A housing full life cycle operation and maintenance management system

Abstract

This invention provides a building lifecycle operation and maintenance management system, belonging to the field of information management technology. The maintenance module of this system performs the following steps: establishing a three-dimensional model of the building based on two-dimensional design drawings and panoramic images; performing finite element analysis on the three-dimensional model to identify multiple areas where the stress exceeds a stress threshold, designated as weak areas; acquiring the building materials and proportions of the weak areas; acquiring the penetrating ultrasonic signals of the weak areas and performing noise reduction, filtering, and enhancement preprocessing to obtain a preprocessed signal; extracting the feature matrix of the preprocessed signal and fusing it with material vectors to form a fusion matrix; using a pre-trained ultrasonic signal-stability model, inputting the fusion matrix to obtain the stability index of the weak areas; if the stability index is lower than the stability index threshold, outputting a maintenance reminder to the user, achieving non-destructive monitoring and high-precision assessment.

Description

Technical Field

[0001] The present invention belongs to the technical field of information management, and in particular, relates to a building full life cycle operation and maintenance management system. Background Art

[0002] Traditional houses have a service life of 50-70 years, and their structural safety is directly related to the safety of life and property. At present, the maintenance and evaluation of house structures are mainly based on regular manual inspections, which have problems such as low inspection frequency and high subjectivity in evaluation. Regarding the issue of house structure safety assessment, existing technologies mainly focus on the following aspects:

[0003] (1) Manual percussion method. Workers use a copper hammer or other physical method to knock on the structure and judge the internal quality by the sound changes. This subjective judgment result has the disadvantages of large variability and poor repeatability.

[0004] (2) Geochemical testing method. Collect trace substances on the surface of the structure for chemical analysis to determine the corrosion situation. This method can only characterize the surface of the material and has poor ductility.

[0005] (3) X-ray detection technology. It uses penetrating rays to detect the internal quality of building materials. This technology has high equipment costs, high radiation levels during on-site operation, and low utilization rate.

[0006] (4) Structural monitoring system. A long-term monitoring device is set up to record the structural response in real time. This device has poor stability, complex maintenance, and is prone to failure after long-term use.

[0007] In summary, the existing various sensing technologies or monitoring systems have common problems such as difficulty in balancing detection depth and accuracy, high equipment costs, and high maintenance complexity. They cannot meet the needs of structural safety assessment throughout the life cycle of a house. Summary of the invention

[0008] In view of this, the present invention provides a building full life cycle operation and maintenance management system, which realizes rapid and accurate evaluation of the building structure, truly reflects the material status of weak areas, and improves the efficiency of safety warning.

[0009] The present invention is achieved in that:

[0010] The present invention provides a housing full life cycle operation and maintenance management system, which includes the following functional modules:

[0011] User management module, used to handle the registration, login, authority allocation, and personal information management of house owners;

[0012] The house archive module is used to store detailed information of each house, including building structure, design drawings, construction materials, and completion date;

[0013] A repair and maintenance module is used to calculate the stability of the house and remind the user to carry out repair and maintenance activities on the house when the stability of the house is lower than the stability index threshold;

[0014] Reporting and analysis module, used to generate comprehensive reports of the house, including usage status and maintenance history;

[0015] Wherein, the maintenance module is used to perform the following steps: Wherein, the maintenance module is used to perform the following steps:

[0016] S10, establishing a three-dimensional model of the house based on the two-dimensional design drawing and the panoramic image of the house;

[0017] S20, performing finite element analysis based on the three-dimensional model of the house, obtaining multiple areas of the house where the stress is greater than the stress threshold, and recording them as weak areas;

[0018] S30, obtaining the building material and proportion of the weak area to form a material vector;

[0019] S40, obtaining a penetrating ultrasonic signal of the weak area, wherein the penetrating ultrasonic signal is emitted by an ultrasonic generator disposed on one side of the weak area and received by an ultrasonic sensor disposed on another side of the weak area, and performing preprocessing of noise reduction, filtering and enhancement on the penetrating ultrasonic signal to obtain a preprocessed signal;

[0020] S50, extracting the characteristic matrix of the preprocessed signal and fusing it with the material vector to form a fusion matrix;

[0021] S60, using a pre-trained ultrasound signal-stability model, inputting the fusion matrix, and obtaining a stability index of the weak area;

[0022] S70: If the stability index is lower than the stability index threshold, output a maintenance reminder to the user.

