Infrared graph diagnosis method and system for thermal defects of building energy-saving transformation

Abstract

The invention discloses an infrared graph diagnosis method and system for thermal defects of building energy-saving reconstruction, and relates to the technical field of building defect diagnosis, and the method comprises the steps: synchronously obtaining an infrared image sequence and wind speed vector data at a corresponding moment; radiation correction and time sequence registration are carried out on the infrared image sequence, and then an external space-time temperature field is extracted; generating an external disturbance characteristic sequence based on the space-time temperature field and the wind speed vector data; the temperature gradient and the time sequence change rate of the internal air layer are obtained, and an internal convection characteristic sequence is generated after fusion; calculating a time sequence correlation coefficient between the external disturbance characteristic sequence and the internal convection characteristic sequence; according to whether the time sequence correlation coefficient is lower than a threshold value or not, an effective internal convection signal with external airflow interference removed is directly determined or obtained through decoupling correction; and finally, identifying a thermotechnical defect area in the building envelope based on the effective internal convection signal. The method has the beneficial effect that the energy-saving reconstruction diagnosis accuracy and reliability of complex enclosure structures such as double-layer curtain walls are improved.

Description

Technical Field

[0001] This invention relates to the field of building defect diagnosis technology, and in particular to an infrared graphic diagnosis method and system for thermal defects in building energy-saving retrofitting. Background Technology

[0002] In the process of building energy-saving renovation, the thermal performance of the building envelope is a key factor affecting energy efficiency. Among them, the detection and repair of thermal defects is the core link of energy-saving renovation.

[0003] Existing building thermal diagnostics typically employ infrared graphic diagnostic technology. By collecting the infrared temperature distribution on the building surface, it can intuitively reflect thermal anomalies such as insulation layer detachment, thermal bridging, and moisture leakage. This technology has the advantages of being non-contact, fast, and visual.

[0004] However, in composite enclosure structures such as double-layer curtain walls, due to the presence of an air layer inside, a complex coupling effect is formed between the natural convection within the air layer and the external wind pressure. External airflow disturbances affect the stability of the flow field within the air layer, resulting in the temperature distribution on the curtain wall surface exhibiting unsteady fluctuation characteristics.

[0005] Existing infrared image diagnostic methods are mainly based on static thermal imaging analysis, which usually assumes that the surface of the structure being tested is under steady-state heat conduction conditions. Therefore, when faced with a complex thermal environment caused by airflow interference and uneven convection within the air layer, it is easy to fail to distinguish between the surface temperature difference caused by external wind pressure fluctuations and the actual thermal defects caused by internal convection instability. This leads to deviations in the assessment results of the curtain wall's energy-saving performance and affects the accuracy of subsequent energy-saving renovation decisions.

[0006] Therefore, an infrared graphic diagnostic method and system for thermal defects in building energy-saving retrofitting is proposed. Summary of the Invention

[0007] In view of the above-mentioned prior art, this application is hereby filed. Embodiments of this application provide an infrared graphic diagnostic method and system for thermal defects in building energy-saving retrofits, which can improve the accuracy and reliability of energy-saving retrofit diagnostics for complex building envelopes such as double-layer curtain walls.

[0008] According to one aspect of this application, an infrared image diagnosis method for thermal defects in building energy-saving retrofitting is provided, comprising: acquiring an infrared image sequence of the exterior of the building envelope and wind speed vector data at corresponding times; extracting the spatiotemporal temperature field of the exterior of the building envelope after performing radiometric correction and temporal registration on the infrared image sequence; generating an external disturbance feature sequence representing the influence of external airflow disturbance on the temperature distribution of the building envelope by normalization processing based on the spatiotemporal temperature field and the wind speed vector data; acquiring the temperature gradient and its temporal rate of change of the air layer inside the building envelope; and sorting the temperature gradient and its temporal rate of change according to... The system uses preset weighted coefficients to fuse and generate an internal convection feature sequence representing the fluctuation of the air convection state within the building envelope. It then calculates the temporal correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence. Next, it determines whether the temporal correlation coefficient is lower than a first preset threshold. If so, the internal convection feature sequence is identified as a valid internal convection signal; otherwise, decoupling correction is performed on the internal convection feature sequence based on the temporal correlation coefficient to obtain the valid internal convection signal free from external airflow interference. Finally, based on the valid internal convection signal, thermal defect areas within the building envelope are identified.

[0009] According to another aspect of this application, an infrared graphic diagnostic system for thermal defects in building energy-saving retrofitting is provided, comprising: an external data acquisition module for acquiring an infrared image sequence of the exterior of the building envelope and corresponding wind speed vector data; an external spatiotemporal temperature field extraction module for extracting the spatiotemporal temperature field of the exterior of the building envelope after radiometric correction and temporal registration of the infrared image sequence; an external disturbance feature generation module for generating an external disturbance feature sequence representing the influence of external airflow disturbance on the temperature distribution of the building envelope based on the spatiotemporal temperature field and the wind speed vector data through normalization processing; an internal temperature feature acquisition module for acquiring the temperature gradient and its temporal rate of change of the air layer inside the building envelope; and an internal convection feature generation module for... The temperature gradient and its temporal rate of change are fused according to a preset weighting coefficient to generate an internal convection feature sequence representing the fluctuation of the convection state of the air layer inside the building envelope; a temporal coherence calculation module is used to calculate the temporal correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence; an effective internal convection signal processing module is used to determine whether the temporal correlation coefficient is lower than a first preset threshold. If so, the internal convection feature sequence is determined as an effective internal convection signal. Otherwise, decoupling correction is performed on the internal convection feature sequence based on the temporal correlation coefficient to obtain the effective internal convection signal after removing external airflow interference; a thermal defect area identification module is used to identify thermal defect areas in the building envelope based on the effective internal convection signal.

