Copper substrate binding force intelligent detection system and method based on Internet of Things

Abstract

The invention discloses a copper substrate binding force intelligent detection system and method based on Internet of Things, and relates to the technical field of Internet of Things intelligent detection, and the system comprises a plastic package coating module, a laser cutting module, a thrust test acquisition module, and a cloud analysis module. The plastic package coating module is used for uniformly coating a plastic package material on the surface of the to-be-tested copper substrate and forming a plastic package layer through heating and curing; the laser cutting module is used for processing regularly arranged plastic package material columnar bodies on the cured plastic package layer by using laser and is used for carrying out thrust test acquisition; the thrust test acquisition module applies thrust to the plastic package material columnar body by using a push broach device, records data in a test process in real time through carried sensor equipment, and transmits the data to the cloud server for analysis; and the cloud analysis module receives the data transmitted by the thrust test acquisition module, and further analyzes the data through an edge computing platform and a convolutional neural network model.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things intelligent detection, and in particular to an Internet of Things-based intelligent detection system and method for the bonding force of a copper substrate. Background Art

[0002] Copper substrates are specialized electronic materials widely used in high-power circuits, power converters, LED lighting, and other fields. They feature high heat dissipation and high precision, making their bonding strength crucial in production. However, conventional methods for testing the bonding strength of copper substrates lack precision, making it difficult to accurately quantify interfacial bonding strength. Furthermore, their applicability to complex structures (such as DCB substrates) is limited, and existing technologies are unable to achieve high-precision detection of local bonding strength. Therefore, this paper proposes a quantitative testing method based on laser processing and mechanical thrust. By designing a standardized columnar structure and correlating it with thrust values, this method enables rapid and accurate assessment of the interfacial bonding strength of copper substrates. Summary of the Invention

[0003] The object of the present invention is to provide a copper substrate bonding force intelligent detection system and method based on the Internet of Things to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] An intelligent detection method for copper substrate bonding strength based on the Internet of Things, characterized in that the method comprises the following steps:

[0006] Step S1, uniformly coating a plastic sealing material on the surface of the copper substrate to be tested, and curing the plastic sealing material by heating to form a plastic sealing layer;

[0007] Step S2: Setting a test plan, using a laser to process regularly arranged columns of plastic encapsulating material on the cured plastic encapsulating layer. The arrangement density is specified in the test plan. The diameter of the columns of plastic encapsulating material must be set according to the arrangement density and the thickness of the plastic encapsulating material. The diameter must be less than half of the thickness of the plastic encapsulating material to prevent the columns of plastic encapsulating material from breaking during testing and affecting the test process and final result data.

[0008] Step S3: Using a pusher device to apply thrust to the columnar plastic encapsulation material, the onboard mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor are used to record data during the test in real time. The data includes thrust value, copper substrate interface microstructure image, fracture surface image, ambient temperature and humidity, air pressure, and current time. The data during the test are timestamped and transmitted to a cloud server for analysis.

[0009] Step S4: Receive the transmitted data, perform preliminary processing on the collected raw data through the edge computing platform, upload the processed data to the cloud, and further analyze the image through the convolutional neural network model in combination with the timestamp, temperature, humidity, and air pressure; statistically analyze the thrust data of a certain number of columns, calculate the average value and dispersion of the copper substrate bonding force, and realize the evaluation of the copper substrate bonding force under the current external environmental conditions.

[0010] Furthermore, the method of heating and curing to form a plastic sealing layer includes the following steps:

[0011] According to the plastic encapsulation thickness required by the process, the plastic encapsulation material is evenly coated on the surface of the copper substrate to be tested. The material of the plastic encapsulation material is also stated in the process requirements;

[0012] The copper substrate to be tested, which is evenly coated with the plastic encapsulation material, is subjected to a heat curing treatment. The heating temperature, time and heating method are determined according to the plastic encapsulation material material stated in the process requirements.

