A method and system for monitoring deformation of a foundation pit

Abstract

This invention relates to the field of foundation pit monitoring technology, and discloses a method and system for foundation pit deformation monitoring. The method includes: performing consistency verification on the original physical quantity monitoring data to obtain a standardized monitoring sequence; acquiring a design parameter set and historical construction condition records, and constructing a dynamic deformation behavior model based on the coupling relationship between the design parameter set and historical construction condition records; inputting the current construction progress into the dynamic deformation behavior model to calculate the expected deformation reference sequence; comparing the standardized monitoring sequence with the expected deformation reference sequence in real time to identify local non-coordinated deformation modes; when a local non-coordinated deformation mode triggers a preset early risk criterion, generating an abnormal structural state warning and a graded response strategy; dynamically adjusting the sampling frequency and spatial density of the distributed monitoring network, and performing rolling optimization on the parameters in the dynamic deformation behavior model; this invention can improve the efficiency of foundation pit deformation monitoring.

Description

Technical Field

[0001] This invention relates to the field of foundation pit monitoring technology, and in particular to a method and system for monitoring foundation pit deformation. Background Technology

[0002] In existing foundation pit deformation monitoring technologies, the original monitoring data is easily affected by factors such as differences in the spatial layout of distributed monitoring networks and asynchronous data acquisition timelines, resulting in problems such as insufficient spatiotemporal continuity and interference from abnormal data. Furthermore, the lack of a systematic consistency verification and data repair mechanism leads to low accuracy and efficiency in generating standardized monitoring sequences, which cannot provide a reliable data foundation for subsequent deformation analysis, thereby restricting the overall accuracy of foundation pit deformation monitoring.

[0003] Traditional foundation pit deformation prediction models often rely on single design parameters or limited historical data, failing to fully explore the coupling relationship between the mechanical constraints in the design parameters and the actual deformation evolution under historical construction conditions. This results in insufficient adaptability of the model to the dynamic deformation behavior of the foundation pit, limited prediction accuracy of the expected deformation reference sequence, and a lack of real-time and targeted identification of deformation anomalies in existing technologies. Furthermore, after triggering risk warnings, it is difficult to quickly trace the root cause of the anomaly and formulate accurate graded response strategies. The sampling frequency and spatial density adjustment of the monitoring network lack dynamic flexibility, and the model parameters cannot be optimized in a timely manner. Consequently, the monitoring system struggles to comprehensively and efficiently address the complex deformation risks during foundation pit construction. Therefore, improving the accuracy of foundation pit deformation monitoring has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for monitoring foundation pit deformation to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for monitoring foundation pit deformation, comprising: S1. Perform consistency verification on the original physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain the standardized monitoring sequence of the target foundation pit; S2. Obtain the design parameter set and historical construction condition records of the target foundation pit, and construct a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. S3. Input the current construction progress of the target foundation pit into the dynamic deformation behavior model to calculate the expected deformation reference sequence of the target foundation pit; S4. Compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit; S5. When the local non-coordinated deformation mode triggers the preset early risk criterion, an abnormal structural state warning of the target foundation pit is generated. Based on the abnormal structural state warning, the real-time construction log and geological conditions of the target foundation pit are traced back to their root causes to obtain the graded response strategy of the target foundation pit. S6. Based on the hierarchical response strategy, dynamically adjust the sampling frequency and spatial density of the distributed monitoring network, and perform rolling optimization on the parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit.

[0006] In a preferred embodiment, the step of performing consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain a standardized monitoring sequence for the target foundation pit includes: Receive raw physical quantity monitoring data collected from the distributed monitoring network in the target foundation pit, and perform time-series alignment on the raw physical quantity monitoring data to obtain the time-series monitoring data stream of the target foundation pit; Based on the spatial deployment coordinates of the distributed monitoring network, a topological relationship diagram of the distributed monitoring network is constructed. Based on the topological correlation diagram, the continuity and rationality of the time-series monitoring data stream in spatial transmission are evaluated, and the spatiotemporal correlation coefficient of the target foundation pit is obtained. The maximum coefficient in the spatiotemporal correlation coefficient is used as the upper limit, and the minimum coefficient in the spatiotemporal correlation coefficient is used as the lower limit to construct the data consistency judgment threshold for the target foundation pit; The degree of deviation in the time-series monitoring data stream is obtained by comparing each monitoring data item in the time-series monitoring data stream. When the deviation exceeds the data consistency judgment threshold, the monitoring data will be marked as a spatially isolated anomaly of the target foundation pit; Spatially isolated outliers were removed, and the missing data was repaired by spatiotemporal interpolation to obtain a standardized monitoring sequence for the target foundation pit.

[0007] In a preferred embodiment, the step of acquiring the design parameter set and historical construction condition records of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records, includes: Obtain the design drawings and historical construction records of the target foundation pit; Extract the structured design parameter entries from the design drawings and perform logical consistency verification on the structured design parameter entries to obtain the preliminary design parameter set of the target foundation pit; Data cleaning was performed on the time-series monitoring data in the historical construction condition records to obtain the standard historical condition dataset of the target foundation pit. The deformation values, deformation rates, and spatial distribution characteristics of the target foundation pit during key construction stages are extracted from the standard historical working condition dataset to form a historical deformation feature library. The preliminary design parameter set is mapped to the mechanical constraint parameter vector of the target foundation pit, and a dynamic deformation behavior model of the target foundation pit is constructed based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector.

[0008] In a preferred embodiment, the step of mapping the preliminary design parameter set to the mechanical constraint parameter vector of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector, includes: The support stiffness-related parameters and excavation geometry parameters are selected from the preliminary design parameter set, and the dimensions of the support stiffness-related parameters and excavation geometry parameters are unified to obtain the current mechanical characteristic vector of the target foundation pit. Statistical quantification of historical working conditions in the historical deformation feature library is performed to obtain the historical deformation feature vector of the target foundation pit; Analyze the comprehensive distance metric between the current mechanical eigenvectors and the historical deformation eigenvectors; Based on the similarity criterion, determine the contribution weight corresponding to the historical working conditions; Based on contribution weights, historical deformation feature vectors are fused to obtain the common deformation pattern vector of the target foundation pit; The common deformation mode vector and the current mechanical characteristic vector are linearly superimposed, and the superimposed vector is nonlinearly corrected to obtain the mechanical-deformation coupling action sequence of the target foundation pit. Based on the mechanical-deformation coupling sequence as the core prediction basis, a dynamic deformation behavior model of the target foundation pit is constructed.

[0009] In a preferred embodiment, the step of inputting the current construction progress of the target foundation pit into the dynamic deformation behavior model and calculating the expected deformation reference sequence of the target foundation pit includes: Obtain the excavation depth, support level number, and application time of the target foundation pit to obtain the current construction progress information of the target foundation pit; Based on the dynamic deformation behavior model, the current construction progress information is converted into the temporal state vector of the target foundation pit; The temporal state vector is input into the dynamic deformation behavior model to drive the coupling relationship in the dynamic deformation behavior model and calculate the preliminary prediction sequence of the target foundation pit. Based on the ambient humidity and temperature of the target foundation pit, the preliminary prediction sequence is corrected in real time to obtain the optimized prediction sequence of the target foundation pit. Short-term trend smoothing is applied to the optimized prediction sequence to obtain the expected deformation reference sequence of the target foundation pit.

