluates the health status and potential maintenance needs of the structure based on the defect change trend data, analyzes potential safety hazards, and obtains the structural health assessment results;

[0112] The defect change trend data is used to evaluate the overall health status of the structure. By analyzing the data, potential safety hazards in the structure are identified. The analysis includes calculating the overall stability of the structure and the potential risk level in the future. During the evaluation process, statistical analysis methods and risk prediction models are used to comprehensively consider all identified defects and output the structural health assessment results.

[0113] See also Figure 2 and Figure 8 ,The execution control module includes a parameter calculation submodule, an operation execution submodule, and a state monitoring submodule;

[0114] The parameter calculation submodule calculates the operation parameters based on the operation control data and the structural health assessment results, adjusts the key parameters of the gate operation, and generates the gate operation optimization parameters;

[0115] Based on the operation control data and structural health assessment results, including various performance indicators of the gate structure and predicted maintenance needs, the input data is analyzed and the key parameters that need to be adjusted for gate operation, such as opening and closing speed, time and force, are calculated. The parameters are adjusted according to the current state of the structure and the predicted health trend to ensure the optimality of the operation. The algorithm takes into account factors such as water flow rate and water level changes. The generated gate operation optimization parameters are a set of precisely adjusted values, which guide the actual gate operation to cope with predicted environmental and structural changes.

[0116] The operation execution submodule automatically controls the opening or closing of the gate according to the gate operation optimization parameters, verifies whether the operation is performed correctly through real-time feedback information, and obtains the operation feedback record;

[0117] The gate operation optimization parameters are used to automatically control the opening or closing of the gate. During the operation, feedback information from the sensor is received in real time. The information includes the opening and closing status of the gate, execution speed and position error. The feedback data is used to verify the correctness and accuracy of the operation. If any deviation is found, adjustments are made immediately to ensure that the operation meets the preset parameters. The operation feedback records obtained record the implementation of each operation and any necessary adjustments in detail, providing important data for subsequent operation optimization and problem diagnosis.

[0118] The status monitoring submodule monitors the gate status and environmental changes in real time based on the operation feedback records, collects key operation data, performs data analysis to verify the operation effect, and generates a control execution log;

[0119] Real-time monitoring is carried out using operation feedback records, which include the physical state of the gate and the surrounding environmental conditions, such as water pressure and flow changes. By processing the collected operation data, the effect of the operation and the response of the environment are verified, any abnormalities or deviations from expected behaviors are identified, and the control strategy is adjusted in time to ensure the continuity and safety of gate operation. The generated control execution log includes records of each operation, environmental responses and any adjustment measures taken.

[0120] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. The water conservancy gate control and monitoring system based on machine vision is characterized by: The system comprises: The visual data acquisition module is based on a camera, which regularly collects images of the gate and its surrounding environment, performs time synchronization processing on continuous image frames, and adds a timestamp to each frame of the image according to the time information of the image frame to obtain a timestamp image set; The visual feature extraction module optimizes image contrast and clarity based on the timestamp image set, extracts visual features of water level and flow velocity using a convolutional neural network, and generates a hydrological feature data set; The time series analysis module converts the hydrological characteristic data set into digital time series data, unifies the data format, analyzes the change trend of the hydrological sequence data, and generates time variation characteristics; the time series analysis module includes a data conversion submodule, a trend analysis submodule, and a characteristic generation submodule; the data conversion submodule performs digital processing based on the hydrological characteristic data set, converts the visual characteristic data into a digital time series, and obtains a unified digital sequence; the trend analysis submodule performs trend analysis on the unified digital sequence, calculates statistical indicators of the data, including the mean and standard deviation, identifies the change trend and periodicity of key variables, and obtains the deviation dynamic analysis results; the characteristic generation submodule refines the key data characteristics according to the deviation dynamic analysis results, extracts the variation characteristics of the time series, and analyzes the behavior pattern of the hydrological data to generate time variation characteristics; The prediction and decision-making module analyzes the time variation characteristics through the Bayesian network model based on the time variation characteristics, predicts the hydrological conditions in the future time period, formulates the gate operation process, and generates operation control data; The fault diagnosis module identifies structural defects in the image data frames, including cracks and corrosion, compares the defect characteristics with historical maintenance data, evaluates the potential risks and maintenance needs of the gate, and generates structural health assessment results; The execution control module calculates and sets the optimal gate operation parameters according to the operation control data and the structural health assessment results, automatically executes the gate opening or closing action, and monitors the gate operation status in real time to obtain a control execution log.

2. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The timestamp image set includes time identification, spatial coordinates, and environmental characteristic indicators; the hydrological characteristic data set specifically includes water level index, flow velocity, and clarity level; the time variation characteristics include periodic change information, trend fluctuation information, and continuous dependency information; the operation control data specifically includes control details, emergency response plans, and operation optimization information; the control execution log includes execution detail records, status update records, and performance adjustment feedback records.

3. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The visual data acquisition module includes an image capture submodule, a time synchronization submodule, and a time marking submodule; The image capture submodule is based on a camera, which regularly collects images of the gate and its surroundings, and simultaneously checks whether the image quality matches the current environment to obtain a complete image data set; The time synchronization submodule performs time synchronization processing of image frames through an internal clock based on the complete image data set, verifies the consistency of the image frames in time, and obtains a set of synchronized frames; The time stamp submodule embeds a verified timestamp into the synchronous frame set, and uses the system clock to mark the acquisition time for each frame of the image, thereby generating a timestamp image set.

4. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The visual feature extraction module includes an image enhancement submodule, a feature recognition submodule, and a data set construction submodule; The image enhancement submodule adjusts the brightness and contrast settings of the image based on the timestamp image set to optimize the image usability and obtain an optimized post-image set; The feature recognition submodule uses a convolutional neural network to extract key visual features, including water level and flow velocity dynamics, based on the optimized post-image set, and captures key visual information affecting hydrological analysis to form a key morphological feature set; The data set construction submodule processes the key morphological feature set, compiles and archives it according to a predetermined data format, and integrates the information to obtain a hydrological feature data set.

5. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The prediction and decision module includes a model application submodule, a prediction analysis submodule, and a decision support submodule; The model application submodule uses the Bayesian network model based on the time variation characteristics, combines historical data and current analysis to make probability predictions, predicts the hydrological conditions in the future time period, and generates a hydrological condition prediction model; The prediction and analysis submodule analyzes the hydrological conditions in the future time period based on the hydrological condition prediction model, identifies potential risks and change trends, and obtains risk trend analysis results; The decision support submodule formulates the gate operation process according to the risk trend analysis results, and verifies whether the operation process matches the current analysis data and prediction results to form operation control data.

6. The machine vision-based water conservancy gate control and monitoring system according to claim 5 is characterized in that: Analyze the hydrological conditions in the future time period to identify potential risks and changing trends, using the formula: Get risk trend analysis results , where Representative time Hydrological parameters The value of Representative time Hydrological parameters The value of represents the adjustment factor, Represents the total number of hydrological parameters.

7. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The fault diagnosis module includes a defect detection submodule, a data comparison submodule, and a risk assessment submodule; The defect detection submodule identifies structural defects in the image data frame, analyzes crack and corrosion features in the image, locates structural defects, and obtains structural defect location records; The data comparison submodule compares the current image data with the records in the historical maintenance data based on the structural defect location records, evaluates the defect change trend and severity, calculates the risk level, and generates defect change trend data; The risk assessment submodule evaluates the health status and potential maintenance needs of the structure based on the defect change trend data, analyzes potential safety hazards, and obtains structural health assessment results.

8. The machine vision-based water conservancy gate control and monitoring system according to claim 7 is characterized in that: The crack and corrosion features in the image are analyzed to locate the structural defects using the formula: Calculate the concentration of defect areas , where Representative The gray value of a pixel, represents the total number of pixels in the analysis area, Represents the total number of pixels within the defect area.

9. The machine vision-based water conservancy gate control and monitoring system according to claim 1 is characterized in that: The execution control module includes a parameter calculation submodule, an operation execution submodule, and a status monitoring submodule; The parameter calculation submodule calculates the operation parameters based on the operation control data and the structural health assessment results, adjusts the key parameters of the gate operation, and generates the gate operation optimization parameters; The operation execution submodule automatically controls the opening or closing action of the gate according to the gate operation optimization parameters, verifies whether the operation is correctly executed through real-time feedback information, and obtains the operation feedback record; The status monitoring submodule monitors the gate status and environmental changes in real time according to the operation feedback records, collects key operation data, performs data analysis to verify the operation effect, and generates a control execution log.

Images