Flexible circuit board quality detection method and system based on data analysis

Abstract

The invention relates to the technical field of circuit board quality detection, and discloses a flexible circuit board quality detection method and system based on data analysis, and the method comprises the steps: collecting high-definition image data, electrical data and mechanical test data of a flexible circuit board, analyzing the surface layer shape and quality misaccuracy, guide topology misaccuracy and film defect degree of the flexible circuit board, and determining the quality of the flexible circuit board according to the surface layer shape and quality misaccuracy, guide topology misaccuracy and film defect degree of the flexible circuit board; determining the surface efficiency loss degree of the flexible circuit board; calculating an electrical signal attenuation rate and an electrical signal distortion rate of the flexible circuit board so as to analyze a topology deviation degree of the electrical appliance performance corresponding to the flexible circuit board; analyzing a fatigue durability limit value and a flexibility decline coefficient of the flexible circuit board so as to analyze the mechanical performance reduction of the flexible circuit board; and performing comprehensive quality analysis on the flexible circuit board according to the surface efficiency loss degree, the topology deviation degree and the mechanical performance reduction amount to obtain a quality detection report. According to the invention, the accuracy of quality detection of the flexible circuit board can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to a flexible circuit board quality detection method and system based on data analysis, and the technical field of circuit board quality detection. BACKGROUND

[0002] The flexible circuit board is a printed circuit board made of flexible insulating substrates such as polyimide or polyester film, which has the characteristics of bendable, foldable and stretchable. Compared with traditional rigid circuit boards, it has the advantages of lightness, thinness, small size, high wiring density and three-dimensional assembly. Therefore, the application field of flexible circuit board is very wide, such as in electronic products, automotive electronics, medical equipment and aerospace field, so it is very important to control the production quality of flexible circuit board.

[0003] At present, the quality detection of flexible circuit board is generally carried out by single-point electrical test method. The contact type measurement is carried out on the preset limited test points (usually only covering 5%-10% of the total line) by using a multimeter, and the line conduction is judged by measuring the direct current resistance value between two points. However, such detection method has the problem of single detection dimension, lacks comprehensive consideration of the quality of flexible circuit board, and leads to insufficient detection accuracy of flexible circuit board. SUMMARY

[0004] The present application provides a flexible circuit board quality detection method and system based on data analysis, which mainly aims to improve the accuracy of flexible circuit board quality detection.

[0005] To achieve the above purpose, the flexible circuit board quality detection method based on data analysis provided by the present application comprises: Collecting high-definition image data, electrical data and mechanical test data of the flexible circuit board, performing double-branch feature fusion on the high-definition image data to obtain efficient fusion features, using the efficient fusion features to analyze the surface quality deviation degree, circuit topology deviation degree and film defect degree of the flexible circuit board, and determining the surface performance loss degree of the flexible circuit board; Constructing a time-frequency heat map of the electrical data, calculating the electrical signal attenuation rate and electrical signal distortion rate of the flexible circuit board by using the time-frequency heat map, and analyzing the topology deviation degree of the corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation rate and the electrical signal distortion rate; Using the mechanical test data to construct a mechanical performance parameter matrix of the flexible circuit board, using the mechanical performance parameter matrix to analyze the fatigue endurance limit value and flexibility decay coefficient of the flexible circuit board, and using the fatigue endurance limit value and the flexibility decay coefficient to analyze the mechanical performance degradation of the flexible circuit board; According to the surface performance loss degree, the topological deviation degree and the mechanical performance reduction amount, a comprehensive quality analysis is performed on the flexible circuit board to obtain a quality detection report.

[0006] Optionally, the high-definition image data is subjected to double-branch feature fusion to obtain high-efficiency fusion features, including: extracting double-branch features of the high-definition image data to obtain local detail features and global structure features; performing cross-branch spatial alignment on the local detail features and the global structure features to obtain aligned double-branch features; performing multi-modal fusion on the aligned double-branch features to obtain preliminary fusion features; performing dimension reduction optimization on the preliminary fusion features to obtain high-efficiency fusion features.

[0007] Optionally, the high-efficiency fusion features are used to analyze surface layer quality misalignment degree, guide path topological misalignment degree and film defect degree of the flexible circuit board to determine the surface performance loss degree of the flexible circuit board, including: using the high-efficiency fusion features to perform defect region segmentation on the flexible circuit board to obtain a surface layer defect distribution map and a film defect distribution map; performing geometric parameter quantization on the surface layer defect distribution map to obtain the surface layer quality misalignment degree; using the high-efficiency fusion features to perform guide path topological registration on the flexible circuit board to obtain a guide path offset heat map; performing spatial error statistics on the guide path offset heat map to obtain the guide path topological misalignment degree; using the film defect distribution map to analyze the film defect degree of the flexible circuit board; performing weighted fusion on the surface layer quality misalignment degree, the guide path topological misalignment degree and the film defect degree to obtain the surface performance loss degree.

[0008] Optionally, the time-frequency heat map is used to calculate the electrical signal attenuation rate and the electrical signal distortion rate of the flexible circuit board, including: extracting a fundamental wave energy curve of the time-frequency heat map, and determining input energy and output energy of the flexible circuit board based on the fundamental wave energy curve; based on the input energy and the output energy, calculating the electrical signal attenuation rate of the flexible circuit board using the following formula:

[0009] wherein, represents the electrical signal attenuation rate, represents the input energy, represents the output energy; Equally interval sampling the signal waveform in the time-frequency heat map to obtain a sampling signal point; Calculating a signal value of the sampling signal point; Based on the signal value, calculating an electrical signal distortion rate of the flexible circuit board by using the following formula:

[0010] Wherein, The electrical signal distortion rate is represented, n represents the number of sampling signal points, The signal value of the i-th sampling signal point is represented, The ideal signal value of the i-th sampling signal point is represented.

[0011] Optionally, based on the electrical signal attenuation rate and the electrical signal distortion rate, analyzing the topological deviation degree of the corresponding electrical performance of the flexible circuit board, comprising: Using the electrical signal attenuation rate and the electrical signal distortion, performing impedance inversion calculation on the flexible circuit board to obtain an equivalent impedance deviation degree and a nonlinear distortion coefficient; According to the equivalent impedance deviation degree and the nonlinear distortion coefficient, performing electrical field distribution reconstruction on the flexible circuit board to obtain an electrical abnormal field; Using the electrical abnormal field, constructing a guide network diagram of the flexible circuit board; Using the guide network diagram, identifying the topological deviation degree of the flexible circuit board.