[0023] Wherein, S10, based on the two-dimensional design drawing of the house and the panoramic image, establish a three-dimensional model of the house;

[0024] The specific steps are:

[0025] Collect two-dimensional design drawings of the house, including floor plans, elevations, and sections;

[0026] Take panoramic images of the real house scene to obtain indoor and outdoor panoramic images;

[0027] Establish a 3D reconstruction algorithm, use structured light and SLAM technology, use the design drawing as the framework, and combine the panoramic image details to reconstruct the 3D point cloud;

[0028] Poisson surface reconstruction and Ball Pivoting algorithm are used to convert the point cloud into a fine mesh model and establish a three-dimensional model of the house.

[0029] S20, performing finite element analysis based on the three-dimensional model of the house, obtaining multiple areas of the house where the stress is greater than the stress threshold, and recording them as weak areas;

[0030] The specific steps are:

[0031] The Delaunay triangulation method is used to mesh the three-dimensional model;

[0032] According to the properties of the material in each direction, determine whether the material is isotropic or anisotropic, and establish a unified structural model of the mixed material. If the material has the same properties in each direction, it is an isotropic material; if the material has different properties in each direction, it is an anisotropic material.

[0033] Set displacement constraints, support constraints, static loads, and dynamic loads;

[0034] Gaussian elimination method and pursuit method are used to solve the stiffness equation and calculate the stress-strain distribution;

[0035] According to the calculated stress-strain distribution, an area where the stress is greater than the stress threshold is obtained and recorded as a weak area.

[0036] S30, obtaining the building material and proportion of the weak area to form a material vector;

[0037] The specific steps are:

[0038] Based on the three-dimensional model, extract the components in the weak area and perform segmentation and classification;

[0039] Querying an engineering database to obtain a material list of the weak area components, wherein the engineering database contains information on material types, specifications, and performance of all components;

[0040] The usage amount or proportion of the component materials in the weak area is counted, a component-material mapping relationship table is established, and a material vector is formed.

[0041] S40, obtaining a penetrating ultrasonic signal of the weak area, wherein the penetrating ultrasonic signal is emitted by an ultrasonic generator disposed on one side of the weak area and received by an ultrasonic sensor disposed on another side of the weak area, and performing preprocessing of noise reduction, filtering and enhancement on the penetrating ultrasonic signal to obtain a preprocessed signal;

[0042] The specific steps are:

[0043] Ultrasonic transmitting and receiving probes are respectively installed on both sides of the weak area;

[0044] Set the frequency and pulse parameters to collect penetration signals;

[0045] Performing noise reduction processing, the noise reduction processing methods include wavelet transform and median filtering;

[0046] Use bandpass filtering to filter the signal; amplify the envelope to enhance the signal; extract signal features to interpret the internal quality of the material.

[0047] S50, extracting the characteristic matrix of the preprocessed signal and fusing it with the material vector to form a fusion matrix;

[0048] The specific steps are:

[0049] Extract the time and frequency domain features of the ultrasonic signal to construct a matrix;

[0050] Perform PCA or LASSO feature selection;

[0051] Concatenate material vectors and feature matrices to integrate new feature expressions;

[0052] Prepare samples with different degradation levels and collect feature matrices and true value samples;

[0053] Establish a structural status to stability assessment model.

[0054] S60, using a pre-trained ultrasound signal-stability model, inputting the fusion matrix, and obtaining a stability index of the weak area;

[0055] The specific steps are:

[0056] The evaluation model is trained using the GBDT algorithm;

[0057] Preprocess the fusion feature matrix;

[0058] Input the model to predict the stability index of the weak area;

[0059] Denormalize to get the actual index value.

[0060] S70: If the stability index is lower than the stability index threshold, output a maintenance reminder to the user.

[0061] The specific steps are:

[0062] Preset stability index threshold;

[0063] comparing the predicted index with the stability index threshold;

[0064] If the stability index is lower than the threshold, a maintenance suggestion is generated to prompt the user;

[0065] Update processing status to prevent duplicate reminders.

[0066] The user management module is used to perform the following:

[0067] Receive and process registration information from property owners, including identity verification and information entry;

[0068] Receive and verify user login requests, ensure the legitimacy of user identities, and use authentication algorithms, including hash algorithms, to securely store and verify user passwords;

[0069] Automatically or manually assigning permissions based on user roles and levels, including using role-based access control or attribute-based access control algorithms;

[0070] Use CRUD operations to manage user information, where CRUD operations include create, read, update, and delete.

