Method and system for detecting thermal insulation material of building external wall

Abstract

The invention relates to the technical field of material detection, and particularly discloses a building external wall thermal insulation material detection method and system.The method comprises the steps that standardized thermal excitation is applied to a standard thermal insulation material model subjected to heat conduction performance inspection through a thermal excitation device, and a thermal infrared image time sequence of the standard thermal insulation material model is synchronously collected; and performing time sequence analysis on the thermal infrared image time sequence of the thermal insulation material model by using a deep learning technology so as to construct thermal response characteristic representation of the thermal insulation material model. Furthermore, the same thermal excitation is performed on the to-be-tested materials in the same batch, the thermal response characteristic characterization of the to-be-tested materials is constructed, and similarity analysis is performed on the thermal response characteristic characterization of the to-be-tested materials and the thermal response characteristic characterization of the thermal insulation material model, so that the heat conduction performance of the to-be-tested materials is quickly and accurately evaluated. Compared with a traditional steady-state detection method, the method is based on unsteady-state thermal response characteristic analysis, the detection efficiency is remarkably improved, and the method is suitable for a large-scale quality detection scene of the building external wall thermal insulation material.

Description

Technical Field

[0001] The present application relates to the technical field of material testing, and more specifically, to a method and system for testing building exterior wall insulation materials. Background Art

[0002] With the increasing global emphasis on building energy conservation and carbon emission control, building exterior wall insulation systems have been widely used and developed as a key technology to improve the thermal performance of building envelope structures and reduce energy consumption. The performance of insulation materials, especially their thermal conductivity, directly determines the energy-saving effect and living comfort of buildings, and is also related to the durability and safety of building structures. Currently, in the field of performance testing of building exterior wall insulation materials, the standard method widely used in the industry is the steady-state hot plate method. This method establishes a stable temperature difference on both sides of the material, waits for the internal heat flow to reach a stable state, and then measures the heat flux density and temperature gradient to accurately calculate the thermal conductivity of the material.

[0003] However, the drawback of the steady-state hot plate method is its reliance on thermal stability. Building insulation materials, such as extruded polystyrene (EPS), are designed to hinder heat transfer, resulting in extremely low thermal diffusion rates and thermal conductivity coefficients typically around 0.04W / (m·K). Consequently, during testing, it takes a long time for the material to reach thermal stability, with a single test cycle often lasting several hours or even longer. This low testing efficiency makes it impractical to individually test all insulation materials in a project in terms of both economic and time costs. Typically, only small-batch sampling tests can be used, making it difficult for the test results to fully and truly reflect the performance uniformity of the entire batch of materials, leaving quality risks for construction projects.

[0004] Therefore, an optimized detection method and system for building exterior wall insulation materials is expected. Summary of the Invention

[0005] In order to solve the above-mentioned technical problems, the present application is proposed. The embodiment of the present application provides a method and system for detecting building exterior wall insulation materials, which applies standardized thermal excitation to a standard insulation material sample that has been tested for thermal conductivity through a thermal excitation device, synchronously collects its thermal infrared image time series, and uses deep learning technology to perform time series analysis on the thermal infrared image time series of the insulation material sample to construct its thermal response characteristic representation. Furthermore, by applying the same thermal excitation to the materials to be tested in the same batch, its thermal response characteristic representation is constructed, and a similarity analysis is performed with the thermal response characteristic representation of the insulation material sample, thereby achieving a rapid and accurate evaluation of the thermal conductivity performance of the material to be tested. Compared with the traditional steady-state detection method, this method is based on non-steady-state thermal response characteristic analysis, which significantly improves the detection efficiency and is suitable for large-scale quality inspection scenarios of building exterior wall insulation materials.

[0006] According to one aspect of the present application, a method for detecting building exterior wall insulation materials is provided, comprising:

[0007] applying standardized thermal excitation to the insulation material sample using a thermal excitation device, and simultaneously collecting a time series of thermal infrared images of the insulation material sample at a predetermined sampling frequency;

[0008] Extracting thermal response features from a time series of thermal infrared images of the thermal insulation material sample to obtain a thermal response feature vector of the thermal insulation material sample;

[0009] Uploading the thermal conductivity coefficient of the thermal insulation material sample and the thermal response characteristic vector of the thermal insulation material sample to a cloud server to establish a thermal insulation material sample database;

[0010] Using the thermal excitation device to apply standardized thermal excitation to the thermal insulation material to be inspected, and simultaneously collecting a time series of thermal infrared images of the thermal insulation material to be inspected at the predetermined sampling frequency;

[0011] Extracting thermal response features from the time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected;

[0012] A similarity analysis is performed on the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample to determine whether the thermal insulation performance of the thermal insulation material to be tested is qualified.

[0013] According to another aspect of the present application, a building exterior wall insulation material detection system is provided, comprising:

[0014] a material thermal excitation and image acquisition module, for applying standardized thermal excitation to the thermal insulation material sample using a thermal excitation device, and simultaneously acquiring a time series of thermal infrared images of the thermal insulation material sample at a predetermined sampling frequency;

[0015] a thermal response feature extraction module, configured to extract thermal response features from a time series of thermal infrared images of the thermal insulation material sample to obtain a thermal response feature vector of the thermal insulation material sample;

[0016] A material template database establishment module is used to upload the thermal conductivity coefficient of the thermal insulation material template and the thermal response characteristic vector of the thermal insulation material template to the cloud server to establish a thermal insulation material template database;

[0017] A thermal excitation and image acquisition module for the material to be tested, configured to apply standardized thermal excitation to the thermal insulation material to be tested using the thermal excitation device, and simultaneously acquire a time series of thermal infrared images of the thermal insulation material to be tested at the predetermined sampling frequency;

[0018] a thermal response feature extraction module for the material to be detected, configured to extract thermal response features from a time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected;

[0019] The similarity analysis module is used to perform similarity analysis on the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample to determine whether the thermal insulation performance of the thermal insulation material to be tested is qualified.

