Intelligent optimization recommendation method and system for efficient house detection and identification

Abstract

The invention relates to an intelligent optimization recommendation method for efficient house detection and identification. The method comprises the steps that house three-dimensional coordinates and timestamp four-dimensional data are collected; segmenting the pixel-level defect through an intelligent algorithm network and outputting a security level; associating historical defects, rendering into an interactive three-dimensional dynamic model, and embedding anchor points; the defect analysis model predicts potential defects, and the optimization model generates a detection recommendation report. And one-time acquisition, second-level analysis and visual reinspection are realized. The method has the effect of remarkably improving the detection efficiency and precision.

Description

Technical Field

[0001] This invention relates to the technical field of intelligent building inspection, and in particular to an intelligent optimization recommendation method and system for efficient building inspection and assessment. Background Technology

[0002] With the acceleration of urbanization, the stock of existing buildings has increased dramatically, leading to an explosive growth in the demand for building safety assessments. Currently, the mainstream building inspection and assessment techniques still rely on manual inspections, single-point measurements using levels or total stations, 2D CAD drawings, and static reports as the core processes. Specifically, inspectors need to bring equipment such as levels, rebound hammers, and infrared thermal imagers to the site multiple times, manually reading and photographing each layer and surface. The discrete data is then imported into CAD or Excel to generate 2D drawings and written reports.

[0003] The existing technologies have the following technical problems: manual recording and data processing are time-consuming and prone to errors, resulting in discrepancies between the house inspection results and the actual situation, and reducing inspection efficiency. Summary of the Invention

[0004] To improve the efficiency of building inspection, this application provides an intelligent optimization recommendation method and system for efficient building inspection and assessment.

[0005] Firstly, the above-mentioned inventive objective of this application is achieved through the following technical solution:

[0006] A smart optimization recommendation method for efficient building inspection and assessment, the method comprising the following steps:

[0007] Collect multi-source four-dimensional data of the target house, including the house's three-dimensional spatial coordinates and corresponding public timestamps;

[0008] The pre-set semantic segmentation model performs pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, generating a corresponding heat map of house defects.

[0009] The building defect heat map is compared with a pre-set basic heat map, and the structural safety level corresponding to the defect area is output based on the comparison results.

[0010] The defect heatmap, building structure type, and historical maintenance records are associated based on the public timestamp to generate a building historical defect dataset.

[0011] The pre-set 3D construction model analyzes the historical defect dataset of the houses based on a self-learning algorithm and renders it into a corresponding interactive 3D dynamic model, and embeds anchor points in the interactive 3D dynamic model;

[0012] The pre-set defect analysis model performs defect analysis on the interactive 3D dynamic model based on a machine self-learning algorithm to generate corresponding potential house defect data;

[0013] The pre-set house optimization model analyzes the potential defect data of the house and the interactive three-dimensional dynamic model based on the machine self-learning algorithm to generate an optimization detection recommendation report.

[0014] By adopting the above technical solution, a single four-dimensional spatiotemporal acquisition replaces the traditional multiple on-site visits, reducing acquisition time and avoiding data recording errors caused by multiple manual acquisitions. Intelligent algorithm models are used to complete defect classification and safety assessment, improving detection accuracy. Interactive three-dimensional dynamic models are generated in real time and embedded with anchor points, achieving a rapid closed loop of "detection-re-inspection." Multiple on-site surveys are unnecessary, significantly reducing the overall process time compared to traditional solutions. This also reduces manpower input and achieves the effects of "efficient acquisition, efficient analysis, and efficient re-inspection," thereby improving the efficiency of building inspection.

[0015] In a preferred embodiment, this application can be further configured as follows: the step of performing defect analysis on the interactive 3D dynamic model based on a machine learning algorithm using a pre-set defect analysis model to generate corresponding potential housing defect data includes the following steps:

[0016] Taking each defect anchor point in the interactive 3D dynamic model as the center, extract the corresponding six-dimensional feature vectors of defect depth, area, spatial coordinates, material strength, historical temperature and humidity load, and vibration spectrum;

[0017] The six-dimensional feature vector is input into the defect propagation prediction sub-model based on graph neural network, and the three-dimensional diffusion probability field of the defect in the future time step is output.

