Building foundation grouting reinforcement full life cycle dynamic risk assessment system based on multi-source sensing fusion

Abstract

The invention relates to the technical field of risk assessment, and discloses a building foundation grouting reinforcement full life cycle dynamic risk assessment system based on multi-source sensing fusion, which comprises a multi-source data acquisition module, a data fusion calibration module, an observation cycle optimization module, a risk pre-judgment and early warning module and a management module. Through multi-source sensing space correlation binding and cooperative calibration, cross validation is carried out through precision complementation of settlement, crack and inclination data, the reliability of monitoring data and the comprehensiveness of evaluation are improved, and risk misjudgment and missed judgment caused by single data are avoided; and meanwhile, the observation period is adjusted by depending on a multi-dimensional risk index, and the risk adaptation frequency is realized in combination with an equipment linkage calibration mechanism, so that the early warning real-time response speed is improved while the resource waste is reduced, and sufficient time is reserved for risk disposal.

Description

Technical Field

[0001] This invention relates to the field of risk assessment technology, and more specifically discloses a dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion. Background Technology

[0002] Risk assessment refers to the process of comprehensively evaluating and providing early warnings on the current and future safety status of a building's foundation grouting reinforcement structure by systematically collecting and analyzing multi-source monitoring data (such as settlement, cracks, and tilt) related to the stability of the foundation grouting reinforcement, combined with engineering specifications and geological conditions.

[0003] The prior art patent document with authorization announcement number CN120174914A discloses an "Automatic Monitoring and Risk Assessment System for Deformation of Building Foundation Grouting Reinforcement", which includes a preset monitoring model generation module, a historical settlement data acquisition module, a settlement data analysis module, a settlement type determination module, and a risk warning module. By determining the span fluctuation range and the complexity change range, and based on the settlement type determination result and real-time monitoring data, the system compares the settlement span to be assessed and the complexity change range to be assessed with the comparison range, and quickly generates a warning signal of the corresponding level. This avoids monitoring only the deformation of the building foundation grouting reinforcement while ignoring the changes in the surface settlement shape of the building foundation grouting reinforcement.

[0004] The patent document with authorization announcement number CN120450412A discloses "A risk assessment system and assessment method for monitoring deep building foundation grouting reinforcement based on big data", which includes: a data collection and storage module for storing building foundation grouting reinforcement experimental data obtained from building foundation grouting reinforcement experimental monitoring and building foundation grouting reinforcement simulation experimental data obtained from building foundation grouting reinforcement simulation experiments; a building foundation grouting reinforcement experimental monitoring and conversion module for monitoring real building foundation grouting reinforcement experiments, and converting and transmitting the monitored building foundation grouting reinforcement experimental data; and a building foundation grouting reinforcement experimental simulation module for simulating building foundation grouting reinforcement risk using existing data.

[0005] While existing technologies can analyze the type of surface settlement during foundation grouting reinforcement of buildings, determine relevant fluctuations and change ranges, and generate early warning signals by combining real-time data, and can also use past monitoring data and real-world experimental data to conduct simulation experiments, incorporating some surrounding environmental factors, thus improving monitoring accuracy and prediction fit to some extent, existing technologies lack collaborative sensing capabilities. The lack of a collaborative calibration mechanism in data processing leads to insufficient accuracy, and the lack of dynamic adaptive capabilities in the observation cycle results in insufficient timeliness of early warnings. Furthermore, risk assessment is based on a post-event feedback model, lacking full life-cycle prediction, and the disconnect between management and technology systems prevents the formation of closed-loop control, making it difficult to achieve accurate, efficient, and proactive prevention and control of risks throughout the entire life cycle of foundation grouting reinforcement of buildings. Summary of the Invention

[0006] The main technical problem solved by this invention is to provide a dynamic risk assessment system for the entire life cycle of grouting reinforcement of building foundations based on multi-source sensor fusion, which can solve the problems proposed in the background art.

