Building outer wall damage detection method based on infrared thermal imaging

Abstract

The invention discloses a building outer wall damage detection method based on infrared thermal imaging, relates to the technical field of building detection, and provides the following scheme: obtaining structure data, influence data and prediction time of the current surface of a building; inputting the structural data into a pre-constructed degradation correlation formula to obtain a predicted heat conduction rate; inputting the influence data and the prediction time into an aging influence factor coefficient formula, and calculating to obtain an aging influence factor coefficient; performing aging influence factor coefficient correction on the predicted heat conduction rate to obtain a corrected predicted heat conduction rate; comparing the corrected predicted heat conduction rate with a material standard heat conduction rate to obtain an aging coefficient; and obtaining the danger level of the target building according to the aging coefficient and a preset comparison table representing the mapping relation between the aging coefficient and the danger level. And a degradation correlation formula and an aging influence factor coefficient formula are used to clearly quantify the comprehensive influence of environmental factors on material aging, so that the evaluation is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of building detection, and in particular to a method for detecting damage to building exterior walls based on infrared thermal imaging. Background Art

[0002] Infrared thermal imagers are widely used in non-destructive testing of building exterior walls. Based on the law of thermal radiation and the differential equation of heat conduction, they can measure the temperature distribution on the surface of an object at a long distance, non-contact, in real time, and quickly, and display the temperature difference in the form of a visual image, thereby achieving intuitive and accurate detection of defects and damage on the building's exterior walls.

[0003] Traditional methods (such as infrared thermal imaging detection) can only obtain the apparent damage data of the building at the current moment (such as crack width and hollow area), and cannot dynamically predict future aging trends by combining environmental variables (temperature, humidity, corrosive medium concentration, etc.) with the time dimension; although there are a few studies on the prediction of future aging trends, existing aging models usually simplify environmental variables into constant values ​​or empirical coefficients, and do not dynamically integrate real-time influencing data (such as acid rain frequency and number of freeze-thaw cycles), resulting in the problem of inaccurate aging predictions. Summary of the Invention

[0004] In view of the existing problems, the present invention provides a method for detecting damage to building exterior walls based on infrared thermal imaging.

[0005] In order to achieve the above object, the technical solution of the present invention is as follows: In a first aspect, the present application discloses a method for detecting damage to a building exterior wall based on infrared thermal imaging, comprising the following steps: Obtain the current surface structural data, impact data and predicted time of the building; Input the structural data into the pre-built degradation correlation formula to obtain the predicted heat transfer rate; Input the impact data and predicted time into the aging impact factor coefficient formula to calculate the aging impact factor coefficient; The predicted heat conduction rate is corrected by the aging influence factor coefficient to obtain the corrected predicted heat conduction rate The aging coefficient is obtained by comparing the corrected predicted thermal conductivity with the standard thermal conductivity of the material; The hazard level of the target building is obtained according to the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the hazard level.

[0006] Secondly, this application introduces a building exterior wall damage detection system based on infrared thermal imaging, which includes the following modules: The building exterior wall aging prediction system is characterized in that the system applies the aforementioned building exterior wall damage detection method based on infrared thermal imaging, and the system includes: Data acquisition module, which is used to obtain the current surface structural data, impact data and prediction time of the building in real time; A data processing module is used to input structural data into a pre-built degradation correlation formula to obtain a predicted heat transfer rate; further used to input impact data and predicted time into an aging impact factor coefficient formula to calculate an aging impact factor coefficient; further used to correct the predicted heat transfer rate by the aging impact factor coefficient to obtain a corrected predicted heat transfer rate; further used to compare the corrected predicted heat transfer rate with a standard heat transfer rate of the material to obtain an aging coefficient; The data analysis module is used to obtain the danger level of the target building based on the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the danger level.

[0007] In a third aspect, the present invention provides a computing device, comprising: Memory, used to store programs; The processor is used to execute computer-executable instructions, which, when executed by the processor, implement the steps of the aforementioned method for detecting damage to building exterior walls based on infrared thermal imaging.