[0018] Threshold segmentation is performed on the diffusion probability field to obtain the coordinate set of potential defect regions, and this coordinate set is mapped back to the three-dimensional dynamic model to generate potential defect data with risk color marks.

[0019] In a preferred embodiment, this application can be further configured as follows: the step of performing defect analysis on the interactive 3D dynamic model based on a machine learning algorithm using a pre-set defect analysis model to generate corresponding potential housing defect data includes the following steps:

[0020] Read the structural safety level corresponding to the defect anchor point in the interactive 3D dynamic model;

[0021] By calling a finite element reduced-order network, based on the current defect parameters and real-time load data, the stress-strain evolution curve of the defect under material nonlinearity conditions is simulated;

[0022] When the evolution curve reaches the preset failure criterion threshold, the load combination that triggers failure and the corresponding failure mode are recorded as early warning information for potential defects in the building.

[0023] In a preferred embodiment, this application can be further configured such that the step of analyzing the potential defect data of the house and the interactive three-dimensional dynamic model based on a machine learning algorithm using a pre-set house optimization model to generate an optimization detection recommendation report includes the following steps:

[0024] It receives the set of potential defect area coordinates output by the defect propagation prediction sub-model and the damage warning information output by the defect evolution simulation sub-model, forming multi-dimensional potential defect data containing spatial coordinates, damage probability, damage mode, and expected occurrence time.

[0025] The multi-dimensional potential defect data, along with the building structure type, historical maintenance records, material properties, and real-time load data from the interactive 3D dynamic model, are input into the reinforcement learning optimization engine. The optimization engine uses "minimizing detection cost, maximizing structural risk reduction, and minimizing detection cycle" as the joint reward function and iteratively solves for the optimal combination of detection actions through a policy gradient algorithm.

[0026] Based on the optimal combination of detection actions, an optimized detection recommendation report is automatically generated, which includes the three-dimensional coordinates of the detection points, detection methods, a list of required equipment, estimated costs, estimated construction period, and dynamic re-inspection anchor points. It also supports the real-time display of the priority and risk mitigation effect of each detection action in an interactive three-dimensional dynamic model using color coding.

[0027] In a preferred embodiment, this application can be further configured as follows: after the step of performing defect analysis on the interactive 3D dynamic model based on a machine learning algorithm using a pre-set defect analysis model to generate corresponding potential housing defect data, the following steps are included:

[0028] Obtain the coordinates of the houses and call the application programming interface of the meteorological bureau to extract daily temperature, humidity, wind speed, precipitation and extreme weather events in recent years to build a weather-time series dataset;

[0029] Obtain measured building material data for the house, and match the measured concrete strength, steel corrosion potential, and masonry moisture content of the house with the pre-set material database;

[0030] The pre-set material degradation prediction model analyzes the weather-time series dataset and the measured building material data based on a machine self-learning algorithm to form a material-performance degradation curve;

[0031] The pre-set house prediction model predicts the structural aging of the target house based on the material-performance degradation curve and generates an aging prediction model.

[0032] The aging prediction model is compared with the interactive three-dimensional dynamic model to generate detection difference data;

[0033] If the detected difference data exceeds the preset deviation threshold, a corresponding re-inspection recommendation report will be generated.

[0034] In a preferred embodiment, this application can be further configured such that, in the step of generating an aging prediction model by performing structural aging prediction on a target house based on the material-performance degradation curve using a pre-set house prediction model, the house prediction model includes the following formula: ,

[0035] Where t is a user-defined time length, and i is the i-th type of material. Let i be the initial properties of the i-th material. The average daily temperature The daily average relative humidity, This represents the average daily precipitation. This is the highest daily wind speed. Let be the composite material degradation coefficient corresponding to the i-th material in the material-performance degradation curve. For the i-th material, the time correction factor is... and All material-property degradation curves were obtained by analyzing the material-property degradation curves using the LSTM regression algorithm.