[0007] To address the aforementioned technical problems, according to one aspect of the present invention, more specifically, a dynamic risk assessment system for the entire lifecycle of grouting reinforcement of building foundations based on multi-source sensor fusion, comprising: a multi-source data acquisition module, a data fusion calibration module, an observation cycle optimization module, a risk prediction and early warning module, and a management module; the multi-source data acquisition module spatially correlates and binds settlement, crack, and tilt monitoring data throughout the entire lifecycle of grouting reinforcement of building foundations, and performs data acquisition and post-acquisition data preprocessing; the data fusion calibration module constructs a core benchmark based on settlement monitoring data, cross-validates multi-source monitoring data, and completes data calibration through adjustment algorithms; the observation cycle optimization module calculates multi-dimensional risk indicators, dynamically adjusts the observation cycle according to the risk indicators, and establishes a linkage between the observation cycle and monitoring equipment; the risk prediction and early warning module configures weights according to different stages of the entire lifecycle of grouting reinforcement of building foundations, performs risk trend extrapolation based on calibrated data, and achieves graded early warning; the management module performs calibration management of monitoring instruments, triggers corresponding safety measures based on early warning information, and standardizes the monitoring results throughout the entire lifecycle.

[0008] Furthermore, the multi-source data acquisition module includes: a sensor deployment module, a synchronous acquisition module, and a data preprocessing module; Sensor deployment module: Φ20 threaded steel bars, stainless steel nails, and reflective sheets are used as markers for settlement, cracks, and tilt observation points, respectively, and are deployed on the exterior wall of the second basement level of the building, 0.4m above the ground and on both sides of the cracks; Synchronous acquisition module: Triggers synchronous data acquisition according to a preset observation period, and adopts a timestamp synchronization mechanism; Data preprocessing module: Cleans and standardizes the collected raw data on settlement, cracks, and tilt, removes invalid data, and marks abnormal data.

[0009] Furthermore, the data fusion calibration module includes: a settlement benchmark construction module, a data cross-validation module, and an adjustment calibration module; Settlement benchmark construction module: Using a Trimble DINI03 electronic level, a closed leveling line was measured according to the first-order settlement observation level, and the average value of the round trip observations was used to construct the core settlement benchmark; Data cross-validation module: Establishes correlation thresholds between crack width and settlement, and between tilt angle and settlement, and performs cross-validation on multi-source data; Adjustment and calibration module: Using Weiyuantu adjustment calculation software, crack observation data and tilt observation data are substituted into the settlement observation rigorous adjustment equation for collaborative adjustment and calibration.

[0010] Furthermore, the observation cycle optimization module includes: a risk indicator calculation module, an observation cycle adjustment module, and a cycle-equipment linkage module; Risk index calculation module: Calculates risk indicators such as settlement rate, crack propagation rate, and tilt rate. Settlement rate is calculated as the ratio of cumulative settlement to observation interval. Observation cycle adjustment module: Based on the ratio of risk indicators to warning values, it is divided into the initial stage, stable stage, early warning stage, and risk escalation stage. Periodic-equipment linkage module: When the observation period is adjusted, the i-angle of the monitoring instrument is synchronously linked, and the comprehensive calibration of the instrument is triggered during the risk escalation stage.

[0011] Furthermore, the risk prediction and early warning module includes: a phased weight configuration module, a trend projection module, and a graded early warning module; Phased weighting configuration module: Configure the weight ratio of settlement, crack width, and tilt angle according to the initial baseline stage, construction impact stage, and operation stability stage respectively; Trend projection module: Using a linear regression algorithm, combined with the geological characteristics of the Yongding River alluvial fan and correction of anti-buoyancy water level data, it predicts the amount of future periodic deformation; Tiered early warning module: A three-level early warning mechanism is set up, corresponding to different ratio thresholds of predicted values ​​and warning values.

[0012] Furthermore, the management module includes: an instrument calibration management module, a safety handling module, and a results management module; Instrument calibration management module: Regularly calibrate the monitoring instruments, and the calibration cycle shall be carried out in accordance with the instrument calibration specifications; Safety Response Module: After an alert is triggered, targeted safety response measures will be pushed out. Results Management Module: Standardizes and archives fused data, calibration records, early warning information, and response plans to form electronic documents and written reports.

[0013] Furthermore, the sensor deployment module uses a DINI03 electronic level for settlement observation, a Leica TM60 total station for tilt observation, and a digital display vernier caliper for crack observation.

