Engineering measurement data intelligent analysis method and system

Abstract

The invention discloses an engineering measurement data intelligent analysis method and system, and relates to the technical field of engineering data analysis, and the method comprises the steps of data access and fingerprint identification, space-time alignment and dynamic reference construction, semantic feature extraction and causal atlas construction, multi-modal data fusion and intelligent diagnosis, abnormity traceability and causal chain generation, and visual feedback and early warning output. According to the method, the total station, the unmanned aerial vehicle, the GNSS and the optical fiber sensor are used for collecting multi-source data and generating fingerprints, constructing a dynamic reference profile, extracting semantic features and combining construction logs to form a causal atlas, multi-modal fusion and anomaly diagnosis can be achieved, a trigger source can be traced, a causal chain is output, finally, risks are visually marked in a three-dimensional BIM model, and early warning is pushed. The integrity, real-time performance and interpretability of monitoring data are effectively improved, and construction safety and management decision support are enhanced.

Description

Technical Field

[0001] This invention relates to the field of engineering data analysis technology, and in particular to intelligent analysis methods and systems for engineering measurement data. Background Technology

[0002] In engineering surveying and construction monitoring, traditional methods typically rely on a single surveying device to acquire point data, which is then processed manually or using offline software. This approach has the following shortcomings: it cannot simultaneously collect multimodal data such as point clouds, settlement, stress, and attitude, resulting in incomplete monitoring results; data is mostly processed offline, lacking real-time performance and failing to reflect the construction status in a timely manner; it relies on fixed benchmarks, making it susceptible to changes in construction progress and terrain conditions, leading to error accumulation; existing methods can usually only output numerical deviations and cannot reveal the construction behavior or environmental factors behind anomalies; alarms are mostly triggered by a single threshold, lacking anomaly chain tracing and visualization, making it difficult to provide management with actionable decision-making basis; and existing technologies are insufficient to meet the needs of multi-source data fusion, real-time diagnosis, and interpretive analysis in complex construction environments.

[0003] Currently, Chinese invention patent application number 202310121922.7 discloses an intelligent engineering information analysis system and method based on big data, including: an engineering information acquisition module, a database, an engineering information analysis module, a construction planning module, and a construction safety management module. The engineering information acquisition module collects construction environment information and construction resource information, and the database stores all the collected data. The engineering information analysis module analyzes the construction environment information and construction resource information. The construction planning module allocates construction resources and selects the optimal allocation method. The construction safety management module manages construction safety inspections. The system analyzes engineering information by establishing radar charts, which reduces the difficulty of engineering information analysis, enables appropriate resource allocation, and improves the construction safety of individual construction areas and the overall construction safety of the project.

[0004] The aforementioned technologies are insufficient for achieving unified spatiotemporal alignment and causal analysis of multi-source measurement data, and are unable to provide interpretable diagnosis and 3D visualization early warning of anomalies in dynamic construction environments. Summary of the Invention

[0005] The technical problem solved by this invention is that existing technologies are unable to achieve unified spatiotemporal alignment and causal analysis of multi-source measurement data, and cannot perform interpretable diagnosis and three-dimensional visualization early warning of anomalies in dynamic construction environments.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: The intelligent analysis method for engineering survey data includes the following steps: Step S1, Data Access and Fingerprint Identification; Step S2, Spatiotemporal alignment and dynamic benchmark construction; Step S3: Semantic feature extraction and causal graph construction; Step S4: Multimodal data fusion and intelligent diagnosis; Step S5: Anomaly tracing and causal chain generation; Step S6: Visual feedback and early warning output.

[0007] Preferably, step S1 includes the following sub-steps: Step S101: Collect point cloud measurement data, surface settlement data, surrounding rock stress data and component attitude data using a total station, UAV mapping system, GNSS base station, distributed fiber optic sensor and inclinometer, and simultaneously record the site environmental conditions, including construction progress, terrain conditions and geological conditions, and output the collected data. Step S102: Generate a source fingerprint for the collected data. The source fingerprint records the collection time, device number, sensor accuracy, operator information, and external environmental parameters. Step S103: The collected data with the source fingerprint is transmitted to a unified data pool, and standardized processing and consistency verification are performed during the transmission process.