[0010] In a preferred embodiment, the step of comparing the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit includes: The standardized monitoring sequence is aligned point-by-point with the expected deformation reference sequence to obtain the data pair sequence of the target network; Error analysis was performed on the data sequence to obtain the directional deviation between the standardized monitoring sequence and the expected deformation reference sequence; By tracking directional deviations along the time axis, potential abnormal development windows of the target network can be identified. The monitoring points in the potential abnormal development window are mapped to the spatial layout map of the distributed monitoring network, and the spatial clustering of the monitoring points is analyzed. By jointly determining the temporal persistence and spatial clustering of potential abnormal development windows, abnormal events in the target network can be obtained. Mechanism identification of abnormal events yields the local non-coordinated deformation mode of the target foundation pit.

[0011] In a preferred embodiment, when the local non-coordinated deformation mode triggers a preset early risk criterion, an abnormal structural state warning for the target foundation pit is generated. Based on the abnormal structural state warning, the real-time construction logs and geological conditions of the target foundation pit are traced back to their root causes to obtain a graded response strategy for the target foundation pit, including: Extract the spatiotemporal feature labels corresponding to the local non-coordinated deformation modes, and encapsulate the spatiotemporal feature labels into structured early warning information for the target foundation pit; By matching the structured early warning information with the preset early risk criteria, the early warning risk level and potential impact type of the target foundation pit can be obtained. Based on the warning risk level and potential impact type, the real-time construction log of the target foundation pit is traced back to obtain the set of suspected construction activities of the target foundation pit; By spatially overlaying and analyzing the spatial distribution range of structured early warning information and the digital geological profile of the target foundation pit, a set of suspected geological conditions for the target foundation pit is obtained. A correlation analysis was conducted on the suspected construction activity set and the suspected geological condition set to obtain the correlation relationship between them. Based on the correlation, strategy mapping is performed on the suspected construction activity set and the suspected geological condition set to obtain the graded response strategy for the target foundation pit.

[0012] In a preferred embodiment, the step of dynamically adjusting the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy includes: By analyzing the graded response strategy, we can obtain the spatial boundary information, risk level identification, and recommended monitoring intensity level of the risk area in the target foundation pit. Based on spatial boundary information, the target adjustment area is delineated in the spatial topology map of the distributed monitoring network; Based on the risk level identification and the sampling frequency of the target adjustment area, generate individualized frequency adjustment instructions for the target foundation pit; Execute individualized frequency adjustment commands to adjust the sampling frequency of the distributed monitoring network; Based on the recommended monitoring intensity level and risk level identifiers, the key nodes and associated edges in the spatial topology map are updated to adjust the spatial density in the distributed monitoring network.

[0013] In a preferred embodiment, the rolling optimization of parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit includes: Collect newly added monitoring data from the updated distributed monitoring network and process the new monitoring data into an optimized verification sequence that is synchronized with the dynamic deformation behavior model in time; By aligning and comparing the optimized verification sequence with the expected deformation reference sequence, the deviation distribution pattern and deviation evolution trend of the target foundation pit can be obtained. Based on the deviation distribution pattern and deviation evolution trend, the parameters in the dynamic deformation behavior model are updated to obtain the optimized deformation behavior model of the target foundation pit. The accuracy of the optimized deformation behavior model is verified in order to achieve complete monitoring of the target foundation pit.

[0014] To address the above problems, the present invention also provides a foundation pit deformation monitoring system, the system comprising: The data acquisition module is used to perform consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit, and obtain the standardized monitoring sequence of the target foundation pit; The model building module is used to obtain the design parameter set and historical construction condition records of the target foundation pit, and to build a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. The data prediction module is used to input the current construction progress of the target foundation pit into the dynamic deformation behavior model and calculate the expected deformation reference sequence of the target foundation pit. The data comparison module is used to compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit. The strategy generation module is used to generate an early warning of structural anomalies in the target foundation pit when the local non-coordinated deformation mode triggers the preset early risk criteria. Based on the early warning of structural anomalies, the module traces the root causes of the target foundation pit's real-time construction logs and geological conditions to obtain a graded response strategy for the target foundation pit. The feedback optimization module is used to dynamically adjust the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy, and to perform rolling optimization of the parameters in the dynamic deformation behavior model in order to achieve complete monitoring of the target foundation pit.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention employs a rigorous data consistency verification process to eliminate spatially isolated anomalies and complete spatiotemporal interpolation repair of missing data, ensuring the reliability and integrity of standardized monitoring sequences and providing high-quality data support for subsequent deformation analysis. Based on the mechanical constraints of design parameters and the deformation evolution patterns of historical construction conditions, a dynamic deformation behavior model is constructed. Combined with the current construction progress, the expected deformation reference sequence is accurately calculated. Through real-time comparison, local non-coordinated deformation patterns are quickly identified, significantly improving the accuracy of foundation pit deformation monitoring and the timeliness of anomaly identification.

[0016] This invention dynamically adjusts the sampling frequency and spatial density of the distributed monitoring network through a hierarchical response strategy, achieving precise allocation and efficient utilization of monitoring resources. This ensures that risk areas receive focused attention, and the dynamic deformation behavior model is continuously optimized by adding new monitoring data to adapt to changes in the foundation pit construction process and geological conditions. This gives the monitoring system good dynamic adaptability, effectively ensuring complete monitoring of foundation pit deformation throughout the entire life cycle and in all aspects, and further strengthening the safety assurance capability during foundation pit construction. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for monitoring foundation pit deformation according to an embodiment of the present invention. Figure 2 This is a functional block diagram of a foundation pit deformation monitoring system provided in an embodiment of the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] This application provides a method for monitoring foundation pit deformation. The execution subject of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the foundation pit deformation monitoring method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0021] Reference Figure 1 The diagram shown is a flowchart illustrating a method for monitoring foundation pit deformation according to an embodiment of the present invention. In this embodiment, the method for monitoring foundation pit deformation includes: S1. Perform consistency verification on the original physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain the standardized monitoring sequence of the target foundation pit; In this embodiment of the invention, the step of performing consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain a standardized monitoring sequence for the target foundation pit includes: Receive raw physical quantity monitoring data collected from the distributed monitoring network in the target foundation pit, and perform time-series alignment on the raw physical quantity monitoring data to obtain the time-series monitoring data stream of the target foundation pit; Based on the spatial deployment coordinates of the distributed monitoring network, a topological relationship diagram of the distributed monitoring network is constructed. Based on the topological correlation diagram, the continuity and rationality of the time-series monitoring data stream in spatial transmission are evaluated, and the spatiotemporal correlation coefficient of the target foundation pit is obtained. The maximum coefficient in the spatiotemporal correlation coefficient is used as the upper limit, and the minimum coefficient in the spatiotemporal correlation coefficient is used as the lower limit to construct the data consistency judgment threshold for the target foundation pit; The degree of deviation in the time-series monitoring data stream is obtained by comparing each monitoring data item in the time-series monitoring data stream. When the deviation exceeds the data consistency judgment threshold, the monitoring data will be marked as a spatially isolated anomaly of the target foundation pit; Spatially isolated outliers were removed, and the missing data was repaired by spatiotemporal interpolation to obtain a standardized monitoring sequence for the target foundation pit.