[0014] The beneficial effects of this invention's dynamic risk assessment system for the entire lifecycle of building foundation grouting reinforcement based on multi-source sensor fusion are as follows: Through spatial correlation binding and collaborative calibration of multi-source sensors, cross-validation is performed using the complementary accuracy of settlement, crack, and tilt data, improving the reliability of monitoring data and the comprehensiveness of assessment, avoiding misjudgments and omissions caused by single data points; simultaneously, by adjusting the observation cycle based on multi-dimensional risk indicators and combining it with an equipment linkage calibration mechanism, "risk-adapted frequency" is achieved, reducing resource waste while improving early warning response speed and reserving sufficient time for risk handling; furthermore, through the trend extrapolation of phased weight configuration and geological condition correction throughout the entire lifecycle, early risk prediction is achieved, transforming passive alarms into proactive prevention and control, significantly reducing the risk of safety accidents during building foundation grouting reinforcement; finally, the integration of instrument calibration, safety handling, and results archiving ensures standardized management and provides scientific and efficient technical support for the safe operation of building foundation grouting reinforcement throughout its entire lifecycle. Attached Figure Description

[0015] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0016] Figure 1 This is a schematic diagram of the system module architecture; Figure 2 This is a schematic diagram of the monitoring points for Building 4. Figure 3 This is a schematic diagram of the monitoring points for Building 5. Figure 4 Schematic diagram of the installation of leveling benchmarks; Figure 5 This is a schematic diagram of nail embedding; Figure 6 This is a schematic diagram of the relevant thresholds. Detailed Implementation

[0017] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0018] According to one aspect of the invention, such as Figures 1-6As shown, a dynamic risk assessment system for the entire lifecycle of building foundation grouting reinforcement based on multi-source sensor fusion is provided. This system includes: a multi-source data acquisition module, which spatially correlates and binds settlement, crack, and tilt monitoring data throughout the entire lifecycle of building foundation grouting reinforcement, and performs data acquisition and post-acquisition data preprocessing. This module includes: Sensor deployment module: Φ20 threaded steel bars, stainless steel nails, and reflective sheets are used as markers for settlement, cracks, and tilt observation points, respectively, and are deployed on the exterior wall of the second basement level of the building, 0.4m above the ground and on both sides of the cracks; First, deploy settlement monitoring points: Select column foundations or structural walls located 0.4m above ground level on the exterior wall of the second basement level of the building. Use Φ20 threaded steel bars as monitoring point markers, drill holes and insert them into the wall, ensuring a firm connection between the steel bars and the wall structure. The exposed length should be controlled to approximately 3cm, and the exposed portion should be ground into an ellipsoidal shape to avoid damage from external impacts. At the same time, clearly mark the monitoring point with a number (e.g., ...). Figure 2 , Figure 3 As shown in the figure, they correspond one-to-one with the layout plan of the settlement monitoring points; Then, deploy crack observation points: For the cracks already discovered in the building walls, place a set of observation markers at the widest point and the end of the crack. Each set of markers consists of two stainless steel nails, fixed on both sides of the crack. Specifically, drill holes of appropriate depth and diameter to the stainless steel nails in the wall surface on both sides of the crack, insert approximately 6cm long stainless steel nails into the holes, ensuring about 2cm of the nails protrude, fill the holes with cement mortar and compact it (e.g., ...). Figure 5 (As shown), after the cement mortar has completely solidified, clean up the debris around the observation point to ensure that the observation field of view is unobstructed; Finally, the deployment of inclined observation points: Inclined observation points are set up at corresponding positions at the top and bottom along the same vertical line of the building. The bottom observation point and the settlement observation point are located on the outer wall of the second basement level, 0.4m above the ground. The top observation point is arranged at the corresponding position on the outer wall of the building roof. Reflective sheets are used as markers. The reflective sheets are firmly attached to the wall with construction adhesive to ensure that the surface of the reflective sheets is flat and not tilted, and that there is no obstruction to the observation line of the Leica TM60 total station, so as to achieve spatial correspondence between the top and bottom observation points.

[0019] Synchronous acquisition module: Triggers synchronous data acquisition according to a preset observation period, and adopts a timestamp synchronization mechanism; The synchronous acquisition is based on the different cycle requirements set by the observation cycle optimization module for the initial stage, stable stage, early warning stage, and risk escalation stage. The system issues a unified acquisition command to trigger the Trimble DINI03 electronic level, Leica TM60 total station, and digital display vernier caliper to start data acquisition simultaneously. For example, in the initial stage, there is an acquisition cycle of 3 days. The system issues acquisition commands at preset time points to ensure that the three data points of the same "three-point linkage sensor group" (1 settlement observation point + 1 corresponding crack observation point + 1 corresponding tilt observation point) are acquired within the same time period. The acquisition time difference is strictly controlled within 30 minutes to ensure the spatial correlation and temporal consistency of the data. The timestamp uses a unified high-precision time reference. When each acquisition device starts acquisition, the acquisition start timestamp is recorded with millisecond accuracy. After acquisition is completed, the timestamp is bound and stored with the corresponding monitoring data. For example, the timestamp error of settlement data, corresponding crack data, and corresponding tilt data does not exceed 10 milliseconds. In subsequent data preprocessing and fusion calibration, the timestamp is used as the basis for data matching to ensure accurate correspondence of multi-source data within the same monitoring period.