[0012] Optionally, using the mechanical test data, constructing a mechanical performance parameter matrix of the flexible circuit board, comprising: Extracting test parameters in the mechanical test data; Using the test parameters, constructing a mechanical performance parameter matrix of the flexible circuit board, wherein the mechanical performance parameter matrix can be represented by using the following matrix:

[0013] Wherein, The mechanical performance parameter matrix is represented, The first test parameter of the elastic modulus in the test parameters is represented, The first test parameter of the tensile strength in the test parameters is represented, The first test parameter of the elongation at break in the test parameters is represented, The first test parameter of the bending stiffness in the test parameters is represented, The Nth test parameter of the elastic modulus in the test parameters is represented, The Nth test parameter of the tensile strength in the test parameters is represented, Nth test parameter representing the tensile elongation of the test parameter, Nth test parameter representing the bending stiffness of the test parameter.

[0014] Optionally, the fatigue endurance limit and the flexibility degradation coefficient of the flexible circuit board are analyzed by using the mechanical property parameter matrix, including: The dynamic stress-strain characteristics of the flexible circuit board are identified by using the mechanical property parameter matrix; Based on the dynamic stress-strain characteristics, a stress-life characteristic set and an elastic modulus attenuation sequence of the flexible circuit board are constructed; The material fatigue curve of the flexible circuit board is constructed by using the stress-life characteristic set; Based on the material fatigue curve, the fatigue endurance limit of the flexible circuit board is calculated; Based on the modulus attenuation sequence, the flexible circuit board is subjected to exponential attenuation fitting to obtain a flexibility degradation coefficient.

[0015] Optionally, the mechanical performance degradation of the flexible circuit board is analyzed by using the fatigue endurance limit and the flexibility degradation coefficient, including: The fatigue endurance limit and the flexibility degradation coefficient are normalized to obtain a standardized life index and a standardized degradation rate; The standardized life index and the standardized degradation rate are weighted and fused to obtain an initial performance degradation; The initial performance degradation is subjected to environmental compensation correction to obtain a corrected performance degradation; The corrected performance degradation is subjected to hierarchical mapping to obtain a mechanical performance degradation.

[0016] Optionally, the flexible circuit board is subjected to comprehensive quality analysis according to the surface performance loss degree, the topological deviation degree, and the mechanical performance degradation to obtain a quality detection report, including: The surface performance loss degree, the topological deviation degree, and the mechanical performance degradation are subjected to data comprehensive processing to obtain a comprehensive data set; The interactive influence matrix of the flexible circuit board is constructed by using the comprehensive data set; The defect coupling coefficient of the flexible circuit board is calculated based on the interactive influence matrix; The defect coupling coefficient is subjected to dynamic weight correction to obtain a corrected coupling coefficient; The comprehensive quality of the flexible circuit board is evaluated based on the corrected coupling coefficient; The quality detection report of the flexible circuit board is constructed based on the evaluation result of the comprehensive quality evaluation.

[0017] To solve the above problems, the application further provides a flexible circuit board quality detection system based on data analysis, which comprises: a surface quality analysis module, which is used for collecting high-definition image data, electrical data and mechanical test data of the flexible circuit board, performing double-branch feature fusion on the high-definition image data to obtain efficient fusion features, using the efficient fusion features to analyze surface performance loss and defect degrees of the flexible circuit board, including surface layer shape error, circuit topology error and film defect, so as to determine surface performance loss and defect degrees of the flexible circuit board; an electrical quality analysis module, which is used for constructing a time-frequency heat map of the electrical data, using the time-frequency heat map to calculate electrical signal attenuation rate and electrical signal distortion rate of the flexible circuit board, and analyzing topology deviation degrees of corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation rate and the electrical signal distortion rate; a mechanical quality analysis module, which is used for constructing a mechanical performance parameter matrix of the flexible circuit board using the mechanical test data, using the mechanical performance parameter matrix to analyze fatigue endurance limit value and flexibility recession coefficient of the flexible circuit board, and using the fatigue endurance limit value and the flexibility recession coefficient to analyze mechanical performance degradation of the flexible circuit board; a comprehensive quality analysis module, which is used for performing comprehensive quality analysis on the flexible circuit board according to the surface performance loss and defect degrees, the topology deviation degrees and the mechanical performance degradation, so as to obtain a quality detection report.

[0018] Compared with the problems described in the background art, the present application first captures potential problems in multiple dimensions by collecting high-definition images, electrical and mechanical test data, then performs double-branch feature fusion on the high-definition image data, integrates local and global features to improve defect detection accuracy, and then uses efficient fusion features to analyze surface quality loss degree, guide way topology loss degree and coating defect degree, and determines the surface performance loss degree after weighted fusion, and accurately quantifies the surface quality state; Further, the present application can quickly locate electrical defects such as signal distortion by constructing a time-frequency heat map of electrical data and visualizing the changes of electrical parameters with time and frequency, and then calculating the electrical signal attenuation rate and distortion rate using the time-frequency heat map to quantitatively evaluate signal transmission problems, and based on the two indicators, impedance inversion calculation, electrical field distribution reconstruction, and guide way network graph construction are performed to finally identify the topology deviation degree, thereby intuitively reflecting the deviation degree of the circuit topology structure from the design standard; Further, the present application can quantitatively evaluate the life threshold and performance decay rate under mechanical stress to intuitively reflect the influence of mechanical stress on the performance of the circuit board by constructing a mechanical performance parameter matrix using mechanical test data to present the mechanical properties of the circuit board under bending, stretching and other working conditions, then identifying dynamic stress-strain characteristics based on the matrix, and constructing a stress-life feature set and an elastic modulus decay sequence to obtain a material fatigue curve and calculate the fatigue endurance limit and flexibility decay coefficient; Further, the present application can intuitively present the correlation strength of each quality index by constructing an interaction matrix through data comprehensive processing of the surface performance loss degree, topology deviation degree and mechanical performance degradation, and then calculating the defect coupling coefficient to represent the interaction degree of multiple defects according to the interaction matrix, and dynamically correcting the weight according to the application scenario, and then performing comprehensive quality analysis on the flexible circuit board. Therefore, the present application can improve the accuracy of flexible circuit board quality detection. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The flowchart of the flexible circuit board quality detection method based on data analysis provided by an embodiment of the present application is shown. Figure 2 The module diagram of the flexible circuit board quality detection method based on data analysis provided by an embodiment of the present application is shown.

