Methods and systems for quality and safety risk management in water conservancy projects

Abstract

This invention discloses a method and system for quality and safety risk management in water conservancy projects, relating to the field of risk management technology. The method includes: acquiring and segmenting the dam surface construction drawings of the water conservancy project; analyzing the multi-dimensional image features of each construction block to determine the construction status; if the construction status is paving, activating the inner layer of a nested management network to perform paving risk analysis; when the paving risk management result meets predetermined constraints, collecting dam surface construction information and activating the outer layer network to analyze compaction risk; and performing quality and safety risk management of the water conservancy project based on the compaction risk management result. This invention solves the technical problem that existing water conservancy project construction quality and safety supervision mainly relies on manual inspection and experience-based judgment, making it difficult to achieve real-time and accurate monitoring of the construction process. It achieves the technical effect of improving the real-time performance and accuracy of construction quality and safety supervision by utilizing an intelligent management network for real-time monitoring and risk assessment of the paving and compaction stages.

Description

Technical Field

[0001] This invention relates to the field of risk management technology, specifically to a method and system for quality and safety risk management in water conservancy projects. Background Technology

[0002] In water conservancy engineering construction, quality control of dam face construction is crucial to ensuring the long-term stability and safety of the project. Traditional quality control methods typically rely on manual inspections and experience-based judgment, which suffer from low efficiency, inability to monitor in real time, and subjective judgment results, making it difficult to effectively cope with complex construction environments and unexpected problems. Furthermore, most existing technologies rely on single sensors or visual inspection systems, failing to integrate data from different stages for intelligent analysis in real time. Especially in the critical stages of paving and compaction, existing systems lack specificity and flexibility, failing to provide accurate and effective construction risk management. Summary of the Invention

[0003] This application provides a method and system for quality and safety risk management in water conservancy projects, which addresses the technical problem that the current supervision of construction quality and safety in water conservancy projects mainly relies on manual inspection and experience-based judgment, making it difficult to achieve real-time and accurate monitoring of the construction process.

[0004] The first aspect of this application provides a method for quality and safety risk management in water conservancy projects. The method includes: acquiring a dam surface construction drawing of the water conservancy project and segmenting the dam surface construction drawing to obtain a segmentation result; analyzing a first multi-dimensional image feature of a first construction block in the segmentation result to obtain a first construction state; if the first construction state is a paving state, activating the inner layer network in a nested management network to analyze the first multi-dimensional image feature to obtain a paving risk management result; when the paving risk management result meets predetermined management constraints, collecting dam surface construction information and activating the outer layer network in the nested management network to analyze the dam surface construction information to obtain a compaction risk management result; and performing quality and safety risk management of the water conservancy project based on the compaction risk management result.

[0005] A second aspect of this application provides a quality and safety risk management system for water conservancy projects. The system includes: a dam surface construction drawing segmentation module, used to acquire dam surface construction drawings of the water conservancy project and segment the dam surface construction drawings to obtain segmentation results; a construction status analysis module, used to analyze the first multi-dimensional image features of the first construction block in the segmentation results to obtain a first construction status; a paving risk management module, used to activate the inner layer network in a nested management network to analyze the first multi-dimensional image features if the first construction status is a paving status, to obtain a paving risk management result; a compaction risk management module, used to collect dam surface construction information and activate the outer layer network in the nested management network to analyze the dam surface construction information when the paving risk management result meets predetermined management constraints, to obtain a compaction risk management result; and a quality and safety risk management module, used to perform quality and safety risk management of the water conservancy project based on the compaction risk management result.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The method and system for quality and safety risk management in water conservancy projects provided in this application relate to the field of risk management technology. By acquiring and segmenting the dam surface construction drawings of the water conservancy project, analyzing the multi-dimensional image features of each construction block, and determining the construction status, if it is in the paving state, the inner layer of the nested management network is activated to analyze paving risks. After the paving risks meet predetermined constraints, the outer layer network is activated to analyze compaction risks and perform quality and safety risk management. This solves the technical problem that existing water conservancy project construction quality and safety supervision mainly relies on manual inspection and experience judgment, making it difficult to achieve real-time and accurate monitoring of the construction process. It achieves real-time monitoring and risk assessment of the paving and compaction stages using an intelligent management network, improving the real-time nature and accuracy of construction quality and safety supervision, thereby enhancing overall construction efficiency and safety. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic diagram of a quality and safety risk management method for water conservancy projects provided in an embodiment of this application;

[0010] Figure 2This is a schematic diagram of the structure of a quality and safety risk management system for water conservancy projects provided in an embodiment of this application.

[0011] Explanation of reference numerals in the attached drawings: 11 Dam surface construction drawing segmentation module, 12 Construction status analysis module, 13 Paving risk management module, 14 Compaction risk management module, 15 Quality and safety risk management module. Detailed Implementation

[0012] This application provides a method and system for quality and safety risk management in water conservancy projects, which addresses the technical problem that the current supervision of construction quality and safety in water conservancy projects mainly relies on manual inspection and experience-based judgment, making it difficult to achieve real-time and accurate monitoring of the construction process.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0015] Example 1, as Figure 1 As shown, this application provides a method for quality and safety risk management in water conservancy projects, the method comprising:

[0016] P10: Obtain the dam surface construction drawing of the water conservancy project, and segment the dam surface construction drawing to obtain the segmentation result.

