An electrospray film forming quality evaluation method based on multi-droplet fusion monitoring

Abstract

The present application belongs to the field of inkjet printing, and particularly relates to an electrohydrodynamic atomization film forming quality evaluation method based on multi-droplet fusion monitoring, comprising: selecting multiple observation areas on a printing area, sequentially taking pictures of each observation area for R rounds to obtain R multi-droplet fusion images of each observation area and the acquisition time of each multi-droplet fusion image; extracting the edge information of fusion droplets in each multi-droplet fusion image of each observation area, and combining the acquisition time of each multi-droplet fusion image to obtain the fusion dynamic information of multi-droplets in each observation area over time; based on the corresponding fusion dynamic information of each observation area, using a pre-constructed film feature prediction model to predict the film feature of the solidified film in the observation area; inputting the film features of the solidified film in each observation area into a pre-constructed film quality classification model to predict the electrohydrodynamic atomization film forming quality classification of the entire printing area. The present application can efficiently and quantitatively analyze the electrohydrodynamic atomization film forming quality.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of inkjet printing, and more particularly relates to an electrospray film formation quality evaluation method based on multi-droplet fusion monitoring. BACKGROUND

[0002] Electrospray printing is a technology that uses electrospray technology to deposit a thin film on the surface of a substrate. It is a non-contact micro-nano printing technology that can achieve direct spraying of nanoscale solutions on flexible or hard substrates. Compared with traditional film production technologies, it does not require a vacuum environment, has low energy and low cost, high material utilization, flexible deposition of materials, patterning, and is suitable for large-scale manufacturing. It is one of the most commonly discussed emerging thin film deposition technologies.

[0003] During the production of thin films, the thickness and uniformity of the thin films usually need to be strictly controlled. However, research has found that many fluctuations in the external environment can cause fluctuations in the electrospray spray pattern, such as changes in the electric field distribution caused by changes in the position of the substrate shape, the relative position of other objects to the nozzle, liquid physical parameters (temperature, electrical conductivity, surface tension, mass density, and viscosity), liquid flow rate, external voltage, etc. These environmental fluctuations can cause instability in the spray pattern, nozzle flow rate, and droplet distribution, resulting in uneven thin film deposition. Therefore, it is necessary to strictly control the above variables and monitor the droplet deposition during production to eliminate possible thin film defects.

[0004] Current electrospray printing thin film defect detection methods mainly include cone detection and solidified film detection: by measuring the cone shape to detect whether the electrospray is normal, or by breaking the solidified film for testing, to determine whether the film thickness and surface roughness meet the requirements. Cone shape detection cannot quantitatively detect film uniformity and thickness. Solidified film needs to go through the steps of leveling and solidification, and the detection process is complicated, and it cannot be determined whether the film defect comes from the electrospray deposition process or from the liquid leveling and solidification process. Therefore, a new electrospray film formation quality evaluation method is needed to efficiently and quantitatively analyze the electrospray film formation quality. SUMMARY

[0005] In view of the defects and improvement needs of the prior art, the present application provides an electrospray film formation quality evaluation method based on multi-droplet fusion monitoring, which aims to provide an evaluation method that can efficiently and quantitatively analyze the electrospray film formation quality.

[0006] To achieve the above-mentioned purpose, according to one aspect of the present application, an electrospray film formation quality evaluation method based on multi-droplet fusion monitoring is provided, comprising:

[0007] After performing the electro-fluidic atomization printing, a plurality of observation areas are selected on the printing area, and R rounds of sequential image capturing are performed on each observation area to obtain R multi-droplet fusion images of each observation area and the capturing time of each multi-droplet fusion image;

[0008] Edge information of the fused droplets in each multi-droplet fusion image of each observation area is extracted, and the capturing time of each multi-droplet fusion image is combined to obtain fusion dynamic information of the multi-droplets in each observation area over time;

[0009] Based on the fusion dynamic information corresponding to each observation area, a pre-constructed film feature prediction model is used to predict the film feature of the solidified film in the observation area;

[0010] The film features of the solidified film in each observation area are input into a pre-constructed film quality classification model to predict the electro-fluidic atomization film formation quality classification of the entire printing area, and the electro-fluidic atomization film formation quality evaluation is completed.

[0011] Further, the plurality of observation areas are uniformly distributed on the printing area of the planar substrate.