[0008] In a fourth aspect, the present invention provides a computer-readable storage medium, comprising: when the program is executed by a processor, the steps of implementing the aforementioned method for detecting damage to building exterior walls based on infrared thermal imaging are implemented.

[0009] Compared with the prior art, the present invention has the following beneficial effects: 1. This application uses a degradation correlation formula to predict thermal conductivity, linking structural data with thermal conductivity, enabling assessments based on physical changes rather than subjective experience. The aging impact factor coefficient formula clearly quantifies the combined impact of environmental factors on material aging, making the assessment more accurate. Comparing the corrected predicted thermal conductivity rate with the material's standard thermal conductivity rate yields a quantitative aging coefficient, providing an objective and measurable indicator of degradation. 2. This application incorporates the time factor into the aging impact factor coefficient formula for calculation, which makes the evaluation result not a static description of the current status, but a prediction of the degree of aging at a specific point in the future. Assessing the hazard level based on the prediction results can identify potential structural safety hazards in advance, providing a valuable time window for preventive maintenance, reinforcement or evacuation decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The disclosure of the present invention is described with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. In the drawings, the same reference numerals are used to refer to the same components. Among them: Figure 1 This is a flow chart of a building exterior wall damage detection method based on infrared thermal imaging introduced by the present invention; Figure 2 Flowchart for constructing the degradation mapping formula; Figure 3 A flowchart for constructing the coefficient formula of aging impact factors; Figure 4 Flowchart of the building exterior wall aging prediction system. DETAILED DESCRIPTION

[0011] It is easy to understand that according to the technical solution of the present invention, without changing the essential spirit of the present invention, a person skilled in the art can propose a variety of interchangeable structural modes and implementation modes. Therefore, the following specific embodiments and drawings are only exemplary descriptions of the technical solution of the present invention and should not be regarded as the entire invention or as a limitation or restriction of the technical solution of the present invention.

[0012] In the prior art, infrared thermal imaging is only used when problems arise with building aging, and is not used to predict the aging of existing buildings.

[0013] In order to solve the above problems, Figure 1 As shown, the present invention proposes a building exterior wall damage detection method based on infrared thermal imaging, comprising the following steps: S1. Obtain the current surface structural data, impact data and predicted time of the building; S2. Input the structural data into a pre-established degradation correlation formula to obtain a predicted heat transfer rate; input the impact data and predicted time into the aging impact factor coefficient formula to calculate the aging impact factor coefficient; S3. Correcting the predicted heat transfer rate by the aging influence factor coefficient to obtain a corrected predicted heat transfer rate; S4. comparing the corrected predicted thermal conductivity with the standard thermal conductivity of the material to obtain an aging coefficient; S5. Obtain the danger level of the target building based on the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the danger level.

[0014] Regarding step S1. Obtain the structural data, impact data, and predicted time of the building's current surface.

[0015] Structural data include temperature field distribution data on the building surface, ambient temperature and humidity near the building, standard thermal conductivity rate of building materials, spatial data of building porosity, and spatial data of salt crystallinity; impact data include historical annual average number of freeze-thaw cycles, historical annual average acid rain frequency, and historical annual extreme high temperature duration data; the prediction time is in years.

[0016] The method to obtain data is as follows: Infrared thermal imaging is used to obtain temperature distribution data on the building surface, identifying localized low-temperature evaporation zones and high-temperature dehydration areas. Temperature and humidity sensors are used to obtain real-time time-series data on ambient temperature and humidity. Ultrasonic tomography is used to quantify the pore connectivity attenuation curve from the surface to a depth of 50 mm, generating spatial porosity data. Ion-sensitive electrodes are embedded at pre-set grid points to obtain salt crystallinity data. The existing building material thermal conductivity database is used to obtain the benchmark temperature transfer rate of the building material. Historical meteorological data is extracted for the historical average annual freeze-thaw cycles, acid rain frequency, and duration of extreme high temperatures in the area where the building is located. The required building forecast time is then obtained.

[0017] Regarding step S2. The function of the degradation correlation formula is to mathematically correlate the building heat transfer rate with the spatial data of material porosity and salt crystallinity.