[0020] The purpose of the present application, functional features and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0021] It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0022] The embodiment of the present application provides a flexible circuit board quality detection method based on data analysis. The execution subject of the flexible circuit board quality detection method based on data analysis includes but is not limited to at least one of the electronic devices such as a server, a terminal and the like which can be configured to execute the method provided by the embodiment of the present application. In other words, the flexible circuit board quality detection method based on data analysis can be executed by the software or hardware installed in the terminal device or the server device. The server includes but is not limited to a single server, a server cluster, a cloud server or a cloud server cluster and the like.

[0023] Embodiment 1 Referring to Figure 1 Fig. 1 shows a flowchart of the flexible circuit board quality detection method based on data analysis provided by an embodiment of the present application. In the embodiment, the flexible circuit board quality detection method based on data analysis includes the following steps. S1, collecting high-definition image data, electrical data and mechanical test data of the flexible circuit board, performing double-branch feature fusion on the high-definition image data to obtain efficient fusion features, and using the efficient fusion features to analyze surface quality deviation, circuit topology deviation and film defect degree of the flexible circuit board to determine surface performance loss degree of the flexible circuit board.

[0024] The embodiment of the present application can comprehensively capture potential problems of the circuit board from multiple dimensions by collecting high-definition image data, electrical data and mechanical test data of the flexible circuit board. For example, high-definition images can help to find surface scratches, stains and other appearance defects; electrical data can help to detect short circuit, open circuit and other performance abnormalities; and mechanical test data can evaluate the flexibility and fatigue resistance of the circuit board, for example, through bending test to check whether there is wrinkle or fracture, thereby providing the most critical data support for accurately positioning quality problems and ensuring product performance.

[0025] The high-definition image data refers to the visualization data of the surface and internal structure of the flexible circuit board, such as physical and mechanical performance test and film integrity of the flexible circuit board. The electrical data refers to signal characteristic data of the circuit board under the condition of power on, such as voltage / current waveform and impedance spectrum. The mechanical test data refers to performance data of the circuit board under mechanical stress, such as bending stiffness and tensile strength.

[0026] Optionally, the high-definition image data can be collected by a high-resolution industrial camera, the electrical data can be obtained by electrical test equipment such as a multimeter and an oscilloscope, and the mechanical test data can be obtained by querying the test data recorded after the physical and mechanical performance test of the flexible circuit board.

[0027] Further, the embodiment of the present application can obtain high-efficiency fusion features by double-branch feature fusion of the high-definition image data, which can integrate local details and global structure features in the image, and improve the detection accuracy of the surface defects of the flexible circuit board.

[0028] The high-efficiency fusion features refer to multi-scale and multi-dimensional image feature representations integrated by the double-branch feature fusion technology, which can capture microscopic defects and macroscopic structure abnormalities of the flexible circuit board.

[0029] As an embodiment of the present application, the double-branch feature fusion of the high-definition image data is performed to obtain high-efficiency fusion features, which includes: extracting double-branch features of the high-definition image data to obtain local detail features and global structure features; aligning the local detail features and the global structure features in cross-branch space to obtain aligned double-branch features; performing multi-modal fusion on the aligned double-branch features to obtain preliminary fusion features; performing dimension reduction optimization on the preliminary fusion features to obtain high-efficiency fusion features.

[0030] The local detail features refer to microscopic physical defect features of the high-definition image extracted by a convolutional neural network, the global structure features refer to macroscopic topological correlation features of the high-definition image extracted based on a visual or graph neural network, and the aligned double-branch features refer to features of the local detail features calibrated in spatial position according to the global structure features.

[0031] In the specific implementation process, a double-branch convolutional neural network can be designed, one branch of which uses a small convolution kernel (such as 3x3) and a multi-level cascade to perform pixel-by-pixel convolution operation on the high-definition image data to extract local detail features such as scratch texture and line edge, and the other branch uses a large convolution kernel (such as 7x7) and a hollow convolution to obtain global structure features such as overall layout and pattern distribution of the circuit board with fewer convolution layers; a spatial transformation network is used to calculate the accurate spatial position transformation parameters of the local detail features in the image according to the overall layout information provided by the global structure features, and then affine transformation, rotation or scaling operation is performed on the local detail features to accurately align them in the spatial dimension and form aligned double-branch features; the local detail feature tensor and the global structure feature tensor after alignment are merged in the channel dimension by using tensor splicing to generate preliminary fusion features; the principal component analysis algorithm is used to calculate the covariance matrix of the preliminary fusion features to extract the main component direction, and then the high-dimensional feature vector is projected into a low-dimensional space, and finally the auto-encoder network is used to learn the compressed representation of the features through the encoding-decoding structure to remove redundant information, and the high-efficiency fusion features are obtained.

[0032] Further, the embodiment of the present application analyzes the surface layer quality misalignment degree, the guide topology misalignment degree and the film defect degree of the flexible circuit board by using the high-efficiency fusion feature to determine the surface performance loss and defect degree of the flexible circuit board. The defect information such as scratches, line short circuits and film bubbles in the image can be quantified as misalignment degree and defect degree indexes, and the surface performance loss and defect degree is obtained through comprehensive evaluation, so that the surface quality state of the circuit board can be accurately judged. For example, in actual production, if the high-efficiency fusion feature shows that there are multiple line short circuits and film bubbles in a batch of circuit boards, it is determined through analysis that the surface performance loss and defect degree is high, which indicates that the surface quality of the batch of products is unqualified.

[0033] The surface layer quality misalignment degree refers to the deviation degree between the surface physical form and the standard design form of the flexible circuit board due to surface defect such as scratches, scratches, stains and indentation. The guide topology misalignment degree refers to the degree of deviation of the circuit connection structure and layout from the design topology structure due to the problems such as short circuit, open circuit and pattern shift of the circuit on the circuit board. The film defect degree refers to the degree of decline of the protection function of the cover film and the insulating layer due to defects such as bubbles, wrinkles, missing and edge lifting. The surface performance loss and defect degree refers to the quantitative evaluation of the damage degree of the overall function and performance of the surface of the flexible circuit board, which directly reflects the influence degree of the surface quality of the circuit board on the normal use performance.

[0034] As an embodiment of the present application, the surface performance loss and defect degree of the flexible circuit board is determined by analyzing the surface layer quality misalignment degree, the guide topology misalignment degree and the film defect degree of the flexible circuit board by using the high-efficiency fusion feature, comprising: The high-efficiency fusion feature is used to segment the defect area of the flexible circuit board to obtain a surface layer defect distribution map and a film defect distribution map; The surface layer defect distribution map is quantified by geometric parameters to obtain the surface layer quality misalignment degree; The high-efficiency fusion feature is used to register the guide topology of the flexible circuit board to obtain a guide offset heat map; The space error of the guide offset heat map is counted to obtain the guide topology misalignment degree; The film defect degree of the flexible circuit board is analyzed by using the film defect distribution map; The surface layer quality misalignment degree, the guide topology misalignment degree and the film defect degree are weighted and fused to obtain the surface performance loss and defect degree.