[0071] The reporting and analysis module is used to perform the following:

[0072] Obtain detailed information about the house and use the historical database, process the data through data aggregation and visualization algorithms, use statistical methods to process and summarize, and generate a comprehensive report on the house;

[0073] Record the maintenance history of each house, including maintenance time, maintenance content, and maintenance personnel information;

[0074] Conduct in-depth analysis of the generated housing comprehensive report through machine learning algorithms, including statistical analysis, trend analysis, and predictive models. Machine learning algorithms include linear regression and time series analysis.

[0075] Machine learning algorithms are used to identify patterns and predict future trends.

[0076] Compared with the prior art, the beneficial effects of the housing full life cycle operation and maintenance management system provided by the present invention are:

[0077] (1) An intelligent detection technology integrating ultrasonic waves and AI algorithms has been established to achieve in-depth structural and non-destructive testing of the internal quality of materials, covering all kinds of common building materials and expanding the scope of application;

[0078] (2) Construct 3D printed samples, fit the actual working environment, train the GBDT model, and achieve high-precision material state assessment;

[0079] (3) The assessment results are interpretable, which can identify the physical location of structural weaknesses and automatically generate treatment suggestions. When serious hidden dangers are detected, the treatment measures are fed back to the construction party and the owner in real time;

[0080] (4) The system is flexible and convenient to deploy. It can be integrated into the routine inspection process or laid out as a continuous monitoring facility, reducing additional costs and eliminating the need for professional maintenance personnel, thus achieving long-term management of the health of the building structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0082] Figure 1 The figure is a flow chart of a house full life cycle operation and maintenance management system. DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0084] like Figure 1 FIG. 1 is a flow chart of a first embodiment of a housing full life cycle operation and maintenance management system provided by the present invention. In this embodiment, the following functional modules are included:

[0085] User management module, used to handle the registration, login, authority allocation, and personal information management of house owners;

[0086] The user management module is used to perform the following:

[0087] Receive and process registration information from property owners, including identity verification and information entry;

[0088] Receive and verify user login requests, ensure the legitimacy of user identities, and use authentication algorithms, including hash algorithms, to securely store and verify user passwords;

[0089] Automatically or manually assigning permissions based on user roles and levels, including using role-based access control or attribute-based access control algorithms;

[0090] Use CRUD operations to manage user information, where CRUD operations include create, read, update, and delete.

[0091] The house archive module is used to store detailed information of each house, including building structure, design drawings, construction materials, and completion date;

[0092] A repair and maintenance module is used to calculate the stability of the house and remind the user to carry out repair and maintenance activities on the house when the stability of the house is lower than the stability index threshold;

[0093] Reporting and analysis module, used to generate comprehensive reports of the house, including usage status and maintenance history;

[0094] The Reporting and Analysis module is used to perform the following:

[0095] Obtain detailed information about the house and use the historical database, process the data through data aggregation and visualization algorithms, use statistical methods to process and summarize, and generate a comprehensive report on the house;

[0096] Record the maintenance history of each house, including maintenance time, maintenance content, and maintenance personnel information;

[0097] Conduct in-depth analysis of the generated housing comprehensive report through machine learning algorithms, including statistical analysis, trend analysis, and predictive models. Machine learning algorithms include linear regression and time series analysis.

[0098] Machine learning algorithms are used to identify patterns and predict future trends.

[0099] The maintenance module is used to perform the following steps:

[0100] S10, establishing a three-dimensional model of the house based on the two-dimensional design drawing and the panoramic image of the house;

[0101] The specific steps are:

[0102] Collect two-dimensional design drawings of the house, including floor plans, elevations, and sections;

[0103] Take panoramic images of the real house scene to obtain indoor and outdoor panoramic images;

[0104] Establish a 3D reconstruction algorithm, use structured light and SLAM technology, use the design drawing as the framework, and combine the panoramic image details to reconstruct the 3D point cloud;

[0105] Poisson surface reconstruction and Ball Pivoting algorithm are used to convert the point cloud into a fine mesh model and establish a three-dimensional model of the house.

[0106] S20, performing finite element analysis based on the three-dimensional model of the house, obtaining multiple areas of the house where the stress is greater than the stress threshold, and recording them as weak areas;

[0107] The specific steps are:

[0108] The Delaunay triangulation method is used to mesh the three-dimensional model;

[0109] According to the properties of the material in each direction, determine whether the material is isotropic or anisotropic, and establish a unified structural model of the mixed material. If the material has the same properties in each direction, it is an isotropic material; if the material has different properties in each direction, it is an anisotropic material.

