System and method for detecting heat insulation performance of heat insulation gel material

Abstract

The invention relates to the technical field of detection, in particular to a thermal insulation performance detection system and method of a thermal insulation gel material.The method comprises the steps that surface temperature data and internal pore distribution data of a thermal insulation gel material sample are obtained and subjected to dynamic heat flow coupling treatment, and a dynamic coupling feature sequence is obtained; constructing a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence and carrying out regularization constraint to obtain a thermal impedance time sequence; extracting a nonlinear thermal shielding index in the thermal impedance time sequence, constructing a high-dimensional feature vector, and generating a probability density distribution parameter set; and according to the probability density distribution parameter set, constructing a multi-dimensional input data set and carrying out iterative optimization, calculating a confidence interval of the thermal insulation performance boundary value, evaluating the thermal insulation performance grade of the thermal insulation gel material based on the confidence interval, and generating a performance evaluation parameter report. The comprehensive heat insulation effect of the heat insulation gel material in actual use can be more comprehensively and accurately reflected.

Description

Technical Field

[0001] The present invention relates to the field of detection technologies, and more particularly, to a detection system and method for the heat insulation performance of a heat insulation gel material. Background Art

[0002] The heat insulation gel material is made of inorganic or organic polymer substances through a special manufacturing process, has high porosity and low thermal conductivity, and its basic feature is to enclose gas or other low thermal conductivity substances inside through a gel network structure, thereby effectively hindering the transfer of heat. However, there are still certain technical challenges in accurately evaluating the heat insulation performance of this material. Currently, the detection methods for heat insulation performance are often too simplistic and cannot comprehensively and accurately reflect the performance of the material in actual use.

[0003] The existing heat insulation performance testing methods are through thermal conductivity measurement technologies, such as the heat flow meter method, the steady state method, etc. These methods mostly adopt static test conditions and evaluate the heat insulation performance by measuring the thermal conductivity of the material at a specific temperature. Although these methods can provide certain information in some cases, due to the strong dynamic characteristics of the heat insulation performance itself, only static testing cannot comprehensively reflect the complex behavior of the heat insulation material in actual use. Especially for the heat insulation gel material, factors such as the pore distribution inside, the heat conduction path, and the dynamic heat response have an important impact on its heat insulation effect, while the existing technologies often ignore these key factors, making the test results often only provide a rough estimate of the heat insulation performance of the material, rather than obtaining an accurate and dynamic evaluation of the heat isolation ability.

[0004] Therefore, there is an urgent need for a new detection method to solve the above technical problems. Summary of the Invention

[0005] The main object of the present invention is to provide a detection method for the heat insulation performance of a heat insulation gel material, aiming to overcome the technical problem that in the detection of the heat insulation performance of the existing heat insulation gel material, the dynamic heat response of the material and the influence of the internal structure on the heat isolation performance cannot be comprehensively evaluated.

[0006] To achieve the above invention problem, the present invention proposes a detection method for the heat insulation performance of a heat insulation gel material, including: Obtain the surface temperature data and the internal pore distribution data of the heat insulation gel material sample, and fuse the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; Perform dynamic heat flow coupling processing on the thermal response baseline matrix to obtain a dynamic coupling characteristic sequence; Construct a multi-dimensional thermal impedance tensor based on the dynamic coupling characteristic sequence, and perform regularization constraint on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series; Extract the non - linear thermal shielding index in the thermal impedance time series, construct a high - dimensional feature vector based on the non - linear thermal shielding index, and generate a set of probability density distribution parameters; Construct a multi - dimensional input data set according to the set of probability density distribution parameters, and perform iterative optimization on the multi - dimensional input data set to obtain the heat insulation performance boundary value; Calculate the confidence interval of the heat insulation performance boundary value, evaluate the heat insulation performance grade of the heat insulation gel material based on the confidence interval, and generate a performance evaluation parameter report.

[0007] Further, the step of obtaining the surface temperature data and the internal pore distribution data of the heat insulation gel material sample and fusing the surface temperature data and the pore distribution data to obtain the thermal response baseline matrix includes: Based on the infrared thermal imager to collect the two - dimensional distribution data of the sample surface temperature changing with time, perform discretization processing on the two - dimensional distribution data to generate the original data stream of the surface temperature containing temperature spatio - temporal evolution information; Decompose the original data stream of the surface temperature into multiple sub - band signals according to the time series, extract the characteristic frequency peaks of each sub - band signal, and generate a set of thermal response frequency components; Obtain the scattering data of the internal pore structure of the heat insulation gel material sample through X - ray scattering processing, calculate the anisotropy eigenvalue of the scattering data, and obtain the pore anisotropy vector; Calculate the boundary fractal dimension of the pore structure according to the pore anisotropy vector, and generate a set of pore fractal parameters containing multi - scale fractal information; Construct an initial feature matrix, map the set of thermal response frequency components and the set of pore fractal parameters to the initial feature matrix, and generate the thermal response baseline matrix.

[0008] Further, the step of performing dynamic heat - flow coupling processing on the thermal response baseline matrix to obtain a dynamic coupling feature sequence includes: Divide the thermal response baseline matrix into multiple sub - regions, and obtain the heat - flow modulation distribution data by applying different amplitude heat - flow modulation signals to each sub - region; Construct a heat - flow probe array, and generate heat - flow distribution data according to the layer - by - layer distribution characteristics of the heat - flow collected from the surface to the inside by the heat - flow probe array; Extract the frequency characteristics of the heat - flow distribution data, analyze the difference in the energy spectral density between the frequency characteristics and the surface temperature data, and generate cross - spectral density distribution data; Perform point - by - point correction on the cross - spectral density distribution data according to the frequency order according to the pore distribution data to generate a dynamic coupling feature sequence.

[0009] Further, the step of generating a dynamic coupling feature sequence by pointwise correcting the cross-spectral density distribution data according to the pore distribution data in frequency order further includes: Performing a weighting process on each frequency point of the cross-spectral density distribution data based on the spatial heterogeneity of the pore distribution data to generate weighted cross-spectral density data; Setting a sliding window with a preset number of digits, controlling the sliding window to slide along the frequency axis on the weighted cross-spectral density data, and extracting local feature vectors within each sliding window; Sorting multiple local feature vectors according to similarity and performing clustering analysis to generate a set of clustering center feature vectors; Globally optimizing the weighted cross-spectral density data according to the set of clustering center feature vectors to obtain an optimized dynamic coupling feature sequence.

[0010] Further, the step of constructing a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence and performing regularization constraints on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series includes: Decomposing the dynamic coupling feature sequence into multiple spatial resolution levels, traversing the fluctuation characteristics of each level, and generating feature distribution data; Fusing the heat flux component and the temperature component of each level in the feature distribution data into a high-order tensor, and stacking the distribution characteristics of different scales layer by layer to generate an initial multi-dimensional tensor; Calculating the singular value distribution of each dimension in the initial multi-dimensional tensor to generate a set of thermal impedance components including multiple types of components, where the set of thermal impedance components includes at least a scalar component and a direction component; Constructing a fractal constraint function according to the pore distribution characteristics in the thermal response baseline matrix, embedding the fractal constraint function into the scalar component and the direction component and adjusting the regularization weight to generate a regularized thermal impedance tensor; Performing an evolution simulation process on the regularized thermal impedance tensor according to the time step to obtain the thermal impedance time series.