[0013] Furthermore, the method for processing regularly arranged columns of plastic packaging material includes the following steps:

[0014] Set up a test plan, which specifies the arrangement of the molding material columns and the required diameter of the molding material columns. The diameter is set based on the density of the arrangement and the thickness of the molding material. It is usually set to half the thickness of the molding material to prevent the molding material columns from breaking during the test and affecting the test process and final results.

[0015] The laser wavelength, power, and pulse frequency are set according to the thickness of the plastic packaging material to achieve a processing depth that only penetrates the plastic packaging layer without damaging the copper substrate. The laser is used to process regularly arranged plastic packaging material columns on the solidified plastic packaging layer.

[0016] Furthermore, the method of applying a thrust to the columnar body of the plastic packaging material using the pusher device includes the following steps:

[0017] A pusher device is used to apply a thrust to the columnar body of the plastic encapsulating material. The thrust is applied in a direction perpendicular to the columnar body of the plastic encapsulating material so that the columnar body of the plastic encapsulating material is evenly stressed during the test. The data during the test are recorded according to the recording frequency set in the test plan by the onboard mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor. The above data includes the thrust value, the microstructure image of the copper substrate interface, the fracture surface image, the ambient temperature and humidity, the air pressure, and the current time.

[0018] When the columnar body of the plastic packaging material is stressed to the point of breaking, recording is stopped and the data recorded in step S3.1 is timestamped. The timestamp is the current time accurately recorded to milliseconds during data recording. All data are recorded in the database.

[0019] Furthermore, the method for performing preliminary processing on the collected raw data through the edge computing platform includes the following steps:

[0020] Perform data cleaning and filtering on the raw data to reduce noise. Use a low-pass filter to filter out high-frequency noise generated by mechanical vibration to reduce the impact on the test results.

[0021] The processed data is uploaded to the cloud for further analysis.

[0022] Furthermore, the method for further analysis using a convolutional neural network model includes the following steps:

[0023] The convolution kernel in the convolutional neural network technology is used to extract the color histogram and color moment of the plastic sealing material column in the image to describe its color change; the gray level co-occurrence matrix and local binary pattern are used to extract the texture of the plastic sealing material column in the image;

[0024] The above feature information is pooled to reduce the amount of data and the computational complexity of the model. The color and texture of all images of the same plastic encapsulating material column are compared to obtain an image at the moment when it is broken by the thrust.

[0025] An image of a column of plastic encapsulating material at the exact moment of fracture is acquired, along with its timestamp. Using this timestamp, the thrust value, ambient temperature, humidity, and air pressure recorded at the moment of fracture are retrieved from a database. The maximum thrust value that the column of plastic encapsulating material can withstand when fractured under the current ambient temperature, humidity, and air pressure conditions is determined and recorded in the database. This data represents the copper substrate bonding strength at the location of the column of plastic encapsulating material under the current ambient conditions.

[0026] Repeat the above method to record the maximum thrust value borne by a number of plastic encapsulation material columns in a certain area of ​​the copper substrate to be tested when they are broken, put the obtained thrust value data into a result set, calculate the average value of all thrust value data in the result set, and the obtained result is the copper substrate bonding strength of the copper substrate to be tested in this area. Set a copper substrate bonding strength threshold. If the obtained average value is lower than the threshold, the copper substrate to be tested fails this time; calculate the standard deviation of all data in the result set based on the average thrust value of the result set. The standard deviation is the dispersion of the data in the result set. Set a dispersion threshold. If the obtained standard deviation is higher than the threshold, the copper substrate to be tested has excessively large local differences in the copper substrate bonding strength in this area, and the copper substrate to be tested fails this time;

[0027] Data with an average value higher than the copper substrate bonding strength threshold and a standard deviation lower than the dispersion threshold is regarded as valid data. The data is recorded in the database together with the measured ambient temperature, humidity and air pressure, and marked as the copper substrate bonding strength of the copper substrate to be tested under the current ambient temperature, humidity and air pressure conditions. This enables the evaluation of the copper substrate bonding strength under the current external environmental conditions and can be used to compare the bonding strength of the copper substrate under other external environmental conditions to optimize the process.