[0018] like Figure 2 As shown, the method for establishing the degradation correlation formula includes the following steps: Synchronously collect structural data on the building surface.

[0019] (1) Data acquisition stage Infrared thermal imaging cameras were used to obtain temperature distribution data on the building surface, identifying localized low-temperature evaporation zones and high-temperature dehydration areas. Temperature and humidity sensors were used to obtain real-time time-series data on ambient temperature and humidity. Ultrasonic tomography was used to quantify the pore connectivity attenuation curve from the surface to a depth of 50 mm, obtaining spatial data on the material's porosity. Ion-sensitive electrodes were embedded at pre-set grid points to obtain spatial data on salt crystallinity. The existing building material thermal conductivity database was used to obtain the benchmark temperature transfer rate for building materials.

[0020] (2) Data processing stage The obtained temperature field distribution data of the building surface, the time series data of the ambient temperature and humidity, the spatial data of the material porosity, and the spatial data of the salt crystallinity are input into the multivariate correlation model, and the temperature and humidity time series fluctuation coefficient is further introduced. (in is the standard deviation, is the mean value), construct the four-element coupling equation:

[0021] in, are the material property constants calibrated by the particle swarm optimization algorithm, is the spatial data of the material porosity, is the salt crystallinity spatial data, is the ambient temperature, is the ambient humidity.

[0022] (3) Data output stage This equation is the degenerate correlation formula.

[0023] The construction of the multivariate association model includes the following steps: (1) Data acquisition stage Infrared thermal imaging cameras were used to obtain temperature distribution data on the building surface, identifying localized low-temperature evaporation zones and high-temperature dehydration areas. Temperature and humidity sensors were used to obtain real-time time series data on ambient temperature and humidity. Ultrasonic tomography was used to quantify the pore connectivity attenuation curve from the surface to a depth of 50 mm, obtaining spatial data on material porosity. Ion-sensitive electrodes were embedded at pre-set grid points to obtain spatial data on salt crystallinity. The benchmark temperature transfer rate of building materials was obtained using an existing database of thermal conductivity of building materials.

[0024] (2) Data processing stage Step 1: Data Preprocessing The infrared temperature field distribution data and porosity spatial data are combined through the spatial registration unit and salt crystallinity spatial data Align to the same coordinate system; the spatial registration unit uses the Kriging interpolation algorithm to convert the discrete porosity data and salt crystallinity data Resample to spatial grid data that matches the resolution of the infrared temperature field. Use the time series alignment unit to synchronize the ambient temperature and humidity data with the temperature field acquisition timestamp.

[0025] Step 2: Constraints Under the conditions of satisfying the dynamic response correlation engine and multi-parameter coupling mechanism, the original data is converted into model parameters to obtain a multivariate correlation model.

[0026] The establishment of a dynamic response correlation engine includes the following steps: a. Thermal insulation performance diagnosis: By intercepting the sudden rise / fall events of the ambient temperature (change slope > 5°C / h), the delay time required for the surface temperature to reach 90% of the steady state is calculated to establish a negative correlation between the delay time and the porosity; If the delay time is shortened by more than a threshold (e.g., 20%), it is determined that the thermal insulation performance has deteriorated; The criteria for determining the degraded area mark are:

[0027] in: is the calculated predicted value of heat transfer rate, is the measured conduction velocity, is the material degradation judgment threshold.

[0028] b. Analysis of salt crystallization area: When the ambient humidity is continuously >80%, the expansion rate of the surface low-temperature zone is tracked. If the rate is proportional to the growth rate of salt concentration, it is marked as a salt non-crystallization zone; During the drying period (humidity <40%), the spatial offset between the coordinates of the salt concentration peak and the highest surface temperature point was detected. When the offset was >10 cm, it was determined to be a salt crystallization area.