[0010] According to another aspect of this application, an electronic device is provided, including a memory and a processor, the memory being used to store computer-executable instructions, and the processor being used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method described above.

[0011] According to another aspect of this application, a computer storage medium is provided that stores computer-executable instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0012] Compared with existing technologies, the infrared graphic diagnostic method and system for thermal defects in building energy-saving retrofits according to the embodiments of this application can achieve quantitative decoupling between external wind disturbances and internal thermal anomalies, thereby improving the accuracy and reliability of energy-saving retrofit diagnostics for complex building envelopes such as double-layer curtain walls. Attached Figure Description

[0013] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0014] Figure 1 This is a flowchart of the infrared graphic diagnosis method for thermal defects in building energy-saving retrofitting according to the present invention.

[0015] Figure 2 This is a block diagram of the infrared graphic diagnostic system for thermal defects in building energy-saving retrofitting according to the present invention.

[0016] Figure 3 This is a block diagram of an electronic device according to the present invention. Detailed Implementation

[0017] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0018] Exemplary methods

[0019] Figure 1 The illustration shows an infrared graphic diagnosis method for thermal defects in building energy-saving retrofitting according to an embodiment of this application, including steps S1 to S8.

[0020] like Figure 1 As shown, in step S1, an infrared image sequence of the exterior of the building envelope and wind speed vector data at the corresponding time are acquired.

[0021] This step is the data foundation of the entire diagnostic method. The infrared image sequence records the dynamic changes in temperature distribution on the outer surface of the building envelope over time, while the wind speed vector data reflects the intensity and direction of external airflow disturbances. The synchronous acquisition of both is a prerequisite for subsequent decoupling analysis.

[0022] In practice, infrared image sequences can be acquired using mid-wave or long-wave infrared cameras, and wind speed vector data can be obtained using an anemometer. It should be noted that, to ensure data consistency, the infrared camera and the anemometer should be triggered through a unified time synchronization system.

[0023] The above scheme yielded time-registered infrared image sequences and wind speed vector data, laying a data foundation for subsequent external disturbance feature extraction and internal-external coupling analysis.

[0024] return Figure 1 In step S2, after radiometric correction and temporal registration of the infrared image sequence, the spatiotemporal temperature field outside the building envelope is extracted.

[0025] This step aims to convert raw infrared images into quantitatively analyzable temperature field data. Radiometric correction eliminates the influence of factors such as atmospheric transmittance, ambient radiation, and lens distortion on infrared imaging, ensuring absolute accuracy in temperature measurement. The temporal alignment criterion precisely aligns infrared images acquired at different times in space, eliminating pixel shifts caused by minor camera shake or changes in viewing angle, thereby guaranteeing the comparability of temperature data from the same physical location at different times.

[0026] The radiation correction includes the following steps: First, based on the calibration parameters of the infrared camera, the infrared radiation intensity is converted into apparent temperature using Planck's radiation law; then, based on the atmospheric temperature, relative humidity, and target distance measured on-site, the atmospheric transmittance is calculated and atmospheric radiation correction is performed; finally, considering the emissivity characteristics of the target surface, the true surface temperature is obtained.

[0027] The temporal registration process employs an image registration algorithm based on feature point matching: First, stable feature points, such as building edges and window frame corners, are extracted from the first frame image; then, the positional changes of these feature points are tracked in subsequent frames; finally, all frames are aligned to the same coordinate system through affine transformation or perspective transformation.

[0028] After the above processing, the spatiotemporal temperature field is obtained. ,in Represents spatial coordinates, This indicates time. The temperature field not only contains information on the temperature distribution of the building surface, but more importantly, it preserves the dynamic characteristics of temperature evolution over time, providing high-quality data input for subsequent time-series analysis.

[0029] return Figure 1 In step S3, based on the spatiotemporal temperature field and wind speed vector data, a sequence of external disturbance characteristics representing the influence of external airflow disturbance on the temperature distribution of the building envelope is generated through normalization processing.

[0030] The core objective of this step is to extract a standardized feature sequence from external observation data that can quantify the intensity of external airflow disturbances. Traditional methods directly use wind speed or temperature change rate as disturbance indicators, but due to the different dimensions and large differences in numerical ranges between wind speed and temperature change rate, effective fusion is difficult. This step solves the problem of dimension inconsistency through normalization and achieves complementarity of multi-source information through weighted fusion, enabling the intensity of external disturbances to be represented by a unified dimensionless scalar sequence.

[0031] Specifically, the generation of the external perturbation feature sequence includes the following steps:

[0032] First, the magnitude of the wind speed vector data is normalized to obtain the normalized wind speed.

[0033] Wind speed vector data It includes information on wind speed magnitude and direction. The wind speed modulus is calculated. Then, the reference wind speed is adopted. Normalize:

[0034] ;

[0035] Among them, reference wind speed Generally, the 90th percentile of wind speed during the measurement period or a typical wind speed value determined based on local meteorological statistics is used. Normalized wind speed. As a dimensionless quantity, its magnitude directly reflects the strength of the wind speed at the current moment relative to the reference state.

[0036] Returning to the generation of the external disturbance feature sequence, we then calculate the temporal rate of change of the average temperature of the outer surface in the spatiotemporal temperature field and normalize it to obtain the normalized temperature rate of change.