[0020] Compared with the existing technology, the method and system for detecting building exterior wall insulation materials provided by this application uses a thermal excitation device to apply standardized thermal excitation to a standard insulation material sample that has been tested for thermal conductivity, synchronously collects its thermal infrared image time series, and uses deep learning technology to perform time series analysis on the thermal infrared image time series of the insulation material sample to construct its thermal response characteristic representation. Furthermore, by applying the same thermal excitation to the materials to be tested in the same batch, its thermal response characteristic representation is constructed, and a similarity analysis is performed with the thermal response characteristic representation of the insulation material sample, thereby achieving a rapid and accurate evaluation of the thermal conductivity performance of the material to be tested. Compared with the traditional steady-state detection method, this method is based on non-steady-state thermal response characteristic analysis, which significantly improves the detection efficiency and is suitable for large-scale quality inspection scenarios of building exterior wall insulation materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0022] Figure 1 This is a flow chart of a method for detecting building exterior wall insulation materials according to an embodiment of the present application.

[0023] Figure 2 This is a data flow diagram of a method for detecting building exterior wall insulation materials according to an embodiment of the present application.

[0024] Figure 3 This is a flowchart of sub-step S2 of the method for detecting building exterior wall insulation materials according to an embodiment of the present application.

[0025] Figure 4 This is a flowchart of sub-step S22 of the method for detecting building exterior wall insulation materials according to an embodiment of the present application.

[0026] Figure 5 This is a flowchart of sub-step S221 of the method for detecting building exterior wall insulation materials according to an embodiment of the present application.

[0027] Figure 6 This is a block diagram of a building exterior wall insulation material detection system according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0029] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0030] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0031] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0032] It is worth noting that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0033] In response to the technical problems described in the above background technology, this application proposes a method for detecting building exterior wall insulation materials, which uses a thermal excitation device to apply standardized thermal excitation to a standard insulation material sample that has been tested for thermal conductivity, and simultaneously collects its thermal infrared image time series. It also uses deep learning technology to perform time series analysis on the thermal infrared image time series of the insulation material sample to construct its thermal response characteristic representation. Furthermore, by applying the same thermal excitation to the materials to be tested in the same batch, its thermal response characteristic representation is constructed, and a similarity analysis is performed with the thermal response characteristic representation of the insulation material sample, thereby achieving a rapid and accurate evaluation of the thermal conductivity performance of the material to be tested. Compared with the traditional steady-state detection method, this method is based on non-steady-state thermal response characteristic analysis, which significantly improves the detection efficiency and is suitable for large-scale quality inspection scenarios of building exterior wall insulation materials.

[0034] Figure 1 This is a flow chart of a method for detecting building exterior wall insulation materials according to an embodiment of the present application. Figure 2 This is a data flow diagram of the building exterior wall insulation material detection method according to the embodiment of the present application. Figure 1 and Figure 2 As shown, the method for detecting building exterior wall insulation materials includes the following steps: S1, using a thermal excitation device to apply standardized thermal excitation to the insulation material sample, and at the same time collecting the time series of thermal infrared images of the insulation material sample at a predetermined sampling frequency; S2, performing thermal response feature extraction on the time series of thermal infrared images of the insulation material sample to obtain a thermal response feature vector of the insulation material sample; S3, uploading the thermal conductivity coefficient of the insulation material sample and the thermal response feature vector of the insulation material sample to a cloud server to establish an insulation material sample database; S4, using the thermal excitation device to apply standardized thermal excitation to the insulation material to be detected, and at the same time collecting the time series of thermal infrared images of the insulation material to be detected at the predetermined sampling frequency; S5, performing thermal response feature extraction on the time series of thermal infrared images of the insulation material to be detected to obtain a thermal response feature vector of the insulation material to be detected; S6, performing similarity analysis on the thermal response feature vector of the insulation material to be detected and the thermal response feature vector of the insulation material sample to determine whether the insulation performance of the insulation material to be detected is qualified.

[0035] In the above-mentioned building exterior wall insulation material detection method, the step S1 uses a thermal excitation device to apply standardized thermal excitation to the insulation material sample, and at the same time collects a time series of thermal infrared images of the insulation material sample at a predetermined sampling frequency. It should be understood that since traditional steady-state detection relies on the material to spontaneously reach thermal equilibrium, the low thermal conductivity of the insulation material (such as EPS is only 0.04W / (m·K)) causes the thermal equilibrium process to be extremely slow, while the material's non-steady-state thermal response process can reflect its thermophysical properties in a short time. Therefore, in order to actively stimulate the dynamic thermal behavior of the material and construct a reproducible detection benchmark, this application is based on the theory of non-steady-state heat conduction. A controllable heat source (such as a pulsed laser or a constant power heating plate) is used to apply standardized thermal excitation with strictly uniform energy, duration, and spatial distribution to the insulation material sample, using external energy input to stimulate the material to produce a transient thermal response, and using an infrared thermal imager to collect the thermal response data of the insulation material sample at a high frame rate in a short time, forming a time series of thermal infrared images of the insulation material sample, so that the thermal response data of the material can be used to characterize its thermal conductivity. In the specific implementation process, the thermal excitation device is set according to the preset parameters (such as power density 500W / m 2 , duration 10s), a standardized heat pulse with constant energy intensity is applied to the surface of the insulation material sample that has been tested by the traditional steady-state hot plate method and confirmed that its thermal conductivity meets the standard (such as the national standard GB / T10294). At the same time as this excitation occurs, an infrared thermal imager is synchronously used to continuously shoot the surface of the insulation material sample at a sampling frequency of ≥30Hz on the opposite side of the thermal excitation application surface, and continuously record the diffusion to attenuation process of the temperature field on the surface of the insulation material sample within a preset time range (greater than the duration of the heat pulse) from the moment the heat pulse is turned on, to generate a time series of thermal infrared images of the insulation material sample. In this way, the entire picture of the non-steady-state thermal response of the material can be captured in a short time, including the initial rapid response stage of heat diffusion and the subsequent slow decay process, avoiding several hours of waiting for thermal equilibrium.