[0110] Set displacement constraints, support constraints, static loads, and dynamic loads;

[0111] Gaussian elimination method and pursuit method are used to solve the stiffness equation and calculate the stress-strain distribution;

[0112] According to the calculated stress-strain distribution, the area where the stress is greater than the stress threshold is obtained and recorded as the weak area.

[0113] S30, obtaining the building materials and proportions of the weak area to form a material vector;

[0114] The specific steps are:

[0115] Based on the 3D model, extract components in weak areas and perform segmentation and classification;

[0116] Query the engineering database to obtain the material details of the components in the weak area. The engineering database contains information on the material type, specifications, and performance of all components;

[0117] Count the usage or proportion of component materials in weak areas, establish a component-material mapping relationship table, and form a material vector.

[0118] S40, obtaining a penetrating ultrasonic signal of the weak area, where the penetrating ultrasonic signal is emitted by an ultrasonic generator disposed on one side of the weak area and received by an ultrasonic sensor disposed on another side of the weak area, and performing preprocessing of noise reduction, filtering and enhancement on the penetrating ultrasonic signal to obtain a preprocessed signal;

[0119] The specific steps are:

[0120] Ultrasonic transmitting and receiving probes are installed on both sides of the weak area;

[0121] Set the frequency and pulse parameters to collect penetration signals;

[0122] Performing noise reduction processing, the noise reduction processing methods include wavelet transform and median filtering;

[0123] Use bandpass filtering to filter the signal; amplify the envelope to enhance the signal; extract signal features to interpret the internal quality of the material.

[0124] S50, extracting the characteristic matrix of the preprocessed signal and fusing it with the material vector to form a fusion matrix;

[0125] The specific steps are:

[0126] Extract the time and frequency domain features of the ultrasonic signal to construct a matrix;

[0127] Perform PCA or LASSO feature selection;

[0128] Concatenate material vectors and feature matrices to integrate new feature expressions;

[0129] Prepare samples with different degradation levels and collect feature matrices and true value samples;

[0130] Establish a structural status to stability assessment model.

[0131] S60, using a pre-trained ultrasound signal-stability model, inputting a fusion matrix, and obtaining a stability index of the weak area;

[0132] The specific steps are:

[0133] Use GBDT algorithm to train the evaluation model;

[0134] Preprocess the fusion feature matrix;

[0135] Input the model to predict the stability index of weak areas;

[0136] Denormalize to get the actual index value.

[0137] S70: If the stability index is lower than the stability index threshold, output a maintenance reminder to the user.

[0138] The specific steps are:

[0139] Preset stability index threshold;

[0140] Comparing the prediction index with the robustness index threshold;

[0141] If the stability index is below the threshold, maintenance suggestions are generated to prompt the user;

[0142] Update processing status to prevent duplicate reminders.

[0143] A second embodiment of a housing full life cycle operation and maintenance management system provided by the present invention includes the following functional modules:

[0144] User management module, used to handle the registration, login, authority allocation, and personal information management of house owners;

[0145] The user management module is used to perform the following:

[0146] Receive and process registration information from property owners, including identity verification and information entry;

[0147] Receive and verify user login requests, ensure the legitimacy of user identities, and use authentication algorithms, including hash algorithms, to securely store and verify user passwords;

[0148] Automatically or manually assigning permissions based on user roles and levels, including using role-based access control or attribute-based access control algorithms;

[0149] Use CRUD operations to manage user information, where CRUD operations include create, read, update, and delete.

[0150] The house archive module is used to store detailed information of each house, including building structure, design drawings, construction materials, and completion date;

[0151] A repair and maintenance module is used to calculate the stability of the house and remind the user to carry out repair and maintenance activities on the house when the stability of the house is lower than the stability index threshold;

[0152] Reporting and analysis module, used to generate comprehensive reports of the house, including usage status and maintenance history;

[0153] The Reporting and Analysis module is used to perform the following:

[0154] Obtain detailed information about the house and use the historical database, process the data through data aggregation and visualization algorithms, use statistical methods to process and summarize, and generate a comprehensive report on the house;

[0155] Record the maintenance history of each house, including maintenance time, maintenance content, and maintenance personnel information;

[0156] Conduct in-depth analysis of the generated housing comprehensive report through machine learning algorithms, including statistical analysis, trend analysis, and predictive models. Machine learning algorithms include linear regression and time series analysis.

[0157] Machine learning algorithms are used to identify patterns and predict future trends.

