Hydrophilic aluminum foil surface hydrophilic layer defect flaw detection system

Abstract

The invention discloses a hydrophilic aluminum foil surface hydrophilic layer defect flaw detection system, and relates to the technical field of hydrophilic aluminum foil detection. The use method of the system comprises the following steps: performing targeted excitation on the hydrophilic aluminum foil to be detected to obtain an excitation interaction signal; and inputting the excitation interaction signal into a feature fusion model, outputting a feature recognition result by the feature fusion model, and determining a defect type according to the feature recognition result. Through a progressive targeting excitation mode of infrared preheating, ultrasonic activation and electric field strengthening, defect signals can be comprehensively captured from four dimensions of interface bonding force, insulativity, thermal conductivity and surface energy, the defect detection range is wide, meanwhile, precise recognition of complex planar defects is achieved, the situations of missed judgment and misjudgment of the defects are effectively reduced, and the defect detection efficiency is improved. And the quantitative judgment of the defect grade is realized by the association of the equivalent size, the actual defect area and the maximum deviation value of the characteristic parameters, so that accurate data support is provided for the optimization of the production process of the hydrophilic aluminum foil.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydrophilic aluminum foil detection, in particular to a hydrophilic aluminum foil surface hydrophilic layer defect detection system. BACKGROUND

[0002] Hydrophilic aluminum foil is a hydrophilic treatment of aluminum foil, which covers a layer of hydrophilic layer on its surface through special process treatment. Condensed water on the hydrophilic aluminum foil will quickly spread out and will not condense into water droplets, thus increasing the heat exchange area and accelerating the refrigeration and heating speed, and also effectively avoiding the noise caused by condensed water hindering air flow. Hydrophilic aluminum foil is widely used in heat exchangers of air conditioners and refrigeration equipment. Once the hydrophilic layer of the hydrophilic aluminum foil has defects such as pinholes, micro-cracks, local peeling, damage or hydrophilic decay, it will directly lead to the decrease of heat exchange efficiency of the heat exchanger and the acceleration of corrosion of the aluminum foil substrate, thus shortening the service life of the equipment and even causing operation failure. Therefore, it is crucial to accurately, non-destructively and industrially detect the defects of the hydrophilic layer. At present, the defect detection method of the hydrophilic layer on the surface of the hydrophilic aluminum foil cannot meet the requirements of non-destructive, high-precision, fast response and quantifiable industrial accurate detection, mainly because of the following defects:

[0003] 1. Insufficient defect recognition sensitivity: The existing detection technology mainly uses single ultrasonic, infrared or electric field excitation method. Single excitation can only capture the characteristics of the hydrophilic layer in one dimension, such as ultrasonic detection interface bonding and infrared detection thermal conductivity. It cannot comprehensively cover the multi-dimensional characteristics of the hydrophilic layer such as insulation, thermal conductivity, interface bonding and surface energy, resulting in low recognition sensitivity to small planar defects such as pinholes below 50 microns and linear micro-cracks, and easy to miss or misjudge the defects, and difficult to distinguish the defect types.

[0004] 2. Insufficient defect quantification capability: The existing detection technology mainly uses general signal parameters such as single vibration amplitude and temperature value to determine defects, and the defect grade mainly depends on manual experience, so the defect grade determination standard is fuzzy, which cannot provide accurate data support for production process optimization, and is out of touch with the actual use demand of industrial quantitative detection. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a hydrophilic aluminum foil surface hydrophilic layer defect detection system, which solves the problems of insufficient defect recognition sensitivity and insufficient defect quantification capability of the current hydrophilic aluminum foil surface hydrophilic layer defect detection method.

[0006] To achieve the above purpose, the present application is realized by the following technical scheme: a hydrophilic aluminum foil surface hydrophilic layer defect detection system, the use method of the system comprising the following steps:

[0007] Targeted excitation is performed on the hydrophilic aluminum foil to be detected to obtain excitation interaction signals;

[0008] The excitation interaction signal is input into a feature fusion model, the feature fusion model outputs a feature recognition result, and the feature recognition result is used to determine a defect type.