[0037] Average temperature of outer surface Defined as a spacetime temperature field Average value in space:

[0038] ;

[0039] in Let be the area of ​​the observation region. Calculate the time derivative of this average temperature to obtain the rate of temperature change. This rate of change reflects how quickly the outer surface temperature fluctuates over time. To eliminate the influence of dimensions, a reference temperature rate of change is used. Normalize:

[0040] ;

[0041] The reference temperature change rate can be determined based on historical data statistics or theoretical heat transfer models. For example, for a typical double-layer curtain wall, under steady-state conditions without wind disturbance, the external surface temperature change rate generally does not exceed 0.1°C / min, and this value can be taken as a reference. When the external wind field suddenly intensifies, due to the increase in the convective heat transfer coefficient, the external surface temperature change rate may reach 0.5°C / min or even higher. At this time, the normalized temperature change rate will increase significantly, reflecting strong external disturbances.

[0042] The advantage of this normalization process is that it transforms the rate of temperature change into a dimensionless quantity of the same order of magnitude as the normalized wind speed, facilitating subsequent weighted fusion. Simultaneously, by standardizing the reference values, data from different projects and under different climatic conditions become comparable.

[0043] Returning to the generation of the external disturbance feature sequence, the normalized wind speed and normalized temperature change rate are finally weighted and fused based on preset weight coefficients to obtain the external disturbance feature sequence.

[0044] External disturbance intensity Defined as a weighted sum of two normalized features:

[0045] ;

[0046] in, , For the preset weighting coefficients, satisfy and and Both are greater than zero, reflecting the relative importance of wind speed and temperature change rates in characterizing external disturbances. In practical applications, they can be determined through fitting experimental data during system calibration or through theoretical analysis based on energy balance equations.

[0047] Generally speaking, The value range is 0.6, because wind speed directly drives convective heat transfer and is the main factor of external disturbance; The value is 0.4. Although the rate of temperature change is a result of the disturbance rather than its cause, it can reflect the cumulative effect of the disturbance. For example, when wind speed fluctuates frequently but with small amplitudes, the rate of temperature change may still remain at a low level; while under the influence of continuous strong winds, even if the wind speed stabilizes for a short period, the cumulative heat exchange effect will keep the rate of temperature change at a high level. Therefore, the fusion of the two can more comprehensively characterize the dynamic characteristics of external disturbances.

[0048] Will As a characteristic sequence of external perturbation : .

[0049] Assuming the normalized wind speed at a certain moment =1.2, normalized rate of temperature change =0.8, take the weighting coefficient =0.7, =0.3, then the intensity of the external disturbance is:

[0050] ;

[0051] This value indicates that the intensity of the external disturbance is slightly higher than the reference state (based on 1.0). The impact of this disturbance on the inner air layer will be assessed in subsequent steps.

[0052] Through the above three sub-steps, a standardized external disturbance feature sequence was successfully constructed. This sequence not only eliminated the dimensional differences, but also enhanced the characterization ability of external airflow disturbances through multi-source information fusion, providing a reliable input for subsequent internal and external coupling analysis.

[0053] return Figure 1 In step S4, the temperature gradient and its temporal rate of change of the air layer inside the building envelope are obtained.

[0054] The purpose of this step is to obtain key physical quantities reflecting the thermal state of the air layer inside the double-walled curtain wall. Temperature gradient. It characterizes the degree of uneven temperature distribution within the air layer and is a direct indicator of convection intensity; time-series variation rate This reflects the dynamic evolution rate of the overall thermal state of the air layer. The combination of these two physical quantities can comprehensively characterize the convective state fluctuations of the internal air layer.

[0055] Traditional methods often neglect the acquisition of the internal temperature field or rely solely on infrared images of the outer surface for inference, leading to inaccurate characterization of the internal thermal state. This application provides two schemes for acquiring internal temperature information, each suitable for different engineering scenarios.

[0056] The first option is a sensor array solution, suitable for new or renovation projects, which includes:

[0057] Acquire time-series temperature sampling data at at least three different altitudes within the air layer;

[0058] The temperature gradient and its temporal rate of change are calculated based on the time-series temperature sampling data.

[0059] This scheme directly measures temperature values ​​at different altitudes by deploying an array of temperature sensors inside the air layer. The sensor deployment follows the minimum three-point principle: at least one high-precision temperature sensor is installed at the top, middle, and bottom of the air layer, and the sampling frequency is consistent with that of the infrared image.

[0060] The specific calculation process is as follows:

[0061] Let the temperature measured by the top sensor be... The temperature measured by the bottom sensor is The vertical distance between the sensors is Then the temperature gradient is approximately:

[0062] ;

[0063] The average temperature of the air layer is:

[0064] ;

[0065] Taking the time derivative of the average temperature, we obtain the rate of change over time:

[0066] ;

[0067] in, This represents the sampling time interval.

[0068] For example, in a double-glazed curtain wall, the vertical spacing of the sensors... Meters, measured at a certain moment =18.2 , The temperature gradient is:

[0069] ;

[0070] If the average temperature 5 seconds ago was The current average temperature is Then the rate of change over time is:

[0071] ;

[0072] The advantages of the above-mentioned sensor array scheme are high measurement accuracy and strong data reliability, making it the preferred solution for new construction projects or buildings that can be renovated.

[0073] The second option is infrared tomography inversion, which is suitable for existing buildings and includes:

[0074] Based on spatiotemporal temperature field and wind speed vector data, the temporal temperature distribution within the air layer is obtained by inversion algorithm based on multilayer radiative transfer model.

[0075] The temperature gradient and its temporal rate of change are calculated based on the time-series temperature distribution.

[0076] This approach is suitable for existing buildings where it is impossible to deploy sensors inside. The basic principle is that although infrared cameras can only observe the surface temperature of the outer curtain wall, this temperature distribution is influenced by the temperature field inside the air layer. By establishing a multi-layer radiative transfer model and combining it with boundary conditions such as external wind speed, the internal temperature field can be inverted and calculated.