[0008] Preferably, step S2 includes the following sub-steps: Step S201: Perform time synchronization on the collected data in the unified data pool and establish a unified time index; During the time synchronization process, the collected data at different sampling frequencies are resampled and interpolated, and the deviation is eliminated by timestamp correction, and the output is multi-source data; Step S202: Perform spatial mapping on the multi-source data to form a unified spatial framework; During the spatial mapping process, coordinate transformation and unified projection are performed on point cloud measurement data, settlement monitoring points and sensor deployment locations according to the project coordinate system, and quality control algorithms are used to identify and correct distorted data. Step S203: Establish a dynamic benchmark profile based on construction progress and terrain conditions, and update it during construction. The dynamic benchmark profile represents the geometric state of the construction section and, combined with geological conditions, forms a multi-temporal benchmark model.

[0009] Preferably, step S3 includes the following sub-steps: Step S301: Convert the point cloud measurement data into semantic features, the semantic features including tunnel cross-section radius offset and segment misalignment. During the point cloud measurement data conversion process, noise filtering, density equalization, and cross-sectional slicing calculations are performed on the point cloud measurement data. By comparing with dynamic benchmark profiles, semantic feature results are output, which include error range and confidence level. Step S302: Map the surrounding rock stress data and surface settlement data onto the temporal and spatial indexes of semantic features to form initial causal nodes, and label the node attributes, including numerical range, trend of change and associated construction section; Step S303: Introduce the construction log data into the causal graph to establish the causal relationship between construction conditions and measurement results; The construction log data includes grouting volume, tunneling speed, and construction shifts, with additional time sequence and operator markings.

[0010] Preferably, the construction log data further includes the start and end times of the operation, material batches, equipment operating status, and records of abnormal events; The construction log data, after being entered in a unified format, corresponds to the nodes in the cause-and-effect graph.

[0011] Preferably, step S4 includes the following sub-steps: Step S401: Perform multimodal fusion on semantic features, settlement time series data, stress time series data and construction log data to generate a fused feature vector; During the fusion process, the semantic features, settlement time series data, stress time series data and construction log data are time-stamp aligned and scale normalized, and feature weighting and redundancy removal are performed. Step S402: Based on the fused feature vector, perform diagnosis to identify potential abnormal patterns; During the diagnostic process, historical comparisons are used to distinguish between temporary fluctuations and persistent deviations, and a list of abnormal patterns is generated. Step S403: Compare the list of abnormal patterns with the construction log activities, and output the abnormal type and corresponding section. By combining the time, section, and operation records of the construction event, determine whether the anomaly is related to a specific construction behavior and output the judgment result.

[0012] Preferably, step S5 includes the following sub-steps: Step S501: Trace the list of abnormal patterns according to the cause-effect graph to locate the triggering factors; During the tracing process, the order of anomalies is determined by combining time series information, and the triggering source is marked. Step S502: Form an abnormal causal chain, wherein the abnormal causal chain records the time of the abnormality, the data source involved, and the construction conditions that triggered it; In the process of forming an abnormal causal chain, abnormal information from different data sources is logically linked together; Step S503: Compare the abnormal causal chain with the construction log activities, and verify the degree of matching between the abnormality and the construction behavior by combining the execution time, operators and construction sections, and output a corresponding relationship report.

[0013] Preferably, step S6 includes the following sub-steps: Step S601: Map the abnormal cause-effect chain, judgment result and corresponding relationship report to the 3D BIM model to generate a 3D diagnostic report; During the mapping process, key nodes of the causal chain are mapped to components and spatial locations in the BIM model and displayed in the form of color coding and timeline markings; Step S602: Mark potential risk areas in the diagnostic report; The marker includes a risk level, probability of occurrence, and historical evolution trend description, and displays the data source and corresponding time period related to the risk; Step S603: Generate early warning information, which includes risk type, scope of impact and recommended handling measures, and push it to the construction management terminal; The warning information is displayed through text, charts, and highlighted areas, and is pushed out in a tiered manner.

