A data processing-based evaluation method for thermal insulation effect of thermal insulation clothing

Abstract

This invention belongs to the field of performance testing and evaluation technology for thermal protection equipment, and particularly relates to a method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing. The method involves synchronously and continuously collecting temperature and heat flux density data from the inner and outer surfaces of the thermal insulation clothing, performing spatiotemporal dual-parameter transient heat conduction model inversion processing, calculating the material's equivalent thermal diffusivity and equivalent heat storage coefficient, and obtaining real-time inner surface thermal resistance data; and acquiring dynamic physiological safety time data; obtaining thermal stability index data based on spectral entropy through spectral entropy calculation, which, together with the dynamic physiological safety time data, constitutes a multi-dimensional evaluation index dataset; and outputting the thermal insulation effect evaluation level. During the quasi-steady-state stage, two different time points are selected to establish a system of simultaneous equations for analytical solution, synchronously calculating the equivalent thermal diffusivity and equivalent heat storage coefficient from the data, overcoming the deficiency of traditional single-parameter models in being unable to characterize the material's heat storage history effect.

Description

Technical Field

[0001] This invention belongs to the field of performance testing and evaluation technology of thermal protection equipment, and in particular relates to a method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing. Background Technology

[0002] Currently, the evaluation of the thermal insulation effect of thermal insulation clothing mainly relies on measuring the time required for the inner surface temperature of the clothing to reach a fixed temperature rise threshold. This method essentially only compares temperature data based on thresholds, resulting in three fundamental flaws: First, dynamic information is severely lost. Information such as the material's heat storage effect, heat wave propagation delay, and dynamic evolution of thermal resistance during the entire unsteady heat transfer process is not utilized, making it impossible to distinguish between clothing with transiently high resistance but rapid failure and clothing with moderate thermal resistance but long-term stability. Second, the evaluation dimension is singular, compressing complex thermal insulation performance into a single time value, failing to reveal the physical essence and dynamic behavior of thermal diffusivity, heat storage coefficient, etc. Third, there is insufficient correlation with physiological safety. There is a lack of quantitative mapping between the fixed temperature rise threshold and the human skin burn process, making it impossible to accurately reflect the true physiological safety boundary.

[0003] Existing technologies employ single-parameter thermal diffusion models for identifying thermophysical parameters, failing to consider the heat storage history effect of materials in unsteady-state stages; they use simple coefficients of variation in thermal stability evaluation, lacking comprehensive quantification of multi-scale fluctuation characteristics and spatial structural uniformity; and they use fixed weights or subjective weighting in multi-index comprehensive evaluation, unable to dynamically adjust weight allocation based on the measurement uncertainty of each index. Therefore, a dynamic comprehensive evaluation method centered on data processing, capable of extracting and fusing multi-dimensional thermal insulation performance information from sensor data, is needed. Summary of the Invention

[0004] This invention addresses the technical problems existing in the performance testing and evaluation of thermal protection equipment by proposing a data processing-based method for evaluating the thermal insulation effect of thermal insulation clothing. This method is reasonable in design, simple in approach, and theoretically sound.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] S1. During the heat exposure process, temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit are collected synchronously and continuously. An improved strong tracking adaptive filter is used to filter the collected temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit. The filtering gain of the improved strong tracking adaptive filter is dynamically adjusted according to the transient change intensity index jointly defined by the temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit. The filtered inner surface temperature data sequence, the filtered outer surface temperature data sequence, and the filtered heat flux density data sequence are output.

[0007] S2. Perform spatiotemporal dual-parameter transient heat conduction model inversion processing on the filtered outer surface temperature data sequence and the filtered heat flux density data sequence output in step S1. Using the filtered outer surface temperature data sequence as time-varying boundary conditions, calculate the equivalent thermal diffusivity and equivalent heat storage coefficient of the material simultaneously from the data using a Bayesian joint inversion algorithm.

[0008] S3. Perform heat conduction forward problem solving and thermal resistance spatiotemporal cumulative effect modeling on the equivalent thermal diffusivity and equivalent heat storage coefficient calculated in step S2, as well as the filtered outer surface temperature data sequence and filtered heat flux density data sequence output in step S1, to reconstruct the spatiotemporal temperature field data inside the material and calculate the real-time inner surface thermal resistance data.

[0009] S4. Perform dynamic burn prediction processing with thermal resistance attenuation compensation on the filtered inner surface temperature data sequence output in step S1 and the real-time inner surface thermal resistance data obtained in step S3. By introducing a thermal resistance attenuation compensation function to correct and calculate the burn integral value, dynamic physiological safety time data is obtained in real time and dynamically updated as the heat exposure process progresses.

[0010] S5. Obtain thermal stability index data based on spectral entropy through the spectral entropy calculation method, and combine it with the dynamic physiological safety time data obtained in step S4 to form a multidimensional evaluation index dataset.

[0011] S6. Perform information entropy-driven adaptive weighted fusion processing on the multidimensional evaluation index dataset obtained in step S5. Dynamically allocate weights according to the information certainty of each index data in the current test, calculate the comprehensive thermal insulation performance index, and output the thermal insulation effect evaluation level.

[0012] Preferably, the transient change intensity index in S1 is defined as follows:

[0013] , ,

[0014] ,

[0015] , , ,

[0016] in, As an index of transient change intensity, For the cooperative modulation factor, This is the normalized component of the external surface temperature abrupt change. This represents the normalized component of the abrupt change in heat flux at the outer surface. This is the normalized component of the internal surface temperature abrupt change. For a width of Within the sliding window, the local Pearson correlation coefficient between the change in external surface temperature and the change in external surface heat flux. To strengthen the index, The width of the sliding window. In order to be in At time t, the first backward difference of the outer surface temperature, In order to be in Instantaneous temperature data collected at all times by sensors attached to the outer surface of the thermal insulation suit. In order to be in Instantaneous temperature data is constantly collected by sensors attached to the inner surface of the thermal insulation suit. In order to be in Instantaneous heat flux density data collected at all times by sensors attached to the outer surface of the thermal insulation suit. This represents the local average of the temperature change on the outer surface within the sliding window. In order to be in The first backward difference of the heat flux density on the outer surface at time t, This represents the local average value of the change in heat flux on the outer surface within the moving window. For a width of Within the sliding window, the local standard deviation of the outer surface temperature data, For a width of Within the sliding window, the local standard deviation of the inner surface temperature data, For a width of Within the sliding window, the local standard deviation of the outer surface heat flux density data, In order to be in The first-order backward difference of the outer surface temperature at time step In order to be in The first-order backward difference of the surface temperature within time t. In order to be in The first backward difference of the heat flux density on the outer surface at time t.