[0009] The present application further provides that the targeted excitation of the hydrophilic aluminum foil to be detected is performed in the order of infrared preheating, ultrasonic activation and electric field strengthening.

[0010] The present application further provides that, before the excitation interaction signal is input into the feature fusion model, the excitation interaction signal is subjected to denoising, substrate signal elimination and normalization processing to obtain a purified excitation interaction signal.

[0011] Feature parameters are extracted according to the purified excitation interaction signal, and the feature parameters include ultrasonic hydrophilic interface combined vibration deviation, electric field hydrophilic insulation distortion coefficient, infrared hydrophilic heat conduction gradient difference, hydrophilic angle deviation, ultrasonic frequency offset, electric field current mutation rate, infrared thermal response lag time, vibration signal variance, electric field signal variance, thermal response signal variance, hydrophilic angle stability and interaction signal synergy coefficient.

[0012] The present application further provides that the feature fusion model includes an input layer, an attention layer, a CNN feature extraction layer, an LSTM time sequence mining layer and a full connection layer.

[0013] The input layer is used to input all the feature parameters.

[0014] The attention layer is used to assign weights to each feature parameter.

[0015] The CNN feature extraction layer is used to extract defect spatial features.

[0016] The LSTM time sequence mining layer is used to mine time sequence features of all the feature parameters.

[0017] The full connection layer is used to fuse the defect spatial features and the time sequence features, and output a fusion feature value as the feature recognition result.

[0018] The present application further provides that the process of determining the defect type according to the feature recognition result is as follows:

[0019] A preset feature alarm threshold is set, and it is determined whether the fusion feature value exceeds the feature alarm threshold, if yes, the defect type is determined, and if not, it is determined that the hydrophilic aluminum foil is defect-free.

[0020] The present application further provides that the determination logic of the defect type is as follows:

[0021] When the electric field hydrophilic insulation distortion coefficient is greater than or equal to 2.0, the absolute value of the hydrophilic angle deviation is greater than 5°, and the absolute value of the infrared hydrophilic heat conduction gradient difference is greater than or equal to 1℃, it is determined that a point surface defect occurs.

[0022] When the absolute value of the ultrasonic frequency deviation rate is ≥5%, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is ≥30%, and the vibration signal variance is ≥0.05, a linear-surface defect is determined to have occurred.

[0023] A sheet-like surface defect is determined to occur when the absolute value of the ultrasonic frequency deviation rate is ≥30%, the infrared thermal response hysteresis time is ≥0.5s, and the interaction signal coordination coefficient is ≤0.3.

[0024] When the electric field hydrophilic insulation distortion coefficient is ≥3.0, the absolute value of the infrared hydrophilic thermal conductivity gradient difference is ≥2℃, and the absolute value of the hydrophilic angle deviation is >10°, an irregular surface defect is determined to have occurred.

[0025] When the absolute value of the hydrophilic angle deviation is >5°, the hydrophilic angle stability is ≥2°, the electric field hydrophilic insulation distortion coefficient is <2.0, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is <30%, and the absolute value of the ultrasonic frequency offset rate is <5%, a regional surface defect is determined to have occurred.

[0026] The present invention is further configured such that the system includes:

[0027] An infrared preheating excitation module is used to preheat the hydrophilic aluminum foil to be tested using infrared technology.

[0028] An ultrasonic activation excitation module is used to ultrasonically activate the hydrophilic aluminum foil to be tested.

[0029] An electric field enhancement excitation module is used to enhance the electric field of the hydrophilic aluminum foil to be tested.

[0030] The excitation interaction acquisition and analysis module is used to acquire excitation interaction signals after infrared preheating, ultrasonic activation, and electric field enhancement excitation, and extract feature parameters.

[0031] A feature fusion model service module is used to calculate fused feature values ​​based on feature parameters.

[0032] A defect type determination module is used to determine the defect type of the hydrophilic aluminum foil to be tested based on the fusion feature value.

[0033] The defect level determination module is used to determine the defect level based on the defect type determination result.

[0034] The present invention is further configured such that: before determining the defect level based on the defect type determination result, the defect location is located based on the coordinates of the probe used to collect the excitation interaction signal and the peak value of the characteristic parameters.