[0036] Specifically, the thermal excitation device uses a controllable heat source, such as a constant power heating plate or a pulsed laser, which can apply heat input with uniform energy intensity and consistent time distribution to the surface of the thermal insulation material according to preset parameters. This form of heat input is different from the traditional steady-state test that relies on the natural heat conduction of the material to establish a temperature difference. Instead, it actively stimulates the transient thermal response behavior inside the material. Specifically, before the experiment begins, the thermal excitation device needs to be calibrated to ensure that its output power density is stable at the set value and can continue to act on the surface of the thermal insulation material within the specified time. This standardized thermal excitation process ensures the consistency of heat input conditions between different samples, thereby providing a physical basis for the comparability and repeatability of subsequent thermal response data.

[0037] At the same time, infrared thermal imagers, as key data acquisition devices, undertake the important task of capturing the dynamic changes in the material surface temperature field. To fully record the entire thermal diffusion process exhibited by the insulation material under stimulation, the infrared thermal imager requires a high sampling frequency and good spatial resolution. Simultaneously with thermal stimulation, the infrared thermal imager should continuously capture images from the side opposite the surface where the stimulation is applied, obtaining a complete image sequence of the temperature evolution of the back surface of the material over time. Because the insulation material itself has a low thermal conductivity and a slow thermal diffusion rate, it is necessary to continue recording the temperature decay process for a period of time after the thermal stimulation ends in order to fully capture the entire time sequence of the heat wave propagating from the stimulation surface to the corresponding area on the back surface. Typically, thermal stimulation lasts for several seconds, and the total recording time of the infrared thermal imager must cover the entire process from the moment the thermal stimulation is initiated to its decay, generally at least several times the duration of the thermal stimulation. The specific duration is adjusted according to the type and thickness of the material.

[0038] During the specific implementation process, the selection and processing of insulation material samples also play a vital role in the accuracy of the experimental results. The insulation material samples used to construct the standard database must be tested by the traditional steady-state hot plate method to confirm that their thermal conductivity meets the requirements of relevant national standards. In addition, the physical parameters of the sample, such as size, shape, and surface state, should be kept consistent to avoid the introduction of additional noise due to sample differences. Before the formal test, the sample needs to be placed in a constant temperature and humidity environment for a certain period of pretreatment to keep it in a stable initial thermodynamic state, thereby reducing the impact of environmental factors on the thermal response process. During the test, the actions between the thermal excitation device and the infrared thermal imager must be highly synchronized to ensure that the application of thermal excitation strictly corresponds to the starting moment of image acquisition to avoid distortion of the thermal response signal due to time misalignment.

[0039] It is worth noting that the parameter setting of standardized thermal excitation not only affects the thermal response intensity of the material, but also determines whether the infrared thermal imager can effectively capture sufficient thermal signal changes. If the thermal excitation intensity is too low, it may cause the surface temperature of the material to rise insignificantly, affecting the signal-to-noise ratio of the thermal response signal; if the intensity is too high, it may cause local thermal damage to the material, thereby changing its original thermal physical properties. Therefore, when determining the thermal excitation parameters, it is necessary to comprehensively consider the basic thermal physical parameters of the material, such as heat capacity, density, and thermal conductivity, and optimize and adjust them in combination with previous experimental data. At the same time, the detection range, focal length adjustment, background radiation compensation and other functions of the infrared thermal imager need to be reasonably configured to ensure that the collected image data truly reflects the temperature distribution on the surface of the material.

[0040] Furthermore, the stability of environmental conditions is also a factor that cannot be ignored. Because the thermal response process is sensitive to ambient temperature fluctuations, testing should be conducted in a closed, temperature-controlled space equipped with a temperature and humidity monitoring system to record changes in environmental parameters in real time during testing and correct them during data analysis. Furthermore, to prevent external light sources or electromagnetic interference from affecting infrared image quality, the test area should be shielded from external interference sources as much as possible, and filters or other optical methods should be used to enhance image contrast when necessary.

[0041] In the above-mentioned method for detecting building exterior wall insulation materials, the step S2 extracts thermal response features from the time series of thermal infrared images of the insulation material template to obtain a thermal response feature vector of the insulation material template. It should be understood that since the time series of the original thermal infrared image of the insulation material template contains a large amount of redundant spatial information and cannot directly characterize the thermal conductivity of the material. Therefore, in order to compress effective information from the time and space dimensions and establish a thermal response characteristic characterization of the insulation material template, this application is based on the powerful ability of deep learning technology in learning time and space data representation, and uses a deep learning model to deeply explore the dynamic evolution of the temperature distribution on the surface of the insulation material template over time, extract key thermal response features, and thus construct a thermal fingerprint of the insulation material template to obtain a thermal response feature vector of the insulation material template. Among them, Figure 3 Flowchart of sub-step S2 of the building exterior wall insulation material detection method according to the embodiment of the present application. Figure 3 As shown, the step S2 includes the steps of: S21, respectively extracting the thermal infrared features of each thermal infrared image of the thermal insulation material template in the time series of the thermal infrared image of the thermal insulation material template to obtain a time series of thermal infrared feature vectors of the thermal insulation material template; S22, performing double-layer information transfer encoding on the time series of the thermal infrared feature vectors of the thermal insulation material template to obtain the thermal response feature vector of the thermal insulation material template.