[0011] Further, the step of extracting the non-linear thermal shielding index in the thermal impedance time series, constructing a high-dimensional feature vector based on the non-linear thermal shielding index, and generating a set of probability density distribution parameters includes: Extracting the non-linear fluctuation characteristics of the thermal impedance time series in the time dimension to generate a non-linear response distribution sequence; Extracting the spectral density peak in the dynamic coupling feature sequence according to the frequency information in the non-linear response distribution sequence, obtaining multiple frequency domain components by Fourier transform of the spectral density peak, and performing spectral reconstruction processing on the multiple frequency domain components to obtain a composite feature spectrum; Map the spectral components of the composite feature spectrum to the fractal space, calculate the fractal dimension features of each spectral component, and obtain the thermal shielding index set; Construct a high-dimensional feature vector based on the thermal shielding index set, where the high-dimensional feature vector includes the thermal shielding index set and its evolution information in the time dimension; Use the kernel density estimation method to estimate the probability density distribution of the high-dimensional feature vector and generate a set of probability density distribution parameters.

[0012] Further, the step of constructing a multi-dimensional input data set based on the set of probability density distribution parameters and iteratively optimizing the multi-dimensional input data set to obtain the heat insulation performance boundary value includes: Split the set of probability density distribution parameters into statistical distribution components of different scales according to a preset scale threshold and integrate them to form a parameter matrix; Set an initial input framework, and perform topological reconstruction processing on the initial input framework according to the parameter matrix to obtain a multi-dimensional input data set; Perform convolution processing on each dimension of the multi-dimensional input data set in turn, extract local correlations and compress redundant information to obtain a compressed feature sequence; Construct an iterative model, adjust the weights and biases of the iterative model based on the distribution characteristics of the compressed feature sequence, and update the parameters layer by layer based on the gradient feedback mechanism to generate an optimized network model; Segment the compressed feature sequence and input it into the fully connected layer of the optimized network model, and perform feature mapping through a non-linear activation function to obtain a boundary prediction feature set; Construct a heat flux response function, perform weighted adjustment on the heat flux response function based on the distribution characteristics of the boundary prediction feature set, and calculate the boundary value point by point along the domain of definition of the weighted-adjusted heat flux response function based on the constraint optimization algorithm to obtain the heat insulation performance boundary value.

[0013] Further, the step of calculating the confidence interval of the heat insulation performance boundary value, evaluating the heat insulation performance level of the heat insulation gel material based on the confidence interval, and generating a performance evaluation parameter report includes: Perform multiple samplings on the heat insulation performance boundary value to generate a series of simulated boundary value data; Calculate the mean and variance of the simulated boundary value data, and determine the confidence interval of the heat insulation performance boundary value according to the mean and variance; Construct a heat insulation performance evaluation model, and classify the heat insulation performance of the heat insulation gel material based on the distribution characteristics of the boundary values within the confidence interval; Generate a performance evaluation parameter report according to the classification result, where the performance evaluation parameter report includes at least the heat insulation performance level, the confidence interval range, and relevant evaluation indicators.

[0014] The present invention also provides a heat insulation performance detection system for a heat insulation gel material, comprising: An acquisition module, configured to acquire the surface temperature data and the pore distribution data inside the heat insulation gel material sample, fuse the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; A coupling module, configured to perform dynamic heat flux coupling processing on the thermal response baseline matrix to obtain a dynamic coupling feature sequence; A regularization module, configured to construct a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence, and perform regularization constraints on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series; An extraction module, configured to extract the non-linear heat shielding index in the thermal impedance time series, construct a high-dimensional feature vector based on the non-linear heat shielding index, and generate a probability density distribution parameter set; An optimization module, configured to construct a multi-dimensional input data set according to the probability density distribution parameter set, and perform iterative optimization on the multi-dimensional input data set to obtain a heat insulation performance boundary value; An output module, configured to calculate the confidence interval of the heat insulation performance boundary value, evaluate the heat insulation performance level of the heat insulation gel material based on the confidence interval, and generate a performance evaluation parameter report.

[0015] The present invention also provides a computer device, which includes a processor, a memory, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above method is implemented.

[0016] Advantageous effects: A heat insulation performance detection method for a heat insulation gel material proposed in this application, by fusing surface temperature data and pore distribution data, adopting dynamic heat flux coupling processing technology, and combining the construction of a thermal response baseline matrix and a multi-dimensional thermal impedance tensor, solves the problem in the prior art that the heat insulation performance of the heat insulation gel material cannot be comprehensively and accurately reflected, fully considers key factors such as the internal pore distribution, heat conduction path, and its dynamic thermal response of the material, so as to more accurately simulate and predict the heat insulation performance of the heat insulation material in actual applications. Further, by constructing a high-dimensional feature vector, a probability density distribution parameter set, and performing iterative optimization on the multi-dimensional input data set, the boundary value of the heat insulation performance can be obtained, and the confidence interval can be further calculated, so as to provide an accurate evaluation of the heat insulation performance of the heat insulation gel material, improve the accuracy and comprehensiveness of the heat insulation performance detection, make the detection process more in line with the actual use conditions, and avoid the deficiency that the static method cannot reflect the dynamic characteristics of the material.

[0017] In summary, the detection method of the present application improves the accuracy of heat insulation performance evaluation, overcomes the problem of insufficient consideration of the dynamic characteristics of materials in the prior art, and can more comprehensively and accurately reflect the comprehensive heat insulation effect of the heat insulation gel material in actual use. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a schematic diagram of the steps of a method for detecting the heat insulation performance of a heat insulation gel material according to an embodiment of the present invention; Figure 2 is a schematic block diagram of the structure of a system for detecting the heat insulation performance of a heat insulation gel material according to an embodiment of the present invention; Figure 3 is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0019] The implementation, functional features, and advantages of the objectives of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] In order to make the objectives, technical solutions, and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0021] Those skilled in the art of the present technology can understand that unless specifically stated otherwise, the singular forms "a", "an", "the above", and "the" used herein may also include the plural forms. It should be further understood that the term "including" used in the specification of the present invention means the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or their groups. It should be understood that when an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any one of the listed items and all combinations of related items.

[0022] Those skilled in the art of the present technology can understand that unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as the general understanding of those of ordinary skill in the field to which the present invention belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless specifically defined as here.

[0023] Refer to Figure 1, an embodiment of the present invention provides a method for detecting the heat insulation performance of a heat insulation gel material, and the method includes: S1: Obtain the surface temperature data and the internal pore distribution data of the heat insulation gel material sample, and fuse the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; In step S1, temperature sensing devices such as an infrared thermal imager or a thermocouple are used to measure and collect the physical properties of the heat insulation gel material. The infrared thermal imager can monitor the temperature distribution on the surface of the sample in a non-contact manner and obtain a two-dimensional temperature field map. In actual operation, multiple measurements can be carried out at different positions and different temperature conditions of the sample to ensure that the obtained temperature data has sufficient representativeness and accuracy. For obtaining the internal pore distribution data of the sample, techniques such as scanning electron microscopy (SEM) or X-ray computed tomography (CT) can be used to obtain the pore structure information inside the heat insulation gel material. The scanning electron microscope can provide high-resolution images and obtain the pore structure inside the sample, including characteristics such as the distribution, morphology, and relative size of the pores. CT scanning can non-destructively obtain the overall internal pore distribution information of the sample through three-dimensional imaging technology, thereby more intuitively showing the three-dimensional pore network structure of the material. When the surface temperature data and the pore distribution data are collected, these two different types of data are fused to obtain a comprehensive thermal response baseline matrix. The matrix contains multi-dimensional information of the material under different thermal response conditions. The fusion process can adopt various data processing techniques, such as principal component analysis (PCA) or multivariate regression analysis, to transform the surface temperature and pore distribution data into a unified mathematical model.

