A method for predicting the lifespan of polyurethane thermal insulation materials

Abstract

This invention provides a method for predicting the lifespan of polyurethane thermal insulation materials, belonging to the field of intelligent prediction technology. It obtains multi-layer temperature field data by radially slicing the infrared temperature data of the insulation layer and spatially matching it with the thickness data through a grid. This multi-layer temperature field data is then spatially aligned with environmental temperature and humidity, and pipeline vibration acceleration data. An aging acceleration model is used to convert humidity penetration depth and vibration stress amplitude into equivalent temperature values, constructing a spatial coupling feature set. The temperature values, humidity equivalent temperature values, and vibration equivalent temperature values ​​at each grid location are weighted and summed to obtain the multi-field coupled equivalent total temperature. Based on this, local aging acceleration zones are sorted and selected. The lifespan decay rate of these local aging acceleration zones is calculated, and failure risk is prioritized to generate partitioned lifespan prediction results for the insulation layer. This invention accurately locates local aging weak points in a multi-field coupled environment, improving the accuracy of lifespan prediction and the applicability of risk assessment.

Description

Technical Field

[0001] This invention relates to the field of intelligent prediction technology, and more specifically, to a method for predicting the lifespan of polyurethane thermal insulation materials. Background Technology

[0002] With the rapid development of technologies such as artificial intelligence, machine learning (ML), and data-driven analytics, these advanced technologies are playing an increasingly significant role in empowering traditional industrial sectors. In polymer material applications, computational technology is widely used in scenarios such as performance prediction, aging assessment, and lifespan management. Polyurethane thermal insulation materials, due to their excellent thermal insulation performance and low thermal conductivity, are widely used in buildings, industrial pipelines, and cold chain transportation. Although these materials exhibit excellent thermal performance, they inevitably undergo aging and degradation under prolonged exposure to complex environments (such as temperature gradients, humidity erosion, and frequent mechanical vibrations), leading to localized failure or functional decline. Traditional material lifespan assessment methods mainly rely on accelerated aging experiments and macroscopic characteristic models, which result in poor adaptability to actual complex environments and an inability to reflect detailed changes under the influence of localized environments.

[0003] To address the shortcomings of existing life prediction technologies for polyurethane thermal insulation materials, the following key issues can be identified: First, current technologies primarily focus on life analysis under single-factor conditions, failing to adequately consider the synergistic effects of multiple factors such as temperature, humidity, and vibration, thus making it difficult to reflect the coupled aging behavior of materials in real-world service environments. Second, existing models have limited spatial resolution, typically estimating life only at an overall level, unable to identify localized aging regions and their corresponding failure risks. Furthermore, the dynamic characteristics of time series data and the response mechanisms of material properties to environmental changes are neglected during spatial data processing and model construction, resulting in a lack of accuracy in describing the material aging process. Therefore, these shortcomings significantly restrict the widespread application of polyurethane materials and the reliability and scientific validity of actual life prediction results. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method for predicting the lifespan of polyurethane thermal insulation materials. This method can, to some extent, solve the problems of large errors in the lifespan prediction results and difficulty in assessing the risk of local failures, caused by the inability to accurately obtain the dynamic characteristics of the temperature field in each region inside the polyurethane insulation layer, the difficulty in identifying local accelerated aging zones due to the coupling effect of multiple physical fields, and the lack of effective support for dynamic calculation of multi-field coupling characteristics in existing lifespan prediction methods.

[0005] According to one aspect of the present invention, a method for predicting the lifespan of polyurethane thermal insulation materials is provided, comprising: Infrared temperature distribution data of the insulation layer is acquired and sliced ​​according to radial depth. Spatial grid matching is performed on the temperature value and thickness measurement data of each depth layer to obtain multi-layer temperature field data. The multi-layer temperature field data is spatially aligned with the environmental temperature and humidity time series data and the pipeline vibration acceleration data. The humidity penetration depth and vibration stress amplitude at each grid location are converted into equivalent temperature values ​​through the aging acceleration model to construct a spatially coupled feature set. The temperature, humidity equivalent temperature, and vibration equivalent temperature values ​​at each grid location within the spatial coupling feature set are weighted and summed to obtain the multi-field coupling equivalent total temperature. Local aging acceleration zones are then sorted and selected based on the multi-field coupling equivalent total temperature. The characteristic data of the local accelerated aging zone are used to calculate the partitioned life decay rate. Based on the decay rate, the failure risk priority of each local area is sorted to generate the partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

[0006] Furthermore, the aging acceleration model includes a first aging acceleration model and a second aging acceleration model; The first aging acceleration model is based on the accelerating effect of humidity on the aging rate of thermal insulation materials, and the formula is expressed as:

[0007] in, This is the humidity-equivalent temperature value; As the reference temperature, The dimensionless penetration depth factor. The dimensionless moisture content factor The dimensionless diffusion time factor Here is the denominator regularization parameter. It is a dimensionless activation energy scaling factor. The intensity attenuation sensitivity coefficient, It is a dimensionless intensity attenuation factor. It is a dimensionless permeability ratio factor. For auxiliary functions, For ambient relative humidity, The ambient temperature.

[0008] Furthermore, the formula for calculating the penetration depth factor is as follows:

[0009] in, To measure the actual humidity penetration depth, This represents the critical penetration depth of the material. The effective penetration depth exceeds the critical depth; The formula for the intensity attenuation factor is expressed as follows:

[0010] in, The current tensile strength of the material. The initial tensile strength of the material. This is the protection threshold for the residual strength of the material.

[0011] Furthermore, the second aging acceleration model is established based on the accelerating effect of vibration fatigue on the aging rate of insulation materials, and the formula is expressed as:

[0012] in, This is the vibration equivalent temperature value; The vibration stress amplitude calculated for the current grid position; As the reference stress amplitude threshold, For reference stress value, As the vibration acceleration factor, This is the cumulative acceleration factor for the cycle. To accumulate the number of vibration cycles, The reference number of cycles is used to make the cumulative effect of cycles dimensionless; This represents the standard deviation of stress.

[0013] Furthermore, the calculation of the equivalent total temperature of the multi-field coupling includes: A normalized weighting coefficient is assigned to each depth layer, with the weighting coefficient of the inner surface depth layer being greater than that of the outer surface depth layer. The depth-weighted average temperature value is obtained by multiplying the temperature values ​​of each depth layer by the corresponding normalized weighting coefficient and then summing them. The percentage of depths affected by humidity is obtained by dividing the number of depth layers marked as humidity-affected layers by the total number of depth layers. Multiplying the humidity equivalent temperature value by the humidity influence depth ratio yields the humidity-weighted equivalent temperature. The multi-field coupling equivalent total temperature is obtained by summing the depth-weighted average temperature, the humidity-weighted equivalent temperature, and the vibration equivalent temperature.