[0028] An intelligent detection system for copper substrate bonding force based on the Internet of Things, including a plastic coating module, a laser cutting module, a thrust test acquisition module, and a cloud analysis module;

[0029] The plastic coating module is used to evenly coat the plastic coating material on the surface of the copper substrate to be tested, and form a plastic coating layer by heating and curing;

[0030] The laser cutting module is used to use laser to process regularly arranged plastic packaging material columns on the cured plastic packaging layer for thrust test collection;

[0031] The thrust test acquisition module is used to use a push knife device to apply thrust to the columnar body of the plastic packaging material, and record the data during the test in real time through the high-precision mechanical sensors, industrial cameras, temperature and humidity sensors, air pressure sensors, and time sensors installed. The data during the test is time-stamped and then transmitted to the cloud server for analysis;

[0032] The cloud analysis module receives the data transmitted by the thrust test acquisition module, performs preliminary processing on the collected raw data through the edge computing platform, uploads the processed data to the cloud, and further analyzes the image through the convolutional neural network model in combination with timestamp, temperature, humidity, and air pressure. The thrust data of a certain number of columns are statistically analyzed, and the average value and dispersion of the copper substrate bonding force are calculated to realize the evaluation of the copper substrate bonding force under different external environmental conditions.

[0033] Furthermore, the thrust test acquisition module includes a blade pushing device unit, a sensor unit, and a data transmission unit;

[0034] The thrust sensor accuracy of the pusher device unit is ≤0.1N, the pushing speed is 0.1-1mm / s, and the end face of the pusher is completely in contact with the top surface of the columnar body;

[0035] The sensor unit is used to obtain and record the thrust value, copper substrate interface microstructure image, fracture surface image, ambient temperature and humidity, air pressure, and current time in real time during the test process;

[0036] The data transmission unit is used to transmit the data acquired by the sensor unit to a cloud server.

[0037] Furthermore, the cloud analysis module includes an edge computing platform unit and a convolutional neural network model unit;

[0038] The edge computing platform unit is used to perform preliminary processing on the collected raw data;

[0039] The convolutional neural network model unit is used to further analyze the collected images.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. Through laser processing, standardized columns are formed on the plastic packaging layer to convert the copper substrate interface bonding force into a quantifiable mechanical thrust value;

[0042] 2. Fixed-point measurement can be performed on local areas;

[0043] 3. Suitable for high-precision bonding force detection needs in the fields of microelectronics packaging, PCB manufacturing, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a flow chart of an intelligent detection method for copper substrate bonding strength based on the Internet of Things of the present invention;

[0045] Figure 2 This is a structural schematic diagram of an intelligent detection system for copper substrate bonding strength based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0046] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0047] Example: Figure 1-Figure 2 As shown, the present invention provides a technical solution, a method for intelligently detecting the bonding strength of a copper substrate based on the Internet of Things, characterized in that the method comprises the following steps:

[0048] Step S1, uniformly coating a plastic sealing material on the surface of the copper substrate to be tested, and curing the plastic sealing material by heating to form a plastic sealing layer;

[0049] Step S2: Setting a test plan, using a laser to process regularly arranged columns of plastic encapsulating material on the cured plastic encapsulating layer. The arrangement density is specified in the test plan. The diameter of the columns of plastic encapsulating material must be set according to the arrangement density and the thickness of the plastic encapsulating material. The diameter must be less than half of the thickness of the plastic encapsulating material to prevent the columns of plastic encapsulating material from breaking during testing and affecting the test process and final result data.