[0077] Specifically, the inversion algorithm is based on the following physical model:

[0078] Infrared radiation intensity of the outer curtain wall surface It consists of three parts: radiation from the outer glass layer itself, radiation transmitted through the air layer, and radiation reflected by the inner glass layer. The radiative transfer equation is established as follows:

[0079] ;

[0080] in, For glass reflectivity, Air layer transmittance, For glass reflectivity, The Stefan-Boltzmann constant is... It is environmental radiation.

[0081] This equation relates to the observable external surface temperature. With the required internal temperature By iteratively solving the inverse problem, we can... Inversion yields The iterative process uses gradient descent.

[0082] ;

[0083] in, The root mean square error residual between the observed temperature and the model-predicted temperature:

[0084] ;

[0085] The iteration step size is determined by multiple iterations until... It converges to below the preset threshold.

[0086] The internal temperature field was obtained through inversion. Then, the calculation method is similar to that of Scheme 1:

[0087] ;

[0088] in, For effective height.

[0089] It should be noted that the accuracy of the inversion scheme is lower than that of the direct measurement scheme, but for existing buildings, it is sufficient to meet diagnostic needs. In engineering practice, it is recommended to prioritize Scheme 1; Scheme 2 should only be used as an alternative when sensors cannot be installed.

[0090] The temperature gradient of the internal air layer was successfully obtained using one of the two methods described above. and its time-series rate of change These two physical quantities are the basic data for constructing the internal convection characteristic sequence.

[0091] return Figure 1 In step S5, the temperature gradient and its temporal rate of change are fused according to a preset weighting coefficient to generate an internal convection characteristic sequence representing the fluctuation of the convection state of the air layer inside the building envelope.

[0092] The core idea of ​​this step is similar to that of step S3, both involving the weighted fusion of multiple physical quantities to construct a comprehensive characteristic index. Temperature gradient It reflects the instantaneous convection intensity within the air layer, while the time-series rate of change... This reflects the dynamic evolution trend of the convection state. The integration of the two can more comprehensively characterize the fluctuation characteristics of internal convection.

[0093] Internal convection intensity Defined as:

[0094] ;

[0095] in , These are the preset weighting coefficients. It should be understood that although the normalization step is not explicitly stated here, similar to the construction of the external perturbation feature sequence, it is also necessary to... and Normalization is performed to eliminate the influence of dimensions.

[0096] The weighting coefficients here , satisfy , , All are greater than zero, and are consistent with the preset weighting coefficients in step 3. , The principle for determining the value is the same. The value of is usually large because the temperature gradient is the direct driving force of the flow, while The values ​​are relatively small because the rate of temperature change characterizes the result of convection. This weighting allocation is consistent with the physical mechanism of convective heat transfer.

[0097] Will As an internal convection characteristic sequence : .

[0098] For example, measuring the temperature gradient at a certain moment. The normalized value is 1.12, representing the rate of temperature change. After normalization, it is 0.67, and the weighting coefficient is taken. , Then the internal convection intensity is:

[0099] ;

[0100] This value is close to 1.0, indicating that the internal convection is at a moderate level.

[0101] The above scheme successfully constructed a standardized internal convection characteristic sequence, which can quantitatively characterize the convection state fluctuations inside the air layer of the double-layer curtain wall, providing key internal signal input for subsequent internal and external coupling diagnosis.

[0102] return Figure 1 In step S6, the temporal correlation coefficient between the external disturbance characteristic sequence and the internal convection characteristic sequence is calculated.

[0103] This step is one of the core innovations of this application. The temporal correlation coefficient quantitatively assesses the degree of correlation between external airflow disturbances and internal air layer convection over time. When the two are highly coherent, i.e., the absolute value of the coherence coefficient is close to 1, it indicates that internal convection fluctuations are mainly driven by external disturbances. In this case, the detected temperature anomaly is likely a spurious signal from external wind disturbances rather than a real thermal defect. Conversely, when the coherence is very low, i.e., the absolute value of the coherence coefficient is close to 0, it indicates that internal convection is relatively independent, and the temperature anomaly is more likely to originate from internal thermal defects.

[0104] Traditional methods lack this quantitative coupling assessment mechanism, often attributing all temperature anomalies to thermal defects, leading to a high false positive rate. This application achieves quantitative decoupling between external disturbances and internal thermal anomalies by introducing temporal coherence analysis.

[0105] The calculation of the time-series correlation coefficient includes the following steps:

[0106] Within a preset sliding time window, calculate the mean and standard deviation of the external disturbance characteristic sequence and the internal convection characteristic sequence respectively;

[0107] The normalized correlation coefficient between the external disturbance characteristic sequence and the internal convection characteristic sequence is calculated based on the mean and standard deviation, and is used as the time series correlation coefficient.

[0108] Specifically, the time-series correlation coefficient uses a sliding window version of the Pearson correlation coefficient. Within a length of... Within the time window, the statistical properties of the two feature sequences are first calculated:

[0109] Mean of the external disturbance feature sequence :

[0110] ;

[0111] Standard deviation of external perturbation feature sequence :

[0112] ;

[0113] Mean and standard deviation of internal convection characteristic sequences , The calculation methods are similar.

[0114] Time series correlation coefficient Defined as:

[0115] ;

[0116] in, For length is The moving average operator, The length of the sliding time window. , These are the mean values ​​of the external disturbance characteristic sequence and the internal convection characteristic sequence, respectively. , These are the standard deviations of the external disturbance characteristic sequence and the internal convection characteristic sequence, respectively. This represents the characteristic sequence of external perturbations. This represents the internal convection characteristic sequence.

[0117] Expanding the moving average operator, the complete calculation formula is as follows:

[0118] ;

[0119] Time series correlation coefficient The range of values ​​is . This indicates a perfect positive correlation, meaning that internal convection increases synchronously when external disturbances intensify. This indicates a completely negative correlation; This indicates that the two are not wirelessly related.