[0029] The output function of the heat transfer anomaly warning report is:

[0030] Where: It is the safety threshold of salt crystallization; Representing coordinates Triggering warning The multi-parameter coupling mechanism includes the following: a. Damage conditions in salt crystallization area: When the spatial overlap between the region with porosity > 25% and the region with salt concentration > 0.6% is ≥ 70%, the periodic salt crystallization zone damage mechanism is activated; b. Introduce time dimension variables: Comparing the porosity monitoring data of the same location over the years, if the pore expansion rate is >2% / year for three consecutive years, the area will be marked as an accelerated degradation zone; When salt is detected to form crystal blockage in the pores, the critical salt concentration threshold of the area is automatically increased by 30%.

[0031] When local The measured value deviates from the model predicted value by more than the threshold When the water level drops below 1000 m, an alarm signal of abnormal porosity or salt crystallization failure is triggered.

[0032] Step 3: Create a temperature conduction rate calculation unit The temperature field distribution data of the building surface and the time series data of the ambient temperature and humidity are obtained through the heat conduction formula to obtain the real-time building material temperature transfer rate, and the local heat conduction rate is solved based on the time gradient and spatial gradient of the temperature field data. :

[0033] in, is the temperature field, For time, is the spatial temperature gradient.

[0034] Step 4: Build the dataset The data set is constructed by combining the spatial data of building material temperature transfer rate, ambient temperature and humidity, material porosity, and salt crystallinity data. The data set is constructed based on the correlation parameter comparison table of material porosity-salt crystallinity-heat conduction rate.

[0035] The associated parameter comparison table is expressed in the following matrix form:

[0036] in, is the discrete porosity sampling value, , is the discrete salt crystallinity sampling value, , is the ambient temperature and humidity combination, , The heat transfer rate predicted by the model.

[0037] Step 5: Dataset Processing The IQR (Interquartile Range) method was used to remove outliers from the data set and perform outlier processing. After removing outliers, each data point was independently normalized using the Z-score.

[0038] Augment the data by adding Gaussian noise of a specified signal-to-noise ratio and expanding the dataset by linear interpolation.

[0039] Based on the random forest algorithm, the preprocessed data set is screened to obtain the processed key and the interference features are extracted.

[0040] The data is weighted to reduce the mean square error (MSE), and the dataset is optimized and trained through a fully convolutional network (FCN).

[0041] The model stability was evaluated by multi-round cross-validation of the key indicator (wavelength mean absolute error) error value.

[0042] (3) Data output stage Constructing a multivariate association model: Porosity , salt crystallinity 、Ambient temperature and humidity ( ) as input variables and output a multivariate association model:

[0043] (in is the model residual, function obtained through support vector machine or neural network training); The aging impact factor formula takes into account the impact of historical environmental data on building materials. By adjusting the aging impact factor coefficient, the resulting aging coefficient maintains high accuracy even when predicting changes over time, reducing detection errors caused by such changes.

[0044] like Figure 3 As shown in Figure 2, the construction of the aging impact factor coefficient formula includes the following steps: (1) Data acquisition stage The impact data of the area are collected through the local historical meteorological database, including the historical annual average number of freeze-thaw cycles, the historical annual average frequency of acid rain, and the historical annual average duration of extreme high temperatures; (2) Data processing stage Step 1: Data Normalization Normalize and calibrate the obtained data, and convert the obtained data into dimensionless parameters in the range of 0-1; Annual freeze-thaw cycles Perform normalization transformation to obtain , divide the measured freeze-thaw times by the preset climate zone benchmark value. The benchmark value is set according to the frigid, temperate, and subtropical zones: ; Annual acid rain frequency Convert to probability value , convert the acid rain frequency percentage to a scale value between 0 and 1: ; The duration of extreme high temperatures throughout the year Perform normalization transformation to obtain , convert the number of days with high temperature into the proportion of days in the whole year, the conversion formula is: ; in, is the number of freeze-thaw cycles per year, The maximum freeze-thaw cycle number benchmark value for the preset area (e.g. 50 for the frigid zone, 80 for the temperate zone, and 100 for the subtropical zone). is the annual acid rain frequency, The duration of extreme high temperatures in a year.

