A Detection Method for the Thickness Uniformity of the P Layer of EPE Film Based on Particle Swarm Optimization

Abstract

The present invention discloses a method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization, specifically relating to the field of EPE film detection, which includes building an optical interference detection system. The optical interference detection system includes a laser light source, a beam splitter, an interference detector, and a data acquisition module; placing an EPE film sample in the measurement area of the optical interference detection system, and then adjusting the optical path parameters of the laser light source and the beam splitter. The interference images at different positions are captured by the interference detector. The interference images include fringe displacement, fringe spacing, and fringe intensity distribution, and the interference images are used to reflect the change information of the thickness of the P layer of the EPE film. Through optical interference detection combined with the particle swarm optimization algorithm, a thickness distribution model of the P layer of the film is established. The model synthesizes the information of fringe displacement, fringe spacing, and fringe intensity distribution, relatively improving the comprehensiveness of thickness uniformity detection, and thus being applicable to the detection requirements of thickness changes in reality.

Description

Technical Field

[0001] The present invention relates to the technical field of EPE film detection. More specifically, the present invention relates to a method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization. Background Art

[0002] The EPE film is a functional film based on polyethylene foamed material, with characteristics such as light weight, buffering, shock resistance, and high light transmittance, and is widely used in the fields of optical components and photovoltaic encapsulation; the P layer of the EPE film is a functional layer in its structure, usually used to provide optical properties, mechanical strength, or protection; the uniformity of the P layer thickness has a relative impact on the overall performance of the film;

[0003] The existing technology has limitations in the measurement method of the P layer thickness of the EPE film, mainly relying on a single measurement means, such as contact measurement or optical scanning, and cannot capture the thickness change characteristics of the sample; at the same time, there is a lack of a mathematical model for the thickness distribution, making it difficult to comprehensively process features such as fringe displacement, fringe spacing, and intensity distribution, resulting in limitations in sample detection. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization. By combining optical interference detection with a particle swarm optimization algorithm, a thickness distribution model of the P layer of the film is established. The model synthesizes the information of fringe displacement, fringe spacing, and fringe intensity distribution to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solution: A method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization, including:

[0006] First, an optical interference detection system is built. The optical interference detection system includes a laser light source, a beam splitter, an interference detector, and a data acquisition module; the EPE film sample is placed in the measurement area of the optical interference detection system. Immediately afterwards, the optical path parameters of the laser light source and the beam splitter are adjusted, and interference images at different positions are captured by the interference detector. The interference images include fringe displacement, fringe spacing, and fringe intensity distribution, and the interference images are used to reflect the thickness change information of the P layer of the EPE film; the data acquisition module is used to record the interference images captured by the interference detector;

[0007] The collected interference images are digitally processed to remove noise interference; immediately afterwards, the characteristic parameters of the fringes are extracted through an image processing algorithm; the image processing algorithm includes an edge detection algorithm;

[0008] According to the physical characteristics of optical interference, the characteristic parameters extracted from the interference image are correlated with the thickness distribution of the P layer of the adhesive film, and a thickness distribution model of the P layer of the adhesive film is established;

[0009] The parameters of the thickness distribution model of the P layer of the adhesive film are iteratively adjusted by the particle swarm optimization algorithm to optimize the fitting accuracy between the thickness distribution model of the P layer of the adhesive film and the actual interference image data, and the analysis and verification of thickness uniformity are completed.

[0010] In a preferred embodiment, the particle swarm optimization algorithm is initialized, the scale of the particle swarm is set, multiple initial solutions are randomly generated in the parameter space, and each particle corresponds to a thickness distribution scheme; at the same time, the velocity and position of the particle are initialized, and the fitness function is defined;

[0011] The particle swarm optimization algorithm is used for iterative calculation. In each iteration, the velocity and position of the particle are updated according to the rules of particle swarm optimization;

[0012] After each iteration, the fitness value of each particle is calculated. The fitness value is used to reflect the error between the prediction result of the thickness distribution model of the P layer of the adhesive film and the actual measured interference image data;

[0013] In each round of iteration, the solution of the current particle individual is recorded, and the global solution is continuously updated. When the fitness value is less than the preset threshold or the number of its iterations reaches the upper limit, the optimization is stopped, and the final thickness distribution model of the P layer of the adhesive film is output.