[0017] Preferably, the spatiotemporal two-parameter transient heat conduction model in S2 is:

[0018] ,

[0019] in, These are spatial coordinates, representing the vertical distance from a point inside the thermal insulation material to its inner surface. The time coordinate represents the real-time time calculated from the instant the heat exposure began. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, In depth ,time The transient temperature inside the material at that location, In depth Historical Moments Local instantaneous heat flux density at the location, The equivalent thermal diffusivity, The equivalent heat storage coefficient, The macroscopic density of the thermal insulation clothing material. The volumetric heat capacity of the material. The specific heat capacity of the thermal insulation material is given by the formulas for calculating the equivalent thermal diffusivity and equivalent heat storage coefficient of the material:

[0020] , ,

[0021] ,

[0022] , ,

[0023] ,

[0024] ,

[0025] in, Let be the temperature response characteristic function, representing the temperature response at time t. The temperature rise response caused by a unit heat flux density input to the outer surface. For the first-order heat flux accumulation moment, This is the second-order heat flux accumulation moment. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, To be at a historical moment The filtered outer surface heat flux density data at that time, The time coordinate represents the real-time time calculated from the instant the heat exposure began. The equivalent thermal diffusivity, The equivalent heat storage coefficient, For the first selected time point, The second selected time point, For at any time The calculated temperature response characteristic function value, For at any time The calculated temperature response characteristic function value, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated second-order heat flux cumulative moment, For at any time The calculated second-order heat flux cumulative moment.

[0026] Preferably, the formula for calculating the real-time inner surface thermal resistance data in S3 is as follows:

[0027] ,

[0028] ,

[0029] in, This is the internal surface thermal resistance data. For spatial discrete indexes, For time-discrete indexes, For the first The time from the start of thermal exposure at each sampling point For the first The depth distance from each discrete point to the inner surface. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, This is the temperature-dependent dynamic decay coefficient. For the interior of the material at depth ,time Local instantaneous heat flux density at the location, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, As the thermal recovery weighting factor, For local thermal recovery function, For the interior of the material at depth Historical Moments The transient temperature at that point, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, This is the nonlinear index of thermal stress.

[0030] Preferably, the formula for calculating the dynamic physiological safety time data in S4 is:

[0031] ,

[0032] , ,

[0033] in, This is the thermal resistance attenuation compensation function. This is the amplitude compensation coefficient. For rate compensation coefficient, This is the steady-state thermal resistance value. For real-time internal surface thermal resistance data, This represents the real-time thermal resistance deviation. For rate nonlinearity, This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, For the corrected burn integral value, To be at a historical moment At that time, the instantaneous temperature of the basal layer of the skin, To be at a historical moment The thermal resistance attenuation compensation function value, The cumulative damage feedback coefficient, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, This is dynamic physiological safety time data.

[0034] Preferably, the calculation formula for the thermal stability index data based on spectral entropy in S5 is as follows:

[0035] ,

[0036] ,

[0037] , ,

[0038] ,

[0039] in, For the first Spectral entropy of time fluctuations at scale. For the first The total number of layer detail coefficients, For the time index of detail coefficients, For the first The first layer of wavelet decomposition A detailed coefficient, For the first The first layer of wavelet decomposition A detailed coefficient, The entropy is the direction of the spatial thermal resistance gradient. For the thermal resistance spatial gradient, It is a set of positive gradients. This represents the total number of discrete points in space. As dummy variables in the summation process, It is a set of negative gradients. For the first Adaptive weights for each time-scale spectral entropy component For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. This represents the total number of layers in the time-scale analysis. For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. The adaptive weights for the entropy components of the spatial gradient direction. The sensitivity factor for the spatial spectral entropy component. This is thermal stability index data.

[0040] Preferably, the output formula for the thermal insulation effect evaluation level in S6 is:

[0041] ,

[0042] , ,

[0043] , ,

[0044] ,

[0045] in, To achieve a comprehensive thermal insulation performance index, For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping Assign an index number to the evaluation indicators. Corresponding steady-state thermal resistance value , Corresponding dynamic physiological safety time data , Corresponding thermal stability index data , This is the index number for the second evaluation indicator. For the first The and the first The interaction coupling coefficient between the indicators For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping For the first The mapping steepness factor of each indicator For the first The current value of each indicator, For the first The benchmark reference value for each indicator, For the first Asymmetric factors of each indicator For the first The normalized uncertainty of each indicator For the first Normalized uncertainty of each indicator Based on the coupling amplitude, For the first The and the first The Spearman rank correlation coefficient of each indicator in multiple historical trials. The coupling strength index is... The evaluation level is based on the insulation performance.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] 1. By establishing a spatiotemporal two-parameter transient heat conduction model, introducing a heat storage source term, and using the first-order and second-order heat flow accumulation moments to construct an analytical expression for the temperature response characteristic function, a simultaneous equation system is established and solved analytically at two different times in the quasi-steady-state stage. The equivalent thermal diffusivity and equivalent heat storage coefficient are calculated simultaneously from the data, overcoming the deficiency of traditional single-parameter models in being unable to characterize the heat storage history effect of materials.

[0048] 2. The thermal resistance sequence is separated into different time scales by wavelet decomposition. The temporal fluctuation spectral entropy and spatial thermal resistance gradient directional entropy of each scale are calculated. The thermal stability index data based on spectral entropy is obtained by weighted fusion using adaptive weights driven by the sensitivity factor.

[0049] 3. By introducing amplitude compensation coefficient, rate compensation coefficient and rate nonlinearity index to construct thermal resistance attenuation compensation function, and introducing cumulative damage feedback coefficient to correct the burn integral value, feedback correction of dynamic physiological safety time data on thermal insulation performance degradation is realized. Attached Figure Description

[0050] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a schematic diagram of a data processing-based method for evaluating the thermal insulation effect of thermal insulation clothing. Detailed Implementation

[0052] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described below in conjunction with the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0053] Numerous specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways than those described herein, and therefore the invention is not limited to the specific embodiments disclosed in the following specification.

[0054] Examples, such as Figure 1As shown, during the heat exposure process, temperature data and heat flux density data of the inner and outer surfaces of the thermal insulation suit are collected synchronously and continuously. An improved strong tracking adaptive filter is used to filter the collected temperature data and heat flux density data of the inner and outer surfaces of the thermal insulation suit. The filtering gain of the improved strong tracking adaptive filter is dynamically adjusted according to the transient change intensity index jointly defined by the temperature data and heat flux density data of the inner and outer surfaces of the thermal insulation suit. The filtered inner surface temperature data sequence, the filtered outer surface temperature data sequence, and the filtered heat flux density data sequence are output. In the thermal exposure test of the thermal insulation suit, data acquisition faces the contradiction of fast change tracking and strong noise suppression. The millisecond-level drastic jumps in temperature and heat flux data at the beginning of the heat exposure need to be completely preserved, while the sensor noise and environmental interference need to be strongly filtered out in the mid-to-late quasi-steady-state stage. Traditional fixed-gain filters cannot meet these two conflicting requirements at the same time, and amplitude detection alone cannot distinguish between real heat exposure events and random noise spikes. To address this, this invention defines a transient change intensity index calculated jointly from temperature and heat flow data, and introduces a local Pearson correlation coefficient to construct a collaborative modulation factor. This factor is used to collaboratively enhance the normalized components of temperature and heat flow changes, ensuring that the filter only enters fast tracking mode when both physical quantities undergo highly correlated changes simultaneously. During stable periods, it automatically switches to a strong denoising mode. This effectively suppresses noise in the stable phase while fully preserving the transient characteristics at the onset of thermal exposure. Furthermore, it provides a high-fidelity data foundation for the accurate identification of the equivalent thermal diffusivity and equivalent heat storage coefficients, reliable reconstruction of the corrected transient thermal resistance spectrum, and accurate prediction of dynamic physiological safety time. The transient change intensity index is defined as follows:

[0055] , ,

[0056] ,

[0057] , , ,

[0058] in, As an index of transient change intensity, For the cooperative modulation factor, This is the normalized component of the external surface temperature abrupt change. This represents the normalized component of the abrupt change in heat flux at the outer surface. This is the normalized component of the internal surface temperature abrupt change. For a width of Within the sliding window, the local Pearson correlation coefficient between the change in external surface temperature and the change in external surface heat flux. To strengthen the index, The width of the sliding window. In order to be in At time t, the first backward difference of the outer surface temperature, In order to be in At any given moment, instantaneous temperature data is collected by sensors attached to the outer surface of the thermal insulation suit. In order to be in At any given moment, instantaneous temperature data is collected by sensors attached to the inner surface of the thermal insulation suit. In order to be in At any given moment, instantaneous heat flux density data is collected by sensors attached to the outer surface of the thermal insulation suit. This is the local average value of the temperature change on the outer surface within the sliding window. In order to be in At time t, the first backward difference of the heat flux density on the outer surface, This represents the local average value of the change in heat flux on the outer surface within the moving window. For a width of Within the sliding window, the local standard deviation of the outer surface temperature data, For a width of Within the sliding window, the local standard deviation of the inner surface temperature data, For a width of Within the sliding window, the local standard deviation of the outer surface heat flux density data, In order to be in At time t, the first backward difference of the outer surface temperature, In order to be in At time t, the first backward difference of the inner surface temperature, In order to be in At any given moment, the first-order backward difference of the external surface heat flux density is used to extract the instantaneous change in the signal by calculating the first-order backward difference of the external surface temperature and external surface heat flux within adjacent sampling intervals. This transforms abrupt change detection from judging absolute amplitude to judging the rate of change. The local standard deviation of each signal within the sliding window is introduced to measure the normal fluctuation amplitude, i.e., the local noise level, in the current local time period. The temperature change and heat flux change are divided by their respective local standard deviations to obtain the normalized components of the external surface temperature change, internal surface temperature change, and external surface heat flux change. This maps physical quantities with different dimensions and different background noise levels to a unified local signal-to-noise ratio scale, making them comparable. Finally, a local Pearson correlation coefficient is introduced between the changes in internal and external surface temperature and the changes in external surface heat flux within the sliding window. This method measures the temporal synchronicity and directional consistency of two physical quantities. Real thermal exposure events physically inevitably lead to simultaneous, unidirectional jumps in temperature and heat flow, resulting in a correlation coefficient close to positive one. Random noise, lacking physical correlation, exhibits low or even negative correlation. By taking the non-negative value of the correlation coefficient and applying a strengthening exponent for nonlinear sharpening, a co-modulation factor is constructed. Low-correlation or negatively correlated noise events are strongly compressed to near zero, retaining only highly correlated real thermal exposure events. The co-modulation factor of the outer and inner surface schemes is multiplied by their corresponding normalized components, and the larger value is used as the transient change intensity index. This ensures sensitive detection regardless of whether thermal exposure first manifests on the outer surface or penetrates to the inner surface. Furthermore, the product form guarantees that only when the amplitude is sufficiently large and physically highly coordinated is it considered a strong change. This index directly drives the fading factor of the improved strong tracking adaptive filter, enabling the filter to automatically enter a fast tracking mode to fully preserve the information of the change frontier during real thermal exposure events, and automatically switch to a strong denoising mode during the stable phase, thus balancing the smoothness and response speed of the filtering process.

[0059] The filtered outer surface temperature data sequence and the filtered heat flux density data sequence are subjected to spatiotemporal two-parameter transient heat conduction model inversion processing. Using the filtered outer surface temperature data sequence as time-varying boundary conditions, the equivalent thermal diffusivity and equivalent heat storage coefficient of the material are simultaneously calculated from the data using a Bayesian joint inversion algorithm. In thermal exposure tests of thermal insulation clothing, traditional methods for identifying thermophysical parameters typically employ a Fourier heat conduction model containing only a single thermal diffusivity. This model simplifies the material as a pure diffusion medium, neglecting the heat storage history effect of the material absorbing heat through its own temperature rise during the unsteady-state phase. This results in an inability to describe the phenomenon that heat is first absorbed and stored by the material itself in the early stages of thermal exposure rather than immediately penetrating to the inner surface, leading to insufficient accuracy in predicting the temperature field and the loss of the heat storage coefficient, an important dimension reflecting the dynamic thermal performance of the material, thus limiting the richness of evaluation information. This invention establishes a spatiotemporal two-parameter transient heat conduction model including diffusion and heat storage term. The heat storage history effect of the material is explicitly incorporated into the governing equations through the cumulative heat flux change rate, enabling the model to simultaneously describe two independent physical mechanisms: heat diffusion and heat storage. Filtered external surface temperature data is used as time-varying boundary conditions, and filtered external surface heat flux density data is used to construct first-order and second-order heat flux accumulation moments. A system of simultaneous equations is established at two different moments in the quasi-steady-state stage using the temperature response characteristic function. The equivalent heat diffusion coefficient and equivalent heat storage coefficient are simultaneously calculated using a Bayesian joint inversion algorithm, and the posterior distribution and confidence intervals of both parameters are given. This overcomes the deficiency of traditional single-parameter models in failing to characterize the heat storage history effect of materials, improving the accuracy of temperature field reconstruction. It also extracts the heat storage coefficient as an independent dimension for evaluating material thermal performance from the data, enriching the information basis for subsequent multi-dimensional evaluation. Simultaneously, the Bayesian inversion provides a quantitative index for the uncertainty of parameter estimation, providing a basis for the confidence assessment of the final evaluation results. The spatiotemporal two-parameter transient heat conduction model is as follows:

[0060] ,

[0061] in, These are spatial coordinates, representing the vertical distance from a point inside the thermal insulation material to its inner surface. The time coordinate represents the real-time time calculated from the instant the heat exposure began. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, In depth ,time The transient temperature inside the material at that location, In depth Historical Moments Local instantaneous heat flux density at the location, The equivalent thermal diffusivity, The equivalent heat storage coefficient, The macroscopic density of the thermal insulation clothing material. The volumetric heat capacity of the material. The specific heat capacity of the thermal insulation material is given by the formulas for calculating the equivalent thermal diffusivity and equivalent heat storage coefficient of the material:

[0062] , ,

[0063] ,

[0064] , ,

[0065] ,

[0066] ,

[0067] in, Let be the temperature response characteristic function, representing the temperature response at time t. The temperature rise response caused by a unit heat flux density input to the outer surface. For the first-order heat flux accumulation moment, This is the second-order heat flux accumulation moment. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, To be at a historical moment The filtered outer surface heat flux density data at that time, The time coordinate represents the real-time time calculated from the instant the heat exposure began. The equivalent thermal diffusivity, The equivalent heat storage coefficient, For the first selected time point, The second selected time point, For at any time The calculated temperature response characteristic function value, For at any time The calculated temperature response characteristic function value, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated second-order heat flux cumulative moment, For at any time The calculated second-order heat flux cumulative moment. In the inversion processing of the spatiotemporal two-parameter transient heat conduction model, spatial coordinates, time coordinates, and integral dummy variables are selected as the basic spatiotemporal framework. The transient temperature inside the material is used as the core unknown quantity for model solution, and the local instantaneous heat flux density is used as the direct driving force for temperature change to fully describe the spatiotemporal evolution of heat transfer along the thickness direction within the material over time. The equivalent thermal diffusivity is selected as the first parameter to be identified, characterizing the rate of heat transfer through the diffusion mechanism and determining the speed at which heat waves propagate from the outer surface to the inner surface. The equivalent heat storage coefficient is selected as the second parameter to be identified. Traditional single-parameter models simplify the material as a pure diffusion medium, ignoring the heat storage history effect of the material absorbing heat through its own temperature rise in the unsteady state stage. Introducing this parameter can independently separate the heat storage capacity from the diffusion rate, allowing the model to simultaneously describe the two independent physical dimensions of heat transfer rate and heat storage capacity. The macroscopic density and specific heat capacity of the material are selected. Their product constitutes the volumetric heat capacity, which is a necessary physical property constant in the heat storage source term that connects the heat storage coefficient with temperature change. In terms of the computational system for parameter identification, a temperature response characteristic function is selected and defined as the ratio of the external surface temperature rise to the external surface heat flux density. This characteristic function transforms the complex inverse problem of spatiotemporal partial differential equations into algebraic equations, significantly reducing computational complexity. Simultaneously, its physical meaning is clear, measuring the material's temperature sensitivity to external thermal exposure. A first-order heat flux cumulative moment is selected and defined as the cumulative integral of the external surface heat flux density from the start of thermal exposure to the current moment. This represents the total heat transferred to the material per unit area, reflecting the total cumulative heat load borne by the material, and constitutes a pure diffusion contribution term in the relational expression. A second-order heat flux cumulative moment is selected and defined as the cumulative integral of the external surface heat flux density weighted by time lag. The first-order moment can only describe the total cumulative heat load and cannot distinguish whether heat is concentrated or uniformly transferred in the early stages. The material's heat storage effect is highly sensitive to the temporal distribution of heat flux; concentrated early transfer will trigger a stronger heat storage response. This temporal information must be captured by the second-order moment, constituting a heat storage contribution term in the relational expression. By establishing a relationship between the temperature response characteristic function and the first and second-order cumulative heat flux moments, the external surface temperature rise response is decomposed into the sum of pure diffusion and heat storage contributions. This decomposition achieves algebraic separation of the two physical mechanisms, allowing the two parameters to be identified to be analytically solved through a system of simultaneous equations at different times, rather than relying on iterative optimization algorithms which may lead to local optima or convergence instability. Two different times in the quasi-steady-state phase are chosen to establish the simultaneous equations because a single time point can only provide one equation, failing to simultaneously determine two independent parameters. However, the first and second-order cumulative heat flux moments differ at the two different times, making the two equations independent and uniquely determining the equivalent thermal diffusivity and equivalent heat storage coefficient. Repeated calculations at multiple times in the quasi-steady-state phase, with the median as the final result, are performed to suppress the bias that may be influenced by random noise at a single time point, improving the robustness of parameter estimation.

[0068] The calculated equivalent thermal diffusivity and equivalent heat storage coefficient, as well as the output filtered outer surface temperature data sequence and filtered heat flux density data sequence, are processed by solving the forward problem of heat conduction and modeling the spatiotemporal cumulative effect of thermal resistance. The spatiotemporal temperature field data inside the material is reconstructed, and the real-time inner surface thermal resistance data is calculated. In the evaluation of the thermal insulation effect of thermal insulation clothing, if the evaluation is carried out directly after obtaining the equivalent thermal diffusivity and equivalent heat storage coefficient through step S2, there are defects. These two parameters are macroscopic equivalent physical properties of the material, which only describe the overall ability of heat transfer and storage, and cannot reveal the spatial distribution difference of thermal resistance along the material thickness direction and its dynamic evolution process over time. The evaluation information remains at the lumped level. Traditional thermal resistance calculation is based only on the simple ratio of the temperature difference between the inner and outer surfaces to the heat flux density, assuming that the material is a homogeneous medium and the thermal resistance is constant. It completely ignores the spatiotemporal cumulative decay effect of thermal resistance caused by the gradual change of the material's microstructure under continuous high temperature, and cannot reflect the gradual degradation of performance. This invention first utilizes the calculated equivalent thermal diffusivity and equivalent heat storage coefficient, using the filtered output outer surface temperature data sequence as time-varying boundary conditions, and the filtered heat flux density data sequence as the surface thermal input to solve the spatiotemporal two-parameter transient heat conduction forward problem, reconstructing the spatiotemporal temperature field data inside the material. Actual measurements can only obtain data from two locations: the inner and outer surfaces. The temperature and heat flux distributions at various depths inside the material cannot be directly obtained through sensors; therefore, the entire internal field information must be inferred from the boundary data through forward problem solving, providing a complete spatiotemporal temperature distribution basis for subsequent layer-by-layer calculations of local thermal resistance. Based on this, the reconstructed spatiotemporal temperature field data inside the material is modeled for the spatiotemporal cumulative effect of thermal resistance. This modeling process introduces a temperature-dependent dynamic attenuation coefficient, which automatically accelerates the attenuation rate in high-temperature regions, reflecting the accelerated degradation effect of high temperatures. It also introduces a thermal stress nonlinear exponent to power-modulate the absolute value of the temperature gradient, nonlinearly amplifying the attenuation contribution in high thermal stress regions and characterizing the material's nonlinear sensitivity to extreme thermal stress. Furthermore, it introduces a local thermal recovery function and a thermal recovery weighting factor to construct a degradation saturation suppression mechanism in the denominator, automatically slowing down the subsequent attenuation rate in regions that have already undergone significant degradation, thus avoiding over-prediction of thermal resistance attenuation under long-term thermal exposure. By coupling these three mechanisms, an exponential cumulative attenuation correction is applied to the classical instantaneous local thermal resistance, resulting in real-time internal surface thermal resistance data. On the one hand, by reconstructing the spatiotemporal temperature field inside the material, the limitation of only being able to obtain surface data is broken through to the acquisition of full-field information, providing data support for the layer-by-layer refined analysis of thermal resistance. On the other hand, by modeling the spatiotemporal cumulative effect of thermal resistance, the coupling effect of three mechanisms—temperature acceleration, thermal stress nonlinearity, and degradation saturation—is explicitly embedded into the thermal resistance calculation, so that real-time internal surface thermal resistance data can truly reflect the gradual performance degradation process of the material under continuous thermal exposure, rather than assuming that the thermal resistance is constant as in traditional methods.This real-time internal surface thermal resistance data provides direct input for the subsequent construction of the thermal resistance attenuation compensation function and the calculation of the dynamic physiological safety time. The calculation formula for the real-time internal surface thermal resistance data is:

[0069] ,

[0070] ,

[0071] in, This is the internal surface thermal resistance data. For spatial discrete indexes, For time-discrete indexes, For the first The time from the start of thermal exposure at each sampling point For the first The depth distance from each discrete point to the inner surface. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, This is the temperature-dependent dynamic decay coefficient. For the interior of the material at depth ,time Local instantaneous heat flux density at the location, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, As the thermal recovery weighting factor, For local thermal recovery function, For the interior of the material at depth Historical Moments The transient temperature at that point, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, The thermal stress nonlinearity index is used. In the real-time calculation of the inner surface thermal resistance data, the material micro-layer closest to the human skin is located using the spatial discrete index and the depth distance at the inner surface, and different time nodes in the thermal exposure process are marked using the temporal discrete index and the sampling time. The difference between the transient internal temperature of the material obtained by reconstruction at adjacent depth layers and the inner surface is used as the numerator, which characterizes the instantaneous temperature potential driving the heat flow of the micro-layer segment; the local instantaneous heat flux density calculated by combining the reconstructed temperature field with Fourier's law and calibrated by filtering the outer surface heat flux density data at the same location is used as the denominator, which represents the instantaneous heat flow rate through the micro-layer segment. The ratio of the two constitutes the classical local instantaneous thermal resistance, reflecting the instantaneous ability of the segment to prevent heat from passing through at the current moment. Building upon this foundation, an exponential cumulative degradation correction factor is introduced, employing a temperature-dependent dynamic degradation coefficient. This coefficient is correlated exponentially with the baseline degradation coefficient, material thermal degradation activation energy, molar gas constant, and local transient temperature. The material's thermal degradation rate exhibits an exponential positive correlation with temperature; higher temperatures result in faster degradation. Fixed constants cannot describe the accelerated degradation phenomenon in high-temperature regions. A thermal stress nonlinear exponent is used to power-modulate the absolute value of the local temperature gradient. Material microstructural damage often accelerates sharply after thermal stress exceeds a certain threshold. The nonlinear exponent can characterize this threshold effect, while linear modulation cannot reflect the accelerated characteristics of high thermal stress regions. A local thermal recovery function is selected and defined as the cumulative integral of the degradation driving force from the start of thermal exposure to a historical moment. This function, along with the thermal recovery weighting factor, is placed in the denominator of the degradation correction factor to form a degradation saturation suppression term. Material performance degradation is not infinitely accelerated; once the surface layer is completely carbonized, further thermal stress no longer changes its thermal resistance at the same rate. Without this mechanism, the thermal resistance degradation under long-term thermal exposure would be over-predicted, losing physical accuracy. By coupling the three mechanisms of temperature acceleration effect, thermal stress nonlinear effect and degradation saturation effect, the calculated real-time internal surface thermal resistance data can truly reflect the complete evolution process of the material under continuous thermal exposure from initial integrity to gradual degradation and then to saturation, providing reliable input for the subsequent construction of thermal resistance attenuation compensation function and accurate calculation of dynamic physiological safety time.

[0072] The filtered inner surface temperature data sequence and the obtained real-time inner surface thermal resistance data are subjected to dynamic burn prediction processing with thermal resistance attenuation compensation. By introducing a thermal resistance attenuation compensation function to correct the burn integral value, dynamic physiological safety time data is obtained in real time and updated dynamically with the heat exposure process. In traditional thermal insulation clothing performance evaluation, skin burn prediction usually adopts an open-loop calculation method, that is, directly inputting the inner surface temperature data into a fixed burn integral model to obtain a static safety time. This method ignores the performance degradation of the thermal insulation clothing itself under continuous high temperature. When the thermal resistance of the material decreases due to fiber shrinkage, carbonization, etc., the same inner surface temperature means that the actual heat penetrating the clothing is increasing rapidly, but traditional open-loop prediction does not react to this, leading to a systematic overestimation of the safety time. This overestimation may cause workers to exceed the safe exposure limit and suffer irreversible thermal damage in actual rescue scenarios. This invention introduces dynamic burn prediction processing with thermal resistance attenuation compensation. The obtained real-time inner surface thermal resistance data is fed back into the burn integral calculation to construct a closed-loop correction loop. The thermal resistance attenuation compensation function is driven by an amplitude compensation coefficient, a rate compensation coefficient, and a rate nonlinear exponent. This allows the function to respond not only to the amplitude of thermal resistance attenuation but also to the instantaneous rate at which thermal resistance is decreasing. The rate response is sharpened by the nonlinear exponent, rapidly amplifying the compensation function when thermal resistance deteriorates sharply. Simultaneously, a cumulative damage feedback coefficient is introduced into the burn integral value calculation, feeding back the accumulated burn damage to the subsequent burn rate calculation, reflecting the physiological characteristic that skin tissue becomes more susceptible to further damage. The corrected burn integral value accumulates rapidly with the real-time attenuation of thermal resistance and the dynamic increase of cumulative damage. The dynamic physiological safety time is the moment when this integral value first reaches the second-degree burn threshold, and it is updated in real-time with the continuous change of thermal resistance during the heat exposure process. A direct closed-loop feedback relationship is established between the performance degradation of thermal insulation clothing and human physiological safety, overcoming the shortcomings of traditional open-loop predictions that overestimate safety time due to neglecting material degradation. The dual-driven compensation of amplitude and rate ensures that the safety time responds reasonably to both slow degradation and rapid deterioration scenarios. The cumulative damage memory effect makes the calculation of the safety time closer to the actual physiological process. The final output dynamic physiological safety time is a conservative estimate that is continuously revised as heat exposure progresses, providing a more reliable early warning basis for on-site safety decisions. The calculation formula for the dynamic physiological safety time data is as follows:

[0073] ,

[0074] , ,

[0075] in, This is the thermal resistance attenuation compensation function. This is the amplitude compensation coefficient. For rate compensation coefficient, This is the steady-state thermal resistance value. Real-time internal surface thermal resistance data, This represents the real-time thermal resistance deviation. For rate nonlinearity, This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, For the corrected burn integral value, To be at a historical moment At that time, the instantaneous temperature of the basal layer of the skin, To be at a historical moment The thermal resistance attenuation compensation function value, The cumulative damage feedback coefficient, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, This study focuses on dynamic physiological safety time data. A thermal resistance attenuation compensation function is selected as the core tool for correcting burn integral values, establishing a direct quantitative bridge between the performance degradation of thermal insulation clothing and human physiological safety, allowing the safety time to be dynamically adjusted according to the actual state of the clothing. In this compensation function, the first compensation term is the product of the amplitude compensation coefficient and the ratio of the real-time thermal resistance deviation to the steady-state thermal resistance value. The real-time thermal resistance deviation quantifies the absolute decrease in thermal resistance. Dividing this by the steady-state thermal resistance value and normalizing it yields the relative attenuation rate, which is then multiplied by the amplitude compensation coefficient. This directly transforms the degree of performance degradation into an amplification factor for the burn integral accumulation rate; the more severe the degradation, the faster the acceleration. The second compensation term is composed of the rate compensation coefficient, the rate nonlinearity exponent, and the effective attenuation rate. Amplitude compensation alone cannot distinguish between slow degradation and rapid deterioration. The effective attenuation rate is the truncation of the negative value of the time derivative of the real-time inner surface thermal resistance data under non-negative conditions. A positive value is taken when the thermal resistance is decreasing to activate rate compensation, and zero is taken when the thermal resistance is stable or rising to avoid misjudging small increases in thermal resistance due to measurement noise as effective attenuation. The rate nonlinearity exponent modulates the effective decay rate exponentially. When thermal resistance decreases slowly, the compensation term is small; however, if a rapid deterioration occurs, the compensation term amplifies dramatically in an exponential manner, significantly shortening the safety time and reflecting the warning logic that accelerated deterioration is more dangerous. Linear responses cannot achieve this sensitive capture of sudden, rapid decay. The steady-state thermal resistance value is used as the normalized denominator in both the first and second compensation terms, providing a unified reference baseline for the amplitude and rate of thermal resistance decay, making compensation calculations comparable for different materials and test conditions. A modified burn integral value is used instead of the traditional fixed burn integral value, which relies solely on skin temperature history and ignores the feedback effect of clothing performance degradation on the heat transfer process. In the calculation of the modified burn integral value, the reciprocal of the characteristic time constant corresponding to the instantaneous temperature of the skin basal layer is used as the basic instantaneous burn accumulation rate. This characteristic time constant represents the theoretical value of the time required for the skin to reach a second-degree burn under constant temperature. Its reciprocal is the core of the classic Henriques model and an indispensable physiological basis for any burn prediction. The cumulative damage feedback coefficient is multiplied by the integral of the accumulated burn score over time and then incorporated into the burn rate. Under continuous heat exposure, the accumulated thermal damage to skin tissue in the early stages leads to a decrease in tissue defense capabilities and the failure of repair mechanisms, resulting in faster damage accumulation under subsequent exposure to the same temperature. Without a feedback mechanism, this physiological characteristic of becoming more susceptible to injury after initial exposure would lead to burn predictions that assume constant skin tolerance, deviating from the actual physiological process. Integral dummy variables and inner-layer integral dummy variables are used to traverse the time history from the start of heat exposure to the current moment and from the start of heat exposure to a historical moment, respectively. This two-layer integral structure ensures that the cumulative damage memory effect can completely trace the burn accumulation process throughout the entire history of heat exposure.The dynamic physiological safety time is defined as the moment when the corrected burn score first reaches the second-degree burn threshold and is updated in real time with the change of the internal surface thermal resistance data. The safety time should not be a one-time static calculation result at the moment of heat exposure, but should be continuously shortened as the performance of clothing deteriorates, so as to provide a more conservative and reliable real-time early warning basis for on-site safety decisions.

[0076] The thermal stability index data based on spectral entropy is obtained through the spectral entropy calculation method, and together with the obtained dynamic physiological safety time data, constitutes a multidimensional evaluation index dataset; the calculation formula for the thermal stability index data based on spectral entropy is as follows:

[0077] ,

[0078] ,

[0079] , ,

[0080] ,

[0081] in, For the first Spectral entropy of time fluctuations at scale. For the first The total number of layer detail coefficients, For the time index of detail coefficients, For the first The first layer of wavelet decomposition A detailed coefficient, For the first The first layer of wavelet decomposition A detailed coefficient, The entropy is the direction of the spatial thermal resistance gradient. For the thermal resistance spatial gradient, It is a set of positive gradients. This represents the total number of discrete points in space. As dummy variables in the summation process, It is a set of negative gradients. For the first Adaptive weights for each time-scale spectral entropy component For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. This represents the total number of layers in the time-scale analysis. For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. The adaptive weights for the entropy components of the spatial gradient direction. The sensitivity factor for the spatial spectral entropy component. This refers to thermal stability index data. In evaluating the thermal stability of thermally insulated clothing, traditional methods typically only calculate the simple coefficient of variation or range of the internal surface thermal resistance time series. This has limitations; a single statistic mixes the fluctuation characteristics of different time scales, failing to distinguish between high-frequency instantaneous disturbances and low-frequency trend-based decay—two distinctly different fluctuation modes based on physical mechanisms. Furthermore, focusing only on the temporal fluctuations at a single location on the internal surface completely ignores the spatial structural differences in thermal resistance between different layers along the material thickness direction, failing to detect localized weak points within the material. To address these issues, this invention employs a multi-scale spatiotemporal-frequency joint spectral entropy calculation method. The internal surface thermal resistance sequence is separated into multiple time scales through wavelet decomposition, and the temporal fluctuation spectral entropy is calculated for the detail coefficients at each scale. The detail coefficients represent the fluctuation components of the thermal resistance at the corresponding time point at that scale; smaller scales correspond to high-frequency rapid fluctuations, while larger scales correspond to low-frequency slow fluctuations. After normalizing the absolute values ​​of the detail coefficients at each scale to a probability distribution, the spectral entropy is calculated. Spectral entropy measures the uniformity of the distribution of wave energy along the time axis. A smaller value indicates that the wave energy is concentrated in a few moments, i.e., sporadic disturbances; a larger value indicates that the wave energy is evenly distributed, i.e., persistent irregular fluctuations. This information-theoretic-based metric can keenly capture non-periodic, non-stationary transient disturbances. Traditional coefficients of variation only reflect the overall dispersion and cannot achieve this precise quantification of the distribution structure. High spectral entropy specifically captures instantaneous disturbances, while low spectral entropy specifically captures trend decay. If multi-scale decomposition is not performed and only a single overall spectral entropy is calculated, the characteristics of both types of fluctuations will be averaged and lose their diagnostic value. Spatial thermal resistance gradient directional entropy is introduced. The thermal resistance spatial gradient between adjacent layers is calculated along the thickness direction using the time mean of thermal resistance in the quasi-steady-state phase of each depth layer, and then divided into positive gradient sets and negative gradient sets according to the sign of the gradient. Positive gradients represent the normal state where thermal resistance increases with depth, i.e., the closer to the heat source, the higher the thermal resistance. Negative gradients represent the abnormal state where thermal resistance decreases with depth, i.e., the surface layer may have degraded. The normalized probability distribution of gradient magnitudes and the corresponding directional entropy within each set are calculated separately. A weighted sum is then performed using the proportion of elements in each set to the total gradients to obtain the spatial thermal resistance gradient directional entropy. Positive entropy measures the uniformity of the increase in normal layer magnitudes, while negative entropy measures the uniformity of the decrease in abnormal layer magnitudes. Their physical meanings are completely different; mixing them would lose the ability to specifically identify degraded layers. The design using the proportion of elements in each set as weights ensures that when there are many normal layers, positive entropy dominates the spatial entropy, and when there are many abnormal layers, negative entropy is activated and amplified, achieving an automatic balance quantification of spatial structural order and local degradation. For the fusion of the temporal and spatial dimensions, an adaptive weight driven by a sensitivity factor is introduced. The ratio of the standard deviation to the mean of the spectral entropy components at each time scale and the spatial spectral entropy components in multiple historical trials is defined as their respective sensitivity factors, and then the sensitivity factors are normalized into adaptive weights.The sensitivity factor reflects the component's ability to distinguish between different materials or experimental conditions. A larger coefficient of variation indicates a higher degree of information discrimination in historical data. The weights are entirely determined by the data's inherent discrimination power; components with strong discrimination power automatically receive higher weights, and those with weaker discrimination power automatically receive lower weights. There is no need to manually set fixed weight values. Furthermore, the weights automatically adapt and adjust when changes in experimental conditions or material systems alter the discrimination power of each component—a feature that traditional fixed-weight methods cannot achieve. Finally, all time-scale spectral entropy components and spatial spectral entropy components are weighted and summed according to their respective adaptive weights to obtain the thermal stability index data based on spectral entropy. A smaller value indicates that the material exhibits less fluctuation across multiple time scales and has an ordered spatial structure, resulting in better overall thermal stability.

