A method for generating household acceptance data based on intelligent measuring equipment

Abstract

The application discloses a kind of based on smart measuring equipment's acceptance data generation method of division, it is related to building acceptance technical field, to the fusion analysis of multi-source building data, building data is divided into normal state data and abnormal state data, extract the key frame and non-key frame of normal state data, non-key frame is compressed storage to dimension reduction, potential uneven area is layered coding, final output is uploaded to acceptance platform and early warning system, the application realizes the fusion analysis and abnormal detection of multi-source building data by smart measuring equipment, improves acceptance efficiency and accuracy.Meanwhile, based on historical defect database training defect growth prediction model, combined with causal diagram analysis, realize the leap from experience acceptance to scientific prevention, defect growth simulation as dynamic verification tool, causal diagram as scientific guidance framework, provide scientific basis for long-term monitoring and prevention of building quality, promote the sustainable development of building industry.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of building acceptance, and in particular to a household acceptance data generation method based on intelligent measurement equipment. BACKGROUND

[0002] With the rapid development of the construction industry, the requirements for construction engineering quality are becoming increasingly stringent. Traditional household acceptance methods mainly rely on manual detection and simple measurement tools, which not only consume time and effort, but also make it difficult to comprehensively and accurately identify potential quality problems. In order to improve acceptance efficiency and accuracy, intelligent measurement equipment has been gradually introduced into the construction engineering acceptance process. These devices can collect real-time three-dimensional point cloud data, environmental temperature, humidity, vibration frequency, and sound wave reflection of the building interior, providing the possibility for comprehensive assessment of building quality.

[0003] At present, the Chinese patent with application number CN202211479094.6 discloses a building material detection and acceptance method, which includes the following steps: S1, establishing a building material information library; S2, when the building materials are put into the warehouse, the delivery detection report of the building materials that meet the warehouse entry conditions is judged for eligibility; S3, after the building materials are put into the warehouse, the inspection period of the building materials is determined according to the first inspection quality level of the building materials; S4, periodically detecting and recording the environmental data inside and outside the warehouse; S5, detecting the storage of individual building materials; S6, when the building materials are taken out of the warehouse, the best use time of the building materials is predicted according to the change amount of the relevant characteristics in the inspection information.

[0004] The related art cannot simulate defect growth based on multi-data fusion and establish a causal diagram, generate a risk warning report in advance, and automatically identify the root cause of quality problems, affecting the acceptance effect. SUMMARY

[0005] The technical problem solved by the present application is that the prior art cannot simulate defect growth based on multi-data fusion and establish a causal diagram, generate a risk warning report in advance, and automatically identify the root cause of quality problems, affecting the acceptance effect.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] A household acceptance data generation method based on intelligent measurement equipment, comprising the following steps:

[0008] Step S1, acquiring multi-source building data, performing fusion analysis on the multi-source building data, and outputting a fusion detection result;

[0009] Step S2, dividing the building data into normal state data and abnormal state data, and analyzing the normal state data;

[0010] Step S3: Extract key frames and non-key frames from the normal state data, and perform dimensionality reduction and compression on the non-key frames for storage.

[0011] Step S4: Trigger high-precision scanning and encoding optimization for abnormal state data, and perform layered encoding on potentially uneven areas;

[0012] Step S5: Upload the final output to the acceptance platform and early warning system;

[0013] Step S1 includes the following sub-steps:

[0014] Step S101, Obtain multi-source building data:

[0015] A laser scanning robot acquires three-dimensional point cloud data of the entire house, including ambient temperature, humidity, vibration frequency, and sound wave reflection data. The mathematical expression for the three-dimensional point cloud data of the entire house is as follows:

[0016] ;

[0017] in, For point cloud data, Let the coordinates of the point cloud be in three-dimensional space. For any coordinate, Let be the intensity of the laser reflection at the coordinates. This represents the total number of points in a single scan. The reflection intensity coefficient;

[0018] If the vibration frequency exceeds the preset vibration frequency threshold If this occurs, the crack detection mode is triggered, and a set of potential defect coordinates is output.

[0019] Step S102: Determine anomalies in the sensor data:

[0020] If the ambient temperature data does not meet the preset ambient temperature threshold or the ambient humidity data does not meet the preset ambient humidity threshold, it is marked as an environmental anomaly.