[0036] Secondly, the above-mentioned inventive objective of this application is achieved through the following technical solutions:

[0037] An intelligent optimization recommendation device for efficient house inspection and assessment, the device comprising: a house data acquisition unit for acquiring multi-source four-dimensional data of the target house;

[0038] The building defect heatmap generation unit is used to generate a corresponding building defect heatmap by performing pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, with a pre-set semantic segmentation model.

[0039] The structural safety level comparison unit is used to compare the building defect heat map with a preset foundation heat map and output the structural safety level corresponding to the defect area based on the comparison result.

[0040] The housing historical defect dataset generation unit is used to associate the defect heat map, housing structure type and historical maintenance records based on the public timestamp to generate a housing historical defect dataset.

[0041] An interactive 3D dynamic model generation unit is used to pre-set a 3D construction model based on a self-learning algorithm to analyze the historical defect dataset of the houses and render it into a corresponding interactive 3D dynamic model, and to embed anchor points in the interactive 3D dynamic model.

[0042] The building potential defect data generation unit is used to perform defect analysis on the interactive three-dimensional dynamic model based on a pre-set defect analysis model and a machine self-learning algorithm to generate corresponding building potential defect data.

[0043] The optimized inspection recommendation report generation unit is used to analyze the potential defect data of the house and the interactive three-dimensional dynamic model based on the pre-set house optimization model using a machine self-learning algorithm to generate an optimized inspection recommendation report.

[0044] Thirdly, the above-mentioned objectives of this application are achieved through the following technical solutions:

[0045] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned intelligent optimization recommendation method for efficient house inspection and assessment.

[0046] Fourthly, the above-mentioned objectives of this application are achieved through the following technical solutions:

[0047] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned intelligent optimization recommendation method for efficient house inspection and assessment. Attached Figure Description

[0048] Figure 1 This is a flowchart of an intelligent optimization recommendation method for efficient house inspection and assessment in one embodiment of this application;

[0049] Figure 2 This is a block diagram illustrating the principle of an intelligent optimization recommendation device for efficient house inspection and assessment according to one embodiment of this application;

[0050] Figure 3 This is a schematic diagram of an electronic device according to an embodiment of this application.

[0051] Icon labels:

[0052] 1. Building data acquisition unit; 2. Building defect heat map generation unit; 3. Structural safety level comparison unit; 4. Building historical defect dataset generation unit; 5. Interactive 3D dynamic model generation unit; 6. Building potential defect data generation unit; 7. Optimization inspection and recommendation report generation unit. Detailed Implementation

[0053] The present application will be further described in detail below with reference to the accompanying drawings.

[0054] In one embodiment, such as Figure 1 As shown, this application discloses an intelligent optimization recommendation method for efficient building inspection and assessment, which specifically includes the following steps:

[0055] S10: Collect multi-source four-dimensional data of the target house, including the three-dimensional spatial coordinates of the house and the corresponding public timestamp;

[0056] Specifically, a drone equipped with a tilting camera is used to capture 360° images of the target building; simultaneously, a lidar is used to perform a 360° full scan of the building, with a pre-set point density of 5 mm in the authorized embodiment; an infrared thermal imager is used to collect thermal images of the exterior facade at 30-second intervals on the ground, and all data is synchronized via GNSS time synchronization to generate four-dimensional data of XYZ point cloud + RGB + temperature with UTC timestamp.

[0057] S20: The pre-set semantic segmentation model performs pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network to generate a corresponding heat map of house defects.

[0058] Specifically, the four-dimensional data is input into the semantic segmentation model. In this embodiment, the semantic segmentation model has a pre-trained CNN+Transformer network (ResNet50-Encoder + Swin-Decoder), which is trained on a dataset with multiple pixel-level labels for cracks, hollow areas, and peeling. The model outputs a defect heatmap, and the crack width can be identified down to 0.2mm.