[0045] Step 2: Physical constraint setting: Set constraints for symbolic regression based on the material aging mechanism: Freeze-thaw damage The power index >0, representing the nonlinear characteristics of damage accumulation; High temperature damage Attenuation coefficient k >0, ensuring that the damage increases monotonically with exposure time; The weight coefficient satisfies the normalization constraint w 1+ w 2+ w 3=1 and wi ≥0.

[0046] Step 3: Symbolic regression execution: Adopting the tree structure coding formula, iterative optimization is carried out according to the following process: (a) Initialization: Randomly generate a population of formulas, the operator set is {+, ×, exp, power}, and the operands are { Nn , An , Hn ,constant}; (b) Evaluation: Calculating fitness ,in is the mean square error, is the constraint satisfaction factor (when all constraints are satisfied =1, otherwise =0.1); (c) Selection and genetic operations: select high fitness individuals through roulette, perform subtree crossover and node mutation; (d) Iteration termination: When the optimal formula Below threshold Or it terminates when it reaches 1000 generations.

[0047] Step 4: Formula Refinement Numerically optimize the optimal formula for symbolic regression output: Use gradient descent with L2 regularization to fine-tune the coefficients { w 1, w 2, w 3, p , k}; Regularization strength λ =0.01, suppressing overfitting and ensuring the physical rationality of the coefficients.

[0048] (3) Data output stage The coefficient formula of the aging influencing factor is obtained by combining symbolic regression with physical constraints to ensure physical rationality, and the coefficient accuracy is improved through linear compensation and constraint optimization.

[0049] Aging impact factor coefficient formula:

[0050] Added a time correction parameter to the company's refined formula.

[0051] Output the final aging impact factor coefficient formula:

[0052] in, is the dimensionless parameter of the annual freeze-thaw cycles, is the dimensionless parameter of acid rain frequency, is the dimensionless parameter of the duration of extreme high temperature, 、 、 is the weight coefficient, and the weight constraint is , power index Characterize the nonlinear characteristics of freeze-thaw damage, attenuation coefficient is the high temperature damage rate, .

[0053] The weight distribution is dynamically adjusted according to the type of building materials; the power exponent and attenuation coefficient need to be achieved through a combination of accelerated aging tests and inversion of actual engineering data.

[0054] The above describes in detail the details of each method step for constructing the aging impact factor coefficient model. The following specifically describes an application scenario of this embodiment: Data for concrete buildings in a temperate region in 2023: Freeze-thaw cycles: 52 (temperate benchmark: 80 → normalized value: 0.65) Acid rain frequency: 22% (normalized value 0.22) High temperature duration: 68 days (annual percentage 0.186) Processing flow: Assume the power exponent is 1.8 and the attenuation coefficient is 3.2 Freeze-thaw damage: 0.5×(0.65)^1.8=0.5×0.47=0.235 Acid rain damage: 0.3×0.22=0.066 High temperature damage: 0.2×(1-e^{-3.2×0.186})=0.2×0.45=0.090 Time correction parameter: 0.5 Comprehensive aging impact factor coefficient: (0.235+0.066+0.090)×0.5=0.1955 Output result: The comprehensive aging impact factor coefficient is 0.1955.

[0055] The process of obtaining time correction parameters includes the following steps: (1) Data acquisition stage The data acquisition steps include: collecting the daily temperature time series data of the target area for consecutive historical years through the local historical meteorological database, with a minimum time span of 15 natural years; The freeze-thaw event status of a single day is determined based on the preset freeze-thaw temperature difference threshold, which is set as the daily maximum temperature ≥ 0℃ and the daily minimum temperature ≤ -3℃; The total number of annual freeze-thaw cycles is generated by aggregation according to the natural year, and the starting month and duration of each freeze-thaw period are recorded simultaneously.

[0056] (2) Data processing stage The data processing steps include: Step 1: Data reconstruction phase: The missing year data were supplemented by spatial weighted interpolation of neighboring meteorological stations, and abnormal fluctuation points were eliminated based on sliding median filtering; Step 2: Feature derivation phase: Extract the long-term trend component, seasonal fluctuation component and interannual variability intensity of the historical series; construct the correlation characteristics of climate driving factors, including: the annual mean of the El Niño index and the winter integral value of the Arctic Oscillation Index; Step 3: Model training phase: The reconstructed feature set is input into a bidirectional temporal encoder, and the cross-year dependencies are captured through the attention mechanism; the network weights are dynamically adjusted with the annual freeze-thaw time offset as the optimization target.