[0035] The present invention is further configured such that: when determining the defect level based on the defect type determination result, the determination logic is as follows:

[0036] The equivalent size at the defect location is <50μm, and the defect area is <1962.5μm. 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency deviation rate) is less than 30%, it is judged as a first-level defect.

[0037] The equivalent size at the defect location is less than 200 μm and 1962.5 μm within the range of 50 μm to 1962.5 μm. 2 ≤Defect area <31400μm 2 When 30%≤max (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, absolute value of ultrasonic frequency deviation rate) <50%, it is judged as a level two defect;

[0038] The equivalent size at the defect location is ≥200μm and the defect area is ≥31400μm. 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency deviation rate) is ≥50%, it is judged as a level three defect.

[0039] The present invention is further configured to: issue a yellow alarm and trigger a marking command when the defect is determined to be a level two defect;

[0040] When a defect is determined to be a Level 3 defect, a red alarm is issued and a shutdown command is triggered.

[0041] This invention provides a flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil. It has the following beneficial effects:

[0042] (1) This invention obtains diverse excitation interaction signals through a progressive targeted excitation method of infrared preheating, ultrasonic activation and electric field enhancement. When extracting the characteristic parameters of the associated hydrophilic layer, it comprehensively captures defect signals from four dimensions: interface bonding force, insulation, thermal conductivity and surface energy. While the defect detection scope is wide, it can achieve accurate identification of complex planar defects and effectively reduce the situation of missed or misjudged defects.

[0043] (2) This invention uses the diameter of the outer circle of the planar defect as the equivalent size, synchronously associates the actual area of ​​the defect, and realizes the quantitative judgment of the defect level through the maximum deviation of the characteristic parameters. While replacing the existing fuzzy judgment method that relies on human experience to determine the defect level, it provides accurate data support for the optimization of the production process of hydrophilic aluminum foil. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0045] Figure 2 This is a system architecture block diagram of the present invention. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0047] Please see Figures 1-2 The present invention provides the following technical solutions:

[0048] Example 1

[0049] A method for detecting defects in the hydrophilic layer on the surface of hydrophilic aluminum foil, the method comprising the following steps:

[0050] The hydrophilic aluminum foil to be tested was subjected to triple-targeted excitation to obtain excitation interaction signals; the triple-targeted excitation included infrared preheating, ultrasonic activation, and electric field enhancement.

[0051] The excitation interaction signal is input into the feature fusion model, the feature fusion model outputs the feature recognition result, and the defect type is determined based on the feature recognition result.

[0052] Therefore, by using a progressive targeted excitation method of infrared preheating, ultrasonic activation and electric field enhancement, diverse excitation interaction signals can be obtained. This method can comprehensively capture defect signals from four dimensions: interfacial bonding force, insulation, thermal conductivity and surface energy. While providing a wide range of defect detection, it can also accurately identify complex planar defects, effectively reducing the occurrence of missed or false defects.

[0053] In one embodiment, before targeted excitation of the hydrophilic aluminum foil to be tested, the hydrophilic aluminum foil is subjected to planarization and surface cleaning treatment, and a benchmark database is established. The process of establishing the benchmark database includes:

[0054] Obtain a confirmed defect-free hydrophilic aluminum foil sample (confirmed by microscopy that the sample is free of defects such as pinholes, cracks, and peeling), and collect the core parameters of the hydrophilic aluminum foil sample, such as the aluminum foil thickness, hydrophilic layer material, and hydrophilic layer thickness.

[0055] The hydrophilic aluminum foil sample was targeted and stimulated according to the stimulation parameters adapted to it (determined based on the core parameters) to obtain the stimulation interaction signals of the left, middle and right regions of the hydrophilic aluminum foil sample.

[0056] After performing noise reduction, substrate signal removal and normalization on the excitation interaction signal in sequence, the mean value of the interaction signal in each region is calculated as the reference signal value of the hydrophilic aluminum foil sample.

[0057] After establishing the correlation between core parameters, excitation parameters, and reference signal values, the reference database is completed. The correlation between core parameters and excitation parameters is shown in Table 1.