[0017] Specifically, the first step is to obtain the dam surface construction drawings for the water conservancy project. These drawings are a crucial outcome of the water conservancy project design phase, detailing the dam's structural layout, construction details, and relevant technical parameters. These drawings are typically stored digitally for easier subsequent processing and analysis. Obtaining these drawings can be achieved in various ways, such as obtaining electronic versions from the water conservancy project's design institute or converting paper drawings to digital format through on-site scanning.

[0018] After obtaining the construction drawings, a crucial step is to segment the dam surface construction drawings. This involves using image processing techniques to divide the complete dam surface construction drawings into several regions or construction blocks with specific characteristics. For example, this can be achieved based on image segmentation algorithms, which can accurately identify different construction areas based on differences in color, texture, shape, and other features within the construction drawings. For instance, edge detection algorithms can identify the boundaries of different areas in the construction drawings, thus achieving effective segmentation. In practice, mature image processing software or programming tools such as MATLAB and Python can be used to implement this function. By analyzing the shape, identifiers, and coordinate information in the construction drawings, different areas can be automatically identified and rationally divided. This process can significantly reduce errors from manual operation while improving segmentation efficiency and accuracy, making the division of each construction unit more scientific and reasonable.

[0019] After segmentation, the result is multiple independent construction zones, each containing specific construction requirements, schedules, risk assessments, and other information. This segmented data not only provides a clearer framework for construction site management but also offers essential data support for subsequent construction monitoring and quality and safety risk management.

[0020] P20: Analyze the first multidimensional image features of the first construction block in the segmentation result to obtain the first construction state.

[0021] Furthermore, step P20 in this embodiment of the application also includes:

[0022] P21: Extract the first texture feature from the first multidimensional image features; P22: Calculate the first deviation value between the first texture feature and the predetermined texture feature, and determine the first construction state based on the first deviation value; wherein, if the first deviation value is at the predetermined deviation threshold, the first construction state is the compaction state; if the first deviation value is not at the predetermined deviation threshold, the first construction state is the paving state; wherein, the predetermined texture feature refers to the dam surface image texture feature of the water conservancy project in the standard state after compaction construction.

[0023] It should be understood that the first multi-dimensional image feature analysis is performed on the first construction block in the segmentation result to obtain the construction status of the block. A key indicator of this analysis process is a series of image features extracted from construction drawings or site images. These image features are digital representations of the surface details of the construction block, including but not limited to texture, color distribution, and surface smoothness. Through the extraction and analysis of these features, the current construction stage or status of the block can be accurately determined, providing a basis for subsequent construction management and risk control.

[0024] Specifically, the first step is to extract the first texture feature from the multidimensional image features. Texture features typically refer to the pattern, shape, or structure of the image surface texture, reflecting the state and degree of processing of the surface material during construction. In dam construction in water conservancy projects, texture features can reflect the quality of construction. For example, the dam surface in the compaction stage will exhibit different texture features than that in the paving stage, usually showing a denser and more uniform surface structure. Image processing algorithms, such as Gray-Level Co-occurrence Matrix (GLCM) and Local Binary Pattern (LBP), can quantify texture features, providing a data foundation for subsequent deviation calculations. Taking the GLCM as an example, it can calculate the image's contrast, correlation, energy, and other statistics, which can comprehensively reflect the texture characteristics of the image.

[0025] Next, the first deviation value between the first texture feature and the predetermined texture feature is calculated. The predetermined texture feature refers to the texture feature of the dam surface image in the standard state after compaction construction. By comparing it with the predetermined texture feature, the difference between the current state and the ideal state of the construction block can be quantified. The first deviation value can be calculated by calculating the Euclidean distance between the two texture features, with the specific formula as follows: ,in, This represents the i-th statistic of the first texture feature. Let represent the i-th statistic of the predetermined texture feature, and n be the total number of statistics. Based on the comparison between the first deviation value and the predetermined deviation threshold, the first construction state is determined: if the first deviation value is within the predetermined deviation threshold, the first construction state is the compaction state, indicating that the texture feature of the construction block is similar to the standard state after compaction, meaning that the construction block has completed compaction and the construction quality meets the requirements; if the first deviation value is not within the predetermined deviation threshold, the first construction state is the paving state, indicating that the texture feature of the construction block differs significantly from the standard state after compaction, meaning that the construction block has not yet completed compaction and may still be in the paving stage.

[0026] During implementation, the characteristics of the construction drawings and construction requirements should be fully considered. Appropriate texture feature extraction methods and deviation value calculation methods should be selected, and suitable deviation thresholds should be set. The predetermined deviation threshold needs to be determined based on actual engineering experience and experimental data. A reasonable deviation threshold range can be determined by conducting experimental analysis on construction blocks in known states. Simultaneously, a corresponding data storage and management mechanism should be established to store and manage the extracted texture features and calculated deviation values, so that subsequent steps can easily access and process this data. For example, a database or file system can be used to store this information, and corresponding data indexes can be established to improve data retrieval efficiency.

[0027] By following the steps above, the construction status of the dam face can be monitored in real time during construction, thereby determining whether it meets the predetermined quality standards. This method provides a precise status monitoring mechanism for the construction process of water conservancy projects, helping project managers to promptly identify problems and avoid subsequent safety and quality hazards caused by substandard construction quality.