[0050] Step S3: Using a pusher device to apply thrust to the columnar plastic encapsulation material, the onboard mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor are used to record data during the test in real time. The data includes thrust value, copper substrate interface microstructure image, fracture surface image, ambient temperature and humidity, air pressure, and current time. The data during the test are timestamped and transmitted to a cloud server for analysis.

[0051] Step S4: Receive the transmitted data, perform preliminary processing on the collected raw data through the edge computing platform, upload the processed data to the cloud, and further analyze the image through the convolutional neural network model in combination with the timestamp, temperature, humidity, and air pressure; statistically analyze the thrust data of a certain number of columns, calculate the average value and dispersion of the copper substrate bonding force, and realize the evaluation of the copper substrate bonding force under the current external environmental conditions.

[0052] Furthermore, the method of heating and curing to form a plastic sealing layer includes the following steps:

[0053] According to the plastic encapsulation thickness required by the process, the plastic encapsulation material is evenly coated on the surface of the copper substrate to be tested. The material of the plastic encapsulation material is also stated in the process requirements;

[0054] The copper substrate to be tested, which is evenly coated with the plastic encapsulation material, is subjected to a heat curing treatment. The heating temperature, time and heating method are determined according to the plastic encapsulation material material stated in the process requirements.

[0055] Furthermore, the method for processing regularly arranged columns of plastic packaging material includes the following steps:

[0056] Set up a test plan, which specifies the arrangement of the molding material columns and the required diameter of the molding material columns. The diameter is set based on the density of the arrangement and the thickness of the molding material. It is usually set to half the thickness of the molding material to prevent the molding material columns from breaking during the test and affecting the test process and final results.

[0057] The laser wavelength, power, and pulse frequency are set according to the thickness of the plastic packaging material to achieve a processing depth that only penetrates the plastic packaging layer without damaging the copper substrate. The laser is used to process regularly arranged plastic packaging material columns on the solidified plastic packaging layer.

[0058] Furthermore, the method of applying a thrust to the columnar body of the plastic packaging material using the pusher device includes the following steps:

[0059] A pusher device is used to apply a thrust to the columnar body of the plastic encapsulating material. The thrust is applied in a direction perpendicular to the columnar body of the plastic encapsulating material so that the columnar body of the plastic encapsulating material is evenly stressed during the test. The data during the test are recorded according to the recording frequency set in the test plan by the onboard mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor. The above data includes the thrust value, the microstructure image of the copper substrate interface, the fracture surface image, the ambient temperature and humidity, the air pressure, and the current time.

[0060] When the columnar body of the plastic packaging material is stressed to the point of breaking, recording is stopped and the data recorded in step S3.1 is timestamped. The timestamp is the current time accurately recorded to milliseconds during data recording. All data are recorded in the database.

[0061] Furthermore, the method for performing preliminary processing on the collected raw data through the edge computing platform includes the following steps:

[0062] Perform data cleaning and filtering on the raw data to reduce noise. Use a low-pass filter to filter out high-frequency noise generated by mechanical vibration to reduce the impact on the test results.

[0063] The processed data is uploaded to the cloud for further analysis.

[0064] Furthermore, the method for further analysis using a convolutional neural network model includes the following steps:

[0065] The convolution kernel in the convolutional neural network technology is used to extract the color histogram and color moment of the plastic sealing material column in the image to describe its color change; the gray level co-occurrence matrix and local binary pattern are used to extract the texture of the plastic sealing material column in the image;

[0066] The above feature information is pooled to reduce the amount of data and the computational complexity of the model. The color and texture of all images of the same plastic encapsulating material column are compared to obtain an image at the moment when it is broken by the thrust.

[0067] An image of a column of plastic encapsulating material at the exact moment of fracture is acquired, along with its timestamp. Using this timestamp, the thrust value, ambient temperature, humidity, and air pressure recorded at the moment of fracture are retrieved from a database. The maximum thrust value that the column of plastic encapsulating material can withstand when fractured under the current ambient temperature, humidity, and air pressure conditions is determined and recorded in the database. This data represents the copper substrate bonding strength at the location of the column of plastic encapsulating material under the current ambient conditions.