[0042] Specifically, the step S21 extracts the thermal infrared features of each thermal infrared image of the thermal insulation material sample in the time series of the thermal infrared images of the thermal insulation material sample to obtain a time series of thermal infrared feature vectors of the thermal insulation material sample. In a specific example of the present application, the step S21 includes: performing thermal infrared feature extraction based on the FPN model on each thermal infrared image of the thermal insulation material sample in the time series of the thermal infrared images of the thermal insulation material sample to obtain a time series of thermal infrared feature vectors of the thermal insulation material sample. Specifically, the present application takes into account that a single-scale convolutional neural network is difficult to simultaneously capture the microscopic hot spot diffusion and macroscopic temperature field evolution on the surface of the material, resulting in an incomplete description of the thermal distribution characteristics of the thermal insulation material sample. Therefore, in order to comprehensively extract the thermal distribution spatial characteristics of the material from each frame of the thermal infrared image, the present application is based on the multi-scale feature fusion technology in the field of computer vision, and introduces a feature pyramid network (FPN) to extract thermal infrared features of each frame of the thermal infrared image of the thermal insulation material sample, so as to comprehensively capture the detailed information and global distribution of the temperature field on the surface of the material. Specifically, the feature pyramid network (FPN) extracts temperature distribution features from a single-frame thermal infrared image of a thermal insulation material template layer by layer through a backbone network (such as ResNet-50), wherein shallow convolution is used to capture high-resolution local details (such as pixel-level temperature gradient changes around thermal excitation points), and deep convolution is used to extract low-resolution global patterns (such as the uniformity of temperature distribution of the entire material). A top-down path aggregation module is then used to fuse deep semantic information with shallow spatial information, and feature dimension compression is used to generate a thermal infrared feature vector of the thermal insulation material template with multi-scale perception capabilities. In this way, the temperature distribution information in each frame of the thermal infrared image of the thermal insulation material template is compressed into a compact representation that retains spatial semantics, forming the initial feature primitives for the temporal thermal response characteristic analysis of the thermal insulation material template, laying the foundation for subsequent temporal modeling.

[0043] Specifically, the step S22 performs double-layer information transfer encoding on the time series of the thermal infrared feature vector of the thermal insulation material template to obtain the thermal response feature vector of the thermal insulation material template. It should be understood that since the single-frame thermal infrared image features extracted by the FPN network only reflect the static thermal state of the thermal insulation material template, the thermal conductivity of the material is reflected in the spatiotemporal correlation of dynamic heat transfer (such as thermal wave penetration rate, temperature field evolution trajectory). Therefore, in order to effectively model the long-term dependencies of the thermal response process of the thermal insulation material template and enhance the expression ability of local key evolution features, the present application proposes a double-layer message transmission mechanism, which first performs local context interaction on the thermal infrared features of the thermal insulation material templates of adjacent frames in the time dimension, transfers the local temperature evolution pattern, and further aggregates the global spatiotemporal features to capture the overall dynamic trend in the thermal response process of the material, and obtains the final thermal response feature vector of the thermal insulation material template. Among them, Figure 4FIG. 1 is a flow chart of sub-step S22 of the method for detecting building exterior wall insulation materials according to an embodiment of the present application. Figure 4 As shown, the step S22 includes the steps of: S221, performing context enhancement perception based on an adaptive local neighborhood window on each thermal infrared feature vector of the thermal insulation material template in the time series of the thermal infrared feature vector of the thermal insulation material template to obtain a time series of local context enhancement coding vectors of the thermal infrared feature of the thermal insulation material template; S222, performing global time domain transfer coding on the time series of the local context enhancement coding vectors of the thermal infrared feature of the thermal insulation material template to obtain the thermal response feature vector of the thermal insulation material template.

[0044] Figure 5 Flowchart of sub-step S221 of the building exterior wall insulation material detection method according to the embodiment of the present application. Figure 5 As shown, the step S221 includes the steps of: S2211, determining the window size of the local context enhanced perception window of each thermal infrared feature vector of the thermal insulation material template based on the characteristic distribution of each thermal infrared feature vector of the thermal insulation material template in the time series of the thermal infrared feature vector of the thermal insulation material template; S2212, performing local neighborhood context enhanced perception on each thermal infrared feature vector of the thermal insulation material template based on all the thermal infrared feature vectors of the thermal insulation material template in the local context enhanced perception window of the thermal infrared feature vector of the thermal insulation material template to obtain the time series of the local context enhanced coding vector of the thermal infrared feature of the thermal insulation material template.

[0045] In a specific example of the present application, step S2211 is expressed as follows:

[0046]

[0047] Where exp(·) represents the exponential function operation with e as the base, h i and h j They represent the i-th and j-th thermal infrared feature vectors of the thermal insulation material sample in the time series of the thermal infrared feature vector of the thermal insulation material sample, ‖·‖ represents the calculation of the Euclidean norm, τ represents the thermodynamic factor, and s ij Indicates h i and h j The feature correlation between is the preset neighborhood, ∈ represents the main regularization term, γ represents the secondary regularization term, k represents the index of the vector, Represents a preset neighborhood The thermal infrared characteristic vector of the jth insulation material sample relative to h i The correlation weight coefficient, log2(·) represents the logarithmic function with base 2, ε i Represents vector hi The neighbor feature distribution entropy, w max Indicates the preset maximum window size, w i Indicates h i The window size of the local context enhanced perception window corresponding to the thermal infrared feature vector of the insulation material sample.

