Multi-level buffer lamination method and device for explosion prevention of door and window glass

Abstract

The invention provides a door and window glass explosion-proof multi-level buffer lamination method and device, relates to the field of multi-level buffer lamination, and solves the technical problems that a conventional lamination structure lacks elasticity modulus gradient design, and defect positions and mechanical responses cannot be associated. The method comprises the following steps: acquiring an optical interference image, an ultrasonic echo signal, an infrared thermal image, a light transmittance value and a chromatic aberration delta E value of the door and window glass buffer layer, and analyzing and processing through five parallel channels of a data analysis module to obtain a defect feature vector and a final defect parameter. A door and window glass explosion-proof multi-level buffer layer is divided into an edge sealing layer, a middle buffer layer, a fiber reinforcement layer and a base body glass layer from outside to inside, a multi-level stress buffer mechanism is formed, and a stress transmission rule that impact force is transmitted to the next layer according to the level attenuation rate larger than or equal to 30% is preset. And constructing a defect and hierarchy incidence matrix, and calculating the security level of an output buffer layer. The method and the device are used in a multi-level buffer lamination process.

Description

TECHNICAL FIELD

[0001] The application relates to the field of multilevel cushion lamination, in particular to a multilevel cushion lamination method and device for preventing explosion of door and window glass. BACKGROUND

[0002] The multilevel cushion lamination process for preventing explosion of door and window glass refers to a technical means for improving the explosion-proof performance of door and window glass by compounding multiple material layers with different functions and mechanical properties in a specific process to form a laminated structure capable of gradually absorbing and dispersing external impact force. The existing cushion lamination process mostly adopts a simple double-layer or triple-layer composite structure, commonly combined as "surface protection film + base glass" or "cushion interlayer + base glass". For example, some technologies use a single thickness of polyvinyl butyral (PVB) interlayer directly combined with tempered glass, or only apply ordinary sealant to the edges of the glass. Therefore, there is a lack of gradient planning of the elastic modulus of each level, and no clear impact force level attenuation rule is established. Usually, only the material properties of a single cushion layer are relied on to achieve energy absorption, and it is impossible to achieve step-by-step reduction of impact force through the synergistic effect of multiple levels. Moreover, due to the simple compounding of double and triple layers, it is impossible to accurately evaluate the safety state of each level through defect information, such as determining whether the delamination defect is located in the key cushion layer affecting the transmission of impact force, or developing targeted optimization or maintenance schemes based on the defect position. Ultimately, the explosion-proof performance stability of laminated glass is insufficient, and the safety evaluation lacks precise spatial basis. SUMMARY

[0003] The application provides a multilevel cushion lamination method and device for preventing explosion of door and window glass, which solves the technical problems of the conventional laminated structure lacking elastic modulus gradient design and the inability to correlate defect position and mechanical response in the prior art.

[0004] To achieve the above-mentioned purposes, the application adopts the following technical solutions: In a first aspect, a multilevel buffer lamination method for preventing explosion of a door and window glass comprises: obtaining an optical interference image, an ultrasonic echo signal, an infrared thermal image, a light transmittance value and a color difference ΔE value of a door and window glass buffer layer to form original data. The original data is input into a data analysis module, and the optical interference image, the ultrasonic echo signal, the infrared thermal image, the light transmittance value and the color difference ΔE value are analyzed and processed in parallel through five channels of the data analysis module, and the bubble area inside the door and window glass buffer layer, the profile of the delamination area, the defect depth and size, the position of the cold spot in the delamination area, the fogging level and the yellowing level are extracted to obtain a defect feature vector and final defect parameters. The multilevel buffer layer for preventing explosion of the door and window glass is divided from the surface to the inside into an edge sealing layer, a middle buffer layer, a fiber reinforced layer and a base glass layer to form a multilevel stress buffer mechanism, and a preset impact force is transmitted to the next layer with a stress transmission rule of a level attenuation rate ≥ 30%. The level parameters in the multilevel stress buffer mechanism are obtained, and the defect feature vector is input into an evaluation module synchronously to construct a defect and level correlation matrix, and the safety level of the buffer layer is calculated and output through a buffer efficiency index.

[0005] Based on the above technical solution, in the multilevel buffer lamination method for preventing explosion of the door and window glass provided in the present application, each layer is allowed to have a preset impact force attenuation rate ≥ 30% through a four-level composite structure, and a spatial coordinate system is constructed through a B-spline surface equation to map the defect coordinates to the level position in an affine manner to form a hierarchical stress buffer and defect spatial mapping, so that the impact force can be attenuated gradually during the lamination process (such as 30% energy consumption of the sealing layer → 35% energy consumption of the PVB layer), the effect of directional energy dissipation is achieved by avoiding stress concentration, and according to the accurate correlation between the defect and the level, the positions of the bubbles and the delamination in the level can be located (such as the cold spot located at the PVB layer Z = 1.2 mm), which provides a spatial basis for safety evaluation and solves the problems of lack of elastic modulus gradient design in conventional lamination structure and inability to correlate the defect position and mechanical response.

[0006] In conjunction with the first aspect mentioned above, in one possible implementation, the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value are analyzed and processed simultaneously through five parallel channels of the data analysis module. This process extracts the bubble area, the outline of the delamination area, the defect depth and size, the location of cold spots in the delamination area, the fogging level, and the yellowing level within the buffer layer of the window glass. The result is a defect feature vector and final defect parameters. This includes: inputting the optical interference image from the original data into the optical interference image processing channel within the data analysis module; performing fringe morphology analysis on the optical interference image; identifying distorted areas of the interference fringes; and calculating the bubble area and the outline coordinates of the delamination area based on the fringe spacing variation. The ultrasonic echo signal from the original data is input into the ultrasonic echo signal processing channel within the data analysis module; performing time-frequency feature extraction on the ultrasonic echo signal; calculating the defect depth using the time difference between the reflected and emitted waves; and quantifying the defect size using the signal attenuation amplitude. The infrared thermal image from the original data is input into the infrared thermal image processing channel within the data analysis module. Temperature gradient analysis is performed on the infrared thermal image to identify temperature anomaly areas and locate the center coordinates and coverage area of ​​cold spots in the delamination area. The transmittance value from the original data is input into the transmittance value processing channel within the data analysis module. The ambient temperature value is obtained through a temperature and humidity sensor. Based on the real-time ambient temperature value, a reference glass transmittance is preset. The transmittance value of the original data is compared with the transmittance of the reference glass. If the difference exceeds 5%, the fogging level is determined to be high-risk; if the difference is between 3% and 5%, it is medium-risk; and if the difference is less than 3%, it is low-risk. The color difference ΔE value from the original data is input into the color difference ΔE value processing channel within the data analysis module. Based on the Lab color space model, the highest and lowest thresholds for color difference aging are adjusted proportionally according to the service life of the glass. When ΔE value > the highest threshold, the yellowing level is determined to be severe aging; when the lowest threshold < ΔE value ≤ the highest threshold, it is determined to be moderate aging; and when ΔE value ≤ the lowest threshold, it is determined to be slight aging. The 5-channel analysis structure in the data analysis module is encoded into a structured vector, denoted as the defect feature vector output. The structure of the defect feature vector is [bubble area, degumming contour, defect depth, defect size, cold spot location, fogging level, yellowing level]. This is based on a confidence-weighted fusion formula. Calculate the final parameter ZL, where w i Let D be the confidence weight of the i-th detection channel. i is the original defect parameter value output by the i-th detection channel, and n is the number of effective detection channels participating in the fusion.

[0007] In conjunction with the first aspect mentioned above, in one possible implementation, the parallel 5-channel synchronous analysis and processing of the optical interference image, ultrasonic echo signal, infrared thermal image, transmittance value, and color difference ΔE value by the data analysis module further includes: when the optical interference channel detects a bubble area > 5mm... 2However, if the ultrasonic channel fails to identify the enhanced acoustic reflection signal at the corresponding coordinates, the infrared channel is activated to re-inspect the area, using the infrared cold spot coverage as the final bubble area. If the transmittance channel determines a high risk of fogging, but the color difference ΔE value is ≤1, the optical interference channel is triggered to check for contamination on the glass surface. When the standard deviation of the confidence-weighted result is >30%, the historical case library is called to match data under the same working conditions, and the case parameters with a similarity >85% are selected as the arbitration output.

[0008] In conjunction with the first aspect mentioned above, in one possible implementation, the multi-layered buffer layer for explosion-proof window glass is divided from the outside in into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a base glass layer. This includes: an edge sealing layer composed of a polymer sealant, 0.5–1.0 mm thick, covering the glass edge area to absorb shear stress and prevent moisture penetration; an intermediate buffer layer composed of polyvinyl butyral (PVB) or ionic polymer solid gluconate (SGP) interlayer material, 1.0–1.8 mm thick, located inside the sealing layer, used to disperse impact energy through viscoelastic deformation; a fiber reinforcement layer composed of a woven mesh of carbon fiber and glass fiber, with a mesh density of 60–80 mesh, embedded between the intermediate buffer layer and the base glass layer to inhibit crack propagation and improve interlayer peel strength; and a base glass layer of tempered glass, ≥5 mm thick, used to withstand residual stress after impact and maintain overall structural integrity. The layers are composited using a hot-pressing process, and the elastic modulus gradient attenuation rate between adjacent layers is ≥30%.

[0009] In conjunction with the first aspect mentioned above, one possible implementation involves constructing a defect-layer correlation matrix, including: building a spatial coordinate system based on the edge sealing layer, intermediate buffer layer, fiber reinforcement layer, and matrix glass layer. The bubble contour coordinates, debonded area contours, and cold spot location coordinates from the defect feature vector are mapped to the layer coordinate system through affine transformation, outputting a defect coordinate set with layer identifiers. For each defect location, a convolution operation is performed between the layer material properties and the defect features to generate the defect-layer correlation matrix.

[0010] In conjunction with the first aspect mentioned above, in one possible implementation, a spatial coordinate system is constructed based on the edge sealing layer, the intermediate buffer layer, the fiber reinforcement layer, and the base glass layer. This includes: based on the coating trajectory of the polymer sealant in the edge sealing layer, scanning the edge of the glass annular surface using a laser profilometer to generate three-dimensional point cloud data of the sealing layer to establish a first sub-coordinate system, and establishing the radial coordinates of the first sub-coordinate system. axial coordinates The origin of the coordinate system is the center line of the sealant coating, where R is the glass radius and T is the center line of the sealant coating. sealThe thickness of the edge sealing layer is 0.5–1.0 mm. Based on the viscoelastic parameters of the PVB or SGP interlayer material of the intermediate buffer layer, a thickness gradient of 1.0–1.8 mm is transformed into a non-uniform grid on the Z-axis and mapped to the first subcoordinate system to form a spatial coordinate system, and the range of the Z-axis is The plane grid resolution is 1 mm×1 mm, where T mid is the thickness of the intermediate buffer layer of 1.0–1.8 mm. Based on the intersection points of carbon fiber or glass fiber grids in the fiber reinforcement layer, through the B-spline surface control equation The actual coordinate points S(u,v) in the three-dimensional space are obtained to map the grid intersection coordinates of the fiber reinforcement layer, where P i,j is the intersection coordinate of the carbon fiber or glass fiber grid, and N i,p (u) is the p-th B-spline basis function, and N j,q (v) is the q-th B-spline basis function along the parameter v direction, and (u,v) is the parameter domain coordinate. A constant interval is set for the Z-axis of the matrix glass layer , and the plane boundary coincides with the glass cutting line for binding the rigid boundary, where T glass is the thickness of the tempered glass base layer.