[0024] S2: Perform dynamic heat flux coupling processing on the thermal response baseline matrix to obtain a dynamic coupling characteristic sequence; In step S2, the dynamic heat flux coupling processing takes into account that during the heat flux change process of the material, the heat flux not only enters through the surface of the material but is also affected by the internal structure of the material, especially the distribution and size of the pores. These factors will change the heat flux path and conduction efficiency. By numerically simulating the dynamic heat flux on the thermal response baseline matrix, it can be achieved by calculating the change of heat flux, heat transfer, and heat dissipation. The numerical methods include finite element analysis (FEA) or computational fluid dynamics (CFD) simulation, etc. The temperature change of the material is captured at different time nodes, and the heat conduction characteristics of the material in a dynamic thermal environment are accurately reflected. Through dynamic heat flux coupling analysis, the thermal response data at each time point can be obtained, and combined with the temperature data under different heat flux conditions, a dynamic coupling characteristic sequence is formed, which reflects the performance change of the heat insulation gel material under dynamic heat flux conditions, includes the temperature response of the material at different time points, and combines the change of the internal heat conduction path of the material.

[0025] S3: Construct a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence, and perform regularization constraints on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series; In step S3, calculate the thermal impedance tensor from the dynamic coupling feature sequence, that is, convert the heat flux and temperature change data at each time point into impedance values. The multi-dimensional thermal impedance tensor is a multi-dimensional matrix, which represents the thermal impedance characteristics of the material in multiple dimensions (such as temperature, pore distribution, thermal conductivity, etc.) under different heat flux conditions. During the construction of the thermal impedance tensor, regularization constraints are applied to smooth the high-dimensional data, making the obtained thermal impedance time series more stable and having better interpretability. When implementing the regularization constraints, it can include L1 regularization (Lasso) or L2 regularization (Ridge), which control the complexity of the model by penalizing overly large coefficients. For example, L2 regularization applies a penalty term to each element in the tensor, so that the magnitude of the thermal impedance value is not too large, thus avoiding performance degradation caused by overfitting. In specific implementation, a numerical optimization algorithm can be used to optimize the multi-dimensional thermal impedance tensor in combination with the actually measured dynamic coupling feature sequence, ensuring that the regularized thermal impedance tensor can reflect the true heat conduction characteristics of the material under various heat flux conditions. The thermal impedance time series obtained after the above regularization optimization steps reflects the thermal insulation performance of the thermal insulation gel material at different time points and under different heat flux conditions.

[0026] S4: Extract the non-linear thermal shielding index from the thermal impedance time series, construct a high-dimensional feature vector based on the non-linear thermal shielding index, and generate a probability density distribution parameter set; In step S4, from the regularized thermal impedance time series, features that can accurately reflect the non-linear changes in the material's heat conduction characteristics are extracted. These features are related to changes in multiple factors such as the material's heat conduction rate, heat flux distribution, and temperature response. Through mathematical modeling, the non-linear response pattern of the material under the action of heat flux can be identified from the time series. This process can use methods such as non-linear regression and neural networks to capture the non-linear trends in the data, obtain the non-linear thermal shielding index, and construct a high-dimensional feature vector based on this index. The feature vector is a data structure that comprehensively describes the material's heat conduction performance, expressing multiple key heat response features in vector form. In this step, by combining the non-linear thermal shielding index with other relevant heat response features (such as pore structure, temperature gradient, etc.), a multi-dimensional feature vector is constructed. This vector can more comprehensively reflect the heat insulation ability of the material under different heat flux conditions, and a set of probability density distribution parameters is generated based on this vector. The probability density distribution reflects the distribution of the material's heat insulation ability under different heat flux conditions. By performing statistical analysis on the high-dimensional feature vector, the corresponding probability density distribution can be generated. This distribution can characterize the fluctuations in the heat insulation performance of the material under different environmental conditions. The generated parameter set contains the probability distribution of the changes in the heat insulation ability of the material under various conditions.

[0027] S5: Construct a multi-dimensional input data set according to the set of probability density distribution parameters, and perform iterative optimization on the multi-dimensional input data set to obtain the boundary value of the heat insulation performance; In step S5, information such as the thermal response data and thermal shielding index in the probability density distribution parameter set is organized into an input form suitable for machine learning or optimization algorithms. Specifically, these data are constructed into a data set containing multiple dimensions (such as temperature, heat flux, porosity, etc.), and each dimension represents an independent physical property or environmental factor, and then enters the iterative optimization stage. The goal of iterative optimization is to optimize the evaluation model of the heat insulation performance based on the constructed multi-dimensional input data set, and finally obtain the boundary value of the heat insulation performance. This process can use some optimization algorithms, such as gradient descent method, genetic algorithm, simulated annealing, etc. These optimization algorithms can search for the optimal boundary value by continuously adjusting the parameters in the input data set in multiple possible solution spaces. In this process, each iteration will adjust the input data according to the existing results, gradually approaching the optimal performance evaluation model. It should be noted that during the optimization process, some constraints need to be considered. For example, some materials may exhibit relatively stable thermal insulation performance under specific environmental conditions, while large fluctuations may occur under other conditions. Therefore, when performing iterative optimization, the optimization algorithm not only needs to find the best boundary value, but also needs to ensure that the obtained boundary value has certain physical meaning and practicality to ensure the authenticity and reliability of the final performance evaluation result. Through this series of iterative optimization processes, the obtained boundary value of the heat insulation performance will provide a quantitative basis for evaluating the actual application performance of the heat insulation gel material, reflecting the heat insulation ability of the material under different heat flux and temperature conditions.

[0028] S6: Calculate the confidence interval of the boundary value of the heat insulation performance, evaluate the heat insulation performance level of the heat insulation gel material based on the confidence interval, and generate a performance evaluation parameter report.