[0020] Data preprocessing module: Cleans and standardizes the collected raw data on settlement, cracks, and tilt, removes invalid data, and marks abnormal data; The cleaning process involves verifying each piece of raw data collected, removing data that exceeds the instrument's measurement accuracy range. For example, the measurement accuracy of a digital display vernier caliper is 0.1mm. If the raw data for crack width shows extreme values ​​below 0.05mm or above 5mm, it is considered invalid data and is removed. For settlement observation data, if the difference between the elevation value collected in a single period and the data from the adjacent period exceeds 0.5mm (exceeding the error requirement for elevation difference of a second-class settlement observation station), and there is no reasonable geological or construction disturbance cause, it is considered invalid data. At the same time, duplicate data and blank data caused by instrument failure or human error are removed to ensure the validity of the remaining data. Standardization involves converting different types of monitoring data into a unified format. Settlement data is in mm and retained to two decimal places. Downward settlement is recorded as "-" and upward rebound as "+". Crack width data is in mm and retained to one decimal place. Tilt angle data is in ° and retained to three decimal places. At the same time, a standardized data table is established according to the observation point number and the collection time stamp to ensure a unified data structure and facilitate subsequent fusion, calibration and calculation. Finally, abnormal data is marked: based on the correlation thresholds between crack width and settlement, and between tilt angle and settlement preset by the data cross-validation module (for example, under moderately compressible soil conditions, when the cumulative settlement reaches 0.001L, the crack width should be ≤0.3mm), such as... Figure 6As shown, in the analysis of standardized data, if the ratio of the crack width data to the corresponding settlement data at a certain observation point exceeds the correlation threshold, or if the conversion result of the tilt angle data and the settlement data (tilt = settlement difference / distance) exceeds the design allowable range, the data will be marked as abnormal data, and the observation point number, collection time, and other related monitoring data will be recorded to provide anomaly prompts for subsequent data fusion and calibration.

[0021] The data fusion and calibration module constructs a core benchmark based on settlement monitoring data, performs cross-validation on multi-source monitoring data, and completes data calibration through adjustment algorithms. This module includes: Settlement benchmark construction module: Using a Trimble DINI03 electronic level, a closed leveling line was measured according to the first-order settlement observation level, and the average value of the round trip observations was used to construct the core settlement benchmark; Specifically, the following steps are taken: First, several (e.g., three) stainless steel wall-mounted leveling benchmarks (numbered BM1, BM2, BM3, etc.) are established outside the area affected by interference around the project. These benchmarks are ensured to be far from roads and vibration zones, in a stable and safe environment, and capable of long-term preservation. The distance between the benchmarks and the observation points is moderate, avoiding both excessive error propagation and the influence of deformation of the observed object (e.g., Figure 4 (As shown in the figure), then according to the requirements of the first-class settlement observation level of the measurement specifications, use the Trimble DINI03 automatic leveling electronic instrument to measure the closed leveling line. During the observation, strictly control the line of sight length between 4m and 30m, the difference between foresight and backsight distance ≤1.0m, the cumulative difference between foresight and backsight distance ≤3.0m, the line of sight height ≥0.65m, the number of round trip observations for each section ≥3, and the number of round trip stations is even. When turning from forward measurement to reverse measurement, the two scales are swapped and the instrument is repositioned, as shown in Table 1 below: Table 1 Data cross-validation module: Establishes correlation thresholds between crack width and settlement, and between tilt angle and settlement, and performs cross-validation on multi-source data; Specifically, the process involves two steps: First, based on clearly defined geological characteristics and the permissible deformation standards for foundation grouting reinforcement in building structural design codes, a reasonable range for crack widths corresponding to different settlement amounts is determined. Simultaneously, based on parameters such as building height and wall structural stiffness, the correlation between tilt angle and settlement is clarified, thus constructing two sets of correlation threshold systems. Second, relying on the spatial association binding of multi-source data acquisition modules, the preprocessed settlement data, crack width data, and tilt angle data are precisely matched according to observation point numbers and timestamps, ensuring that each set of verification data comes from the same monitoring period within the same observation area. Simultaneously, threshold comparisons are performed on each set of matched data. If the settlement at a certain observation point is within a specific range, and the corresponding crack width is within the reasonable threshold range corresponding to that settlement, and the deviation between the settlement difference after tilt angle conversion and the actual settlement data is within the allowable range, then the set of multi-source data is deemed to have passed verification and is considered valid data. If the crack width exceeds the threshold range corresponding to the settlement, or the deviation between the tilt angle conversion result and the settlement data is too large, then the set of data is deemed to have an anomaly and is marked as data to be calibrated. This provides a clear indication of the anomaly data for the subsequent adjustment and calibration module, ensuring the consistency and reliability of the multi-source data.