[0014] Preferably, the three-dimensional diagnostic report includes detailed data source and timestamp information; The data sources include the acquisition device number, sampling mode, operator and environmental parameter descriptions. The timestamp information records the precise acquisition time, time zone and calibration status, and is displayed in a layered format together with risk markers.

[0015] The intelligent analysis system for engineering survey data includes a data identification module, a benchmark construction module, a map construction module, a fusion diagnosis module, a source tracing module, and a feedback and early warning module. The data identification module is used for data access and fingerprint identification; The benchmark construction module is used for spatiotemporal alignment and dynamic benchmark construction; The graph construction module is used for semantic feature extraction and causal graph construction; The fusion diagnostic module is used for multimodal data fusion and intelligent diagnosis; The source tracing module is used for anomaly tracing and causal chain generation; The feedback and early warning module is used for visual feedback and early warning output.

[0016] The beneficial effects of this invention are as follows: This invention collects multi-source data and generates fingerprints by using total stations, UAVs, GNSS and fiber optic sensors, constructs dynamic benchmark profiles, extracts semantic features and combines them with construction logs to form causal graphs, which can realize multimodal fusion and anomaly diagnosis, trace trigger sources and output causal chains, and finally intuitively mark risks and push early warnings in the three-dimensional BIM model, effectively improving the integrity, real-time and interpretability of monitoring data, and enhancing construction safety and management decision support. Attached Figure Description

[0017] Figure 1 A flowchart illustrating the steps of an intelligent analysis method for engineering measurement data provided in one embodiment of the present invention; Figure 2 This is a basic flowchart of an intelligent analysis system for engineering measurement data provided in one embodiment of the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0019] Example 1, referring to Figure 1 It provides an intelligent analysis method for engineering survey data, including the following steps: Step S1, Data Access and Fingerprint Identification.

[0020] Step S2, Spatiotemporal alignment and dynamic benchmark construction.

[0021] Step S3: Semantic feature extraction and causal graph construction.

[0022] Step S4: Multimodal data fusion and intelligent diagnosis.

[0023] Step S5: Anomaly tracing and causal chain generation.

[0024] Step S6: Visual feedback and early warning output.

[0025] This invention collects multi-source data and generates fingerprints using total stations, drones, GNSS, and fiber optic sensors. It constructs dynamic benchmark profiles, extracts semantic features, and combines them with construction logs to form causal graphs. This enables multimodal fusion and anomaly diagnosis, traces trigger sources, and outputs causal chains. Finally, it visually marks risks and pushes early warnings in a 3D BIM model, effectively improving the integrity, real-time performance, and interpretability of monitoring data, and enhancing construction safety and management decision support.

[0026] Step S1 includes the following sub-steps: Step S101: Collect point cloud measurement data, surface settlement data, surrounding rock stress data, and component attitude data using a total station, UAV mapping system, GNSS base station, distributed fiber optic sensor, and inclinometer, and simultaneously record on-site environmental conditions, including construction progress, terrain conditions, and geological conditions. The output is the collected data.

[0027] Step S101 involves collecting point cloud measurement data, surface settlement data, surrounding rock stress data, and component attitude data, while simultaneously recording construction progress, terrain conditions, and geological conditions. This ensures that the acquired data comprehensively reflects the on-site construction conditions and guarantees the authenticity and coverage of the monitoring results.

[0028] Step S102: Generate source fingerprint for the collected data. The source fingerprint records the collection time, device number, sensor accuracy, operator information, and external environmental parameters.

[0029] Step S102 generates a source fingerprint for the collected data, recording the collection time, device number, sensor accuracy, operator information, and external environmental parameters to achieve unique identification and traceability of the data, providing a basis for subsequent anomaly tracing and quality verification.

[0030] Step S103: The collected data with the source fingerprint is transmitted to a unified data pool, and standardized processing and consistency verification are performed during the transmission process.