[0011] Combined with the first aspect above, in a possible implementation, the safety level of the buffer layer is calculated through the buffer efficiency index, including: according to the association matrix of defects and levels, the defect density factor δ and stress sensitivity coefficient ε corresponding to each level are extracted. Based on the preset rules of the multi-level stress buffer mechanism, the impact force attenuation rate η of each level is obtained. Based on the defect density factor δ, stress sensitivity coefficient ε and impact force attenuation rate η, through the buffer efficiency index formula The overall safety value BEI is calculated. The safety level is output according to the BEI value range. BEI≤0.25 is the allowable service safety level, 0.25<BEI≤0.5 is the warning observation safety level, and BEI>0.5 is the high-risk replacement safety level.

[0012] Combined with the first aspect above, in a possible implementation, after the safety level of the buffer layer is output, it further includes: obtaining the historical experimental impact force waveform data associated with the safety level and the material constitutive model of the multi-level stress buffer mechanism. Based on the association matrix of defects and levels, buffer efficiency index BEI and material constitutive model, a finite element model of the multi-level buffer layer with defects is reconstructed in ANSYS / LS-DYNA. The historical experimental impact force waveform is input as the boundary condition into the finite element model of the multi-level buffer layer, and explicit dynamic simulation is performed to dynamically output the stress distribution nephogram of each level and identify the stress concentration area. According to the comparison between the stress concentration area and the preset stress threshold, the weak level and its stress overrun ratio are identified.

[0013] In conjunction with the first aspect mentioned above, one possible implementation method, after identifying the weak layers, further includes: extracting the peak stress σmax and location coordinates of the stress concentration region, and calculating the actual impact force attenuation rate of each layer. If the actual impact force attenuation rate is <30% or the stress exceedance ratio is >5%, based on the identification and quantitative evaluation results of the weak layers, a targeted set of layer parameter optimization suggestions is generated. The set of layer parameter optimization suggestions includes: if the edge sealing layer is a weak layer, it is recommended to increase the coating thickness of its polymer sealant or use a sealing material with a higher elastic modulus; if the intermediate buffer layer is a weak layer, it is recommended to adjust the thickness of its PVB or SGP interlayer or optimize the viscoelastic constitutive model parameters; if the fiber reinforcement layer is a weak layer, it is recommended to increase the density of the carbon fiber or glass fiber mesh or increase the diameter of a single fiber; if the matrix glass layer is a weak layer, it is recommended to increase the thickness of its tempered glass base layer.

[0014] Secondly, a multi-level buffer lamination device for explosion-proof window and door glass is provided, comprising: a communication unit and a processing unit; the communication unit is used to acquire optical interference images, ultrasonic echo signals, infrared thermal images, transmittance values, and color difference ΔE values ​​of the window and door glass buffer layer to form raw data; the processing unit is used to input the raw data into a data analysis module, and to analyze and process the optical interference images, ultrasonic echo signals, infrared thermal images, transmittance values, and color difference ΔE values ​​through parallel 5-channel synchronous analysis, extracting the bubble area, the outline of the delamination area, the defect depth and size, the location of cold spots in the delamination area, the fogging level, and the yellowing level inside the window and door glass buffer layer, to obtain the defect feature vector and the final defect parameters. The multi-level buffer layer for explosion-proof window and door glass is divided from the surface to the inside into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a base glass layer, forming a multi-level stress buffer mechanism, and a preset stress transmission rule that the impact force is transmitted to the next layer according to a layer attenuation rate ≥30%. The parameters of each level in the multi-level stress buffer mechanism are obtained and synchronously input into the evaluation module along with the defect feature vector. The correlation matrix between defects and levels is constructed, and the safety level of the buffer layer is calculated and output through the buffer effectiveness index.

[0015] Thirdly, this application provides a multi-level buffer lamination device for explosion-proof window and door glass, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the method described in the first aspect and any possible implementation thereof. This multi-level buffer lamination device for explosion-proof window and door glass can be an electronic device or a chip within an electronic device.

[0016] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on the multi-level buffer lamination device for explosion-proof window and door glass, cause the multi-level buffer lamination device to perform the method described in the first aspect and any possible implementation thereof.

[0017] Fifthly, this application provides a computer program product containing instructions that, when the computer program product is run on the multi-level buffer lamination device for explosion-proof window and door glass, causes the multi-level buffer lamination device for explosion-proof window and door glass to perform the method described in the first aspect and any possible implementation thereof.

[0018] This application provides a multi-level buffer lamination method and device for explosion-proof window and door glass. It utilizes a four-layer composite structure to pre-determine an impact force attenuation rate of ≥30% for each layer. Furthermore, a spatial coordinate system is constructed using B-spline surface equations, affinely mapping defect coordinates to layer positions, forming a hierarchical stress buffer and defect spatial mapping. This allows for progressive attenuation of impact force during the lamination process (e.g., 30% energy dissipation in the sealing layer → 35% energy dissipation in the PVB layer), avoiding stress concentration and achieving directional energy dissipation. Simultaneously, based on the precise correlation between defects and layers, the location of bubbles and delamination within the layers can be pinpointed (e.g., cold spots located in the PVB layer at Z=1.2mm), providing spatial basis for safety assessment. This addresses the problem of conventional lamination structures lacking elastic modulus gradient design and the inability to correlate defect locations with mechanical responses.

[0019] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0020] Figure 1 A system architecture diagram of a multi-level buffer lamination system for explosion-proof door and window glass provided in this application embodiment; Figure 2A flowchart illustrating a multi-level buffer lamination method for explosion-proof window and door glass provided in this application embodiment; Figure 3 A flowchart illustrating another multi-level buffer lamination method for explosion-proof window and door glass provided in this application embodiment; Figure 4 A flowchart illustrating another multi-level buffer lamination method for explosion-proof window and door glass provided in this application embodiment; Figure 5 A flowchart illustrating another multi-level buffer lamination method for explosion-proof window and door glass provided in this application embodiment; Figure 6 This is a schematic diagram of a multi-level buffer lamination device for explosion-proof door and window glass provided in an embodiment of this application. Detailed Implementation

[0021] In the description of this application, unless otherwise stated, "" means "or," for example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The words "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0022] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0023] To address the lack of elastic modulus gradient design and the absence of a multi-layered collaborative stress buffering mechanism in existing conventional laminated structures, simple two- or three-layer composite structures are often employed. These structures fail to achieve orderly transmission of impact force through the gradient attenuation of elastic modulus between adjacent layers, relying solely on the material properties of a single layer to absorb energy. This easily leads to localized stress concentration and fails to meet the requirement of effective impact force attenuation across layers, resulting in limited buffering efficiency. Furthermore, the lack of a multi-dimensional, synchronous defect detection system, with some technologies employing only single detection methods (such as optical image recognition of surface defects or ultrasonic detection of internal defects), fails to comprehensively extract multi-dimensional defect features such as bubble area, delamination contour, defect depth, cold spot location, fogging level, and yellowing level, resulting in low accuracy in defect parameter extraction. Furthermore, since defects cannot be associated with specific layers, it is impossible to construct a correlation matrix between defects and layers. Consequently, it is impossible to establish a connection between defect characteristics and the mechanical response of each layer, and it is impossible to assess the stress buffering capacity of each layer through defect information. This leads to a lack of scientific basis for safety assessment. This application provides a multi-level buffer lamination method for explosion-proof door and window glass. This method uses a four-layer composite structure to preset an impact force attenuation rate of ≥30% for each layer. A spatial coordinate system is constructed using B-spline surface equations, and defect coordinates are affinely mapped to layer positions, forming a hierarchical stress buffer and defect spatial mapping. This allows the impact force to be gradually attenuated during the lamination process (e.g., the sealing layer consumes 30% of the energy → the PVB layer consumes another 35%), avoiding stress concentration and achieving directional energy dissipation. At the same time, based on the precise correlation between defects and layers, the positions of bubbles and delamination in the layers can be located (e.g., cold spots are located in the PVB layer at Z=1.2mm), providing a spatial basis for safety assessment. This solves the problem that conventional lamination structures lack elastic modulus gradient design and that defect positions cannot be correlated with mechanical responses.

[0024] like Figure 1 As shown in the embodiment of this application, a multi-level buffer lamination method for explosion-proof window and door glass includes: Step 101: Obtain the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value and color difference ΔE value of the window glass buffer layer to form the original data.

[0025] Among them, the optical interferometric image is a fringe pattern captured by an optical interferometer, showing the optical wave interference distortion inside the buffer layer, and is used to identify bubbles and delamination areas; the ultrasonic echo signal is a sound wave signal emitted and received by an ultrasonic probe, and the time difference and attenuation amplitude of the reflected wave reflect the depth and size of the defect; the infrared thermal image is a temperature gradient distribution image acquired by an infrared thermal imager, used to locate the cold spot position in the delamination area; the transmittance value is a value that measures the light transmission performance of the glass, and the fogging level is determined by comparing it with a reference value; the color difference ΔE value is a color difference value based on the Lab color space model, used to quantify the degree of yellowing and aging; the raw data is the initial dataset formed by the above multi-source measurement data, which serves as the input for subsequent analysis.

[0026] In some implementations, optical interferometers (such as laser profilometers) are used to perform non-contact scanning of the buffer layer of window and door glass. Optical interferometric images are obtained by analyzing the distorted areas of the light wave interference fringes to identify bubble outlines and delamination areas. Then, an ultrasonic probe (frequency such as 2.5MHz or 5MHz) emits ultrasonic pulses towards the buffer layer and receives the reflected echo signals. The defect depth is calculated by measuring the time difference between the emitted and reflected waves, while the defect size is quantified using the signal attenuation amplitude. Next, an infrared thermal imager (such as a device equipped with temperature gradient analysis capabilities) is deployed to capture the temperature distribution on the surface of the buffer layer, generating an infrared thermal image. The system is used to locate the center coordinates and coverage area of ​​cold spots in the debonding area; simultaneously, a transmittance meter (combined with a temperature and humidity sensor) is used to measure the transmittance of the glass, and the transmittance is compared with that of a reference glass preset with real-time ambient temperature to determine whether the difference exceeds a threshold and thus determine the level of fogging risk; in addition, a colorimeter is used to measure the color difference ΔE value in the Lab color space model, and the aging threshold is dynamically adjusted according to the service life of the glass to assess the yellowing level; finally, all acquired optical interference images, ultrasonic echo signals, infrared thermograms, transmittance values, and color difference ΔE values ​​are integrated into a structured dataset to form the raw data output for subsequent processing.