[0120] Based on practical engineering experience and typical time scales of wind speed fluctuations, the sliding time window length is... The preferred value is 30 seconds, corresponding to 120 data points. If the window is too small, there will be insufficient data points, the statistical estimate will be unstable, and the coherence coefficient will fluctuate drastically. If the window is too large, it contains too much time span and cannot capture rapid changes in the external wind field.

[0121] For example, suppose that within a 45-second time window, the mean of the external disturbance feature sequence is... 1.05, standard deviation =0.15; Mean of the internal convection characteristic sequence =0.92, standard deviation =0.12. Calculating the numerator of the covariance (the sum of the products after centering) yields 0.0162, therefore:

[0122] ;

[0123] This result indicates that the external disturbance is highly coherent with the internal convection at this moment, meaning that the fluctuations in internal convection at the current moment are mainly driven by the external wind field, rather than by internal thermal defects.

[0124] It is important to emphasize that the time-series correlation coefficient It changes dynamically over time. When the external wind speed remains stable, It may decrease; however, when wind speed changes abruptly or gusts occur, It will rise rapidly. This dynamic tracking capability enables this method to adapt to complex and ever-changing external weather conditions.

[0125] The time-series correlation coefficient was obtained through the above calculations. This coefficient quantitatively characterizes the degree of influence of external disturbances on internal convection, providing a basis for subsequent decoupling decisions.

[0126] return Figure 1 In step S7, it is determined whether the time-series correlation coefficient is lower than the first preset threshold. If so, the internal convection feature sequence is determined as an effective internal convection signal. Otherwise, the internal convection feature sequence is decoupled and corrected based on the time-series correlation coefficient to obtain an effective internal convection signal that removes external airflow interference.

[0127] This step achieves automatic separation between external disturbances and internal thermal anomalies. This is achieved by setting a first preset threshold. (Preferred value: 0.3~0.5), different working conditions are divided into two categories for processing:

[0128] The first type of situation is Low coherence case:

[0129] When the absolute value of the time series correlation coefficient is lower than the first preset threshold, it indicates that external disturbances and internal convection are basically independent. At this point, fluctuations in internal convection are mainly determined by the internal thermal state, and the influence of the external wind field can be ignored. In this case, the internal convection characteristic sequence can be directly used as the effective signal. .

[0130] For example, assuming a first preset threshold The coherence coefficient is 0.3, indicating that under calm or light wind conditions at night, external disturbances are very weak, and the calculated coherence coefficient is [value missing]. At this point, it is determined that internal convection is independent of external disturbances, and the internal convection characteristic sequence is... It can be directly used for subsequent thermal defect identification without the need for decoupling correction.

[0131] The second type of situation is High coherence case:

[0132] When the absolute value of the time series correlation coefficient is not lower than the first preset threshold, it indicates that the external disturbance has a significant impact on the internal convection. At this time, the internal convection characteristic sequence... The spurious signal mixed with external disturbances must be decoupled and corrected in order to extract the true internal thermal signal.

[0133] The core idea of ​​decoupling correction is to subtract the components caused by external disturbances from the internal convection characteristic sequence, while retaining the components caused by internal thermal defects. The formula for decoupling correction is:

[0134] ;

[0135] in, The decoupling coefficient is adaptively determined by the ratio of the external disturbance characteristic sequence to the internal convection characteristic sequence within a preset historical time window. Represents the time-series correlation coefficient. This represents the characteristic sequence of external perturbations. Represents the internal convection characteristic sequence. This indicates an effective internal convection signal.

[0136] The physical meaning of this formula is as follows:

[0137] Represents the original strength of the external disturbance. The coefficient representing the transfer of external disturbances to internal convection, i.e., the coherence intensity. This represents the response component induced by external disturbances in internal convection. This represents the pure internal signal after removing the influence of external disturbances.

[0138] Among them, the decoupling coefficient There are two methods for determining it:

[0139] One approach is to use a historical window adaptive method, which statistically analyzes the actual impact of external disturbances on internal convection within a historical time window.

[0140] ;

[0141] This method estimates the strength of external disturbances by comparing the corrected signal with the original signal within a historical window, and then uses this difference for decoupling at the current time. Its advantage lies in its ability to adapt to slowly changing conditions.

[0142] It should be noted that during the cold start phase, when there is not enough historical data, a fixed calibration coefficient is used for decoupling.

[0143] The second method is the calibration coefficient method, which involves determining the decoupling coefficients under typical operating conditions through experimental data or simulation analysis. Then, adjust dynamically based on the current coherence:

[0144] ;

[0145] This method is computationally simple and suitable for relatively stable scenarios.

[0146] Using the previous example, let's assume the current time... , , The decoupling coefficient is calculated using method two, and is taken as follows: =0.6, then:

[0147] ;

[0148] ;

[0149] Effective internal convection signal after decoupling =0.460 is significantly lower than the original signal of 0.985, indicating that about 53% of the original signal is caused by external disturbances. The decoupling correction successfully eliminated this part of the spurious signal and retained about 47% of the real internal thermal signal.

[0150] It should be noted that decoupling correction is not a simple signal subtraction, but a directional correction based on physical mechanisms. Coherence coefficient The introduction of this ensures that corrections are only made when the two are indeed correlated; decoupling coefficient The adaptive adjustment ensures that the correction strength matches the actual coupling degree, avoiding over-correction or under-correction.

[0151] Through the above steps, regardless of changes in external weather conditions, the system can automatically determine whether decoupling is necessary and perform corresponding correction operations, ultimately obtaining an effective internal convection signal free from external airflow interference. This signal accurately reflects the thermal state of the internal air layer and serves as a reliable basis for subsequent thermal defect identification.

[0152] return Figure 1In step S8, thermal defect areas in the building envelope are identified based on effective internal convection signals.