[0021] If the sound wave reflection data is less than the preset sound wave reflection threshold, it is determined to be a hollow defect;

[0022] The output fusion detection results include a set of potential defect coordinates, ambient temperature data, ambient humidity data, and acoustic wave reflection data.

[0023] Step S2 includes the following sub-steps:

[0024] Step S201: Divide the data according to the fusion detection results:

[0025] Extract the point cloud data corresponding to the timestamp of the environmental anomaly and output it as the abnormal state data; output the remaining data as the normal state data.

[0026] Step S202: Perform dynamic behavior analysis on the normal state data:

[0027] The mathematical expression for calculating wall flatness based on curvature differences is:

[0028] ;

[0029] in, Due to differences in wall curvature, The area to be detected;

[0030] If the wall surface flatness is greater than the preset wall surface flatness threshold, it is marked as a potential uneven area; if the wall surface flatness is less than the preset wall surface flatness threshold, it is marked as weak motion data.

[0031] Step S4 includes the following sub-steps:

[0032] Step S401: For abnormal data, trigger a high-resolution rescan.

[0033] The laser scanning accuracy is improved from the first accuracy to the second accuracy, and the scanning frequency is improved from the first frequency to the second frequency;

[0034] Step S402: Perform layered coding on the potential uneven areas:

[0035] Bitrate allocation based on preset defect severity weights:

[0036] ;

[0037] in, For layered coding bitrate, base bitrate, This is the bit rate gain coefficient. This represents the severity weight of the defect.

[0038] Preferably, step S3 includes the following sub-steps:

[0039] Step S301: Extract keyframes from the weak motion data;

[0040] The difference between point clouds in adjacent frames is calculated using the Hausdorff distance:

[0041] ;

[0042] in, For any point cloud collection, To and Adjacent point cloud sets, The distance to Hausdorff. For any point, For any other point, For point and points Euclidean distance, For the upper bound, The infimum;

[0043] like A frame is marked as a keyframe; otherwise, it is a non-keyframe. The preset keyframe determination threshold;

[0044] Step S302: Perform dimensionality reduction and compression on non-key frames:

[0045] Principal component analysis was used to reduce the point cloud dimensions from Down to Output the compressed data and store it.

[0046] Preferably, step S5 includes the following sub-steps:

[0047] Step S501: If the abnormal state data contains abrupt values ​​of structural cracks or curvature of irregular hollow areas, it is marked as a high-risk event and a real-time alarm is triggered.

[0048] The remaining data is uploaded to the acceptance platform to generate acceptance data, which includes whole-house 3D point cloud data, abnormal status data, and potential uneven areas.

[0049] Step S502: Compare the acceptance data with the BIM model and calculate the error matrix. :

[0050] ;

[0051] in, For any coordinate point, Let be the error matrix of the coordinate points, if It automatically triggers design correction suggestions and updates the construction standard library. This is the preset maximum total error.

[0052] Preferably, step S5 further includes:

[0053] Step S503: Write the hash value of the key data into the distributed ledger. The key data includes three-dimensional point cloud data, ambient temperature data, ambient humidity data, structural crack mutation value, and curvature of regular hollow areas.

[0054] Preferably, the method further includes step S6, generating adversarial examples based on a historical defect database and training a defect growth prediction model;

[0055] Constructing a cause-effect graph ,node Indicates process parameters, edge Indicates the weight of causal relationships.

[0056] Preferably, the mathematical expression for step S6 is:

[0057] Defect growth prediction model:

[0058] ;

[0059] in, For the defect area over time Changes, The material degradation coefficient, The stress distribution gradient, Input data;

[0060] Preferably, the final output includes key data and a defect growth prediction model.

[0061] The beneficial effects of this invention are as follows: This invention achieves fusion analysis and anomaly detection of multi-source building data through intelligent measurement equipment, improving the efficiency and accuracy of acceptance testing. Simultaneously, by training a defect growth prediction model based on a historical defect database and combining it with causal graph analysis, it achieves a leap from experience-based acceptance to scientific prevention. Defect growth simulation serves as a dynamic verification tool, and causal graphs serve as a scientific guiding framework, together providing a scientific basis for long-term monitoring and prevention of building quality, and promoting the sustainable development of the construction industry. Attached Figure Description

[0062] Figure 1 The flowchart illustrates the steps of a method for generating household acceptance data based on intelligent measurement equipment, as provided in one embodiment of the present invention. Detailed Implementation

[0063] 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.

