Real-time monitoring and early warning system and method for foundation settlement of ancient building
Through multiple settlement sensor nodes, data is collected and features are extracted using deep learning technology, the problems of data analysis difficulty and long monitoring cycle of traditional monitoring methods are solved, real-time monitoring and early warning of foundation settlement of ancient buildings are realized, and the protection of ancient buildings is provided. More accurate and efficient support.
Patent Information
- Application Number
- CN202411057668.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-12-06
AI Technical Summary
Traditional ancient building foundation settlement monitoring methods have difficulty in data analysis, long monitoring cycle, high cost and limitations, and cannot achieve real-time and comprehensive monitoring and early warning.
Multiple settlement sensor nodes are used to collect data, and the multi-scale correlation characteristics of settlement data are extracted through deep learning technology to form a self-reinforced settlement data multi-scale correlation feature map for real-time monitoring and early warning.
Real-time monitoring and early warning of the settlement of ancient building foundations has been achieved, the workload of manual data processing has been reduced, data analysis efficiency has been improved, and more accurate and efficient protection and restoration technical support has been provided.
Smart Images

Figure CN119091603A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent monitoring and early warning technology, and more specifically, to a real-time monitoring and early warning system and method for ancient building foundation settlement. Background Art
[0002] Ancient buildings are important carriers of history and culture and have irreplaceable value. However, due to factors such as natural disasters, human destruction, and environmental changes, the problem of ancient building foundation settlement is becoming increasingly serious, posing a huge challenge to the protection and restoration of ancient buildings. Traditional methods for monitoring the settlement of ancient building foundations have many shortcomings, such as the difficulty in analyzing monitoring data.
[0003] Therefore, an optimized real-time monitoring and early warning scheme for ancient building foundation settlement is desired. Summary of the invention
[0004] In order to solve the above technical problems, the present invention is proposed. The embodiment of the present invention provides a real-time monitoring and early warning system and method for the settlement of ancient building foundations, which obtains the settlement data of the monitored ancient building foundations for multiple days within a predetermined time period collected by multiple settlement sensor nodes; performs data structuring processing on the settlement data of the monitored ancient building foundations for multiple days within a predetermined time period collected by the multiple settlement sensor nodes to obtain a global matrix of the settlement data of the monitored ancient building foundations; extracts settlement features from the global matrix of the settlement data of the monitored ancient building foundations to obtain a multi-scale correlation feature map of self-enhanced settlement data; and, based on the multi-scale correlation feature map of self-enhanced settlement data, determines whether to generate an early warning prompt. In this way, the multi-scale correlation features of the settlement data can be extracted through deep learning technology, thereby realizing real-time monitoring and early warning of the settlement of the ancient building foundations, thereby providing technical support for the protection and restoration of ancient buildings.
[0005] In a first aspect, a real-time monitoring and early warning method for ancient building foundation settlement is provided, which comprises:
[0006] Acquire settlement data of the monitored ancient building foundation collected by multiple settlement sensor nodes for multiple days within a predetermined time period;
[0007] Performing data structured processing on the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the plurality of settlement sensor nodes to obtain a global matrix of the settlement data of the monitored ancient building foundation;
[0008] Extracting settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of self-reinforced settlement data; and
[0009] Based on the multi-scale correlation characteristic graph of the self-enhanced settlement data, it is determined whether to generate an early warning prompt.
[0010] In the second aspect, a real-time monitoring and early warning system for ancient building foundation settlement is provided, which includes:
[0011] A settlement data acquisition module is used to acquire settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes;
[0012] A data structured processing module is used to perform data structured processing on the settlement data of the monitored ancient building foundation collected by the plurality of settlement sensor nodes for multiple days within a predetermined time period to obtain a global matrix of the settlement data of the monitored ancient building foundation;
[0013] A settlement feature extraction module, used for extracting settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of self-reinforced settlement data; and
[0014] The early warning prompt determination module is used to determine whether to generate an early warning prompt based on the multi-scale correlation feature map of the self-reinforced settlement data. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0016] Figure 1 The present invention is a flowchart of a method for real-time monitoring and early warning of ancient building foundation settlement according to an embodiment of the present invention.
[0017] Figure 2 Schematic diagram of the architecture of a real-time monitoring and early warning method for ancient building foundation settlement according to an embodiment of the present invention.
[0018] Figure 3 4 is a block diagram of a real-time monitoring and early warning system for ancient building foundation settlement according to an embodiment of the present invention.
[0019] Figure 4 Schematic diagram of a scenario of a real-time monitoring and early warning method for ancient building foundation settlement according to an embodiment of the present invention. DETAILED DESCRIPTION
[0020] The technical solutions in the embodiments of the present invention will be described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0021] Unless otherwise specified, all technical and scientific terms used in the embodiments of the present invention have the same meanings as those commonly understood by those skilled in the art of the present invention. The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the scope of the present invention.