[0022] The system receives raw physical quantity monitoring data collected from multiple monitoring nodes in the distributed monitoring network within the target foundation pit. Each monitoring node acquires data according to a preset acquisition cycle. During data transmission, the acquisition timestamp corresponding to each data point is recorded synchronously. All data is centrally aggregated to the processing terminal through a dedicated data transmission channel. Using the earliest acquisition time among all monitoring nodes as the starting benchmark, the raw data of each monitoring node is regularized at fixed time intervals. The data collected by different nodes within the same time interval are matched one-to-one to form a time-series monitoring data stream with a unified time dimension and continuous data.

[0023] The planar coordinates and elevation information of each monitoring node in the distributed monitoring network are obtained through professional surveying. These coordinates accurately correspond to the specific location of the actual construction area of ​​the target foundation pit. Each monitoring node is used as an independent vertex in the topology graph. Based on the actual spatial distance and relative orientation between nodes, the correlation of data transmission between nodes is determined. Spatially adjacent nodes with logical data transmission relationships are connected by line segments. Finally, a topology graph that can clearly reflect the spatial structure of the monitoring network and the logic of node association is constructed.

[0024] Based on the connection paths of each node in the topological association diagram, and following the logical direction of natural data transmission, the data changes of adjacent nodes in the time-series monitoring data stream at corresponding time points are checked one by one. This verifies whether the data of the next node can be reasonably deduced from the data change trend of the previous node. At the same time, the consistency of node data on different transmission paths within the same time period is checked. Through quantitative evaluation of the continuity of data transmission and the rationality of deduction, a spatiotemporal correlation coefficient that can accurately reflect the degree of spatiotemporal correlation of data is obtained.

[0025] From all the calculated spatiotemporal correlation coefficients, the coefficient with the largest value is selected and set as the upper limit of the data consistency judgment threshold. The coefficient with the smallest value is selected and set as the lower limit of the data consistency judgment threshold. The upper and lower limits together constitute a clear range of values, forming the data consistency judgment threshold used to determine whether the monitoring data meets the spatiotemporal consistency requirements.

[0026] According to the chronological order, each data point in the time-series monitoring data stream is compared one by one with the monitoring data of nodes that are spatially related at the same time point. The specific difference between the individual data and the related data is calculated. Combined with the overall data fluctuation of the time interval in which the data is located and the distribution characteristics of adjacent data in the spatial location, the degree of deviation of the data relative to the overall time-series monitoring data stream is determined, that is, the deviation degree of the time-series monitoring data stream is obtained.

[0027] The deviation of each data point is directly compared with the established data consistency judgment threshold. If the deviation of a certain data point is higher than the upper limit of the threshold or lower than the lower limit of the threshold, it indicates that the data cannot reasonably correspond with other normal data in the spatiotemporal correlation dimension and does not have the continuity and rationality of data transmission. The data point is directly marked as a spatially isolated anomaly point of the target foundation pit.

[0028] All data marked as spatially isolated anomalies were completely removed from the time-series monitoring data stream. For the gaps in the data stream after the anomalies were removed, normal monitoring data at adjacent time points before and after the gaps, as well as normal data at the same time points of monitoring nodes spatially adjacent to the corresponding nodes at the gaps, were used to calculate reasonable data for the gaps according to the temporal variation patterns and spatial distribution characteristics of the data. This process was then used to fill the gaps, ultimately forming a standardized monitoring sequence that is complete, meets spatiotemporal consistency standards, and can be directly used for subsequent analysis.

[0029] The beneficial effects are that the above process systematically processes the original physical quantity monitoring data, effectively eliminating abnormal data without spatiotemporal correlation, and filling data gaps through precise interpolation repair, ensuring the data integrity and reliability of the standardized monitoring sequence, providing a high-quality data foundation for subsequent steps such as the construction of dynamic deformation behavior models and the calculation of expected deformation reference sequences, and ensuring the accuracy and effectiveness of the overall process of foundation pit deformation monitoring.

[0030] S2. Obtain the design parameter set and historical construction condition records of the target foundation pit, and construct a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. In this embodiment of the invention, the step of obtaining the design parameter set and historical construction condition records of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records, includes: Obtain the design drawings and historical construction records of the target foundation pit; Extract the structured design parameter entries from the design drawings and perform logical consistency verification on the structured design parameter entries to obtain the preliminary design parameter set of the target foundation pit; Data cleaning was performed on the time-series monitoring data in the historical construction condition records to obtain the standard historical condition dataset of the target foundation pit. The deformation values, deformation rates, and spatial distribution characteristics of the target foundation pit during key construction stages are extracted from the standard historical working condition dataset to form a historical deformation feature library. The preliminary design parameter set is mapped to the mechanical constraint parameter vector of the target foundation pit, and a dynamic deformation behavior model of the target foundation pit is constructed based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector.

[0031] The process of mapping the preliminary design parameter set to the mechanical constraint parameter vector of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector, includes: The support stiffness-related parameters and excavation geometry parameters are selected from the preliminary design parameter set, and the dimensions of the support stiffness-related parameters and excavation geometry parameters are unified to obtain the current mechanical characteristic vector of the target foundation pit. Statistical quantification of historical working conditions in the historical deformation feature library is performed to obtain the historical deformation feature vector of the target foundation pit; Analyze the comprehensive distance metric between the current mechanical eigenvectors and the historical deformation eigenvectors; Based on the similarity criterion, determine the contribution weight corresponding to the historical working conditions; Based on contribution weights, historical deformation feature vectors are fused to obtain the common deformation pattern vector of the target foundation pit; The common deformation mode vector and the current mechanical characteristic vector are linearly superimposed, and the superimposed vector is nonlinearly corrected to obtain the mechanical-deformation coupling action sequence of the target foundation pit. Based on the mechanical-deformation coupling sequence as the core prediction basis, a dynamic deformation behavior model of the target foundation pit is constructed.

[0032] Obtain the design drawings and historical construction records of the target foundation pit. The design drawings are retrieved from the construction unit's technical archives and contain key information such as the dimensions of the support structure, material strength, and excavation depth. The historical construction records are extracted from the project construction management system and cover complete information such as the time nodes, construction techniques, and monitoring data records of each stage of past construction.

[0033] Extract the structural design parameters from the design drawings, including the cross-sectional dimensions of the support structure, the elastic modulus of the material, the support spacing, the excavation layer thickness, and the slope gradient. Perform logical consistency checks on these structural parameter items, verify the matching relationship between each parameter, eliminate contradictory or illogical parameter items, and compile the remaining valid parameters to form the preliminary design parameter set for the target foundation pit.

[0034] Data cleaning was performed on the time-series monitoring data in the historical construction condition records to remove abnormal data that exceeded the reasonable monitoring range, correct data with recording errors, and supplement monitoring data missing at key time nodes. At the same time, the time-series monitoring data was classified and organized according to the construction stage to ensure that the monitoring data of each construction stage is continuous, complete and accurate, and finally the standard historical condition dataset of the target foundation pit was obtained.

[0035] Deformation values, deformation rates, and spatial distribution characteristics of the target foundation pit during key construction stages are extracted from the standard historical working condition dataset. Key construction stages include the initial stage of foundation pit excavation, the middle stage of excavation, the completion of the support structure, and the pre-sealing stage of the foundation pit. Deformation values ​​are extracted from the actual deformation values ​​of different monitoring points at each stage. The deformation rate is obtained by calculating the ratio of the deformation difference between adjacent monitoring time points to the time interval. Spatial distribution characteristics are obtained by analyzing the spatial correlation of deformation differences and deformation trends at different monitoring points. These characteristics are classified and stored according to construction stage and deformation type to form a historical deformation feature library.