[0158] The maintenance module is used to perform the following steps:

[0159] S10, establishing a three-dimensional model of the house based on the two-dimensional design drawing and the panoramic image of the house;

[0160] The goal of this step is to build a 3D model of the house. The specific process is as follows:

[0161] (1) Collect two-dimensional drawings of house design;

[0162] It is represented as a design atlas G = {g1, g2, ..., g n}, where g iRepresents the i-th design two-dimensional drawing, mainly including plan view, elevation view, section view, etc.

[0163] (2) Obtaining a panoramic image of the house;

[0164] Use panoramic camera equipment to collect indoor and outdoor panoramic images to obtain a panoramic image set P = {p1, p2, ..., p m}, where p j represents the jth panoramic image.

[0165] (3) 3D reconstruction;

[0166] Use the structured light 3D reconstruction algorithm to construct the overall 3D framework through the design graph G. The structured light algorithm is based on Gray code theory and can quickly obtain a relatively complete 3D framework.

[0167] Using the visual SLAM algorithm, the 3D scene details are supplemented through the panoramic image P. The SLAM algorithm can track image features in real time and incrementally build a 3D model.

[0168] The combined 3D reconstruction result is the point cloud model M c ={m1, m2, ..., m l}, containing l three-dimensional points.

[0169] (4) 3D reconstruction;

[0170] Point cloud data cannot fully express the surface geometry information of three-dimensional objects. Poisson surface reconstruction algorithm is used to achieve high-quality meshing.

[0171] The Poisson surface reconstruction function is Recon(M c , k), where k is the number of target patches.

[0172] The algorithm obtains the surface normal vector information by solving the Poisson equation and fitting the original sample points.

[0173] Convert point clouds into progressive triangulated mesh models.

[0174] The final reconstructed three-dimensional model is M = {V, E, F}, where V, E, F represent the point set, edge set, and face set.

[0175] It should be noted that the variables are explained as follows:

[0176] S10, including the following variables:

[0177] G stands for design atlas, which means the collected two-dimensional design drawings;

[0178] g i represents the i-th two-dimensional design drawing;

[0179] P represents a panoramic image set, which represents the collected panoramic images of the house;

[0180] p j represents the jth panoramic image;

[0181] M c Represents the reconstructed three-dimensional point cloud model;

[0182] m i Represents the i-th 3D point in the model;

[0183] M represents the reconstructed three-dimensional mesh model;

[0184] V, E, and F represent the point set, edge set, and face set in the model respectively.

[0185] S20, performing finite element analysis based on the three-dimensional model of the house, obtaining multiple areas of the house where the stress is greater than the stress threshold, and recording them as weak areas;

[0186] The purpose of this step is to construct a finite element model and perform structural stress analysis. The main process is as follows:

[0187] (1) Grid discretization;

[0188] The surface model M is input and voxelized to obtain volume data T.

[0189] Apply Delaunay triangulation to mesh the volume data:

[0190] T -1 (M) = Mesh(T);

[0191] Among them, the operator T -1 Represents volume data conversion function, and Mesh represents triangulation function.

[0192] Get the discrete mesh model Mesh.

[0193] (2) Material modeling;

[0194] Define the material parameter matrix:

[0195] Mat={e, v, ρ,...};

[0196] Where E is Young's modulus, v is Poisson's ratio, and ρ is density. The matrix Mat is set according to different material types.

[0197] (3) Set constraints and loads;

[0198] Includes displacement constraints, support constraints, and static and dynamic loads. Use functions to apply constraints and loads:

[0199] BL=Load(Mesh,Constraints);

[0200] Where BL represents the boundary condition list.

[0201] (4) Solve stress and strain;

[0202] Assembly stiffness equation:

[0203] [K]{U}={F};

[0204] Where [K] is the global stiffness matrix, {U} is the node displacement vector, and {F} is the load vector.

[0205] The displacement solution is solved by a linear equation solver.

[0206] Calculate the stress and strain distribution of each element.

[0207] It should be noted that the variables are explained as follows:

[0208] S20, including the following variables:

[0209] T represents volume data, which represents the volume model after voxelization;

[0210] Mesh represents a gridded discrete model;

[0211] Mat represents the material parameter matrix, including parameters such as Young's modulus and Poisson's ratio;

[0212] BL represents the boundary condition list, which includes various constraints and loads;

[0213] [k] represents the stiffness matrix;

[0214] {U} represents the node displacement vector;

[0215] {F} represents the load vector.