[0035] In the implementation process, the high-efficiency fusion feature can be input into a semantic segmentation network (such as U-Net DeepLabV3+) to perform pixel-level classification on the image, identify the positions of surface defects (such as scratches and stains) and film defects (such as bubbles and wrinkles), and then output a binary defect distribution map; morphological analysis is performed on the surface defect distribution map to calculate the area, perimeter and aspect ratio of each defect region, and then the defects are graded according to a preset threshold (such as a scratch length > 5 mm or a stain area > 0.5 mm²), for example, a scratch length > 7 mm corresponds to a surface quality inaccuracy of level 1, and a scratch length > 9 mm corresponds to a surface quality inaccuracy of level 2, and the specific threshold can be set according to actual application or process requirements; the line feature in the high-efficiency fusion feature is matched with a standard design template to calculate the spatial deviation of the actual line from the ideal line, and then the deviation value is mapped to a heat map, with a deeper color indicating a more serious deviation (such as red indicating serious deviation and green indicating normality) to obtain a guide path deviation heat map; the guide path center line is extracted from the guide path deviation heat map, and the deviation of each position relative to the design map is calculated to quantify the guide path topology inaccuracy, for example, a deviation of 0-20 μm is normal, a deviation of 21-40 μm is general inaccuracy, and a deviation of 41-60 μm is high inaccuracy; connected component analysis is performed on the film defect distribution map to identify the types of defects such as bubbles, wrinkles and missing, and calculate the specific defect area ratio, and then different weights are assigned according to the defect type and position (such as an edge defect weight > an internal defect), and finally the film defect degree is obtained by weighted summation, for example, 5 internal bubbles (total area ratio 3%, weight 0.8), 2 edge wrinkles (area ratio 1.5%, weight 1.2) and 1 center area film missing (area ratio 2%, weight 1.0) are identified in the film defect distribution map, then the film defect degree = 3% x 0.8 + 1.5% x 1.2 + 2% x 1.0 = 6.2%; set the weight coefficient (such as surface quality inaccuracy 0.4, guide path topology inaccuracy 0.4 and film defect degree 0.2), and calculate the final index by weighted summation formula: surface performance loss degree = surface quality inaccuracy x 0.4 + guide path topology inaccuracy x 0.4 + film defect degree x 0.2.

[0036] S2, construct a time-frequency heat map of the electrical data, calculate the electrical signal attenuation rate and the electrical signal distortion rate of the flexible circuit board by using the time-frequency heat map, and analyze the topology deviation degree of the performance of the corresponding electrical appliance based on the electrical signal attenuation rate and the electrical signal distortion rate.

[0037] The embodiment of the present application can present the changes of the electrical parameters (such as current, voltage and signal frequency) of the flexible circuit board with time and frequency in the form of a visual heat map by constructing a time-frequency heat map of the electrical data, which can quickly locate the electrical defects such as signal distortion and abnormal fluctuation, and intuitively reflect the circuit performance.

[0038] The time-frequency heat map refers to a visualization tool for presenting information of electrical data changing with time and frequency in the form of a two-dimensional heat image.

[0039] Optionally, the time-frequency heat map can be obtained by performing short-time Fourier transform or wavelet transform on the electrical data of the flexible circuit board to convert the time-domain signal into time-frequency domain data, and then mapping the energy value with color depth according to the signal intensity at different time points and frequencies, so as to construct a visualization map directly presenting the dynamic change of the electrical data.

[0040] Further, by using the time-frequency heat map, the electrical signal attenuation rate and the electrical signal distortion rate of the flexible circuit board can be calculated to quantitatively evaluate the loss degree of signal strength and the deformation of waveform of the circuit board in the signal transmission process, and accurately locate the electrical performance defects. For example, in the detection of a smart watch heart rate monitoring module, it is found that the signal attenuation rate of a certain guide path is 15% (standard <10%), and the distortion rate is 8% (standard <5%), which determines that there is impedance abnormality in the region and it needs to be repaired.

[0041] The electrical signal attenuation rate refers to the weakening proportion of the signal strength compared with the original input signal strength during the transmission of the electrical signal of the flexible circuit board, and the electrical signal distortion rate refers to the degree of deviation of the waveform of the electrical signal transmitted by the flexible circuit board from the original standard waveform.

[0042] As an embodiment of the present application, the electrical signal attenuation rate and the electrical signal distortion rate of the flexible circuit board are calculated by using the time-frequency heat map, which includes: extracting the fundamental wave energy curve of the time-frequency heat map, and determining the input energy and the output energy of the flexible circuit board based on the fundamental wave energy curve; calculating the electrical signal attenuation rate of the flexible circuit board by using the following formula based on the input energy and the output energy:

[0043] wherein, represents the electrical signal attenuation rate, represents the input energy, represents the output energy; equally interval sampling the signal waveform in the time-frequency heat map to obtain a sampling signal point; calculating the signal value of the sampling signal point; calculating the electrical signal distortion rate of the flexible circuit board by using the following formula based on the signal value:

[0044] wherein, denotes the electrical signal distortion rate, n denotes the number of sampling signal points, denotes the signal value of the i-th sampling signal point, denotes the ideal signal value of the i-th sampling signal point.

[0045] wherein the fundamental wave energy curve refers to a curve reflecting the trend of the energy of the fundamental wave component in the flexible circuit board transmission signal changing with time, and the signal waveform refers to the form of the electrical signal transmitted by the flexible circuit board in the time domain and the frequency domain, such as a sine wave and a square wave.

[0046] In the specific implementation process, the fundamental wave component in the time-frequency heat map is extracted by two-dimensional Fourier transform, a curve of the energy of the fundamental wave changing with time is constructed, and the energy integrals of the starting point (input end) and the ending point (output end) of the curve are calculated respectively as the input energy and the output energy; sampling is performed on the time domain axis of the time-frequency heat map at a fixed time interval Δt, for example, 100ps for a 1GHz signal, and 100 sampling signal points can be obtained within 10ns.

[0047] Further, the embodiment of the present application can convert the loss and deformation in the signal transmission process into a quantitative index by analyzing the topological deviation degree of the corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation rate and the electrical signal distortion rate, directly reflecting the deviation degree of the circuit topological structure from the design standard, and quickly positioning potential electrical performance problems. For example, in the detection of the flexible circuit board of the 5G communication module, it is found by calculation that the signal attenuation rate is too high and the distortion is obvious, and it is analyzed that the topological deviation degree is over-standard, indicating that the circuit has problems such as impedance mismatch or unreasonable line layout.