[0036] The support stiffness-related parameters and excavation geometry parameters are selected from the preliminary design parameter set. The support stiffness-related parameters include the material elastic modulus, the moment of inertia of the support structure section, and the support spacing. The excavation geometry parameters include the excavation depth, excavation width, layer excavation thickness, and slope angle. These parameters in different units are converted into a unified measurement standard, and then all parameters with unified dimensions are arranged in a preset order to obtain the current mechanical characteristic vector of the target foundation pit.

[0037] For each historical working condition in the historical deformation feature database, statistical quantification is performed. The average value of deformation, the maximum value of deformation, the peak value of deformation rate, and the deformation stabilization time of the key construction stages under the working condition are statistically analyzed. At the same time, the spatial distribution characteristics are converted into quantifiable descriptive parameters, such as the difference coefficient of deformation in different regions and the consistency index of deformation trend. These quantified features are combined in a fixed order to obtain the historical deformation feature vector of the target foundation pit corresponding to each historical working condition.

[0038] By analyzing the comprehensive distance metric between the current mechanical feature vector and each historical deformation feature vector, and by comparing the differences in the corresponding parameters between the two, and combining the importance of each parameter to the deformation of the foundation pit, the overall difference quantification result that can intuitively reflect the similarity between the current mechanical conditions of the foundation pit and the mechanical conditions of historical working conditions is calculated.

[0039] The contribution weights corresponding to historical working conditions are determined based on similarity criteria. The core basis of the similarity criteria is the comprehensive distance metric. The smaller the comprehensive distance metric, the more similar the historical working conditions are to the current foundation pit conditions, and the greater the corresponding contribution weight. The larger the comprehensive distance metric, the smaller the contribution weight of the historical working conditions. The importance of each historical working condition in constructing the dynamic deformation behavior model is clarified and a corresponding weight is assigned.

[0040] The formula for calculating the contribution weight is as follows: ; in, Indicates the first Each contribution weight Indicates the first The comprehensive distance metric corresponding to each historical operating condition This represents the minimum value of the overall distance metric. This represents the maximum value of the overall distance metric. This represents the preset weight distribution adjustment factor. This represents the total number of historical operating conditions. Index variables representing historical operating conditions. Indicates the first The comprehensive distance metric corresponding to each historical operating condition This represents the natural exponential function.

[0041] The current mechanical feature vector is obtained by selecting support stiffness-related parameters and excavation geometric parameters from the preliminary design parameter set and unifying the dimensions of these two types of parameters. The historical deformation feature vector is obtained by statistically quantifying historical working conditions in the historical deformation feature library. The comprehensive distance metric is obtained by analyzing the relationship between the current mechanical feature vector and the historical deformation feature vector of the historical working condition. The minimum value of the comprehensive distance metric is the minimum value selected from the comprehensive distance metrics corresponding to all historical working conditions. The maximum value of the comprehensive distance metric is the maximum value selected from the comprehensive distance metrics corresponding to all historical working conditions. The preset weight distribution adjustment factor is a fixed value set in advance to adjust the weight distribution. The total number of historical working conditions is the specific number obtained by statistically analyzing the historical working conditions contained in the historical deformation feature library.

[0042] When calculating the contribution weight of each historical working condition, first calculate the result of subtracting the minimum value of the comprehensive distance metric from the result of the comprehensive distance metric corresponding to that historical working condition, and then calculate the result of subtracting the minimum value of the comprehensive distance metric from the result of the maximum value of the comprehensive distance metric. Divide the former result by the latter result to obtain a ratio. Multiply this ratio by a negative preset weight distribution adjustment factor to obtain a product. Substitute this product into the natural exponential function to calculate the natural exponential function result corresponding to that historical working condition. Perform the same calculation process for all historical working conditions to obtain the natural exponential function result corresponding to each historical working condition. Summate the natural exponential function results corresponding to all historical working conditions to obtain a total result. Divide the natural exponential function result corresponding to a single historical working condition by the total result to obtain the value of the contribution weight of that historical working condition. The core of this calculation process is to allocate contribution weights based on the similarity between the historical working condition and the current mechanical feature vector. The higher the similarity, the greater the contribution weight of the historical working condition, which makes the constructed dynamic deformation behavior model more consistent with the actual deformation law of the target foundation pit.

[0043] When the preset weight distribution adjustment factor remains fixed, the larger the comprehensive distance metric value corresponding to a certain historical working condition, the smaller the calculated natural exponential function result. The smaller the ratio of this natural exponential function result to the sum of the natural exponential function results of all historical working conditions, the smaller the corresponding contribution weight. Conversely, the smaller the comprehensive distance metric value corresponding to a certain historical working condition, the larger the calculated natural exponential function result. The larger the ratio of this natural exponential function result to the sum of the natural exponential function results of all historical working conditions, the larger the corresponding contribution weight. The sum of the contribution weights of all historical working conditions always equals 1. The contribution weight is maximized when the comprehensive distance metric value equals the minimum value, and minimized when the comprehensive distance metric value equals the maximum value.

[0044] Based on the contribution weight fusion of historical deformation feature vectors, a weighted summation operation is performed according to the contribution weight corresponding to each historical deformation feature vector. This allows historical deformation feature vectors with large contribution weights to dominate the fusion process, while those with small contribution weights play an auxiliary and supplementary role. Through this weighted fusion, common deformation features in different historical working conditions are extracted, resulting in a common deformation pattern vector of the target foundation pit.

[0045] The common deformation mode vector and the current mechanical characteristic vector are linearly superimposed, and the corresponding elements in the two are directly added by value to obtain the preliminary superimposed vector. Then, combined with the actual mechanical law of the foundation pit deformation, and based on the influence of material nonlinearity and geometric nonlinearity on deformation, the elements in the preliminary superimposed vector that do not conform to the nonlinear deformation characteristics are adjusted to obtain the mechanical-deformation coupling action sequence of the target foundation pit.

[0046] Based on the mechanical-deformation coupling sequence as the core prediction basis, the correspondence between each element in the sequence and the deformation amount, deformation rate and deformation trend of the foundation pit is clarified. A dynamic model structure that can simulate and predict the deformation development of the foundation pit according to the current mechanical conditions and construction process is established, and a dynamic deformation behavior model of the target foundation pit is constructed.

[0047] The beneficial effects are that by extracting design parameters and historical working condition characteristics through the system, and through multi-stage verification, cleaning and quantification, the accuracy and completeness of the basic data for model construction are ensured. At the same time, by combining current mechanical conditions and historical common deformation characteristics, and through a scientific process of fusion, superposition and correction, a coupling action sequence that can accurately reflect the relationship between mechanical conditions and deformation is formed. The constructed dynamic deformation behavior model has the characteristics of conforming to the actual deformation law of the foundation pit, which provides solid support for the accurate calculation of the subsequent expected deformation reference sequence, and effectively improves the scientificity and foresight of foundation pit deformation monitoring.