[0059] S30: Compare the building defect heat map with the preset foundation heat map, and output the structural safety level corresponding to the defect area based on the comparison result;

[0060] Specifically, the system performs pixel-level difference calculations between the current heat map and the preset "basic heat map". In this embodiment, when the crack area is greater than 10 cm², it is marked as Grade B, and when it is greater than 50 cm², it is marked as Grade C; when the hollow area is greater than 1 m², it is marked as Grade D, directly triggering a high-risk alarm. In this embodiment, the safety level is divided into AE levels. In other embodiments, it can be divided according to other rules.

[0061] Step S30 can quickly generate a safety level distribution map of the entire building, allowing homeowners to visually see the "red, yellow, and green" risk zones.

[0062] S40: Based on the public timestamp, associate the defect heatmap, building structure type and historical maintenance records to generate a building historical defect dataset;

[0063] Specifically, the heat map is aligned with the building structure type (frame / brick-concrete / steel structure) and the maintenance records (crack repair, grouting, reinforcement) of the past 5 years by timestamp to form a CSV: {Defect ID, Date, Type, Area, Structure, Last Repair}. It should be noted that the maintenance period of the maintenance record can be other limited time lengths.

[0064] S50: The pre-set 3D construction model analyzes the historical defect dataset of the house based on a self-learning algorithm and renders it into a corresponding interactive 3D dynamic model, and embeds anchor points in the interactive 3D dynamic model;

[0065] Specifically, the NVIDIA Omniverse real-time rendering engine is used to extract point cloud data features and defect data from the historical defect dataset of the building to generate a rotatable WebGL model; a QR code anchor is embedded in the center of each defect, and scanning the code will bring up historical photos and maintenance records.

[0066] S60: The pre-set defect analysis model performs defect analysis on the interactive three-dimensional dynamic model based on a machine self-learning algorithm to generate corresponding potential house defect data;

[0067] Specifically, by inputting defect coordinates, structural parameters, and historical loads into the LSTM-GNN hybrid model, the model predicts the average crack growth data and void propagation probability data in recent years, and outputs the coordinates of potential defects and the probability of failure, thereby identifying potential hazards in advance and avoiding sudden accidents.

[0068] Specifically, this application uses the NVIDIA Omniverse real-time rendering engine to generate a rotatable WebGL model from point cloud and defect data; a QR code anchor is embedded in the center of each defect, and scanning the code will bring up historical photos and repair records.

[0069] Homeowners and supervisors can use a mobile app to check the interior condition of a target property, click on defects to view details, and effectively improve communication efficiency.

[0070] S70: The pre-set house optimization model analyzes the potential defect data of the house and the interactive three-dimensional dynamic model based on the machine self-learning algorithm to generate an optimization detection recommendation report;

[0071] Specifically, in this embodiment, for the target house, after the house optimization model completes the defect anchor point modeling, it immediately outputs a comprehensive inspection recommendation report covering the entire process: First, a three-dimensional roadmap is constructed using drone flight paths, laser base stations, and handheld verification points; then, a list of methods for rebound verification, infrared verification, drone re-inspection, and tilt monitoring is listed layer by layer according to risk level, and the required equipment, budget Gantt, and QR code anchor point layout plan are given simultaneously to achieve paperless on-site guidance; finally, through simulations of three working conditions—typhoon, earthquake, and long-term load—suggestions for crack propagation, tilt risk, and future re-inspection are given, significantly compressing the inspection cycle, reducing the re-inspection positioning error to the centimeter level, and allowing the owner to view the three-dimensional progress and risk warnings in real time, greatly improving decision-making efficiency.

[0072] In summary, according to steps S10-S70, millimeter-level four-dimensional spatiotemporal data is simultaneously collected using drones, LiDAR, and infrared thermal imagers, replacing the traditional method of multiple rounds of manual on-site inspections with measuring tapes and levels. Subsequently, a CNN+Transformer hybrid network is used to complete pixel-level defect segmentation within seconds and output an AE-level safety rating, improving efficiency by an order of magnitude compared to the manual interpretation of existing static reports. Furthermore, defects, structural types, and historical maintenance records are linked based on public timestamps to form a traceable historical defect dataset, avoiding information gaps caused by fragmented traditional archives. Next, a 3D model is used to render the dataset into an interactive dynamic model, embedding QR codes / RFID anchors into each defect to achieve centimeter-level in-situ comparison during re-inspection, whereas existing solutions still require secondary surveys and manual comparisons. Finally, the building optimization model provides an optimization report based on potential defects and real-time loads, including inspection points, methods, equipment, budget, and re-inspection paths. This allows for direct execution without further on-site inspection, shortening the overall inspection cycle, reducing labor costs, and truly achieving the efficient inspection goal of "one-time collection, second-level analysis, and one-click re-inspection."