[0022] Adjustment and calibration module: Using Weiyuantu adjustment calculation software, crack observation data and tilt observation data are substituted into the settlement observation rigorous adjustment equation to perform collaborative adjustment and calibration; First, after cross-validation, the data marked as normal and the data marked as abnormal are imported into the Weiyuantu adjustment calculation software. The settlement observation data are used to establish a rigorous adjustment equation according to the first-order settlement observation adjustment standard. The core settlement benchmark data that has been constructed is input (such as the assumed elevation of BM1, BM2, and BM3 and the average elevation difference of forward and backward observations). The target for controlling the station elevation difference error is set to <0.15mm. Then, the crack observation data is converted into "relative elevation correlation data" corresponding to the settlement observation (based on the spatial binding relationship between crack observation points and settlement observation points, the corresponding small elevation changes are inferred by back-calculating the crack width measured by vernier calipers). The tilt observation data is converted into elevation change data by three-dimensional coordinates collected by Leica TM60 total station. All of these are substituted into the rigorous adjustment equation of settlement observation to construct a multi-source data collaborative adjustment model. Finally, the adjustment software was run for iterative calculations. The high precision of the settlement data was used to correct the systematic errors of the crack and tilt data. For the abnormal data marked by cross-validation, invalid data was further eliminated through residual analysis in the adjustment equation, and the valid data after calibration was retained. Finally, the fused calibration data of each observation point after adjustment was output to ensure that the mean error of the height difference of the observation point station is <0.50mm and the consistency of the multi-source data meets the design requirements.

[0023] The observation cycle optimization module calculates multi-dimensional risk indicators, dynamically adjusts the observation cycle based on these indicators, and establishes a linkage between the observation cycle and monitoring equipment. This module includes: Risk index calculation module: Calculates risk indicators such as settlement rate, crack propagation rate, and tilt rate. Settlement rate is calculated as the ratio of cumulative settlement to observation interval. First, select the valid data output by the data fusion calibration module after collaborative adjustment to ensure that the basic data used to calculate risk indicators are free of systematic errors and meet the consistency standards, so as to avoid the original data or abnormal data from affecting the accuracy of the indicators. Then, for each settlement observation point, the cumulative settlement from the initial baseline stage to the current monitoring cycle is extracted (absolute value, unit mm). At the same time, the time interval between the current cycle and the previous effective collection cycle is determined (unit day). The settlement rate is calculated according to the formula "settlement rate = cumulative settlement / time interval". For example, if the cumulative settlement of a certain observation point is 3 mm and the observation interval is 3 days, then the settlement rate is 1 mm / day. Finally, the crack propagation rate and tilt rate are calculated: the crack propagation rate is the difference between the crack width after calibration in the current cycle and the crack width after calibration in the previous cycle, divided by the observation interval. For example, if the crack width in the previous cycle was 0.2 mm, and in the current cycle it is 0.3 mm, with an observation interval of 3 days, then it is as follows: Crack propagation rate = (0.3mm - 0.2mm) ÷ 3 days ≈ 0.03mm / day Rounded to two decimal places, the tilt rate is the difference between the tilt angle after calibration in the current cycle and the tilt angle after calibration in the previous cycle, divided by the observation interval. For example, if the tilt angle in the previous cycle was 0.012°, the current cycle is 0.015°, and the observation interval is 3 days, then it would be as follows: Inclination rate = (0.015° - 0.012°) ÷ 3 days = 0.001° / day Three decimal places are retained, and all calculations are based on spatially correlated data from the same observation point to ensure the correspondence of indicators.