[0022] In the description of the embodiments of the present invention, it should be noted that, unless otherwise specified and limited, the term "connection" should be understood in a broad sense. For example, it can be an electrical connection or a connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.
[0023] It should be noted that the terms "first\second\third" involved in the embodiments of the present invention are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein.
[0024] Ancient buildings refer to representative buildings built in historical periods, which usually have high artistic, historical and cultural values. They are important carriers of historical culture, reflecting the characteristics of social, economic, political and religious aspects of different periods, and are also the crystallization of human wisdom and creativity. Ancient buildings can be divided into different types, such as palaces, temples, ancient pagodas, gardens, houses, etc. Each type of ancient building has its own unique characteristics and style, reflecting the architectural technology, aesthetic concepts and cultural traditions of the time.
[0025] Ancient buildings usually use traditional building materials and techniques, such as wood, masonry, blue bricks, glazed tiles, etc. They focus on structural stability and aesthetic expression, integrating the natural environment and humanistic spirit. The architectural form, layout and decoration of ancient buildings reflect the social system, religious beliefs and aesthetic concepts of the time.
[0026] The value of ancient buildings lies in their historicity, artistry and culture. First, ancient buildings carry rich historical information, record the social development and changes in the past, and are an important basis for studying history. Secondly, ancient buildings demonstrate human artistic creativity and aesthetic pursuits, and are an important part of the history of architectural art. Finally, ancient buildings embody the cultural traditions and spirits of different regions and nationalities, are an important part of cultural heritage, and are also a symbol of national identity and pride.
[0027] Protecting and inheriting ancient buildings is an important task in protecting historical and cultural heritage. However, the problem of ancient building foundation settlement is an important challenge faced by the protection and restoration of ancient buildings. Foundation settlement may be caused by a variety of factors, including natural disasters, geological conditions, human destruction, and environmental changes.
[0028] First, natural disasters are one of the common causes of foundation settlement of ancient buildings. Natural disasters such as earthquakes, floods, and storms can damage and deform the foundations of ancient buildings, leading to foundation settlement. Especially in areas where earthquakes occur frequently, the foundations of ancient buildings are easily affected by vibrations, which in turn causes settlement problems. Secondly, geological conditions can also affect the stability of the foundations of ancient buildings. If the foundation is in soft soil or in an area with large changes in groundwater levels, the risk of foundation settlement will increase. This is because these geological conditions can easily lead to soil settlement or foundation sinking, which can cause foundation settlement problems in ancient buildings. In addition, human destruction is also an important cause of foundation settlement of ancient buildings. For a long time, ancient buildings may be improperly repaired and renovated, or suffer from theft, destruction and other behaviors, which will directly affect the stability of their foundations and lead to foundation settlement problems. Finally, environmental changes may also have an impact on the foundations of ancient buildings. Factors such as climate change and land use changes may change the surrounding environmental conditions, thereby affecting the stability of the foundations of ancient buildings. For example, an increase or decrease in rainfall and land settlement may cause deformation and settlement of the foundations of ancient buildings.
[0029] In the face of the problem of ancient building foundation settlement, protection and restoration work needs to take into account a variety of factors. First, it is necessary to conduct foundation surveys and monitoring to understand the actual situation and changing trends of the foundation. Secondly, take appropriate repair and reinforcement measures, such as foundation reinforcement and soil improvement, to improve the stability of the foundation. In addition, it is also very important to strengthen the maintenance and management of ancient buildings to avoid human damage.
[0030] Traditional methods for monitoring the settlement of ancient building foundations mainly include the following:
[0031] Precision leveling is a commonly used monitoring method that uses tools such as leveling instruments and measuring rods to measure the elevation of ancient buildings and their surrounding ground. Through regular measurements and comparisons, foundation settlement can be detected. This method requires professional measurement equipment and technology, and requires a certain amount of time and human resources.
[0032] Building tilt measurement is carried out by installing inclinometers or tilt sensors and other equipment to measure the tilt of different parts of ancient buildings. By monitoring the tilt of the building, it is possible to determine whether there is a settlement problem in the foundation. This method is relatively simple, but it requires close monitoring and timely recording of tilt data.
[0033] Groundwater level monitoring: changes in groundwater levels have a direct impact on foundation stability. By installing water level meters or downhole measuring equipment, the changes in groundwater levels around ancient buildings can be monitored. If the groundwater level changes significantly, it may cause foundation settlement problems.
[0034] Soil sampling and testing, by collecting soil samples around the foundation and conducting experimental analysis of the physical and mechanical properties, can evaluate the soil's bearing capacity and stability. This method requires professional experimental equipment and technology and can provide some indicative information on foundation settlement.
[0035] Historical document and archive research: By studying historical documents and archive materials, we can understand the construction age, restoration history, foundation engineering and other information of ancient buildings, and preliminarily judge the possibility of foundation settlement. This method can provide some background information and clues, but it needs to be combined with other monitoring methods for comprehensive analysis.