[0073] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0074] In one embodiment, an intelligent optimization recommendation device for efficient building inspection and assessment is provided, which corresponds one-to-one with the intelligent optimization recommendation method for efficient building inspection and assessment described in the above embodiments. For example... Figure 2 As shown, the intelligent optimization recommendation device for efficient house inspection and assessment includes:

[0075] House data acquisition unit 1 is used to collect multi-source four-dimensional data of the target house;

[0076] The building defect heat map generation unit 2 is used to generate a corresponding building defect heat map by performing pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, with a pre-set semantic segmentation model.

[0077] The structural safety level comparison unit 3 is used to compare the building defect heat map with the preset foundation heat map, and output the structural safety level corresponding to the defect area based on the comparison result.

[0078] The housing history defect dataset generation unit 4 is used to associate the defect heat map, housing structure type and historical maintenance records based on the public timestamp to generate a housing history defect dataset.

[0079] The interactive 3D dynamic model generation unit 5 is used to pre-set a 3D construction model based on a self-learning algorithm to analyze the historical defect dataset of the house and render it into a corresponding interactive 3D dynamic model, and to embed anchor points in the interactive 3D dynamic model.

[0080] The building potential defect data generation unit 6 is used to pre-set a defect analysis model to perform defect analysis on the interactive three-dimensional dynamic model based on a machine self-learning algorithm, so as to generate corresponding building potential defect data.

[0081] The optimization inspection recommendation report generation unit 7 is used to generate an optimization inspection recommendation report by analyzing the potential defect data and interactive three-dimensional dynamic model of the house based on a pre-set house optimization model using a machine self-learning algorithm.

[0082] Specific limitations regarding the intelligent optimization recommendation device for efficient building inspection and assessment can be found in the limitations of the intelligent optimization recommendation method for efficient building inspection and assessment described above, and will not be repeated here. Each module in the aforementioned intelligent optimization recommendation device for efficient building inspection and assessment can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the electronic device, or stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of each module.

[0083] In one embodiment, an electronic device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown, the electronic device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores the database. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements an intelligent optimization recommendation method for efficient building inspection and assessment.

[0084] In one embodiment, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0085] Collect multi-source four-dimensional data of the target house, including the house's three-dimensional spatial coordinates and corresponding public timestamps;

[0086] The pre-set semantic segmentation model performs pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, generating a corresponding heat map of house defects.

[0087] The building defect heat map is compared with a pre-set basic heat map, and the structural safety level corresponding to the defect area is output based on the comparison results.

[0088] The defect heatmap, building structure type, and historical maintenance records are associated based on the public timestamp to generate a building historical defect dataset.

[0089] The pre-set 3D construction model analyzes the historical defect dataset of the houses based on a self-learning algorithm and renders it into a corresponding interactive 3D dynamic model, and embeds anchor points in the interactive 3D dynamic model;

[0090] The pre-set defect analysis model performs defect analysis on the interactive 3D dynamic model based on a machine self-learning algorithm to generate corresponding potential house defect data;

[0091] The pre-set house optimization model analyzes the potential defect data of the house and the interactive three-dimensional dynamic model based on the machine self-learning algorithm to generate an optimization detection recommendation report.

[0092] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0093] Collect multi-source four-dimensional data of the target house, including the house's three-dimensional spatial coordinates and corresponding public timestamps;

[0094] The pre-set semantic segmentation model performs pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, generating a corresponding heat map of house defects.