[0014] In a preferred embodiment, using the parameters of the optimized thickness distribution model of the P layer of the adhesive film, a thickness distribution map of the EPE adhesive film sample is generated, and the thickness change of different regions of the EPE adhesive film sample is shown through the thickness distribution map;

[0015] Based on the comparison between the thickness distribution map of the EPE adhesive film sample and the actual detection value, the fitting accuracy of the model is evaluated.

[0016] In a preferred embodiment, the characteristics of the interference image include fringe displacement, fringe spacing and fringe intensity distribution;

[0017] The fringe displacement is expressed as: the offset of the fringe center in the plane;

[0018] ;

[0019] where , are the coordinates of the current fringe center on the axis and the axis respectively; , are the coordinates of the reference fringe center on the axis and the Coordinates on the axis; , are the displacement changes of the fringes in two directions respectively;

[0020] The fringe spacing is expressed as: the distance between the centers of adjacent fringes;

[0021] ;

[0022] where is the Euclidean distance between the fringe centers;

[0023] The fringe intensity distribution is expressed as: the intensity value of each point in the interference image is determined by the optical path difference;

[0024] ;

[0025] ;

[0026] where represents the fringe intensity at point ; is the initial intensity of the light source; is the optical path difference between the reference beam and the measurement beam; is the wavelength of the laser light source; represents the refractive index of the EPE film sample at point ; represents the thickness of the EPE film sample at the point.

[0027] In a preferred embodiment, a thickness distribution model of the P layer of the film is constructed by comprehensively considering the fringe displacement, fringe spacing, and fringe intensity distribution; represents the thickness of the EPE film sample at point ;

[0028] ;

[0029] ;

[0030] where represents the normalized parameter of the fringe spacing; represents the Gaussian attenuation factor of the fringe intensity due to the change in fringe spacing; is the total number of fringes; is the th Euclidean distance between the fringe centers; is the reference fringe spacing.

[0031] In a preferred embodiment, the particle positions of the particle swarm optimization algorithm are initialized; in the parameter space of the thickness distribution model of the P layer of the film, randomly generate A particle, each particle representing a thickness distribution scheme; formulate is the particle at the thickness value of the position, and ;

[0032] ;

[0033] Among them is the thickness parameter set of the particle; is the total number of positions after discretization of the EPE film sample;

[0034] Initialize the particle velocity of the particle swarm optimization algorithm; formulate is the particle at the thickness value of the position, and ; The velocity of each particle is randomly initialized as:

[0035] ;

[0036] Among them is the velocity set of the particle;

[0037] Define the fitness function, and evaluate the fitting degree between the solution of the particle and the actual interference image data through the fitness function:

[0038] ;

[0039] Among them is the fitness value of the particle; is the intensity value of the actually measured interference image at the point ; represents the predicted intensity value of the interference image of the particle.

[0040] In a preferred embodiment, in each iteration, update the velocity according to the particle swarm optimization rule:

[0041] ;

[0042] Among them is the velocity vector of the particle at the th iteration; is the velocity vector of the particle at the th iteration; is the inertia weight; is the acceleration coefficient, Control the weights of the individual historical solutions and the global solution respectively; is a random factor, and its value range is [0, 1]; is the historical optimized solution of particle ; is the global optimized solution among all particles; is particle at the th iteration;

[0043] Update the position of the particle according to the velocity:

[0044] ;

[0045] where is the position of particle at the th iteration;

[0046] Establish an iteration stop condition; the iteration process of the particle swarm optimization algorithm terminates when condition one or condition two is satisfied;

[0047] Condition one is: is less than the preset threshold;

[0048] Condition two is: the number of iterations reaches the upper limit.