[0027] It should be noted that the synchronization of all measuring devices and the consistency of data must be ensured.

[0028] Step 102: Input the original data into the data analysis module. The data analysis module analyzes and processes the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value and color difference ΔE value in parallel through 5 channels. It extracts the bubble area, the outline of the delamination area, the defect depth and size, the location of cold spots in the delamination area, the fogging level and yellowing level, and obtains the defect feature vector and the final defect parameters.

[0029] The data analysis module refers to the algorithm system that receives the original data and performs multi-channel processing; the 5 channels refer to the optical interference image processing channel, ultrasonic echo signal processing channel, infrared thermal image processing channel, transmittance value processing channel, and color difference ΔE value processing channel; the bubble area is the cavity area calculated from the fringe distortion area in the optical interference image; the debonding area contour is the coordinate of the bonding failure boundary identified by optical interference or ultrasonic waves; the defect depth is the vertical position of the defect converted from the time difference between the ultrasonic reflected wave and the emitted wave; the defect size is the physical size of the defect quantified by the attenuation amplitude of the ultrasonic signal; the cold spot position in the debonding area is the center coordinate and coverage area of ​​the temperature anomaly area in the infrared thermal image; the fogging level is the risk level (low, medium, high risk) divided by the difference between transmittance and the reference value; the yellowing level is the degree of aging (slight, moderate, severe aging) determined by comparing the color difference ΔE value with the dynamic aging threshold; the defect feature vector is a structured vector; and the final defect parameter is the comprehensive defect value calculated by the confidence-weighted fusion formula.

[0030] In some implementations, after the raw data is input into the data analysis module, the optical interference image processing channel first performs fringe morphology analysis on the optical interference image to identify distorted interference fringe regions, and calculates the bubble area and the contour coordinates of the debonded area based on the fringe spacing variation. Simultaneously, the ultrasonic echo signal processing channel extracts the time-frequency characteristics of the ultrasonic signal, calculates the defect depth through the time difference between the reflected and emitted waves, and quantifies the defect size based on the signal attenuation amplitude. The infrared thermal image processing channel performs temperature gradient analysis on the infrared thermal image to locate the center coordinates and coverage area of ​​the cold spot in the debonded area. The transmittance numerical processing channel calls the temperature and humidity sensor. The instrument acquires the ambient temperature value, presets the transmittance of the reference glass, and compares the measured transmittance value with the reference value. If the difference exceeds 5%, it is judged as high-risk fogging; if the difference is between 3% and 5%, it is medium-risk; and if it is less than 3%, it is low-risk. The color difference ΔE value processing channel is based on the Lab color space model and dynamically adjusts the color difference aging threshold according to the service life of the glass. When the ΔE value is greater than the highest threshold, it is judged as severe aging; when it is between the lowest and highest thresholds, it is judged as moderate aging; and when it is less than or equal to the lowest threshold, it is judged as slight aging. The output results of each channel are encoded into a structured defect feature vector, and finally the final defect parameters are calculated by the confidence weighted fusion formula.

[0031] The multi-layered explosion-proof buffer layer for doors and windows is divided from the outside in into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a base glass layer, including: The edge sealing layer is made of polymer sealant with a thickness of 0.5–1.0 mm, covering the glass edge area to absorb shear stress and prevent water vapor penetration.

[0032] The intermediate buffer layer is made of polyvinyl butyral (PVB) or ionic polymer SGP interlayer material, with a thickness of 1.0–1.8 mm, and is located inside the sealing layer to disperse impact energy through viscoelastic deformation.

[0033] The fiber reinforcement layer is composed of a mesh woven from a mixture of carbon fiber and glass fiber, with a mesh density of 60–80 mesh. It is embedded between the intermediate buffer layer and the matrix glass layer to suppress crack propagation and improve interlayer peel strength.

[0034] The base glass layer is tempered glass with a thickness of ≥5mm, used to withstand residual stress after impact and maintain the integrity of the overall structure. The layers are bonded together using a hot-pressing process, and the elastic modulus gradient attenuation rate between adjacent layers is ≥30%.

[0035] In some implementations, the explosion-proof buffer layer of window glass is divided into four independent functional zones (edge ​​sealing layer, intermediate buffer layer, fiber reinforcement layer, and base glass layer). Each layer is specifically designed in terms of material, thickness, and function to form a synergistic buffering mechanism. This allows the edge sealing layer to absorb shear stress (such as edge stress caused by wind pressure or impact), block moisture and ultraviolet penetration, and prevent edge delamination and corrosion. The intermediate buffer layer disperses impact energy (such as from hail or impact) through viscoelastic deformation, converting the shock wave into multi-directional stress and reducing energy transfer inwards. The fiber reinforcement layer inhibits crack propagation (the grid structure blocks crack paths) and improves interlayer peel strength (fiber bridging effect), preventing delamination failure. The base glass layer bears residual impact stress, maintaining the overall structural rigidity and integrity, and preventing glass shattering and splashing. Furthermore, during the actual lamination process, the impact force is gradually attenuated layer by layer (e.g., the sealing layer consumes 30% of the energy, and the buffer layer consumes another 35%), achieving directional energy dissipation. This also ensures that the modulus gradient between layers avoids abrupt stress changes.

[0036] Step 103: Divide the multi-level buffer layer of the explosion-proof glass of doors and windows into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer and a base glass layer from the outside to the inside, forming a multi-level stress buffer mechanism, and preset the stress transmission rule that the impact force is transmitted to the next layer according to the layer attenuation rate ≥30%.

[0037] The edge sealing layer refers to a 0.5–1.0 mm thick polymer sealant layer covering the glass edge area to absorb shear stress and prevent water vapor penetration. The intermediate buffer layer refers to a 1.0–1.8 mm thick polyvinyl butyral (PVB) or ionic polymer SGP interlayer material located inside the sealing layer, dispersing impact energy through viscoelastic deformation. The fiber reinforcement layer refers to a 60–80 mesh mesh layer woven from carbon fiber and glass fiber, embedded between the intermediate buffer layer and the base glass layer, to inhibit crack propagation and improve interlayer peel strength. The base glass layer refers to a tempered glass substrate with a thickness of ≥5 mm, which bears residual impact stress and maintains structural integrity. The multi-level stress buffering mechanism is an energy attenuation system formed by a four-level composite structure. The impact force attenuation rate of ≥30% means that the impact force transmitted from each level to the next level is reduced by at least 30%.

[0038] In some implementations, the window glass buffer layer first undergoes a hot-pressing process to composite an edge sealing layer. A polymer sealant is applied to the glass edge in a circular pattern, with a thickness controlled between 0.5 and 1.0 mm to absorb shear stress. Next, an intermediate buffer layer is composited, using PVB or SGP interlayer material with a thickness of 1.0 to 1.8 mm. This layer disperses impact energy into multi-directional stress waves through viscoelastic deformation. Subsequently, a fiber reinforcement layer is embedded, consisting of carbon fiber and glass fiber woven into a mesh structure at a density of 60-80 mesh, covering the surface of the intermediate buffer layer to inhibit lateral crack propagation. Finally, a composite matrix glass layer serves as the supporting matrix, with tempered glass of ≥5 mm thickness providing the final load-bearing capacity. The elastic modulus gradient attenuation rate between each layer is ≥30%, forming a mechanism for the attenuation and transmission of impact force at different levels: when the impact force acts on the edge sealing layer, the force transmitted to the intermediate buffer layer attenuates by ≥30%, then from the intermediate buffer layer to the fiber reinforcement layer by ≥30%, and finally from the fiber reinforcement layer to the matrix glass layer by ≥30%, achieving multi-level energy dissipation.

[0039] Step 104: Obtain the parameters of each level in the multi-level stress buffering mechanism, input them synchronously with the defect feature vector into the evaluation module, construct the correlation matrix between defects and levels, and calculate and output the safety level of the buffer layer through the buffer effectiveness index.

[0040] Among them, the hierarchical parameters refer to material properties such as the thickness of the edge sealing layer, the viscoelastic coefficient of the intermediate buffer layer, the mesh density of the fiber reinforcement layer, and the thickness of the matrix glass layer; the evaluation module refers to the algorithm unit that processes the association between the hierarchy and the defect; the correlation matrix between the defect and the hierarchy is the result of the convolution operation that maps the defect location and the hierarchy attributes to the spatial coordinate system; the buffer effectiveness index BEI is the calculated safety value; the safety level is divided into three levels: permissible service, early warning observation, and high-risk replacement.

[0041] In some implementations, the parameters of each level in the multi-level stress buffering mechanism are first obtained: the thickness of the polymer sealant in the edge sealing layer (0.5–1.0 mm), the viscoelastic modulus of the PVB / SGP in the intermediate buffer layer (1.0–1.8 mm thick), the carbon or glass fiber mesh density in the fiber-reinforced layer (60–80 mesh), and the thickness of the tempered glass in the matrix glass layer (≥5 mm). After synchronously inputting the level parameters and defect feature vectors into the evaluation module, a first sub-coordinate system is established based on the coating trajectory of the edge sealing layer, and the Z-axis range of the intermediate buffer layer is expanded. The fiber-reinforced layer mesh can then be mapped through the B-spline surface equation. The coordinates of the intersection points are determined, and the Z-axis interval is set for the substrate glass layer. The coordinates of the bubble outline, cold spot position, etc. of the defect feature vector are mapped to the hierarchical coordinate system through affine transformation to generate a defect coordinate set with hierarchical labels. For each defect location, the convolution operation of the hierarchical material properties (such as the elastic modulus of the sealant) and the defect features (such as the bubble area) is calculated to generate the correlation matrix between the defect and the hierarchy. Finally, the defect density factor δ and stress sensitivity coefficient ε of each hierarchy are extracted from the correlation matrix. Combined with the preset impact force attenuation rate η (≥30%), the overall safety value is calculated through the buffer effectiveness index formula and the corresponding safety level is output.

[0042] Based on the above technical solution, a four-layer composite structure is used to pre-determine an impact force attenuation rate of ≥30% for each layer. A spatial coordinate system is constructed using B-spline surface equations, and the defect coordinates are affinely mapped to the layer positions, forming a hierarchical stress buffer and defect spatial mapping. This allows the impact force to be gradually attenuated during the lamination process (e.g., the sealing layer consumes 30% of the energy → the PVB layer consumes another 35%), avoiding stress concentration and achieving directional energy dissipation. At the same time, based on the precise correlation between defects and layers, the positions of bubbles and delamination within the layers can be located (e.g., cold spots are located in the PVB layer at Z=1.2mm), providing spatial basis for safety assessment and solving the problems of conventional laminated structures lacking elastic modulus gradient design and the inability to correlate defect positions with mechanical responses.