[0048] That is, based on the characteristic distribution of the thermal infrared feature vectors of each insulation material template in the time series, a perception window that can accurately capture the semantically relevant local environment is dynamically defined for each insulation material template thermal infrared feature vector, so that the window size is adapted to the semantic relevance strength of the insulation material template thermal infrared feature vector, thereby achieving a refined local context understanding of each insulation material template thermal infrared feature vector, avoiding the interference of semantically irrelevant information or the omission of key semantic information caused by a fixed window. In this way, through the adaptively determined window size, it is possible to effectively focus on the local context information that is most relevant to the semantics of the current insulation material template thermal infrared feature vector, thereby improving the accuracy of the entire thermal response feature extraction process in characterizing the thermal conductivity performance of the material.

[0049] In a specific example of the present application, step S2212 includes: inputting all thermal infrared feature vectors of the thermal infrared feature vector of the thermal insulation material sample in the local context enhancement perception window into the local neighborhood context enhancement fusion network based on the attention mechanism to obtain the thermal infrared feature local context enhancement encoding vector of the thermal insulation material sample, which is expressed as follows:

[0050]

[0051] Among them, softmax(·) represents the normalized exponential function, W q 、W k and W g represents the different weight parameter matrices in the local neighborhood context enhancement fusion network, (·) T represents the transpose of the vector, and d represents h i The characteristic dimension, α ij Indicates h j The corresponding attention weight factor, Sigmoid(·) represents the sigmoid activation function, Indicates h i The corresponding local context enhanced encoding vector of thermal infrared features of the insulation material sample.

[0052] That is, a local neighborhood context enhancement fusion network based on the attention mechanism is used to fuse the thermal infrared feature vectors of all insulation material templates within the local context enhancement perception window, and the weights are dynamically allocated through the attention mechanism to focus on the thermal response features with close semantic associations. The thermal infrared feature vectors of the insulation material templates at each time step are promoted to local context enhancement coding vectors of the thermal infrared features of the insulation material templates that contain rich local thermal response semantic information, and a time series of local context enhancement coding vectors of the thermal infrared features of the insulation material templates is generated, so as to accurately capture the local key features of the evolution of the material surface temperature field under thermal excitation, so that each local context enhancement coding vector of the thermal infrared features of the insulation material template can effectively absorb the key semantic information in the neighborhood.

[0053] In particular, here, since the thermodynamic factor τ, the main regularization term ∈ and the secondary regularization term γ are introduced in the determination of the window size of the local semantic enhancement perception window, the W g 、W q and W k For example, it is preferred that the quantization of deformation geometric features should also be considered to improve the contextual semantic embedding under the predetermined local semantic enhancement perception window. Based on this, in a preferred example of the present application, the step S2212 includes: first, the weight parameter matrix of the local neighborhood context enhancement fusion network based on the attention mechanism is optimized based on the weight self-adjustment guided by the local perception field to obtain an updated weight parameter matrix; then, based on the updated weight parameter matrix, all the thermal infrared feature vectors of the thermal infrared feature vector of the thermal insulation material template in the local context enhancement perception window of the thermal infrared feature vector of the thermal insulation material template are subjected to the context enhancement fusion based on the attention mechanism to obtain the thermal infrared feature local context enhancement coding vector of the thermal infrared feature of the thermal insulation material template.

[0054] Specifically, first, for the deformation association vector (τ,∈,γ), it is piecewise linearly interpolated to have the same length as the thermal infrared feature vector of the insulation material template, and then a one-dimensional convolution is performed to perform non-planar curvature transformation to obtain the deformation feature connection vector V α , and the weight matrix W g 、W q and W k As the mapping target, the curvature deformation contact response vector is obtained:

[0055] V α1 =V α W g +V α

[0056] V α2 =V α W q +V α

[0057] V α3 =V α W k +V α

[0058] Among them, V α Denotes the deformation feature connection vector, V α1 、V α2 and V α3 Represents the weight parameter matrix W g 、W q and W k The corresponding curvature deformation contact response vector.

[0059] Then, considering the direct correlation with the curvature deformation, the dominant mechanism of the weighted deformation curvature correlation can be simplified, that is, the high-order nonlinear strain under the non-intrinsic term can be ignored. Therefore, the above curvature deformation correlation vector is respectively correlated with the eigenspace basis vector of the weight parameter matrix, that is, V in the following formula: eg 、V eq and V ek Associate to modify each weight parameter matrix, expressed as:

[0060] W g ′=(V α1 T V eg )⊙W g

[0061] W q ′=(V α2 T V eq )⊙W q

[0062] W k ′=(V α3 T V ek )⊙W k

[0063] Among them, V eg 、V eq and V ek Respectively represent W g 、W q and W k The eigenspace basis vectors, ⊙ represents the dot multiplication operation, W′ g , W′ q and W′ k Respectively represent W g 、W q and W k The corresponding updated weight parameter matrix.

[0064] In this way, the local semantic embedding enhancement based on the weight matrix can take into account the introduction of deformation geometric feature effects in the determination of the window size of the local semantic enhancement perception window, so that when the context enhancement fusion based on the attention mechanism is performed, the context semantic embedding expression effect of the local context enhancement coding vector of the thermal infrared features of each insulation material template is enhanced based on the predetermined local semantic enhancement perception window.

[0065] More specifically, step S222 performs global time-domain transfer coding on the time series of the local context-enhanced coding vectors of the thermal infrared features of the thermal insulation material sample to obtain the thermal response feature vector of the thermal insulation material sample. In a specific example of the present application, step S222 includes: inputting the time series of the local context-enhanced coding vectors of the thermal infrared features of the thermal insulation material sample into an information global transfer coding network based on the Transformer architecture to obtain the thermal response feature vector of the thermal insulation material sample, which is expressed as follows:

[0066]

[0067] in, represents the time series of the local context enhanced encoding vector of the thermal infrared feature of the insulation material sample, and Respectively The first, second and i-th insulation material template thermal infrared feature local context enhanced encoding vector, Transformer(·) represents the Transformer architecture, v f Represents the thermal response eigenvector of the insulation material template.