[0153] This step is the ultimate goal of the entire diagnostic method: to transform the effective signals obtained from all the preceding processing steps into a spatial distribution map of thermal defects that can directly guide energy-saving retrofits. This step includes the following four sub-steps:

[0154] The first step is to construct the initial temperature distribution field of the air layer based on the temperature gradient and its temporal rate of change.

[0155] Although steps S4 and S5 yield scalarized temperature gradients and temporal rates of change, this information can be spatially interpolated to reconstruct a two-dimensional temperature field. The specific method is as follows:

[0156] For sensor array schemes, based on the known temperatures at discrete measurement points, the temperature distribution of the entire air layer is reconstructed using radial basis function (RBF) or kriging interpolation methods.

[0157] ;

[0158] in Let be the measurement value of the i-th sensor. Let be the interpolation weighting function, representing the spatial position of the i-th sensor. The contribution weight of the temperature value This represents the number of sensors.

[0159] For the inversion scheme, the temperature field obtained from the inversion in step S4 can be used directly. .

[0160] It should be noted that the initial temperature distribution field may still contain residual effects of external disturbances, especially in spatial boundary areas such as vents, so further corrections are needed in the next step.

[0161] The second step involves correcting the initial temperature distribution field based on the effective internal convection signal and the preset spatial weighting function to obtain the effective internal temperature field. The spatial weighting function characterizes the difference in the degree of influence of external airflow disturbance on different areas of the building envelope.

[0162] Although step S7 has decoupled the internal convection feature sequence in the temporal dimension, this decoupling is for the overall feature sequence and does not consider the non-uniformity of spatial distribution. In reality, the impact of external disturbances on different regions of the internal temperature field is uneven: areas near vents or building corners are more affected by the external wind field, while the central area of ​​the building is relatively stable. The introduction of a spatial weighting function enables spatial adaptive correction.

[0163] Spatial weight function Defined as:

[0164] ;

[0165] in, The temperature distribution on the outer surface is provided by the spatiotemporal temperature field of step S1. This represents the spatial gradient of the outer surface temperature distribution. The physical meaning of this weighting function is: regions with a large outer surface temperature gradient are more significantly affected by external wind disturbances, and therefore should be assigned a higher correction weight; conversely, regions with a smooth outer surface temperature are less affected by external disturbances and should be assigned a lower correction weight.

[0166] The correction formula for the temperature distribution field is:

[0167] ;

[0168] in, For an effective internal temperature field, This is a correction factor.

[0169] To prevent overcorrection or non-physical negative values ​​during sudden changes in external wind disturbances, a saturation constraint mechanism is introduced. The correction factor is calculated using a saturation function:

[0170] ;

[0171] The saturation function is defined as:

[0172] ;

[0173] The input value will be restricted to Within the range. Saturation threshold. The value is typically between 0.8 and 0.9.

[0174] Continuing with the previous example, suppose there is a certain location At the initial temperature Decoupling coefficient coherence coefficient Spatial weight saturation threshold ,but:

[0175] Calculate the original value of the correction factor: = .

[0176] Applying saturation constraints, because exist Within the range, no truncation is required, therefore, .

[0177] This sub-step yields the effective internal temperature field after temporal decoupling and spatial adaptive correction. This temperature field truly reflects the thermal state inside the double-layer curtain wall, eliminating the influence of external airflow disturbances.

[0178] The third step is to calculate the heat flux density distribution of the effective internal temperature field.

[0179] Heat flux density is a key indicator for evaluating the thermal insulation performance of building envelopes, as its magnitude directly reflects the intensity of heat transfer. The calculation of heat flux density is based on Fourier's law of thermal conductivity:

[0180] ;

[0181] in, This is the heat flux density vector (unit: W / m²). For the effective spatial gradient of the internal temperature field, It is the equivalent thermal conductivity.

[0182] For equivalent thermal conductivity In traditional methods, this is usually treated as a constant and calculated using a steady-state heat transfer model:

[0183] ;

[0184] in , These refer to the thicknesses of the air layer and the glass layer of the double-walled curtain wall. , These are the thermal conductivity coefficients of air and glass, respectively.

[0185] However, for double-layer curtain walls, when external wind disturbances increase, the convective heat transfer coefficient inside the air layer of the double-layer curtain wall will increase significantly, leading to an increase in the equivalent thermal conductivity. Therefore, a transient thermal conductivity correction is introduced here to make the heat flux density calculation more accurate:

[0186] ;

[0187] in, This is the convection enhancement factor, ranging from 0.25 to 0.5, reflecting the enhanced convective heat transfer effect in the air layer under wind disturbance conditions. This factor is calibrated through wind tunnel experiments or CFD (Computational Fluid Dynamics) simulations.

[0188] Its physical mechanism is as follows: When At that time, external disturbances are minimal, the air layer is nearly still, and convective heat transfer is weak. ≈ ;when As the external wind field increases, it drives turbulence within the air layer, enhancing convective heat transfer and increasing the equivalent thermal conductivity. .

[0189] When calculating the heat flux density distribution, it is necessary to determine the spatial gradient of the effective internal temperature field:

[0190] ;

[0191] In the discretization implementation, a central difference scheme can be used:

[0192] ;

[0193] in, For grid node indexing, This represents the grid spacing.

[0194] The heat flux density distribution q is finally obtained. This distribution clearly shows the heat transfer intensity in different areas of the building envelope. In well-insulated areas, the heat flux density is lower; while in areas with thermal bridges, damaged insulation layers, or abnormal air circulation, the heat flux density is significantly higher.

[0195] The fourth step is to identify thermal defect regions based on abrupt changes in the spatial gradient of the heat flux density distribution.