[0024] Observation cycle adjustment module: Based on the ratio of risk indicators to warning values, it is divided into the initial stage, stable stage, early warning stage, and risk escalation stage. Specifically, preset risk indicator warning values ​​are used (e.g., settlement rate warning value is ±1.00mm / day, crack propagation rate warning value is 0.10mm / day, tilt rate warning value is 0.005° / day). The stage is determined by comparing the ratio of multi-dimensional risk indicators with the corresponding warning values: for example, the initial stage is the first 3 monitoring cycles after the system starts up, with no historical risk indicator comparison basis, and the fixed observation cycle is 3 days; the stable stage is when any risk indicator is ≤50% of the corresponding warning value, and the observation cycle is adjusted to 7 days; the early warning stage is when 50% < any risk indicator ≤80% of the corresponding warning value, and the observation cycle is adjusted to 2 days; the risk escalation stage is when any risk indicator >80% of the corresponding warning value, and the observation cycle is adjusted to 1 day. In addition, a phase switching judgment rule is established: if all risk indicators are within the stable phase threshold range for multiple consecutive monitoring periods, the system switches from the initial phase / early warning bud phase to the stable phase; if a risk indicator exceeds the stable phase threshold but does not reach the risk escalation phase threshold in one monitoring period, the system switches from the stable phase / initial phase to the early warning bud phase; if a risk indicator exceeds the early warning bud phase threshold in any monitoring period, the system immediately switches to the risk escalation phase, ensuring that the observation period dynamically adapts to the risk level.

[0025] Periodic-equipment linkage module: When the observation period is adjusted, the i-angle of the monitoring instrument is synchronously linked, and a comprehensive instrument calibration is triggered during the risk escalation stage; The i-angle synchronous linkage calibration is as follows: When switching observation cycles, the system sends i-angle calibration commands to the Trimble DINI03 electronic level and the Leica TM60 total station. After receiving the command, the instruments start the self-test calibration process: The level detects the i-angle deviation through its built-in compensator. If the deviation exceeds the ±15" threshold, electronic compensation calibration is performed. After calibration, the i-angle data (≤±15") is uploaded to the system. The total station detects the i-angle deviation through its dual-axis compensator. If the deviation exceeds the ±15" threshold, laser alignment calibration is started. After calibration, the i-angle data (≤±15") is archived synchronously. If the calibration fails, the system issues an instrument fault alarm and suspends data acquisition until manual intervention is required for repair.

[0026] The risk prediction and early warning module assigns weights according to different stages of the entire life cycle of building foundation grouting reinforcement, and performs risk trend extrapolation based on calibrated data to achieve tiered early warning. This module includes: Phased weighting configuration module: Configure the weight ratio of settlement, crack width, and tilt angle according to the initial baseline stage, construction impact stage, and operation stability stage respectively; First, the system identifies the current life cycle stage of the building foundation grouting reinforcement. This is determined by associating the monitoring point deployment time, the status of surrounding construction activities, and the stability characteristics of the data output from the data fusion calibration module. For example, when the monitoring points have just been deployed, it is determined to be in the initial baseline stage. At this time, the focus is on calibrating the baseline through settlement data, configuring settlement as the dominant weight and crack width and tilt angle as auxiliary weights. When construction disturbance-related characteristics are detected in the surrounding area, the system determines that it has entered the construction impact stage, switching to tilt angle as the dominant weight and settlement and crack width as auxiliary weights, focusing on the overall instability risk prevention and control. When construction activities stop and the fused data continues to be in a stable fluctuation state, the system determines that it has entered the operational stability stage, adjusting the weight configuration to be equal to that of settlement and crack width, with tilt angle as an auxiliary weight, focusing on the risk of local deformation.

[0027] Trend projection module: Using a linear regression algorithm, combined with the geological characteristics of the Yongding River alluvial fan and correction of anti-buoyancy water level data, it predicts the amount of future periodic deformation; First, continuous multi-cycle fused settlement, crack width, and tilt angle data are extracted from the output data to form a complete three-dimensional risk indicator historical sequence, ensuring the continuity and reliability of the data used for simulation. Then, the system calls the built-in linear regression algorithm to fit the historical data sequence to the trend, build a basic trend model, and intuitively present the changing pattern of deformation data. Next, the basic trend model is corrected by combining the preset stratigraphic characteristics of the Yongding River alluvial fan and anti-buoyancy water level data to eliminate the interference of geological and hydrological factors on the deformation trend and improve the accuracy of prediction. During the correction process, the system will associate the correspondence between stratigraphic characteristics and deformation in the historical data to ensure that the correction logic fits the actual engineering geological conditions. Finally, based on the revised trend model, the system calculates and outputs predicted values ​​of settlement, crack width, and tilt angle for multiple future monitoring periods. The prediction results are transmitted to the graded early warning module in real time, providing core data support for early warning judgment. At the same time, the data correlation and correction basis in the prediction process are recorded and archived to ensure that the simulation process is traceable.