[0049] In a preferred embodiment, use the parameters in the optimized to generate the thickness distribution map of the final EPE film sample;

[0050] ;

[0051] where is the finally optimized thickness distribution value; the thickness distribution map of the EPE film sample is presented by corresponding to each position of the EPE film sample, visually showing the spatial variation of the thickness;

[0052] Compare the optimized thickness distribution value with the actually measured interference image data and calculate the final error:

[0053] ;

[0054] where is the final verification error; is the total number of sampling points; represents the predicted intensity based on the optimized thickness distribution.

[0055] The technical effects and advantages of the present invention:

[0056] 1. By combining optical interference detection with the particle swarm optimization algorithm, a thickness distribution model of the P layer of the adhesive film is established. The model synthesizes the information of fringe displacement, fringe spacing, and fringe intensity distribution, relatively enhancing the comprehensiveness of thickness uniformity detection, and thus being applicable to the detection requirements of thickness changes in reality;

[0057] 2. The particle swarm optimization algorithm converges to the optimal solution of the thickness distribution. The iterative process controls the search path through the update formulas of velocity and position, thereby reducing the time for parameter optimization and ensuring the global fitting accuracy of the model;

[0058] 3. By marking the optimized thickness distribution values, a thickness distribution map of the EPE adhesive film sample is generated, intuitively showing the thickness changes in different regions of the sample, providing a clear basis for quality control and process optimization. Description of the Drawings

[0059] Figure 1 This is the flowchart of the present invention. Detailed Embodiments

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0061] Referring to the attached Figure 1 description, a method for detecting the thickness uniformity of the P layer of the EPE adhesive film based on particle swarm optimization according to an embodiment of the present invention includes:

[0062] First, an optical interference detection system is built to form a detection environment for the thickness of the P layer of the EPE adhesive film. The optical interference detection system includes a laser light source, a beam splitter, an interference detector, and a data acquisition module; the EPE adhesive film sample is placed in the measurement area of the optical interference detection system. Immediately afterwards, the optical path parameters of the laser light source and the beam splitter are adjusted, and interference images at different positions are captured by the interference detector. The interference images include fringe displacement, fringe spacing, and fringe intensity distribution, and the interference images are used to reflect the thickness change information of the P layer of the EPE adhesive film; the data acquisition module is used to record the interference images captured by the interference detector and transmit them to the computing system after digital processing for subsequent analysis and processing;

[0063] Among them, the laser light source is used to provide a stable and highly monochromatic light source, that is, to generate the coherent light beam required for interference; the beam splitter is used to divide the laser light source into a reference beam and a measurement beam to ensure that the two beams of light can form interference on the sample surface and the reference surface; the interference detector is used to capture the interference image; in addition, adjusting the optical path parameters of the laser light source and the beam splitter includes but is not limited to: by adjusting the incident angle and output intensity of the laser light source, ensuring that the light beam enters the beam splitter at the optimal angle and matches the reflectivity of the sample surface, while avoiding too weak or saturated detector signals; secondly, optimizing the beam splitting ratio of the beam splitter to balance the energy distribution of the reference beam and the measurement beam and generate clear interference fringes; in addition, adjusting the light beam path length to keep the optical path difference within the coherence range and ensuring that the light beam is parallel through the collimating lens to improve coherence; accurately focusing on the sample surface to avoid distortion of the fringes due to offset or tilt. Through the adjustment of these optical path parameters, the quality and signal-to-noise ratio of the interference fringes can be improved, providing reliable data for subsequent thickness detection;

[0064] The collected interference image is digitally processed to remove noise interference; then, the characteristic parameters of the fringes are extracted through an image processing algorithm, and the characteristic parameters include fringe displacement, fringe spacing, and fringe intensity distribution; the image processing algorithm includes an edge detection algorithm;

[0065] According to the physical characteristics of optical interference, the characteristic parameters extracted from the interference image are correlated with the thickness distribution of the P layer of the adhesive film, and a thickness distribution model of the P layer of the adhesive film is established; the thickness distribution model of the P layer of the adhesive film is used to calculate the thickness change of the P layer of the EPE adhesive film at different positions;

[0066] The parameters of the thickness distribution model of the P layer of the adhesive film are iteratively adjusted through the particle swarm optimization algorithm to optimize the fitting accuracy between the thickness distribution model of the P layer of the adhesive film and the actual interference image data, and the analysis and verification of thickness uniformity are completed.