[0036] These traditional ancient building foundation settlement monitoring methods can help detect and evaluate foundation settlement problems to a certain extent. However, there are some shortcomings in the traditional ancient building foundation settlement monitoring methods. The monitoring data obtained by the traditional monitoring methods are usually a large number of numbers and measurement values, which require complex data processing and analysis, which may be difficult for non-professionals and require professional knowledge and experience to interpret and analyze the monitoring data. Traditional monitoring methods usually require periodic monitoring to obtain long-term monitoring data, which means that the monitoring cycle is long, and sometimes it takes many years or even longer to obtain reliable results. This may not be timely and effective for situations where it is necessary to understand the foundation settlement in a timely manner and take corresponding measures. Traditional monitoring methods usually require professional equipment and human resources to carry out monitoring work, which will increase the cost of monitoring, especially for some ancient building protection and restoration projects, the monitoring cost may become a constraint. Traditional monitoring methods can usually only monitor specific parts or local areas of ancient buildings, and cannot fully understand the foundation settlement of the entire ancient building, which may lead to limitations in monitoring results and cannot fully evaluate the foundation stability of ancient buildings.
[0037] In order to overcome the shortcomings of traditional monitoring methods, the present invention provides an optimized real-time monitoring and early warning solution for ancient building foundation settlement.
[0038] Figure 1 The present invention is a flowchart of a method for real-time monitoring and early warning of ancient building foundation settlement according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the architecture of a real-time monitoring and early warning method for ancient building foundation settlement according to an embodiment of the present invention. Figure 1 and Figure 2As shown, the real-time monitoring and early warning method for ancient building foundation settlement includes: 110, acquiring the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes; 120, performing data structuring processing on the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes to obtain a global matrix of the monitored ancient building foundation settlement data; 130, performing settlement feature extraction on the global matrix of the monitored ancient building foundation settlement data to obtain a self-enhanced settlement data multi-scale correlation feature map; and, 140, determining whether to generate an early warning prompt based on the self-enhanced settlement data multi-scale correlation feature map.
[0039] In step 110, the settlement data of the monitored ancient building foundation collected by multiple settlement sensor nodes for multiple days within a predetermined time period is obtained. Ensure that the sensor nodes are correctly installed and accurately calibrated to obtain reliable settlement data. The monitoring time period should be selected according to the specific situation, usually long enough to obtain reliable monitoring results. Through the data collected by multiple sensor nodes, the settlement conditions of different positions of the ancient building foundation can be obtained, providing more comprehensive monitoring information.
[0040] In step 120, the settlement data of the monitored ancient building foundation collected by the multiple settlement sensor nodes for multiple days within a predetermined time period are subjected to data structuring processing to obtain a global matrix of the settlement data of the monitored ancient building foundation. The collected settlement data are sorted, sorted and archived to ensure the accuracy and integrity of the data, and data structuring processing can be performed using data processing software or programming tools. Through data structuring processing, the data collected by multiple sensor nodes can be integrated into a global matrix, which is convenient for subsequent data analysis and processing.
[0041] In step 130, the global matrix of the monitored ancient building foundation settlement data is subjected to settlement feature extraction to obtain a multi-scale correlation feature map of the self-reinforced settlement data. According to specific needs and analysis purposes, a suitable feature extraction method is selected, such as average settlement, maximum settlement, settlement rate, etc. Feature extraction can be performed using methods such as statistical analysis, signal processing, or machine learning. Through settlement feature extraction, key information about foundation settlement can be extracted from the global matrix to form a multi-scale correlation feature map to help further analyze and evaluate the foundation settlement.
[0042] In step 140, it is determined whether to generate an early warning prompt based on the multi-scale correlation feature map of the self-reinforced settlement data. According to the pre-set early warning indicators and thresholds, the multi-scale correlation feature map is analyzed and compared to determine whether the early warning conditions are met or exceeded. The setting of the early warning prompt should take into account the impact of foundation settlement on the safety of the ancient building structure. Through the analysis based on the multi-scale correlation feature map, abnormal foundation settlement can be discovered in time and early warning prompts can be generated, which is helpful to take timely protection and repair measures to avoid further damage to the ancient building structure.
[0043] In response to the above technical problems, the technical concept of the present invention is to use multiple settlement sensor nodes to collect settlement data of the monitored ancient building foundation, and extract multi-scale correlation features of the settlement data through deep learning technology, so as to realize real-time monitoring and early warning of the settlement of the ancient building foundation, thereby providing technical support for the protection and restoration of ancient buildings.