[0031] Step S103 transmits the collected data with the source fingerprint to a unified data pool and performs standardization processing and consistency verification to ensure the uniformity of the data in terms of format, timestamp and coordinate system, providing a high-quality data foundation for subsequent analysis.

[0032] Step S1 involves collecting engineering measurement data from multiple sources, generating source fingerprints, and transmitting the data to a unified data pool to ensure the integrity, traceability, and standardization of the collected data, providing reliable input for subsequent spatiotemporal alignment and multimodal fusion.

[0033] Step S2 includes the following sub-steps: Step S201: Perform time synchronization on the collected data in the unified data pool and establish a unified time index.

[0034] During the time synchronization process, the collected data at different sampling frequencies are resampled and interpolated, and the deviation is eliminated by timestamp correction, resulting in multi-source data output.

[0035] Step S201 achieves alignment and unification of data with different sampling frequencies by performing time synchronization on the collected data and establishing a unified time index. After resampling, interpolation and timestamp correction, the output multi-source data has a unified time scale, eliminating time deviations caused by device differences and transmission delays.

[0036] Step S202: Perform spatial mapping on the multi-source data to form a unified spatial framework.

[0037] During the spatial mapping process, coordinate transformation and unified projection are performed on point cloud measurement data, settlement monitoring points and sensor deployment locations based on the project coordinate system, and quality control algorithms are used to identify and correct distorted data.

[0038] Step S202 forms a unified spatial framework by performing spatial mapping on multi-source data. After coordinate transformation, unified projection and quality control algorithm processing, point cloud measurement data, settlement monitoring points and sensor positions can be accurately matched, ensuring accurate matching of multi-source data in spatial dimensions.

[0039] Step S203: Establish a dynamic benchmark profile based on construction progress and terrain conditions, and update it during construction.

[0040] The dynamic benchmark profile represents the geometric state of the construction section and, combined with geological conditions, forms a multi-temporal benchmark model.

[0041] Step S203 establishes a benchmark profile that is dynamically updated with the progress of construction and topographic conditions, reflecting the geometric state of the construction section, and combines it with geological conditions to form a multi-temporal benchmark model, providing a continuous and effective benchmark for the comparison of monitoring data and anomaly identification.

[0042] Step S2 performs unified temporal and spatial correction on the collected data in the unified data pool and constructs a dynamic benchmark profile to achieve comparability and consistency of data from different sources and frequencies under the same benchmark framework, providing a stable reference basis for subsequent semantic feature extraction and anomaly diagnosis.

[0043] Step S3 includes the following sub-steps: Step S301: Convert the point cloud measurement data into semantic features, which include the tunnel cross-section radius offset and the segment misalignment.

[0044] During the point cloud measurement data conversion process, noise filtering, density equalization, and cross-sectional slicing calculations are performed on the point cloud measurement data. By comparing with dynamic benchmark profiles, semantic feature results are output, which include error range and confidence level.

[0045] Step S301 processes the point cloud measurement data through noise filtering, density equalization, and cross-sectional slicing calculation to obtain semantic features such as tunnel cross-sectional radius offset and segment misalignment. The semantic feature results output after comparison with the dynamic benchmark profile have error range and confidence level, enabling the point cloud data to reflect the construction status in the form of structured indicators.

[0046] Step S302: Map the surrounding rock stress data and surface settlement data onto the temporal and spatial indexes of semantic features to form initial causal nodes, and label the node attributes, including numerical range, trend of change and associated construction section.

[0047] Step S302 maps the surrounding rock stress data and surface settlement data to the temporal and spatial indexes of semantic features, generates initial causal nodes, and marks attributes such as numerical range, trend of change and associated construction sections, so that the logical connection between different data sources can be clearly expressed, providing a foundation for constructing causal chains.

[0048] Step S303: Introduce the construction log data into the causal graph to establish the causal relationship between construction conditions and measurement results.

[0049] The construction log data includes grouting volume, tunneling speed, and construction shifts, with additional time sequence and operator markings.