[0064] Example, refer to Figure 1 This paper provides a method for generating household acceptance data based on intelligent measurement equipment, including the following steps:

[0065] Step S1: Obtain multi-source building data, perform fusion analysis on the multi-source building data, and output the fusion detection results.

[0066] Step S2: Divide the building data into normal state data and abnormal state data, and analyze the normal state data.

[0067] Step S3: Extract key frames and non-key frames from the normal state data, and perform dimensionality reduction and compression storage on the non-key frames.

[0068] Step S4: Trigger high-precision scanning and encoding optimization for abnormal state data, and perform layered encoding on potentially uneven areas.

[0069] Step S5: Upload the final output to the acceptance platform and early warning system.

[0070] Step S1 includes the following sub-steps:

[0071] Step S101, Obtain multi-source building data:

[0072] A laser scanning robot acquires three-dimensional point cloud data of the entire house, including ambient temperature, humidity, vibration frequency, and sound wave reflection data. The mathematical expression for the three-dimensional point cloud data of the entire house is as follows:

[0073] ;

[0074] in, For point cloud data, Let the coordinates of the point cloud be in three-dimensional space. For any coordinate, Let be the intensity of the laser reflection at the coordinates. This represents the total number of points in a single scan. This is the reflection intensity coefficient.

[0075] If the vibration frequency exceeds the preset vibration frequency threshold If this is triggered, the crack detection mode will be activated, and a set of potential defect coordinates will be output.

[0076] Step S101 efficiently collects three-dimensional point cloud data of the entire house using a laser scanning robot, while simultaneously monitoring ambient temperature, humidity, vibration frequency, and sound wave reflection, ensuring the comprehensiveness and accuracy of the data. In particular, the monitoring of vibration frequency enables the timely detection and location of potential cracks and defects, providing crucial clues for subsequent detailed inspection.

[0077] Step S102: Determine anomalies in the sensor data:

[0078] If the ambient temperature data does not meet the preset ambient temperature threshold or the ambient humidity data does not meet the preset ambient humidity threshold, it is marked as an environmental anomaly.

[0079] If the sound wave reflection data is less than the preset sound wave reflection threshold, it is determined to be a hollow defect.

[0080] The output fusion detection results include a set of potential defect coordinates, ambient temperature data, ambient humidity data, and acoustic wave reflection data.

[0081] Step S102 involves a detailed analysis of the sensor data, which accurately determines whether environmental parameters exceed normal ranges and whether defects such as voids exist. The output of this step, namely the fused detection result, not only includes the specific location of potential defects but also covers key environmental parameter information, providing an important basis for subsequent data fusion and analysis.

[0082] Step S1 utilizes a laser scanning robot and multiple sensors to comprehensively acquire three-dimensional structural information and environmental parameters within the building, enabling a preliminary diagnosis of the building's condition. It can not only identify potential defect locations but also determine whether environmental factors are abnormal, providing a solid foundation for subsequent data analysis and processing.

[0083] Step S2 includes the following sub-steps:

[0084] Step S201: Divide the data according to the fusion detection results:

[0085] Extract the point cloud data corresponding to the timestamp of the environmental anomaly and output it as the abnormal state data; output the remaining data as the normal state data.

[0086] Step S201 accurately extracts the point cloud data corresponding to the abnormal timestamps based on the fusion detection results and marks them as abnormal state data, while treating other data as normal state data. This step achieves accurate data classification, providing a clear data foundation for subsequent analysis and processing.

[0087] Step S202: Perform dynamic behavior analysis on the normal state data:

[0088] The mathematical expression for calculating wall flatness based on curvature differences is:

[0089] ;

[0090] in, Due to differences in wall curvature, The area to be detected.

[0091] If the wall surface flatness is greater than the preset wall surface flatness threshold, it is marked as a potential uneven area; if the wall surface flatness is less than the preset wall surface flatness threshold, it is marked as weak motion data.

[0092] Step S202 involves dynamic behavior analysis of normal state data, particularly calculating wall flatness based on curvature differences, which accurately identifies potential uneven areas on the wall. Simultaneously, data where wall flatness is within a preset threshold is marked as weak motion data. This helps distinguish between normal minor deformations and potential defects, improving the accuracy and reliability of the data analysis.