[0028] Tiered early warning module: Sets up a three-level early warning mechanism, corresponding to different ratio thresholds of predicted values ​​and warning values; First, the system presets the corresponding warning value parameters and clarifies the triggering criteria for the three-level warning. Then the system receives the measured data output by the data fusion calibration module and the predicted data output by the trend inference module in real time. It compares the two types of data with the preset warning value and the three-level warning judgment standard respectively, and triggers the corresponding level of warning according to the data compliance status. Finally, upon triggering different warning levels, the system executes the corresponding response process: For example, when a Level 1 warning is triggered, a routine report is generated, along with trend analysis content, clearly explaining the predicted direction of deformation and potential risk points. This report is then pushed to relevant management departments through pre-set channels. When a Level 2 warning is triggered, the system prioritizes sending warning information to project leaders, safety management personnel, and other relevant parties, followed by a complete monitoring results report. Simultaneously, based on the core risk indicators triggered by the warning, targeted on-site safety inspection priorities are extracted from pre-set safety management measures and pushed to the terminals of on-site inspection personnel. When a Level 3 warning is triggered, the system immediately initiates the highest-level response, notifying all relevant responsible parties as soon as possible, completing the preparation and submission of a special report, and sharing monitoring data, warning information, and related technical materials through the system's built-in collaborative platform, providing support for multiple parties to quickly formulate response measures.

[0029] The management module manages the calibration of monitoring instruments, triggers corresponding safety measures based on early warning information, and standardizes the management of monitoring results throughout their entire lifecycle. This module includes: Instrument calibration management module: Regularly calibrate the monitoring instruments, and the calibration cycle shall be carried out in accordance with the instrument calibration specifications; Specifically, based on the equipment type—Tempron DINI03 electronic level, Leica TM60 total station, and digital display vernier caliper—corresponding periodic calibration cycles are matched, calibration plans are generated, and calibration reminders are sent to equipment management personnel in advance to ensure that calibration work is carried out on time. During calibration, the instruments complete self-checks and corrections according to the built-in calibration process, and the calibration results are uploaded to the system in real time. After being reviewed and confirmed by the technical supervisor, they serve as a core prerequisite for the validity of subsequent data acquisition. Instruments that fail to complete calibration as required or whose calibration is unqualified will have their data acquisition function restricted by the system, and the acquired data will not be able to enter the multi-source data fusion calibration process.

[0030] Safety Response Module: After an alert is triggered, targeted safety response measures will be pushed out. Specifically, when the graded early warning module triggers different levels of early warning, the system identifies the core risk indicators corresponding to the early warning (such as excessive settlement rate, abnormal expansion of crack width, or excessive tilt angle), and accurately matches targeted treatment plans from the measures library: for example, for an abnormal settlement warning, measures such as "prohibiting water accumulation around the foundation and avoiding local load concentration" are pushed; for an abnormal crack warning, measures such as "prohibiting heavy operations around the observation point and strictly prohibiting collisions with the walls of the observation area" are pushed; for an abnormal tilt warning, measures such as "restricting construction disturbance around the building and strengthening the inspection of the overall structural stability" are pushed. The measures are pushed simultaneously through multiple channels such as the system platform, on-site inspection terminals, and management personnel's mobile APP, while recording the time of receipt of the measures and the implementation feedback, forming a closed-loop management of "early warning-push-implementation-feedback" to ensure that safety measures are implemented effectively.

[0031] Results Management Module: Standardizes and archives fused data, calibration records, early warning information, and response plans, forming electronic documents and written reports; Specifically, the fused data output from the multi-source data fusion calibration module is stored according to observation point number and monitoring cycle; calibration instructions, calibration results, and audit records from the instrument calibration management module are archived according to equipment type; the warning level, trigger time, and data basis of the risk prediction and early warning module are recorded; and the measures, implementation status, and effect feedback of the safety handling module are simultaneously retained. The archived data is organized into three forms: text description, charts, and electronic documents. The text description clearly explains the data source, processing flow, and risk conclusions; the charts intuitively present the deformation trend, early warning nodes, and handling process; and the electronic documents are stored according to a unified naming standard. This not only meets the needs of project process management but also provides data support for subsequent similar projects.