[0014] Furthermore, the screening of the locally accelerated aging region includes: The coupling feature groups are arranged in descending order of the equivalent total temperature of the multi-field coupling. Starting from the second pair in the sorted list, calculate the equivalent total temperature difference between each pair and the previous pair in the multi-field coupling, and record the position number of the pair with the largest temperature difference in the list as the natural break point number. The preliminary screening ratio is obtained by dividing the natural breakpoint number by the total number of coupling feature groups. If the preliminary screening ratio is greater than 0.3, the screening ratio parameter is 0.3; if it is less than 0.1, the screening ratio parameter is 0.1; if it is between 0.1 and 0.3, the preliminary screening ratio is directly used. Multiply the total number of the coupled feature groups by the screening ratio parameter and round up to obtain the screening quantity threshold. Mark the top-ranked coupling feature groups with the highest screening quantity threshold as local aging acceleration zones.

[0015] Furthermore, the formula for calculating the lifetime decay rate is as follows:

[0016] in, This represents the lifetime decay rate at the currently traversed spatial grid location. This refers to the design service life of the insulation material under standard environmental conditions. It is a dimensionless temperature sensitivity coefficient. The increase in temperature For reference temperature, The temperature and humidity coupling enhancement coefficient is... The humidity level at the currently traversed spatial grid location affects the depth ratio. The humidity acceleration factor, The vibration acceleration coefficient, This represents the vibration equivalent temperature value at the currently traversed spatial grid location. These are vibration saturation characteristic parameters.

[0017] According to another aspect of the present invention, a life prediction system for polyurethane thermal insulation materials is provided, comprising: The temperature reconstruction module is used to slice the acquired infrared temperature distribution data of the insulation layer according to radial depth, and perform spatial grid matching on the temperature value of each depth layer and the thickness measurement data to obtain multi-layer temperature field data. The coupling construction module is used to spatially align the multi-layer temperature field data with the environmental temperature and humidity time series data and the pipeline vibration acceleration data, and to convert the humidity penetration depth and vibration stress amplitude at each grid location into equivalent temperature values ​​through the aging acceleration model, thereby constructing a spatial coupling feature set. The aging positioning module is used to perform weighted summation of the temperature value, humidity equivalent temperature value and vibration equivalent temperature value at each grid position in the spatial coupling feature set to obtain the multi-field coupling equivalent total temperature, and sort and filter out the local aging acceleration zone according to the multi-field coupling equivalent total temperature. The risk ranking module is used to calculate the partitioned life decay rate of the feature data of the local aging acceleration zone, and to rank the failure risk priority of each local area according to the decay rate, generating a partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

[0018] Compared with existing technologies, this invention acquires and processes infrared temperature distribution data of polyurethane insulation layers, combining it with environmental temperature and humidity data and pipeline vibration acceleration data to achieve spatial alignment and fusion of multi-physics data, significantly improving the ability to refine the modeling of material aging behavior. Based on an accelerated aging model, it converts humidity penetration depth and vibration stress amplitude into equivalent temperature values, and accurately calculates the equivalent total temperature of multi-field coupling through the construction and weighted summation of multi-field coupling feature sets, thereby effectively identifying local accelerated aging regions. Simultaneously, by calculating the partitioned lifetime decay rate of characteristic data in accelerated aging regions, it prioritizes local failure risks, providing efficient and accurate technical support for insulation layer lifetime prediction and maintenance strategies. It demonstrates significant advantages in addressing localized aging problems in complex service environments, improving lifetime prediction accuracy, and enhancing the applicability of risk assessment. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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. In the drawings: Figure 1 A flowchart of a method for predicting the lifespan of polyurethane thermal insulation materials according to an embodiment of the present invention; Figure 2 This is a radar chart comparing the detection rate of local accelerated zones under different aging scenarios according to embodiments of the present invention. Detailed Implementation

[0020] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0021] As mentioned in the background section above, the existing technology has two main problems: First, because it is impossible to fully obtain multi-field synergistic data of polyurethane thermal insulation materials under complex environmental conditions, the existing life prediction methods are based on modeling only on a single physical field or macroscopic experimental data, which makes it difficult to truly reflect the aging behavior of materials in actual service environments; Second, the existing technology lacks the ability to accurately identify local accelerated aging areas and fails to effectively integrate multi-physical field data such as temperature, humidity and vibration, resulting in insufficient spatial resolution of life assessment results and difficulty in quantifying the risk of local failure in advance.

[0022] Figure 1 This is a system block diagram of a method for predicting the lifespan of polyurethane thermal insulation materials according to an embodiment of the present invention. Figure 1 and Figure 2 As shown, the method for predicting the lifespan of polyurethane thermal insulation materials includes: S1: Obtain infrared temperature distribution data of the insulation layer and perform layered slicing according to radial depth. Spatial grid matching is performed on the temperature value and thickness measurement data of each depth layer to obtain multi-layer temperature field data.

[0023] Based on the collected infrared temperature distribution data of the insulation layer surface, multiple slices are set along the radial depth direction. The specific operation is as follows: Read the insulation layer thickness measurement data, iterate through the thickness values ​​of all measurement points, find the maximum thickness value by comparing the values, and record the maximum thickness value as the layer reference thickness.

[0024] Determine the thickness range of the layer reference thickness: If the layer reference thickness is between 50 mm and 100 mm, divide the layer reference thickness into ten depth layers; if the layer reference thickness is between 100 mm and 150 mm, divide the layer reference thickness into twenty depth layers; if the layer reference thickness exceeds 150 mm, divide the layer reference thickness into thirty depth layers.

[0025] The specific execution method of the equal division operation is as follows: Divide the layer reference thickness value by the number of layers to obtain the interval distance between each depth layer. Starting from the outer surface of the insulation layer, the depth position of the first layer is zero, the depth position of the second layer is one interval distance, the depth position of the third layer is twice the interval distance, and so on, until the depth position of the last layer is equal to the layer reference thickness value. This forms a series of discrete depth levels from the outer surface to the inner surface, with each depth layer identified by its depth value from the outer surface.

[0026] For each depth layer, the temperature value corresponding to the depth layer is calculated. Since the infrared thermal imager can only collect the temperature distribution on the outer surface of the insulation layer, the temperature of each depth layer inside needs to be estimated through the thermal conduction relationship.