[0012] Further, the film features of the solidified film in each observation area include the average film thickness of the solidified film and the surface roughness of the solidified film.

[0013] Further, the fusion dynamic information of the multi-droplets in each observation area over time is the edge dynamic feature of the fused droplets in the observation area, and the construction method is as follows:

[0014] The start time and the end time of the electro-fluidic atomization printing are obtained;

[0015] Based on the start time and the end time, the position of each observation area on the printing path is combined to determine the printing time at the observation area, and the time difference between the capturing time of each multi-droplet fusion image of the observation area and the printing time at the observation area is calculated.

[0016] The edge information of the fused droplets in the R multi-droplet fusion images of each observation area is extracted, and the edge information corresponding to the R multi-droplet fusion images of the observation area and the time difference are recorded in matrix form and the matrix is reduced in dimension to obtain an edge feature vector as the edge dynamic feature of the fused droplets in the observation area.

[0017] Further, the implementation of determining the printing time at each observation area is as follows:

[0018] The center of the i-th observation area is determined to be the closest to each local shortest point on the printing path of the nozzle, and the two points closest to the center of the i-th observation area are selected from all the points;

[0019] The printing time t_type at the corresponding printing position is calculated respectively in combination with the nozzle moving distance from the initial printing to the two positions i and t_type i ′, as the two printing times at the observation area:

[0020]

[0021]

[0022] In the formula, T0 is the initial time, T1 is the end time, S0 is the total nozzle moving distance for the electro-fluidic atomization printing on the planar substrate, S i is the nozzle moving distance from the initial printing to the printing of the first of the two positions, S i ′ is the nozzle moving distance from the initial printing to the printing of the second of the two positions, and i is an integer ranging from 1 to N.

[0023] Further, the fusion dynamic information of the multiple droplets over time in each observation area is the dynamic characteristics of the pores of the fused droplets in the observation area, and the construction manner is:

[0024] The binarization and edge detection are performed on each multiple-droplet fusion image of the i-th observation area, and the deposition area and the pore area are segmented, wherein i is an integer ranging from 1 to N.

[0025] Based on the segmentation results of the j-th and j+1-th multiple-droplet fusion images img i,j and img i,j+1 of the i-th observation area, the edges of the fused droplets in img i,j and img i,j+1 are determined, Q contour points are randomly selected on the edge of the fused droplets in img i,j , the nearest distance D i,j,q of the q-th contour point to the edge of the fused droplets in img i,j+1 is recorded, and the corresponding edge flow velocity i,j is calculated. i,j+1 Wherein, j is an integer ranging from 1 to R-1, Q is determined according to actual production requirements, and q is an integer ranging from 1 to Q; t_pic i,j , t_pic i,j+1 are the acquisition times of the j-th and j+1-th multiple-droplet fusion images of the i-th observation area, respectively.

[0026] Based on the segmentation results of the j-th and j+1-th multiple-droplet fusion images img i,j and img i,j+1 of the i-th observation area, the edges of the fused droplets in img i,j and imgi,j+1 The proportion of porous regions in p i,j and p i,j+1 To calculate the fusion speed

[0027] Construct the pore feature matrix H of the i-th observation region i :

[0028]

[0029] The pore feature matrix H i Dimensionality reduced to n-dimensional column vector H i We obtain the pore feature vector, which serves as the pore dynamic feature of the fused droplet in the i-th observation region.

[0030] This invention also provides a system for evaluating the quality of electrofluid atomization film formation based on multi-droplet fusion monitoring, used to perform the electrofluid atomization film formation quality evaluation method based on multi-droplet fusion monitoring as described above, including:

[0031] The downward-looking observation system includes a camera, a matching lens, and a matching coaxial light source, used to acquire multi-droplet fusion images;

[0032] Control unit, used to control the camera to capture images;

[0033] The data processing unit is used to process multi-droplet fusion images, extract features, predict the characteristics of the cured film in each observation area, and classify the electrofluid atomization film quality of the entire printing area.