[0028] Furthermore, step P21 in this embodiment of the application also includes:

[0029] P21-1: Perform Discrete Cosine Transform (DCT) on the first construction block to obtain the first DCT coefficients; P21-2: Read the predetermined texture factor and adjust the first AC coefficient in the first DCT coefficients using the predetermined texture factor as the weight to obtain the first texture value; P21-3: Use the first texture value to characterize the first texture feature.

[0030] Optionally, the extraction process of the first texture feature of the first construction block can be further refined. Specifically, the first construction block is first processed by Discrete Cosine Transform (DCT) to obtain the first DCT coefficients. Discrete Cosine Transform is a mathematical transformation method that converts an image from the spatial domain to the frequency domain. It concentrates the energy in the image into a few coefficients, facilitating subsequent feature extraction and analysis. Through DCT processing, the image data of the first construction block can be converted into DCT coefficients, which reflect the energy distribution of the image at different frequencies.

[0031] Next, a predetermined texture factor is read, and the first AC coefficient in the first DCT coefficients is adjusted using this predetermined texture factor as a weight to obtain the first texture value. The predetermined texture factor is a weight parameter pre-set based on the texture characteristics of the dam surface construction in the hydraulic engineering project; it can highlight important frequency components related to the construction texture. The first AC coefficient refers to the AC component in the DCT coefficients, which contains the texture information of the image. By weighting and adjusting the first AC coefficient, a first texture value that better reflects the texture characteristics of the construction block can be obtained.

[0032] Finally, the first texture value is used to characterize the first texture feature. The first texture value is a weighted DCT coefficient that can effectively characterize the texture features of the first construction block. The texture features extracted in this way not only consider the frequency distribution of the image but also incorporate predetermined texture factors, thus improving the representativeness and discriminative power of the texture features and providing a more accurate basis for subsequent construction status identification.

[0033] In practical applications, the discrete cosine transform can be calculated using the Fast Fourier Transform (FFT) algorithm to improve computational efficiency. The setting of the predetermined texture factors needs to be optimized based on the specific texture characteristics of the dam surface construction in the hydraulic engineering project. Appropriate weight parameters can be determined through the analysis of a large amount of sample data. Simultaneously, a corresponding data storage and management mechanism should be established to store and manage the extracted first texture values, so that subsequent steps can easily call and process these data. For example, a database or file system can be used to store this information, and corresponding data indexes can be established to improve data retrieval efficiency.

[0034] In summary, through the discrete cosine transform processing, weighted adjustment, and texture value representation steps described above, the first texture feature of the first construction block can be accurately extracted. This process not only provides an important basis for subsequent construction status identification but also improves the accuracy and reliability of texture feature extraction.

[0035] P30: If the first construction state is the paving state, activate the inner network in the nested management network to analyze the features of the first multidimensional image and obtain the paving risk management result.

[0036] Furthermore, step P30 in this embodiment of the application also includes:

[0037] P31: Extract the first hue feature from the first multidimensional image features; P32: Weight the first hue feature and the first texture feature to obtain the first feature value of the first construction block; P33: Extract the first saliency index from the first multidimensional image features; P34: Use the first feature value and the first saliency index as input variables of the paving risk identification model in the inner network to obtain output information; P35: Use the output information as the paving risk management result; wherein, the paving risk identification model refers to an intelligent model obtained by supervised machine learning on a training dataset based on historical dam surface construction drawings, and the training dataset includes the feature values, saliency index, and identifiers of whether paving risk exists in the historical dam surface construction drawings.

[0038] It should be understood that determining whether the first construction state is the paving state is crucial. If the first construction state is determined to be the paving state, the inner network in the nested management network is activated to further analyze the first multi-dimensional image features, thereby obtaining the paving risk management results during the paving process. The paving stage is one of the key stages in the construction of dam faces in water conservancy projects, and the paving quality directly affects the subsequent compaction effect. Therefore, the management of paving risks is of paramount importance.

[0039] Specifically, the first step is to extract the first hue feature from the first multi-dimensional image features. Hue features describe the distribution of colors in the image and are typically used to reflect the surface uniformity during dam construction. For example, during paving, uneven hue distribution may indicate uneven laying of the paving material or problems in the construction process (such as uneven material distribution). Extracting these hue features can provide a preliminary basis for judging the paving quality. Methods for extracting hue features include converting the image from the RGB color space to other color spaces (such as HSV or LAB) and then calculating the statistics of the hue components, such as the mean and variance. These statistics can quantify the hue features and provide data support for subsequent risk assessment.

[0040] Next, the first hue feature and the first texture feature are weighted to obtain a comprehensive first feature value. The weighting process can assign different weights to the hue and texture features based on their importance in paving risk management. For example, if the texture feature is more critical in identifying paving risks, it is given a higher weight. Through weighting, both hue and texture features can be considered comprehensively to obtain a more comprehensive feature value for subsequent risk assessment.

[0041] Furthermore, a first saliency index is extracted from the first multidimensional image features. The saliency index refers to the prominence of certain regions or features in an image relative to other regions; it can reflect potential anomalies or critical areas within the construction block. There are various methods for extracting the saliency index, such as methods based on visual saliency models, which can identify regions in the image that attract human attention, providing crucial information about potential problem areas within the construction block for subsequent risk identification.