[0027] Under steady-state operation of the pipeline, the calculation logic is based on the physical law that the temperature inside the insulation layer exhibits a linear gradient distribution from the outer surface to the inner surface. The specific calculation method is as follows: For a certain location on the insulation layer surface, the outer surface temperature value is read, and simultaneously, the corresponding pipe inner wall temperature value is read. The pipe inner wall temperature is taken as an approximation of the pipe medium temperature, or calculated inversely from the pipe outer wall temperature through heat conduction. The outer surface temperature value is subtracted from the pipe inner wall temperature value to obtain the total temperature difference at the surface location. For a certain depth layer below the surface location, the depth value of the depth layer from the outer surface is read. The depth value is divided by the total insulation layer thickness at the surface location to obtain the depth ratio. The total temperature difference is multiplied by the depth ratio to obtain the temperature drop of the depth layer relative to the outer surface. The temperature drop value is subtracted from the outer surface temperature value to obtain the temperature value of the depth layer. For each surface location, the above calculations are performed sequentially from the first layer to the last layer to obtain the temperature value sequence of the surface location across all depth layers.

[0028] Spatial gridding matching is performed between the temperature values ​​of each depth layer and the insulation layer thickness measurement data. Specifically, the insulation layer surface is divided into uniformly distributed square grids, with a side length of 20 mm. Based on the actual dimensions of the insulation layer surface, the number of grids is calculated in both the horizontal and vertical directions. The horizontal grid number equals the total horizontal length of the insulation layer divided by 20 mm and rounded up; the vertical grid number equals the total vertical length of the insulation layer divided by 20 mm and rounded up. For each grid cell, the spatial coordinates of the grid cell's center point are calculated. The horizontal center point coordinates equal to the grid cell's horizontal index number multiplied by 20 mm plus 10 mm; the vertical center point coordinates equal to the grid cell's vertical index number multiplied by 20 mm plus 10 mm. The pixel closest to the center point coordinates is found in the infrared temperature distribution data, and its temperature value is used as the surface temperature value of the grid cell. The measurement point closest to the center point coordinates is found in the thickness measurement data, and its thickness value is used as the insulation layer thickness value of the grid cell. The method for finding the closest point is as follows: traverse all data points, calculate the Euclidean distance between each data point and the center point of the grid cell, and select the data point with the smallest distance value as the matching point.

[0029] Furthermore, due to the differences in insulation layer thickness at different spatial locations, the depth layer at each grid location needs to be corrected. The correction method is as follows: For a given grid location, read the actual thickness value and divide it by the layer reference thickness value to obtain the thickness scaling ratio. For each depth layer at the grid location, read the depth position value of the depth layer under the layer reference and multiply it by the thickness scaling ratio to obtain the corrected depth position value of the depth layer at the grid location. After correction, the temperature values ​​at each corrected depth layer at the grid location are recalculated. The calculation method is as follows: Read the outer surface temperature value and the pipe inner wall temperature value at the grid location, and subtract the pipe inner wall temperature value from the outer surface temperature value to obtain the total temperature difference. Divide the corrected depth position value by the actual thickness value at the grid location to obtain the corrected depth ratio. Multiply the total temperature difference by the corrected depth ratio to obtain the temperature drop value of the corrected depth layer relative to the outer surface. Subtract the temperature drop value from the outer surface temperature value to obtain the temperature value of the corrected depth layer at the grid location.

[0030] Temperature values ​​from all grid cells across all depth layers are integrated to obtain multi-layer temperature field data. This multi-layer temperature field data is stored as a three-dimensional array. The first dimension of the array corresponds to the horizontal position number of the grid cells, ranging from one to the total number of horizontal grid cells. The second dimension corresponds to the vertical position number of the grid cells, ranging from one to the total number of vertical grid cells. The third dimension corresponds to the depth layer number, ranging from one to the total number of depth layers.

[0031] Accessing an element in the 3D array requires specifying three index values. The element stores the temperature values ​​for the corresponding horizontal position, vertical position, and depth layer. Simultaneously, a 2D array is created to store the actual insulation layer thickness and outer surface temperature for each grid location. The first dimension index of the 2D array corresponds to the grid's horizontal position number, and the second dimension index corresponds to the grid's vertical position number. Each array element contains two values: the first is the actual insulation layer thickness at the grid location, and the second is the outer surface temperature at that location.

[0032] S2: Spatially align the multi-layer temperature field data with the environmental temperature and humidity time series data and the pipeline vibration acceleration data, and convert the humidity penetration depth and vibration stress amplitude at each grid location into equivalent temperature values ​​through the aging acceleration model to construct a spatially coupled feature set.

[0033] Spatial alignment is performed on the multi-layer temperature field data, ambient temperature and humidity time-series data, and pipeline vibration acceleration data. The specific operations of spatial alignment include: Read the grid division information in the multi-layer temperature field data, obtain the horizontal position number, vertical position number and corresponding spatial coordinate value of each grid cell, and use the two-dimensional coordinate system established with a fixed point on the surface of the insulation layer as the origin as the reference. Read the humidity sensor placement information from the ambient temperature and humidity time series data. Each humidity sensor has its installation coordinates on the pipe surface, which are also labeled based on the same two-dimensional coordinate system. Read the accelerometer placement information from the pipeline vibration acceleration data; each accelerometer also has its installation coordinates on the pipeline surface. For each grid cell in the multi-layer temperature field data, the following alignment operation is performed: For each grid cell, the installation coordinates of all humidity sensors are traversed, the Euclidean distance between the sensor coordinates and the grid center point coordinates is calculated, and the nearest sensor is selected, with its collected humidity data sequence used as the environmental humidity data for that grid location. Similarly, all accelerometers are traversed, and the nearest sensor is selected, with its collected vibration acceleration data sequence used as the vibration data for that grid location. Through these operations, each grid location is assigned a set of temperature data, a set of humidity data, and a set of vibration data, achieving spatial alignment and matching of multi-source data.

[0034] Furthermore, the humidity penetration depth is calculated from the environmental humidity data. This calculation is based on the hygroscopic properties of the insulation material and the humidity diffusion pattern. Specifically: for a given grid location, the corresponding environmental humidity data sequence is read, and the cumulative duration (in hours) of relative humidity exceeding 80% within the past month is extracted. The humidity diffusion coefficient of the insulation material is then read. Based on the type of insulation material at that grid location, the humidity diffusion coefficient value corresponding to that material type is retrieved from a preset material property parameter database. The humidity diffusion coefficient is multiplied by the cumulative duration, and the square root is taken to obtain a preliminary value for the humidity penetration depth. If the preliminary value exceeds the actual thickness of the insulation layer at that grid location, the humidity penetration depth is set to the actual thickness value; otherwise, the preliminary value is used directly.