[0029] In step S6, the confidence interval is a way to infer the range of population parameters based on sample data, which can determine the probability that the actual heat insulation performance boundary value falls within a certain interval at a certain confidence level. When calculating the confidence interval, the statistical methods used can include the normal distribution hypothesis, t-distribution, etc. Specifically, assuming that the distribution of the heat insulation performance boundary value satisfies the normal distribution or an approximate normal distribution, using the mean, standard deviation, and sample size of the sample, the confidence interval can be calculated through a formula. For example, using the z-distribution or t-distribution calculation formula, the confidence interval of the heat insulation performance boundary value can be deduced. This calculation process includes the mean of the data, the volatility of the data, and the sample size, and evaluates the heat insulation performance level of the heat insulation gel material through the confidence interval. According to the obtained boundary value and its confidence interval, the performance of the material within a certain range can be judged. For example, assuming that the heat insulation performance boundary value obtained through step S5 is 10 W / m·K, and the confidence interval obtained through statistical analysis is 9.5 W / m·K to 10.5 W / m·K, that is, at a 95% confidence level, the heat insulation performance boundary value is between 9.5 W / m·K and 10.5 W / m·K. This analysis result shows that even under different experimental environments and conditions, the change range of the heat insulation performance of the material can be effectively controlled within this interval, so it has high stability and reliability. Based on this confidence interval, the heat insulation performance level of the heat insulation gel material can be further evaluated, which can be achieved by quantitatively grading the performance of the material. The evaluation criteria can be based on different industrial standards or experimental requirements, such as dividing different grades according to different thermal resistance value intervals. For example, a heat insulation performance of 10 W / m·K can be set as a Class A heat insulation material, and other values within the confidence interval of 9.5 W / m·K to 10.5 W / m·K may correspond to Class B or Class C materials. In this way, the final evaluation result provides a clear reference for the performance grade of the material. Based on this performance evaluation, a performance evaluation parameter report is generated, which includes the heat insulation performance boundary value within the confidence interval, and assigns a clear performance grade to the heat insulation gel material according to industry standards or a preset evaluation model. The report can also include other detailed data on the material performance, such as the performance change under different test conditions, the heat response characteristics of the material, etc., providing a theoretical basis for subsequent material optimization or improvement.

[0030] In one embodiment, the step of obtaining the surface temperature data and the internal pore distribution data of the heat insulation gel material sample and fusing the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix includes: Based on the two-dimensional distribution data of the sample surface temperature changing with time collected by an infrared thermal imager, discretize the two-dimensional distribution data to generate a raw data stream of the surface temperature containing temperature spatio-temporal evolution information; Decompose the original surface temperature data stream into multiple sub-band signals according to the time series, extract the characteristic frequency peaks of each sub-band signal, and generate a set of thermal response frequency components; Obtain the scattering data of the pore structure inside the thermal insulation gel material sample through X-ray scattering processing, calculate the anisotropy eigenvalue of the scattering data, and obtain the pore anisotropy vector; Calculate the boundary fractal dimension of the pore structure according to the pore anisotropy vector, and generate a set of pore fractal parameters containing multi-scale fractal information; Construct an initial feature matrix, map the set of thermal response frequency components and the set of pore fractal parameters to the initial feature matrix, and generate the thermal response baseline matrix.

[0031] In the above embodiments, when obtaining the surface temperature data of the thermal insulation gel material sample through an infrared thermal imager, the infrared thermal imager has the ability to capture the surface temperature distribution of an object in real time, and the collected data is a two-dimensional temperature distribution that changes over time. The two-dimensional data here represents the temperature values of each region on the sample surface at different times, reflecting the spatio-temporal evolution of the surface temperature of the sample under the action of heat flux. The discretization process converts the original continuous temperature data into discrete numerical points, forming the original data stream of the surface temperature. In this process, the data is gridded or quantified through appropriate mathematical methods to ensure appropriate resolution in both the spatial and temporal dimensions, including the absolute value of the temperature change, and the trend and spatial distribution pattern of the temperature change over time. The original temperature data stream is decomposed into multiple sub-band signals, and these signals represent the temperature change characteristics at different time scales. Through this decomposition, rapid and slow temperature fluctuations can be captured. The characteristic frequency peak of each sub-band signal reflects the main periodic component of the temperature change within that frequency band. After extracting these frequency peaks, a set of thermal response frequency components is obtained, including the thermal response characteristics of the sample at different time scales. The pore distribution data inside the sample is obtained through X-ray scattering technology. In this embodiment, X-rays are directed at the sample, and the pore structure inside the sample scatters the X-rays. These scattering data can reveal the morphology, size, and arrangement of the pores in different directions, and analyze the degree of asymmetry or heterogeneity of the pore structure of the material in different directions. Through analysis, a pore anisotropy vector is obtained, representing the directional characteristics of the pore structure inside the sample. The boundary fractal dimension of the pore structure is calculated based on the pore anisotropy vector. The fractal dimension is a parameter used to describe the geometric shape with self-similarity and complex structure. Through specific mathematical methods, the fractal dimension of the pore structure boundary can be calculated, thereby quantifying the complexity of the pores. This dimension can reflect the complexity of the pore morphology at different scales. The larger the fractal dimension, the stronger the surface area and irregularity of the pores, thus affecting heat transfer and thermal insulation effects. An output set of pore fractal parameters is obtained, including multi-scale fractal information of the pore structure. The set of thermal response frequency components and the set of pore fractal parameters are mapped into the initial feature matrix as different characteristic dimensions. Each dimension represents a specific performance index, combining temperature response data and pore structure data, generating a thermal response baseline matrix. This matrix provides a multi-dimensional reference framework for the thermal insulation performance detection of the thermal insulation gel material, and can comprehensively evaluate the thermal insulation ability of the material from both the thermal response and material structure aspects.

[0032] In one embodiment, the step of performing dynamic heat flux coupling processing on the thermal response baseline matrix to obtain a dynamic coupling characteristic sequence includes: Dividing the thermal response baseline matrix into multiple sub-regions, and obtaining heat flux modulation distribution data by applying heat flux modulation signals with different amplitudes to each sub-region; Construct a heat flux probe array, and generate heat flux distribution data according to the layer-by-layer distribution characteristics of the heat flux transmitted from the surface to the interior collected by the heat flux probe array; Extract the frequency characteristics of the heat flux distribution data, analyze the difference in the energy spectral density between the frequency characteristics and the surface temperature data, and generate cross-spectral density distribution data; Perform point-by-point correction on the cross-spectral density distribution data according to the pore distribution data in the order of frequency to generate a dynamic coupling feature sequence.

[0033] In the above embodiments, the thermo-responsive baseline matrix is divided into multiple sub-regions, and the thermo-responsive characteristics of the entire sample are subdivided into several parts, which can be carried out according to the physical properties of the sample or the geometric shape of the material. For example, the division can be made according to the thickness, porosity, surface morphology, etc. of the material, and heat flux modulation signals with different amplitudes are applied. The heat flux modulation signal refers to simulating different heat load conditions by changing the intensity or frequency of the heat flux, which can stimulate the thermo-responses of different regions of the material, and the responses of each sub-region under the action of the heat flux are recorded. Specifically, the heat flux modulation signals with different amplitudes applied will cause temperature changes on the surface and inside of the sample, thereby forming a series of heat flux modulation distribution data. A heat flux probe array is constructed. The heat flux probe array consists of multiple detectors, and these detectors are arranged on the surface or inside of the sample in the form of a sensor array for measuring the heat flux characteristics transmitted from the surface of the sample to the inside, that is, the thickness direction. Through these probe arrays, the heat flux transfer process at different depth positions inside the sample can be obtained, the propagation characteristics of the heat flux between different layers can be obtained, and the heat insulation effect of the material is reflected. Frequency features are extracted from these heat flux distribution data. The frequency features are periodic information describing the change of the heat flux signal over time, indicating the intensity of different frequency components in the heat flux transfer process. Signal processing methods such as Fourier transform can be used to convert the time-domain signal into a frequency-domain signal, and then the energy distribution at different frequencies can be obtained. Analyze the difference in the energy spectral density between these frequency features and the surface temperature data. The energy spectral density is a measure describing the distribution of the signal in the frequency domain, reflecting the energy information contained in different frequency components. In this embodiment, analyze the difference between the energy spectral density of the heat flux signal and the energy spectral density of the surface temperature signal, identify which frequency components play a dominant role in the heat conduction behavior of the sample during the thermo-response process, and which components may reflect the characteristics of the heat insulation gel material, and obtain the cross-spectral density distribution data. The cross-spectral density distribution data indicates the interaction relationship between the heat flux and the surface temperature signal, and these data are for the energy transfer characteristics between different frequency components and the heat transfer characteristics of different layers inside the material. The cross-spectral density distribution data is corrected point by point according to the pore distribution data obtained in the foregoing steps. Specifically, a decay weight function is constructed to apply weighted adjustment to the cross-spectral density distribution, and the cross-spectral density is corrected point by point in the order of frequency, highlighting the non-linear heat flux effect, extracting key coupling features, generating a smooth dynamic coupling feature sequence, further improving the coupling relationship between the heat flux and the temperature, and making the corrected data more in line with the actual heat transfer situation.