[0068] Repeat the above method to record the maximum thrust value borne by a number of plastic encapsulation material columns in a certain area of ​​the copper substrate to be tested when they are broken, put the obtained thrust value data into a result set, calculate the average value of all thrust value data in the result set, and the obtained result is the copper substrate bonding strength of the copper substrate to be tested in this area. Set a copper substrate bonding strength threshold. If the obtained average value is lower than the threshold, the copper substrate to be tested fails this time; calculate the standard deviation of all data in the result set based on the average thrust value of the result set. The standard deviation is the dispersion of the data in the result set. Set a dispersion threshold. If the obtained standard deviation is higher than the threshold, the copper substrate to be tested has excessively large local differences in the copper substrate bonding strength in this area, and the copper substrate to be tested fails this time;

[0069] Data with an average value higher than the copper substrate bonding strength threshold and a standard deviation lower than the dispersion threshold is regarded as valid data. The data is recorded in the database together with the measured ambient temperature, humidity and air pressure, and marked as the copper substrate bonding strength of the copper substrate to be tested under the current ambient temperature, humidity and air pressure conditions. This enables the evaluation of the copper substrate bonding strength under the current external environmental conditions and can be used to compare the bonding strength of the copper substrate under other external environmental conditions to optimize the process.

[0070] The specific implementation scenario is to test the copper substrate bonding strength of a batch of DCB substrates used for high-power LED heat dissipation.

[0071] The size of the DCB substrate is 50mm×50mm×1.0mm, and the plastic packaging material is a certain type of epoxy resin with good fluidity and curing properties at 150°C.

[0072] Epoxy resin was evenly coated on the surface of the DCB substrate using a screen printing process, with a coating thickness of 0.6 mm. After coating, the substrate was placed in an oven and heated to 150°C at a rate of 5°C / min. The temperature was maintained at this temperature for 2 hours to fully cure the epoxy resin and form a plastic seal layer.

[0073] The cured plastic layer is 0.6mm thick, and a pulsed laser with a wavelength of 1064nm is used for processing. Based on experience and previous experiments, the laser power is set to 10W, the pulse frequency is 20kHz, and the processing speed is 5mm / s. This ensures that the processing depth only penetrates the plastic layer without damaging the copper substrate, and that the bottom of the plastic column is in close contact with the copper substrate.

[0074] To ensure the accuracy and representativeness of the test results, the columns of plastic encapsulation material were designed to be arranged in a 5×5 matrix in the center area of ​​the substrate. Based on the estimated copper substrate interface strength range, the column spacing was set to 3mm. Taking into account the arrangement density and the thickness of the plastic encapsulation material, the column diameter was set to 1.2mm after calculation and simulation. This can effectively avoid the test results being affected by the column's own fracture in subsequent thrust tests. A high-precision pusher device was used, which was equipped with a mechanical sensor with a resolution of 0.01N, an industrial camera with a pixel count of 12 million, a temperature and humidity sensor with an accuracy of ±0.5℃ and ±2%RH, an air pressure sensor with an accuracy of ±0.1hPa, and a time sensor with an accuracy of 0.001s.

[0075] The DCB substrate, after being machined into a cylindrical shape, is secured to the test platform of a pusher mechanism, with the pusher's direction perpendicular to the plastic encapsulating material column. The pusher mechanism is activated, applying a thrust to the column at a rate of 0.1 mm / s. During the test, data is recorded in real time using a high-precision mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor. The mechanical sensor records the thrust value in real time, while the industrial camera captures images of the copper substrate's interface microstructure and fracture surface every 0.1 seconds. The temperature and humidity sensor and air pressure sensor also collect ambient temperature, humidity, and air pressure data every 0.1 seconds. The time sensor also records the current time.