[0035] Furthermore, the vibration stress amplitude is calculated from the vibration acceleration data. For a specific grid location, the corresponding vibration acceleration data sequence is read, and the acceleration values ​​in the three orthogonal directions over the past week are extracted. For each time point, the square root of the sum of the squares of the accelerations in the three directions is calculated to obtain the composite acceleration amplitude, and the maximum value of the composite acceleration amplitude within that week is identified as the peak acceleration. The density value of the insulation material at that grid location is read. If a field-sampled and measured density value exists at that grid location, it is directly read; otherwise, the nominal density value is retrieved from a preset material property parameter database based on the type of insulation material at that grid location. The actual thickness of the insulation layer at that grid location is read. The peak acceleration is multiplied by the density value, and then by the thickness value to obtain the inertial force per unit area. The inertial force per unit area is used as the quasi-static stress amplitude characterizing the level of periodic vibration load, and is denoted as the vibration stress amplitude. The vibration stress amplitude is compared with the material's elastic modulus: if the vibration stress amplitude exceeds the material's elastic range, the upper limit of the elastic range is taken as the final vibration stress amplitude; if the vibration stress amplitude does not exceed the material's elastic range, the inertial force value is directly used as the final vibration stress amplitude.

[0036] Furthermore, the temperature values, humidity penetration depth, and vibration stress amplitude corresponding to the same coordinates are constructed into a coupling feature group. For each grid location, the temperature value sequence at each depth layer is read, along with the humidity penetration depth and vibration stress amplitude values ​​calculated in the previous steps. The grid's horizontal position number, vertical position number, temperature value sequence at each depth layer, humidity penetration depth value, and vibration stress amplitude value are combined into a coupling feature group. For each depth layer, it is determined whether its depth position is less than or equal to the humidity penetration depth. If so, it is marked as a humidity-affected layer; otherwise, it is marked as a non-humidity-affected layer. This marking information is also stored in the coupling feature group.

[0037] The humidity-equivalent temperature value is calculated based on the humidity penetration depth in a preset first aging acceleration model. The first aging acceleration model is established based on the accelerating effect of humidity on the aging rate of insulation materials, and the formula is expressed as:

[0038] in, This is the humidity-equivalent temperature value; As the reference temperature, The dimensionless penetration depth factor. The dimensionless moisture content factor The dimensionless diffusion time factor Here is the denominator regularization parameter. It is a dimensionless activation energy scaling factor. The intensity attenuation sensitivity coefficient, It is a dimensionless intensity attenuation factor. It is a dimensionless permeability ratio factor. For auxiliary functions, For ambient relative humidity, The ambient temperature.

[0039] Furthermore, the formula for the penetration depth factor is expressed as:

[0040] in, To measure the actual humidity penetration depth, This represents the critical penetration depth of the material. The effective penetration depth is beyond the critical depth.

[0041] The formula for the moisture content factor is expressed as:

[0042] in, This refers to the moisture content of the material when it reaches equilibrium under the current humidity conditions. The saturated water content of the material. The baseline moisture content is used.

[0043] The formula for the diffusion time factor is expressed as:

[0044] in, To measure the effective humidity diffusion coefficient, The baseline humidity diffusivity, This represents the cumulative duration of time the material has been exposed to a humid environment. The baseline exposure duration is used.

[0045] The formula for the intensity attenuation factor is expressed as:

[0046] in, The current tensile strength of the material. The initial tensile strength of the material. This is the protection threshold for the residual strength of the material.

[0047] The formula for the permeability ratio factor is expressed as:

[0048] in, To measure the actual humidity penetration depth, This represents the total thickness of the insulation layer.

[0049] The formula for the activation energy proportionality factor is expressed as:

[0050] in, The apparent activation energy of humidity-induced aging represents the energy barrier that the aging reaction of a material must overcome under humid conditions. The apparent activation energy for pure thermal aging represents the energy barrier that must be overcome for the aging reaction of a material under pure temperature conditions.

[0051] The formula for the auxiliary function is expressed as:

[0052] in, For ambient relative humidity, For ambient temperature, This is the humidity response sensitivity coefficient. Critical relative humidity for a material represents the humidity threshold at which its hygroscopic behavior changes significantly. saturated relative humidity, For humidity protection parameters, The reference temperature is the temperature at which the material does not undergo significant aging, while the lower limit of the temperature at which the aging rate of the material is negligible is indicated. This is a temperature protection parameter, in Kelvin. This represents the temperature coupling index.

[0053] It should be noted that, The moisture absorption equilibrium curve of the material is determined based on the current relative humidity of the environment. The moisture absorption equilibrium curve is pre-determined through a constant temperature and humidity chamber test. The adsorption was determined by step adsorption tests of the material under different humidity conditions; Tensile tests were conducted using on-site samples. Determined through accelerated aging tests under different humidity conditions; The results were determined by thermogravimetric analysis under different temperature conditions.

[0054] Furthermore, the boundary protection conditions are as follows:

[0055] in, The effective penetration depth exceeds the critical depth. To measure the actual humidity penetration depth, This is the critical penetration depth at which the material begins to absorb moisture significantly.

[0056] Boundary protection conditions ensure that only when the moisture penetration depth exceeds the critical depth at which the material begins to absorb moisture significantly is the excess portion included in the effective penetration depth for subsequent aging acceleration calculations. When the penetration depth does not exceed the critical depth, the effective penetration depth is set to zero, indicating that the moisture has not yet produced an identifiable accelerating effect on the material's aging, thereby avoiding misjudgments in aging assessments caused by minor fluctuations in penetration depth.

[0057] Based on the calculated humidity-equivalent temperature value, an upper bound constraint is further applied to constrain it, as expressed by the formula:

[0058] in, This represents the final humidity-equivalent temperature value after applying the upper bound constraint. This represents the unconstrained humidity-equivalent temperature value. This is the upper limit of the equivalent temperature limit, representing the maximum multiple of the humidity-accelerated effect relative to the reference temperature. This is the reference temperature.

[0059] The vibration equivalent temperature value is calculated based on the vibration stress amplitude in a preset second aging acceleration model. The second aging acceleration model is established based on the accelerating effect of vibration fatigue on the aging rate of the insulation material, and the formula is expressed as:

[0060] in, This is the vibration equivalent temperature value; The vibration stress amplitude calculated for the current grid position; As the reference stress amplitude threshold, For reference stress value, As the vibration acceleration factor, This is the cumulative acceleration factor for the cycle. To accumulate the number of vibration cycles, The reference number of cycles is used to make the cumulative effect of cycles dimensionless; This represents the standard deviation of stress.

[0061] It should be noted that, The calculation method is as follows: The resultant acceleration amplitude at each time point within a week is converted into a stress value sequence using the same conversion relationship as the vibration stress amplitude. Specifically, the resultant acceleration amplitude at each time point is multiplied by the material density and then by the insulation layer thickness to obtain the quasi-static stress value at each time point. The standard deviation of this stress value sequence is then calculated. .