[0082] The obtained multidimensional evaluation index dataset is subjected to information entropy-driven adaptive weighted fusion processing. Weights are dynamically assigned based on the information certainty of each index data in the current test, and the comprehensive thermal insulation performance index is calculated to output the thermal insulation effect evaluation level. The output formula for the thermal insulation effect evaluation level is as follows:

[0083] ,

[0084] , ,

[0085] , ,

[0086] ,

[0087] in, To achieve a comprehensive thermal insulation performance index, For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping Assign an index number to the evaluation indicators. Corresponding steady-state thermal resistance value , Corresponding dynamic physiological safety time data , Corresponding thermal stability index data , This is the index number for the second evaluation indicator. For the first The and the first The interaction coupling coefficient between the indicators For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping For the first The mapping steepness factor of each indicator For the first The current value of each indicator, For the first The benchmark reference value for each indicator, For the first Asymmetric factors of each indicator For the first Normalized uncertainty of each indicator For the first Normalized uncertainty of each indicator Based on the coupling amplitude, For the first The and the first The Spearman rank correlation coefficient of each indicator in multiple historical trials. The coupling strength index is... This invention evaluates the thermal insulation effect by assigning a grade. Traditional methods for comprehensive evaluation of thermal insulation effects typically use fixed weights or subjective weighting to linearly weight each indicator. This approach has drawbacks: the measurement reliability of each indicator varies, but fixed weights cannot reflect differences in indicator confidence levels caused by measurement noise or model errors. Low-quality indicators are treated equally with high-quality indicators. Linear mapping assumes that the contribution of each indicator to the baseline is equal, failing to reflect the asymmetric requirement in safety evaluation that performance degradation should be punished more severely than excellent performance. Furthermore, it assumes that each indicator is independent, ignoring potential physical correlations between indicators and failing to utilize synergistic or antagonistic information to correct the evaluation results. This invention introduces a multi-element collaborative adaptive weighted fusion process. Dynamic confidence weights are selected and defined as the ratio of the exponential functions of the normalized uncertainties of each indicator, rather than directly using information entropy or subjective weights. The normalized uncertainty of each indicator originates from the measurement or model uncertainty of its preceding calculation steps. The uncertainty of the steady-state thermal resistance value originates from the fitting residual of the corrected transient thermal resistance spectrum reconstruction. The uncertainty of the dynamic physiological safety time originates from the parameter uncertainty of the burn prediction model. The uncertainty of the thermal stability index originates from the statistical uncertainty of the multi-scale spectral entropy calculation. The weights directly reflect the data quality of each indicator in the current single test, rather than relying on the statistical distribution of historical data or expert experience. Indicators with lower uncertainty automatically receive higher weights, allowing reliable data to play a dominant role in the evaluation, while the contribution of unreliable data is automatically suppressed, fundamentally solving the defect that traditional fixed weights cannot distinguish data quality. An exponential function is used to convert uncertainty into a confidence factor, causing the weight of high-uncertainty indicators to be non-linearly compressed and non-linearly decreased, further strengthening the penalty for low-quality data. An asymmetric logistic mapping function is used to perform non-linear mapping on each indicator, introducing an asymmetric factor. When the indicator value is better than or equal to the benchmark value, the asymmetric factor is one; when the indicator value is worse than the benchmark value, the asymmetric factor is a deterioration-side acceleration coefficient greater than one. Simultaneously, a mapping steepness factor is introduced to control the slope of the curve near the benchmark point, and a scaling factor is introduced to normalize deviations of different dimensions. In the field of security protection, the tolerance for performance deficiencies is far lower than the requirement for performance redundancy. For the same deviation, the penalty for performance degradation should be greater than the bonus for performance improvement, reflecting the principle of conservative security. Traditional symmetric or linear mappings cannot achieve this asymmetric penalty. The larger the acceleration coefficient on the degradation side, the more severe the penalty for performance degradation, making the evaluation results more biased towards the conservative security side. An interaction coupling coefficient is selected and defined as the Spearman rank correlation coefficient based on pairwise indicators from multiple historical experiments, multiplied by the basic coupling amplitude, and nonlinearly modulated with a coupling strength exponent. The basic coupling amplitude controls the maximum influence of the interaction term on the total score, while the coupling strength exponent concentrates the interaction effect on strongly correlated indicators, while weakly correlated indicators compress the interaction term.If two indicators have a stable positive physical correlation, the interaction term amplifies their joint contribution when both are better than the benchmark, and amplifies the penalty when both are worse than the benchmark. If a negative correlation exists, it indicates a potential physical trade-off between the two indicators, and the interaction term suppresses unidirectional contributions to prevent over-evaluation. This overcomes the limitation of traditional linear weighting's assumption of independent indicators, utilizing the physical correlation information between indicators to correct evaluation results. The comprehensive thermal insulation performance index is obtained by multiplying the sum of the independent contributions of each indicator and the interactive coupling contributions of all pairwise combinations by a percentage coefficient. This ensures that the final score reflects both the independent performance of each indicator and includes corrective information on synergistic or antagonistic interactions between indicators. The thermal insulation effect evaluation level discretizes the continuous comprehensive thermal insulation performance index into four levels: excellent, good, average, and poor, facilitating rapid judgment and comparison in engineering applications.

[0088] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments for application in other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing, characterized in that, Includes the following steps: S1. During the heat exposure process, temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit are collected synchronously and continuously. An improved strong tracking adaptive filter is used to filter the collected temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit. The filtering gain of the improved strong tracking adaptive filter is dynamically adjusted according to the transient change intensity index jointly defined by the temperature data and heat flux density data of the inner and outer surfaces of the heat insulation suit. The filtered inner surface temperature data sequence, the filtered outer surface temperature data sequence, and the filtered heat flux density data sequence are output. S2. Perform spatiotemporal dual-parameter transient heat conduction model inversion processing on the filtered outer surface temperature data sequence and the filtered heat flux density data sequence output in step S1. Using the filtered outer surface temperature data sequence as time-varying boundary conditions, calculate the equivalent thermal diffusivity and equivalent heat storage coefficient of the material simultaneously from the data using a Bayesian joint inversion algorithm. S3. Perform heat conduction forward problem solving and thermal resistance spatiotemporal cumulative effect modeling on the equivalent thermal diffusivity and equivalent heat storage coefficient calculated in step S2, as well as the filtered outer surface temperature data sequence and filtered heat flux density data sequence output in step S1, to reconstruct the spatiotemporal temperature field data inside the material and calculate the real-time inner surface thermal resistance data. S4. Perform dynamic burn prediction processing with thermal resistance attenuation compensation on the filtered inner surface temperature data sequence output in step S1 and the real-time inner surface thermal resistance data obtained in step S3. By introducing a thermal resistance attenuation compensation function to correct and calculate the burn integral value, dynamic physiological safety time data is obtained in real time and dynamically updated as the heat exposure process progresses. S5. Obtain thermal stability index data based on spectral entropy through the spectral entropy calculation method, and combine it with the dynamic physiological safety time data obtained in step S4 to form a multidimensional evaluation index dataset. S6. Perform information entropy-driven adaptive weighted fusion processing on the multidimensional evaluation index dataset obtained in step S5. Dynamically allocate weights according to the information certainty of each index data in the current test, calculate the comprehensive thermal insulation performance index, and output the thermal insulation effect evaluation level.

2. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The transient change intensity index in S1 is defined as follows: ， , , ， ， , in, As an index of transient change intensity, For the cooperative modulation factor, This is the normalized component of the external surface temperature abrupt change. This represents the normalized component of the abrupt change in heat flux at the outer surface. This is the normalized component of the internal surface temperature abrupt change. For a width of Within the sliding window, the local Pearson correlation coefficient between the change in external surface temperature and the change in external surface heat flux. To strengthen the index, The width of the sliding window. In order to be in At time t, the first backward difference of the outer surface temperature, In order to be in At any given moment, instantaneous temperature data is collected by sensors attached to the outer surface of the thermal insulation suit. In order to be in At any given moment, instantaneous temperature data is collected by sensors attached to the inner surface of the thermal insulation suit. In order to be in At any given moment, instantaneous heat flux density data is collected by sensors attached to the outer surface of the thermal insulation suit. This represents the local average of the temperature change on the outer surface within the sliding window. In order to be in At time t, the first backward difference of the heat flux density on the outer surface, This represents the local average value of the change in heat flux on the outer surface within the moving window. For a width of Within the sliding window, the local standard deviation of the outer surface temperature data, For a width of Within the sliding window, the local standard deviation of the inner surface temperature data, For a width of Within the sliding window, the local standard deviation of the outer surface heat flux density data, In order to be in At time t, the first backward difference of the outer surface temperature, In order to be in At time t, the first backward difference of the inner surface temperature, In order to be in At time t, the first backward difference of the heat flux density on the outer surface.

3. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The spatiotemporal two-parameter transient heat conduction model in S2 is as follows: , in, These are spatial coordinates, representing the vertical distance from a point inside the thermal insulation material to its inner surface. The time coordinate represents the real-time time calculated from the instant the heat exposure began. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, In depth ,time The transient temperature inside the material at that location, In depth Historical Moments Local instantaneous heat flux density at the location, The equivalent thermal diffusivity, The equivalent heat storage coefficient, The macroscopic density of the thermal insulation clothing material. The volumetric heat capacity of the material. The specific heat capacity of the thermal insulation material is given by the formulas for calculating the equivalent thermal diffusivity and equivalent heat storage coefficient of the material: ， ， ， ， ， ， ， in, Let be the temperature response characteristic function, representing the temperature response at time t. Temperature rise response per unit heat flux input to the outer surface For the first-order heat flux accumulation moment, This is the second-order heat flux accumulation moment. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, To be at a historical moment The filtered outer surface heat flux density data at that time, The time coordinate represents the real-time time calculated from the instant the heat exposure began. The equivalent thermal diffusivity, The equivalent heat storage coefficient, For the first selected time point, The second selected time point, For at any time The calculated temperature response characteristic function value, For at any time The calculated temperature response characteristic function value, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated first-order heat flux accumulation moment, For at any time The calculated second-order heat flux cumulative moment, For at any time The calculated second-order heat flux cumulative moment.

4. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The formula for calculating the real-time inner surface thermal resistance data in S3 is as follows: ， ， in, This is the internal surface thermal resistance data. For spatial discrete indexes, For time-discrete indexes, For the first The time from the start of thermal exposure at each sampling point For the first The depth distance from each discrete point to the inner surface. This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, This is the temperature-dependent dynamic decay coefficient. For the interior of the material at depth ,time Local instantaneous heat flux density at the location, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, For the interior of the material at depth ,time The transient temperature at that point, As the thermal recovery weighting factor, For local thermal recovery function, For the interior of the material at depth Historical Moments The transient temperature at that point, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, This is the nonlinear index of thermal stress.

5. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The formula for calculating the dynamic physiological safety time data in S4 is as follows: ， ， ， in, This is the thermal resistance attenuation compensation function. This is the amplitude compensation coefficient. For rate compensation coefficient, This is the steady-state thermal resistance value. Real-time internal surface thermal resistance data, This represents the real-time thermal resistance deviation. For rate nonlinearity, This is an integral dummy variable, representing the time from the start of heat exposure to the current moment. At any point in history between, For the corrected burn integral value, To be at a historical moment At that time, the instantaneous temperature of the basal layer of the skin, To be at a historical moment The thermal resistance attenuation compensation function value, The cumulative damage feedback coefficient, The inner integral dummy variable represents the time from the start of thermal exposure to a historical moment. At any time between, This is dynamic physiological safety time data.

6. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The calculation formula for the thermal stability index data based on spectral entropy in S5 is as follows: ， ， ， ， ， in, For the first Spectral entropy of time fluctuations at scale. For the first The total number of layer detail coefficients, For the time index of detail coefficients, For the first The first layer of wavelet decomposition A detailed coefficient, For the first The first layer of wavelet decomposition A detailed coefficient, The entropy is the direction of the spatial thermal resistance gradient. For the thermal resistance spatial gradient, It is a set of positive gradients. This represents the total number of discrete points in space. As dummy variables in the summation process, It is a set of negative gradients. For the first Adaptive weights for each time-scale spectral entropy component For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. This represents the total number of layers in the time-scale analysis. For the first Sensitivity factor for spectral entropy components at each time scale. For time-scale indexing. The adaptive weights for the entropy components of the spatial gradient direction. The sensitivity factor for the spatial spectral entropy component. This is thermal stability index data.

7. The method for evaluating the thermal insulation effect of thermal insulation clothing based on data processing according to claim 1, characterized in that, The output formula for the thermal insulation effect evaluation level in S6 is: ， ， ， ， ， ， in, To achieve a comprehensive thermal insulation performance index, For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping Assign an index number to the evaluation indicators. Corresponding steady-state thermal resistance value , Corresponding dynamic physiological safety time data , Corresponding thermal stability index data , This is the index number for the second evaluation indicator. For the first The and the first The interaction coupling coefficient between the indicators For the first The dynamic confidence weight of each indicator For the first The contribution values ​​of each indicator after asymmetric logistic mapping For the first The mapping steepness factor of each indicator For the first The current value of each indicator, For the first The benchmark reference value for each indicator, For the first Asymmetric factors of each indicator For the first Normalized uncertainty of each indicator For the first Normalized uncertainty of each indicator Based on the coupling amplitude, For the first The and the first The Spearman rank correlation coefficient of each indicator in multiple historical trials. The coupling strength index is... The evaluation level is based on the insulation performance.

Images