[0034] In one embodiment, the step of correcting the cross-spectral density distribution data point by point in the order of frequency according to the pore distribution data to generate a dynamic coupling feature sequence further includes: Perform weighted processing on each frequency point of the cross-spectrum density distribution data based on the spatial heterogeneity of the pore distribution data to generate weighted cross-spectrum density data; Set a sliding window with a preset number of digits, control the sliding window to slide along the frequency axis on the weighted cross-spectrum density data, and extract local feature vectors within each sliding window; Sort multiple local feature vectors according to similarity and perform clustering analysis to generate a set of clustering center feature vectors; Perform global optimization on the weighted cross-spectrum density data according to the set of clustering center feature vectors to obtain an optimized dynamic coupling feature sequence.

[0035] In the above embodiment, weighted processing is performed on each frequency point of the cross-spectrum density distribution data based on the spatial heterogeneity of the pore distribution data. By assigning different weights to different frequency points, the cross-spectrum density distribution data receives more attention and correction in regions with strong spatial heterogeneity, thus better reflecting the thermal response characteristics of these regions. Perform sliding window analysis on the weighted cross-spectrum density data. Set a sliding window with a preset number of digits and control the sliding window to slide along the frequency axis on the weighted cross-spectrum density data. By controlling the number of digits of the sliding window, the data range extracted each time can be adjusted so that each window on the frequency axis can capture local frequency characteristics. As the window slides along the frequency axis, local feature vectors in different frequency bands can be gradually extracted. These feature vectors represent the change trend and characteristics of the cross-spectrum density in that frequency band and reflect the heat transfer characteristics of the heat flow in that frequency range. Sort the local feature vectors according to similarity and perform clustering analysis. The extracted local feature vectors are divided into several categories according to similarity, and each category represents a set of frequency ranges with similar heat conduction characteristics. Specifically, the similarity between local feature vectors can be measured by calculating the distance between them (such as Euclidean distance or cosine similarity). Then, sort these feature vectors according to similarity to find sets of feature vectors with high similarity in heat transfer characteristics. Similar feature vectors will be aggregated to form a set of clustering center feature vectors, and these clustering center feature vectors represent typical heat conduction behaviors in different frequency ranges. Further, use this set of clustering center feature vectors to perform global optimization on the weighted cross-spectrum density data. By combining the feature vectors of the clustering center with the original weighted cross-spectrum density data, the final obtained dynamic coupling feature sequence can more comprehensively and accurately reflect the heat conduction characteristics of the material. During the optimization process, the deviation in the data can be continuously adjusted through an iterative update method to ensure that the optimized feature sequence can better balance the characteristics of each frequency point and obtain more practical heat insulation performance data.

[0036] In one embodiment, the step of constructing a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence and performing regularization constraint on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series includes: Decompose the dynamic coupling feature sequence into multiple spatial resolution levels, traverse the fluctuation characteristics of each level, and generate feature distribution data; Fuse the heat flux component and temperature component of each level in the feature distribution data into a high-order tensor, and stack the distribution characteristics of different scales layer by layer to generate an initial multi-dimensional tensor; Calculate the singular value distribution of each dimension in the initial multi-dimensional tensor to generate a set of thermal impedance components including multiple types of components, and the set of thermal impedance components includes at least scalar components and directional components; Construct a fractal constraint function according to the pore distribution characteristics in the thermal response baseline matrix, embed the fractal constraint function into the scalar components and directional components and adjust the regularization weights to generate a regularized thermal impedance tensor; Perform evolutionary simulation processing on the regularized thermal impedance tensor according to the time step to obtain the thermal impedance time series.

[0037] In the above embodiments, the dynamically coupled feature sequence is decomposed into multiple spatial resolution levels, and each level represents the feature performance at different scales, covering the heat conduction characteristics from the macroscopic to the microscopic level. Among these levels, the fluctuation characteristics at different scales reflect the changes in heat flow and temperature of the material at different resolutions, and the fluctuation characteristics of each level are traversed to generate feature distribution data. Further, the heat flow component and the temperature component in each level are combined together to form a high-order tensor. The heat flow component and the temperature component are two basic quantities in the description of the heat conduction process, respectively reflecting the intensity and directionality of the heat flow and the state of the temperature distribution. In each level, the two are integrated through fusion into a more complex tensor, which reflects the mutual relationship between the heat flow and the temperature. By stacking the distribution characteristics at different scales layer by layer, the multi-level local information is combined together, and then an initial multi-dimensional tensor is obtained, which represents the heat response behavior of the material at different spatial resolutions. Calculate the singular value distribution of each dimension in the initial multi-dimensional tensor. The singular value is the weight factor of each dimension in the tensor, revealing the importance of different dimensions to the entire heat conduction process. Through the calculation of the singular value distribution, different components of the tensor can be classified, and a set of heat impedance components containing various types of components is generated. Among these components, there are at least a scalar component and a direction component. The scalar component represents the magnitude of the heat impedance, and the direction component reflects the distribution of the heat impedance in different directions. A fractal constraint function is constructed according to the pore distribution characteristics in the heat response baseline matrix. By embedding the fractal constraint function into the scalar component and the direction component, the distribution characteristics of the pores can be introduced into the calculation process of the heat impedance, so as to ensure that the heat conduction characteristics can more realistically reflect the microstructure of the material. On this basis, the heat impedance tensor is further optimized by adjusting the regularization weight. The purpose of regularization is to reduce the overfitting phenomenon and ensure that the model has better generalization ability. The regularized heat impedance tensor needs to be processed by evolution simulation according to the time step, and the heat insulation performance of the material during actual use is reflected by simulating the dynamic change of the heat flow. Through simulation, the obtained heat impedance time series can accurately describe the heat response change of the material over a period of time.