[0044] By collecting data through multiple sensor nodes and using deep learning technology for real-time processing, real-time monitoring of the settlement of the foundation of ancient buildings can be achieved, which enables the monitoring results to be fed back to relevant personnel in a timely manner so that timely measures can be taken to protect the ancient buildings. Deep learning technology can extract multi-scale correlation features from a large amount of settlement data. These features can reveal the laws of settlement changes in different locations and time periods, and can better understand the settlement of the foundation of ancient buildings. By analyzing the multi-scale correlation features of settlement data, an early warning model can be established to detect abnormal settlement in a timely manner. Once the monitoring data exceeds the preset threshold or shows abnormal changes, the system can issue an early warning prompt to remind relevant personnel to take necessary measures to prevent further damage to the foundation of the ancient building. Deep learning technology can automatically learn and extract features. Compared with traditional data analysis methods, it can process a large amount of monitoring data more efficiently, which helps to reduce the workload of manual data processing and improve the efficiency of data analysis.
[0045] By using multiple sensor nodes to collect settlement data of ancient building foundations and extracting multi-scale correlation features through deep learning technology, real-time monitoring and early warning of settlement of ancient building foundations can be achieved, providing more accurate and efficient technical support for the protection and restoration of ancient buildings.
[0046] Based on this, in the technical solution of the present invention, firstly, the settlement data of the monitored ancient building foundation collected by multiple settlement sensor nodes for multiple days in a predetermined time period are obtained; and the settlement data of the monitored ancient building foundation collected by the multiple settlement sensor nodes for multiple days in a predetermined time period are arranged into a global matrix of the settlement data of the monitored ancient building foundation according to the time dimension and the sample dimension. That is, the discretely distributed settlement data is processed into a data structure to convert it into a vector representation that is convenient for subsequent model reading and recognition.
[0047] In a specific embodiment of the present invention, the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes are subjected to data structuring processing to obtain a global matrix of the settlement data of the monitored ancient building foundation, including: arranging the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes according to the time dimension and the sample dimension respectively into the global matrix of the settlement data of the monitored ancient building foundation.
[0048] Arranging the settlement data collected by multiple sensor nodes by time dimension can integrate the data of different sensors in chronological order to form a unified data format, which makes data management and processing more convenient and efficient. By arranging the settlement data by sample dimension, the data at different time points can be integrated to form a global matrix, which can provide an overall grasp of the settlement of the ancient building foundation and better understand the changing trends and laws of the foundation. The formation of the global matrix of settlement data makes it more convenient to analyze and compare the data of different time points and samples. Through statistical analysis, data visualization and other methods, the settlement of the ancient building foundation can be more intuitively observed and evaluated, so as to timely discover abnormalities or changing trends. The global matrix of settlement data can save historical data and use it as a reference for backtracking and analysis. Through the analysis of historical data, the settlement trend of the ancient building foundation can be understood, and trend prediction based on data can be provided, which can provide a reference for future protection and restoration work.
[0049] Arranging the ancient building foundation settlement data collected by multiple sensor nodes into a global settlement data matrix according to the time dimension and sample dimension will help integrate, analyze and compare the data, provide a global perspective and historical retrospection, and provide more comprehensive and accurate technical support for the protection and restoration of ancient buildings.
[0050] Then, the global matrix of the monitored ancient building foundation settlement data is subjected to settlement feature extraction to obtain a multi-scale correlation feature map of the self-reinforced settlement data. It should be understood that the global matrix of the monitored ancient building foundation settlement data can reflect the settlement of the ancient building foundation at different times and locations, and contains various characteristics of the ancient building foundation settlement, such as settlement rate, settlement amplitude, settlement distribution, settlement trend, etc. By performing multi-scale feature mining on it, the implicit laws and abnormal patterns of the ancient building foundation settlement can be captured.
[0051] In one embodiment of the present invention, settlement features are extracted from the global matrix of the monitored ancient building foundation settlement data to obtain a self-reinforced multi-scale correlation feature map of the settlement data, including: using a deep learning network model to extract settlement features in the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data; and, passing the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a feature autocorrelation attention enhancement module to obtain the self-reinforced multi-scale correlation feature map of the settlement data.
[0052] In a specific embodiment of the present invention, a deep learning network model is used to extract settlement features in the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data, including: passing the global matrix of the monitored ancient building foundation settlement data through a foundation settlement data multi-scale correlation feature extractor comprising a first convolutional layer and a second convolutional layer to obtain the multi-scale correlation feature map of the monitored ancient building foundation settlement data.
[0053] The first convolution layer and the second convolution layer respectively use dilated convolution kernels with different dilation rates.
[0054] In a specific example of the present invention, the encoding process of extracting settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a self-reinforced settlement data multi-scale correlation feature map includes: first, passing the global matrix of the monitored ancient building foundation settlement data through a foundation settlement data multi-scale correlation feature extractor comprising a first convolutional layer and a second convolutional layer to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data, wherein the first convolutional layer and the second convolutional layer respectively use void convolution kernels with different void rates; and then passing the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a feature autocorrelation attention enhancement module to obtain a self-reinforced settlement data multi-scale correlation feature map.
[0055] That is, the multi-scale correlation feature extractor for foundation settlement data including the first convolution layer and the second convolution layer is used to extract and identify the settlement state features of the monitored ancient building by using the first convolution layer and the second convolution layer respectively using the dilated convolution kernels with different dilation rates, and then the feature autocorrelation attention enhancement module is used to aggregate the complete settlement state feature information of the monitored ancient building based on the similarity relationship between each pixel in the feature distribution.