[0076] When testing the plastic encapsulation material column in row 3, column 2, force is applied to the column until it breaks from the DCB substrate. All recorded data is immediately timestamped. This time-stamped data is quickly transmitted to a cloud server via industrial Ethernet for subsequent analysis. This process is repeated for each plastic encapsulation material column until all 25 columns have broken under force. All data is then transferred to the cloud server for analysis.

[0077] After receiving the raw data from the island, the edge computing platform within the cloud server first cleans it to remove erroneous data caused by sensor anomalies and other factors. It then applies a low-pass filter with a cutoff frequency of 50 Hz to the raw thrust data collected by the force sensor to reduce noise, removing high-frequency noise caused by mechanical vibration and improving data quality and processing efficiency. The processed data is then uploaded to the cloud.

[0078] The cloud uses a convolutional neural network model to further analyze the image data. The convolution kernel extracts the color histogram and color moments of the plastic encapsulation material cylinder image to describe its color variations. The gray-level co-occurrence matrix and local binary pattern are used to extract texture features from the image. Pooling is performed on this feature information to reduce the data size and computational complexity of the model. By comparing the color and texture of all images of the plastic encapsulation material cylinder in row 3, column 2 at different times, the image at the moment of fracture under thrust is accurately obtained, with a timestamp of 10:25:30.567. Based on this timestamp, the thrust value recorded at that moment is accurately extracted from the database, which is 50.23N. The corresponding ambient temperature is 25.1°C, the relative humidity is 49.8% RH, and the air pressure is 1013.1hPa. The external environmental conditions at the time of fracture are determined and recorded in the database.

[0079] Statistical analysis of data from 25 columns of plastic encapsulation material within the center region of the DCB substrate revealed an average copper-to-substrate bonding force of 48.5 N, with a dispersion of 3.2 N. By combining statistical data from external environmental conditions recorded at different times, we can provide a detailed assessment of the copper-to-substrate bonding force of DCB substrates under varying environmental conditions, providing robust data support for optimizing product performance and improving quality.

[0080] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. An intelligent detection method for copper substrate bonding strength based on the Internet of Things, characterized in that: The method comprises the following steps: Step S1, uniformly coating a plastic sealing material on the surface of the copper substrate to be tested, and curing the plastic sealing material by heating to form a plastic sealing layer; Step S2: Setting a detection scheme, using a laser to process regularly arranged columns of the plastic encapsulating material on the cured plastic encapsulating layer, wherein the arrangement density is specified in the detection scheme, and the diameter of the columns of the plastic encapsulating material is set according to the arrangement density and the plastic encapsulating thickness determined in step S1; Step S3: Using a pusher device to apply thrust to the columnar plastic encapsulation material, the onboard mechanical sensor, industrial camera, temperature and humidity sensor, air pressure sensor, and time sensor are used to record data during the test in real time. The data includes thrust value, copper substrate interface microstructure image, fracture surface image, ambient temperature and humidity, air pressure, and current time. The data during the test are timestamped and transmitted to a cloud server for analysis. Step S4: Receive the data transmitted in step S3, perform preliminary processing on the collected raw data through the edge computing platform, upload the processed data to the cloud, and further analyze the image through the convolutional neural network model in combination with the timestamp, temperature, humidity, and air pressure; statistically analyze the thrust data of a number of columns, calculate the average value and dispersion of the copper substrate bonding force, and realize the evaluation of the copper substrate bonding force under the current external environmental conditions.

2. The method for intelligent detection of copper substrate bonding strength based on the Internet of Things according to claim 1, characterized in that: The method for heating and curing to form a plastic sealing layer in step S1 comprises the following steps: Step S1.1: According to the thickness of the plastic seal required by the process, the plastic seal material is evenly coated on the surface of the copper substrate to be tested. The material of the plastic seal material is also stated in the process requirements. Step S1.2: The copper substrate to be tested, which is evenly coated with the plastic encapsulation material, is subjected to a heating and curing treatment. The heating temperature, time and heating method are determined according to the plastic encapsulation material material stated in the process requirements of step S1.