[0034] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects:

[0035] (1) The method for evaluating the quality of the film formed by the electrohydrodynamic atomization based on the monitoring of the fusion of multiple droplets, after the electrohydrodynamic atomization printing is performed, a plurality of observation areas are selected on the printing area, R rounds of sequential image acquisition are performed on each observation area, R pieces of multiple-droplet fusion images of each observation area and the acquisition time of each piece of multiple-droplet fusion image are obtained, the fusion process of each observation area is monitored based on the multiple fusion images and the acquisition time, and the prediction results of the film thickness and the film thickness consistency of the plurality of observation areas are comprehensively considered. Compared with the current measurement method of destructive testing after solidification, the method does not need to wait for the solidification of the liquid film, does not need to destroy the film, has higher detection efficiency, and is more accurate in positioning the process of generating defects; in addition, compared with the current cone detection, the method does not focus on the huge amount of small droplets generated by the electrohydrodynamic atomization, greatly reduces the calculation amount of the image processing process, improves the system processing speed, and greatly reduces the required precision of the observation system; at the same time, for the solidified film quality evaluation, the prediction results of the film thickness and the film thickness consistency of the plurality of observation areas are comprehensively considered, the observation and data processing amount is reduced, the accuracy of the film evaluation result is ensured, and the method is suitable for the detection of the quality of a large-size film.

[0036] (2) The dynamic information of the fusion of multiple droplets in each observation area with time is the dynamic characteristics of the pores and / or the edges of the fused droplets in the observation area, and a specific construction method is given, the correlation between the liquid flow speed and the droplet distribution of the fused droplets and the liquid film surface roughness and the liquid film height after a certain flow leveling time is comprehensively considered, and the analysis of the solidified film thickness and uniformity is realized with high precision and quantitatively.

[0037] (3) The method for evaluating the quality of the film formed by the electrohydrodynamic atomization based on the monitoring of the fusion of multiple droplets, for the film feature prediction model, the input features of the model are reduced in dimension in the data preprocessing process, the calculation amount of the model is reduced, and at the same time, the model is suitable for different observation area numbers, different contour point numbers and different image acquisition round numbers of working scenes. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A flow chart of the method for evaluating the quality of the film formed by the electrohydrodynamic atomization based on the monitoring of the fusion of multiple droplets is provided for the embodiments of the present application;

[0039] Figure 2 An image acquisition process schematic diagram is provided for the embodiments of the present application;

[0040] Figure 3 A film feature prediction model schematic diagram based on the monitoring of the fusion of multiple droplets is provided for the embodiments of the present application;

[0041] Figure 4 A multiple-droplet fusion process schematic diagram is provided for the embodiments of the present application;

[0042] Figure 5 This is a schematic diagram of pore dynamic feature extraction provided in an embodiment of the present invention;

[0043] Figure 6 This is a schematic diagram of edge dynamic feature extraction provided in an embodiment of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0045] Example 1

[0046] A method for evaluating the quality of electrofluid atomization film formation based on multi-droplet fusion monitoring, such as... Figure 1 As shown, it includes:

[0047] After performing electrofluid atomization printing, multiple observation areas are selected on the printing area, and images are sequentially acquired in R rounds for each observation area to obtain R multi-droplet fusion images for each observation area and the acquisition time of each multi-droplet fusion image.

[0048] The edge information of the fused droplets in each multi-droplet fusion image of each observation area is extracted, and combined with the acquisition time of each multi-droplet fusion image, the dynamic information of multi-droplet fusion over time in each observation area is obtained.

[0049] Based on the fused dynamic information corresponding to each observation area, the characteristics of the cured film in that observation area are predicted using a pre-built thin film feature prediction model.

[0050] The characteristics of the cured film in each observation area are input into a pre-built film quality classification model to predict the electrofluid atomization film quality classification of the entire printing area, thus completing the electrofluid atomization film quality assessment.

[0051] According to existing coating flow models, liquid flow velocity and droplet distribution are closely related to the surface roughness and height of the liquid film after a certain leveling time. The electrohydrodynamic atomization printing film evaluation method proposed in this paper, based on droplet flow process observation, predicts the cured film thickness and thickness uniformity from the perspective of multi-droplet fusion. In other words, this method is a process observation method that can quantitatively analyze the cured film thickness and uniformity, and perform defect detection before the film curing step.