[0043] In one possible implementation of the embodiments of this application, combined with Figure 1 ,like Figure 2 As shown, the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value are analyzed and processed simultaneously through 5 parallel channels of the data analysis module. The bubble area, outline of the delamination area, defect depth and size, location of cold spot in the delamination area, fogging level and yellowing level are extracted from the buffer layer of the window glass. The defect feature vector and final defect parameters can be obtained through the following steps 201 to 207, which are explained in detail below: Step 201: Input the optical interference image from the original data into the optical interference image processing channel in the data analysis module, perform fringe morphology analysis on the optical interference image, identify the distorted areas of the interference fringes, and calculate the bubble area and the contour coordinates of the degummed area based on the fringe spacing change.

[0044] The optical interference image processing channel is a submodule specifically designed for processing such images. Fringe morphology analysis is the process of analyzing the shape and pattern of interference fringes. Distortion regions of interference fringes are the parts of the fringes that exhibit abnormal deformation. Variations in fringe spacing represent the differences in distance between fringes. Bubble area is the size of the area occupied by the bubble. Debonded region contour coordinates are the coordinate values ​​of the boundary of the debonded region.

[0045] In some implementations, the optical interference image from the raw data is input into the optical interference image processing channel within the data analysis module. Then, the image is analyzed for fringe morphology in this channel, including detecting the direction, density, and shape changes of the fringes. Based on the analysis results, the distorted regions of the interference fringes are identified, which usually indicate the presence of defects. Subsequently, the area size of the bubble is calculated based on the change in the fringe spacing, and the contour coordinates of the degummed area are further determined.

[0046] It should be noted that the optical interference image input to the optical interference image processing channel meets the analysis requirements, clearly displays the fringe pattern, and the processing channel is pre-configured for this type of analysis, providing a reference value for the standard fringe spacing. This makes the calculation process dependent on the quantification of the fringe spacing variation, but only based on the given features, without involving additional preprocessing or postprocessing steps.

[0047] Step 202: Input the ultrasonic echo signal from the original data into the ultrasonic echo signal processing channel in the data analysis module, extract the time-frequency features of the ultrasonic echo signal, calculate the defect depth by the time difference between the reflected wave and the emitted wave, and quantify the defect size by the signal attenuation amplitude.

[0048] The ultrasonic echo signal processing channel is a submodule specifically designed for processing this type of signal. Time-frequency feature extraction is the process of simultaneously analyzing the time and frequency domain features of the signal. The reflected wave is the ultrasonic wave that returns after encountering a defect. The emitted wave is the initially emitted ultrasonic wave. The time difference is the time interval between the reflected and emitted waves. The defect depth is the location and depth of the defect within the material. The signal attenuation amplitude is the degree to which the ultrasonic energy weakens during propagation. The defect size is a measure of the size of the defect.

[0049] In some implementations, the ultrasonic echo signal from the raw data is input into the ultrasonic echo signal processing channel within the data analysis module. Then, time-frequency features of the ultrasonic echo signal are extracted in this channel to obtain the comprehensive characteristics of the signal in time and frequency. Next, the extracted features are used to identify the reflected wave, and the depth of the defect is calculated based on the propagation speed of the ultrasonic wave in the material by calculating the time difference between the reflected wave and the emitted wave. Finally, the size of the defect is quantified by analyzing the attenuation amplitude of the ultrasonic signal.

[0050] It should be noted that historical data is used to obtain a constant propagation speed of ultrasound in the material, and a quantifiable relationship exists between the signal attenuation amplitude and the defect size. The optical interferometry image processing channel only processes the given signal characteristics and does not involve additional steps such as environmental calibration or material property compensation.

[0051] Step 203: Input the infrared thermal image from the original data into the infrared thermal image processing channel in the data analysis module, perform temperature gradient analysis on the infrared thermal image, identify abnormal temperature areas, and locate the center coordinates and coverage area of ​​the cold spot in the degummed area.

[0052] The infrared thermal imaging processing channel is a specific submodule dedicated to processing this type of image. Temperature gradient analysis is the process of calculating and analyzing the rate of temperature change in the image. Temperature anomaly regions are areas whose temperature is significantly higher or lower than the surrounding background temperature. Debonded cold spots refer to low-temperature areas caused by abnormal heat conduction due to debonding defects. Center coordinates are the coordinates of the center point of the cold spot region. Coverage range is the size or area of ​​the cold spot region.

[0053] In some implementations, the infrared thermal image from the raw data is input into the infrared thermal image processing channel in the data analysis module. Then, temperature gradient analysis is performed on the input image in this channel to depict the temperature change trend of each pixel in the image. Next, based on the analysis results, temperature abnormal areas in the image are identified and delineated. Then, specific cold spots caused by degumming defects are further located in these abnormal areas. Finally, the center coordinates of the cold spot are calculated and its coverage area is determined.

[0054] It should be noted that the infrared thermal imaging processing channel is based on the principle that debonding defects can cause detectable temperature anomalies (cold spots) on the material surface. Therefore, the infrared thermal imaging processing channel only processes the temperature information in the image and does not involve additional operations such as ambient temperature compensation or emissivity correction of the thermal image.

[0055] Step 204: Input the transmittance value from the original data into the transmittance value processing channel in the data analysis module. Obtain the ambient temperature value through the temperature and humidity sensor. Based on the real-time ambient temperature value, preset the transmittance of the reference glass. Compare the transmittance value of the original data with the transmittance of the reference glass. If the difference exceeds 5%, the fogging level is determined to be high risk. If the difference is between 3% and 5%, it is medium risk. If the difference is less than 3%, it is low risk.

[0056] The transmittance data processing channel is a specific submodule dedicated to processing this type of data. The temperature and humidity sensor is a device used to measure ambient temperature and humidity. The detected ambient temperature value is the temperature reading of the surrounding environment during the test. The reference glass transmittance is a preset standard glass transmittance value at a specific ambient temperature. The fogging rating is a classification of the risk level of fogging on the glass surface, divided into three levels: high risk, medium risk, and low risk.

[0057] In some implementations, the transmittance value from the raw data is input into the transmittance value processing channel within the data analysis module. Simultaneously, the current ambient temperature value is acquired in real time via a temperature and humidity sensor. Based on this real-time ambient temperature value, a corresponding reference glass transmittance standard value is preset. Then, the input raw transmittance value is compared with the preset reference glass transmittance to calculate the percentage difference between the two. Finally, the atomization level is determined based on the magnitude of this difference: if the difference exceeds 5%, it is considered high risk; if the difference is between 3% and 5%, it is considered medium risk; and if the difference is less than 3%, it is considered low risk.

[0058] Step 205: Input the color difference ΔE value from the original data into the color difference ΔE value processing channel in the data analysis module. Based on the Lab color space model, adjust the highest and lowest thresholds for color difference aging proportionally according to the service life of the glass. When the ΔE value > the highest threshold, the yellowing level is determined to be severe aging. When the lowest threshold < the ΔE value ≤ the highest threshold, it is determined to be moderate aging. When the ΔE value ≤ the lowest threshold, it is determined to be slight aging.

[0059] The color difference ΔE value processing channel is a specific submodule dedicated to processing this type of data. The Lab color space model is a color model based on human visual perception, where L represents lightness, and a and b represent color opposites. Glass service life refers to the number of years the glass has been in use. Color difference aging is the color change phenomenon that occurs in glass over time. The highest and lowest thresholds are pre-set critical values ​​used to judge the degree of aging. Yellowing level is a classification of the degree of yellowing aging of glass, divided into three levels: severe aging, moderate aging, and slight aging.

[0060] In some implementations, the color difference ΔE value in the raw data is input into the color difference ΔE value processing channel in the data analysis module. Then, the data is standardized based on the Lab color space model. Next, according to the specific service life of the glass, the highest and lowest thresholds for judging the degree of aging are dynamically adjusted in a proportional relationship. Then, the input ΔE value is compared with the adjusted threshold range: when the ΔE value is greater than the highest threshold, the yellowing level is judged as severe aging; when the ΔE value is between the lowest and highest thresholds, it is judged as moderate aging; and when the ΔE value is less than or equal to the lowest threshold, it is judged as slight aging.

[0061] Step 206: Encode the 5-channel analysis structure in the data analysis module into a structured vector, denoted as the defect feature vector output. The structure of the defect feature vector is [bubble area, degumming outline, defect depth, defect size, cold spot location, fogging level, yellowing level].

[0062] A structured vector is a data structure formed by arranging multiple data elements in a specific order. A defect feature vector is a structured vector used to comprehensively describe defect features. Its specific structure includes seven dimensions: bubble area, degumming contour, defect depth, defect size, cold spot location, fogging level, and yellowing level.

[0063] It should be noted that the structural order of the defect feature vector is fixed, and the data at each position depends entirely on the output of the corresponding processing channel.

[0064] Step 207: Confidence-weighted fusion formula Calculate the final parameter ZL, where w i Let D be the confidence weight of the i-th detection channel. i is the original defect parameter value output by the i-th detection channel, and n is the number of effective detection channels participating in the fusion.

[0065] The confidence-weighted fusion formula is a mathematical expression that calculates confidence by assigning confidence weights to different data sources and then weighting them.

[0066] In some implementations, based on the various parameters contained in the defect feature vectors obtained from the previous processing channels, a preset confidence weight value is assigned to each parameter. Then, each parameter is multiplied by its corresponding weight value according to the confidence weighted fusion formula. Finally, all weighted values ​​are summed, and the comprehensive final parameter ZL is obtained by calculating the weighted sum.

[0067] The data analysis module performs parallel 5-channel synchronous analysis and processing of optical interferometric images, ultrasonic echo signals, infrared thermograms, transmittance values, and color difference ΔE values. This also includes: When the optical interference channel detects a bubble area >5mm 2 However, if the ultrasonic channel does not identify the enhanced sound wave reflection signal at the corresponding coordinates, the infrared channel is activated to re-inspect the area, and the coverage area of ​​the infrared cold spot is used as the final bubble area.

[0068] If the transmittance channel determines that there is a high risk of fogging, but the color difference ΔE value is ≤1, then the optical interference channel is triggered to check for contamination on the glass surface.

[0069] When the standard deviation of the confidence-weighted result is greater than 30%, the historical case library is called to match data under the same working conditions, and the case parameters with a similarity of greater than 85% are selected as the arbitration output.

[0070] Based on the above technical solution, a multi-sensor fusion structure consisting of five parallel processing channels (optical interference image processing channel, ultrasonic echo signal processing channel, infrared thermal image processing channel, transmittance value processing channel, and color difference ΔE value processing channel) achieves multi-dimensional collaborative detection and cross-verification of glass defects. This avoids the problem that a single detection method cannot simultaneously cover surface and internal defects, and that each detection result exists independently and cannot form mutual verification. Simultaneously, in conjunction with inter-channel trigger re-inspection rules (such as triggering infrared re-inspection when optical and ultrasonic detection conflict, and triggering optical verification when transmittance and color difference detection contradict each other), a closed-loop detection effect of multi-source data mutual verification is achieved. Furthermore, through a dynamic preset of reference transmittance based on ambient temperature and a proportional adjustment mechanism of color difference threshold based on service life, the detection standard achieves adaptive optimization with environmental conditions and material state. Compared to a fixed threshold judgment system, this improves the adaptability to variables such as environmental temperature and humidity and material aging, avoiding the disconnect between the detection standard and actual working conditions.