[0093] Step S2 effectively identified abnormal conditions and potential uneven areas in the building by meticulously dividing the fused detection results and conducting in-depth dynamic behavior analysis on the normal state data. This step not only improved the targeting and efficiency of data processing but also provided important reference for subsequent detailed inspection and repair work.

[0094] Step S3 includes the following sub-steps:

[0095] Step S301: Extract keyframes from the weak motion data.

[0096] The difference between point clouds in adjacent frames is calculated using the Hausdorff distance:

[0097] ;

[0098] in, For any point cloud collection, To and Adjacent point cloud sets, The distance to Hausdorff. For any point, For any other point, For point and points Euclidean distance, For the upper bound, The infimum;

[0099] like A frame is marked as a keyframe; otherwise, it is a non-keyframe. The preset keyframe determination threshold.

[0100] Step S301 accurately identifies key frames by calculating the difference in point clouds between adjacent frames and using Hausdorff distance as the metric. These key frames contain important information about changes in building status and are crucial for subsequent analysis and monitoring. Meanwhile, marking frames that do not meet the key frame criteria as non-key frames helps reduce the storage and processing of redundant data.

[0101] Step S302: Perform dimensionality reduction and compression on non-key frames:

[0102] Principal component analysis was used to reduce the point cloud dimensions from Down to Output the compressed data and store it.

[0103] Step S302 uses principal component analysis to reduce the dimensionality of non-key frames, which significantly reduces data dimensionality and storage requirements while preserving key information. This step not only improves data processing efficiency but also helps reduce data loss during transmission and storage, ensuring data integrity and accuracy. The dimensionality-reduced data is easier to store and manage, facilitating subsequent data analysis and applications.

[0104] Step S3 effectively reduces the cost of data storage and processing by extracting key frames and compressing non-key frames from weak motion data, while retaining key information, providing an efficient and accurate data foundation for subsequent data analysis and applications.

[0105] Step S4 includes the following sub-steps:

[0106] Step S401: For abnormal data, trigger a high-resolution rescan.

[0107] The laser scanning accuracy is improved from the first accuracy to the second accuracy, and the scanning frequency is improved from the first frequency to the second frequency.

[0108] Step S401 addresses abnormal state data by increasing the precision and frequency of laser scanning, enabling more detailed capture of minute changes on the building surface and thus more accurately identifying potential defects. This step provides a more precise data foundation for subsequent layered coding, contributing to improved accuracy and reliability of defect detection.

[0109] Step S402: Perform layered coding on the potential uneven areas:

[0110] Bitrate allocation based on preset defect severity weights:

[0111] ;

[0112] in, For layered coding bitrate, base bitrate, This is the bit rate gain coefficient. This represents the severity weight of the defect.

[0113] Step S402 involves hierarchical coding of potential uneven areas, allocating the code rate based on a preset defect severity weight. This coding method not only effectively records the specific location and shape of defects but also categorizes them according to their severity, providing clearer and more targeted information for subsequent analysis and repair work. The implementation of this step helps improve the efficiency and accuracy of defect management.

[0114] Step S4, by triggering a high-resolution rescan on abnormal data and performing layered coding on potentially uneven areas, achieves accurate identification and detailed recording of building defects. This step not only improves the accuracy of defect detection but also provides comprehensive information support for subsequent repair and maintenance work.

[0115] Step S5 includes the following sub-steps:

[0116] Step S501: If the abnormal status data contains abrupt changes in structural crack values ​​or irregular hollow area curvature, it is marked as a high-risk event and a real-time alarm is triggered.

[0117] The remaining data is uploaded to the acceptance platform to generate acceptance data, which includes whole-house 3D point cloud data, abnormal status data, and potential uneven areas.

[0118] Step S501 can quickly identify abrupt changes in structural crack values ​​or irregular hollow area curvature in abnormal state data, mark them as high-risk events, and immediately trigger a real-time alarm so that timely measures can be taken to avoid potential safety hazards. At the same time, the remaining data is uploaded to the acceptance platform to generate comprehensive acceptance data including whole-house 3D point cloud data, abnormal state data, and potential uneven areas, providing a solid foundation for subsequent analysis and processing.