[0196] This step employs edge detection to identify abrupt changes in the heat flux density field. Normal regions should exhibit a smooth heat flux density distribution, while thermal defects will cause sharp changes in heat flux density. The spatial gradient of the heat flux density is then calculated.

[0197] ;

[0198] ;

[0199] Define the magnitude of the heat flux density gradient:

[0200] ;

[0201] Set threshold , This can be determined through statistical analysis, typically using the 95th percentile of the heat flux density gradient distribution. When the heat flux density gradient at a certain location exceeds this threshold, it is considered a thermal defect.

[0202] ;

[0203] in, 1 indicates the presence of a thermal defect. 0 indicates that there are no thermal defects.

[0204] It should be noted that in engineering practice, to improve the stability of recognition, the recognition results at multiple time points can be voted on.

[0205] ;

[0206] That is, only when it is judged as a defect in more than 70% of the time is it finally confirmed as a thermal defect area.

[0207] In summary, this application achieves a complete transformation from raw infrared images and wind speed data to final thermal defect diagnosis. Compared with traditional methods, the core innovation of this application lies in: achieving quantitative decoupling between external wind disturbance and internal thermal anomalies, fundamentally solving the problem of misjudgment in infrared diagnosis of double-layer curtain walls.

[0208] Exemplary System

[0209] Figure 2 The illustration shows an infrared graphic diagnostic system for thermal defects in building energy-saving retrofitting according to an embodiment of this application, comprising: an external data acquisition module for acquiring an infrared image sequence of the exterior of the building envelope and corresponding wind speed vector data; an external spatiotemporal temperature field extraction module for extracting the spatiotemporal temperature field of the exterior of the building envelope after radiometric correction and temporal registration of the infrared image sequence; an external disturbance feature generation module for generating an external disturbance feature sequence representing the influence of external airflow disturbance on the temperature distribution of the building envelope based on the spatiotemporal temperature field and wind speed vector data through normalization processing; an internal temperature feature acquisition module for acquiring the temperature gradient and its temporal rate of change of the air layer inside the building envelope; and an internal convection feature generation module for... The system integrates the temperature gradient and its temporal rate of change with a preset weighting coefficient to generate an internal convection feature sequence representing the fluctuation of the convection state of the air layer inside the building envelope; a temporal coherence calculation module is used to calculate the temporal correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence; an effective internal convection signal processing module is used to determine whether the temporal correlation coefficient is lower than a first preset threshold. If so, the internal convection feature sequence is determined as an effective internal convection signal; otherwise, decoupling correction is performed on the internal convection feature sequence based on the temporal correlation coefficient to obtain an effective internal convection signal that removes external airflow interference; and a thermal defect area identification module is used to identify thermal defect areas in the building envelope based on the effective internal convection signal.

[0210] In one example, obtaining the temperature gradient and its temporal rate of change of the air layer inside the building envelope includes: obtaining temporal temperature sampling data at at least three different heights within the air layer; and calculating the temperature gradient and its temporal rate of change based on the temporal temperature sampling data.

[0211] In one example, obtaining the temperature gradient and its temporal rate of change of the air layer inside the building envelope includes: based on spatiotemporal temperature field and wind speed vector data, the temporal temperature distribution within the air layer is obtained by inversion algorithm based on a multi-layer radiative transfer model; and the temperature gradient and its temporal rate of change are calculated based on the temporal temperature distribution.

[0212] In one example, the generation of the external disturbance feature sequence includes: normalizing the magnitude of the wind speed vector data to obtain the normalized wind speed; calculating and normalizing the temporal rate of change of the average temperature of the outer surface in the spatiotemporal temperature field to obtain the normalized temperature rate of change; and weighting and fusing the normalized wind speed and the normalized temperature rate of change based on preset weight coefficients to obtain the external disturbance feature sequence.

[0213] In one example, calculating the time-series correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence includes: calculating the mean and standard deviation of the external disturbance feature sequence and the internal convection feature sequence respectively within a sliding time window of a preset length; and calculating the normalized correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence as the time-series correlation coefficient based on the mean and standard deviation.

[0214] ;

[0215] in, For length is The moving average operator, The length of the sliding time window. , These are the mean values ​​of the external disturbance characteristic sequence and the internal convection characteristic sequence, respectively. , These are the standard deviations of the external disturbance characteristic sequence and the internal convection characteristic sequence, respectively. This represents the characteristic sequence of external perturbations. This represents the internal convection characteristic sequence.

[0216] In one example, the formula for decoupling correction is:

[0217] ;

[0218] in, The decoupling coefficient is adaptively determined by the ratio of the external disturbance characteristic sequence to the internal convection characteristic sequence within a preset historical time window. Represents the time-series correlation coefficient. This represents the characteristic sequence of external perturbations. Represents the internal convection characteristic sequence. This indicates an effective internal convection signal.

[0219] In one example, identifying thermal defect areas in a building envelope includes: constructing an initial temperature distribution field of the air layer based on the temperature gradient and its temporal rate of change; correcting the initial temperature distribution field based on the effective internal convection signal and a preset spatial weighting function to obtain the effective internal temperature field, wherein the spatial weighting function characterizes the difference in the degree of influence of external airflow disturbance on different areas of the building envelope; calculating the heat flux density distribution of the effective internal temperature field; and identifying thermal defect areas based on abrupt changes in the spatial gradient of the heat flux density distribution.

[0220] Exemplary electronic devices

[0221] Figure 3 A block diagram of an electronic device according to an embodiment of this application is illustrated.

[0222] like Figure 3 As shown, the electronic device includes one or more processors and memory.

[0223] A processor can be a central processing unit (CPU) or other form of processing unit with data processing and / or instruction execution capabilities, and can control other components in an electronic device to perform desired functions.