[0042] Subsequently, the first eigenvalue and the first significance index are used as input variables for the paving risk identification model in the inner network to obtain output information. The paving risk identification model is an intelligent model obtained through supervised machine learning on a training dataset based on historical dam construction drawings. This training dataset contains eigenvalues, significance indices, and indicators of the presence or absence of paving risk from the historical dam construction drawings. Through machine learning algorithms (such as support vector machines and neural networks), the model can learn the relationship between eigenvalues, significance indices, and paving risk, thereby achieving accurate identification of paving risk. By inputting the first eigenvalue and the first significance index into the model, the model outputs information about paving risk based on its learned knowledge.

[0043] Finally, the output information is used as the paving risk management result. The paving risk management result includes the identification of potential problems in the paved pavement (such as cracks, looseness, ruts, etc.). If the model output shows that there are paving risks, such as cracks, looseness, or ruts, then corresponding risk control measures need to be taken; if the model output indicates that there are no paving risks, then the next step, namely the compaction stage, can be carried out.

[0044] In practical applications, the extraction methods for hue and texture features, as well as the calculation methods for salience indices, should be rationally selected based on the specific construction conditions of the dam surface in the water conservancy project. Simultaneously, the paving risk identification model needs thorough training and validation to ensure its accuracy and reliability. Furthermore, a corresponding data storage and management mechanism should be established to store and manage the extracted feature values, salience indices, and model output information, so that subsequent steps can easily access and process this data. For example, a database or file system can be used to store this information, and corresponding data indexes can be established to improve data retrieval efficiency.

[0045] In summary, through the above steps, the system can monitor the quality status in real time during the paving process. By combining image features and intelligent models, it can automatically identify and report potential risks during the paving process, providing guidance for subsequent construction operations and ensuring the stability and safety of the dam surface construction quality in water conservancy projects.

[0046] Furthermore, step P33 in this embodiment of the application also includes:

[0047] P33-1: Extract features from the first construction block using a predetermined feature dimension through a feature extraction channel to obtain the first extracted feature; wherein, the predetermined feature dimension includes brightness, red tone, green tone, blue tone, and direction. P33-2: Perform a center-periphery difference analysis on the first extracted feature to obtain a first feature comparison map; P33-3: Normalize the area of ​​the salient region in the first feature comparison map to obtain the first salience index.

[0048] Optionally, the extraction process of the first saliency index of the first construction block can be further refined. Specifically, the first construction block is first subjected to feature extraction of predetermined feature dimensions through a feature extraction channel to obtain the first extracted features. In this process, the predetermined feature dimensions include brightness, red hue, green hue, blue hue, and direction. Brightness features can reflect the lightness and darkness of the construction block, while red hue, green hue, and blue hue features help to capture the color information of the construction materials. Directional features can reveal the directionality of lines or textures in the construction block. By integrating these feature dimensions, the visual characteristics of the first construction block can be comprehensively described, laying the foundation for subsequent saliency analysis.

[0049] Subsequently, a center-periphery difference analysis is performed on the first extracted features to generate a first feature comparison map. Center-periphery difference analysis is a commonly used method in image processing, primarily used to highlight the differences between the central and surrounding areas of an image. In construction quality monitoring, the central area is often a critical area for construction, where any unevenness or defects will be apparent. Therefore, through this difference analysis, for each pixel in the first extracted features, the differences between it and surrounding pixels in terms of features such as brightness, hue, and orientation can be calculated. This allows for the clear identification of potential problems in the construction area, especially during paving, where the differences between the central and surrounding areas may reflect unevenness in construction materials or deficiencies in construction operations. The generated first feature comparison map will display the difference information of each area in the image, helping managers to more clearly understand the spatial distribution of construction quality.

[0050] Next, the areas of salient regions in the first feature comparison map are normalized to obtain the first significance index. The purpose of normalization is to adjust the values ​​of the salient region areas to a uniform range, such as [0,1]. This facilitates subsequent comparison and analysis and helps eliminate the influence of size differences between different construction blocks. Specifically, the proportion of the salient region area to the total area of ​​the construction block can be calculated, and this proportion can be used as the first significance index. The higher the significance index, the larger the proportion of salient regions in the construction block, and the more attention should be paid to potential problems or anomalies.

[0051] In practical applications, feature extraction channels can be implemented based on various image processing techniques and algorithms. For example, grayscale processing can be used for brightness feature extraction; for hue feature extraction, the image can be converted from the RGB color space to the HSV color space, and then the hue components can be extracted. Directional feature extraction can utilize edge detection algorithms, such as the Sobel or Canny operators, to detect edge directions in the image. Center-periphery difference analysis can be achieved by calculating the feature differences between a pixel and its neighboring pixels, while normalization can be completed through simple mathematical operations.

[0052] Furthermore, to ensure the accuracy and reliability of the saliency index, the feature extraction channels and difference analysis algorithms need to be thoroughly tested and optimized. Experimental analysis of images with known saliency regions can be conducted to adjust algorithm parameters and improve saliency detection performance. Simultaneously, a corresponding data storage and management mechanism should be established to store and manage the extracted saliency index, enabling convenient retrieval and processing of this data in subsequent steps. For example, a database or file system can be used to store this information, and corresponding data indexes can be established to improve data retrieval efficiency.