[0057] (3) Data output stage The data output steps include: (a) Execute freeze-thaw time pattern prediction for the next 1-5 natural years, and output the core parameter as the annual time adjustment coefficient , whose physical definition is:

[0058] in, is the ordinal number of the start date of the predicted annual freeze-thaw period (annual cumulative days), is the ordinal number of the day corresponding to the historical base year; Regarding step S3. The predicted heat transfer rate is corrected by the aging influence factor coefficient to obtain the corrected predicted heat transfer rate. The formula is:

[0059] in, To predict the heat transfer rate, is the aging impact factor coefficient.

[0060] Regarding step S4. S4. comparing the corrected predicted thermal conductivity with the standard thermal conductivity of the material to obtain an aging coefficient;

[0061] in, is the predicted heat transfer rate after correction, is the standard thermal conductivity of the material.

[0062] Regarding step S5. A static correspondence table between the discrete intervals of the aging coefficient and the hazard level is established, where each threshold boundary value is determined based on the critical point of the structural failure probability specified in the industry standards. The comparison table is as follows: Discretize the continuous aging coefficient interval [0,1] into preset danger levels, including: Level I (safety zone): aging coefficient ≤ 0.3, corresponding to a state with no significant structural risk; Level II (warning zone): 0.3<aging coefficient≤0.5, corresponding local components need to be monitored; Level III (hazardous area): 0.5<aging coefficient≤0.7, corresponding to the state of decreased overall structural reliability; Level IV (high-risk): Aging coefficient > 0.7, requiring immediate intervention; In addition, since this application mainly focuses on the aging of buildings, various conditions can affect the aging of buildings, such as water seepage. Therefore, when collecting data, if it is necessary to know whether a building has water seepage, the structural data collected by this application can be used for analysis.

[0063] For example, the spatial data of building porosity and spatial data of salt crystallinity in the structural data collected in step 1 of this application, because the water seepage area is prone to salt precipitation crystallization, and causes crack morphologies such as straight lines and grids. When the spatial data of building porosity and spatial data of salt crystallinity in a certain area are significantly different from those in other areas or are very different from the data collected historically, there is reason to suspect that there is water seepage in the area. It can be further verified by combining the temperature field distribution data on the surface of the building obtained by scanning with an infrared thermal imager. Therefore, if there is water seepage, the temperature of the water seepage area will be lower than the temperature of the non-water seepage area. If the temperature of the area is lower than other temperatures, it means that there is water seepage in the area. If the water seepage is serious, it can be observed from the appearance of the building. There will be discoloration, stains, water stains, wet spots, cracks, and even peeling, tilting or deformation of the wall.

[0064] The severity of water seepage is analyzed based on collected data and preset thresholds. The system categorizes water seepage into three levels: severe (widespread) with the risk of structural deformation, moderate (widespread dampness) affecting building durability, and minor (localized dampness) primarily causing surface degradation. For minor water seepage, the leaking area is monitored more frequently during the aging prediction process than other areas. For severe or moderate water seepage, the relevant departments are notified to initiate maintenance.

[0065] Since one of the main factors of water seepage is the existence of hollowing in buildings, the hollowing area is mainly manifested as local bulges or wrinkles on the surface. This area often shows obvious color difference due to changes in temperature and humidity, and is accompanied by secondary discoloration caused by water seepage. The edge tissue of the hollowing area is mostly in a stretched state, and the surrounding area is often accompanied by radial or annular cracks. The crack morphology includes various types such as linear, serrated and mesh. The hollowing area is often accompanied by derivative phenomena such as water moisture, structural damage and abnormal vibration. At the same time, it can be divided into three levels according to the size of the hollowing area: severe hollowing (>0.5m 2 ) There is a safety hazard of falling off, medium hollowing (0.1-0.5m 2 ) affects the appearance and waterproof performance, slight hollowing (<0.1m 2 ) has no direct impact on structural safety. Therefore, while determining whether a building has water seepage, it can also simultaneously determine whether the water seepage is caused by hollowing, thereby assisting maintenance personnel in accurately locating the damage.