[0056] More specifically, in an embodiment of the present invention, the encoding process of the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a feature autocorrelation attention reinforcement module to obtain a self-reinforced settlement data multi-scale correlation feature map includes: firstly passing the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a first convolution layer to obtain a dimensionality reduction feature map; then passing the dimensionality reduction feature map through a second convolution layer to obtain an efficient correlation structure map; then, using a cosine similarity operation to calculate a relationship matrix of the efficient correlation structure map; then, using a Softmax function to normalize the relationship matrix to obtain a normalized relationship matrix; next, The normalized relationship matrix is used to model the relationship between any two eigenvalues in the reduced-dimensionality feature map by using an element-by-element multiplication operation to obtain a correlation feature map; then, the correlation feature map is deconvolved to obtain a deconvolved correlation feature map; further, the deconvolved correlation feature map is element-by-element added to the reduced-dimensionality feature map to obtain a preliminary result feature map; the preliminary result feature map is channel-expanded to obtain an expanded preliminary result feature map, and then the expanded preliminary result feature map and the multi-scale correlation feature map of the monitored ancient building foundation settlement data are residually connected to obtain the multi-scale correlation feature map of the self-reinforced settlement data.
[0057] It should be understood that the deep learning network model can learn the abstract features in the data and extract the features related to the settlement from the global matrix of the monitored ancient building foundation settlement data, including the correlation features between different time points and samples. These features can better understand the settlement of the ancient building foundation. The features extracted by the deep learning network model can generate a multi-scale correlation feature map of the monitored ancient building foundation settlement data. The multi-scale correlation feature map can show the correlation between different time points and samples and reveal the spatiotemporal variation of foundation settlement. By observing and analyzing these feature maps, we can have a more comprehensive understanding of the settlement of the ancient building foundation. The deep learning network model can efficiently process a large amount of data and extract useful features from it, which can greatly reduce the workload of manual data processing and improve the efficiency and accuracy of data analysis. The correlation in the multi-scale correlation feature map can be further strengthened by using the feature autocorrelation attention enhancement module. This module can automatically learn the correlation between features and focus more on key features, thereby improving the quality and expression of the feature map.
[0058] The deep learning network model is used to extract features from the global matrix of the monitored ancient building foundation settlement data, and the multi-scale correlation feature map of the self-reinforced settlement data is obtained through the feature autocorrelation attention reinforcement module. This can improve the understanding and analysis capabilities of the ancient building foundation settlement situation and provide more accurate and efficient technical support for the protection and restoration of ancient buildings.
[0059] In a specific embodiment of the present invention, based on the multi-scale correlation feature map of the self-enhanced settlement data, determining whether to generate a warning prompt includes: passing the multi-scale correlation feature map of the self-enhanced settlement data through a classifier to obtain a classification result, and the classification result is used to indicate whether to generate a warning prompt.
[0060] By inputting the multi-scale correlation feature map of self-reinforced settlement data into the classifier, the settlement of the foundation of the ancient building can be classified and judged. The classification result can be used to indicate whether an early warning prompt is generated, that is, to judge whether there is an abnormality in the foundation or exceeds the predetermined safety range, so that potential problems can be discovered in time and appropriate measures can be taken for repair or protection, thereby avoiding possible disaster risks. Automatic classification of settlement data using a classifier can reduce the impact of manual intervention and subjective judgment, improve the objectivity and accuracy of decision-making, and speed up the response to settlement problems. By inputting the multi-scale correlation feature map of self-reinforced settlement data into the classifier, real-time monitoring and early warning can be achieved. The classifier can quickly process data, give corresponding classification results, and issue early warning prompts in time, which is crucial for the protection of the foundation of ancient buildings. It can help to discover potential problems early and take appropriate measures for repair or protection, thereby avoiding further damage.
[0061] In a specific embodiment of the present invention, the real-time monitoring and early warning method for ancient building foundation settlement also includes: a training step for training the multi-scale correlation feature extractor of foundation settlement data including the first convolutional layer and the second convolutional layer, the feature autocorrelation attention enhancement module and the classifier; wherein the training step includes: acquiring training data, the training data including: training settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes; arranging the training settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes according to the time dimension and the sample dimension into a global matrix of training monitored ancient building foundation settlement data; arranging the training global matrix of monitored ancient building foundation settlement data through the foundation settlement data including the first convolutional layer and the second convolutional layer A multi-scale correlation feature extractor is used to obtain a multi-scale correlation feature map of the foundation settlement data of the monitored ancient building for training; the multi-scale correlation feature map of the foundation settlement data of the monitored ancient building for training is passed through a feature autocorrelation attention enhancement module to obtain a multi-scale correlation feature map of the self-reinforced settlement data for training; the multi-scale correlation feature map of the self-reinforced settlement data for training is passed through a classifier to obtain a classification loss function value; based on the multi-scale correlation feature map of the self-reinforced settlement data for training, a multi-scale correlation loss function value of the self-reinforced settlement data is calculated; a weighted sum of the classification loss function value and the multi-scale correlation loss function value of the self-reinforced settlement data is used as the final loss function value and the multi-scale correlation feature extractor for foundation settlement data comprising a first convolutional layer and a second convolutional layer, the feature autocorrelation attention enhancement module and the classifier are trained through the propagation direction of gradient descent.