[0053] Through these steps, the system can not only extract and analyze multi-dimensional features in construction images, but also delve into spatial difference information in the images, further improving the detection accuracy of paving quality.

[0054] Furthermore, if the first construction state is the paving state, the inner network in the nested management network is activated to analyze the features of the first multidimensional image to obtain the paving risk management result, which then includes:

[0055] P31a: Continuously monitor the operation of the paver to obtain paving operation information; P32a: Compare the paving operation information with the predetermined paving operation information and analyze to obtain the paving health status; P33a: Verify the paving risk management results based on the paving health status.

[0056] In one possible embodiment of this application, after obtaining the paving risk management results, in order to further ensure the accuracy and reliability of the results, it is necessary to monitor and analyze the operating status of the paver.

[0057] First, continuous monitoring of the paver's operation is conducted to obtain paving operation information. This information encompasses various operating parameters of the paver during construction, such as paving speed, paving temperature, hopper material level, auger distributor speed, and screed temperature. These operating parameters can be collected in real time by various sensors installed on the paver, enabling continuous monitoring of the paver's operating status. Sensor types include speed sensors, temperature sensors, and material level sensors, which convert the collected physical quantities into electrical signals, which are then transmitted to the monitoring system for processing and analysis.

[0058] Next, the collected paving operation information is compared with the predetermined paving operation information and analyzed to obtain the paving health status. The predetermined paving operation information consists of parameter ranges or standard values ​​pre-set based on the paver's normal operating status; it reflects the paver's operation under ideal construction conditions. By comparing the actual paving operation information with the predetermined paving operation information, potential anomalies during paver operation can be identified. For example, if the paving speed is lower than the predetermined speed range, it may indicate insufficient power or poor material supply; if the paving temperature is too high or too low, it may affect the performance of the construction materials and the paving quality. By comprehensively analyzing the comparison results of various operating parameters, the health status of the paver, i.e., the paving health status, can be assessed. The assessment of paving health status can use quantitative methods, such as assigning weights to each operating parameter and calculating a health score based on its deviation from the predetermined value.

[0059] Finally, the paving health score is used to verify the paving risk management results. The verification of paving health can be based on the following rule: paving problems, such as cracks, looseness, and ruts, are more likely to occur when equipment malfunctions. If the paving health score is low, it indicates that the paver is not operating well. Therefore, even if the paving risk management results indicate no paving risk, the construction quality should be reassessed, as equipment failure may lead to undetected problems during construction. Conversely, if the paving health score is high and the paving risk management results also indicate no paving risk, then the construction quality can be considered more confident, and the next construction stage, namely compaction, can be proceeded to.

[0060] In practical applications, continuous operation monitoring of pavers can be achieved through a sensor network installed on the paver. These sensors should possess high precision and reliability to ensure the accuracy of the collected operational information. Simultaneously, a real-time monitoring system is needed to rapidly process and analyze the collected operational data and promptly provide feedback on paving health information. Furthermore, the parameter ranges and standard values ​​for predetermined paving operation information should be reasonably set according to specific construction requirements and paver models, and the monitoring system should be calibrated and maintained regularly to ensure its long-term stable operation.

[0061] This process not only improves the accuracy of paving risk management, but also provides a basis for timely detection and resolution of paver operation problems, thereby further ensuring the quality and safety of dam surface construction in water conservancy projects.

[0062] P40: When the paving risk management result meets the predetermined management constraints, the dam face construction information is collected, and the outer network in the nested management network is activated to analyze the dam face construction information to obtain the compaction risk management result. The dam face construction information includes the number of compaction passes, vibration frequency, vehicle speed, driving direction, and compaction value (CMV).

[0063] Furthermore, step P40 in this embodiment of the application also includes:

[0064] P41: The dam surface construction information is analyzed using the compaction risk identification model in the outer network to obtain the predicted relative density of the dam surface; P42: The predicted relative density of the dam surface is weighted by a compaction coefficient to obtain the dam surface compaction quality index; P43: The dam surface compaction quality index is used to characterize the compaction risk management result.

[0065] It should be understood that when the paving risk management results meet the predetermined management constraints, it indicates that there are no significant risks in the paving construction phase. At this point, the system will collect information related to the dam surface construction and activate the outer network in the nested management network to further analyze the collected dam surface construction information, thereby obtaining the compaction risk management results. The dam surface construction information includes multiple parameters, such as the number of compaction passes, vibration frequency, vehicle speed, travel direction, and compaction volume value (CMV). These parameters are key indicators for evaluating the quality of compaction construction and can reflect the execution and effectiveness of the compaction process. The number of compaction passes and vibration frequency directly reflect the working condition of the compactor during construction, while vehicle speed and travel direction affect the compaction quality. The compaction volume value (CMV) is an important indicator for assessing the degree of compaction of the soil or construction surface.

[0066] Specifically, firstly, the collected dam surface construction information is analyzed using a compaction risk identification model within the outer network to obtain a predicted relative density of the dam surface. This compaction risk identification model is an intelligent model trained on historical data, capable of predicting the relative density of the dam surface based on current construction information. Relative density is an important indicator for measuring the degree of soil compaction; it reflects the relationship between the compacted density and maximum compaction of the soil after compaction. By predicting the relative density of the dam surface, the quality of the compaction construction can be preliminarily assessed.