[0048] wherein the topological deviation degree refers to the deviation degree of the actual electrical topological structure of the flexible circuit board from the design topology, which is usually represented by graph theory characteristic parameters.

[0049] As an embodiment of the present application, based on the electrical signal attenuation rate and the electrical signal distortion rate, the topological deviation degree of the corresponding electrical performance of the flexible circuit board is analyzed, including: Using the electrical signal attenuation rate and the electrical signal distortion, impedance inversion calculation is performed on the flexible circuit board to obtain an equivalent impedance deviation degree and a nonlinear distortion coefficient; According to the equivalent impedance deviation degree and the nonlinear distortion coefficient, the electrical field distribution of the flexible circuit board is reconstructed to obtain an electrical abnormal field; Using the electrical abnormal field, a guide network diagram of the flexible circuit board is constructed; Using the guide network diagram, the topological deviation degree of the flexible circuit board is identified.

[0050] The equivalent impedance deviation degree refers to the relative deviation degree of the actual equivalent impedance value of the flexible circuit board and the design nominal value, the nonlinear distortion coefficient refers to the proportion of harmonic components caused by material nonlinearity (such as dielectric constant changing with field strength) or structural defects (such as wire oxidation) when the flexible circuit board transmits electrical signals, the electrical abnormal field refers to the abnormal area in the surface electric field distribution of the flexible circuit board reconstructed by electromagnetic field simulation that does not conform to the design standard, and the guide network diagram refers to a graph theory model constructed by taking the high field strength area in the electrical abnormal field as a node and the actual current path as an edge.

[0051] In the implementation process, the attenuation rate and distortion rate can be substituted into the impedance inversion algorithm based on the transmission line theory (such as Newton-Raphson iteration method) to solve the actual impedance value of each line segment, and then the relative deviation (such as ΔZ / Z0×100%) of the actual impedance and the design impedance is calculated as the equivalent impedance deviation. Then, through harmonic analysis, the proportion of 3rd and higher harmonic components is extracted as the nonlinear distortion coefficient; the impedance parameters are introduced into the electromagnetic field simulator (such as CST Studio) to reconstruct the electric field intensity distribution of the circuit board, and then compared with the standard electric field distribution to mark the abnormal field strength area (such as the area with field strength exceeding the design value ±10%, which should be set according to the actual application), and generate an electrical abnormal field heat map; the high field strength area in the electrical abnormal field is taken as a node, and the shortest path algorithm (such as Dijkstra algorithm) is used to connect adjacent nodes, and each edge is assigned a weight (such as resistance value, field strength gradient) to quantitatively represent the electrical connection strength between nodes, thereby constructing a guide network diagram reflecting the current path; the node degree centrality, clustering coefficient and average path length of the guide network diagram are extracted, and each parameter is quantitatively calculated through graph theory algorithm to obtain the actual network characteristic parameters. Then, the difference between the actual network and the standard network is standardized, and the deviation degree of each parameter is calculated by the Euclidean distance formula. Finally, the parameter weights are set differently, for example, the node degree centrality is given a weight of 0.4, the clustering coefficient is given a weight of 0.3, and the average path length is given a weight of 0.3. Finally, the weighted Euclidean distance is used to obtain the quantitative index representing the deviation degree of the topology structure, i.e. the topology deviation degree.

[0052] S3, constructing a mechanical performance parameter matrix of the flexible circuit board using the mechanical test data, analyzing the fatigue endurance limit and flexibility decay coefficient of the flexible circuit board using the mechanical performance parameter matrix, and analyzing the mechanical performance degradation of the flexible circuit board using the fatigue endurance limit and the flexibility decay coefficient.

[0053] The mechanical performance parameter matrix of the flexible circuit board is constructed by using the mechanical test data, so that the test data can be integrated into a structured matrix, the mechanical properties of the circuit board under bending, stretching and other working conditions are quantitatively represented, and standardized data support is provided for mechanical reliability evaluation.

[0054] The mechanical performance parameter matrix refers to a quantitative representation system in which mechanical property parameters of the flexible circuit board are organized in a matrix form.

[0055] As an embodiment of the present application, the mechanical performance parameter matrix of the flexible circuit board is constructed by using the mechanical test data, including: extracting test parameters in the mechanical test data; constructing the mechanical performance parameter matrix of the flexible circuit board by using the test parameters, wherein the mechanical performance parameter matrix can be represented by the following matrix:

[0056] wherein, the mechanical performance parameter matrix is represented by, a first test parameter of the elastic modulus in the test parameters is represented by, a first test parameter of the tensile strength in the test parameters is represented by, a first test parameter of the elongation at break in the test parameters is represented by, a first test parameter of the bending stiffness in the test parameters is represented by, an Nth test parameter of the elastic modulus in the test parameters is represented by, an Nth test parameter of the tensile strength in the test parameters is represented by, an Nth test parameter of the elongation at break in the test parameters is represented by, an Nth test parameter of the bending stiffness in the test parameters is represented by.

[0057] The elastic modulus refers to a physical quantity of the ability of a material to resist elastic deformation, that is, the ratio of stress to strain in the elastic deformation stage of the material, the elongation at break refers to the percentage of the elongation of the material at the time of tensile fracture to the original length, the tensile strength refers to the maximum tensile stress that can be borne by the material during the tensile process, that is, the ultimate tension borne by the unit cross-sectional area before fracture, and the bending stiffness refers to a physical quantity of the ability of a structure or material to resist bending deformation.

[0058] The fatigue endurance limit value and the flexibility decay coefficient of the flexible circuit board are analyzed by using the mechanical performance parameter matrix, so that the life threshold and the performance decay speed of the circuit board under repeated bending and other mechanical stresses can be quantitatively evaluated, and data support is provided for mechanical performance quality evaluation of the flexible circuit board.

[0059] It should be further noted that the mechanical property parameter matrix not only contains parameters such as elastic modulus, elongation at break, and bending stiffness, but also contains parameters such as fatigue life, damping factor, Poisson's ratio, and storage modulus, as well as test loading rate and test temperature.

[0060] The fatigue endurance limit refers to the maximum number of cycles or time threshold that the flexible circuit board can maintain its functional integrity under specific mechanical load conditions (such as bending and stretching); the flexibility decay coefficient refers to the quantitative characterization of the degree of decay of flexibility (such as bending stiffness and elastic deformation ability) with the number of uses under the action of mechanical load cycles.