[0048] S3. Input the current construction progress of the target foundation pit into the dynamic deformation behavior model to calculate the expected deformation reference sequence of the target foundation pit; In this embodiment of the invention, the step of inputting the current construction progress of the target foundation pit into the dynamic deformation behavior model and calculating the expected deformation reference sequence of the target foundation pit includes: Obtain the excavation depth, support level number, and application time of the target foundation pit to obtain the current construction progress information of the target foundation pit; Based on the dynamic deformation behavior model, the current construction progress information is converted into the temporal state vector of the target foundation pit; The temporal state vector is input into the dynamic deformation behavior model to drive the coupling relationship in the dynamic deformation behavior model and calculate the preliminary prediction sequence of the target foundation pit. Based on the ambient humidity and temperature of the target foundation pit, the preliminary prediction sequence is corrected in real time to obtain the optimized prediction sequence of the target foundation pit. Short-term trend smoothing is applied to the optimized prediction sequence to obtain the expected deformation reference sequence of the target foundation pit.

[0049] The current excavation depth data of the target foundation pit is read in real time by depth sensors deployed at the construction site. The support level number is confirmed from the construction technical archives and on-site support structure identification. The application time node of the support structure is extracted from the real-time records of the construction management system. These three types of information are organized and integrated according to the time logic and data type of the construction process to obtain the current construction progress information of the target foundation pit.

[0050] The dynamic deformation behavior model incorporates a mapping rule between construction progress information and vector data. According to this rule, the excavation depth value, the standardized code corresponding to the support level number, and the time quantization data of the application time node in the current construction progress information are arranged into an ordered data combination, which is then converted into a time-series state vector of the target foundation pit that can be directly processed by the dynamic deformation behavior model.

[0051] After the temporal state vector is input into the dynamic deformation behavior model, the construction progress information in the vector will trigger the coupling relationship between the pre-established mechanical constraints and deformation evolution law inside the model. Based on this coupling relationship, the model simulates the development path of the foundation pit deformation under the current construction conditions and calculates the foundation pit deformation prediction value corresponding to different time nodes. These prediction values ​​are arranged in chronological order to calculate the preliminary prediction sequence of the target foundation pit.

[0052] By uniformly deploying temperature and humidity monitoring equipment around the target foundation pit, environmental humidity and temperature data are collected in real time. The influence of these two environmental parameters on the physical properties of the foundation pit soil and the mechanical performance of the support structure is analyzed. Based on this law, the deformation prediction value at each time node in the preliminary prediction sequence is fine-tuned to correct the prediction deviation that may be caused by environmental factors, and the optimized prediction sequence of the target foundation pit is obtained.

[0053] The moving average method is used to smooth the short-term trend of the optimized prediction sequence. The arithmetic mean of three consecutive adjacent data points in the optimized prediction sequence is calculated and the average value is used to replace the data point in the middle position of the sequence. The entire optimized prediction sequence is processed point by point to eliminate random fluctuations in the sequence, making the deformation trend more stable and consistent, and thus obtaining the expected deformation reference sequence of the target foundation pit.

[0054] The beneficial effects are that by accurately collecting key information on the construction process and converting it into a vector form that the model can recognize, and combining it with environmental factors for targeted correction and trend smoothing, the expected deformation reference sequence can be closely matched with the actual construction conditions and environmental conditions of the foundation pit. It has extremely high accuracy and reliability, and provides an accurate benchmark for the comparison between the subsequent standardized monitoring sequence and the expected deformation reference sequence, effectively ensuring the accuracy of local non-coordinated deformation pattern recognition.

[0055] S4. Compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit; In this embodiment of the invention, the step of comparing the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit includes: The standardized monitoring sequence is aligned point-by-point with the expected deformation reference sequence to obtain the data pair sequence of the target network; Error analysis was performed on the data sequence to obtain the directional deviation between the standardized monitoring sequence and the expected deformation reference sequence; By tracking directional deviations along the time axis, potential abnormal development windows of the target network can be identified. The monitoring points in the potential abnormal development window are mapped to the spatial layout map of the distributed monitoring network, and the spatial clustering of the monitoring points is analyzed. By jointly determining the temporal persistence and spatial clustering of potential abnormal development windows, abnormal events in the target network can be obtained. Mechanism identification of abnormal events yields the local non-coordinated deformation mode of the target foundation pit.

[0056] Using timestamps as the sole matching criterion, monitoring data corresponding to the same monitoring time point and the same monitoring location in the standardized monitoring sequence and the expected deformation reference sequence are paired one by one. Each pair of data contains the standardized monitoring value and the expected deformation reference value at that time point and location. All paired data are arranged in chronological order of monitoring time to obtain the data pair sequence of the target network.

[0057] For each pair of data in the data pair sequence, the corresponding expected deformation reference value is subtracted from the standardized monitoring value to obtain the absolute error value of each data pair. At the same time, the positive and negative attributes of the error value are recorded. By analyzing the positive and negative changes of the error and the increase and decrease trend of the error value in multiple consecutive data pairs, the deviation of the standardized monitoring sequence from the expected deformation reference sequence in the deformation direction is clarified, and the directional deviation between the standardized monitoring sequence and the expected deformation reference sequence is obtained.

[0058] Using the timeline as a guide, the changes in directional deviation are continuously tracked in the order of monitoring time. A fixed time window length standard is set. When the directional deviations of multiple consecutive monitoring time points maintain the same deviation trend and the deviation values ​​do not return to a reasonable range, this continuous time interval is defined as a period that needs to be focused on, thus identifying potential abnormal development windows of the target network.

[0059] Extract the spatial coordinates of all monitoring points with directional deviations within the potential anomaly development window, accurately map the coordinates of these monitoring points onto the spatial layout map of the distributed monitoring network, and determine whether these monitoring points are concentrated in a specific spatial range by counting the number of abnormal monitoring points in the area and calculating the average distance between abnormal monitoring points, thus analyzing the spatial clustering of monitoring points.

[0060] Set time-duration criteria and spatial clustering criteria. The time-duration criteria is that the duration of the potential anomaly development window reaches the preset duration, and the spatial clustering criteria is that the clustering density of the anomaly monitoring points reaches the preset density value. When the potential anomaly development window meets both criteria at the same time, it is confirmed that an unexpected deformation phenomenon has occurred in the time period and spatial area, and the anomaly event of the target network is obtained.

[0061] Collect information on the construction stage, geological conditions, and deformation characteristics of monitoring points corresponding to the abnormal event. Analyze the intrinsic relationship between abnormal deformation and construction operations and geological structure changes. Clarify the causes of abnormal deformation, the rate of deformation development, and the characteristics of spatial influence range. Summarize the unique deformation manifestations corresponding to the abnormal event, identify the mechanism of the abnormal event, and obtain the local non-coordinated deformation mode of the target foundation pit.

[0062] The beneficial effects are that through a systematic process of point-by-point alignment, error analysis, time-series tracking, spatial aggregation analysis, and joint judgment, the accurate location and comprehensive identification of foundation pit deformation anomalies are achieved. The mechanism identification process further clarifies the essential characteristics of local non-coordinated deformation, ensuring the accuracy and pertinence of the identification results. This provides a reliable basis for the generation of subsequent structural anomaly warnings and the formulation of graded response strategies, effectively improving the accuracy and effectiveness of foundation pit deformation risk monitoring.