[0067] Next, the compaction coefficient is introduced to weight the predicted relative density of the dam surface, resulting in a dam surface compaction quality index. The compaction coefficient is a parameter set based on engineering experience and specific construction requirements; it reflects the degree of influence of compaction quality on the final construction quality. By weighting the predicted relative density of the dam surface, a dam surface compaction quality index that comprehensively considers the compaction coefficient can be obtained, which more comprehensively reflects the quality status of the roller compaction construction.

[0068] Finally, the dam surface compaction quality index is used to characterize the results of rolling risk management. This index is a quantitative indicator that can intuitively reflect the quality and risk status of rolling construction. If the dam surface compaction quality index reaches the predetermined standard, it indicates that the rolling construction quality is good and there is no significant risk; conversely, if the index is lower than the predetermined standard, corresponding risk control measures need to be taken, such as adjusting rolling parameters or increasing the number of rolling passes.

[0069] Furthermore, the construction process of the crush risk identification model includes:

[0070] P41-1a: Collect historical dam surface compaction construction records and extract the first historical record from the historical dam surface compaction construction records; P41-2a: Extract the first historical dam surface construction information and the first historical dam surface relative density from the first historical record and form a first dataset; P41-3a: Train the first dataset and verify the obtained compaction risk identification model.

[0071] Specifically, the process begins by collecting historical records of dam surface compaction construction and extracting the first historical records. These records contain detailed information about past construction processes and form a crucial data foundation for model building. Then, the first historical dam surface construction information and the first historical dam surface relative density are extracted from these records to create the first dataset. This dataset provides rich samples for model training, enabling the model to learn the intrinsic relationship between construction information and dam surface relative density. The first dataset is then used for training, and the compaction risk identification model is obtained through validation. The training process can employ machine learning algorithms such as linear regression, support vector machines, or neural networks to ensure good predictive performance and generalization ability. The validation process uses methods such as cross-validation to evaluate the accuracy and reliability of the model, resulting in an effective compaction risk identification model.

[0072] In practical applications, to ensure the accuracy and reliability of compaction risk management results, precise collection and recording of dam surface construction information are necessary. Simultaneously, the construction and training of the compaction risk identification model should be based on a large amount of high-quality historical data and employ advanced machine learning algorithms and optimization techniques. Furthermore, the compaction coefficient should be adjusted according to specific engineering requirements and construction conditions to ensure that the dam surface compaction quality index truly reflects the construction quality. Through these measures, the quality and risks of compaction construction can be effectively assessed, providing strong support for the construction of water conservancy projects.

[0073] In summary, through a series of data collection, feature extraction, weighted processing, and intelligent model analysis, the system can accurately assess the risks during the compaction process and provide corresponding quality control suggestions. Real-time verification of the compaction risk management results ensures that the quality of dam construction in water conservancy projects meets design standards during the compaction stage, thereby improving the overall stability and safety of the project.

[0074] P50: Conduct quality and safety risk management for the water conservancy project based on the results of the compaction risk management.

[0075] Specifically, based on the compaction risk management results obtained from the above steps, the overall quality and safety of the water conservancy project are comprehensively managed. This process is a key link in the entire quality and safety risk management of the water conservancy project. By integrating the risk assessment results of the paving and compaction construction stages in the early stages, the quality and safety of the entire construction process are ensured.

[0076] First, a comprehensive analysis of the compaction risk management results is conducted. These results are presented as a dam surface compaction quality index, which integrates dam surface construction information (such as the number of compaction passes, vibration frequency, vehicle speed, driving direction, and compaction volume measurement (CMV)) with the compaction coefficient. If the dam surface compaction quality index meets or exceeds the predetermined standard, it indicates good compaction quality with no significant risks, allowing subsequent construction procedures to proceed. Conversely, if the index falls below the predetermined standard, it indicates potential quality problems in the compaction process, such as insufficient or uneven compaction, requiring appropriate risk control measures.

[0077] Next, based on the results of the compaction risk management, corresponding quality and safety risk control strategies are formulated. If risks exist, the causes need to be analyzed, and targeted measures should be taken for rectification. For example, if the compaction quality index is low, it may be due to insufficient compaction passes, inappropriate vibration frequency, or excessive vehicle speed. To address these issues, measures such as increasing the number of compaction passes, adjusting the vibration frequency, or reducing the vehicle speed can be taken. Simultaneously, for the already compacted sections, supplementary compaction or re-compaction may be necessary to ensure that the compaction quality of the dam surface meets the design requirements.

[0078] After implementing risk control measures, the effectiveness of the rectification needs to be verified. This can be achieved by collecting dam surface construction information again and reassessing the dam surface compaction quality index using a compaction risk identification model. If the compaction quality index after rectification meets the predetermined standard, it can be confirmed that the risk has been effectively controlled, and subsequent construction can continue. If the rectification effect is not ideal, further analysis of the causes is needed, and more effective measures should be taken for rectification until the risk is completely controlled.

[0079] Furthermore, it is necessary to integrate the risk management results from the paving and compaction stages to form a comprehensive assessment of the overall quality and safety of the water conservancy project. This includes comprehensive consideration of potential quality problems such as cracks, loosening, and rutting during construction, as well as monitoring and evaluating the operational status of construction equipment. This integrated management approach ensures that every stage of the water conservancy project's construction meets quality requirements, thereby guaranteeing the overall quality and safety of the project.