[0062] Here, the global matrix of the training monitored ancient building foundation settlement data is passed through a multi-scale correlation feature extractor for foundation settlement data comprising a first convolutional layer and a second convolutional layer, and each feature matrix of the multi-scale correlation feature map of the training monitored ancient building foundation settlement data is used to express the local correlation characteristics of the time series-sample cross-dimension of the training settlement data at different scales, and the feature matrices follow the channel distribution of the multi-scale correlation feature extractor for foundation settlement data. Moreover, after the multi-scale correlation feature map of the monitored ancient building foundation settlement data is trained through the feature autocorrelation attention enhancement module, the autocorrelation between feature matrices can be enhanced based on the feature representation of each feature matrix in the channel dimension. However, considering that the scale difference of the void convolution kernels with different void rates will bring significant distribution differences to the feature representation of each feature matrix, the multi-scale correlation feature map of the trained self-enhanced settlement data will still have a sparse local feature distribution complexity, that is, a sparse sub-manifold outside the complex distribution of the overall high-dimensional feature manifold, which makes the multi-scale correlation feature map of the trained self-enhanced settlement data have a relatively obvious problem of insufficient understanding of the complex feature correlation distribution expression classification regression, affecting the accuracy of the classification results obtained by the classifier.
[0063] Based on this, the present application further introduces a loss function for promoting classification understanding based on the complex feature expression of the training self-reinforced sedimentation data multi-scale correlation feature map, that is, based on the training self-reinforced sedimentation data multi-scale correlation feature map, the self-reinforced sedimentation data multi-scale correlation loss function value is calculated. Wherein, based on the training self-reinforced sedimentation data multi-scale correlation feature map, the self-reinforced sedimentation data multi-scale correlation loss function value is calculated, including: first, by calculating the training self-reinforced sedimentation data multi-scale correlation feature vector obtained after the training self-reinforced sedimentation data multi-scale correlation feature map is expanded, for example, the mean and variance between the feature values of any two positions recorded as V are obtained to obtain the mean weight matrix M μ and the variance weight matrix M σ , and then the multi-scale correlation feature vector V of the training self-enhanced settlement data is respectively combined with the mean weight matrix M μ and the variance weight matrix M σ After query multiplication, the Frobenius norm of the autocorrelation matrix of the two eigenvectors is calculated, and the product of the Frobenius norm of the autocorrelation matrix of the training self-enhanced sedimentation data multi-scale correlation eigenvector V and the weight as a hyperparameter is subtracted to obtain the multi-scale correlation loss function value of the self-enhanced sedimentation data.
[0064] For example, expressed as
[0065]
[0066] Wherein, V represents the multi-scale correlation feature vector of the training self-enhanced sedimentation data obtained after the multi-scale correlation feature map of the training self-enhanced sedimentation data is expanded, and M μ is the mean weight matrix, M σ is the variance weight matrix, T represents the transpose of the vector, represents the matrix multiplication operation, α is the weight as a hyperparameter, ‖·‖ F represents the Frobenius norm of the matrix, Loss represents the multi-scale correlation loss function value of the self-enhanced settlement data, and as described above:
[0067]
[0068] Among them, v i represents the i-th eigenvalue of the multi-scale correlation eigenvector of the training self-enhanced sedimentation data, v j represents the jth eigenvalue of the multi-scale correlation eigenvector of the training self-enhanced settlement data, M μ (i,j) represents the eigenvalue of the (i,j)th position of the mean weight matrix, M σ (i,j) represents the eigenvalue at the (i,j)th position of the variance weight matrix.
[0069] That is, the multi-scale association loss function value of the self-reinforced sedimentation data is based on determining the overall distributed structural bottleneck of the training self-reinforced sedimentation data multi-scale association feature graph through the multi-dimensional interaction of the detail groups of the training self-reinforced sedimentation data multi-scale association feature graph, and then simulates the correlation between the structural details of the training self-reinforced sedimentation data multi-scale association feature graph and the macro-structured feature representation behavior through the low-rank structured reasoning of the training self-reinforced sedimentation data multi-scale association feature graph. In this way, by training the classifier with the multi-scale association loss function value of the self-reinforced sedimentation data, the structural detail dependency relative to the macro-complexity can be modeled in the classification scenario, thereby promoting the understanding of the macro-complex feature expression based on the structural details of the training self-reinforced sedimentation data multi-scale association feature graph, and improving the accuracy of the classification results.