[0063] S5. When the local non-coordinated deformation mode triggers the preset early risk criterion, an abnormal structural state warning of the target foundation pit is generated. Based on the abnormal structural state warning, the real-time construction log and geological conditions of the target foundation pit are traced back to their root causes to obtain the graded response strategy of the target foundation pit. In this embodiment of the invention, when the local non-coordinated deformation mode triggers a preset early risk criterion, an abnormal structural state warning for the target foundation pit is generated. Based on the abnormal structural state warning, the real-time construction logs and geological conditions of the target foundation pit are traced back to their root causes to obtain a graded response strategy for the target foundation pit, including: Extract the spatiotemporal feature labels corresponding to the local non-coordinated deformation modes, and encapsulate the spatiotemporal feature labels into structured early warning information for the target foundation pit; By matching the structured early warning information with the preset early risk criteria, the early warning risk level and potential impact type of the target foundation pit can be obtained. Based on the warning risk level and potential impact type, the real-time construction log of the target foundation pit is traced back to obtain the set of suspected construction activities of the target foundation pit; By spatially overlaying and analyzing the spatial distribution range of structured early warning information and the digital geological profile of the target foundation pit, a set of suspected geological conditions for the target foundation pit is obtained. A correlation analysis was conducted on the suspected construction activity set and the suspected geological condition set to obtain the correlation relationship between them. Based on the correlation, strategy mapping is performed on the suspected construction activity set and the suspected geological condition set to obtain the graded response strategy for the target foundation pit.

[0064] Extract the spatiotemporal feature labels corresponding to the local non-coordinated deformation patterns. The temporal feature labels include specific information such as the start time, duration, and rhythm of deformation anomalies. The spatial feature labels include the coordinates of the monitoring points where abnormal deformation occurs, the coverage area, and the spatial distribution pattern. These temporal and spatial feature labels are classified and organized according to a preset data structure and uniformly packaged into structured early warning information for the target foundation pit containing complete spatiotemporal information.

[0065] The structured early warning information is matched with the preset early risk criteria. The preset early risk criteria include corresponding rules for different combinations of spatiotemporal features and risk levels and impact types. The spatiotemporal features in the structured early warning information are compared with the feature combinations in the criteria one by one to find the perfect or closest matching item. Based on the item, the early warning risk level and potential impact type of the target foundation pit are determined. The early warning risk level clarifies the urgency of the risk, and the potential impact type clarifies the type of damage that deformation may cause to the foundation pit structure.

[0066] Based on the warning risk level and potential impact type, the real-time construction log of the target foundation pit was reviewed. The real-time construction log recorded in detail the construction content, construction technology, operation parameters, personnel and equipment configuration, etc. of each time period and construction area. According to the scope of concern defined by the warning risk level and the relevant construction links pointed to by the potential impact type, the construction activities carried out during the period of abnormal deformation, in the abnormal area and surrounding area were screened out. These construction activities were summarized and organized to obtain the suspected construction activities set of the target foundation pit.

[0067] The spatial distribution range of structured early warning information is transformed into precise coordinate boundary data. A digital geological profile of the target foundation pit is retrieved. This profile contains detailed geological information such as soil layer distribution, rock and soil properties, and groundwater depth throughout the foundation pit area. Through spatial coordinate alignment technology, the coordinate boundaries of the structured early warning information are superimposed on the digital geological profile. Geological condition data within the superimposed area are accurately extracted, and geological factors that may induce non-coordinated deformation are screened out to obtain a set of suspected geological conditions for the target foundation pit.

[0068] A correlation analysis was conducted on the suspected construction activity set and the suspected geological condition set. The interaction relationship between each suspected construction activity and each suspected geological condition was checked one by one to determine whether the construction activity would trigger or aggravate the deformation anomaly under the corresponding geological condition. At the same time, it was confirmed whether the geological condition would amplify the impact of the construction activity on the foundation pit deformation. It was determined which suspected construction activities and which suspected geological conditions were directly related, indirectly related, or synergistically related, and the correlation relationship between the suspected construction activity set and the suspected geological condition set was sorted out.

[0069] Based on the correlation, a strategy mapping is performed on the suspected construction activity set and the suspected geological condition set. A database corresponding to the correlation and response strategy is established in advance. The database contains construction adjustment plans, risk control measures, emergency response procedures, etc. for different correlation scenarios. Based on the established correlation, the appropriate response strategy is retrieved from the database. Combined with the execution priority and implementation intensity of the strategy for classifying the early warning risk level, a hierarchical response strategy for the target foundation pit covering multiple aspects such as risk control, problem rectification, and continuous monitoring is formed.

[0070] The beneficial effects are that by extracting spatiotemporal feature labels and encapsulating structured early warning information, the early warning information is presented in a standardized manner, providing a clear basis for subsequent risk assessment and root cause tracing. After accurately determining the risk level and impact type through pattern matching, targeted backtracking of construction logs and geological spatial overlay analysis are carried out to efficiently identify suspected factors and clarify their correlations. Finally, a graded response strategy that fits the actual situation is generated through strategy mapping, ensuring the pertinence and effectiveness of response measures, providing strong protection for the safety of the foundation pit structure, and significantly improving the efficiency and quality of handling and managing foundation pit deformation risks.

[0071] S6. Based on the hierarchical response strategy, dynamically adjust the sampling frequency and spatial density of the distributed monitoring network, and perform rolling optimization on the parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit.

[0072] In this embodiment of the invention, the step of dynamically adjusting the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy includes: By analyzing the graded response strategy, we can obtain the spatial boundary information, risk level identification, and recommended monitoring intensity level of the risk area in the target foundation pit. Based on spatial boundary information, the target adjustment area is delineated in the spatial topology map of the distributed monitoring network; Based on the risk level identification and the sampling frequency of the target adjustment area, generate individualized frequency adjustment instructions for the target foundation pit; Execute individualized frequency adjustment commands to adjust the sampling frequency of the distributed monitoring network; Based on the recommended monitoring intensity level and risk level identifiers, the key nodes and associated edges in the spatial topology map are updated to adjust the spatial density in the distributed monitoring network.

[0073] The rolling optimization of parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit includes: Collect newly added monitoring data from the updated distributed monitoring network and process the new monitoring data into an optimized verification sequence that is synchronized with the dynamic deformation behavior model in time; By aligning and comparing the optimized verification sequence with the expected deformation reference sequence, the deviation distribution pattern and deviation evolution trend of the target foundation pit can be obtained. Based on the deviation distribution pattern and deviation evolution trend, the parameters in the dynamic deformation behavior model are updated to obtain the optimized deformation behavior model of the target foundation pit. The accuracy of the optimized deformation behavior model is verified in order to achieve complete monitoring of the target foundation pit.

[0074] The hierarchical response strategy is analyzed. This strategy includes specific control requirements for different risk scenarios. By breaking down the structured content of the strategy one by one, the specific coordinate boundaries of the risk area in the target foundation pit, the clear risk level indicators such as high risk, medium risk, and low risk, and the corresponding suggested monitoring intensity levels such as high intensity, medium intensity, and normal intensity are accurately extracted to ensure that the information obtained is complete and fully matches the strategy requirements.

[0075] Based on the extracted spatial boundary information of the risk area, which is defined by precise plane coordinates and elevation range, these coordinate data are unified with the spatial topology map of the distributed monitoring network. By outlining the area contour that perfectly matches the spatial boundary information in the topology map, the target adjustment area that needs to be adjusted for monitoring parameters is clearly defined.

[0076] Based on the risk level identification and the current basic sampling frequency of the target adjustment area, combined with the preset risk level and sampling frequency correspondence rules, such as shortening the sampling interval by 50% for high risk level, shortening the sampling interval by 30% for medium risk level, and keeping the sampling interval unchanged for low risk level, the new sampling interval to be adopted for the target adjustment area is determined. The new sampling interval, the scope of the target adjustment area, the execution time and other information are integrated to generate an individualized frequency adjustment instruction for the target foundation pit.