[0052] Among them, such as Figure 2 As shown, multi-droplet fusion image acquisition can be performed in the following manner:

[0053] a1, selecting N observation areas on the printing area on the planar substrate, the N observation areas are preferably uniformly distributed on the printing area on the planar substrate, the size of N is determined according to actual production requirements; the size of a single observation area is slightly smaller than the observation field of view of the observation system, so as to ensure that only one multi-droplet fusion image is collected for each observation area each time;

[0054] a2, during image collection, the downward-looking camera is turned on, and multi-droplet fusion images of each observation area are collected in turn, and the time of collecting each multi-droplet fusion image is recorded respectively;

[0055] a3, after the multi-droplet fusion images of all observation areas are collected, the downward-looking camera returns to the initial position, and the step a2 is repeated to perform the next round of image collection; the total number of rounds of image collection is R, R is greater than or equal to 3; the time of collecting the jth round of multi-droplet fusion image of the ith observation area is denoted as t_pic i,j , the multi-droplet fusion image is denoted as img i,j , the value range of i is an integer from 1 to N, and the value range of j is an integer from 1 to R.

[0056] The above plurality of observation areas are uniformly distributed on the printing area of the planar substrate, so as to ensure uniform sampling.

[0057] As an optimization, the cured film thin film characteristics of each observation area include the average film thickness of the cured film and the surface roughness of the cured film.

[0058] For this purpose, the electrohydrodynamic atomization printing thin film characteristic prediction model can be constructed in the following manner:

[0059] s1, calculating the pore characteristic vector and / or the edge characteristic vector of each observation area;

[0060] s2, after the printing is completed and the leveling time leveltime is elapsed, the printing sample is moved to the curing device for curing to obtain a cured film, and the surface morphology of the cured film is detected by a white light interferometer to obtain the average film thickness of the cured film and the surface roughness of the cured film, wherein the leveling time leveltime is determined by process requirements;

[0061] s3, training the thin film characteristic prediction model, as shown in the following formula: Figure 3 , wherein the input of the model training is a vector regression_x w , regression_x w is an n+m-dimensional column vector spliced by H w ' and / or M' w , wherein H w ' is the pore characteristic vector of the wth observation area, and M' wLet w be the edge feature vector of the w-th observation region, and let regression_y be the target vector of the model prediction. w regression_y w For FT w With Ra w The FT is a 2D column vector. w Ra is the average film thickness of the cured film in the w-th observation area. w Let w be the surface roughness of the cured film in the w-th observation area, and W be the sample size used to train the model. The value of w is determined based on actual production and ranges from 1 to W.

[0062]

[0063] regression_x w =[h1 h2 … h n ε1 ε2 … ε m ] T

[0064]

[0065] regression_y w =[FT w Ra w ] T .

[0066] Furthermore, the thin film quality classification model can be constructed as follows:

[0067] Based on process requirements, the cured films are evaluated. Films that meet the process requirements are classified as 1, and films that do not meet the requirements are classified as 0. The classification model input vector is `classification_x`, which is derived from the average film thickness (FT) of the cured films. i With surface roughness Ra i The composition is as follows: i takes the value of an integer from 1 to N, where N is the number of observation areas.

[0068] classification_x=[FT1…FT N Ra1 … Ra N ] T

[0069] Classification models can use SVM, DT, or ANN.

[0070] Multi-droplet fusion process such as Figure 4 As shown, as a preferred embodiment, the fusion dynamic information of multiple droplets in each observation area over time represents the porosity dynamic characteristics of the fused droplets within that observation area, and its construction method is as follows:

[0071] The j-th round multi-droplet fusion image of the i-th observation region (img) i,j Binarization and edge detection are performed to segment the deposition area and the porosity area, resulting in an image. i,j The edge of the fused droplet, where i is an integer ranging from 1 to N, and j is an integer ranging from 1 to R-1;

[0072] Based on the segmentation results, in the multi-droplet fusion image (img) i,j Q contour points are randomly selected on the edge, and the value of Q is determined according to actual production needs;

[0073] The (j+1)th round multi-droplet fusion image of the i-th observation region (img) i,j+1 Edge detection is performed to segment the deposition area and the porosity area, resulting in an image. i,j+1 At the edge of the fused droplet, record the q-th contour point and the img i,j+1 The nearest distance D to the edge i,j,q (like Figure 5 (As shown) and calculate the corresponding edge flow velocity. Where i is an integer ranging from 1 to N, j is an integer ranging from 1 to R-1, q is an integer ranging from 1 to Q, and t_pic i,j Let t_pic be the acquisition time of the j-th round of multi-droplet fusion image for the i-th observation region. i,j+1 The acquisition time of the (j+1)th round of multi-droplet fusion image for the i-th observation area;