[0050] The construction log data further includes the start and end times of the work, material batches, equipment operating status, and records of abnormal events.

[0051] After the construction log data is entered in a standardized format, it corresponds to the nodes in the cause-and-effect graph.

[0052] Step S303 combines construction log data with causal nodes to establish a causal relationship between construction conditions and measurement results. By recording information such as grouting volume, tunneling speed, construction shifts, start and end times of operations, material batches, equipment operating status, and abnormal events, and corresponding them with causal nodes, the integrity and traceability of the causal graph are enhanced.

[0053] Step S3 performs semantic processing and causal modeling on point cloud measurement data, surrounding rock stress data, surface settlement data, and construction log data to realize the logical relationship between measurement indicators and construction conditions, forming a causal graph, which provides interpretable support for subsequent anomaly diagnosis and source tracing.

[0054] Step S4 includes the following sub-steps: Step S401: Perform multimodal fusion on semantic features, settlement time series data, stress time series data and construction log data to generate a fused feature vector.

[0055] During the fusion process, the semantic features, settlement time series data, stress time series data and construction log data are time-stamp aligned and scale normalized, and feature weighting and redundancy removal are performed.

[0056] Step S401 performs multimodal fusion of semantic features, settlement time series data, stress time series data, and construction log data to generate a fused feature vector. During the fusion process, timestamp alignment, scale normalization, feature weighting, and redundancy removal are completed, so that data from different sources complement each other under a unified benchmark, resulting in stable and accurate comprehensive features, laying the foundation for subsequent diagnosis.

[0057] Step S402: Diagnosis is performed based on the fused feature vector to identify potential abnormal patterns.

[0058] During the diagnostic process, historical comparisons are used to distinguish between temporary fluctuations and persistent deviations, and a list of abnormal patterns is generated.

[0059] Step S402 performs diagnosis based on fused feature vectors, identifies potential abnormal patterns, distinguishes between temporary fluctuations and persistent deviations by comparing with historical data, and generates a list of abnormal patterns, enabling the system to accurately locate possible anomalies and form a candidate set to be verified.

[0060] Step S403: Compare the list of abnormal patterns with the construction log activities, and output the abnormal type and corresponding section.

[0061] By combining the time, section, and operation records of the construction event, determine whether the anomaly is related to a specific construction behavior and output the judgment result.

[0062] Step S403 compares the list of abnormal patterns with the construction log activities, outputs the abnormal type and corresponding section, and combines the time, section and operation record of the construction event to determine whether the abnormality is related to a specific construction behavior and outputs a clear judgment result, thereby providing a reliable basis for subsequent abnormal source tracing.

[0063] Step S4 performs multimodal fusion and intelligent diagnosis on semantic features, settlement time series data, stress time series data and construction log data, which can identify abnormal patterns and verify them in conjunction with construction activities. Finally, it outputs the abnormal type and corresponding section, realizing the comprehensive identification and interpretation of abnormalities.

[0064] Step S5 includes the following sub-steps: Step S501: Trace the list of abnormal patterns according to the cause-effect graph to locate the triggering factors.

[0065] During the tracing process, the order in which anomalies occur is determined by combining time series information, and the triggering source is marked.

[0066] Step S501 traces the list of abnormal patterns based on the cause-effect graph, determines the order of abnormal occurrence by combining time series information, marks the trigger source, and enables the system to clearly identify the leading factors and key nodes that cause the abnormality.

[0067] Step S502: Form an abnormal causal chain. The abnormal causal chain records the time of the abnormality, the data source involved, and the construction conditions that triggered it.

[0068] In the process of forming an abnormal causal chain, abnormal information from different data sources is logically linked.

[0069] Step S502 forms an abnormal causal chain, recording the time of the abnormality, the data sources involved, and the construction conditions that triggered it. In the process of forming the chain, abnormal information from different data sources is logically linked together to obtain a chain with complete causal logic, providing a structured basis for the interpretation of the abnormality.

[0070] Step S503: Compare the abnormal causal chain with the construction log activities, and verify the degree of matching between the abnormality and the construction behavior by combining the execution time, operators and construction sections, and output a corresponding relationship report.