[0070] In summary, a real-time monitoring and early warning method for settlement of ancient building foundations based on an embodiment of the present invention is explained, and multiple settlement sensor nodes are used to collect settlement data of the monitored ancient building foundations, and multi-scale correlation features of the settlement data are extracted through deep learning technology, thereby realizing real-time monitoring and early warning of settlement of ancient building foundations, thereby providing technical support for the protection and restoration of ancient buildings.
[0071] In one embodiment of the present invention, Figure 3 This is a block diagram of a real-time monitoring and early warning system for ancient building foundation settlement according to an embodiment of the present invention. Figure 3As shown, the real-time monitoring and early warning system 200 for settlement of ancient building foundations according to an embodiment of the present invention includes: a settlement data acquisition module 210, which is used to obtain the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes; a data structured processing module 220, which is used to perform data structured processing on the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes to obtain a global matrix of settlement data of the monitored ancient building foundation; a settlement feature extraction module 230, which is used to extract settlement features from the global matrix of settlement data of the monitored ancient building foundation to obtain a multi-scale correlation feature map of self-reinforced settlement data; and an early warning prompt determination module 240, which is used to determine whether to generate an early warning prompt based on the multi-scale correlation feature map of self-reinforced settlement data.
[0072] In the real-time monitoring and early warning system for ancient building foundation settlement, the data structured processing module is used to arrange the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes according to the time dimension and the sample dimension respectively into a global matrix of the monitored ancient building foundation settlement data.
[0073] In the real-time monitoring and early warning system for ancient building foundation settlement, the settlement feature extraction module includes: a feature map extraction unit, which is used to use a deep learning network model to extract the settlement features in the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data; and an autocorrelation attention enhancement unit, which is used to pass the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a feature autocorrelation attention enhancement module to obtain the self-enhanced multi-scale correlation feature map of the self-enhanced settlement data.
[0074] Here, those skilled in the art can understand that the specific functions and operations of each unit and module in the above-mentioned ancient building foundation settlement real-time monitoring and early warning system have been referenced above. Figure 1 to Figure 2 The description of the real-time monitoring and early warning method for ancient building foundation settlement has been introduced in detail, and therefore, its repeated description will be omitted.
[0075] As described above, the ancient building foundation settlement real-time monitoring and early warning system 200 according to the embodiment of the present invention can be implemented in various terminal devices, such as a server for real-time monitoring and early warning of ancient building foundation settlement. In one example, the ancient building foundation settlement real-time monitoring and early warning system 200 according to the embodiment of the present invention can be integrated into the terminal device as a software module and / or a hardware module. For example, the ancient building foundation settlement real-time monitoring and early warning system 200 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the ancient building foundation settlement real-time monitoring and early warning system 200 can also be one of the many hardware modules of the terminal device.
[0076] Alternatively, in another example, the ancient building foundation settlement real-time monitoring and early warning system 200 and the terminal device may also be separate devices, and the ancient building foundation settlement real-time monitoring and early warning system 200 may be connected to the terminal device via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.
[0077] Figure 4 This is a schematic diagram of a scenario of a real-time monitoring and early warning method for ancient building foundation settlement according to an embodiment of the present invention. Figure 4 As shown, in this application scenario, first, the settlement data of the monitored ancient building foundation collected by multiple settlement sensor nodes for multiple days within a predetermined time period are obtained (for example, Figure 4 Then, the acquired settlement data is input into a server (for example, Figure 4 In S) shown in , the server can process the settlement data based on the real-time monitoring and early warning algorithm of the ancient building foundation settlement to determine whether to generate an early warning prompt.
[0078] It should also be noted that in the device, apparatus and method of the present invention, each component or each step can be decomposed and / or reassembled, and such decomposition and / or reassembly should be regarded as an equivalent solution of the present invention.
[0079] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
[0080] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.
[0081] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present invention to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.
Claims
1. A real-time monitoring and early warning method for ancient building foundation settlement, characterized in that: include: Acquire settlement data of the monitored ancient building foundation collected by multiple settlement sensor nodes for multiple days within a predetermined time period; Performing data structured processing on the settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the plurality of settlement sensor nodes to obtain a global matrix of the settlement data of the monitored ancient building foundation; Extracting settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the self-reinforced settlement data; as well as Based on the multi-scale correlation characteristic graph of the self-enhanced settlement data, it is determined whether to generate an early warning prompt.
2. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 1 is characterized in that: The settlement data of the monitored ancient building foundation collected by the plurality of settlement sensor nodes for multiple days within a predetermined time period are subjected to data structured processing to obtain a global matrix of the settlement data of the monitored ancient building foundation, including: The settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes are arranged according to the time dimension and the sample dimension to form a global matrix of the settlement data of the monitored ancient building foundation.
3. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 2 is characterized in that: Extracting settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of self-reinforced settlement data includes: Using a deep learning network model to extract the settlement features in the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data; and The multi-scale correlation feature map of the monitored ancient building foundation settlement data is passed through a feature autocorrelation attention enhancement module to obtain the self-enhanced multi-scale correlation feature map of the self-enhanced settlement data.
4. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 3 is characterized in that: The deep learning network model is used to extract the settlement features in the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data, including: The global matrix of the monitored ancient building foundation settlement data is passed through a foundation settlement data multi-scale correlation feature extractor comprising a first convolutional layer and a second convolutional layer to obtain a multi-scale correlation feature map of the monitored ancient building foundation settlement data.
5. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 4 is characterized in that: The first convolution layer and the second convolution layer respectively use dilated convolution kernels with different dilation rates.
6. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 5 is characterized in that: The multi-scale correlation feature map of the monitored ancient building foundation settlement data is passed through a feature autocorrelation attention enhancement module to obtain the self-enhanced multi-scale correlation feature map of the settlement data, including: Passing the multi-scale correlation feature map of the monitored ancient building foundation settlement data through a first convolutional layer to obtain a dimension reduction feature map; Passing the dimension-reduced feature map through a second convolutional layer to obtain an efficient association structure map; Calculating the relationship matrix of the efficient association construction graph using a cosine similarity operation; Using a Softmax function to normalize the relationship matrix to obtain a normalized relationship matrix; The normalized relationship matrix is used to model the relationship between any two eigenvalues in the reduced-dimensional feature map by using an element-by-element multiplication operation to obtain a correlation feature map; Performing a deconvolution operation on the associated feature map to obtain a deconvolved associated feature map; Adding the deconvolution associated feature map and the dimension reduction feature map element by element to obtain a preliminary result feature map; and After channel expansion is performed on the preliminary result feature map to obtain an expanded preliminary result feature map, the expanded preliminary result feature map and the monitored ancient building foundation settlement data multi-scale correlation feature map are residually connected to obtain the self-reinforced settlement data multi-scale correlation feature map.
7. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 6 is characterized in that: Based on the multi-scale correlation characteristic graph of the self-enhanced settlement data, determining whether to generate an early warning prompt includes: The multi-scale correlation feature map of the self-enhanced settlement data is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether an early warning prompt is generated.
8. The real-time monitoring and early warning method for ancient building foundation settlement according to claim 7 is characterized in that: Also includes: A training step for training the multi-scale correlation feature extractor for foundation settlement data comprising a first convolutional layer and a second convolutional layer, the feature autocorrelation attention enhancement module, and the classifier; Wherein, the training step includes: Acquiring training data, the training data comprising: training settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes; Arrange the training settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes into a global matrix of the training monitored ancient building foundation settlement data according to the time dimension and the sample dimension; The global matrix of the training monitored ancient building foundation settlement data is passed through a foundation settlement data multi-scale correlation feature extractor comprising a first convolutional layer and a second convolutional layer to obtain a training monitored ancient building foundation settlement data multi-scale correlation feature map; The multi-scale correlation feature map of the foundation settlement data of the monitored ancient building is passed through a feature autocorrelation attention reinforcement module to obtain a multi-scale correlation feature map of the training self-reinforced settlement data; Passing the multi-scale correlation feature map of the training self-enhanced sedimentation data through a classifier to obtain a classification loss function value; Based on the multi-scale correlation feature graph of the training self-reinforced sedimentation data, calculating the multi-scale correlation loss function value of the self-reinforced sedimentation data; The weighted sum of the classification loss function value and the self-reinforced settlement data multi-scale correlation loss function value is used as the final loss function value and the multi-scale correlation feature extractor of foundation settlement data including the first convolutional layer and the second convolutional layer, the feature autocorrelation attention enhancement module and the classifier are trained through the propagation direction of gradient descent.
9. A real-time monitoring and early warning system for ancient building foundation settlement, characterized in that: include: A settlement data acquisition module is used to acquire settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by multiple settlement sensor nodes; A data structured processing module is used to perform data structured processing on the settlement data of the monitored ancient building foundation collected by the plurality of settlement sensor nodes for multiple days within a predetermined time period to obtain a global matrix of the settlement data of the monitored ancient building foundation; A settlement feature extraction module is used to extract settlement features from the global matrix of the monitored ancient building foundation settlement data to obtain a multi-scale correlation feature map of self-reinforced settlement data; as well as The early warning prompt determination module is used to determine whether to generate an early warning prompt based on the multi-scale correlation feature map of the self-reinforced settlement data.
10. The ancient building foundation settlement real-time monitoring and early warning system according to claim 9 is characterized in that: The data structured processing module is used to: The settlement data of the monitored ancient building foundation for multiple days within a predetermined time period collected by the multiple settlement sensor nodes are arranged according to the time dimension and the sample dimension to form a global matrix of the settlement data of the monitored ancient building foundation.
Citation Information
Cited By
Building overall settlement calculation method and system based on multi-source data fusion
CN121351245A