[0066] Example 2 This embodiment introduces a building exterior wall damage detection system based on infrared thermal imaging, including: The data acquisition module 100 is used to obtain the current surface structural data, impact data and prediction time of the building in real time; The data processing module 200 is used to input the structural data into a pre-established degradation correlation formula to obtain a predicted heat transfer rate; further used to input the impact data and the predicted time into the aging impact factor coefficient formula to calculate the aging impact factor coefficient; further used to correct the predicted heat transfer rate by the aging impact factor coefficient to obtain a corrected predicted heat transfer rate; further used to compare the corrected predicted heat transfer rate with the material standard heat transfer rate to obtain the aging coefficient; The data analysis module 300 is used to obtain the danger level of the target building according to the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the danger level.

[0067] Example 3 This embodiment introduces a computing device, including: Memory, used to store programs; The processor is used to execute computer-executable instructions, which, when executed by the processor, implement a building exterior wall damage detection method based on infrared thermal imaging.

[0068] Example 4 This embodiment introduces a computer-readable storage medium storing a program. When the program is executed by a processor, a method for detecting damage to a building exterior wall based on infrared thermal imaging is implemented.

[0069] The storage medium proposed in this embodiment and the building exterior wall damage detection method based on infrared thermal imaging proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0070] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or in other words, it contributes to the existing technology. Parts may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disk, and include a number of instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0071] The technical scope of the present invention is not limited to the contents of the above description. Those skilled in the art can make various deformations and modifications to the above embodiments without departing from the technical idea of ​​the present invention, and these deformations and modifications should all fall within the protection scope of the present invention.

Claims

1. A building exterior wall damage detection method based on infrared thermal imaging, characterized in that: include: Obtain the current surface structural data, impact data and predicted time of the building; Inputting the structural data into a pre-built degradation correlation formula to obtain a predicted heat transfer rate; Input the impact data and predicted time into the aging impact factor coefficient formula to calculate the aging impact factor coefficient; Correcting the predicted heat transfer rate by the aging influence factor coefficient to obtain a corrected predicted heat transfer rate; The aging coefficient is obtained by comparing the corrected predicted thermal conductivity with the standard thermal conductivity of the material; The hazard level of the target building is obtained according to the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the hazard level.

2. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 1, characterized in that: The structural data include temperature field distribution data on the building surface, time series data of ambient temperature and humidity, porosity spatial distribution data, salt crystallinity spatial data, and building material benchmark temperature transfer rate; the impact data include historical annual average number of freeze-thaw cycles, acid rain frequency, and extreme high temperature duration data.

3. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 2, characterized in that: The method for establishing the degenerate correlation formula comprises the following steps: Obtain structural data on the building surface; The obtained structural data is input into the multivariate association model, and the temperature and humidity time series fluctuation coefficient is introduced , construct the four-element coupling equation and obtain the degenerate correlation formula: in, are the material property constants calibrated by the particle swarm optimization algorithm, is the spatial data of the material porosity, is the salt crystallinity spatial data, is the ambient temperature, is the ambient humidity.

4. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 3, characterized in that: The construction of the multivariate association model includes the following steps: aligning the structural data to the same coordinate system through a spatial registration unit; Establish a multivariable correlation model under the conditions of meeting the dynamic response correlation engine and multi-parameter coupling mechanism; constructing a data set based on the correlation between the structural data and the material porosity, salt crystallinity and thermal conductivity; Perform outlier processing, standardization, and data enhancement on the data set; Extract interference features based on random forest algorithm; The data is weighted to optimize the training process of the data set; Input the optimized trained data set into the multivariate association model to obtain the trained multivariate association model: in, is the model residual, function Obtained through support vector machine or neural network training.

5. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 1, characterized in that: The aging impact factor coefficient formula construction includes the following steps: Obtain impact data for the building's location; Normalize and calibrate the obtained data, and convert the obtained data into dimensionless parameters in the range of 0-1; Set constraints for symbolic regression based on the material aging mechanism: Adopt tree structure to encode formula, randomly generate formula population, calculate formula fitness, select high fitness individuals through roulette, perform subtree crossover and node mutation, and when the optimal formula When the value is lower than the threshold, it is terminated and the coefficient logarithm is optimized by using the gradient descent method with L2 regularization; The coefficient formula of aging impact factor is obtained by symbolic regression combined with physical constraints. Aging impact factor coefficient formula: Add the time correction parameter to the refined formula; Output the final aging impact factor coefficient formula: in, is the dimensionless parameter of the annual freeze-thaw cycles, is the dimensionless parameter of acid rain frequency, is the dimensionless parameter of the duration of extreme high temperature, 、 、 is the weight coefficient, and the weight constraint is , power index Characterize the nonlinear characteristics of freeze-thaw damage, attenuation coefficient is the high temperature damage rate, , is the time correction parameter, is the forecast time in years.

6. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 5, characterized in that: The obtained data are normalized and calibrated, and the method of converting the obtained data into dimensionless parameters in the range of 0-1 is as follows: Annual freeze-thaw cycles Perform normalization transformation to obtain : ; Annual acid rain frequency Convert to a scale value between 0 and 1 : ; Duration of annual extreme high temperatures Perform normalization transformation to obtain , the conversion formula is: ; in, It is the benchmark value of the maximum number of freeze-thaw cycles in the preset area.

7. The method for detecting damage to building exterior walls based on infrared thermal imaging according to claim 5, characterized in that: The process of obtaining the time correction parameters includes: Obtain the daily temperature time series data for the area where the building is located for consecutive years in history, determine the occurrence of a single-day freeze-thaw event based on the preset freeze-thaw temperature difference threshold, and generate a series of the total number of annual freeze-thaw cycles by aggregating them according to the natural year; Spatial weighted interpolation is used to fill in missing year data, and abnormal fluctuation points are eliminated based on sliding median filtering; Extract the long-term trend component, seasonal fluctuation component and interannual variability intensity of the historical series to construct the correlation characteristics of climate driving factors; The reconstructed feature set is input into a bidirectional temporal encoder, and the cross-year dependencies are captured through an attention mechanism. The network weights are dynamically adjusted based on the annual freeze-thaw time offset. Execute the prediction of freeze-thaw time pattern in the future natural year, and output the core parameter as the annual time adjustment coefficient , whose physical definition is: in, is the ordinal number of the start date of the predicted annual freeze-thaw period (annual cumulative days), It is the ordinal number of the day corresponding to the historical base year.

8. Building exterior wall damage detection system based on infrared thermal imaging, characterized in that: The system applies the building exterior wall damage detection method based on infrared thermal imaging according to any one of claims 1 to 7, and the system includes: Data acquisition module, which is used to obtain the current surface structural data, impact data and prediction time of the building in real time; a data processing module for inputting the structural data into a pre-established degradation correlation formula to obtain a predicted heat conduction rate; for inputting the impact data and the predicted time into an aging impact factor coefficient formula to calculate an aging impact factor coefficient; for correcting the predicted heat conduction rate by the aging impact factor coefficient to obtain a corrected predicted heat conduction rate; and for comparing the corrected predicted heat conduction rate with a standard heat conduction rate of the material to obtain an aging coefficient; The data analysis module is used to obtain the danger level of the target building based on the aging coefficient and a preset comparison table representing the mapping relationship between the aging coefficient and the danger level.

9. An electronic device, characterized in that: include: Memory, used to store programs; A processor is used to load the program to execute the steps of the building exterior wall damage detection method based on infrared thermal imaging as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a program, characterized in that: When the program is executed by a processor, the steps of the building exterior wall damage detection method based on infrared thermal imaging as described in any one of claims 1 to 7 are implemented.

Images