[0067] Regarding the above scheme, it should be specifically noted that the collected interference image is first digitally processed by the data acquisition module. The core of this step is to convert the analog signal into a discrete digital signal for convenient subsequent analysis and processing; due to the influence of ambient light interference, light source fluctuations, or electronic noise, the quality of the digitized image may deteriorate, so it is necessary to remove the noise; common methods include Gaussian filtering and median filtering; Gaussian filtering weakens the high-frequency noise in the image through smoothing processing while retaining the main edge features of the image, making the brightness transition of the fringes more uniform; median filtering is more suitable for removing random pulse noise and helps to eliminate isolated strong noise points; after these processing steps, the clarity of the fringes can be enhanced, providing high-quality input for subsequent feature extraction;

[0068] In the denoised interference image, the fringes are formed by the alternating bright and dark curves caused by the light intensity change. Its characteristic parameters include fringe displacement, fringe spacing, and fringe intensity distribution. The core of feature extraction is to identify and locate the fringe center through image processing algorithms. By using edge detection algorithms, the positions of gray value changes in the image are calculated to identify the edges of the fringes, and then the fringe center position is located through fitting techniques to calculate the displacement information of the fringes. The fringe spacing is the distance between adjacent fringe centers, and the fringe spacing distribution is obtained by calculating the Euclidean distance between adjacent fringes point by point. The fringe intensity distribution reflects the light intensity characteristics under interference conditions by statistically analyzing the light intensity change information within the fringe region and extracting the light intensity values of each fringe.

[0069] The edge detection algorithm is a part of feature extraction, whose function is to identify the edges of the fringes and further extract the fringe center position. During the processing, first, the interference image is smoothed to reduce the interference of noise on edge detection. Then, the gradient calculation method is used to detect the regions with drastic gray value changes in the image, and the redundant weak gradient information is filtered out through non-maximum suppression technology, only retaining the relatively significant edge features. Finally, by setting a threshold, the insignificant edge noise is filtered out, and the fringe edges are fitted to extract the center position of the fringes. This process ensures the accuracy of the calculation of fringe displacement and fringe spacing and provides an accurate basis for the regional division of the fringe intensity distribution statistics.

[0070] Initialize the particle swarm optimization algorithm, set the size of the particle swarm, and randomly generate multiple initial solutions in the parameter space. Each particle corresponds to a possible thickness distribution scheme. At the same time, initialize the velocity and position of the particles and define the fitness function, which is used to evaluate the quality of the solutions of each particle.

[0071] Use the particle swarm optimization algorithm for iterative calculation. In each iteration, update the velocity and position of the particles according to the rules of particle swarm optimization. The particles gradually approach the optimal solution in the search space. The update of the particles is affected by the historical best solution and the global best solution, enabling the particle swarm optimization algorithm to search for the optimal solution of the thickness distribution model globally.

[0072] After each iteration, calculate the fitness value of each particle. The fitness value is used to reflect the error between the predicted result of the thickness distribution model of the P layer of the adhesive film and the actual measured interference image data. The smaller the fitness value, the closer the solution of the particle is to the actual distribution.

[0073] In each round of iteration, record the optimal solution of the current particle individual and continuously update the global optimal solution. When the fitness value is less than the preset threshold or the number of iterations reaches the upper limit, stop the optimization and output the final thickness distribution model of the P layer of the adhesive film.