[0095] The building defect heat map is compared with a pre-set basic heat map, and the structural safety level corresponding to the defect area is output based on the comparison results.

[0096] The defect heatmap, building structure type, and historical maintenance records are associated based on the public timestamp to generate a building historical defect dataset.

[0097] The pre-set 3D construction model analyzes the historical defect dataset of the houses based on a self-learning algorithm and renders it into a corresponding interactive 3D dynamic model, and embeds anchor points in the interactive 3D dynamic model;

[0098] The pre-set defect analysis model performs defect analysis on the interactive 3D dynamic model based on a machine self-learning algorithm to generate corresponding potential house defect data;

[0099] The pre-set house optimization model analyzes the potential defect data of the house and the interactive three-dimensional dynamic model based on the machine self-learning algorithm to generate an optimization detection recommendation report.

[0100] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0101] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0102] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An intelligent optimization recommendation method for efficient building inspection and assessment, characterized in that, The method includes the steps of: collecting multi-source four-dimensional data of the target house, wherein the multi-source four-dimensional data includes the three-dimensional spatial coordinates of the house and the corresponding public timestamp; The pre-set semantic segmentation model performs pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, generating a corresponding heat map of house defects. The building defect heat map is compared with a pre-set basic heat map, and the structural safety level corresponding to the defect area is output based on the comparison results. The defect heatmap, building structure type, and historical maintenance records are associated based on the public timestamp to generate a building historical defect dataset. The pre-set 3D construction model analyzes the historical defect dataset of the houses based on a self-learning algorithm and renders it into a corresponding interactive 3D dynamic model, and embeds anchor points in the interactive 3D dynamic model; The pre-set defect analysis model performs defect analysis on the interactive 3D dynamic model based on a machine self-learning algorithm to generate corresponding potential house defect data; The pre-set house optimization model analyzes the potential defect data of the house and the interactive three-dimensional dynamic model based on the machine self-learning algorithm to generate an optimization detection recommendation report.

2. The intelligent optimization recommendation method for efficient building inspection and assessment according to claim 1, characterized in that, The step of performing defect analysis on the interactive 3D dynamic model based on a pre-set defect analysis model using a machine self-learning algorithm to generate corresponding potential building defect data includes the following steps: Taking each defect anchor point in the interactive 3D dynamic model as the center, extract the corresponding six-dimensional feature vectors of defect depth, area, spatial coordinates, material strength, historical temperature and humidity load, and vibration spectrum; The six-dimensional feature vector is input into the defect propagation prediction sub-model based on graph neural network, and the three-dimensional diffusion probability field of the defect in the future time step is output. Threshold segmentation is performed on the diffusion probability field to obtain the coordinate set of potential defect regions, and this coordinate set is mapped back to the three-dimensional dynamic model to generate potential defect data with risk color marks.

3. The intelligent optimization recommendation method for efficient building inspection and assessment according to claim 2, characterized in that, The step of performing defect analysis on the interactive 3D dynamic model based on a pre-set defect analysis model using a machine self-learning algorithm to generate corresponding potential building defect data includes the following steps: Read the structural safety level corresponding to the defect anchor point in the interactive 3D dynamic model; By calling a finite element reduced-order network, based on the current defect parameters and real-time load data, the stress-strain evolution curve of the defect under material nonlinearity conditions is simulated; When the evolution curve reaches the preset failure criterion threshold, the load combination that triggers failure and the corresponding failure mode are recorded as early warning information for potential defects in the building.

4. The intelligent optimization recommendation method for efficient building inspection and assessment according to claim 3, characterized in that, The step of analyzing the potential defect data and interactive 3D dynamic model of the house based on a machine learning algorithm in a pre-set house optimization model to generate an optimization detection and recommendation report includes the following steps: It receives the set of potential defect area coordinates output by the defect propagation prediction sub-model and the damage warning information output by the defect evolution simulation sub-model, forming multi-dimensional potential defect data containing spatial coordinates, damage probability, damage mode, and expected occurrence time. The multi-dimensional potential defect data, along with the building structure type, historical maintenance records, material properties, and real-time load data from the interactive 3D dynamic model, are input into the reinforcement learning optimization engine. The optimization engine uses "minimizing detection cost, maximizing structural risk reduction, and minimizing detection cycle" as the joint reward function and iteratively solves for the optimal combination of detection actions through a policy gradient algorithm. Based on the optimal combination of detection actions, an optimized detection recommendation report is automatically generated, which includes the three-dimensional coordinates of the detection points, detection methods, a list of required equipment, estimated costs, estimated construction period, and dynamic re-inspection anchor points. It also supports the real-time display of the priority and risk mitigation effect of each detection action in an interactive three-dimensional dynamic model using color coding.