[0061] As an embodiment of the present application, the fatigue endurance limit and flexibility decay coefficient of the flexible circuit board are analyzed using the mechanical property parameter matrix, including: The dynamic stress-strain characteristics of the flexible circuit board are identified using the mechanical property parameter matrix; Based on the dynamic stress-strain characteristics, the stress-life characteristic set and the elastic modulus decay sequence of the flexible circuit board are constructed; The material fatigue curve of the flexible circuit board is constructed using the stress-life characteristic set; The fatigue endurance limit of the flexible circuit board is calculated based on the material fatigue curve; Based on the modulus decay sequence, the flexible circuit board is subjected to exponential decay fitting to obtain the flexibility decay coefficient.

[0062] The dynamic stress-strain characteristics refer to the stress-strain response characteristics of the material under dynamic load, the stress-life characteristic set refers to a group of fatigue life data points representing the material under different stress amplitudes, usually in the form of discrete characteristic points of the "stress amplitude-failure cycle number" (S-N) curve, the elastic modulus decay sequence refers to the change sequence of the elastic modulus of the material with the number of loads under cyclic load, and the material fatigue curve is a mathematical curve used to describe the relationship between stress amplitude and fatigue life of the material under alternating stress.

[0063] In the implementation process, the data of stress, strain, damping factor and storage modulus under different loading speeds and temperature conditions can be extracted from the mechanical property parameter matrix, the response of the flexible circuit board under dynamic load is tested by using a dynamic mechanical analyzer, a stress-strain curve is drawn, the hysteresis loop area and slope change of the curve are analyzed, and the dynamic stress-strain characteristics are obtained; according to the fatigue life parameters in the matrix under different stress amplitudes, the stress amplitude and the corresponding failure cycle number data pair are arranged to form a stress-life characteristic set, and the elastic modulus data after each interval of a certain number of bending times are extracted from the matrix and arranged in order to form an elastic modulus attenuation sequence; the stress amplitude and the failure cycle number data in the stress-life characteristic set are nonlinearly fitted by using the Basquin formula, the formula parameters are adjusted by using a mathematical calculation software (such as Origin) to make the fitting curve as close to the data points as possible, and the material fatigue curve is obtained; the stress amplitude working condition in the actual use of the flexible circuit board is determined, the stress value is substituted into the constructed material fatigue curve equation, and the corresponding failure cycle number is calculated, which is the fatigue endurance limit of the flexible circuit board under this working condition; the data in the elastic modulus attenuation sequence are fitted by using an exponential decay function, the function parameters are determined by using an optimization algorithm such as the least square method, and the exponential equation of the elastic modulus changing with the bending times is obtained, and the value reflecting the attenuation speed of the elastic modulus is extracted from the equation parameters to quantify the flexibility degradation coefficient.

[0064] By using the fatigue endurance limit and the flexibility degradation coefficient, the mechanical performance degradation of the flexible circuit board is analyzed, and the attenuation degree of key performance indicators such as tensile strength and elastic modulus of the circuit board relative to the initial value after a certain number of bending times is quantified, so that the influence of mechanical stress on the performance of the circuit board is intuitively evaluated.

[0065] The mechanical performance degradation refers to the decrease of the original mechanical performance (such as strength, elastic modulus, plasticity, hardness, etc.) of the material or component relative to the initial state due to the action of load, environmental factors (such as temperature, humidity, corrosion) or fatigue damage during use.

[0066] As an embodiment of the present application, the mechanical performance degradation of the flexible circuit board is analyzed by using the fatigue endurance limit and the flexibility degradation coefficient, including: The fatigue endurance limit and the flexibility degradation coefficient are normalized to obtain a standardized life index and a standardized degradation rate; The standardized life index and the standardized degradation rate are weighted and fused to obtain an initial performance degradation; The initial performance degradation is modified by environmental compensation to obtain a modified performance degradation; The modified performance degradation is hierarchically mapped to obtain a mechanical performance degradation.

[0067] In the implementation process, the fatigue endurance limit and the flexibility degradation coefficient can be normalized by using a normalization function; the standardized life index and the standardized degradation rate can be assigned weight coefficients according to the application scenario of the flexible circuit board, and then a weighted sum is calculated to obtain an initial performance degradation; according to the actual use environment (temperature T, humidity H, and vibration frequency f, etc.), an environmental influence factor matrix F=[fT, fH, ff, ] is constructed, and a modified performance degradation is calculated by a function P=P0 ∏ifi; the modified performance degradation P is mapped to a preset performance grade interval (for example, P∈[0,0.3] corresponds to'slight degradation', P∈[0.3,0.7] corresponds to'moderate degradation', and P∈[0.7,1] corresponds to'severe degradation') to determine the final mechanical performance degradation grade.

[0068] S4, according to the surface performance loss degree, the topological deviation degree and the mechanical performance degradation, the flexible circuit board is comprehensively analyzed to obtain a quality detection report.

[0069] The embodiment of the application can intuitively reflect the overall quality of the product in terms of electrical performance, structural rationality and mechanical reliability by comprehensively analyzing the flexible circuit board according to the surface performance loss degree, the topological deviation degree and the mechanical performance degradation to obtain a quality detection report, and finally generating a quality detection report to provide data support for product improvement and quality control. For example, in the detection of a certain foldable screen mobile phone FPC, the surface performance loss degree shows that the signal attenuation is out of standard, the topological deviation degree indicates that the key node connection is abnormal, and the mechanical performance degradation prompts that the stiffness change after bending is too large. After comprehensive analysis, the report clearly points out that the circuit layout needs to be optimized and the material toughness needs to be improved to improve the product quality.

[0070] The quality detection report refers to a systematic quality evaluation file formed based on the comprehensive analysis of the multi-dimensional quantitative data of the surface performance loss degree, the topological deviation degree and the mechanical performance degradation of the flexible circuit board.

[0071] As an embodiment of the application, according to the surface performance loss degree, the topological deviation degree and the mechanical performance degradation, the flexible circuit board is comprehensively analyzed to obtain a quality detection report, which includes: The surface performance loss degree, the topological deviation degree and the mechanical performance degradation are comprehensively processed to obtain a comprehensive data set; The comprehensive data set is used to construct an interaction influence matrix of the flexible circuit board; Based on the interaction influence matrix, a defect coupling coefficient of the flexible circuit board is calculated; The defect coupling coefficient is dynamically weighted and corrected to obtain a corrected coupling coefficient; Based on the corrected coupling coefficient, a comprehensive quality assessment of the flexible circuit board is performed; Based on the assessment result of the comprehensive quality assessment, a quality detection report of the flexible circuit board is constructed.