[0058]

[0059] Table 1

[0060] In one embodiment, the process of targeted excitation of the hydrophilic aluminum foil to be tested is as follows:

[0061] Obtain the core parameters of the hydrophilic aluminum foil to be tested, and obtain the excitation parameters of the hydrophilic aluminum foil to be tested from the benchmark database;

[0062] Based on the excitation parameters, the hydrophilic aluminum foil to be tested is subjected to progressive targeted excitation in the order of infrared preheating, ultrasonic activation, and electric field enhancement.

[0063] Based on this, excitation interaction signals are obtained, including vibration interaction signals, electric field interaction signals, thermal response interaction signals, and hydrophilic interaction signals of the surface of the hydrophilic aluminum foil to be tested.

[0064] In one embodiment, before inputting the excitation interaction signal into the feature fusion model, the db4 wavelet filtering denoising algorithm is used to denoise the vibration interaction signal, electric field interaction signal, thermal response interaction signal and hydrophilic interaction signal, and then the substrate signal is removed and normalized to obtain the purified excitation interaction signal.

[0065] Twelve feature parameters were extracted from the purified excitation interaction signal. The feature parameters are as follows:

[0066] The ultrasonic hydrophilic interface bonding vibration deviation represents the vibration parameter of the bonding surface between the hydrophilic layer and the aluminum foil substrate under ultrasonic activation excitation. It is used to reflect the tightness of the bonding between the hydrophilic layer and the aluminum foil substrate. Its calculation formula is: ultrasonic hydrophilic interface bonding vibration deviation = (measured vibration amplitude - reference vibration amplitude) / reference vibration amplitude.

[0067] The electric field hydrophilic insulation distortion coefficient is used to reflect the insulation integrity of the hydrophilic layer under electric field enhancement excitation. Its calculation formula is: electric field hydrophilic insulation distortion coefficient = measured electric field intensity gradient / reference electric field intensity gradient;

[0068] The infrared hydrophilic thermal conductivity gradient difference is used to reflect the thermal conductivity continuity of the hydrophilic layer under infrared preheating excitation. Its calculation formula is: Infrared hydrophilic thermal conductivity gradient difference = measured hot spot temperature difference - reference temperature difference.

[0069] The hydrophilicity angle deviation is used to judge the quality of hydrophilicity of the hydrophilic layer. Its calculation formula is: hydrophilicity angle deviation = measured hydrophilicity angle - reference hydrophilicity angle;

[0070] Ultrasonic frequency offset is used to capture fine structural damage in the hydrophilic layer. Its calculation formula is: ultrasonic frequency offset = (measured ultrasonic frequency - reference ultrasonic frequency) / reference ultrasonic frequency.

[0071] The electric field current mutation rate is used to amplify the signal of minute damage to the hydrophilic layer. Its calculation formula is: Electric field current mutation rate = (Measured electric field current - Reference electric field current) / Reference electric field current;

[0072] Infrared thermal response lag time is used to detect the structural integrity of the hydrophilic layer. Its calculation formula is: electric field current mutation rate = measured thermal response lag time - reference thermal response lag time;

[0073] Vibration signal variance is used to determine the stability of vibration signals. It is obtained by measuring the variance of vibration amplitude data within the acquisition period.

[0074] The variance of the electric field signal is used to capture the fluctuation characteristics of the electric field signal. It is obtained by measuring the variance of the electric field intensity data within the acquisition period.

[0075] The variance of the thermal response signal is used to reflect the uniformity of temperature distribution. It is obtained by measuring the variance of temperature data within the acquisition period.

[0076] Hydrophilic angle stability is used to determine the consistency of the surface energy of the hydrophilic layer. It is obtained by measuring the standard deviation of the hydrophilic angle in at least 5 consecutive measurements.

[0077] The interaction signal coordination coefficient is used to verify the correlation between the corresponding excitation interaction signals of infrared preheating, ultrasonic activation, and electric field enhancement. This interaction signal coordination coefficient is the consistency coefficient of the change trend of the three interaction signals, and its value range is [0, 1].

[0078] Furthermore, a feature fusion model was built based on the TensorFlow 2.1 framework. This feature fusion model includes an input layer, an attention layer, a CNN feature extraction layer, an LSTM temporal mining layer, and a fully connected layer.