[0119] Step S502: Compare the acceptance data with the BIM model and calculate the error matrix. :

[0120] ;

[0121] in, For any coordinate point, Let be the error matrix of the coordinate points, if It automatically triggers design correction suggestions and updates the construction standard library. This is the preset maximum total error.

[0122] Step S502 compares the acceptance data with the BIM model to accurately calculate the error matrix, thereby identifying discrepancies between the design and actual construction. When the error exceeds the set maximum total error, design correction suggestions are automatically triggered, and the construction standard library is updated to ensure the quality and efficiency of subsequent design and construction. This step helps achieve precise alignment between design and construction, improving the overall quality of the building.

[0123] Step S503: Write the hash value of the key data into the distributed ledger. The key data includes three-dimensional point cloud data, ambient temperature data, ambient humidity data, structural crack mutation value, and curvature of regular hollow areas.

[0124] Step S503, writing the hash values ​​of key data into the distributed ledger, effectively ensures data security and traceability. The decentralized nature of the distributed ledger makes data difficult to tamper with, ensuring its authenticity and integrity. Simultaneously, the uniqueness of the hash value makes data records in the ledger highly identifiable, facilitating subsequent data retrieval and analysis. This step provides strong support for the long-term preservation and efficient utilization of building data.

[0125] Step S5 involves real-time monitoring and analysis of building data to promptly identify high-risk events and trigger alarms, while ensuring the comprehensiveness and accuracy of acceptance data. By comparing BIM models, errors between design and construction can be identified, and corrective suggestions can be automatically proposed to optimize design and construction standards. Furthermore, the distributed ledger recording of key data further enhances data security and traceability.

[0126] This method also includes step S6, which generates adversarial examples based on a historical defect database and trains a defect growth prediction model.

[0127] Constructing a cause-effect graph ,node Indicates process parameters, edge Indicates the weight of causal relationships.

[0128] The mathematical expression for step S6 is:

[0129] Defect growth prediction model:

[0130] ;

[0131] in, For the defect area over time Changes, The material degradation coefficient, The stress distribution gradient, For input data.

[0132] Step S6 generates adversarial examples using a historical defect database and trains a defect growth prediction model based on these examples, significantly enhancing the model's ability to predict defect evolution. This step not only improves the model's accuracy in identifying known defect types but also enables it to better adapt to predicting unknown or complex defects, providing strong technical support for the early detection and timely handling of building defects.

[0133] The final output includes key data and a defect growth prediction model.

[0134] This method uses intelligent measurement equipment to collect multi-source building data in real time, and performs fusion analysis and anomaly detection to quickly and accurately identify potential quality problems such as cracks and hollow areas. Simultaneously, it performs dynamic behavior analysis on normal state data, extracts key and non-key frames, and performs dimensionality reduction and compression storage on non-key frames, effectively reducing data storage and processing costs and improving acceptance efficiency.

[0135] Adversarial examples are generated based on a historical defect database to train a defect growth prediction model, which can simulate the defect growth process and provide a scientific basis for long-term monitoring and prevention of building quality. Simultaneously, the hash values ​​of key data are written into a distributed ledger to ensure data immutability and traceability. Based on this, a causal graph is established to correlate the defect growth process with various influencing factors, forming a "scientific guidance framework" for the simulation system.

[0136] Defect growth simulation serves as a "dynamic verification tool" for cause-effect diagrams. By comparing simulation results with actual detection data, the cause-effect diagrams are continuously optimized and improved, enhancing the accuracy and reliability of the prediction model. The cause-effect diagrams, in turn, become a "scientific guidance framework" for the simulation system, providing a scientific basis for the prevention and maintenance of building quality. This transformation not only improves the scientific rigor and accuracy of acceptance testing but also promotes the shift in the construction industry from experience-based acceptance to scientific prevention, laying a solid foundation for the sustainable development of construction projects.

[0137] 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.