[0071] In one possible implementation of this application embodiment, the correlation matrix between defects and levels can be constructed through the following steps 301 to 303, which are described in detail below: Step 301: Construct a spatial coordinate system based on the edge sealing layer, the intermediate buffer layer, the fiber reinforcement layer, and the matrix glass layer.

[0072] A spatial coordinate system is constructed based on an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a matrix glass layer, including: Based on the coating trajectory of the polymer sealant in the edge sealing layer, the edge of the glass annular surface is scanned by a laser profilometer to generate three-dimensional point cloud data of the sealing layer, establishing a first sub-coordinate system. The radial coordinates of the first sub-coordinate system... axial coordinates The origin of the coordinate system is the center line of the sealant coating, where R is the glass radius and T is the center line of the sealant coating. seal The edge sealing layer thickness is 0.5–1.0 mm.

[0073] Based on the viscoelastic parameters of the PVB or SGP interlayer material in the intermediate buffer layer, a thickness gradient of 1.0–1.8 mm is transformed into a non-uniform Z-axis mesh mapped onto the first sub-coordinate system to form a spatial coordinate system, with the Z-axis range being [missing information]. The planar grid resolution is 1mm × 1mm, where T mid The thickness of the intermediate buffer layer is 1.0–1.8 mm.

[0074] Based on the intersections of carbon fiber or glass fiber meshes in the fiber reinforcement layer, the B-spline surface governing equations are used. The actual coordinates of point S(u,v) in three-dimensional space are mapped to the coordinates of the mesh intersection of the fiber reinforcement layer, where P i,j N represents the coordinates of the intersection points of the carbon fiber or glass fiber mesh. i,p (u) is a p-th degree B-spline basis function, N j,q (v) is a B-spline basis function of degree q along the direction of parameter v, and (u,v) are the coordinates in the parameter domain.

[0075] Set a constant Z-axis range for the substrate glass layer Furthermore, the planar boundary coincides with the glass cutting line, used to bind the rigid boundary, where T glass This refers to the thickness of the tempered glass substrate.

[0076] It should be noted that a 3D point cloud is generated based on the scanning of the sealant coating trajectory by a laser profilometer, and radial constraints are established. and axial constraints The adaptive sub-coordinate system not only solves the coordinate drift problem caused by uneven thickness of the sealing layer at the edge of curved glass, but also addresses the positioning distortion caused by the thickness fluctuation (0.5–1.0 mm) of the sealant at the edge of irregularly shaped glass. This ensures that the origin of the coordinate system always aligns with the actual coating centerline, providing a geometric reference for defect mapping. Simultaneously, the viscoelastic parameters (PVB, SGP) are transformed into a non-uniform Z-axis mesh (1 mm × 1 mm resolution), ensuring the coordinate system remains within a certain range. The dynamic range within the interval reflects material gradient changes, enabling precise spatial characterization of energy absorption paths. This allows the coordinate grid density to match the material property distribution (grid refinement in high-stress areas), supporting the quantified execution of stress transfer rules with an impact force attenuation rate ≥30% in space. Furthermore, by mapping the fiber mesh intersection coordinates to three-dimensional space using B-spline control equations, the surface topology of 60-80 mesh carbon and glass fiber meshes (such as wavy weaves) can be accurately expressed. This transforms abstract mesh parameters into calculable physical locations, suppressing spatial modeling errors in crack propagation paths and overcoming the limitation of traditional planar projections in representing changes in fiber weave angles. Finally, a constant Z-axis interval is set. Furthermore, by aligning the plane with the glass cutting line and applying forced constraints to the boundary conditions of the finite element analysis, distortion of the boundary conditions in the residual stress analysis can be avoided.

[0077] Step 302: Map the bubble contour coordinates, degummed area contours, and cold spot location coordinates in the defect feature vector to the hierarchical coordinate system through affine transformation, and output a defect coordinate set with hierarchical labels.

[0078] Affine transformation refers to the geometric mapping between coordinate systems achieved through linear transformations (translation, rotation, scaling). A hierarchical coordinate system refers to a three-dimensional spatial reference system constructed based on the physical boundaries of the edge sealing layer, intermediate buffer layer, fiber reinforcement layer, and matrix glass layer. A defect coordinate set with hierarchical identifiers refers to a set of defect location data that, after mapping, is labeled with its specific hierarchical affiliation (e.g., "intermediate buffer layer").

[0079] In some implementations, the polygon vertex coordinates of the bubble contour, the boundary point sequence of the degummed area, and the center coordinates and radius range of the cold spot are extracted from the defect feature vector. Subsequently, physical parameters based on the hierarchical coordinate system (such as the radial range of the edge sealing layer) are used. Z-axis interval of intermediate buffer layer Affine transformation matrix is ​​constructed, including the translation parameters of the coordinate system origin (e.g., offset of the sealing layer centerline), the scaling factors for each layer thickness (e.g., the thickness ratio of the matrix glass layer), and the plane rotation angle (e.g., the angle between the fiber mesh and the coordinate axes). Then, the original defect coordinates are multiplied by the affine transformation matrix to achieve coordinate mapping: for example, the bubble contour vertex (x...)... raw ,y raw） The coordinates are converted to hierarchical coordinates (x′, y′, z′), where the z-value is automatically assigned to the corresponding hierarchical level based on the defect depth and the Z-axis range (e.g., a depth of 1.2mm falls into the intermediate buffer layer, Z∈[0.8mm, 2.6mm]). Finally, a hierarchical label (e.g., "fiber reinforcement layer") is added to each mapped coordinate point, outputting a structured defect coordinate set with the data format (x, y, z, hierarchical), (x, y, z, hierarchical). It should be noted that if the defect spans multiple layers (such as a through crack), the layer classification should be marked according to coordinate segments. At the same time, during mapping, it is necessary to check whether the coordinates exceed the physical boundaries of the layers (such as the cutting line range of the substrate glass layer), and the excess part should be automatically corrected to the nearest boundary point.

[0080] Step 303: Calculate the convolution operation between the layer material properties and the defect features for each defect location to generate the correlation matrix between the defect and the layer.

[0081] Among them, the hierarchical material properties refer to the inherent physical parameters of each layer, including the elastic modulus of the polymer sealant in the edge sealing layer, the viscoelastic coefficient of the PVB / SGP in the intermediate buffer layer, the mesh density (60-80 mesh) of the fiber reinforcement layer, and the tempering strength of the matrix glass layer. Convolutional transport refers to the sliding window multiplication and summation operation on the material property matrix and defect feature values ​​in the spatial domain. The correlation matrix is ​​a two-dimensional data table generated by convolution calculation, where rows represent defect types, columns represent hierarchical parameters, and element values ​​represent the correlation strength between defect features and hierarchical properties.

[0082] In some implementations, the three-dimensional coordinates and layer labels of each defect location are first extracted from a set of defect coordinates with layer labels. Then, the material properties of the corresponding layer are retrieved: the edge sealing layer uses the sealant elastic modulus (e.g., 3.5 MPa), the intermediate buffer layer uses the PVB viscoelastic coefficient (e.g., 0.4 GPa·s), the fiber reinforcement layer uses carbon and glass fiber mesh density (e.g., 70 mesh), and the matrix glass layer uses tempering strength (e.g., 120 MPa), constructing a layered material property matrix. Next, parameters matching the location in the defect feature vector (e.g., bubble area 5 mm²) are used... 2 The input value is used as the convolution kernel. Then, a 3×3 grid (1mm×1mm resolution) is defined centered on the defect location, and convolution is performed in the corresponding region of the hierarchical material property matrix: the material property value of each cell in the grid is multiplied by the defect feature value and then summed (formula: M = D / M), where M is the material property value and D is the defect feature value. Finally, the convolution result is filled into the correlation matrix according to the defect type (bubble, delamination, cold spot) as rows and the hierarchical parameters (elastic modulus, viscoelasticity, grid density) as columns to form a structured output.

[0083] It should be noted that when the defect is located near the layer boundary, only the current layer mesh region is calculated (e.g., if the defect is at Z=1.7mm in the intermediate buffer layer, it does not cross into the fiber reinforcement layer). Furthermore, the values ​​of the correlation matrix elements characterize the sensitivity of the defect to the layer's function; for example, the higher the convolution value of the bubble area and the viscoelastic coefficient of the intermediate buffer layer, the greater the weakening of the energy absorption capacity by the defect.

[0084] Based on the above technical solution, point cloud data is generated by scanning the edge of the sealing layer using a laser profilometer. This data is then combined with the B-spline surface control equations to map the intersections of the fiber mesh, constructing a non-uniform mesh coordinate system. This allows for hierarchical spatial positioning of defects, ensuring a strict correspondence between the impact force transmission path (e.g., from the sealing layer to the buffer layer) and the defect location (e.g., an air bubble in the buffer layer at Z=1.2mm), guaranteeing directional energy dissipation of the multi-level stress buffering mechanism (with a preset impact attenuation rate of ≥30%). Simultaneously, the defect feature vector coordinates are processed using affine transformation matrices (including coordinate translation to the sealing layer centerline, pixel-to-physical-size scaling, and planar rotation correction), outputting a coordinate set with hierarchical labels. This enables real-time correlation between defects and mechanical responses, binding discrete defect locations detected by optical / infrared methods (e.g., cold spot center coordinates) to hierarchical material properties (e.g., the viscoelasticity of the buffer layer), providing a spatial reference for stress analysis. Finally, a 3×3 grid (1mm×1mm resolution) is defined at the defect location. The hierarchical attribute matrix (such as the viscoelastic value of the buffer layer) and the defect characteristic value (such as the bubble area) are multiplied and summed to generate an association matrix with the defect type as the row and the hierarchical parameter as the column. This allows for a quantitative assessment of defect risk and drives safety level decisions.

[0085] In a possible implementation of the embodiment of the present application, calculating the safety level of the output buffer layer through the buffer efficiency index can be achieved through the following steps 401 to step 403, which are specifically described below: Step 401: According to the defect and layer correlation matrix, extract the defect density factor δ and stress sensitivity coefficient ε corresponding to each layer.

[0086] Among them, the defect and layer correlation matrix is a matrix generated by mapping the coordinates in the defect feature vector to the layer coordinate system through affine transformation, and is used to quantify the correlation strength between defects and layer material properties; the defect density factor δ is a parameter extracted from the correlation matrix, representing the distribution density or concentration degree of defects in a specific layer; the stress sensitivity coefficient ε is a parameter extracted from the correlation matrix, representing the sensitivity of the layer material to stress changes and affecting the degree of defect response to stress.

[0087] Step 402: Based on the preset rules of the multi-layer stress buffer mechanism, obtain the impact force attenuation rate η of each layer.