[0074] Using the parameters of the optimized thickness distribution model of the P layer of the adhesive film, generate the thickness distribution map of the EPE adhesive film sample, and display the thickness changes in different regions of the EPE adhesive film sample through the thickness distribution map;

[0075] Based on the comparison between the thickness distribution map of the EPE adhesive film sample and the actual measured values, evaluate the fitting accuracy of the model; in the actual application process, if the fitting accuracy does not meet the requirements, the parameters of the particle swarm optimization algorithm (such as the particle swarm size, the number of iterations, etc.) can be adjusted or the data acquisition and preprocessing methods can be improved. For different sample materials or process conditions, repeat the above process to verify the applicability and robustness of the algorithm, and ensure that this method can be widely applied to the detection of the thickness uniformity of the P layer of the EPE adhesive film.

[0076] The characteristics of the interference image include fringe displacement, fringe spacing, and fringe intensity distribution;

[0077] The fringe displacement is expressed as: the offset of the fringe center in the plane;

[0078] ;

[0079] where , are the coordinates of the current fringe center on the axis and axis respectively; , are the coordinates of the reference fringe center on the axis and axis respectively; , are the displacement changes of the fringe in two directions, used to reflect the local surface changes of the thickness distribution of the P layer of the EPE adhesive film;

[0080] The fringe spacing is expressed as: the distance between adjacent fringe centers;

[0081] ;

[0082] where is the Euclidean distance between the fringe centers, used to reflect the uniformity of the fringe distribution;

[0083] The fringe intensity distribution is expressed as: the intensity value of each point in the interference image is determined by the optical path difference;

[0084] ;

[0085] ;

[0086] where represents the fringe intensity at the point ; is the initial intensity of the light source; is the optical path difference between the reference beam and the measurement beam; is the wavelength of the laser light source; represents the EPE film sample at point the refractive index at; represents the EPE film sample at point the thickness at; additionally, in the coefficient in the formula is obtained by converting the wavelength of the laser light source to the angular frequency; the coefficient in the formula represents the double-pass path of the beam through the sample surface.

[0087] Construct a thickness distribution model of the P layer of the film by synthesizing fringe displacement, fringe spacing, and fringe intensity distribution; represents the EPE film sample at point the thickness at;

[0088] ;

[0089] ;

[0090] where is the normalized parameter representing the fringe spacing, used to control the influence range of the fringe spacing on the intensity attenuation; is the Gaussian attenuation factor of the fringe intensity due to the change in fringe spacing; is the total number of fringes; is the Euclidean distance between the centers of the th fringe; the reference fringe spacing is used to calibrate the standard value of the spacing;

[0091] Regarding the formation of the above formula, it should be noted that:

[0092] the effect of the fringe intensity distribution on the thickness; is determined by the cosine relationship of the optical path difference, and the basic formula is:

[0093] ;

[0094] can be inversely deduced through the inverse cosine function ;

[0095] the correction of the thickness by the fringe displacement; the fringe displacement , reflects the non-uniformity of the local area of the sample surface, and its change will cause the asymmetry of the thickness distribution. The displacement is calculated through its components as:

[0096] ;

[0097] Modulation of fringe spacing on intensity; as the fringe spacing changes, the intensity of the interference fringes gradually decays. Therefore, a Gaussian attenuation factor for the fringe intensity due to the change in fringe spacing is introduced into the intensity distribution formula.

[0098] ;

[0099] Finally, the fringe intensity distribution, displacement correction, and spacing modulation are combined to obtain the final thickness distribution model formula.

[0100] Initialize the particle positions of the particle swarm optimization algorithm; in the parameter space of the thickness distribution model of the P layer of the adhesive film, randomly generate particles, and each particle represents a possible thickness distribution scheme; designate as the thickness value of particle at the th position, representing the thickness parameter corresponding to the EPE adhesive film sample at the discrete point , and ;

[0101] ;

[0102] where is the set of thickness parameters of the th particle; is the total number of discretized positions of the EPE adhesive film sample, that is, the number of grid points; where represents the position index after discretization of the sample;

[0103] Initialize the particle velocities of the particle swarm optimization algorithm; designate as the velocity value of particle at the th position, representing the rate of thickness adjustment, and ; the velocities of each particle are randomly initialized and expressed as:

[0104] ;

[0105] where is the set of velocities of the th particle;