[0138] 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. A method for generating household acceptance data based on intelligent measurement equipment, characterized in that, Includes the following steps: Step S1: Acquire multi-source building data, perform fusion analysis on the multi-source building data, and output the fusion detection results; Step S2: Divide the building data into normal state data and abnormal state data, and analyze the normal state data; Step S3: Extract key frames and non-key frames from the normal state data, and perform dimensionality reduction and compression on the non-key frames for storage. Step S4: Trigger high-precision scanning and encoding optimization for abnormal state data, and perform layered encoding on potentially uneven areas; Step S5: Upload the final output to the acceptance platform and early warning system; Step S1 includes the following sub-steps: Step S101, Obtain multi-source building data: A laser scanning robot acquires three-dimensional point cloud data of the entire house, including ambient temperature, humidity, vibration frequency, and sound wave reflection data. The mathematical expression for the three-dimensional point cloud data of the entire house is as follows: ； in, For point cloud data, Let the coordinates of the point cloud be in three-dimensional space. For any coordinate, Let be the intensity of the laser reflection at the coordinates. This represents the total number of points in a single scan. The reflection intensity coefficient; If the vibration frequency exceeds the preset vibration frequency threshold If this occurs, the crack detection mode is triggered, and a set of potential defect coordinates is output. Step S102: Determine anomalies in the sensor data: If the ambient temperature data does not meet the preset ambient temperature threshold or the ambient humidity data does not meet the preset ambient humidity threshold, it is marked as an environmental anomaly. If the sound wave reflection data is less than the preset sound wave reflection threshold, it is determined to be a hollow defect; The output fusion detection results include a set of potential defect coordinates, ambient temperature data, ambient humidity data, and acoustic wave reflection data. Step S2 includes the following sub-steps: Step S201: Divide the data according to the fusion detection results: Extract the point cloud data corresponding to the timestamp of the environmental anomaly and output it as the abnormal state data; output the remaining data as the normal state data. Step S202: Perform dynamic behavior analysis on the normal state data: The mathematical expression for calculating wall flatness based on curvature differences is: ； in, Due to differences in wall curvature, The area to be detected; If the wall surface flatness is greater than the preset wall surface flatness threshold, it is marked as a potential uneven area; if the wall surface flatness is less than the preset wall surface flatness threshold, it is marked as weak motion data. Step S4 includes the following sub-steps: Step S401: For abnormal data, trigger a high-resolution rescan. The laser scanning accuracy is improved from the first accuracy to the second accuracy, and the scanning frequency is improved from the first frequency to the second frequency; Step S402: Perform layered coding on the potential uneven areas: Bitrate allocation based on preset defect severity weights: ； in, For layered coding bitrate, base bitrate, This is the bit rate gain coefficient. This represents the severity weight of the defect.

2. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 1, characterized in that, Step S3 includes the following sub-steps: Step S301: Extract keyframes from the weak motion data; The difference between point clouds in adjacent frames is calculated using the Hausdorff distance: ； in, For any point cloud collection, To and Adjacent point cloud sets, The distance to Hausdorff. For any point, For any other point, For point and points Euclidean distance, For the upper bound, The infimum; like A frame is marked as a keyframe; otherwise, it is a non-keyframe. The preset keyframe determination threshold; Step S302: Perform dimensionality reduction and compression on non-key frames: Principal component analysis was used to reduce the point cloud dimensions from Down to Output the compressed data and store it.

3. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 2, characterized in that, Step S5 includes the following sub-steps: Step S501: If the abnormal state data contains abrupt values ​​of structural cracks or curvature of irregular hollow areas, it is marked as a high-risk event and a real-time alarm is triggered. The remaining data is uploaded to the acceptance platform to generate acceptance data, which includes whole-house 3D point cloud data, abnormal status data, and potential uneven areas. Step S502: Compare the acceptance data with the BIM model and calculate the error matrix. : ； in, For any coordinate point, Let be the error matrix of the coordinate points, if It automatically triggers design correction suggestions and updates the construction standard library. This is the preset maximum total error.

4. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 3, characterized in that, Step S5 further includes: Step S503: Write the hash value of the key data into the distributed ledger. The key data includes three-dimensional point cloud data, ambient temperature data, ambient humidity data, structural crack mutation value, and curvature of regular hollow areas.

5. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 4, characterized in that, It also includes step S6, which generates adversarial examples based on the historical defect database and trains the defect growth prediction model; Constructing a cause-effect graph ,node Indicates process parameters, edge Indicates the weight of causal relationships.

6. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 5, characterized in that, The mathematical expression for step S6 is: Defect growth prediction model: ； in, For the defect area over time Changes, The material degradation coefficient, The stress distribution gradient, For input data.

7. The method for generating household acceptance data based on intelligent measurement equipment as described in claim 6, characterized in that, The final output includes key data and a defect growth prediction model.

Images