[0062] The coupling feature sets for all grid locations are aggregated to obtain the spatial coupling feature set. The feature set is stored in list form, where each element corresponds to a coupling feature set for a grid location.

[0063] Each coupling feature group contains the following complete information: grid horizontal position number, grid vertical position number, original temperature value sequence of each depth layer at that position, humidity penetration depth value at that position, vibration stress amplitude value at that position, humidity equivalent temperature value at that position, vibration equivalent temperature value at that position, and humidity influence marker sequence of each depth layer at that position.

[0064] The spatial coupling feature set fully records the temperature field information of the insulation layer in spatial distribution, as well as the influence of two non-temperature factors, humidity and vibration, providing a unified data foundation for subsequent comprehensive evaluation.

[0065] S3: The temperature value, humidity equivalent temperature value and vibration equivalent temperature value of each grid position in the spatial coupling feature set are weighted and summed to obtain the multi-field coupling equivalent total temperature. The local aging acceleration zone is sorted and selected according to the multi-field coupling equivalent total temperature.

[0066] For each spatial location within the spatial coupling feature set, calculate the equivalent total temperature of multi-field coupling. Traverse each coupling feature group in the spatial coupling feature set, read the horizontal and vertical position numbers of the mesh stored in the feature group, and determine the current processing position. Read the original temperature value sequence of each depth layer at the current position. The depth layers are arranged sequentially from the outer surface to the inner surface, with the first number corresponding to the outermost layer and the maximum number corresponding to the innermost layer.

[0067] Because of the temperature gradient within the insulation layer, the inner region near the high-temperature pipe experiences higher temperatures and a faster aging rate. Therefore, a greater weight needs to be assigned to the inner region when calculating the representative temperature. The weighting rule is as follows: Count the total number of depth layers divided into which the insulation layer at the current location is divided, denoted as N. Assign a sequence number to each depth layer, starting from 1 on the inner surface and increasing sequentially towards the outer surface, with the outermost layer's sequence number equal to N. Calculate the initial weight value for each depth layer: the initial weight for the inner surface depth layer is N, the next innermost layer's initial weight is N minus one, and so on, with the outermost layer's initial weight being 1. Sum the initial weight values ​​of all depth layers to obtain the initial weight sum. Divide the initial weight value of each depth layer by the initial weight sum to obtain the corresponding normalized weight coefficient. Multiply the temperature value of each depth layer by the corresponding normalized weight coefficient, sum the weighted temperature values ​​of all depth layers, and obtain the depth-weighted average temperature value at the current location as the original representative temperature value.

[0068] Read the humidity-affected layer marker sequence stored in the feature group, and count the number of depth layers marked as Boolean truth values, denoted as the humidity-affected layer count M. Read the humidity equivalent temperature value stored in the feature group. When M is greater than zero, divide M by the total number of depth layers N to obtain the humidity-affected depth ratio, and then multiply the humidity equivalent temperature value by the humidity-affected depth ratio to obtain the humidity-weighted equivalent temperature. When M equals zero, the humidity-weighted equivalent temperature is zero. Read the vibration equivalent temperature value stored in the feature group. Calculate the arithmetic sum of the depth-weighted average temperature value, the humidity-weighted equivalent temperature, and the vibration equivalent temperature value to obtain the multi-field coupling equivalent total temperature at the current location.

[0069] The above calculation operation is performed on all coupling feature groups in the spatial coupling feature set one by one. The corresponding multi-field coupling equivalent total temperature value is calculated for each spatial location, and the multi-field coupling equivalent total temperature value and the humidity influence depth ratio value are stored back into the coupling feature group at the corresponding location.

[0070] Furthermore, a temporary storage list is created. Each coupling feature group in the spatial coupling feature set is traversed, and the equivalent total temperature value of multi-field coupling is read. This equivalent total temperature value and the complete coupling feature group information are added as a pair to the temporary storage list. A quicksort is performed on the temporary storage list, with the following comparison rules: The equivalent total temperature value of the multi-field coupling of two pairs is compared, and the pair with the larger value is ranked first; if the temperature values ​​of two pairs are equal, the ratio of their stored humidity influence depth is compared, and the pair with the larger ratio is ranked first. After sorting, the pairs in the list are arranged in descending order of temperature value.

[0071] The total number of coupling feature groups in the spatial coupling feature set is denoted as S; a preset screening ratio parameter is read. The screening ratio parameter is adaptively determined as follows: starting from the second pair in the sorted list, the equivalent total temperature difference between each pair and the previous pair is calculated sequentially. The position number of the pair with the largest temperature difference in the list is recorded as the natural break point number. The natural break point number is divided by S to obtain the preliminary screening ratio. If the preliminary screening ratio is greater than 0.3, the screening ratio parameter is set to 0.3; if the preliminary screening ratio is less than 0.1, the screening ratio parameter is set to 0.1; if the preliminary screening ratio is between 0.1 and 0.3, the preliminary screening ratio is directly used as the screening ratio parameter.

[0072] Multiply S by the filtering ratio parameter to obtain a floating-point result of the filtering quantity. Round the floating-point result up to obtain the filtering quantity threshold K. Create an empty local aging acceleration zone feature data set. Starting from the beginning of the sorted list, read the paired items in the list sequentially, extract the coupling feature group information from the paired items, add an aging acceleration zone identifier field to the coupling feature group and set it to a Boolean true value, and then add the marked coupling feature group to the local aging acceleration zone feature data set. After adding each coupling feature group, count the number of coupling feature groups already contained in the current set, and determine whether the threshold K has been reached. If not, continue to read the next paired item from the list and perform the marking and addition operation; if the threshold K has been reached, stop adding.

[0073] For all remaining pairs in the sorted list from position K plus one to the end of the list, extract the coupling feature group information, add an aging acceleration zone identifier field to the feature group and set it to a boolean false value.

[0074] Create a two-dimensional data table with the number of rows equal to the maximum value of the longitudinal position numbers on the insulation layer surface, and the number of columns equal to the maximum value of the transverse position numbers. Iterate through each coupling feature group in the spatial coupling feature set, reading the transverse and longitudinal position numbers. Locate the cell at the intersection of the corresponding row and column in the table, and fill in the multi-field coupling equivalent total temperature value, the humidity influence depth ratio value, and the aging acceleration zone identifier status. When the aging acceleration zone identifier is true, the cell displays the equivalent total temperature value with a red background and an asterisk after the value, along with the humidity influence depth ratio in percentage format. When the aging acceleration zone identifier is false, the cell displays the equivalent total temperature value with a white background, without an asterisk, and the humidity influence depth ratio is displayed in small gray text.