[0068] That is, the self-attention mechanism of the Transformer architecture is used to perform global information interaction and dependency modeling on the time series of the local context enhanced coding vectors of the thermal infrared features of the insulation material templates, so that each local context enhanced coding vector of the thermal infrared features of the insulation material template can dynamically integrate the information of all other vectors in the sequence, thereby capturing the long-distance dependencies and complex evolution patterns of the thermal response features in the time dimension, and integrating the local semantic refinement with the global structural insight to form a thermal response feature vector of the insulation material template that can comprehensively characterize the thermal response characteristics of the material. It not only retains the thermal response detail features in the local context enhanced coding vector of the thermal infrared features of the insulation material template, but also incorporates the macroscopic evolution trend of the thermal infrared features in the entire time series, realizing a comprehensive characterization of the thermal conductivity performance of the material from local dynamics to global trends, and providing a feature benchmark with both detail accuracy and global vision for the subsequent evaluation of the thermal insulation performance of the material to be tested through similarity analysis.

[0069] In the above-mentioned method for detecting building exterior wall insulation materials, in step S3, the thermal conductivity coefficient and thermal response characteristic vector of the insulation material sample are uploaded to the cloud server to establish an insulation material sample database. It should be understood that this application takes into account that various types and thicknesses of insulation materials may be encountered in the project, with different thermal response characteristics, and the reference data of the insulation material sample needs to be called at different locations. Therefore, this application is further based on cloud computing and database technology to manage the benchmark information of all qualified samples by constructing an expandable and remotely accessible insulation material sample database. Specifically, for each insulation material sample processed by the above steps, the thermal conductivity coefficient measured by the traditional steady-state method and the thermal response characteristic vector of the insulation material sample generated based on the above method are bound as a data pair, uploaded to the cloud server, and stored in a structured insulation material sample database. The insulation material sample database is indexed and classified according to attributes such as material type, thickness, manufacturer, and production batch. In this way, when a specific material needs to be tested, the thermal response characteristic vector of the corresponding insulation material template can be easily retrieved and downloaded from the cloud as a comparison benchmark, thereby providing platform support for the large-scale application of material insulation performance testing methods.

[0070] In the above-mentioned method for detecting building exterior wall insulation materials, in step S4, the thermal excitation device is used to apply standardized thermal excitation to the insulation material to be detected, and at the same time, a time series of thermal infrared images of the insulation material to be detected is collected at the predetermined sampling frequency. Specifically, since the materials to be detected on the production line need to be compared and analyzed for thermal response behavior under completely consistent thermal boundary conditions with the insulation material template. Therefore, in order to ensure the comparability of the test results, this application is based on the principle of experimental condition reproduction, integrates automated thermal excitation tooling on the production line, triggers synchronous heating and image acquisition after the material is positioned by the conveyor belt, applies thermal excitation parameters (including heat source position, power, duration) and environmental temperature and humidity control that are exactly the same as those of the template to the insulation material to be detected, and simultaneously starts the infrared thermal imager to continuously capture the temperature change process of the surface of the insulation material to be detected with the same focal length, field of view angle and predetermined sampling frequency, so as to obtain a time series of thermal infrared images of the insulation material to be detected that can be compared with the template data. Based on this, through strict experimental condition control, it is helpful to eliminate the feature offset caused by differences in experimental conditions and ensure the reliability of the thermal response behavior similarity analysis.

[0071] In the above-mentioned method for detecting thermal insulation materials for building exterior walls, the step S5 is to extract thermal response features from the time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected. It should be understood that in order to ensure that the thermal response behavior characteristics of the thermal insulation material to be detected are compared with the template data in the same mathematical space to achieve isomorphic alignment of the feature space, the present application performs the same feature extraction process on the time series of thermal infrared images of the thermal insulation material to be detected as that of the thermal insulation material template, and directly deploys the above-mentioned deep learning model to the edge computing device (such as an industrial GPU industrial computer), and outputs a thermal response feature vector of the thermal insulation material to be detected with the same feature dimension as the template data through completely consistent feature pyramid hierarchical encoding and timing message passing operations, thereby ensuring that the thermal response behavior characteristics of all materials are mapped to a unified feature hyperplane, avoiding evaluation bias introduced by model inconsistency.

[0072] In the above-mentioned method for detecting thermal insulation materials for building exterior walls, the step S6 performs a similarity analysis on the thermal response characteristic vector of the thermal insulation material to be detected and the thermal response characteristic vector of the thermal insulation material template to determine whether the thermal insulation performance of the thermal insulation material to be detected is qualified. That is, by calculating the similarity score between the thermal response characteristic vector of the thermal insulation material to be detected and the thermal response characteristic vector of the thermal insulation material template in the feature space, the degree of similarity of the thermal response behaviors of the two is quantified, thereby evaluating whether the thermal insulation performance of the thermal insulation material to be detected meets the standards. Specifically, first, the thermal response characteristic vector of the thermal insulation material template with the same attributes as the thermal insulation material to be detected, such as type, thickness, manufacturer and production batch, is retrieved from the thermal insulation material template database stored in the cloud server as a comparison benchmark. Then, considering that in the vector space, the geometric distance between the characteristic vectors directly reflects the similarity of the thermal response behavior of the materials. Therefore, in order to quantitatively determine the consistency between the thermal insulation material to be tested and the qualified template, this application is based on the vector space metric principle, by calculating the cosine similarity between the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material template, and based on the comparison between the cosine similarity and the preset qualified threshold, determining whether the thermal insulation performance of the thermal insulation material to be tested is qualified. Among them, the preset qualified threshold is statistically derived based on a large amount of experimental data and engineering application experience, ensuring that it can effectively distinguish between qualified and unqualified materials and adapt to normal production fluctuations. If the cosine similarity is higher than or equal to the preset qualified threshold (for example, 0.95), the thermal insulation performance of the thermal insulation material to be tested is determined to be qualified; conversely, if the similarity is lower than the preset qualified threshold, its thermal insulation performance is determined to be unqualified. In this way, not only the efficiency of thermal conductivity detection of thermal insulation materials is improved, but also it helps to achieve large-scale quality monitoring of thermal insulation materials, providing strong support for quality control of thermal insulation materials.