[0072] The interaction influence matrix is a square matrix based on a comprehensive data set, which is used to intuitively present the correlation strength of each quality indicator of the flexible circuit board. The defect coupling coefficient is a quantitative indicator obtained by eigenvalue decomposition or principal component analysis of the interaction influence matrix, which is used to represent the interaction degree between multiple defects of the flexible circuit board.

[0073] In the specific implementation process, the surface performance loss defect degree, topological deviation degree and mechanical performance reduction degree can be uniformly converted to the 0-1 interval to eliminate the dimension influence, and then the processed data is combined to form a comprehensive data set containing multiple indicators. Pearson correlation analysis is performed on the comprehensive data set to calculate the correlation coefficient between the indicators, which is filled in the corresponding position of the 3×3 matrix. For example, the measured data of a certain flexible circuit board is: surface performance loss defect degree (12%), topological deviation degree (0.25), and mechanical performance reduction degree (18 MPa). After calculation, the correlation coefficient between the surface performance loss defect degree and the topological deviation degree is 0.6, and the correlation coefficient between the surface performance loss defect degree and the mechanical performance reduction degree is 0.4. The correlation coefficient between the topological deviation degree and the mechanical performance reduction degree is 0.3. The following matrix can be used to represent it: Interaction influence matrix

[0074] Further, the interaction influence matrix can be subjected to principal component analysis (PCA), and main components with a cumulative contribution rate of more than 85% are extracted, and the defect coupling coefficient is calculated by weighting the variance contribution rate. For example, after PCA analysis, two main components are extracted, and the variance contribution rates are 55% and 30%, respectively. The defect coupling coefficient is 0.72, which is obtained by combining the main component load vector with the interaction influence matrix. For high-temperature application scenarios (such as automobile engine compartments, 85°C), the weight is adjusted using the particle swarm optimization algorithm (PSO). The mechanical performance degradation weight is increased from 0.3 to 0.5, and the surface performance loss defect weight is reduced from 0.4 to 0.3. The coupling coefficient is recalculated. The corrected coupling coefficient is compared with the preset standard: excellent (<0.5), good (0.5-0.65), qualified (0.65-0.8), and unqualified (>0.8) to determine the quality level. The original data, processing results, and corrected coupling coefficient of the surface performance loss defect, topological deviation, and mechanical performance degradation are first presented in tabular form, and then the interaction influence relationship is visually displayed using charts. Combined with the quality level judgment, detailed analysis conclusions and improvement suggestions are written, and finally a complete quality detection report is formed.

[0075] Embodiment 2: As Figure 2 shown, it is a functional module diagram of the flexible circuit board quality detection system based on data analysis of the application.

[0076] The flexible circuit board quality detection system based on data analysis 200 can be installed in an electronic device. According to the realized functions, the flexible circuit board quality detection system based on data analysis can include a surface quality analysis module 201, an electrical quality analysis module 202, a mechanical quality analysis module 203, and a comprehensive quality analysis module 204. The modules of the application can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can complete a fixed function, and are stored in the memory of the electronic device.

[0077] In the embodiments of the application, the functions of each module / unit are as follows: The surface quality analysis module 201 is used to collect high-definition image data, electrical data, and mechanical test data of the flexible circuit board, perform double-branch feature fusion on the high-definition image data, obtain efficient fusion features, analyze the surface quality deviation, circuit topology deviation, and film defect degree of the flexible circuit board using the efficient fusion features, and determine the surface performance loss defect degree of the flexible circuit board. The electrical quality analysis module 202 is configured to construct a time-frequency heat map of the electrical data, calculate an electrical signal attenuation rate and an electrical signal distortion rate of the flexible circuit board by using the time-frequency heat map, and analyze a topological deviation degree of the corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation rate and the electrical signal distortion rate. The mechanical quality analysis module 203 is configured to construct a mechanical performance parameter matrix of the flexible circuit board by using the mechanical test data, analyze a fatigue endurance limit value and a flexibility recession coefficient of the flexible circuit board by using the mechanical performance parameter matrix, and analyze a mechanical performance degradation of the flexible circuit board by using the fatigue endurance limit value and the flexibility recession coefficient. The comprehensive quality analysis module 204 is configured to perform comprehensive quality analysis on the flexible circuit board according to the surface performance loss degree, the topological deviation degree and the mechanical performance degradation, and obtain a quality detection report.

[0078] In detail, the modules in the flexible circuit board quality detection system 200 based on data analysis in the embodiments of the present application adopt the same technical means as the flexible circuit board quality detection method based on data analysis in the above-mentioned Figure 1 , and can produce the same technical effects, which will not be described here again.

[0079] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application.

Claims

1. A flexible circuit board quality detection method based on data analysis, characterized in that, The method comprises: Collecting high-definition image data, electrical data and mechanical test data of the flexible circuit board, performing double-branch feature aggregation on the high-definition image data to obtain efficient fusion features, and using the efficient fusion features to analyze surface shape misalignment degree, guide path topology misalignment degree and film defect degree of the flexible circuit board to determine surface performance loss degree of the flexible circuit board; Constructing a time-frequency heat map of the electrical data, using the time-frequency heat map to calculate electrical signal attenuation rate and electrical signal distortion rate of the flexible circuit board, and analyzing topology deviation degree of corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation rate and the electrical signal distortion rate; Using the mechanical test data to construct a mechanical performance parameter matrix of the flexible circuit board, using the mechanical performance parameter matrix to analyze fatigue endurance limit and flexibility recession coefficient of the flexible circuit board, and using the fatigue endurance limit and the flexibility recession coefficient to analyze mechanical performance reduction of the flexible circuit board; According to the surface performance loss degree, the topology deviation degree and the mechanical performance reduction, performing comprehensive quality analysis on the flexible circuit board to obtain a quality detection report.

2. The data analysis-based flexible circuit board quality inspection method of claim 1, wherein, Performing double-branch feature aggregation on the high-definition image data to obtain efficient fusion features, comprising: Extracting double-branch features of the high-definition image data to obtain local detail features and global structure features; Performing cross-branch space alignment on the local detail features and the global structure features to obtain aligned double-branch features; Performing multi-modal fusion on the aligned double-branch features to obtain preliminary fusion features; Performing dimension reduction optimization on the preliminary fusion features to obtain efficient fusion features.