[0075] In one embodiment, a comparative test was conducted on the detection of local accelerated aging zones using both the existing empirical threshold partitioning method and the method of the present invention. The test results are as follows: Figure 2 As shown.

[0076] Figure 2 The radar chart compares the detection rates of local accelerated zones under different aging scenarios. The horizontal axis includes six typical scenarios: high temperature zone, high humidity zone, high vibration zone, temperature-humidity coupling zone, temperature-vibration coupling zone, and three-field coupling zone. The vertical axis represents the detection rate percentage. The red curve in the figure represents the detection rate of the empirical threshold partitioning method in the prior art, and the green curve represents the detection rate of the method of the present invention.

[0077] from Figure 2As can be seen, in high-temperature, high-humidity, and high-vibration scenarios dominated by a single factor, the detection rates of the method of this invention are 96%, 94%, and 92%, respectively, which are 8, 49, and 40 percentage points higher than the empirical threshold method, respectively. In multi-factor coupled scenarios such as temperature-humidity coupled zone, temperature-vibration coupled zone, and three-field coupled zone, the detection rates of the method of this invention are 95%, 97%, and 93%, respectively, which are 57, 55, and 65 percentage points higher than the empirical threshold method, respectively.

[0078] The improved detection rate stems from step S2, where humidity penetration depth and vibration stress amplitude are converted into humidity equivalent temperature values ​​and vibration equivalent temperature values ​​using humidity aging acceleration model and vibration aging acceleration model, respectively. In step S3, these values ​​are arithmetically summed with depth-weighted average temperature values ​​to obtain the multi-field coupling equivalent total temperature. Then, an adaptive screening mechanism driven by natural breakpoint number is used to determine the screening ratio parameter and screening quantity threshold K, avoiding the systematic omissions in multi-field coupling scenarios when using fixed temperature or humidity thresholds in existing technologies.

[0079] S4: Calculate the partitioned life decay rate of the characteristic data of the local aging accelerated zone, sort the failure risk priority of each local area according to the decay rate, and generate the partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

[0080] Obtain the design service life of the insulation material under standard environmental conditions from the materials database. Failure temperature Dimensionless temperature sensitivity coefficient Temperature and humidity coupling enhancement coefficient Humidity acceleration coefficient Vibration acceleration coefficient and vibration saturation characteristic parameters .

[0081] Among them, failure temperature The temperature value at which the material mass loss rate reaches 5%, and the temperature sensitivity coefficient. The parameters are calculated by dividing the material's thermal aging activation energy by the gas constant and then by the reference temperature. All parameters are pre-determined through accelerated aging tests or vibration fatigue tests and stored in the material database. A standard ambient temperature is set. The reference temperature is 20 degrees Celsius. .

[0082] Iterate through each coupled feature group in the feature data set of the local aging acceleration zone to obtain the equivalent total temperature of multi-field coupling. Humidity affects depth ratio and vibration equivalent temperature value Calculate the temperature rise. .when When the value is greater than 0, the lifetime decay rate of the currently traversed spatial grid position under the current multi-field coupling condition is calculated according to the following formula:

[0083] when When the temperature acceleration factor is less than or equal to 0, it is set to one, and the lifetime decay rate is calculated using the following simplified formula:

[0084] in, This represents the lifetime decay rate at the currently traversed spatial grid location. This refers to the design service life of the insulation material under standard environmental conditions. It is a dimensionless temperature sensitivity coefficient. The increase in temperature For reference temperature, The temperature and humidity coupling enhancement coefficient is... The humidity level at the currently traversed spatial grid location affects the depth ratio. The humidity acceleration factor, The vibration acceleration coefficient, This represents the vibration equivalent temperature value at the currently traversed spatial grid location. These are vibration saturation characteristic parameters.

[0085] The calculated lifetime decay rate Store it back into the currently traversed coupled feature group. Perform the above calculation on each coupled feature group in the local aging acceleration region feature data set.

[0086] A temporary sorting list is created. The feature data set of the local accelerated aging region is traversed, and the lifetime decay rate stored in each coupled feature group and its corresponding coupled feature group information are added as pairs to the temporary sorting list. A sorting operation is performed on the temporary sorting list, comparing the lifetime decay rate values ​​of any two pairs in the list; the pair with the larger lifetime decay rate value is ranked first. When the lifetime decay rate values ​​of two pairs are equal, the multi-field coupling equivalent total temperature values ​​corresponding to the two pairs are compared; the pair with the higher multi-field coupling equivalent total temperature value is ranked first.

[0087] Read the pre-set first and second boundary point ratios. The first and second boundary point ratios are determined as follows: Count the total number of all spatial locations on the insulation layer surface; calculate the total number of locations that can be covered in this maintenance based on the number of maintenance teams available for this maintenance and the number of maintenance areas each team can cover per shift; divide the total number of coverable locations by the total number of spatial locations on the insulation layer surface to obtain the first boundary point ratio. The first boundary point ratio represents the maximum proportion of high-risk locations that must be prioritized in this round of maintenance. Multiply the first boundary point ratio by two to obtain the second boundary point ratio, which represents the maximum proportion of locations that are recommended for inspection in the near future in this round of maintenance. Count the total number of paired items in the temporary sorting list; multiply the total number by the first boundary point ratio and round up to obtain the threshold for the number of first-level high-priority locations; multiply the total number by the second boundary point ratio and round up to obtain the threshold for the number of second-level medium-priority locations.

[0088] Iterate through the temporary sorted list and assign a sort number to each pair of items in the list, with the sort numbers increasing sequentially from the beginning. When the sorting sequence number of the pairing item is less than or equal to the threshold for the number of first-level high-priority positions, add a failure risk priority field to the coupling feature group corresponding to the pairing item and set its value to the first-level identifier. When the sorting number of the pairing item is greater than the threshold for the number of first-level high-priority positions and less than or equal to the threshold for the number of second-level medium-priority positions, the failure risk priority field value in the coupling feature group corresponding to the pairing item is set to the second-level identifier. When the sorting number of the pairing item is greater than the threshold for the number of priority positions in the second level, the failure risk priority field in the coupling feature group corresponding to the pairing item is set to the third level identifier.

[0089] For each coupling feature group with assigned priority identifiers, the expected remaining lifetime is obtained by dividing the value of one hundred by the lifetime decay rate stored in the corresponding coupling feature group, and the expected remaining lifetime is stored in the corresponding coupling feature group.