[0071] Step S503 compares the abnormal causal chain with the construction log activities, and verifies the degree of matching between the abnormality and the construction behavior by combining the execution time, operators and construction section, and outputs a corresponding relationship report, so as to establish a clear causal relationship between the cause of the abnormality and the actual construction operation, supporting decision-making and risk prevention and control.

[0072] Step S5 involves tracing the cause and effect of the abnormal pattern list to gradually locate the trigger source and form an abnormal cause and effect chain. This chain is then compared with the construction log to obtain a report on the correspondence between the abnormality and the construction behavior, thereby achieving the interpretability and source tracing analysis of the abnormality source.

[0073] Step S6 includes the following sub-steps: Step S601: Map the abnormal causal chain, judgment result and corresponding relationship report to the three-dimensional BIM model to generate a three-dimensional diagnostic report.

[0074] During the mapping process, key nodes of the causal chain are mapped to components and spatial locations in the BIM model and displayed in the form of color coding and timeline markings.

[0075] Step S601 maps the abnormal causal chain, judgment result and corresponding relationship report to the three-dimensional BIM model to generate a three-dimensional diagnostic report. In the mapping, the correspondence between abnormalities and components and spatial locations is intuitively displayed through color coding and time axis marking, making the risk visible and structured.

[0076] Step S602: Mark potential risk areas in the diagnostic report.

[0077] The label includes the risk level, probability of occurrence, and historical trend description, and displays the data source and corresponding time period related to the risk.

[0078] Step S602 marks potential risk areas in the three-dimensional diagnostic report. The marking content includes the risk level, probability of occurrence and historical evolution trend, and is accompanied by relevant data source and time period information, so that managers can quickly locate the risk range and understand the risk evolution process.

[0079] Step S603: Generate early warning information, which includes the risk type, scope of impact, and recommended handling measures, and push it to the construction management terminal.

[0080] Warning information is displayed through text, charts, and highlighted areas, and is pushed out in a tiered manner.

[0081] The 3D diagnostic report includes detailed data sources and timestamp information.

[0082] The data sources include the acquisition device number, sampling mode, operator and environmental parameter descriptions. The timestamp information records the precise acquisition time, time zone and calibration status, and is displayed in a layered format together with risk markers.

[0083] Step S603 generates early warning information containing risk type, scope of impact and recommended disposal measures, and pushes it to the construction management terminal. The early warning information is displayed through text, charts and highlighted areas, and supports hierarchical push, so that managers at different levels can obtain the corresponding risk warnings in a timely manner.

[0084] The data source and timestamp information included in the 3D diagnostic report covers equipment number, sampling mode, operator and environmental parameters, as well as precise acquisition time, time zone and calibration status, and is displayed in layers corresponding to risk markers to ensure the traceability and verifiability of diagnostic results.

[0085] Step S6 maps the abnormal causal chain with the diagnostic results to the 3D BIM model to generate a visual diagnostic report containing risk sections and early warning information, thereby enabling intuitive display of anomalies and risk warning push, and enhancing the interpretability and management operability of engineering monitoring.

[0086] Example 2, refer to Figure 2 It provides an intelligent analysis system for engineering measurement data, including a data identification module, a benchmark construction module, a map construction module, a fusion diagnosis module, a source tracing module, and a feedback and early warning module.

[0087] The data identification module is used for data access and fingerprint identification.

[0088] The benchmark building module is used for spatiotemporal alignment and dynamic benchmark building.

[0089] The graph construction module is used for semantic feature extraction and causal graph construction.

[0090] The fusion diagnostic module is used for multimodal data fusion and intelligent diagnosis.

[0091] The source tracing module is used for anomaly tracing and causal chain generation.

[0092] The feedback and early warning module is used for visual feedback and early warning output.