[0073] In summary, a method for detecting building exterior wall insulation materials based on an embodiment of the present application is explained, which applies standardized thermal excitation to a standard insulation material sample that has been tested for thermal conductivity through a thermal excitation device, synchronously collects its thermal infrared image time series, and uses deep learning technology to perform time series analysis on the thermal infrared image time series of the insulation material sample to construct its thermal response characteristic representation. Furthermore, by applying the same thermal excitation to the materials to be tested in the same batch, its thermal response characteristic representation is constructed, and a similarity analysis is performed with the thermal response characteristic representation of the insulation material sample, thereby achieving a rapid and accurate evaluation of the thermal conductivity performance of the material to be tested. Compared with the traditional steady-state detection method, this method is based on non-steady-state thermal response characteristic analysis, which significantly improves the detection efficiency and is suitable for large-scale quality inspection scenarios of building exterior wall insulation materials.

[0074] Furthermore, a building exterior wall insulation material detection system is also provided.

[0075] Figure 6 FIG. 1 is a block diagram of a building exterior wall insulation material detection system according to an embodiment of the present application. Figure 6 As shown, the building exterior wall insulation material detection system 100 according to the embodiment of the present application includes: a material thermal excitation and image acquisition module 110, which is used to use a thermal excitation device to apply standardized thermal excitation to the insulation material sample, and at the same time acquire a time series of thermal infrared images of the insulation material sample at a predetermined sampling frequency; a thermal response feature extraction module 120, which is used to extract thermal response features from the time series of thermal infrared images of the insulation material sample to obtain a thermal response feature vector of the insulation material sample; a material sample database establishment module 130, which is used to upload the thermal conductivity coefficient and thermal response feature vector of the insulation material sample to a cloud server to establish a thermal insulation material sample database. a thermal excitation and image acquisition module 140 for the material to be detected, for applying standardized thermal excitation to the thermal insulation material to be detected using the thermal excitation device, and at the same time acquiring a time series of thermal infrared images of the thermal insulation material to be detected at the predetermined sampling frequency; a thermal response feature extraction module 150 for the material to be detected, for performing thermal response feature extraction on the time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected; a similarity analysis module 160, for performing similarity analysis on the thermal response feature vector of the thermal insulation material to be detected and the thermal response feature vector of the thermal insulation material template to determine whether the thermal insulation performance of the thermal insulation material to be detected is qualified.

[0076] Here, those skilled in the art will appreciate that the specific operations of each module in the above-mentioned building exterior wall insulation material detection system have been described in detail in the referenced examples. Figures 1 to 5 The above description has been introduced in detail in the description of the detection method of building exterior wall insulation materials, and therefore, its repeated description will be omitted.

[0077] The basic principles of the present invention have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in the present invention are merely illustrative and non-limiting, and should not be construed as necessarily possessed by each embodiment of the present invention. Furthermore, the specific details of the above embodiments are provided for illustrative purposes and to facilitate understanding, and are not intended to be limiting. These details do not necessarily limit the present invention to being implemented using these specific details.

[0078] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0079] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be encompassed therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0080] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0081] Finally, it should be noted that the above description has been provided for the purpose of illustration and description. In addition, the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to be limiting. Although the technical solutions may be modified or replaced with equivalents with reference to the preferred embodiments, they do not depart from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for detecting building exterior wall insulation materials, characterized in that: include: applying standardized thermal excitation to the insulation material sample using a thermal excitation device, and simultaneously collecting a time series of thermal infrared images of the insulation material sample at a predetermined sampling frequency; Extracting thermal response features from a time series of thermal infrared images of the thermal insulation material sample to obtain a thermal response feature vector of the thermal insulation material sample; Uploading the thermal conductivity coefficient of the thermal insulation material sample and the thermal response characteristic vector of the thermal insulation material sample to a cloud server to establish a thermal insulation material sample database; Using the thermal excitation device to apply standardized thermal excitation to the thermal insulation material to be inspected, and simultaneously collecting a time series of thermal infrared images of the thermal insulation material to be inspected at the predetermined sampling frequency; Extracting thermal response features from the time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected; A similarity analysis is performed on the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample to determine whether the thermal insulation performance of the thermal insulation material to be tested is qualified.

2. The method for detecting building exterior wall insulation materials according to claim 1, characterized in that: Performing a similarity analysis on the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample to determine whether the thermal insulation performance of the thermal insulation material to be tested is qualified, including: The cosine similarity between the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample is calculated, and based on the comparison between the cosine similarity and a preset qualified threshold, it is determined whether the thermal insulation performance of the thermal insulation material to be tested is qualified.

3. The method for detecting building exterior wall insulation materials according to claim 1, characterized in that: Extracting thermal response features from the time series of the thermal infrared image of the thermal insulation material sample to obtain a thermal response feature vector of the thermal insulation material sample includes: Respectively extracting thermal infrared features of each thermal infrared image of the thermal insulation material sample in the time series of the thermal infrared images of the thermal insulation material sample to obtain a time series of thermal infrared feature vectors of the thermal insulation material sample; Double-layer information transfer encoding is performed on the time series of the thermal infrared feature vector of the thermal insulation material sample to obtain the thermal response feature vector of the thermal insulation material sample.