[0074] Calculate the j-th round multi-droplet fusion image (img) for the i-th observation region. i,j The proportion of mesoporous regions p i,j and the (j+1)th round multi-droplet fusion image of the i-th observation region (img) i,j+1 The proportion of mesoporous regions p i,j+1 And calculate the fusion speed. Where i is an integer ranging from 1 to N, and j is an integer ranging from 1 to R-1;

[0075] Construct a matrix V with length R-1 and width Q. i and a vector U of dimension R-1 i ′, and matrix V i and vector U i By splicing the two halves together, a matrix H with a width of R-1 and a length of Q+1 is formed. i :

[0076]

[0077] U i ′=[U i,1 Ui,2 … U i,R-1 ];

[0078]

[0079] matrix H i Dimensionality reduced to n-dimensional column vector H i ′, H i ′=[h1 h2 … h n The value of n is determined based on the actual situation, H i ′ is the pore feature vector of the i-th observation region, which serves as the pore dynamic feature of the fused droplet within the i-th observation region.

[0080] Further integration Figure 4 The multi-droplet fusion process shown is a preferred embodiment. The fusion dynamic information of multiple droplets in each observation area over time is the edge dynamic feature of the fused droplets in that observation area, and its construction method is as follows:

[0081] Obtain the start and end times of electrofluid atomization printing;

[0082] Based on the start time and the end time, and combined with the position of each observation area on the printing path, the printing time at the observation area is determined, and the time difference between the acquisition time of each multi-droplet fusion image in the observation area and the printing time at the observation area is calculated.

[0083] Edge information of fused droplets is extracted from R multi-droplet fusion images of each observation area. The edge information and time difference corresponding to the R multi-droplet fusion images of that observation area are recorded as a matrix M. i The data is dimensionality reduced to obtain the main features of the original matrix, thereby reducing the computational cost of the model, avoiding overfitting, and obtaining the edge dynamic features of the fused droplets within the observation area, such as... Figure 6 As shown, the specific steps are as follows:

[0084] S1, the j-th round multi-droplet fusion image of the i-th observation region (img) i,j Binarization and edge detection are performed to segment the deposition area and the porosity area, resulting in an image. i,j The edge information of the droplet is fused, where i is an integer ranging from 1 to N, and j is an integer ranging from 1 to R.

[0085] S2. Record the edge information and corresponding time differences of the R multi-droplet fusion images of the i-th observation area as a matrix M. i Matrix M i Width is H imgi,j The length is W img The height is R, H img W represents the image height.img is the image width, R is the number of image acquisition rounds, i is an integer ranging from 1 to N;

[0086] M i =

[0087] [Edge i,1 Edge i,1 …Edge i,R Edge i,R ][Δt i,1 [0] Δt i,1 [1] …Δt i,R [0] Δt i,R [1] ] T ; wherein, Edge i,j is the edge information matrix of the jth multi-droplet fusion image of the ith observation area, and the edge pixels in the matrix are represented by 1, and other pixels in the image are represented by 0; Δt i,j [0] and Δt i,j [1] are two time differences corresponding to the jth multi-droplet fusion image of the ith observation area.

[0088] S3, reducing M i to an m-dimensional column vector M i ', as the edge feature vector of the ith observation area, to complete the construction of the edge dynamic characteristics of the fused droplets in the ith observation area.

[0089] It should be noted that the feature reduction described in this embodiment can use PCA dimension reduction, decision tree induction or Auto-Encoder data dimension reduction.

[0090] Further preferably, the implementation of the printing time at each observation area is determined as follows:

[0091] Determine the center distance of the ith observation area to each local shortest point on the printing path of the nozzle, and select the two point positions with the shortest distance to the center of the ith observation area from all point positions;

[0092] Combine the nozzle moving distance from the initial printing to the two point positions, and calculate the printing time t_type i and t_type i ' at the corresponding printing positions, respectively, as the two printing times at the observation area:

[0093]

[0094]

[0095] wherein, T0 is the start time, T1 is the end time, S0 is the total distance of the print head movement for electrohydrodynamic atomization printing on the planar substrate, S i is the distance of the print head movement from the initial printing to the time when the print head prints the first point of the two points, S i is the distance of the print head movement from the initial printing to the time when the print head prints the second point of the two points, and i is an integer ranging from 1 to N.