[0090] When the expected remaining useful life of a coupled feature group is less than three years, an emergency warning flag of Boolean type true value is added to the corresponding coupled feature group; when the expected remaining useful life of a coupled feature group is greater than or equal to three years, an emergency warning flag of Boolean type false value is added to the corresponding coupled feature group.

[0091] A data table for predicting the lifespan of the thermal insulation layer partitions is created using a hash table data structure. The key of the hash table is the spatial location index, and the value is risk assessment information. The feature data set of the local aging acceleration zone, for which priority identifiers have been assigned, is traversed. The horizontal and vertical position numbers of the grid for each coupled feature group are combined into a spatial location index. The multi-field coupling equivalent total temperature, lifespan decay rate, failure risk priority identifier, estimated remaining service life, and emergency warning flag stored in the corresponding coupled feature group are read. These five data items are encapsulated into a risk assessment information structure, which is then stored in the thermal insulation layer partition lifespan prediction data table using the spatial location index as the key and the risk assessment information structure as the value.

[0092] Iterate through each coupling feature group marked as false in the aging acceleration region of the original spatial coupling feature set, and combine the horizontal and vertical position numbers of the mesh of the currently traversed coupling feature group into a spatial position index. Obtain the multi-field coupling equivalent total temperature from the currently traversed coupling feature group. Humidity affects depth ratio and vibration equivalent temperature value Temperature rise The same life decay rate formula as that used in high-risk areas is adopted. The positive / negative judgment logic calculates the lifetime decay rate at the currently traversed spatial grid location. The estimated remaining useful life is obtained by dividing the value by one hundred by the lifetime decay rate, and the failure risk priority is set to Level 4, the routine monitoring level. When the estimated remaining useful life is less than three years, the emergency warning flag is set to a Boolean true value; when the estimated remaining useful life is greater than or equal to three years, the emergency warning flag is set to a Boolean false value. The combined spatial location index, along with risk assessment information including multi-field coupling equivalent total temperature, lifetime decay rate, failure risk priority flag, estimated remaining useful life, and emergency warning flag, is added as key-value pairs to the insulation layer zonal lifetime prediction result data table.

[0093] Create a structured document to present the prediction results. The structured document includes a list of high-risk areas, a visualization table of regional lifetime predictions, and a priority grouping statistical report.

[0094] In the high-risk area list, all first-level high-priority locations, second-level medium-priority locations, and third-level low-priority locations are listed in descending order of failure risk priority. Each location entry includes a spatial location index, multi-field coupling equivalent total temperature, lifetime decay rate, and expected remaining lifetime. Locations marked as having an emergency warning have a red exclamation mark and the word "emergency" added to the end of the corresponding entry.

[0095] In the partitioned lifespan prediction visualization table, create a two-dimensional table with the number of rows equal to the maximum value of the vertical position number and the number of columns equal to the maximum value of the horizontal position number. Traverse all key-value pairs in the partitioned lifespan prediction result data table of the insulation layer, parse the spatial position index in each key-value pair to extract the horizontal and vertical numbers, and fill in the failure risk priority indicator and the estimated remaining service life in the cell at the intersection of the corresponding row and column in the table. Use a dark red background for the first-level high priority, an orange background for the second-level medium priority, a yellow background for the third-level low priority, and a green background for the fourth-level routine monitoring level. The remaining service life is displayed in white font, and cells marked as true for emergency warning are marked with a thick red border.

[0096] In the priority grouping statistical report, the total number of locations marked with priority level 1, 2, 3, and 4 and their percentage of the total number of locations are counted. All locations marked with emergency warning as true are compiled to form an emergency warning list. The average lifetime decay rate of level 1, 2, and 3 locations and the average lifetime decay rate of level 4 location within the local aging acceleration zone are calculated. By comparing the two average decay rates, the failure risk amplification factor of the local aging acceleration zone relative to the normal area is quantified.

[0097] The structured document is organized and formatted in the order of a high-risk area list, a zonal life prediction visualization table, and a priority group statistical report to form an output file of the insulation layer zonal life prediction results. The output file is saved in both electronic document format and printable paper format. The electronic document format allows clicking on any cell in the visualization table to pop up detailed risk assessment information including multi-field coupling equivalent total temperature, values ​​of each acceleration factor, and the percentage of humidity influence depth. The printable paper format optimizes the table layout and font size. On-site maintenance personnel can locate the specific locations that need to be repaired based on the grid coordinates in the table and arrange the order of maintenance work according to the priority labels and remaining life values.

[0098] Here, those skilled in the art will understand that the specific operations of each step in the above-described method for predicting the lifespan of polyurethane thermal insulation materials have been referenced above. Figures 1 to 2 The method for predicting the lifespan of polyurethane thermal insulation materials has been described in detail in the previous section, and therefore, its repeated description will be omitted.

[0099] According to another aspect of the present invention, a life prediction system for polyurethane thermal insulation materials is provided, comprising: The temperature reconstruction module is used to slice the acquired infrared temperature distribution data of the insulation layer according to radial depth, and perform spatial grid matching on the temperature value of each depth layer and the thickness measurement data to obtain multi-layer temperature field data. The coupling construction module is used to spatially align the multi-layer temperature field data with the environmental temperature and humidity time series data and the pipeline vibration acceleration data, and to convert the humidity penetration depth and vibration stress amplitude at each grid location into equivalent temperature values ​​through the aging acceleration model, thereby constructing a spatial coupling feature set. The aging positioning module is used to perform weighted summation of the temperature value, humidity equivalent temperature value and vibration equivalent temperature value at each grid position in the spatial coupling feature set to obtain the multi-field coupling equivalent total temperature, and sort and filter out the local aging acceleration zone according to the multi-field coupling equivalent total temperature. The risk ranking module is used to calculate the partitioned life decay rate of the feature data of the local aging acceleration zone, and to rank the failure risk priority of each local area according to the decay rate, generating a partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

[0100] In summary, the life prediction method for polyurethane thermal insulation materials based on the embodiments of the present invention has been clarified. By acquiring and processing infrared temperature distribution data of the polyurethane insulation layer, and combining it with environmental temperature and humidity data and pipeline vibration acceleration data, spatial alignment and fusion of multi-physics field data are achieved, significantly improving the ability to refine the modeling of material aging behavior. Based on the aging acceleration model, humidity penetration depth and vibration stress amplitude are converted into equivalent temperature values, and the equivalent total temperature of multi-field coupling is accurately calculated through the construction and weighted summation of multi-field coupling feature sets, thereby effectively identifying local aging acceleration regions. Simultaneously, by calculating the partitioned life decay rate of the characteristic data of the aging acceleration region, the priority ranking of local failure risks is achieved, providing efficient and accurate technical support for the life prediction and maintenance strategy of the insulation layer. It demonstrates significant advantages in addressing local aging problems in complex service environments, improving life prediction accuracy, and enhancing the applicability of risk assessment.