[0106] Define the fitness function, and evaluate the fitting degree of the solution of the particle to the actual interference image data through the fitness function:

[0107] ;

[0108] where is the The fitness value of a particle, where a smaller value indicates a higher fitting accuracy; is the intensity value of the actually measured interference image at point and is obtained by recording through an interference detector, representing the optical characteristics at this position; represents the intensity value of the interference image predicted by particle and is calculated through the thickness distribution model of the P layer of the glue film; represents the square of the error between the actually measured intensity and the predicted intensity at point and is used to reflect the local fitting deviation at this point; For further explanation of the fitness function:

[0109] ;

[0110] ;

[0111] Combined with the actually measured values, calculate the sum of the squares of the differences between the two to obtain the fitting error.

[0112] In each iteration, update the velocity according to the particle swarm optimization rule:

[0113] ;

[0114] where is the velocity vector of particle at the th iteration; is the velocity vector of particle at the th iteration; is the inertia weight, which is used to control the search range of the particle; is the acceleration coefficient, respectively controlling the weights of the individual historical solution and the global solution; is the random factor, and its value range is [0, 1]; is the historical best optimization solution of particle ; is the global best optimization solution among all particles; is the position of particle at the th iteration;

[0115] Update the position of the particle according to the velocity:

[0116] ;

[0117] where is the position of particle at the th iteration;

[0118] Establish an iterative stopping condition; the iterative process of the particle swarm optimization algorithm terminates when Condition 1 or Condition 2 is satisfied;

[0119] Condition 1 is: Less than a preset threshold;

[0120] Condition 2 is: the number of iterations reaches the upper limit.

[0121] Using the optimized parameters in, generate the thickness distribution map of the final EPE film sample;

[0122] ;

[0123] Wherein is the finally optimized thickness distribution value; the thickness distribution map of the EPE film sample is presented by marking correspondingly to each position of the EPE film sample, visually showing the spatial variation of the thickness;

[0124] Compare the optimized thickness distribution value with the actually measured interference image data, and calculate the final error:

[0125] ;

[0126] Wherein is the final verification error, and the final verification error represents the average error between the model prediction result and the actual data, is the total number of sampling points; represents the predicted intensity based on the optimized thickness distribution;

[0127] If the final verification error meets the accuracy requirement, the model optimization is completed; otherwise, adjust the parameters and re-optimize.