[0096] The jth multi-droplet fusion image of the ith observation area corresponds to two time differences, which are represented as follows:

[0097] Δt i,j [0] = t_pic i,j -t_type i,j ;

[0098] Δt i,j [1] = t_pic i,j -t_type i ′.

[0099] It should be noted that, since the printing path is a meandering curve, there are multiple vertical distances from the center point of each observation area to the printing path, and therefore the distance from the center of the observation area to the printing path presents a periodic law of from far to near and then to far again, and the nearest distance of each period corresponds to a printing path point as one of the above local minimum points.

[0100] In general, the method can be divided into the following steps: a, electrohydrodynamic atomization printing is performed to deposit atomized droplets on a planar substrate, and the printing time is recorded; after the printing is completed, N observation areas are selected on the printed area on the planar substrate, a downward observation system is used to take multiple images of each observation area, and the time difference between the printing time and the image taking time is calculated; b, image processing is performed on the multi-droplet fusion image, the pore area without droplet deposition is divided, and the pore features and / or image edge features are extracted; c, based on the pore features and / or image edge features, a pre-constructed film feature prediction model is used to predict the thickness and roughness of the solidified film of each observation area; d, the measurement results of each observation area are comprehensively considered, a pre-constructed film quality classification model is used to complete the film quality classification.

[0101] The application provides an electrohydrodynamic atomization film forming quality evaluation method based on multi-droplet fusion monitoring. After electrohydrodynamic atomization printing is performed, a plurality of observation areas are selected on a printing area, R rounds of sequential image acquisition are performed on each observation area, R multi-droplet fusion images of each observation area and acquisition times of each multi-droplet fusion image are obtained, a multi-droplet fusion process of each observation area is monitored based on the plurality of fusion images and the acquisition times, and predicted results of film thickness and film thickness consistency of the plurality of observation areas are comprehensively considered. Compared with a current measurement method of destructive testing after solidification, the method does not need to wait for the liquid film to solidify, does not need to destroy the film, has higher detection efficiency, and is more accurate in positioning a process causing defects. In addition, compared with current cone detection, the method does not focus on a large amount of small droplets generated by electrohydrodynamic atomization, greatly reduces the calculation amount of the image processing process, improves the system processing speed, and greatly reduces the required precision of the observation system. Meanwhile, for solidified film quality evaluation, the application comprehensively considers the predicted results of film thickness and film thickness consistency of the plurality of observation areas, reduces the observation and data processing amount, guarantees the accuracy of the film evaluation result, and is suitable for detection of large-size film quality.

[0102] Embodiment two

[0103] An electrohydrodynamic atomization film forming quality evaluation system based on multi-droplet fusion monitoring is used for performing the electrohydrodynamic atomization film forming quality evaluation method based on multi-droplet fusion monitoring in the above embodiment one, and comprises the following.

[0104] A downward observation system comprises a camera, a matched lens and a matched coaxial light source, and is used for acquiring the multi-droplet fusion image.

[0105] A control unit is used for controlling the camera to acquire images.

[0106] A data processing unit is used for processing the multi-droplet fusion image, feature extraction, predicting film features of the solidified film of each observation area and classifying electrohydrodynamic atomization film forming quality of the whole printing area.

[0107] The related technical solution is the same as that in the embodiment one, and will not be described here.

[0108] It is easy for those skilled in the art to understand that the above description is only the preferred embodiment of the application, and is not used to limit the application, and any modification, equivalent replacement and improvement made within the spirit and principle of the application should be included in the protection scope of the application.