[0032] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention are also within the protection scope of the present invention.

Claims

1. A dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion, characterized in that, include: Multi-source data acquisition module, data fusion and calibration module, observation cycle optimization module, risk prediction and early warning module, management module; The multi-source data acquisition module spatially correlates and binds settlement, crack, and tilt monitoring data throughout the entire life cycle of building foundation grouting reinforcement, and performs data acquisition and post-acquisition data preprocessing; the data fusion and calibration module constructs a core benchmark based on settlement monitoring data, cross-validates multi-source monitoring data, and completes data calibration through adjustment algorithms. The observation cycle optimization module calculates multi-dimensional risk indicators, dynamically adjusts the observation cycle based on the risk indicators, and establishes a linkage between the observation cycle and the monitoring equipment. The risk prediction and early warning module configures weights according to different stages of the entire life cycle of building foundation grouting reinforcement, performs risk trend extrapolation based on calibrated data, and realizes graded early warning. The management module performs calibration management of monitoring instruments, triggers corresponding safety measures based on early warning information, and standardizes the management of monitoring results throughout the entire life cycle.

2. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 1, characterized in that: The multi-source data acquisition module includes: a sensor deployment module, a synchronous acquisition module, and a data preprocessing module; Sensor deployment module: Φ20 threaded steel bars, stainless steel nails, and reflective sheets are used as markers for settlement, cracks, and tilt observation points, respectively, and are deployed on the exterior wall of the second basement level of the building, 0.4m above the ground and on both sides of the cracks; Synchronous acquisition module: Triggers synchronous data acquisition according to a preset observation period, and adopts a timestamp synchronization mechanism; Data preprocessing module: Cleans and standardizes the collected raw data on settlement, cracks, and tilt, removes invalid data, and marks abnormal data.

3. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 1, characterized in that: The data fusion calibration module includes: a settlement benchmark construction module, a data cross-validation module, and an adjustment calibration module; Settlement benchmark construction module: Using a Trimble DINI03 electronic level, a closed leveling line was measured according to the first-order settlement observation level, and the average value of the round trip observations was used to construct the core settlement benchmark; Data cross-validation module: Establishes correlation thresholds between crack width and settlement, and between tilt angle and settlement, and performs cross-validation on multi-source data; Adjustment and calibration module: Using Weiyuantu adjustment calculation software, crack observation data and tilt observation data are substituted into the settlement observation rigorous adjustment equation for collaborative adjustment and calibration.

4. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 1, characterized in that: The observation cycle optimization module includes: a risk indicator calculation module, an observation cycle adjustment module, and a cycle-equipment linkage module; Risk index calculation module: Calculates risk indicators such as settlement rate, crack propagation rate, and tilt rate. Settlement rate is calculated as the ratio of cumulative settlement to observation interval. Observation cycle adjustment module: Based on the ratio of risk indicators to warning values, it is divided into the initial stage, stable stage, early warning stage, and risk escalation stage. Periodic-equipment linkage module: When the observation period is adjusted, the i-angle of the monitoring instrument is synchronously linked, and the comprehensive calibration of the instrument is triggered during the risk escalation stage.

5. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 1, characterized in that: The risk prediction and early warning module includes: a phased weight configuration module, a trend projection module, and a graded early warning module; Phased weighting configuration module: Configure the weight ratio of settlement, crack width, and tilt angle according to the initial baseline stage, construction impact stage, and operation stability stage respectively; Trend projection module: Using a linear regression algorithm, combined with the geological characteristics of the Yongding River alluvial fan and correction of anti-buoyancy water level data, it predicts the amount of future periodic deformation; Tiered early warning module: A three-level early warning mechanism is set up, corresponding to different ratio thresholds of predicted values ​​and warning values.

6. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 1, characterized in that: The management module includes: an instrument calibration management module, a safety handling module, and a results management module; Instrument calibration management module: Regularly calibrate the monitoring instruments, and the calibration cycle shall be carried out in accordance with the instrument calibration specifications; Safety Response Module: After an alert is triggered, targeted safety response measures will be pushed out. Results Management Module: Standardizes and archives fused data, calibration records, early warning information, and response plans to form electronic documents and written reports.

7. The dynamic risk assessment system for the entire life cycle of building foundation grouting reinforcement based on multi-source sensor fusion as described in claim 2, characterized in that: The sensor deployment module uses a DINI03 electronic level for settlement observation, a Leica TM60 total station for tilt observation, and a digital display vernier caliper for crack observation.

Images