[0077] The individualized frequency adjustment command is sent to all monitoring nodes within the target adjustment area of ​​the distributed monitoring network through a dedicated data transmission channel. After receiving the command, the monitoring nodes immediately stop the execution of the original sampling frequency and start data acquisition according to the new sampling interval specified in the command, thus completing the adjustment of the sampling frequency of the distributed monitoring network.

[0078] Based on the recommended monitoring intensity level and risk level identification, and following the adjustment rules of high-intensity monitoring corresponding to encrypted nodes, medium-intensity monitoring corresponding to appropriately supplemented nodes, and regular-intensity monitoring corresponding to maintaining core nodes, key nodes within the target adjustment area are added or retained in the spatial topology map. At the same time, the associated edges between nodes are added or adjusted accordingly to match the distribution of associated edges with the node density. Through the updating of nodes and associated edges, the spatial density in the distributed monitoring network is precisely adjusted.

[0079] After the update, the distributed monitoring network collects new monitoring data according to the new sampling frequency and spatial density. This data is then time-aligned to ensure that the timestamp of each data point is completely synchronized with the time scale of the dynamic deformation behavior model. Invalid interference information in the data is then removed, and the processed valid data is arranged in chronological order to obtain an optimized verification sequence that is time-synchronized with the dynamic deformation behavior model.

[0080] Using timestamps as the sole alignment criterion, the data from the optimized verification sequence and the expected deformation reference sequence corresponding to the same monitoring time point and the same monitoring location are paired and compared one by one. The deviation value of each pair of paired data is calculated. By statistically analyzing the distribution of deviation values ​​in different time intervals and different spatial locations, a deviation distribution pattern of the target foundation pit is formed. At the same time, the increase or decrease of deviation values ​​over time is tracked, and the trend of deviation evolution is summarized.

[0081] Based on the deviation distribution pattern, the mechanical constraint parameters and deformation coupling coefficients of the corresponding construction stages in the dynamic deformation behavior model are adjusted for the time intervals and spatial regions where deviations are concentrated. Combining the deviation evolution trend, if the deviation shows a continuous increasing trend, the prediction logic related parameters in the model are optimized; if the deviation shows a fluctuating convergence trend, the stability related parameters in the model are fine-tuned. Through targeted parameter updates, the optimized deformation behavior model of the target foundation pit is obtained.

[0082] New monitoring data that were not included in the parameter update of the optimized deformation behavior model were selected as verification data. The verification data were input into the optimized deformation behavior model to obtain the verification prediction sequence. The verification prediction sequence was compared with the corresponding actual monitoring data, and the degree of agreement between the two was calculated. When the degree of agreement reached the preset qualified standard, the model optimization was confirmed to be effective, thereby achieving complete monitoring of the target foundation pit.

[0083] The beneficial effects are that by analyzing the hierarchical response strategy, the sampling frequency and spatial density of the monitoring network can be precisely adjusted, realizing the on-demand allocation of monitoring resources. This allows risk areas to be monitored in a focused manner while avoiding resource waste. Meanwhile, the rolling optimization of the dynamic deformation behavior model continuously improves the model's predictive accuracy, enabling the monitoring system to dynamically adapt to changes in foundation pit construction. This ensures the continuity, accuracy, and integrity of the entire monitoring process, providing a continuous and reliable guarantee for the safety of foundation pit construction.

[0084] like Figure 2 The diagram shown is a functional block diagram of a foundation pit deformation monitoring system provided in an embodiment of the present invention.

[0085] The foundation pit deformation monitoring system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the foundation pit deformation monitoring system 100 may include a data acquisition module 101, a model building module 102, a data prediction module 103, a data comparison module 104, a strategy generation module 105, and a feedback optimization module 106. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0086] In this embodiment, the functions of each module / unit are as follows: The data acquisition module 101 is used to perform consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit, and obtain the standardized monitoring sequence of the target foundation pit. The model building module 102 is used to obtain the design parameter set and historical construction condition records of the target foundation pit, and to build a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. The data prediction module 103 is used to input the current construction progress of the target foundation pit into the dynamic deformation behavior model and calculate the expected deformation reference sequence of the target foundation pit. The data comparison module 104 is used to compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit. The strategy generation module 105 is used to generate an early warning of structural anomalies in the target foundation pit when the local non-coordinated deformation mode triggers the preset early risk criteria. Based on the early warning of structural anomalies, the real-time construction log and geological conditions of the target foundation pit are traced back to their root causes to obtain the graded response strategy of the target foundation pit. The feedback optimization module 106 is used to dynamically adjust the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy, and to perform rolling optimization of the parameters in the dynamic deformation behavior model in order to achieve complete monitoring of the target foundation pit.

[0087] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0088] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0089] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0090] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0091] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for monitoring the deformation of a foundation pit, characterized in that, The method includes: S1. Perform consistency verification on the original physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain the standardized monitoring sequence of the target foundation pit; S2. Obtain the design parameter set and historical construction condition records of the target foundation pit, and construct a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. S3. Input the current construction progress of the target foundation pit into the dynamic deformation behavior model to calculate the expected deformation reference sequence of the target foundation pit; S4. Compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit; S5. When the local non-coordinated deformation mode triggers the preset early risk criterion, an abnormal structural state warning of the target foundation pit is generated. Based on the abnormal structural state warning, the real-time construction log and geological conditions of the target foundation pit are traced back to their root causes to obtain the graded response strategy of the target foundation pit. S6. Based on the hierarchical response strategy, dynamically adjust the sampling frequency and spatial density of the distributed monitoring network, and perform rolling optimization on the parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit.

2. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The process of performing consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit to obtain the standardized monitoring sequence of the target foundation pit includes: Receive raw physical quantity monitoring data collected from the distributed monitoring network in the target foundation pit, and perform time-series alignment on the raw physical quantity monitoring data to obtain the time-series monitoring data stream of the target foundation pit; Based on the spatial deployment coordinates of the distributed monitoring network, a topological relationship diagram of the distributed monitoring network is constructed. Based on the topological correlation diagram, the continuity and rationality of the time-series monitoring data stream in spatial transmission are evaluated, and the spatiotemporal correlation coefficient of the target foundation pit is obtained. The maximum coefficient in the spatiotemporal correlation coefficient is used as the upper limit, and the minimum coefficient in the spatiotemporal correlation coefficient is used as the lower limit to construct the data consistency judgment threshold for the target foundation pit; The degree of deviation in the time-series monitoring data stream is obtained by comparing each monitoring data item in the time-series monitoring data stream. When the deviation exceeds the data consistency judgment threshold, the monitoring data will be marked as a spatially isolated anomaly of the target foundation pit; Spatially isolated outliers were removed, and the missing data was repaired by spatiotemporal interpolation to obtain a standardized monitoring sequence for the target foundation pit.

3. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The process involves acquiring the design parameter set and historical construction condition records of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution laws in the historical construction condition records. This includes: Obtain the design drawings and historical construction records of the target foundation pit; Extract the structured design parameter entries from the design drawings and perform logical consistency verification on the structured design parameter entries to obtain the preliminary design parameter set of the target foundation pit; Data cleaning was performed on the time-series monitoring data in the historical construction condition records to obtain the standard historical condition dataset of the target foundation pit. The deformation values, deformation rates, and spatial distribution characteristics of the target foundation pit during key construction stages are extracted from the standard historical working condition dataset to form a historical deformation feature library. The preliminary design parameter set is mapped to the mechanical constraint parameter vector of the target foundation pit, and a dynamic deformation behavior model of the target foundation pit is constructed based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector.

4. The method for monitoring foundation pit deformation as described in claim 3, characterized in that, The process of mapping the preliminary design parameter set to the mechanical constraint parameter vector of the target foundation pit, and constructing a dynamic deformation behavior model of the target foundation pit based on the coupling mapping relationship between the feature sequence in the historical deformation feature library and the mechanical constraint parameter vector, includes: The support stiffness-related parameters and excavation geometry parameters are selected from the preliminary design parameter set, and the dimensions of the support stiffness-related parameters and excavation geometry parameters are unified to obtain the current mechanical characteristic vector of the target foundation pit. Statistical quantification of historical working conditions in the historical deformation feature library is performed to obtain the historical deformation feature vector of the target foundation pit; Analyze the comprehensive distance metric between the current mechanical eigenvectors and the historical deformation eigenvectors; Based on the similarity criterion, determine the contribution weight corresponding to the historical working conditions; Based on contribution weights, historical deformation feature vectors are fused to obtain the common deformation pattern vector of the target foundation pit; The common deformation mode vector and the current mechanical characteristic vector are linearly superimposed, and the superimposed vector is nonlinearly corrected to obtain the mechanical-deformation coupling action sequence of the target foundation pit. Based on the mechanical-deformation coupling sequence as the core prediction basis, a dynamic deformation behavior model of the target foundation pit is constructed.

5. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The step of inputting the current construction progress of the target foundation pit into the dynamic deformation behavior model to calculate the expected deformation reference sequence of the target foundation pit includes: Obtain the excavation depth, support level number, and application time of the target foundation pit to obtain the current construction progress information of the target foundation pit; Based on the dynamic deformation behavior model, the current construction progress information is converted into the temporal state vector of the target foundation pit; The temporal state vector is input into the dynamic deformation behavior model to drive the coupling relationship in the dynamic deformation behavior model and calculate the preliminary prediction sequence of the target foundation pit. Based on the ambient humidity and temperature of the target foundation pit, the preliminary prediction sequence is corrected in real time to obtain the optimized prediction sequence of the target foundation pit. Short-term trend smoothing is applied to the optimized prediction sequence to obtain the expected deformation reference sequence of the target foundation pit.

6. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The step of comparing the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit includes: The standardized monitoring sequence is aligned point-by-point with the expected deformation reference sequence to obtain the data pair sequence of the target network; Error analysis was performed on the data sequence to obtain the directional deviation between the standardized monitoring sequence and the expected deformation reference sequence; By tracking directional deviations along the time axis, potential abnormal development windows of the target network can be identified. The monitoring points in the potential abnormal development window are mapped to the spatial layout map of the distributed monitoring network, and the spatial clustering of the monitoring points is analyzed. By jointly determining the temporal persistence and spatial clustering of potential abnormal development windows, abnormal events in the target network can be obtained. Mechanism identification of abnormal events yields the local non-coordinated deformation mode of the target foundation pit.

7. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, When the local non-coordinated deformation mode triggers the preset early risk criterion, an abnormal structural state warning for the target foundation pit is generated. Based on the abnormal structural state warning, the root cause of the abnormality is traced through the real-time construction log and geological conditions of the target foundation pit to obtain a graded response strategy for the target foundation pit, including: Extract the spatiotemporal feature labels corresponding to the local non-coordinated deformation modes, and encapsulate the spatiotemporal feature labels into structured early warning information for the target foundation pit; By matching the structured early warning information with the preset early risk criteria, the early warning risk level and potential impact type of the target foundation pit can be obtained. Based on the warning risk level and potential impact type, the real-time construction log of the target foundation pit is traced back to obtain the set of suspected construction activities of the target foundation pit; By spatially overlaying and analyzing the spatial distribution range of structured early warning information and the digital geological profile of the target foundation pit, a set of suspected geological conditions for the target foundation pit is obtained. A correlation analysis was conducted on the suspected construction activity set and the suspected geological condition set to obtain the correlation relationship between them. Based on the correlation, strategy mapping is performed on the suspected construction activity set and the suspected geological condition set to obtain the graded response strategy for the target foundation pit.

8. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The step of dynamically adjusting the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy includes: By analyzing the graded response strategy, we can obtain the spatial boundary information, risk level identification, and recommended monitoring intensity level of the risk area in the target foundation pit. Based on spatial boundary information, the target adjustment area is delineated in the spatial topology map of the distributed monitoring network; Based on the risk level identification and the sampling frequency of the target adjustment area, generate individualized frequency adjustment instructions for the target foundation pit; Execute individualized frequency adjustment commands to adjust the sampling frequency of the distributed monitoring network; Based on the recommended monitoring intensity level and risk level identifiers, the key nodes and associated edges in the spatial topology map are updated to adjust the spatial density in the distributed monitoring network.

9. The method for monitoring foundation pit deformation as described in claim 1, characterized in that, The rolling optimization of parameters in the dynamic deformation behavior model to achieve complete monitoring of the target foundation pit includes: Collect newly added monitoring data from the updated distributed monitoring network and process the new monitoring data into an optimized verification sequence that is synchronized with the dynamic deformation behavior model in time; By aligning and comparing the optimized verification sequence with the expected deformation reference sequence, the deviation distribution pattern and deviation evolution trend of the target foundation pit can be obtained. Based on the deviation distribution pattern and deviation evolution trend, the parameters in the dynamic deformation behavior model are updated to obtain the optimized deformation behavior model of the target foundation pit. The accuracy of the optimized deformation behavior model is verified in order to achieve complete monitoring of the target foundation pit.

10. A foundation pit deformation monitoring system, characterized in that, The system for implementing the foundation pit deformation monitoring method according to claim 1 includes: The data acquisition module is used to perform consistency verification on the raw physical quantity monitoring data collected by the distributed monitoring network in the target foundation pit, and obtain the standardized monitoring sequence of the target foundation pit; The model building module is used to obtain the design parameter set and historical construction condition records of the target foundation pit, and to build a dynamic deformation behavior model of the target foundation pit based on the coupling relationship between the mechanical constraints in the design parameter set and the actual deformation evolution law in the historical construction condition records. The data prediction module is used to input the current construction progress of the target foundation pit into the dynamic deformation behavior model and calculate the expected deformation reference sequence of the target foundation pit. The data comparison module is used to compare the standardized monitoring sequence with the expected deformation reference sequence in real time to identify the local non-coordinated deformation mode of the target foundation pit. The strategy generation module is used to generate an early warning of structural anomalies in the target foundation pit when the local non-coordinated deformation mode triggers the preset early risk criteria. Based on the early warning of structural anomalies, the module traces the root causes of the target foundation pit's real-time construction logs and geological conditions to obtain a graded response strategy for the target foundation pit. The feedback optimization module is used to dynamically adjust the sampling frequency and spatial density of the distributed monitoring network according to the hierarchical response strategy, and to perform rolling optimization of the parameters in the dynamic deformation behavior model in order to achieve complete monitoring of the target foundation pit.

Images