[0088] Among them, the impact force attenuation rate is the percentage or ratio by which the impact force decreases when passing through each buffer layer, representing the buffer effect.

[0089] In some implementations, first retrieve the layer structure of the buffer system, and simulate or experiment with the application of impact force through algorithms or parameters in the preset rules, such as stress-strain models or energy absorption formulas; then measure the impact force values at the input and output points of each layer and calculate the attenuation rate, that is, obtain the ratio using formulas (output force, input force) or similar methods; finally, record and analyze the attenuation rate data of each layer to evaluate the overall performance of the buffer mechanism.

[0090] Step 403: Based on the defect density factor δ, stress sensitivity coefficient ε, and impact force attenuation rate η, calculate the overall safety value BEI through the buffer efficiency index formula Calculate the overall safety value BEI. Output the safety level according to the BEI value range. BEI≤0.25 is the allowable service safety level, 0.25<BEI≤0.5 is the warning observation safety level, and BEI>0.5 is the high-risk replacement safety level.

[0091] Among them, the buffer efficiency index BEI is calculated and is an index value used to comprehensively evaluate the safety performance of buffer materials or structures.

[0092] In some implementations, the defect density factor δ, stress sensitivity coefficient ε, and impact force attenuation rate η are used as input variables and substituted into the calculation formula of the buffer effectiveness index BEI for calculation, ultimately obtaining a specific BEI value. Finally, this calculation result is compared with a preset safety level threshold: if the BEI value is less than or equal to 0.25, the safety level is determined to be "allowed to serve"; if the BEI value is greater than 0.25 but less than or equal to 0.5, the safety level is determined to be "early warning observation"; if the BEI value is greater than 0.5, the safety level is determined to be "high-risk replacement", thus completing the complete technical process from parameter acquisition to safety level output.

[0093] Based on the above technical solution, the defect density factor, stress sensitivity coefficient, and impact force attenuation rate are used as input variables. Through specific functional relationships, a single index value representing the overall safety status is output. This integrates complex information from multiple dimensions into an easy-to-understand and operate indicator (BEI value). Then, through the preset BEI value range (0.25, 0.5), the three-level accurate classification of the safety status is achieved, enabling rapid assessment.

[0094] In one possible implementation of this application embodiment, after the security level of the output buffer layer, steps 501 to 504 are further included, which are described in detail below: Step 501: Obtain historical experimental impact waveform data and material constitutive models of multi-level stress buffering mechanisms that are associated with the safety level.

[0095] Among them, the material constitutive model is a mathematical model that describes the mechanical behavior of materials (such as stress-strain relationship), including elastic modulus, viscoelastic parameters, etc., and is used for finite element simulation analysis.

[0096] In some implementations, impact force waveform data associated with the current safety level are retrieved from historical experimental databases. This data usually comes from impact tests under similar working conditions, such as recording the waveform of impact force over time using sensors, including parameters such as peak force and duration. At the same time, constitutive models of each layer of materials in the multi-layer stress buffering mechanism are obtained, such as the elastic modulus of the polymer sealant in the edge sealing layer, the viscoelastic coefficient of the PVB or SGP material in the middle buffer layer, the carbon fiber mesh density of the fiber reinforcement layer, and the strength parameters of the tempered glass in the matrix glass layer.

[0097] Step 502: Based on the correlation matrix between defects and levels, the buffer efficiency index BEI, and the material constitutive model, reconstruct the multi-level buffer layer finite element model containing defects in ANSYS / LS-DYNA.

[0098] ANSYS / LS-DYNA is a finite element analysis software used for nonlinear dynamic event simulation. The multi-level buffer layer finite element model is a computer model representing a buffer system with multiple hierarchical structures, used to simulate its protective performance.

[0099] In some implementations, based on the correlation matrix between defects and levels, matrix operations are used to determine the distribution of defects (such as cracks or pores) in the buffer layer and their impact on each level. Then, the buffer effectiveness index (BEI) is used to calculate the buffer effectiveness of each level to evaluate the overall performance. Next, the material constitutive model is combined with the input stress-strain relationship parameters of the material to define the mechanical behavior of each layer. Finally, in ANSYS / LS-DYNA software, the correlation matrix, BEI, and constitutive model data are integrated to construct a multi-level buffer layer finite element model containing defects. Boundary conditions and loads are set for dynamic simulation to analyze the impact response and failure mechanism of the model.

[0100] Step 503: Input the historical experimental impact waveform as the boundary condition into the multi-level buffer layer finite element model, perform explicit dynamic simulation, dynamically output the stress distribution cloud map of each level and mark the stress concentration area.

[0101] Explicit dynamics simulation is a computational technique that uses explicit time integration algorithms to solve dynamic equations, suitable for simulating high-speed transient events. Stress distribution contour maps are graphical outputs that visualize the magnitude and variation of stress in the model through color gradients. Stress concentration regions are areas in the model where the local stress is significantly higher than the surrounding area, usually caused by abrupt changes in geometry or material inhomogeneity.

[0102] In some implementations, historical experimental impact waveform data is input as boundary conditions into a multi-layered buffer layer finite element model and applied to specific nodes or surfaces of the model. Then, a multi-layered buffer layer finite element model is constructed, including defining the material properties, geometric dimensions, and interlayer contact conditions of each layer to ensure the model accurately reflects the actual structure. Next, explicit dynamic simulation is performed, using explicit integration methods such as the central difference method to solve the dynamic equations and simulate the dynamic response process under impact loads. During the simulation, stress distribution data for each layer is calculated and output in real time, generating dynamic stress contour maps. These contour maps visually display the stress magnitude and trend through color changes. Finally, in the post-processing stage, stress concentration areas are automatically or manually identified, and potential high-stress points are highlighted based on stress thresholds or gradient analysis.

[0103] Step 504: Based on the comparison between the stress concentration area and the preset stress threshold, identify the weak layer and its stress over-limit ratio.

[0104] The preset stress threshold is the maximum allowable stress value set in advance based on material strength or safety standards. A weak layer refers to a layer in a layered or multi-layered structure where stress is prone to exceed the allowable range. The stress exceedance ratio is the degree to which the actual stress value exceeds the preset stress threshold, usually expressed as a percentage or ratio.

[0105] In some implementations, stress distribution data in the structure is obtained through stress analysis techniques such as finite element analysis or experimental measurements to identify stress concentration areas. Then, the stress values ​​in these areas are compared one by one with preset stress thresholds to determine whether they exceed the limits. Based on the comparison results, it is then determined which levels are weak levels, i.e., those levels where stress continues or significantly exceeds the limits. Finally, the stress exceedance ratio of each weak level is calculated, and the degree of exceedance is quantified by dividing the exceedance stress value by the threshold and multiplying by 100%, thus completing the identification.

[0106] After identifying the weak levels, the following also includes: Extracting the peak stress σ in the stress concentration region max Using the location coordinates, calculate the actual impact force attenuation rate for each level.

[0107] If the actual impact force attenuation rate is <30% or the stress exceedance ratio is >5%, a targeted set of optimization suggestions for layer parameters is generated based on the identification and quantitative assessment results of the weak layer. The set of optimization suggestions includes: if the edge sealing layer is a weak layer, it is recommended to increase the coating thickness of its polymer sealant or use a sealing material with a higher elastic modulus; if the intermediate buffer layer is a weak layer, it is recommended to adjust the thickness of its PVB or SGP interlayer or optimize the viscoelastic constitutive model parameters; if the fiber reinforcement layer is a weak layer, it is recommended to increase the density of the carbon fiber or glass fiber mesh or increase the diameter of a single fiber; if the base glass layer is a weak layer, it is recommended to increase the thickness of its tempered glass base layer.

[0108] Based on the above technical solution, by creating a correlation matrix between defects and levels, "defect types" (such as bubbles, cracks, and debonding) are linked to "material levels" (such as sealing layers, buffer layers, reinforcement layers, and matrix layers), encompassing quantitative indicators such as defect size, density, and location. Simultaneously, combined with a material constitutive model, a "defect-containing" finite element model is reconstructed in ANSYS / LS-DYNA, thereby digitizing and quantifying unavoidable material defects and manufacturing flaws, significantly improving the reliability and prediction accuracy of simulation results. Furthermore, by using historical experimental impact waveform data as boundary condition input for the simulation, it automatically runs in explicit dynamic simulation, outputting stress cloud maps and automatically identifying stress concentration areas. The stress values ​​are automatically compared with preset stress thresholds, and the "stress over-limit ratio" and "actual impact force attenuation rate" are calculated, forming an automated, data-driven analysis loop. This greatly reduces manual intervention and subjective judgment, making the entire evaluation process repeatable, efficient, and objective. It also maximizes the value of experimental data, directly driving and validating new design iterations.

[0109] The above primarily describes the solutions of the embodiments of this application from the perspective of device implementation. It is understood that each device, for example, a multi-level buffer lamination device for explosion-proof window glass, includes at least one of the hardware structures and software modules corresponding to each function in order to achieve the above-mentioned functions. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0110] This application embodiment can divide the multi-level buffer lamination device for explosion-proof window glass into functional units based on the above method example. For example, each function can be divided into its own functional unit, or two or more functions can be integrated into the same processing unit. The integrated unit can be implemented in hardware or software. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0111] When using integrated units, Figure 6A possible structural schematic diagram of a multi-level buffer lamination device for explosion-proof window and door glass (referred to as a multi-level buffer lamination device 60 for explosion-proof window and door glass) involved in the above embodiments is shown. The multi-level buffer lamination device 60 for explosion-proof window and door glass includes a processing unit 601 and a communication unit 602, and may also include a storage unit 603. Figure 6 The structural diagram shown can be used to illustrate the structure of the multi-level buffer lamination device for explosion-proof window and door glass involved in the above embodiments.

[0112] when Figure 6 The schematic diagram shown illustrates the structure of the multi-level buffer lamination device for explosion-proof window and door glass involved in the above embodiments. The processing unit 601 is used to control and manage the operation of the multi-level buffer lamination device for explosion-proof window and door glass. The communication unit 602 is used for the multi-level buffer lamination device for explosion-proof window and door glass to communicate with other devices. The storage unit 603 is used to store the program code and data of the multi-level buffer lamination device for explosion-proof window and door glass.

[0113] For example, the communication unit 602 is used to acquire the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value of the buffer layer of the window glass to form raw data; the processing unit 601 is used to input the raw data into the data analysis module, and to analyze and process the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value through parallel 6 channels simultaneously, to extract the bubble area, the outline of the delamination area, the defect depth and size, the location of the cold spot in the delamination area, the fogging level, and the yellowing level inside the buffer layer of the window glass, and to obtain the defect feature vector and the final defect parameters. The multi-level buffer layer of the explosion-proof window glass is divided from the outside to the inside into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a base glass layer, forming a multi-level stress buffer mechanism, and a preset stress transmission rule that the impact force is transmitted to the next layer according to the layer attenuation rate ≥30%. The parameters of each level in the multi-level stress buffer mechanism are obtained and synchronously input into the evaluation module along with the defect feature vector. The correlation matrix between defects and levels is constructed, and the safety level of the buffer layer is calculated and output through the buffer effectiveness index.