[0093] The intelligent analysis method for engineering measurement data proposed in this invention, which combines multi-source data acquisition, dynamic benchmarking, semantic feature extraction, and causal graph analysis, can achieve the following beneficial effects: Point cloud measurement data, surface settlement data, surrounding rock stress data, and component attitude data are collected by combining equipment such as total stations, UAVs, GNSS, distributed fiber optic sensors, and inclinometers, and source fingerprint information is added to ensure the traceability and integrity of the data.

[0094] Establish a dynamic benchmark profile that is updated as construction progresses and terrain conditions change, to avoid distortion caused by drift of the fixed benchmark and enhance the accuracy of data comparison.

[0095] Point cloud data is transformed into semantic features after noise filtering and cross-sectional slicing. Combined with settlement, stress and construction log data, a causal graph is constructed to achieve interpretive analysis of anomalies.

[0096] By generating fused feature vectors through multimodal fusion, temporary fluctuations and persistent deviations are identified, and an abnormal pattern list is output. Based on the causal graph, the triggering source of the anomaly is traced to form a causal chain and clarify the correspondence between the anomaly and the construction behavior.

[0097] By mapping the causal chain and diagnostic results to the 3D BIM model, a visual diagnostic report with risk zone markers and a timeline is generated, which intuitively shows the distribution and evolution of anomalies; and early warning information including risk type, scope of impact and disposal measures is pushed to provide real-time and hierarchical decision support for construction management.

[0098] In summary, this invention can not only analyze engineering measurement data in real time and comprehensively, but also realize the causal explanation of anomalies and three-dimensional visualization early warning, significantly improving the scientific nature, controllability and safety of construction monitoring, which is something that existing technologies cannot achieve.

[0099] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] 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, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An intelligent analysis method for engineering survey data, characterized in that, Includes the following steps: Step S1, Data Access and Fingerprint Identification; Step S2, Spatiotemporal alignment and dynamic benchmark construction; Step S3: Semantic feature extraction and causal graph construction; Step S4: Multimodal data fusion and intelligent diagnosis; Step S5: Anomaly tracing and causal chain generation; Step S6: Visual feedback and early warning output.

2. The intelligent analysis method for engineering survey data as described in claim 1, characterized in that, Step S1 includes the following sub-steps: Step S101: Collect point cloud measurement data, surface settlement data, surrounding rock stress data and component attitude data using a total station, UAV mapping system, GNSS base station, distributed fiber optic sensor and inclinometer, and simultaneously record the site environmental conditions, including construction progress, terrain conditions and geological conditions, and output the collected data. Step S102: Generate a source fingerprint for the collected data. The source fingerprint records the collection time, device number, sensor accuracy, operator information, and external environmental parameters. Step S103: The collected data with the source fingerprint is transmitted to a unified data pool, and standardized processing and consistency verification are performed during the transmission process.

3. The intelligent analysis method for engineering survey data as described in claim 2, characterized in that, Step S2 includes the following sub-steps: Step S201: Perform time synchronization on the collected data in the unified data pool and establish a unified time index; During the time synchronization process, the collected data at different sampling frequencies are resampled and interpolated, and the deviation is eliminated by timestamp correction, and the output is multi-source data; Step S202: Perform spatial mapping on the multi-source data to form a unified spatial framework; During the spatial mapping process, coordinate transformation and unified projection are performed on point cloud measurement data, settlement monitoring points and sensor deployment locations according to the project coordinate system, and quality control algorithms are used to identify and correct distorted data. Step S203: Establish a dynamic benchmark profile based on construction progress and terrain conditions, and update it during construction. The dynamic benchmark profile represents the geometric state of the construction section and, combined with geological conditions, forms a multi-temporal benchmark model.

4. The intelligent analysis method for engineering survey data as described in claim 3, characterized in that, Step S3 includes the following sub-steps: Step S301: Convert the point cloud measurement data into semantic features, the semantic features including tunnel cross-section radius offset and segment misalignment. During the point cloud measurement data conversion process, noise filtering, density equalization, and cross-sectional slicing calculations are performed on the point cloud measurement data. By comparing with dynamic benchmark profiles, semantic feature results are output, which include error range and confidence level. Step S302: Map the surrounding rock stress data and surface settlement data onto the temporal and spatial indexes of semantic features to form initial causal nodes, and label the node attributes, including numerical range, trend of change and associated construction section; Step S303: Introduce the construction log data into the causal graph to establish the causal relationship between construction conditions and measurement results; The construction log data includes grouting volume, tunneling speed, and construction shifts, with additional time sequence and operator markings.