[0038] In one embodiment, the steps of extracting the non-linear heat shielding index in the heat impedance time series, constructing a high-dimensional feature vector based on the non-linear heat shielding index, and generating a set of probability density distribution parameters include: Extract the non-linear fluctuation characteristics of the heat impedance time series in the time dimension to generate a non-linear response distribution sequence; Extract the spectral density peak in the dynamically coupled feature sequence according to the frequency information in the non-linear response distribution sequence, perform Fourier transform on the spectral density peak to obtain multiple frequency domain components, and perform spectral reconstruction processing on the multiple frequency domain components to obtain a composite feature spectrum; Map the spectral components of the composite feature spectrum to the fractal space, calculate the fractal dimension features of each spectral component, and obtain the thermal shielding index set; Construct a high-dimensional feature vector based on the thermal shielding index set, where the high-dimensional feature vector includes the thermal shielding index set and its evolution information in the time dimension; Use the kernel density estimation method to estimate the probability density distribution of the high-dimensional feature vector and generate a set of probability density distribution parameters.

[0039] In the above embodiment, the non-linear fluctuation features are extracted from the thermal impedance time series, and a non-linear response distribution sequence is generated. The thermal impedance time series contains various fluctuation characteristics in different time dimensions during the heat transfer process. By analyzing these fluctuation features, the non-linear features in the heat conduction process can be captured. Based on the frequency information in the obtained non-linear response distribution sequence, the spectral density peaks in the dynamic coupling feature sequence are extracted, which are the main energy concentration regions in the heat conduction process. These peaks are transformed from the time-domain data to the frequency domain through Fourier transform to obtain multiple frequency-domain components. Combining multiple frequency-domain components together forms a more comprehensive feature description, constituting a composite spectral feature description, and mapping the composite feature spectrum to the fractal space to calculate the fractal dimension features of each spectral component. The fractal space is a mathematical space for analyzing complex structures. By mapping the spectral components to the fractal space, the fractal dimension features of each spectral component can be calculated, which can effectively quantify the complexity of the material's thermal response and generate the thermal shielding index set. By combining the time evolution information with the thermal shielding index, a complete high-dimensional feature vector can be obtained, which can more comprehensively describe various dynamic behaviors of the material during the heat conduction process. Performing probability density estimation on the high-dimensional feature vector gives a set of probability density distribution parameters, which can be used to describe the distribution of the material's heat conduction characteristics in the entire feature space and identify the potential performance of the material under different conditions.

[0040] In one embodiment, the step of constructing a multi-dimensional input data set based on the set of probability density distribution parameters and iteratively optimizing the multi-dimensional input data set to obtain the thermal insulation performance boundary value includes: Split the set of probability density distribution parameters into statistical distribution components of different scales according to a preset scale threshold and integrate them to form a parameter matrix; Set an initial input framework, and perform topological reconstruction processing on the initial input framework according to the parameter matrix to obtain a multi-dimensional input data set; Perform convolution processing on each dimension of the multi-dimensional input data set in turn, extract local correlations and compress redundant information to obtain a compressed feature sequence; Construct an iterative model, adjust the weights and biases of the iterative model based on the distribution characteristics of the compressed feature sequence, update the parameters layer by layer based on the gradient feedback mechanism, and generate an optimized network model; Input the compressed feature sequence in segments into the fully connected layer of the optimized network model, perform feature mapping through a non-linear activation function, and obtain a boundary prediction feature set; Construct a heat flux response function, perform weighted adjustment on the heat flux response function based on the distribution characteristics of the boundary prediction feature set, and calculate the boundary values point by point along the domain of definition of the weighted adjusted heat flux response function based on the constrained optimization algorithm to obtain the heat insulation performance boundary values.

[0041] In the above embodiments, the parameter set is split into statistical distribution components of different scales according to a preset scale threshold, different statistical features are extracted from the overall probability distribution, and they are integrated to form a parameter matrix. By splitting and integrating the probability density distribution parameter set, the multi-level understanding of the thermal performance can be refined. Specifically, multiple preset scale thresholds can be set, such as small scale, medium scale, and large scale. Each scale threshold corresponds to different statistical distribution characteristics, so that the changes in the heat insulation performance of the heat insulation gel material under different temperature gradients can be captured. Based on this parameter matrix, an initial input framework is set, and the input framework is processed through topological reconstruction to adjust the structure of the original data to meet the requirements of the optimization algorithm, so that the data set can better express the heat conduction characteristics of the material in the multi-dimensional space. The multi-dimensional input data set is subjected to convolution processing to extract local correlations from complex data and identify the most representative local patterns and heat response features in the data. At the same time as convolution, redundant information is compressed to reduce unnecessary data volume, so that the model can focus on the features most influential on the heat insulation performance. After obtaining the compressed feature sequence, an iterative model is constructed. The model adjusts the weights and biases of the model based on the distribution characteristics of the compressed feature sequence. The iterative model updates the parameters layer by layer through the gradient feedback mechanism to optimize the prediction ability of the model. The gradient feedback mechanism calculates the error of each layer and adjusts the parameters based on these errors, thereby gradually reducing the deviation between the prediction result and the actual result. Through repeated adjustment and optimization, an optimized network model is generated, which can accurately reflect the thermal performance of the heat insulation material. The fully connected layer of the optimized network model inputs the compressed feature sequence in segments, and each segment input is processed by a non-linear activation function for feature mapping. The role of the non-linear activation function is to introduce non-linear features, enabling the network to better handle complex relationships, and through this mapping, the input data is transformed into a set of features suitable for subsequent prediction. Through this process, the optimized network model can generate a boundary prediction feature set, which describes the heat conduction boundary of the heat insulation gel material. A heat flow response function is constructed. The initial heat flow response function can be constructed by fitting the heat conduction equation or based on historical experimental data. According to the distribution characteristics of the boundary prediction feature set, the heat flow response function is weighted and adjusted, and different weights are assigned to different parts of the heat flow response, so that the model can more accurately predict the heat insulation performance. After the heat flow response function is adjusted, a constrained optimization algorithm is used to calculate the boundary values of the heat insulation performance point by point along its domain. The constrained optimization algorithm is an algorithm for solving the optimal solution under specific constraint conditions, and finds the boundary value most in line with the actual situation in the complex heat flow response function.

[0042] In another embodiment, the calculation expression of the above embodiment is: ; Among them, B is the boundary value of heat insulation performance, representing the thermal conductivity limit of the heat insulation material under specific conditions (the unit can be thermal resistance or heat flux density); is the integral from the initial temperature (0) to the target temperature (T), representing the cumulative calculation of the heat flux response over the entire temperature range, simulating the dynamic process of thermal performance changing with temperature, and capturing the global characteristics of the boundary value; T is the upper limit of the target temperature (unit: °C); Φ is a non-linear activation function, representing the non-linear mapping function in the optimized network model; is to accumulate the features of N dimensions in the multi-dimensional input data set; N is the total number of dimensions, representing the complexity of the multi-dimensional input data set (an integer, for example, small, medium, and large scales correspond to N = 3); is the scale weight coefficient, representing the weight of the statistical distribution component of the i-th dimension (range: 0 to 1); is the statistical variance, representing the variance of the statistical distribution component of the i-th dimension; is the exponential decay term, the decay factor based on the gradient feedback, representing the compression effect of local correlation in the convolution process; β is the decay rate (a positive real number), is the convolution gradient of the i-th dimension, characterizing the change rate of local correlation; is the heat flux response weighting function, the heat flux response weight varying with the temperature T (range: 0 to 1); λ is the constraint adjustment coefficient, that is, the adjustment factor of the constraint optimization algorithm (a positive real number); is the hyperbolic tangent correction term, the non-linear correction based on the partial derivative of the heat flux response function H with respect to the feature set F; H is the heat flux response function, describing the change of heat flux with the boundary characteristics (unit: W / m²); F is the boundary prediction feature set, generated by the optimized network model; is the partial derivative of the heat flux response with respect to the feature set, reflecting the sensitivity of the feature change to the heat flux; dT is the temperature differential, that is, the temperature increment in the integral. This expression is used to calculate the heat insulation performance boundary value B of the heat insulation gel material. By integrating the multi-scale features of the probability density distribution, the local correlation after convolution compression, the feature mapping of the iterative optimized network, and the weighted adjustment of the heat flux response function, the boundary performance prediction at different temperature gradients is finally output.