Claims

1. A method for evaluating the quality of electrohydrodynamic atomization film formation based on multi-droplet fusion monitoring, characterized in that, include: After performing electrofluid atomization printing, multiple observation areas are selected on the printing area, and images are sequentially acquired in R rounds for each observation area, resulting in R multi-droplet fusion images for each observation area and the acquisition time of each multi-droplet fusion image. The edge information of the fused droplets in each multi-droplet fusion image of each observation area is extracted, and combined with the acquisition time of each multi-droplet fusion image, the dynamic information of multi-droplet fusion over time in each observation area is obtained. Based on the fused dynamic information corresponding to each observation area, the characteristics of the cured film in that observation area are predicted using a pre-constructed thin film feature prediction model. The characteristics of the cured film in each observation area are input into a pre-built film quality classification model to predict the electrofluid atomization film quality classification of the entire printing area, thus completing the electrofluid atomization film quality assessment. Among them, the fusion dynamic information of multiple droplets in each observation area over time is the porosity dynamic characteristic of the fused droplets in that observation area, and its construction method is as follows: For each multi-droplet fusion image of the i-th observation area, binarization and edge detection are performed to segment the deposition area and the pore area, where the value of i is an integer from 1 to N; Based on the j-th and (j+1)-th multi-droplet fusion images of the i-th observation region (img) i,j ,img i,j+1 The segmentation results are used to determine the img values ​​respectively. i,j and img i,j+1 The edge of the fused droplet, in img i,j Q contour points are randomly selected on the edge of the fused droplet, and the relationship between the q-th contour point and the image is recorded. i,j+1 The closest distance D to the edge of the fused droplet i,j,q And calculate the corresponding edge flow velocity. Where j takes the value of an integer from 1 to R-1, Q takes the value determined according to actual production needs, and q takes the value of an integer from 1 to Q; t_pic i,j t_pic i,j+1 These are the acquisition times of the j-th and (j+1)-th multi-droplet fusion images of the i-th observation area, respectively; Based on the j-th and (j+1)-th multi-droplet fusion images of the i-th observation region (img) i,j ,img i,j+1 The segmentation results are used to calculate the img values ​​respectively. i,j and img i,j+1 The proportion of porous regions in p i,j and p i,j+1 To calculate the fusion speed Construct the pore feature matrix H of the i-th observation region i : The pore feature matrix H i Dimensionality reduced to n-dimensional column vector H i ′, and obtain the pore feature vector of the i-th observation area, which serves as the pore dynamic feature of the fused droplet in the i-th observation area.

2. The method for evaluating the quality of electrofluid atomization film formation according to claim 1, characterized in that, The multiple observation areas are evenly distributed on the printing area of ​​the planar substrate.

3. The method for evaluating the quality of electrofluid atomization film formation according to claim 1, characterized in that, The characteristics of the cured film include the average thickness of the cured film and the surface roughness of the cured film.

4. The method for evaluating the quality of electrofluid atomization film formation according to claim 1, characterized in that, The fusion dynamics of multiple droplets over time in each observation region constitute the edge dynamics of the fused droplets within that region, and their construction method is as follows: Obtain the start and end times of the electrofluid atomization printing; Based on the start time and the end time, and combined with the position of each observation area on the printing path, the printing time at the observation area is determined, and the time difference between the acquisition time of each multi-droplet fusion image in the observation area and the printing time at the observation area is calculated. Edge information of fused droplets is extracted from R multi-droplet fusion images of each observation area. The edge information and time difference corresponding to the R multi-droplet fusion images of the observation area are recorded as a matrix and the data dimensionality is reduced to obtain the edge feature vector, which serves as the edge dynamic feature of the fused droplets in the observation area.

5. The electrofluid atomization film quality evaluation method according to claim 4, characterized in that, The method for determining the printing time at each observation zone is as follows: Determine the first i The shortest distance between the center of each observation area and the local point on the print path of the nozzle is selected, and among all points, the one that is closest to the first point is chosen. i The two points with the shortest distance between the centers of the observation area; Based on the nozzle travel distance from the initial printing point to the two printing positions, calculate the printing time t_type at each corresponding printing position. i and t_type i ′, as two printing times in that observation area: In the formula, T0 is the start time, T1 is the end time, S0 is the total distance the print head travels during electrohydraulic atomization printing on the planar substrate, and S... i S is the printhead travel distance from the initial printing to the point where the printhead prints the first of the two points. i ′ represents the distance the printhead travels from the initial printing to the second of the two points, and the value of i is an integer ranging from 1 to N.

6. A electrofluid atomization film formation quality assessment system based on multi-droplet fusion monitoring, characterized in that, A method for evaluating the quality of electrofluid atomization film formation based on multi-droplet fusion monitoring as described in any one of claims 1 to 5, comprising: The downward-looking observation system includes a camera, a matching lens, and a matching coaxial light source, used to acquire multi-droplet fusion images; Control unit, used to control the camera to capture images; The data processing unit is used to process multi-droplet fusion images, extract features, predict the characteristics of the cured film in each observation area, and classify the electrofluid atomization film quality of the entire printing area.

Images