[0080] In summary, the embodiments of this application have at least the following technical effects:

[0081] This application achieves real-time monitoring of construction states such as paving and compaction through segmentation of dam surface construction drawings and multi-dimensional image feature analysis. This enables timely identification of anomalies and risks during construction, ensuring construction progress and quality. Combined with a nested management network, intelligent risk analysis is performed during the paving and compaction stages, automatically assessing construction quality and potential risks, providing data-driven decision support, and optimizing quality control during construction. By collecting dam surface construction information in real time (such as the number of compaction passes and vibration frequency) and analyzing it in conjunction with historical construction data, the comprehensive utilization efficiency of construction data is improved, ensuring the accuracy of construction quality assessment. Through timely identification and effective management of risks, potential quality hazards and safety accidents during construction are reduced, ensuring the smooth progress of water conservancy project construction and improving overall construction efficiency and safety.

[0082] This technology achieves the goal of improving the real-time nature and accuracy of construction quality and safety supervision by utilizing intelligent management networks for real-time monitoring and risk assessment during the paving and compaction stages, thereby enhancing overall construction efficiency and safety.

[0083] Example 2, based on the same inventive concept as the quality and safety risk management method for water conservancy projects in the foregoing examples, such as... Figure 2 As shown, this application provides a quality and safety risk management system for water conservancy projects. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0084] The dam surface construction drawing segmentation module 11 is used to obtain the dam surface construction drawing of the water conservancy project and segment the dam surface construction drawing to obtain the segmentation result.

[0085] The construction status analysis module 12 is used to analyze the first multidimensional image features of the first construction block in the segmentation result to obtain the first construction status.

[0086] The paving risk management module 13 is used to activate the inner network in the nested management network to analyze the features of the first multidimensional image if the first construction state is the paving state, and obtain the paving risk management result.

[0087] The compaction risk management module 14 is used to collect dam surface construction information and activate the outer network in the nested management network to analyze the dam surface construction information when the paving risk management result meets the predetermined management constraints, so as to obtain the compaction risk management result.

[0088] The quality and safety risk management module 15 is used to manage the quality and safety risks of the water conservancy project based on the results of the compaction risk management.

[0089] Furthermore, the construction status analysis module 12 is also used to perform the following steps:

[0090] Extract the first texture feature from the first multidimensional image features; calculate the first deviation value between the first texture feature and the predetermined texture feature, and determine the first construction state based on the first deviation value; wherein, if the first deviation value is at the predetermined deviation threshold, the first construction state is the compaction state; if the first deviation value is not at the predetermined deviation threshold, the first construction state is the paving state; wherein, the predetermined texture feature refers to the dam surface image texture feature of the water conservancy project in the standard state after compaction construction.

[0091] Furthermore, the construction status analysis module 12 is also used to perform the following steps:

[0092] The first construction block is subjected to discrete cosine transform to obtain the first DCT coefficients; a predetermined texture factor is read, and the first AC coefficient in the first DCT coefficients is adjusted with the predetermined texture factor as the weight to obtain the first texture value; the first texture value is used to characterize the first texture feature.

[0093] Furthermore, the paving risk management module 13 is also used to perform the following steps:

[0094] Extract the first hue feature from the first multidimensional image features; weight the first hue feature and the first texture feature to obtain the first feature value of the first construction block; extract the first saliency index from the first multidimensional image features; use the first feature value and the first saliency index as input variables of the paving risk identification model in the inner network to obtain output information; use the output information as the paving risk management result; wherein, the paving risk identification model refers to an intelligent model obtained by supervised machine learning on a training dataset based on historical dam surface construction drawings, and the training dataset includes the feature values, saliency index, and identifiers indicating whether paving risk exists in the historical dam surface construction drawings.

[0095] Furthermore, the paving risk management module 13 is also used to perform the following steps:

[0096] The first construction block is subjected to feature extraction using a feature extraction channel, which extracts features of predetermined feature dimensions to obtain a first extracted feature. The predetermined feature dimensions include brightness, red hue, green hue, blue hue, and direction. A center-periphery difference analysis is performed on the first extracted feature to obtain a first feature comparison map. The area of ​​salient regions in the first feature comparison map is normalized to obtain a first saliency index.

[0097] Furthermore, the paving risk management module 13 is also used to perform the following steps:

[0098] The paver is continuously monitored to obtain paving operation information; the paving operation information is compared with the predetermined paving operation information, and the paving health is analyzed to obtain the paving health status; the paving health status is used to verify the paving risk management results.

[0099] Furthermore, the crushing risk management module 14 is also used to perform the following steps:

[0100] The dam surface construction information is analyzed using the compaction risk identification model in the outer network to obtain the predicted relative density of the dam surface. A compaction coefficient is introduced to weight the predicted relative density of the dam surface, resulting in a dam surface compaction quality index. This index represents the compaction risk management result. The construction process of the compaction risk identification model includes: collecting historical dam surface compaction construction records and extracting a first historical record from these records; extracting the first historical dam surface construction information and the first historical dam surface relative density from the first historical record to form a first dataset; training the first dataset and verifying the obtained compaction risk identification model. The dam surface construction information includes the number of compaction passes, vibration frequency, vehicle speed, driving direction, and compaction volume measurement (CMV).