[0128] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization, characterized in that, Including: First, build an optical interference detection system. The optical interference detection system includes a laser light source, a beam splitter, an interference detector, and a data acquisition module. Place the EPE film sample in the measurement area of the optical interference detection system. Immediately adjust the optical path parameters of the laser light source and the beam splitter, and capture interference images at different positions through the interference detector. The interference images include fringe displacement, fringe spacing, and fringe intensity distribution, and the interference images are used to reflect the change information of the thickness of the P layer of the EPE film. The data acquisition module is used to record the interference images captured by the interference detector. Perform digital processing on the collected interference images to remove noise interference. Immediately extract the characteristic parameters of the fringes through an image processing algorithm. The image processing algorithm includes an edge detection algorithm. According to the physical characteristics of optical interference, correlate the characteristic parameters extracted from the interference images with the thickness distribution of the P layer of the film, and establish a thickness distribution model of the P layer of the film. Iteratively adjust the parameters of the thickness distribution model of the P layer of the film through the particle swarm optimization algorithm, optimize the fitting accuracy between the thickness distribution model of the P layer of the film and the actual interference image data, and complete the analysis and verification of thickness uniformity. Initialize the particle swarm optimization algorithm, set the particle swarm size, randomly generate multiple initial solutions in the parameter space, and each particle corresponds to a thickness distribution scheme. At the same time, initialize the velocity and position of the particles, and define the fitness function. Use the particle swarm optimization algorithm for iterative calculation. In each iteration, update the velocity and position of the particles according to the rules of particle swarm optimization. After each iteration, calculate the fitness value of each particle. The fitness value is used to reflect the error between the prediction result of the thickness distribution model of the P layer of the film and the actual measured interference image data. In each round of iteration, record the solution of the current particle individual, and continuously update the global solution. When the fitness value is less than the preset threshold or the number of its iterations reaches the upper limit, stop the optimization and output the final thickness distribution model of the P layer of the film. Use the parameters of the optimized thickness distribution model of the P layer of the film to generate a thickness distribution map of the EPE film sample, and display the thickness change conditions of different regions of the EPE film sample through the thickness distribution map. Based on the comparison between the thickness distribution map of the EPE film sample and the actual detection value, evaluate the fitting accuracy of the model. The characteristic composition of the interference image includes fringe displacement, fringe spacing, and fringe intensity distribution. The fringe displacement is expressed as: the offset of the fringe center in the plane; ； wherein , are respectively the coordinates of the current fringe center on the axis and the axis; , are respectively the coordinates of the reference fringe center on the axis and the axis; , are respectively the displacement changes of the fringes in two directions; The fringe spacing is expressed as: the distance between the centers of adjacent fringes. ； Among them is the Euclidean distance between the stripe centers; The fringe intensity distribution is expressed as: the intensity value of each point in the interference image is determined by the optical path difference. ； wherein represents the fringe intensity at point ; is the initial intensity of the light source; is the optical path difference between the reference beam and the measurement beam; is the wavelength of the laser light source; represents the refractive index of the EPE film sample at point ; represents the thickness of the EPE film sample at point ; Construct a thickness distribution model of the P layer of the adhesive film by comprehensively considering stripe displacement, stripe spacing, and stripe intensity distribution; denote the thickness of the EPE adhesive film sample at the point ; ； wherein represents the normalized parameter of the fringe pitch; represents the Gaussian attenuation factor of the fringe intensity varying with the fringe pitch; is the total number of fringes; is the Euclidean distance between the centers of the th fringe; is the reference fringe pitch.

2. The method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization according to claim 1, wherein: Initialize the particle positions of the particle swarm optimization algorithm; randomly generate particles in the parameter space of the thickness distribution model of the adhesive film P layer, and each particle represents a thickness distribution scheme; specify as the thickness value of the particle at the th position, and ; ； wherein is the thickness parameter set of the th particle; is the total number of positions after discretization of the EPE film sample; Initialize the particle velocity of the particle swarm optimization algorithm; formulate as the particle at the th position of the velocity value, and ; The velocity of each particle is randomly initialized as: ； Among them is the velocity set of the th particle; Define the fitness function, and evaluate the fitting degree between the solution of the particle and the actual interference image data through the fitness function. ； Among them is the fitness value of the th particle; is the intensity value of the actually measured interference image at point ; represents the predicted intensity value of the interference image of particle .

3. The method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization according to claim 2, wherein: In each iteration, update the velocity according to the particle swarm optimization rules. ； Among them is the velocity vector of the particle at the -th iteration; is the velocity vector of the particle at the -th iteration; is the inertia weight; is the acceleration coefficient, controlling the weights of the individual historical solution and the global solution respectively; is the random factor, whose value range is [0, 1]; is the historical optimal solution of the particle ; is the global optimal solution among all particles; is the position of the particle at the -th iteration; Update the position of the particle according to the velocity. ； wherein is the particle at the position at the th iteration; Establish an iteration stop condition. The iteration process of the particle swarm optimization algorithm terminates when condition one or condition two is satisfied. Condition 1 is that: less than a preset threshold value; Condition two is: the number of iterations reaches the upper limit.

4. A method for detecting the thickness uniformity of the P layer of an EPE film based on particle swarm optimization according to claim 3, characterized in that: Using the optimized parameters in, generate the thickness distribution map of the final EPE film sample; ； Among them is the finally optimized thickness distribution value; the thickness distribution diagram of the EPE film sample is presented by marking correspondingly to each position of the EPE film sample, intuitively showing the spatial variation of the thickness; Compare the optimized thickness distribution value with the actually measured interference image data, and calculate the final error: ； wherein is the final verification error; is the total number of sampling points; represents the predicted strength based on the optimized thickness distribution.

Images