[0079] The input layer is used to input twelve feature parameters;

[0080] The attention layer is used to assign weights to each feature parameter;

[0081] The CNN feature extraction layer uses 3 convolutional layers and 2 pooling layers, with 3×3 convolutional kernels and a stride of 1, and 2×2 pooling kernels. This CNN feature extraction layer is used to extract spatial features of defects.

[0082] The LSTM temporal mining layer uses two hidden layers, each with 128 neurons, to mine temporal features of twelve feature parameters.

[0083] The fully connected layer is used to fuse spatial and temporal features of defects and output fused feature values ​​as feature recognition results.

[0084] As a detailed explanation, the training samples of this feature fusion model include 1,500 sets of defect-free hydrophilic aluminum foil samples and 1,500 sets of hydrophilic aluminum foil samples containing point-like surface defects, line-like surface defects, sheet-like surface defects, irregular surface defects, and regional surface defects. Each set of defects corresponds to 300 sets of hydrophilic aluminum foil samples. After defect labeling, the samples are divided into training and validation sets in a 7:3 ratio.

[0085] Once the training accuracy, validation accuracy, recall, and F1 score of the feature fusion model reach the preset standards, the training of the feature fusion model is completed.

[0086] Based on this, the process for determining the defect type according to the feature recognition results is as follows:

[0087] After inputting twelve feature parameters into the input layer of the feature fusion model, the fully connected layer outputs the fused feature value that fuses the spatial features and temporal features of the defect.

[0088] A preset feature alarm threshold is set to determine whether the fused feature value exceeds the feature alarm threshold. If so, the defect type is determined; otherwise, the hydrophilic aluminum foil is determined to be defect-free.

[0089] The logic for determining the defect type is as follows:

[0090] When the electric field hydrophilic insulation distortion coefficient is ≥2.0, the absolute value of the hydrophilic angle deviation is >5°, and the absolute value of the infrared hydrophilic thermal conductivity gradient difference is ≥1℃, point-like surface defects, such as pinhole defects, are identified.

[0091] When the absolute value of the ultrasonic frequency deviation rate is ≥5%, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is ≥30%, and the vibration signal variance is ≥0.05, a linear or planar defect, such as a microcrack defect, is determined to have occurred.

[0092] When the absolute value of the ultrasonic frequency offset rate is ≥30%, the infrared thermal response hysteresis time is ≥0.5s, and the interaction signal coordination coefficient is ≤0.3, a sheet-like surface defect, such as a local peeling defect, is determined to have occurred.

[0093] When the electric field hydrophilic insulation distortion coefficient is ≥3.0, the absolute value of the infrared hydrophilic thermal conductivity gradient difference is ≥2℃, and the absolute value of the hydrophilic angle deviation is >10°, an irregular surface defect, such as a broken defect, is determined to have occurred.

[0094] When the absolute value of the hydrophilic angle deviation is >5°, the hydrophilic angle stability is ≥2°, the electric field hydrophilic insulation distortion coefficient is <2.0, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is <30%, and the absolute value of the ultrasonic frequency offset rate is <5%, a regional surface defect, such as a hydrophilic attenuation defect, is determined to have occurred.

[0095] In one embodiment, the defect location is calculated by the coordinates of the probe used to acquire the excitation interaction signal and the peak position of the characteristic parameter (i.e., the precise offset of the defect within the coverage area of ​​the corresponding probe).

[0096] Therefore, obtaining defect location information provides data support for determining the defect level.

[0097] In one embodiment, after determining the defect type and locating the defect position, the defect level is determined based on the characteristic parameter deviation and the defect area. The defect level determination logic is as follows:

[0098] For defects with an equivalent size (diameter of the circumcircle of the defect) < 50 μm and a defect area < 1962.5 μm... 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency deviation rate) is less than 30%, it is judged as a level one defect. At this time, the defect has no effect on the performance of hydrophilic aluminum foil (such as heat dissipation and hydrophilicity), and it is only recorded and archived without alarm.