5. The intelligent optimization recommendation method for efficient building inspection and assessment according to claim 4, characterized in that, After the step of performing defect analysis on the interactive 3D dynamic model based on a machine learning algorithm using a pre-set defect analysis model to generate corresponding potential housing defect data, the following steps are included: Obtain the coordinates of the houses and call the application programming interface of the meteorological bureau to extract daily temperature, humidity, wind speed, precipitation and extreme weather events in recent years to build a weather-time series dataset; Obtain measured building material data for the house, and match the measured concrete strength, steel corrosion potential, and masonry moisture content of the house with the pre-set material database; The pre-set material degradation prediction model analyzes the weather-time series dataset and the measured building material data based on a machine self-learning algorithm to form a material-performance degradation curve; The pre-set house prediction model predicts the structural aging of the target house based on the material-performance degradation curve and generates an aging prediction model. The aging prediction model is compared with the interactive three-dimensional dynamic model to generate detection difference data; If the detected difference data exceeds the preset deviation threshold, a corresponding re-inspection recommendation report will be generated.

6. The intelligent optimization recommendation method for efficient building inspection and assessment according to claim 5, characterized in that, In the step of generating an aging prediction model by predicting the structural aging of a target house based on the material-performance degradation curve using a pre-set house prediction model, the house prediction model includes the following formula: , Where t is a user-defined time length, and i is the i-th type of material. Let i be the initial properties of the i-th material. The average daily temperature The daily average relative humidity, This represents the average daily precipitation. This is the highest daily wind speed. Let be the composite material degradation coefficient corresponding to the i-th material in the material-performance degradation curve. For the i-th material, the time correction factor is... and All material-property degradation curves were obtained by analyzing the material-property degradation curves using the LSTM regression algorithm.

7. An intelligent optimization recommendation device for efficient building inspection and assessment, applied to the intelligent optimization recommendation method for efficient building inspection and assessment as described in any one of claims 1-6, characterized in that, The device includes: The housing data acquisition unit (1) is used to collect multi-source four-dimensional data of the target housing; The building defect heat map generation unit (2) is used to generate a corresponding building defect heat map by performing pixel-level defect semantic segmentation on the four-dimensional data based on a self-learning hybrid network, with a pre-set semantic segmentation model. The structural safety level comparison unit (3) is used to compare the building defect heat map with the preset foundation heat map and output the structural safety level corresponding to the defect area based on the comparison result. The housing history defect dataset generation unit (4) is used to associate the defect heat map, housing structure type and historical maintenance records based on the public timestamp to generate a housing history defect dataset. An interactive three-dimensional dynamic model generation unit (5) is used to pre-set a three-dimensional construction model based on a self-learning algorithm to analyze the historical defect dataset of the house and render it into a corresponding interactive three-dimensional dynamic model, and to embed anchor points in the interactive three-dimensional dynamic model. The building potential defect data generation unit (6) is used to pre-set a defect analysis model to perform defect analysis on the interactive three-dimensional dynamic model based on a machine self-learning algorithm, so as to generate corresponding building potential defect data. The optimization inspection recommendation report generation unit (7) is used to generate an optimization inspection recommendation report by analyzing the potential defect data and interactive three-dimensional dynamic model of the house based on the machine self-learning algorithm of the pre-set house optimization model.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent optimization recommendation method for efficient house inspection and assessment as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent optimization recommendation method for efficient house inspection and assessment as described in any one of claims 1 to 7.

Images