[0216] S30, obtaining the building materials and proportions of the weak area to form a material vector;

[0217] The purpose of this step is to determine the material composition ratio of the weak area. The implementation process is:

[0218] (1) Segmentation of weak areas;

[0219] Let the weak patch set be R = {r1, r2, ..., r n},but:

[0220] r i ∈Mesh,Stress(r i )>AllowableStress;

[0221] That is, the sheet with excessive stress belongs to the weak area.

[0222] (2) Mapping material information;

[0223] Query the engineering material database to obtain the material mapping of the weak piece:

[0224] MTable = Query(R);

[0225] Among them, MTable represents the material proportion mapping table.

[0226] (3) Generate material vector;

[0227] Count the usage ratio of each material and generate the material vector:

[0228] MatVector=Percentage(MTable)

[0229] ={m1:p1, m2:p2,…};

[0230] Among them, m i Indicates the type of material, p i Indicates the proportion.

[0231] It should be noted that the variables are explained as follows:

[0232] S30, including the following variables:

[0233] R represents the weak area patch set;

[0234] r i represents the i-th weak piece;

[0235] AllowableStress represents the allowable stress;

[0236] MTable represents the material proportion mapping table;

[0237] MatVector represents the material vector, which indicates the proportion of each material;

[0238] m i , p i Represents the i-th material and the corresponding proportion.

[0239] S40, obtaining a penetrating ultrasonic signal of the weak area, where the penetrating ultrasonic signal is emitted by an ultrasonic generator disposed on one side of the weak area and received by an ultrasonic sensor disposed on another side of the weak area, and performing preprocessing of noise reduction, filtering and enhancement on the penetrating ultrasonic signal to obtain a preprocessed signal;

[0240] This step uses ultrasonic signals to analyze the internal quality of the weak area. The specific implementation is as follows:

[0241] (1) System installation;

[0242] Set up a transmitting probe Te and a receiving probe Tr on the surface of the weak area:

[0243] Loc(T e ,Surface(R)),Loc(T r , Surface(R));

[0244] (2) Signal acquisition;

[0245] Transmit ultrasonic pulses and collect penetration signals:

[0246] X(t)=Emit(T e , Signal);

[0247] Y(t)=Receive(T r , X(t));

[0248] Where X(t) and Y(t) are the transmitted and received signals respectively. Repeat the acquisition N times.

[0249] (3) Signal processing;

[0250] 1. Wavelet Noise Reduction:

[0251] Y′(t)=WaveletDenosing(Y(t));

[0252] 2. Bandpass filtering:

[0253] Y″(t)=Bandpass(Y′(t));

[0254] 3. Signal envelope detection:

[0255] E(t) = Envelope(Y″(t));

[0256] 4. Amplitude amplification:

[0257] S(t) = Amp(E(t));

[0258] Finally, the preprocessed signal S(t) is obtained.

[0259] (4) Feature analysis;

[0260] 1. Extract signal features from the frequency domain perspective.

[0261] 2. Quantitative correspondence between analytical characteristics and material defects.

[0262] It should be noted that the variables are explained as follows:

[0263] S40, including the following variables:

[0264] T e , T r Indicates the transmitting probe and the receiving probe;

[0265] X(t), Y(t) represent the transmitted and received signals;

[0266] Y′(t) represents the signal after wavelet denoising;

[0267] Y″(t) represents the signal after bandpass filtering;

[0268] E(t) represents the signal envelope;

[0269] S(t) represents the signal after amplitude amplification.

[0270] S50, extracting the characteristic matrix of the preprocessed signal and fusing it with the material vector to form a fusion matrix;

[0271] This step establishes an evaluation model between materials, signal characteristics and final stability. The specific implementation is as follows:

[0272] (1) Feature fusion;

[0273] 1. Signal feature vector:

[0274] F s ={f1, f2, ... f m};

[0275] 2.Material vector MatVector;

[0276] 3. Feature Fusion:

[0277] F = Concat(F s , MatVector);

[0278] (2) Sample preparation;

[0279] 1. Consider materials and loads as major factors.

[0280] 2. Prepare samples based on orthogonal array design. The two factors are:

[0281] Material level: M = {good, bad};

[0282] Load level: L = {low, medium, high};

[0283] Obtain a planned sample combination / condition matrix.

[0284] (3) Model sample acquisition;

[0285] For each sample condition, obtain:

[0286] The ultrasonic signal feature is F; the stability label obtained by actual measurement is S;

[0287] Construct sample sets (F1, S1), (F2, S2)...(F N , S N ).