5. The intelligent analysis method for engineering survey data as described in claim 4, characterized in that, The construction log data further includes the start and end times of the operation, material batches, equipment operating status, and records of abnormal events. The construction log data, after being entered in a unified format, corresponds to the nodes in the cause-and-effect graph.

6. The intelligent analysis method for engineering survey data as described in claim 5, characterized in that, Step S4 includes the following sub-steps: Step S401: Perform multimodal fusion on semantic features, settlement time series data, stress time series data and construction log data to generate a fused feature vector; During the fusion process, the semantic features, settlement time series data, stress time series data and construction log data are time-stamp aligned and scale normalized, and feature weighting and redundancy removal are performed. Step S402: Based on the fused feature vector, perform diagnosis to identify potential abnormal patterns; During the diagnostic process, historical comparisons are used to distinguish between temporary fluctuations and persistent deviations, and a list of abnormal patterns is generated. Step S403: Compare the list of abnormal patterns with the construction log activities, and output the abnormal type and corresponding section. By combining the time, section, and operation records of the construction event, determine whether the anomaly is related to a specific construction behavior and output the judgment result.

7. The intelligent analysis method for engineering survey data as described in claim 6, characterized in that, Step S5 includes the following sub-steps: Step S501: Trace the list of abnormal patterns according to the cause-effect graph to locate the triggering factors; During the tracing process, the order of anomalies is determined by combining time series information, and the triggering source is marked. Step S502: Form an abnormal causal chain, wherein the abnormal causal chain records the time of the abnormality, the data source involved, and the construction conditions that triggered it; In the process of forming an abnormal causal chain, abnormal information from different data sources is logically linked together; Step S503: Compare the abnormal causal chain with the construction log activities, and verify the degree of matching between the abnormality and the construction behavior by combining the execution time, operators and construction sections, and output a corresponding relationship report.

8. The intelligent analysis method for engineering survey data as described in claim 7, characterized in that, Step S6 includes the following sub-steps: Step S601: Map the abnormal cause-effect chain, judgment result and corresponding relationship report to the 3D BIM model to generate a 3D diagnostic report; During the mapping process, key nodes of the causal chain are mapped to components and spatial locations in the BIM model and displayed in the form of color coding and timeline markings; Step S602: Mark potential risk areas in the diagnostic report; The marker includes a risk level, probability of occurrence, and historical evolution trend description, and displays the data source and corresponding time period related to the risk; Step S603: Generate early warning information, which includes risk type, scope of impact and recommended handling measures, and push it to the construction management terminal; The warning information is displayed through text, charts, and highlighted areas, and is pushed out in a tiered manner.

9. The intelligent analysis method for engineering survey data as described in claim 8, characterized in that, The three-dimensional diagnostic report includes detailed data source and timestamp information; The data sources include the acquisition device number, sampling mode, operator and environmental parameter descriptions. The timestamp information records the precise acquisition time, time zone and calibration status, and is displayed in a layered format together with risk markers.

10. An intelligent analysis system for engineering surveying data, applied in the intelligent analysis method for engineering surveying data as described in any one of claims 1-9, characterized in that, It includes a data identification module, a benchmark construction module, a map construction module, a fusion diagnosis module, a source tracing module, and a feedback and early warning module; The data identification module is used for data access and fingerprint identification; The benchmark construction module is used for spatiotemporal alignment and dynamic benchmark construction; The graph construction module is used for semantic feature extraction and causal graph construction; The fusion diagnostic module is used for multimodal data fusion and intelligent diagnosis; The source tracing module is used for anomaly tracing and causal chain generation; The feedback and early warning module is used for visual feedback and early warning output.

Images