4. The method for detecting building exterior wall insulation materials according to claim 3, characterized in that: Extracting thermal infrared features of each thermal infrared image of the thermal insulation material sample in the time series of the thermal infrared images of the thermal insulation material sample to obtain a time series of thermal infrared feature vectors of the thermal insulation material sample includes: Thermal infrared feature extraction based on the FPN model is performed on each thermal infrared image of the thermal insulation material sample in the time series of the thermal infrared images of the thermal insulation material sample to obtain a time series of thermal infrared feature vectors of the thermal insulation material sample.

5. The method for detecting building exterior wall insulation materials according to claim 4, characterized in that: Performing double-layer information transfer encoding on the time series of the thermal infrared feature vector of the thermal insulation material sample to obtain the thermal response feature vector of the thermal insulation material sample includes: Performing context-enhanced perception based on an adaptive local neighborhood window on each thermal infrared feature vector of the thermal insulation material sample in the time series of the thermal infrared feature vector of the thermal insulation material sample to obtain a time series of local context-enhanced coding vectors of the thermal infrared feature of the thermal insulation material sample; Global time domain transfer coding is performed on the time series of the local context enhanced coding vector of the thermal infrared feature of the thermal insulation material sample to obtain the thermal response feature vector of the thermal insulation material sample.

6. The method for detecting building exterior wall insulation materials according to claim 5, characterized in that: The method includes performing context-enhanced perception based on an adaptive local neighborhood window on each thermal infrared feature vector of the thermal insulation material sample in the time series of the thermal infrared feature vector of the thermal insulation material sample to obtain a time series of local context-enhanced coding vectors of the thermal infrared feature of the thermal insulation material sample, including: Determining the window size of the local context enhanced perception window of each thermal infrared feature vector of the thermal insulation material sample based on the characteristic distribution of each thermal infrared feature vector of the thermal insulation material sample in the time series of the thermal infrared feature vector of the thermal insulation material sample; Based on all the thermal infrared feature vectors of the thermal insulation material templates in the local context enhanced perception window of the thermal infrared feature vectors of each thermal insulation material template, local neighborhood context enhanced perception is performed on each thermal infrared feature vector of the thermal insulation material template to obtain a time series of the local context enhanced coding vector of the thermal infrared feature of the thermal insulation material template.

7. The method for detecting building exterior wall insulation materials according to claim 6, characterized in that: Based on all thermal infrared feature vectors of the thermal infrared feature vectors of the thermal insulation material templates in the local context enhanced perception window of the thermal infrared feature vectors of the thermal insulation material templates, local neighborhood context enhanced perception is performed on each thermal infrared feature vector of the thermal insulation material templates to obtain a time series of local context enhanced encoding vectors of the thermal infrared features of the thermal insulation material templates, including: All thermal infrared feature vectors of the thermal insulation material template in the local context enhancement perception window of the thermal infrared feature vector of the thermal insulation material template are input into the local neighborhood context enhancement fusion network based on the attention mechanism to obtain the local context enhancement encoding vector of the thermal infrared feature of the thermal insulation material template.

8. The method for detecting building exterior wall insulation materials according to claim 7, characterized in that: Inputting all thermal infrared feature vectors of the thermal insulation material sample in the local context enhancement perception window of the thermal infrared feature vector of the thermal insulation material sample into the local neighborhood context enhancement fusion network based on the attention mechanism to obtain the thermal infrared feature local context enhancement encoding vector of the thermal insulation material sample, including: Performing weight self-adjustment optimization based on local perception field guidance on the weight parameter matrix of the local neighborhood context enhancement fusion network based on the attention mechanism to obtain an updated weight parameter matrix; Based on the updated weight parameter matrix, all thermal infrared feature vectors of the thermal insulation material template in the local context enhancement perception window of the thermal infrared feature vector of the thermal insulation material template are subjected to context enhancement fusion based on the attention mechanism to obtain the local context enhancement encoding vector of the thermal infrared feature of the thermal insulation material template.

9. The method for detecting building exterior wall insulation materials according to claim 5, characterized in that: Performing global time domain transfer coding on the time series of the local context enhanced coding vector of the thermal infrared feature of the thermal insulation material sample to obtain the thermal response feature vector of the thermal insulation material sample, including: The time series of the local context enhanced coding vector of the thermal infrared feature of the thermal insulation material sample is input into the information global transfer coding network based on the Transformer architecture to obtain the thermal response feature vector of the thermal insulation material sample.

10. A building exterior wall insulation material detection system, characterized in that: include: a material thermal excitation and image acquisition module, for applying standardized thermal excitation to the thermal insulation material sample using a thermal excitation device, and simultaneously acquiring a time series of thermal infrared images of the thermal insulation material sample at a predetermined sampling frequency; a thermal response feature extraction module, configured to extract thermal response features from a time series of thermal infrared images of the thermal insulation material sample to obtain a thermal response feature vector of the thermal insulation material sample; A material template database establishment module is used to upload the thermal conductivity coefficient of the thermal insulation material template and the thermal response characteristic vector of the thermal insulation material template to the cloud server to establish a thermal insulation material template database; A thermal excitation and image acquisition module for the material to be tested, configured to apply standardized thermal excitation to the thermal insulation material to be tested using the thermal excitation device, and simultaneously acquire a time series of thermal infrared images of the thermal insulation material to be tested at the predetermined sampling frequency; a thermal response feature extraction module for the material to be detected, configured to extract thermal response features from a time series of thermal infrared images of the thermal insulation material to be detected to obtain a thermal response feature vector of the thermal insulation material to be detected; The similarity analysis module is used to perform similarity analysis on the thermal response characteristic vector of the thermal insulation material to be tested and the thermal response characteristic vector of the thermal insulation material sample to determine whether the thermal insulation performance of the thermal insulation material to be tested is qualified.

Images