[0099] Within 50μm ≤ equivalent size < 200μm, and 1962.5μm 2 ≤Defect area <31400μm 2 When 30%≤max (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, absolute value of ultrasonic frequency deviation rate) <50%, it is judged as a level 2 defect. At this time, the defect has a slight impact on the performance of the hydrophilic aluminum foil, a yellow alarm is issued, and a marking command is triggered.

[0100] With an equivalent size ≥ 200 μm and a defect area ≥ 31400 μm 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency offset rate) is ≥50%, it is judged as a level three defect. At this time, the defect has a serious impact on the performance of the hydrophilic aluminum foil, a red alarm is issued, and the machine is stopped for maintenance.

[0101] The following is an exemplary description of the above-mentioned method for detecting defects in the hydrophilic layer on the surface of hydrophilic aluminum foil, using a methodological flowchart. (Refer to...) Figure 1 Specifically, it includes the following steps:

[0102] S1. Establish a benchmark database (see the benchmark database establishment process above). Based on the core parameters of the hydrophilic aluminum foil to be tested, retrieve the corresponding excitation parameters and benchmark signal values.

[0103] S2. After the hydrophilic aluminum foil to be tested is flattened and cleaned, the hydrophilic aluminum foil to be tested is subjected to progressive targeted excitation of infrared preheating, ultrasonic activation and electric field strengthening according to the excitation parameters.

[0104] S3. Collect the excitation interaction parameters and extract the feature parameters by combining them with the reference signal value (see the twelve feature parameters mentioned above).

[0105] S4. Input the feature parameters into the feature fusion model to obtain the fused feature values;

[0106] S5. After setting the feature alarm threshold, determine the defect type based on the fused feature value (see the above judgment logic for determining the defect type).

[0107] S6. After locating the defect, determine the defect level based on the characteristic parameter deviation and the defect area (see the defect level determination logic above).

[0108] Example 2

[0109] A flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil, applied to the method provided in Example 1, the system comprising:

[0110] Infrared preheating excitation module: The infrared preheating excitation module is used to preheat the hydrophilic aluminum foil to be tested using infrared technology.

[0111] An ultrasonic activation excitation module is used to ultrasonically activate the hydrophilic aluminum foil to be tested.

[0112] The electric field enhancement excitation module is used to enhance the electric field of the hydrophilic aluminum foil to be tested.

[0113] The excitation interaction acquisition and analysis module is used to acquire excitation interaction signals after infrared preheating, ultrasonic activation, and electric field enhancement excitation, and to extract feature parameters.

[0114] The feature fusion model service module is used to calculate fused feature values ​​based on feature parameters.

[0115] The defect type determination module is used to determine the defect type of the hydrophilic aluminum foil to be tested based on the fusion feature value.

[0116] The defect level determination module is used to determine the defect level based on the defect type determination result.

[0117] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0118] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil, characterized in that, The system can be used in the following ways: Targeted excitation was performed on the hydrophilic aluminum foil to be tested to obtain excitation interaction signals; The excitation interaction signal is input into the feature fusion model, the feature fusion model outputs the feature recognition result, and the defect type is determined based on the feature recognition result.

2. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 1, characterized in that, The targeted excitation method for the hydrophilic aluminum foil to be tested is as follows: the hydrophilic aluminum foil to be tested is targetedly excited in the order of infrared preheating, ultrasonic activation, and electric field strengthening.

3. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 1, characterized in that, Before inputting the excitation interaction signal into the feature fusion model, the excitation interaction signal is denoised, the substrate signal is removed and normalized to obtain the purified excitation interaction signal. Feature parameters are extracted from the purified excitation interaction signal. The feature parameters include ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, infrared hydrophilic thermal conduction gradient difference, hydrophilic angle deviation, ultrasonic frequency shift, electric field current abrupt change rate, infrared thermal response hysteresis time, vibration signal variance, electric field signal variance, thermal response signal variance, hydrophilic angle stability, and interaction signal coordination coefficient.

4. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 3, characterized in that, The feature fusion model includes an input layer, an attention layer, a CNN feature extraction layer, an LSTM temporal mining layer, and a fully connected layer. The input layer is used to input all feature parameters; The attention layer is used to assign weights to each feature parameter; The CNN feature extraction layer is used to extract defect spatial features; The LSTM temporal mining layer is used to mine the temporal features of all feature parameters; The fully connected layer is used to fuse spatial and temporal features of defects and output fused feature values ​​as feature recognition results.

5. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 4, characterized in that, The process for determining the defect type based on the feature recognition results is as follows: A preset feature alarm threshold is set to determine whether the fused feature value exceeds the feature alarm threshold. If so, the defect type is determined; otherwise, the hydrophilic aluminum foil is determined to be defect-free.

6. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 5, characterized in that, The logic for determining the defect type is as follows: When the electric field hydrophilic insulation distortion coefficient is ≥2.0, the absolute value of the hydrophilic angle deviation is >5°, and the absolute value of the infrared hydrophilic thermal conductivity gradient difference is ≥1℃, a point-like surface defect is determined to have occurred. When the absolute value of the ultrasonic frequency deviation rate is ≥5%, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is ≥30%, and the vibration signal variance is ≥0.05, a linear-surface defect is determined to have occurred. A sheet-like surface defect is determined to occur when the absolute value of the ultrasonic frequency deviation rate is ≥30%, the infrared thermal response hysteresis time is ≥0.5s, and the interaction signal coordination coefficient is ≤0.

3. When the electric field hydrophilic insulation distortion coefficient is ≥3.0, the absolute value of the infrared hydrophilic thermal conductivity gradient difference is ≥2℃, and the absolute value of the hydrophilic angle deviation is >10°, an irregular surface defect is determined to have occurred. When the absolute value of the hydrophilic angle deviation is >5°, the hydrophilic angle stability is ≥2°, the electric field hydrophilic insulation distortion coefficient is <2.0, the absolute value of the ultrasonic hydrophilic interface bonding vibration deviation is <30%, and the absolute value of the ultrasonic frequency offset rate is <5%, a regional surface defect is determined to have occurred.

7. A flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to any one of claims 1-6, characterized in that, The system includes: An infrared preheating excitation module is used to preheat the hydrophilic aluminum foil to be tested using infrared technology. An ultrasonic activation excitation module is used to ultrasonically activate the hydrophilic aluminum foil to be tested. An electric field enhancement excitation module is used to enhance the electric field of the hydrophilic aluminum foil to be tested. The excitation interaction acquisition and analysis module is used to acquire excitation interaction signals after infrared preheating, ultrasonic activation, and electric field enhancement excitation, and extract feature parameters. A feature fusion model service module is used to calculate fused feature values ​​based on feature parameters. A defect type determination module is used to determine the defect type of the hydrophilic aluminum foil to be tested based on the fusion feature value. The defect level determination module is used to determine the defect level based on the defect type determination result.

8. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 7, characterized in that, Before determining the defect level based on the defect type determination result, the defect location is located based on the coordinates of the probe used to collect the excitation interaction signal and the peak value of the characteristic parameters.

9. The flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 1, characterized in that, When determining the defect level based on the defect type determination result, the determination logic is as follows: The equivalent size at the defect location is <50μm, and the defect area is <1962.5μm. 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency deviation rate) is less than 30%, it is judged as a first-level defect. The equivalent size at the defect location is less than 200 μm and 1962.5 μm within the range of 50 μm to 1962.5 μm. 2 ≤Defect area <31400μm 2 When 30%≤max (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, absolute value of ultrasonic frequency deviation rate) <50%, it is judged as a level two defect; The equivalent size at the defect location is ≥200μm and the defect area is ≥31400μm. 2 When the maximum value (absolute value of ultrasonic hydrophilic interface bonding vibration deviation, electric field hydrophilic insulation distortion coefficient, and absolute value of ultrasonic frequency deviation rate) is ≥50%, it is judged as a level three defect.

10. A flaw detection system for the hydrophilic layer defect on the surface of hydrophilic aluminum foil according to claim 9, characterized in that, When a defect is determined to be a Level 2 defect, a yellow alarm is issued and a marking command is triggered. When a defect is determined to be a Level 3 defect, a red alarm is issued and a shutdown command is triggered.

Images