[0114] In one possible implementation, the processing unit 601 is further configured to simultaneously analyze and process the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value through six parallel channels of the data analysis module. This process extracts the bubble area, the outline of the delamination area, the defect depth and size, the location of the cold spot in the delamination area, the fogging level, and the yellowing level from the buffer layer of the window glass. The result is a defect feature vector and final defect parameters. This includes: inputting the optical interference image from the original data into the optical interference image processing channel within the data analysis module; performing fringe morphology analysis on the optical interference image; identifying the distorted areas of the interference fringes; and calculating the bubble area and the outline coordinates of the delamination area based on the fringe spacing variation. Additionally, the processing unit 601 inputs the ultrasonic echo signal from the original data into the ultrasonic echo signal processing channel within the data analysis module; extracting the time-frequency features of the ultrasonic echo signal; calculating the defect depth through the time difference between the reflected and emitted waves; and quantifying the defect size through the signal attenuation amplitude. The infrared thermal image from the original data is input into the infrared thermal image processing channel within the data analysis module. Temperature gradient analysis is performed on the infrared thermal image to identify abnormal temperature areas and locate the center coordinates and coverage area of ​​cold spots in the delamination zone. The transmittance value from the original data is input into the transmittance value processing channel within the data analysis module. The ambient temperature value is obtained through a temperature and humidity sensor. A reference glass transmittance is preset based on the real-time ambient temperature value. The transmittance value of the original data is compared with the transmittance of the reference glass. If the difference exceeds 6%, the fogging level is determined to be high-risk; if the difference is between 3% and 6%, it is medium-risk; and if the difference is less than 3%, it is low-risk. The color difference ΔE value from the original data is input into the color difference ΔE value processing channel within the data analysis module. Based on the Lab color space model, the highest and lowest thresholds for color difference aging are adjusted proportionally according to the service life of the glass. When ΔE value > the highest threshold, the yellowing level is determined to be severe aging; when the lowest threshold < ΔE value ≤ the highest threshold, it is determined to be moderate aging; and when ΔE value ≤ the lowest threshold, it is determined to be slight aging. The 6-channel analysis structure in the data analysis module is encoded into a structured vector, denoted as the defect feature vector output. The structure of the defect feature vector is [bubble area, degumming contour, defect depth, defect size, cold spot location, fogging level, yellowing level]. This is based on a confidence-weighted fusion formula. Calculate the final parameter ZL, where w i Let D be the confidence weight of the i-th detection channel. i is the original defect parameter value output by the i-th detection channel, and n is the number of effective detection channels participating in the fusion.

[0115] In one possible implementation, the processing unit 601 is further used for the parallel 6-channel synchronous analysis and processing of optical interference images, ultrasonic echo signals, infrared thermal images, transmittance values, and color difference ΔE values ​​by the data analysis module. This also includes: when the optical interference channel detects a bubble area > 6 mm... 2However, if the ultrasonic channel fails to identify the enhanced acoustic reflection signal at the corresponding coordinates, the infrared channel is activated to re-inspect the area, using the infrared cold spot coverage as the final bubble area. If the transmittance channel determines a high risk of fogging, but the color difference ΔE value is ≤1, the optical interference channel is triggered to check for contamination on the glass surface. When the standard deviation of the confidence-weighted result is >30%, the historical case library is called to match data under the same working conditions, and the case parameters with a similarity >86% are selected as the arbitration output.

[0116] In one possible implementation, the processing unit 601 is further used to divide the multi-layered buffer layer of the explosion-proof window glass into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a base glass layer from the outside in. The edge sealing layer is composed of a polymer sealant with a thickness of 0.6–1.0 mm, covering the glass edge area to absorb shear stress and prevent water vapor penetration. The intermediate buffer layer is composed of a polyvinyl butyral (PVB) or ionic polymer solid gluconate (SGP) interlayer material with a thickness of 1.0–1.8 mm, located inside the sealing layer, to disperse impact energy through viscoelastic deformation. The fiber reinforcement layer is composed of a mesh woven from a mixture of carbon fiber and glass fiber with a mesh density of 60–80 mesh, embedded between the intermediate buffer layer and the base glass layer to inhibit crack propagation and improve interlayer peel strength. The base glass layer is a tempered glass substrate with a thickness ≥6 mm, used to withstand residual stress after impact and maintain the overall structural integrity. The layers are composited using a hot-pressing process, and the elastic modulus gradient attenuation rate between adjacent layers is ≥30%.

[0117] In one possible implementation, the processing unit 601 is further configured to construct a defect-layer correlation matrix, including: constructing a spatial coordinate system based on the edge sealing layer, intermediate buffer layer, fiber reinforcement layer, and matrix glass layer; mapping the bubble contour coordinates, debonded area contours, and cold spot location coordinates in the defect feature vector to the layer coordinate system through affine transformation, and outputting a defect coordinate set with layer identifiers; and calculating the convolution operation between the layer material properties and defect features for each defect location to generate the defect-layer correlation matrix.

[0118] In one possible implementation, the processing unit 601 is further configured to construct a spatial coordinate system based on the edge sealing layer, the intermediate buffer layer, the fiber reinforcement layer, and the matrix glass layer, including: based on the coating trajectory of the polymer sealant of the edge sealing layer, scanning the edge of the glass annular surface using a laser profilometer to generate three-dimensional point cloud data of the sealing layer to establish a first sub-coordinate system, and the radial coordinates of the first sub-coordinate system. axial coordinates The origin of the coordinate system is the center line of the sealant coating, where R is the glass radius and T is the center line of the sealant coating. sealThe thickness of the edge sealing layer is 0.6–1.0 mm. Based on the viscoelastic parameters of the PVB or SGP interlayer material of the intermediate buffer layer, a thickness gradient of 1.0–1.8 mm is transformed into a non-uniform grid on the Z-axis and mapped to the first subcoordinate system to form a spatial coordinate system, and the range of the Z-axis is The plane grid resolution is 1 mm×1 mm, where T mid is the thickness of the intermediate buffer layer of 1.0–1.8 mm. Based on the intersection points of carbon fiber or glass fiber grids in the fiber reinforced layer, through the B-spline surface control equation The actual coordinate points S(u,v) in the three-dimensional space are obtained and mapped to the grid intersection coordinates of the fiber reinforced layer, where P i,j are the intersection coordinates of the carbon fiber or glass fiber grids, and N i,p (u) is the p-th B-spline basis function, and N j,q (v) is the q-th B-spline basis function along the parameter v direction, and (u,v) are the parameter domain coordinates. A constant interval on the Z-axis is set for the matrix glass layer , and the plane boundary coincides with the glass cutting line for binding the rigid boundary, where T glass is the thickness of the tempered glass base layer.

[0119] In a possible implementation, the processing unit 601 is further configured to calculate and output the safety level of the buffer layer through the buffer effectiveness index, including: extracting the defect density factor δ and stress sensitivity coefficient ε corresponding to each level according to the association matrix of defects and levels. Based on the preset rules of the multi-level stress buffer mechanism, obtain the impact force attenuation rate η of each level. Based on the defect density factor δ, stress sensitivity coefficient ε and impact force attenuation rate η, through the buffer effectiveness index formula Calculate the overall safety value BEI. Output the safety level according to the BEI value range. BEI≤0.26 is the allowable service safety level, 0.26<BEI≤0.6 is the warning observation safety level, and BEI>0.6 is the high-risk replacement safety level.

[0120] In a possible implementation, after the processing unit 601 outputs the safety level of the buffer layer, it further includes: obtaining the historical experimental impact force waveform data associated with the safety level and the material constitutive model of the multi-level stress buffer mechanism. Based on the association matrix of defects and levels, buffer effectiveness index BEI and material constitutive model, reconstruct the finite element model of the multi-level buffer layer with defects in ANSYS / LS-DYNA. Input the historical experimental impact force waveform as the boundary condition into the finite element model of the multi-level buffer layer, perform explicit dynamic simulation, and dynamically output the stress distribution nephogram of each level and identify the stress concentration area. Compare the stress concentration area with the preset stress threshold to identify the weak level and its stress overrun ratio.

[0121] The processing unit 601 can be a processor or a controller, and the communication unit 602 can be a communication interface, transceiver, transceiver circuit, transceiver device, etc. The term "communication interface" is a general term and may include one or more interfaces. The storage unit 603 can be a memory. When the multi-level buffer lamination device 60 for explosion-proof window and door glass is a chip, the processing unit 601 can be a processor or a controller, and the communication unit 602 can be an input interface and / or an output interface, pins, or circuits, etc. The storage unit 603 can be a storage unit within the chip (e.g., a register, cache, etc.) or a storage unit located outside the chip (e.g., read-only memory (ROM), random access memory (RAM, etc.)).

[0122] The communication unit can also be called a transceiver unit. The antenna and control circuit with transceiver functions in the multi-level buffer lamination device 60 for explosion-proof window glass can be considered as the communication unit 602 of the multi-level buffer lamination device 60 for explosion-proof window glass, and the processor with processing functions can be considered as the processing unit 601 of the multi-level buffer lamination device 60 for explosion-proof window glass. Optionally, the device in the communication unit 602 used to implement the receiving function can be considered as the communication unit, which is used to execute the receiving steps in the embodiments of this application. The communication unit can be a receiver, a receiver circuit, etc. The device in the communication unit 602 used to implement the transmitting function can be considered as the transmitting unit, which is used to execute the transmitting steps in the embodiments of this application. The transmitting unit can be a transmitter, a transmitter, a transmitting circuit, etc.

[0123] Figure 6 If the integrated units in the process are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. Storage media for storing computer software products include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0124] The processor in this application may include, but is not limited to, at least one of the following: a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a microcontroller unit (MCU), or an artificial intelligence processor, etc., and other computing devices that run software. Each computing device may include one or more cores for executing software instructions to perform calculations or processing. The processor may be a standalone semiconductor chip or integrated with other circuits into a single semiconductor chip. For example, it may be integrated with other circuits (such as encoding / decoding circuits, hardware acceleration circuits, or various bus and interface circuits) to form a System-on-a-Chip (SoC), or it may be integrated as a built-in processor within an ASIC. The ASIC with the integrated processor may be packaged separately or together with other circuits. In addition to the cores for executing software instructions to perform calculations or processing, the processor may further include necessary hardware accelerators, such as field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), or logic circuits that implement dedicated logic operations.

[0125] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely illustrative descriptions of the application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and variations.