3. The data analysis-based flexible circuit board quality inspection method of claim 1, wherein, Using the efficient fusion features to analyze surface shape misalignment degree, guide path topology misalignment degree and film defect degree of the flexible circuit board to determine surface performance loss degree of the flexible circuit board, comprising: Using the efficient fusion features to perform defect region segmentation on the flexible circuit board to obtain surface defect distribution map and film defect distribution map; Performing geometric parameter quantization on the surface defect distribution map to obtain surface shape misalignment degree; Using the efficient fusion features to perform guide path topology registration on the flexible circuit board to obtain guide path offset heat map; Performing spatial error statistics on the guide path offset heat map to obtain guide path topology misalignment degree; Using the film defect distribution map to analyze film defect degree of the flexible circuit board; Performing weighted fusion on the surface shape misalignment degree, the guide path topology misalignment degree and the film defect degree to obtain surface performance loss degree.

4. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, Using the time-frequency heat map to calculate electrical signal attenuation rate and electrical signal distortion rate of the flexible circuit board, comprising: Extracting fundamental energy curve of the time-frequency heat map, and determining input energy and output energy of the flexible circuit board based on the fundamental energy curve; Based on the input energy and the output energy, using the following formula to calculate the electrical signal attenuation rate of the flexible circuit board: wherein, represents the rate of electrical signal decay, represents the input energy, represents the output energy; Performing equal-interval sampling on the signal waveform in the time-frequency heat map to obtain sampling signal points; Calculating signal values of the sampling signal points; Based on the signal value, the electrical signal distortion rate of the flexible circuit board is calculated by using the following formula: wherein denotes the electrical signal distortion rate, n denotes the number of sampling signal points, denotes the signal value of the i-th sampling signal point, denotes the ideal signal value of the i-th sampling signal point.

5. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, Based on the electrical signal attenuation rate and the electrical signal distortion rate, the topological deviation degree of the corresponding electrical performance of the flexible circuit board is analyzed, including: Using the electrical signal attenuation rate and the electrical signal distortion, the impedance inversion calculation of the flexible circuit board is performed to obtain the equivalent impedance deviation degree and the nonlinear distortion coefficient; According to the equivalent impedance deviation degree and the nonlinear distortion coefficient, the electrical field distribution reconstruction of the flexible circuit board is performed to obtain the electrical abnormal field; Using the electrical abnormal field, the guide network diagram of the flexible circuit board is constructed; Using the guide network diagram, the topological deviation degree of the flexible circuit board is identified.

6. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, Using the mechanical test data, the mechanical performance parameter matrix of the flexible circuit board is constructed, including: Extracting the test parameters in the mechanical test data; Using the test parameters, the mechanical performance parameter matrix of the flexible circuit board is constructed, wherein the mechanical performance parameter matrix can be represented by using the following matrix: wherein represents a matrix of mechanical performance parameters, represents a first test parameter of the test parameters, the elastic modulus, represents a first test parameter of the test parameters, the tensile strength, represents a first test parameter of the test parameters, the elongation at break, represents a first test parameter of the test parameters, the bending stiffness, represents an Nth test parameter of the test parameters, the elastic modulus, represents an Nth test parameter of the test parameters, the tensile strength, represents an Nth test parameter of the test parameters, the elongation at break, represents an Nth test parameter of the test parameters, the bending stiffness.

7. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, Using the mechanical performance parameter matrix, the fatigue endurance limit value and the flexibility recession coefficient of the flexible circuit board are analyzed, including: Using the mechanical performance parameter matrix, the dynamic stress-strain characteristics of the flexible circuit board are identified; Based on the dynamic stress-strain characteristics, the stress-life characteristic set and the elastic modulus attenuation sequence of the flexible circuit board are constructed; Using the stress-life characteristic set, the material fatigue curve of the flexible circuit board is constructed; Based on the material fatigue curve, the fatigue endurance limit value of the flexible circuit board is calculated; Based on the modulus attenuation sequence, the exponential attenuation fitting of the flexible circuit board is performed to obtain the flexibility recession coefficient.

8. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, Using the fatigue endurance limit value and the flexibility recession coefficient, the mechanical performance degradation of the flexible circuit board is analyzed, including: The fatigue endurance limit value and the flexibility recession coefficient are normalized to obtain the standardized life index and the standardized recession rate; The standardized life index and the standardized recession rate are weighted and fused to obtain the initial performance degradation; The initial performance degradation is environment compensated and corrected to obtain the corrected performance degradation; The corrected performance degradation is classified and mapped to obtain the mechanical performance degradation.

9. The data analysis based flexible circuit board quality inspection method of claim 1, wherein, According to the surface performance damage degree, the topological deviation degree and the mechanical performance degradation, the comprehensive quality analysis of the flexible circuit board is performed to obtain the quality detection report, including: The surface performance damage degree, the topological deviation degree and the mechanical performance degradation are data-comprehensively processed to obtain a comprehensive data set; Using the comprehensive data set, the interaction influence matrix of the flexible circuit board is constructed; Based on the interaction influence matrix, the defect coupling coefficient of the flexible circuit board is calculated; The defect coupling coefficient is dynamically weighted and corrected to obtain a corrected coupling coefficient; Based on the corrected coupling coefficient, the comprehensive quality evaluation of the flexible circuit board is performed; Based on the evaluation result of the comprehensive quality evaluation, the quality detection report of the flexible circuit board is constructed.

10. A flexible circuit board quality inspection system based on data analysis, characterized by, The system includes: The surface quality analysis module is configured to collect high-definition image data, electrical data and mechanical test data of the flexible circuit board, perform double-branch feature fusion on the high-definition image data to obtain high-efficiency fusion features, analyze surface performance loss and defect degrees of the flexible circuit board by using the high-efficiency fusion features, and determine surface performance loss and defect degrees of the flexible circuit board. The electrical quality analysis module is configured to construct a time-frequency heat map of the electrical data, calculate electrical signal attenuation and distortion rates of the flexible circuit board by using the time-frequency heat map, and analyze topological deviation degrees of corresponding electrical performance of the flexible circuit board based on the electrical signal attenuation and distortion rates. The mechanical quality analysis module is configured to construct a mechanical performance parameter matrix of the flexible circuit board by using the mechanical test data, analyze fatigue endurance limit values and flexibility recession coefficients of the flexible circuit board by using the mechanical performance parameter matrix, and analyze mechanical performance reduction of the flexible circuit board by using the fatigue endurance limit values and the flexibility recession coefficients. The comprehensive quality analysis module is configured to perform comprehensive quality analysis on the flexible circuit board according to the surface performance loss and defect degrees, the topological deviation degrees and the mechanical performance reduction, and obtain a quality detection report.

Images