[0043] In one embodiment, the steps of calculating the confidence interval of the heat insulation performance boundary value, evaluating the heat insulation performance grade of the heat insulation gel material based on the confidence interval, and generating a performance evaluation parameter report include: Sampling the heat insulation performance boundary value multiple times to generate a series of simulated boundary value data; Calculating the mean and variance of the simulated boundary value data, and determining the confidence interval of the heat insulation performance boundary value according to the mean and variance; Build a heat insulation performance evaluation model, and classify the heat insulation performance of the heat insulation gel material based on the distribution characteristics of the boundary values within the confidence interval; Generate a performance evaluation parameter report according to the classification result, and the performance evaluation parameter report at least includes the heat insulation performance grade, the confidence interval range, and relevant evaluation indicators.

[0044] In the above embodiment, multiple samplings are performed on the heat insulation performance boundary values. Here, the boundary values refer to the maximum or minimum values that the heat insulation performance may reach under different conditions. Through multiple samplings, a series of simulated boundary value data can be generated, and these data can reflect the changes in heat insulation performance under different conditions, generating simulated boundary value data. It should be noted that the simulated boundary value data will be adjusted based on the actual measured values under different experimental conditions to better simulate the changes in heat insulation performance in various different situations. Based on these simulated boundary value data, the mean and variance are calculated. The mean is a measure of the central tendency of the data set and represents an average level of heat insulation performance under different experimental conditions. The variance is an important indicator to measure the volatility and dispersion of the data, reflecting the range of variations in heat insulation performance in multiple experiments. According to the mean and variance of the simulated data, a normal distribution or other appropriate statistical distribution model is used to calculate an interval range that includes the true value of the heat insulation performance boundary value. Usually, a confidence level (such as 95% or 99%) can be set to determine this confidence interval, that is, within this interval, there is a 95% or 99% probability that the true boundary value of the heat insulation performance falls into it. In this way, the uncertainty in the experiment can be taken into account, and a more reliable heat insulation performance evaluation result can be obtained. Build a heat insulation performance evaluation model based on the distribution characteristics of the boundary values within the confidence interval, that is, within the confidence interval, how the boundary values change and how to match the relationship with the heat insulation performance grade. The core goal of the evaluation model is to correspond different boundary value ranges to different heat insulation performance grades, thus forming a quantifiable heat insulation performance evaluation standard. The classification of the heat insulation performance grade is based on preset standards and industry specifications. For example, according to the boundary values within a certain specific range, it can be classified into different grades such as excellent, good, medium, and poor. Generate a performance evaluation parameter report according to the determined heat insulation performance grade. This report is the final document summarizing and generalizing the performance of the heat insulation gel material and can provide comprehensive performance data for R & D personnel, manufacturers, or end users. The performance evaluation parameter report at least includes the heat insulation performance grade, the confidence interval range, and relevant evaluation indicators. The heat insulation performance grade provides an intuitive evaluation of the heat insulation effect of the material, the confidence interval range shows the stability and reliability of the heat insulation performance of the material under different experimental conditions, and the relevant evaluation indicators include quantifiable thermal parameters such as temperature difference and heat flux density. These indicators help to deeply understand the heat insulation characteristics of the material and guide the selection and use of the material in practical applications.

[0045] Reference Figure 2 , a heat insulation performance detection system for a heat insulation gel material, which is applied to the above detection method. The heat insulation performance detection system includes: An acquisition module 100, configured to acquire surface temperature data and internal pore distribution data of a heat insulation gel material sample, fuse the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; A coupling module 200, configured to perform dynamic heat flow coupling processing on the thermal response baseline matrix to obtain a dynamic coupling feature sequence; A regularization module 300, configured to construct a multi-dimensional thermal impedance tensor based on the dynamic coupling feature sequence, and perform regularization constraints on the multi-dimensional thermal impedance tensor to obtain a thermal impedance time series; An extraction module 400, configured to extract a non-linear heat shielding index from the thermal impedance time series, construct a high-dimensional feature vector based on the non-linear heat shielding index, and generate a probability density distribution parameter set; An optimization module 500, configured to construct a multi-dimensional input data set according to the probability density distribution parameter set, perform iterative optimization on the multi-dimensional input data set to obtain a heat insulation performance boundary value; An output module 600, configured to calculate a confidence interval of the heat insulation performance boundary value, evaluate the heat insulation performance level of the heat insulation gel material based on the confidence interval, and generate a performance evaluation parameter report.

[0046] Reference Figure 3 , an embodiment of the present application further provides a computer device, which may be a server, and its internal structure may be as Figure 3 shown. The computer device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data and other information related to the present application. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a heat insulation performance detection method for a heat insulation gel material.

[0047] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for detecting the thermal insulation performance of a thermal insulation gel material, characterized in that: The method comprises: Acquiring surface temperature data and internal pore distribution data of a thermal insulation gel material sample, and fusing the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; Performing dynamic heat flow coupling processing on the thermal response baseline matrix to obtain a dynamic coupling characteristic sequence; Constructing a multidimensional thermal impedance tensor based on the dynamic coupling characteristic sequence, and performing regularization constraints on the multidimensional thermal impedance tensor to obtain a thermal impedance time series; Extracting a nonlinear thermal shielding index from the thermal impedance time series, constructing a high-dimensional feature vector based on the nonlinear thermal shielding index, and generating a probability density distribution parameter set; Constructing a multidimensional input data set according to the probability density distribution parameter set, and iteratively optimizing the multidimensional input data set to obtain a thermal insulation performance boundary value; The confidence interval of the thermal insulation performance boundary value is calculated, the thermal insulation performance level of the thermal insulation gel material is evaluated based on the confidence interval, and a performance evaluation parameter report is generated.

2. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of obtaining surface temperature data and internal pore distribution data of the thermal insulation gel material sample, and fusing the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix includes: Collecting two-dimensional distribution data of sample surface temperature changing with time based on an infrared thermal imager, discretizing the two-dimensional distribution data, and generating a surface temperature raw data stream containing temperature spatiotemporal evolution information; Decomposing the surface temperature raw data stream into multiple sub-band signals in time series, extracting the characteristic frequency peak of each sub-band signal, and generating a thermal response frequency component set; Acquire scattering data of the pore structure inside the thermal insulation gel material sample through X-ray scattering processing, calculate the anisotropic characteristic value of the scattering data, and obtain the pore anisotropy vector; Calculating the boundary fractal dimension of the pore structure according to the pore anisotropy vector to generate a pore fractal parameter set containing multi-scale fractal information; An initial characteristic matrix is ​​constructed, and the thermal response frequency component set and the pore fractal parameter set are mapped to the initial characteristic matrix to generate the thermal response baseline matrix.

3. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of subjecting the thermal response baseline matrix to dynamic heat flow coupling processing to obtain a dynamic coupling characteristic sequence comprises: Dividing the thermal response baseline matrix into a plurality of sub-regions, obtaining heat flux modulation signals of different amplitudes applied to each sub-region, and generating heat flux modulation distribution data; Constructing a heat flow probe array, collecting layer-by-layer distribution characteristics of heat flow transferred from the surface to the inside according to the heat flow probe array, and generating heat flow distribution data; Extracting the frequency characteristics of the heat flux distribution data, analyzing the difference in energy spectrum density between the frequency characteristics and the surface temperature data, and generating cross-spectral density distribution data; The cross-spectral density distribution data is corrected point by point in frequency order according to the pore distribution data to generate a dynamic coupling characteristic sequence.

4. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 3, characterized in that: The step of performing point-by-point correction on the cross-spectral density distribution data in frequency order according to the pore distribution data to generate a dynamic coupling characteristic sequence also includes: Performing weighted processing on each frequency point of the cross-spectral density distribution data based on the spatial heterogeneity of the pore distribution data to generate weighted cross-spectral density data; Setting a sliding window of a preset number of bits, controlling the sliding window to slide along the frequency axis on the weighted cross-spectral density data, and extracting a local feature vector in each sliding window; Sorting the plurality of local feature vectors according to similarity, and performing cluster analysis to generate a cluster center feature vector set; The weighted cross-spectral density data is globally optimized according to the cluster center feature vector set to obtain an optimized dynamic coupling feature sequence.

5. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of constructing a multidimensional thermal impedance tensor based on the dynamic coupling characteristic sequence and performing regularization constraints on the multidimensional thermal impedance tensor to obtain a thermal impedance time series includes: Decomposing the dynamic coupling feature sequence into multiple spatial resolution levels, traversing the fluctuation characteristics of each level, and generating feature distribution data; The heat flow component and the temperature component of each level in the characteristic distribution data are merged into a high-order tensor, and the distribution features of different scales are superimposed layer by layer to generate an initial multidimensional tensor; Calculating the singular value distribution of each dimension in the initial multidimensional tensor to generate a thermal impedance component set including multiple types of components, wherein the thermal impedance component set includes at least a scalar component and a directional component; Constructing a fractal constraint function according to the pore distribution characteristics in the thermal response baseline matrix, embedding the fractal constraint function into the scalar component and the directional component and adjusting the regularization weight to generate a regularized thermal impedance tensor; The regularized thermal impedance tensor is subjected to evolution simulation processing according to a time step to obtain the thermal impedance time series.

6. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of extracting the nonlinear thermal shielding index from the thermal impedance time series, constructing a high-dimensional feature vector based on the nonlinear thermal shielding index, and generating a probability density distribution parameter set comprises: Extracting the nonlinear fluctuation characteristics of the thermal impedance time series in the time dimension to generate a nonlinear response distribution sequence; Extracting the spectrum density peak in the dynamic coupling characteristic sequence according to the frequency information in the nonlinear response distribution sequence, obtaining a plurality of frequency domain components by Fourier transforming the spectrum density peak, and performing spectrum reconstruction processing on the plurality of frequency domain components to obtain a composite characteristic spectrum; Mapping the spectral components of the composite characteristic spectrum to the fractal space, calculating the fractal dimension characteristics of each spectral component, and obtaining a heat shielding index set; Constructing a high-dimensional feature vector according to the thermal shielding index set, wherein the high-dimensional feature vector includes the thermal shielding index set and its evolution information in the time dimension; A kernel density estimation method is used to estimate the probability density distribution of the high-dimensional feature vector to generate a probability density distribution parameter set.

7. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of constructing a multidimensional input data set according to the probability density distribution parameter set, iteratively optimizing the multidimensional input data set, and obtaining a thermal insulation performance boundary value comprises: Splitting the probability density distribution parameter set into statistical distribution components of different scales according to a preset scale threshold and integrating them to form a parameter matrix; An initial input frame is set, and topological reconstruction is performed on the initial input frame according to the parameter matrix to obtain a multi-dimensional input data set; Performing convolution processing on each dimension of the multidimensional input data set in sequence, extracting local correlation and compressing redundant information to obtain a compressed feature sequence; Constructing an iterative model, adjusting the weight and bias of the iterative model based on the distribution characteristics of the compressed feature sequence, updating the parameters layer by layer based on a gradient feedback mechanism, and generating an optimized network model; Inputting the compressed feature sequence into the fully connected layer of the optimized network model in segments, performing feature mapping through a nonlinear activation function, and obtaining a boundary prediction feature set; A heat flow response function is constructed, the heat flow response function is weighted and adjusted based on the distribution characteristics of the boundary prediction feature set, and the boundary value is calculated point by point along the definition domain of the weighted adjusted heat flow response function based on a constrained optimization algorithm to obtain the boundary value of the thermal insulation performance.

8. The method for detecting the thermal insulation performance of the thermal insulation gel material according to claim 1, characterized in that: The step of calculating the confidence interval of the thermal insulation performance boundary value, evaluating the thermal insulation performance level of the thermal insulation gel material based on the confidence interval, and generating a performance evaluation parameter report includes: Sampling the thermal insulation performance boundary value multiple times to generate a series of simulated boundary value data; Calculating the mean and variance of the simulated boundary value data, and determining the confidence interval of the thermal insulation performance boundary value according to the mean and variance; Constructing a thermal insulation performance evaluation model, and grading the thermal insulation performance of the thermal insulation gel material based on the boundary value distribution characteristics within the confidence interval; A performance evaluation parameter report is generated according to the grading result, and the performance evaluation parameter report at least includes the thermal insulation performance grade, confidence interval range and related evaluation indicators.

9. A thermal insulation performance detection system for thermal insulation gel materials, applied to the method according to any one of claims 1 to 8, characterized in that: include: An acquisition module is used to acquire surface temperature data and internal pore distribution data of the thermal insulation gel material sample, and fuse the surface temperature data and the pore distribution data to obtain a thermal response baseline matrix; A coupling module, used for performing dynamic heat flow coupling processing on the thermal response baseline matrix to obtain a dynamic coupling feature sequence; A regularization module, used for constructing a multidimensional thermal impedance tensor based on the dynamic coupling characteristic sequence, and performing regularization constraints on the multidimensional thermal impedance tensor to obtain a thermal impedance time series; An extraction module, used for extracting a nonlinear thermal shielding index from the thermal impedance time series, constructing a high-dimensional feature vector based on the nonlinear thermal shielding index, and generating a probability density distribution parameter set; An optimization module, used for constructing a multidimensional input data set according to the probability density distribution parameter set, and iteratively optimizing the multidimensional input data set to obtain a thermal insulation performance boundary value; The output module is used to calculate the confidence interval of the thermal insulation performance boundary value, evaluate the thermal insulation performance level of the thermal insulation gel material based on the confidence interval, and generate a performance evaluation parameter report.

10. A computer device, characterized in that: The method comprises a processor, a memory and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 8 when executing the computer program.

Images