[0101] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0102] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0103] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for quality and safety risk management in water conservancy projects, characterized in that, include: Obtain the dam surface construction drawing of the water conservancy project, and segment the dam surface construction drawing to obtain the segmentation result; The first multidimensional image features of the first construction block in the segmentation result are analyzed to obtain the first construction state; If the first construction state is the paving state, the inner network in the nested management network is activated to analyze the features of the first multidimensional image and obtain the paving risk management result; When the paving risk management result meets the predetermined management constraints, the dam surface construction information is collected, and the outer network in the nested management network is activated to analyze the dam surface construction information to obtain the compaction risk management result. The quality and safety risks of the water conservancy project shall be managed based on the results of the compaction risk management. Analyzing the first multidimensional image features of the first construction block in the segmentation result yields the first construction state, including: Extract the first texture feature from the first multidimensional image features; Calculate the first deviation value between the first texture feature and the predetermined texture feature, and determine the first construction state based on the first deviation value; This includes: If the first deviation value is within a predetermined deviation threshold, then the first construction state is a compaction state; If the first deviation value is not within the predetermined deviation threshold, then the first construction state is the paving state; The predetermined texture features refer to the texture features of the dam surface image of the water conservancy project in the standard state after compaction construction. If the first construction state is the paving state, the inner network in the nested management network is activated to analyze the features of the first multidimensional image, and the paving risk management results are obtained, including: Extract the first tone feature from the first multidimensional image features; The first feature value of the first construction block is obtained by weighting the first hue feature and the first texture feature; Extract the first saliency index from the first multidimensional image features; The first feature value and the first significance index are used as input variables for the paving risk identification model in the inner network to obtain output information; The output information is used as the result of the paving risk management. The paving risk identification model refers to an intelligent model obtained by supervised machine learning on a training dataset based on historical dam surface construction drawings. The training dataset includes the feature values, significance index, and indicators of whether paving risk exists in the historical dam surface construction drawings. Extracting the first saliency index from the first multidimensional image features includes: The first construction block is subjected to feature extraction of a predetermined feature dimension through the feature extraction channel to obtain the first extracted feature; Perform a center-periphery difference analysis on the first extracted feature to obtain a first feature comparison map; The area of ​​the salient region in the first feature comparison map is normalized to obtain the first saliency index.

2. The method for quality and safety risk management in water conservancy projects as described in claim 1, characterized in that, Extracting the first texture feature from the first multidimensional image features includes: The first construction block is subjected to discrete cosine transform to obtain the first DCT coefficients; Read the predetermined texture factor, and use the predetermined texture factor as a weight to adjust the first AC coefficient in the first DCT coefficient to obtain the first texture value; The first texture value is used to characterize the first texture feature.

3. The method for quality and safety risk management in water conservancy projects as described in claim 1, characterized in that, The predetermined feature dimensions include brightness, red hue, green hue, blue hue, and direction.

4. The method for quality and safety risk management in water conservancy projects as described in claim 1, characterized in that, If the first construction state is the paving state, the inner network in the nested management network is activated to analyze the features of the first multidimensional image to obtain the paving risk management result, and then the process further includes: Continuous monitoring of the paver's operation is conducted to obtain paving operation information; The paving operation information is compared with the planned paving operation information, and the paving health is analyzed to obtain the paving health status; The paving health status is used to verify the paving risk management results.

5. The method for quality and safety risk management in water conservancy projects as described in claim 1, characterized in that, When the paving risk management results meet the predetermined management constraints, the dam face construction information is collected, and the outer network in the nested management network is activated to analyze the dam face construction information to obtain the compaction risk management results, including: The dam surface construction information is analyzed using the compaction risk identification model in the outer network to obtain the predicted relative density of the dam surface; By introducing a compaction coefficient to weight the predicted relative density of the dam surface, the dam surface compaction quality index is obtained. The compaction quality index of the dam surface is used to characterize the results of the compaction risk management. The construction process of the crushing risk identification model includes: Collect historical records of dam surface compaction construction and extract the first historical record from these records. Extract the first historical dam surface construction information and the first historical dam surface relative density from the first historical record, and form the first dataset; The crush risk identification model is trained on the first dataset and then tested.

6. The method for quality and safety risk management in water conservancy projects as described in claim 1, characterized in that, The dam surface construction information includes the number of compaction passes, vibration frequency, vehicle speed, driving direction, and compaction value (CMV).

7. A quality and safety risk management system for water conservancy projects, characterized in that, The system is used to implement the quality and safety risk management method for water conservancy projects according to any one of claims 1-6, the system comprising: A dam surface construction drawing segmentation module is used to acquire dam surface construction drawings of water conservancy projects and segment the dam surface construction drawings to obtain segmentation results. The construction status analysis module is used to analyze the first multidimensional image features of the first construction block in the segmentation result to obtain the first construction status. The paving risk management module is used to activate the inner network in the nested management network to analyze the features of the first multidimensional image if the first construction state is the paving state, and obtain the paving risk management result. The compaction risk management module is used to collect dam surface construction information and activate the outer network in the nested management network to analyze the dam surface construction information when the paving risk management result meets the predetermined management constraints, so as to obtain the compaction risk management result. A quality and safety risk management module is used to manage the quality and safety risks of the water conservancy project based on the results of the compaction risk management.

Images