Claims

1. A method for predicting the lifespan of polyurethane thermal insulation materials, characterized in that, include: Infrared temperature distribution data of the insulation layer is acquired and sliced ​​according to radial depth. Spatial grid matching is performed on the temperature value and thickness measurement data of each depth layer to obtain multi-layer temperature field data. The multi-layer temperature field data is spatially aligned with the environmental temperature and humidity time series data and the pipeline vibration acceleration data. The humidity penetration depth and vibration stress amplitude at each grid location are converted into equivalent temperature values ​​through the aging acceleration model to construct a spatially coupled feature set. The temperature, humidity equivalent temperature, and vibration equivalent temperature values ​​at each grid location within the spatial coupling feature set are weighted and summed to obtain the multi-field coupling equivalent total temperature. Local aging acceleration zones are then sorted and selected based on the multi-field coupling equivalent total temperature. The characteristic data of the local accelerated aging zone are used to calculate the partitioned life decay rate. Based on the decay rate, the failure risk priority of each local area is sorted to generate the partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

2. The polyurethane thermal insulation material life prediction method according to claim 1, characterized by, The aging acceleration model includes a first aging acceleration model and a second aging acceleration model; The first aging acceleration model is based on the accelerating effect of humidity on the aging rate of thermal insulation materials, and the formula is expressed as: wherein, is a humidity equivalent temperature value; is a baseline reference temperature, is a dimensionless penetration depth factor, is a dimensionless moisture content factor, is a dimensionless diffusion time factor, is a denominator regularization parameter, is a dimensionless activation energy scaling factor, is a strength decay sensitivity coefficient, is a dimensionless strength decay factor, is a dimensionless penetration scaling factor, is an auxiliary function, is an ambient relative humidity, is an ambient temperature.

3. The polyurethane thermal insulation lifetime prediction method according to claim 2, characterized in that, The formula for calculating the penetration depth factor is as follows: wherein, is the measured humidity penetration depth, is the critical penetration depth of the material, is the effective penetration depth beyond the critical depth; The formula for the intensity attenuation factor is expressed as follows: wherein, is the current tensile strength of the material, is the initial tensile strength of the material, is the residual strength protection threshold of the material.

4. The polyurethane thermal insulation lifetime prediction method according to claim 3, characterized in that, The second aging acceleration model is based on the accelerating effect of vibration fatigue on the aging rate of thermal insulation materials, and the formula is expressed as follows: wherein, is a vibration equivalent temperature value; is a vibration stress amplitude calculated for the current grid position; is a reference stress amplitude threshold value, is a reference stress value, is a vibration acceleration factor, is a cyclic accumulation acceleration coefficient, is a cumulative number of vibration cycles, is a reference number of cycles for non-dimensionalizing the cyclic accumulation effect; is a stress standard deviation.

5. The method for predicting the lifespan of polyurethane thermal insulation materials according to claim 4, characterized in that, The calculation of the equivalent total temperature of the multi-field coupling includes: A normalized weighting coefficient is assigned to each depth layer, with the weighting coefficient of the inner surface depth layer being greater than that of the outer surface depth layer. The depth-weighted average temperature value is obtained by multiplying the temperature values ​​of each depth layer by the corresponding normalized weighting coefficient and then summing them. The percentage of depths affected by humidity is obtained by dividing the number of depth layers marked as humidity-affected layers by the total number of depth layers. Multiplying the humidity equivalent temperature value by the humidity influence depth ratio yields the humidity-weighted equivalent temperature. The multi-field coupling equivalent total temperature is obtained by summing the depth-weighted average temperature, the humidity-weighted equivalent temperature, and the vibration equivalent temperature.

6. The method for predicting the lifespan of polyurethane thermal insulation materials according to claim 5, characterized in that, The screening of the localized accelerated aging region includes: The coupling feature groups are arranged in descending order of the equivalent total temperature of the multi-field coupling. Starting from the second pair in the sorted list, calculate the equivalent total temperature difference between each pair and the previous pair in sequence, and record the position number of the pair with the largest temperature difference in the list as the natural break point number. The preliminary screening ratio is obtained by dividing the natural break point number by the total number of coupling feature groups. If the preliminary screening ratio is greater than 0.3, the screening ratio parameter is 0.3; if it is less than 0.1, the screening ratio parameter is 0.1; if it is between 0.1 and 0.3, the preliminary screening ratio is directly used. Multiply the total number of the coupled feature groups by the screening ratio parameter and round up to obtain the screening quantity threshold. Mark the top-ranked coupling feature groups with the highest screening quantity threshold as local aging acceleration zones.

7. The method for predicting the lifespan of polyurethane thermal insulation materials according to claim 6, characterized in that, The formula for calculating the lifetime decay rate is: in, This represents the lifetime decay rate at the currently traversed spatial grid location. This refers to the design service life of the insulation material under standard environmental conditions. It is a dimensionless temperature sensitivity coefficient. The increase in temperature For reference temperature, The temperature and humidity coupling enhancement coefficient is... The humidity level at the currently traversed spatial grid location affects the depth ratio. The humidity acceleration factor, The vibration acceleration coefficient, This represents the vibration equivalent temperature value at the currently traversed spatial grid location. These are vibration saturation characteristic parameters.

8. A lifespan prediction system for polyurethane thermal insulation materials, characterized in that, include: The temperature reconstruction module is used to slice the acquired infrared temperature distribution data of the insulation layer according to radial depth, and perform spatial grid matching between the temperature value of each depth layer and the thickness measurement data to obtain multi-layer temperature field data. The coupling construction module is used to spatially align the multi-layer temperature field data with the environmental temperature and humidity time series data and the pipeline vibration acceleration data, and to convert the humidity penetration depth and vibration stress amplitude at each grid location into equivalent temperature values ​​through the aging acceleration model, thereby constructing a spatial coupling feature set. The aging positioning module is used to perform weighted summation of the temperature value, humidity equivalent temperature value and vibration equivalent temperature value at each grid position in the spatial coupling feature set to obtain the multi-field coupling equivalent total temperature, and sort and filter out the local aging acceleration zone according to the multi-field coupling equivalent total temperature. The risk ranking module is used to calculate the partitioned life decay rate of the feature data of the local aging acceleration zone, and to rank the failure risk priority of each local area according to the decay rate, thereby generating a partitioned life prediction result of the insulation layer containing spatial location index and priority identifier.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method for predicting the lifespan of polyurethane thermal insulation materials as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for predicting the lifespan of polyurethane thermal insulation materials as described in any one of claims 1 to 7.

Images