[0224] The memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0225] In one example, the electronic device may also include input devices and output devices, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0226] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device may include any other suitable components depending on the specific application.

[0227] Exemplary computer-readable media

[0228] Embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps described in the "Exemplary Methods" section above according to the various embodiments of this application.

[0229] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0230] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0231] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0232] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0233] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0234] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for diagnosing infrared patterns of thermal defects in building energy-saving reconstruction, characterized in that, The method comprises: obtaining an infrared image sequence of the exterior of a building envelope and wind speed vector data corresponding to the time; extracting a space-time temperature field of the exterior of the building envelope after radiation correction and time sequence registration of the infrared image sequence; based on the space-time temperature field and the wind speed vector data, generating an external disturbance feature sequence representing the influence of external airflow disturbance on the temperature distribution of the building envelope through normalization processing; obtaining the temperature gradient of the air layer inside the building envelope and its time sequence change rate; fusing the temperature gradient and its time sequence change rate according to a preset weighting coefficient to generate an internal convection feature sequence representing the fluctuation of the air layer inside the building envelope; calculating the time sequence correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence; determining whether the time sequence correlation coefficient is lower than a first preset threshold, if yes, determining the internal convection feature sequence as an effective internal convection signal, otherwise, performing decoupling correction on the internal convection feature sequence based on the time sequence correlation coefficient to obtain the effective internal convection signal removed from external airflow interference; based on the effective internal convection signal, identifying a thermal defect area in the building envelope.

2. The method according to claim 1, wherein, The method comprises: obtaining time sequence temperature sampling data of at least three different heights in the air layer; calculating the temperature gradient and its time sequence change rate according to the time sequence temperature sampling data.

3. The method of claim 1, wherein the method further comprises: The method comprises: based on the space-time temperature field and the wind speed vector data, the time sequence temperature distribution in the air layer is obtained by inversion algorithm based on multi-layer radiation transmission model; the temperature gradient and its time sequence change rate are calculated according to the time sequence temperature distribution.

4. The method of claim 1, wherein the method further comprises: The method comprises: normalizing the modulus of the wind speed vector data to obtain a normalized wind speed; calculating the time sequence change rate of the average temperature of the outer surface in the space-time temperature field and performing normalization processing to obtain a normalized temperature change rate; based on a preset weight coefficient, the normalized wind speed and the normalized temperature change rate are weighted and fused to obtain the external disturbance feature sequence.

5. The infrared graphic diagnosis method for thermal defects in building energy-saving retrofitting according to claim 1, characterized in that, The method comprises: calculating the mean and standard deviation of the external disturbance feature sequence and the internal convection feature sequence in a sliding time window of a preset length; calculating the normalized correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence as the time sequence correlation coefficient according to the mean and standard deviation: wherein is a moving average operator of length , is the length of the moving time window, , are the mean values of the external disturbance feature sequence and the internal convection feature sequence, respectively, , are the standard deviations of the external disturbance feature sequence and the internal convection feature sequence, respectively, denotes the external disturbance feature sequence, denotes the internal convection feature sequence.

6. The method according to claim 1 or 5, wherein, The formula of the decoupling correction is: wherein, is a decoupling coefficient, determined adaptively by a ratio of the external disturbance feature sequence to the internal convection feature sequence within a preset historical time window, denotes the time series correlation coefficient, denotes the external disturbance feature sequence, denotes the internal convection feature sequence, denotes the effective internal convection signal.

7. The method of claim 1, wherein the method further comprises: determining a thermal resistance of the building based on the infrared image and the thermal model. The method comprises: constructing an initial temperature distribution field of the air layer according to the temperature gradient and its time sequence change rate; correct the initial temperature distribution field based on the effective internal convection signal and a preset spatial weight function, to obtain an effective internal temperature field, wherein the spatial weight function represents a difference in influence of external airflow disturbance on different regions of the building envelope; calculate a heat flux density distribution of the effective internal temperature field; identify the thermal defect region according to a spatial gradient mutation of the heat flux density distribution.

8. The infrared pattern diagnosis system for thermal defects in building energy-saving reconstruction, characterized in that, The method comprises the following steps: an external data acquisition module configured to acquire an infrared image sequence and wind speed vector data corresponding to each time point of the building envelope; an external time-space temperature field extraction module configured to extract a time-space temperature field of the building envelope after performing radiation correction and time sequence registration on the infrared image sequence; an external disturbance feature generation module configured to generate an external disturbance feature sequence representing the influence of external airflow disturbance on the temperature distribution of the building envelope by normalization processing based on the time-space temperature field and the wind speed vector data; an internal temperature feature acquisition module configured to acquire a temperature gradient and a time sequence change rate of the air layer inside the building envelope; an internal convection feature generation module configured to fuse the temperature gradient and the time sequence change rate according to a preset weighting coefficient, to generate an internal convection feature sequence representing the fluctuation of the convection state of the air layer inside the building envelope; a time sequence coherence calculation module configured to calculate a time sequence correlation coefficient between the external disturbance feature sequence and the internal convection feature sequence; an effective internal convection signal processing module configured to determine whether the time sequence correlation coefficient is lower than a first preset threshold value, if yes, determine the internal convection feature sequence as an effective internal convection signal, otherwise, perform decoupling correction on the internal convection feature sequence based on the time sequence correlation coefficient, to obtain the effective internal convection signal free of external airflow disturbance; a thermal defect region identification module configured to identify a thermal defect region in the building envelope based on the effective internal convection signal. 9.An electronic device comprising a memory and a processor, the electronic device characterized by: The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the method according to any one of claims 1-7.

10. A computer storage medium having stored thereon computer- executable instructions, comprising: The computer executable instructions, when executed by the processor, implement the steps of the method according to any one of claims 1-7.

Images