Claims

1. A multi-level buffer lamination method for explosion-proof window and door glass, characterized in that, include: The optical interference image, ultrasonic echo signal, infrared thermal image, transmittance value and color difference ΔE value of the buffer layer of the door and window glass are acquired to form the original data; The original data is input into the data analysis module, and the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value and color difference ΔE value are analyzed and processed simultaneously through 5 parallel channels of the data analysis module. The bubble area, the outline of the delamination area, the defect depth and size, the location of the cold spot in the delamination area, the fogging level and the yellowing level are extracted from the buffer layer of the door and window glass to obtain the defect feature vector and the final defect parameters. The multi-level buffer layer of the explosion-proof door and window glass is divided from the outside to the inside into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer and a base glass layer, forming a multi-level stress buffer mechanism, and a pre-set stress transmission rule that the impact force is transmitted to the next layer according to the layer attenuation rate ≥30%. The parameters of each level in the multi-level stress buffering mechanism are obtained and synchronously input into the evaluation module along with the defect feature vector. A correlation matrix between defects and levels is constructed, and the safety level of the buffer layer is calculated and output through the buffer effectiveness index.

2. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 1, characterized in that, The data analysis module performs parallel 5-channel synchronous analysis and processing of the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value to extract the bubble area, delamination area contour, defect depth and size, cold spot location, haze level, and yellowing level within the window glass buffer layer, thus obtaining the defect feature vector and final defect parameters, including: The optical interference image in the original data is input into the optical interference image processing channel in the data analysis module. The fringe morphology of the optical interference image is analyzed to identify the distorted areas of the interference fringes. The bubble area and the contour coordinates of the degummed area are calculated based on the change in the fringe spacing. The ultrasonic echo signal in the original data is input into the ultrasonic echo signal processing channel in the data analysis module. The time-frequency features of the ultrasonic echo signal are extracted, the defect depth is calculated by the time difference between the reflected wave and the emitted wave, and the defect size is quantified by the signal attenuation amplitude. The infrared thermal image from the original data is input into the infrared thermal image processing channel in the data analysis module. Temperature gradient analysis is performed on the infrared thermal image to identify abnormal temperature areas and locate the center coordinates and coverage of cold spots in the degummed area. The transmittance value in the original data is input into the transmittance value processing channel in the data analysis module. The ambient temperature value is obtained through the temperature and humidity sensor. The transmittance of the reference glass is preset based on the real-time ambient temperature value. The transmittance value of the original data is compared with the transmittance of the reference glass. If the difference exceeds 5%, the fogging level is determined to be high risk. If the difference is between 3% and 5%, it is medium risk. If the difference is less than 3%, it is low risk. The color difference ΔE value in the original data is input into the color difference ΔE value processing channel in the data analysis module. Based on the Lab color space model, the highest and lowest thresholds of color difference aging are adjusted proportionally according to the service life of the glass. When the ΔE value is greater than the highest threshold, the yellowing level is determined to be severe aging. When the lowest threshold is less than the ΔE value and less than the highest threshold, it is determined to be moderate aging. When the ΔE value is less than the lowest threshold, it is determined to be slight aging. The analysis structure of the 5 channels in the data analysis module is encoded into a structured vector, denoted as the defect feature vector output. The structure of the defect feature vector is [bubble area, degumming outline, defect depth, defect size, cold spot location, fogging level, yellowing level]; Based on the confidence-weighted fusion formula Calculate the final parameter ZL, where w i Let D be the confidence weight of the i-th detection channel. i is the original defect parameter value output by the i-th detection channel, and n is the number of effective detection channels participating in the fusion.

3. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 2, characterized in that, The data analysis module, which performs parallel 5-channel synchronous analysis and processing of the optical interference image, ultrasonic echo signal, infrared thermogram, transmittance value, and color difference ΔE value, also includes: When the optical interference channel detects a bubble area > 5 mm², but the ultrasonic channel does not identify the enhanced acoustic wave reflection signal at the corresponding coordinates, the infrared channel is activated to re-inspect the area, and the coverage area of ​​the infrared cold spot is used as the final bubble area. If the transmittance channel determines that there is a high risk of fogging, but the color difference ΔE value is ≤1, then the optical interference channel is triggered to check for contamination on the glass surface. When the standard deviation of the confidence weighted result is greater than 30%, the historical case library is called to match data under the same working conditions, and the case parameters with a similarity of greater than 85% are selected as the arbitration output.

4. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 3, characterized in that, The multi-layered buffer layer for explosion-proof door and window glass is divided from the outside to the inside into an edge sealing layer, a middle buffer layer, a fiber reinforcement layer, and a base glass layer, including: The edge sealing layer is made of polymer sealant with a thickness of 0.5–1.0 mm, covering the glass edge area to absorb shear stress and prevent water vapor penetration. The intermediate buffer layer is made of polyvinyl butyral (PVB) or ionic polymer SGP interlayer material, with a thickness of 1.0–1.8 mm, and is located inside the sealing layer to disperse impact energy through viscoelastic deformation. The fiber reinforcement layer is composed of a mesh woven from a mixture of carbon fiber and glass fiber, with a mesh density of 60–80 mesh. It is embedded between the intermediate buffer layer and the matrix glass layer to suppress crack propagation and improve interlayer peel strength. The base glass layer is a tempered glass substrate with a thickness of ≥5mm, which is used to withstand residual stress after impact and maintain the integrity of the overall structure. The layers are composited through a hot-pressing process, and the elastic modulus gradient attenuation rate of adjacent layers is ≥30%.

5. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 4, characterized in that, The correlation matrix between the constructed defects and the hierarchy includes: A spatial coordinate system is constructed based on an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a matrix glass layer; The bubble contour coordinates, degummed area contours, and cold spot location coordinates in the defect feature vector are mapped to a hierarchical coordinate system through an affine transformation, and a defect coordinate set with hierarchical labels is output. For each defect location, a convolution operation is performed between the layer material properties and the defect features to generate the correlation matrix between the defect and the layer.

6. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 5, characterized in that, The spatial coordinate system constructed based on the edge sealing layer, intermediate buffer layer, fiber reinforcement layer, and matrix glass layer includes: Based on the coating trajectory of the polymer sealant in the edge sealing layer, a laser profilometer is used to scan the edge of the glass annular surface to generate three-dimensional point cloud data of the sealing layer, establishing a first sub-coordinate system. The radial coordinates of the first sub-coordinate system... axial coordinates The origin of the coordinate system is the center line of the sealant coating, where R is the glass radius and T is the center line of the sealant coating. seal The edge sealing layer thickness is 0.5–1.0 mm; Based on the viscoelastic parameters of the PVB or SGP interlayer material in the intermediate buffer layer, the thickness gradient of 1.0–1.8 mm is transformed into a non-uniform Z-axis mesh and mapped onto the first sub-coordinate system to form a spatial coordinate system, with the Z-axis range being [missing information]. The planar grid resolution is 1mm × 1mm, where T mid The intermediate buffer layer has a thickness of 1.0–1.8 mm; Based on the intersections of carbon fiber or glass fiber meshes in the fiber reinforcement layer, the B-spline surface governing equations are used. The actual coordinates of point S(u,v) in three-dimensional space are mapped to the coordinates of the mesh intersection of the fiber reinforcement layer, where P i,j N represents the coordinates of the intersection points of the carbon fiber or glass fiber mesh. i,p (u) is a p-th degree B-spline basis function, N j,q (v) is the q-th order B-spline basis function along the direction of parameter v, and (u,v) are the coordinates in the parameter domain; Set a constant Z-axis range [T] for the substrate glass layer total T glass ,T total ], and the planar boundary coincides with the glass cutting line, used to bind the rigid boundary, where T glass This refers to the thickness of the tempered glass substrate.

7. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 6, characterized in that, The calculation of the security level of the output buffer layer using the buffer performance index includes: Extract the defect density factor δ and stress sensitivity coefficient ε corresponding to each level according to the association matrix of the defects and levels; Obtain the impact force attenuation rate η of each level based on the preset rules of the multi-level stress buffer mechanism; Based on the defect density factor δ, stress sensitivity coefficient ε, and impact force attenuation rate η, the buffer effectiveness index formula is used. Calculate the overall safety value (BEI); Output the safety level according to the BEI value range. When BEI ≤ 0.25, it is the allowable service safety level; when 0.25 < BEI ≤ 0.5, it is the warning observation safety level; when BEI > 0.5, it is the high-risk replacement safety level.

8. The multi-level buffer lamination method for explosion-proof window and door glass according to claim 7, characterized in that, After outputting the safety level of the output buffer layer, it further includes: Obtain the historical experimental impact force waveform data associated with the safety level and the material constitutive model of the multi-level stress buffer mechanism; Based on the association matrix of the defects and levels, the buffer efficiency index BEI, and the material constitutive model, reconstruct the finite element model of the multi-level buffer layer with defects in ANSYS / LS-DYNA; Input the historical experimental impact force waveform as the boundary condition into the finite element model of the multi-level buffer layer, perform explicit dynamic simulation, and dynamically output the stress distribution nephogram of each level and mark the stress concentration area; Compare the stress concentration area with the preset stress threshold to identify the weak level and its stress overrun ratio.

9. A multi-level buffer lamination method for explosion-proof window and door glass according to claim 8, characterized in that, After identifying the weak level, it further includes: Extract the peak stress σmax and position coordinates of the stress concentration area, and calculate the actual impact force attenuation rate of each level; If the actual impact force attenuation rate < 30% or the stress overrun ratio > 5%, generate a targeted set of optimization suggestions for level parameters according to the identification of the weak level and the quantitative evaluation results. The set of optimization suggestions for level parameters includes that if the edge sealing layer is the weak level, it is recommended to increase the coating thickness of its polymer sealant or use a sealant material with a higher elastic modulus; if the intermediate buffer layer is the weak level, it is recommended to adjust the thickness of its PVB or SGP interlayer or optimize the viscoelastic constitutive model parameters; if the fiber reinforcement layer is the weak level, it is recommended to increase the density of the carbon fiber or glass fiber grid or increase the diameter of a single fiber; if the matrix glass layer is the weak level, it is recommended to increase the thickness of its tempered glass base layer.

10. A multi-level buffer lamination device for explosion-proof door and window glass, characterized in that, The device includes: a communication unit and a processing unit; The communication unit is used to obtain the optical interference image, ultrasonic echo signal, infrared thermal image, light transmittance value, and color difference ΔE value of the door and window glass buffer layer to form original data; The processing unit is used to input the original data into the data analysis module, and analyze and process the optical interference image, ultrasonic echo signal, infrared thermal image, light transmittance value, and color difference ΔE value through parallel 5 channels synchronously, and extract the bubble area inside the door and window glass buffer layer, the contour of the delamination area, the defect depth and size, the position of the cold spot in the delamination area, the atomization level, and the yellowing level, and obtain the defect feature vector and the final defect parameters; Divide the multi-level buffer layer for explosion protection of the door and window glass from the outside to the inside into an edge sealing layer, an intermediate buffer layer, a fiber reinforcement layer, and a matrix glass layer to form a multi-level stress buffer mechanism, and preset the stress transfer rule that the impact force decays by a rate ≥ 30% to the next level; The parameters of each level in the multi-level stress buffering mechanism are obtained and synchronously input into the evaluation module along with the defect feature vector. A correlation matrix between defects and levels is constructed, and the safety level of the buffer layer is calculated and output through the buffer effectiveness index.

Images