[0288] (4) Model training;

[0289] Use GBDT to train the stability evaluation model.

[0290] Mdl = GBDT (SampleSet);

[0291] It should be noted that the variables are explained as follows:

[0292] S50, including the following variables:

[0293] F s represents the signal feature vector;

[0294] f i represents the i-th signal feature;

[0295] F represents the fused feature vector;

[0296] M, L represent the factor vectors of material level and load level;

[0297] SampleSet represents the model training sample set;

[0298] Mdl represents the trained GBDT evaluation model.

[0299] S60, using a pre-trained ultrasound signal-stability model, inputting a fusion matrix, and obtaining a stability index of the weak area;

[0300] Based on the established GBDT evaluation model, the stability index of components in weak areas is predicted. The specific process is:

[0301] (1) Model input;

[0302] Extract feature vector F from weak area.

[0303] (2) Index prediction;

[0304] SI = Predict(Mdl, F);

[0305] Where SI represents the predicted stability index.

[0306] (3) Result analysis;

[0307] Parse and visualize the prediction results.

[0308] Draw a stability cloud map and mark the weak locations.

[0309] It should be noted that the variables are explained as follows:

[0310] S60, including the following variants:

[0311] SI stands for the predicted stability index.

[0312] S70: If the stability index is lower than the stability index threshold, output a maintenance reminder to the user.

[0313] By comparing with the threshold, the stability is determined. The specific implementation is as follows:

[0314] (1) Setting the threshold;

[0315] The stability threshold is determined empirically:

[0316] TH=60;

[0317] (2) comparative judgment;

[0318] For each weak area i:

[0319]

[0320] Finally, the alarm mark set Alert = {Alert1, Alert2, ...} is obtained.

[0321] (3) Maintenance reminder;

[0322] Output alarm mark information to the user.

[0323] It should be noted that the variables are explained as follows:

[0324] S70, including the following variables: TH represents the stability judgment threshold; Alert represents the alarm mark set.

[0325] Specifically, the principle of the present invention is: the core of the present invention is to develop a matching ultrasonic detection and data analysis method to achieve in-depth structural and non-destructive monitoring, and effectively solve the contradiction between the existing technology that is difficult to balance the detection depth and quality:

[0326] (1) Beam steering technology is used to enable ultrasonic signals to propagate in a direction, avoiding surface noise interference and ensuring that they can penetrate deep into the internal structure for detection;

[0327] (2) Beamforming technology based on pulse phased arrays can accurately locate weak areas in the structure and clearly identify the physical location of the fault;

[0328] (3) Extract the dispersion characteristics of soft and hard tissues and establish a matching library between material types and characteristic spectra. Determine the integrity of the tissue based on the corresponding frequency components of the material;

[0329] (4) Integrate material characteristics and acoustic parameters, map them to the final safety rating, and use the GBDT model to correctly judge the structural state to the safety level;

[0330] The application of the above-mentioned technology conforms to the law of wave propagation, can realize the quantitative measurement of the internal quality of the structure, overcome the blind spots of existing technologies, and achieve the technical effect of non-destructive monitoring and high-precision assessment.

Claims

1. A house full life cycle operation and maintenance management system, characterized in that: Includes the following functional modules: User management module, used to handle the registration, login, authority allocation, and personal information management of house owners; The house archive module is used to store detailed information of each house, including building structure, design drawings, construction materials, and completion date; A repair and maintenance module is used to calculate the stability of the house and remind the user to carry out repair and maintenance activities on the house when the stability of the house is lower than the stability index threshold; Reporting and analysis module, used to generate comprehensive reports of the house, including usage status and maintenance history; Wherein, the maintenance module is used to perform the following steps: S10, establishing a three-dimensional model of the house based on the two-dimensional design drawing and the panoramic image of the house; S20, performing finite element analysis based on the three-dimensional model of the house, obtaining multiple areas of the house where the stress is greater than the stress threshold, and recording them as weak areas; S30, obtaining the building material and proportion of the weak area to form a material vector; S40, obtaining a penetrating ultrasonic signal of the weak area, wherein the penetrating ultrasonic signal is emitted by an ultrasonic generator disposed on one side of the weak area and received by an ultrasonic sensor disposed on another side of the weak area, and performing preprocessing of noise reduction, filtering and enhancement on the penetrating ultrasonic signal to obtain a preprocessed signal; S50, extracting the characteristic matrix of the preprocessed signal and fusing it with the material vector to form a fusion matrix; S60, using a pre-trained ultrasound signal-stability model, inputting the fusion matrix, and obtaining a stability index of the weak area; S70: If the stability index is lower than the stability index threshold, output a maintenance reminder to the user.

2. A housing full life cycle operation and maintenance management system according to claim 1, characterized in that: The specific steps of S10 are: Collect two-dimensional design drawings of the house, including floor plans, elevations, and sections; Take panoramic images of the real house scene to obtain indoor and outdoor panoramic images; Establish a 3D reconstruction algorithm, use structured light and SLAM technology, use the design drawing as the framework, and combine the panoramic image details to reconstruct the 3D point cloud; Poisson surface reconstruction and Ball Pivoting algorithm are used to convert the point cloud into a fine mesh model and establish a three-dimensional model of the house.

3. A housing full life cycle operation and maintenance management system according to claim 2, characterized in that: The specific steps of S20 are: The Delaunay triangulation method is used to mesh the three-dimensional model; According to the properties of the material in each direction, determine whether the material is isotropic or anisotropic, and establish a unified structural model of the mixed material. If the material has the same properties in each direction, it is an isotropic material; if the material has different properties in each direction, it is an anisotropic material. Set displacement constraints, support constraints, static loads, and dynamic loads; Gaussian elimination method and pursuit method are used to solve the stiffness equation and calculate the stress-strain distribution; According to the calculated stress-strain distribution, an area where the stress is greater than the stress threshold is obtained and recorded as a weak area.

4. A housing full life cycle operation and maintenance management system according to claim 3, characterized in that: The specific steps of S30 are: Based on the three-dimensional model, extract the components in the weak area and perform segmentation and classification; Querying an engineering database to obtain a material list of the weak area components, wherein the engineering database contains information on material types, specifications, and performance of all components; The usage amount or proportion of the component materials in the weak area is counted, a component-material mapping relationship table is established, and a material vector is formed.

5. A housing full life cycle operation and maintenance management system according to claim 4, characterized in that: The specific steps of S40 are: Ultrasonic transmitting and receiving probes are respectively installed on both sides of the weak area; Set the frequency and pulse parameters to collect penetration signals; Performing noise reduction processing, the noise reduction processing methods include wavelet transform and median filtering; Use bandpass filtering to filter the signal; amplify the envelope to enhance the signal; Extract signal features to interpret the internal quality of the material.

6. A housing full life cycle operation and maintenance management system according to claim 5, characterized in that: The specific steps of S50 are: Extract the time and frequency domain features of the ultrasonic signal to construct a matrix; Perform PCA or LASSO feature selection; Concatenate material vectors and feature matrices to integrate new feature expressions; Prepare samples with different degradation levels and collect feature matrices and true value samples; Establish a structural status to stability assessment model.

7. A housing full life cycle operation and maintenance management system according to claim 6, characterized in that: The specific steps of S60 are: The evaluation model is trained using the GBDT algorithm; Preprocess the fusion feature matrix; Input the model to predict the stability index of the weak area; Denormalize to get the actual index value.

8. A housing full life cycle operation and maintenance management system according to claim 7, characterized in that: The specific steps of S70 are: Preset stability index threshold; comparing the predicted index with the stability index threshold; If the stability index is lower than the threshold, a maintenance suggestion is generated to prompt the user; Update processing status to prevent duplicate reminders.

9. A housing full life cycle operation and maintenance management system according to claim 1, characterized in that: The user management module is used to perform the following: Receive and process registration information from property owners, including identity verification and information entry; Receive and verify user login requests, ensure the legitimacy of user identities, and use authentication algorithms, including hash algorithms, to securely store and verify user passwords; Automatically or manually assigning permissions based on user roles and levels, including using role-based access control or attribute-based access control algorithms; Use CRUD operations to manage user information, where CRUD operations include create, read, update, and delete.

10. A housing full life cycle operation and maintenance management system according to claim 1, characterized in that: The reporting and analysis module is used to perform the following: Obtain detailed information about the house and use the historical database, process the data through data aggregation and visualization algorithms, use statistical methods to process and summarize, and generate a comprehensive report on the house; Record the maintenance history of each house, including maintenance time, maintenance content, and maintenance personnel information; Conduct in-depth analysis of the generated housing comprehensive report through machine learning algorithms, including statistical analysis, trend analysis, and predictive models. Machine learning algorithms include linear regression and time series analysis. Machine learning algorithms are used to identify patterns and predict future trends.

Images