1.

3. The method for intelligent detection of copper substrate bonding strength based on the Internet of Things according to claim 2, characterized in that: The method for processing the regularly arranged plastic encapsulation material columns in step S2 comprises the following steps: Step S2.1, setting a test plan, which specifies the arrangement of the molding material columns and the required diameter of the molding material columns in this test, wherein the diameter is set based on the density of the arrangement and the thickness of the molding material in step S1.1; Step S2.2: setting the laser wavelength, power, and pulse frequency according to the thickness of the plastic packaging material in step S1.1, and using the laser to process regularly arranged plastic packaging material columns on the cured plastic packaging layer.

4. The method for intelligently detecting the bonding strength of copper substrates based on the Internet of Things according to claim 3, wherein: The method of using the pusher device to apply a thrust to the columnar body of the plastic packaging material in step S3 includes the following steps: Step S3.1, using a pusher device to apply a thrust to the columnar body of the plastic encapsulating material, with the thrust applied in a direction perpendicular to the columnar body of the plastic encapsulating material, and recording data from the test process according to a recording frequency using a mechanical sensor, an industrial camera, a temperature and humidity sensor, an air pressure sensor, and a time sensor. The recording frequency is set in the detection scheme of step S2.

1. The above data includes the thrust value, the microstructure image of the copper substrate interface, the fracture surface image, the ambient temperature and humidity, the air pressure, and the current time; Step S3.2: When the columnar plastic material is stressed to the point of breaking, stop recording and add a timestamp to the data recorded in step S3.

1. The timestamp is the current time accurately recorded to the millisecond when the data is recorded. All data are recorded in the database.

5. The method for intelligent detection of copper substrate bonding strength based on the Internet of Things according to claim 1, characterized in that: The method for performing preliminary processing on the collected raw data by the edge computing platform in step S4 includes the following steps: Step S4.1, clean the raw data, filter and reduce noise, and use a low-pass filter to remove high-frequency noise generated by mechanical vibration; Step S4.2: Upload the data processed in step S4.1 to the cloud for further analysis.

6. The method for intelligent detection of copper substrate bonding strength based on the Internet of Things according to claim 1, characterized in that: The method of further analyzing the convolutional neural network model in step S4 includes the following steps: Step S4.3: Using the convolution kernel in the convolutional neural network technology, the color histogram and color moments of the columnar plastic material in the image are extracted to describe its color variation; and the texture of the columnar plastic material in the image is extracted using the gray-level co-occurrence matrix and local binary pattern. The above feature information is pooled to reduce the amount of data and the computational complexity of the model. The color and texture of all images of the same plastic encapsulating material column are compared to obtain an image at the moment when it is broken by the thrust. Step S4.4: Acquire an image of the moment when a column of plastic encapsulating material is broken in step S4.3, obtain its timestamp, and obtain the thrust value, ambient temperature, humidity, and air pressure recorded when the column of plastic encapsulating material is broken from the database using the timestamp. Determine the maximum thrust value that the column of plastic encapsulating material can withstand when it breaks under the current external ambient temperature, humidity, and air pressure conditions, and record it in the database. This data represents the copper substrate bonding strength of the copper substrate to be tested at the point where the column of plastic encapsulating material is located under the current external ambient conditions. Step S4.5, repeat the method of step S4.4, record the maximum thrust value borne when a number of plastic encapsulation material columns in a certain area of ​​the copper substrate to be tested are broken, put the obtained thrust value data into a result set, calculate the average value of all thrust value data in the result set, and the obtained result is the copper substrate bonding force of the copper substrate to be tested in the area, set the copper substrate bonding force threshold, if the obtained average value is lower than the threshold, the copper substrate to be tested is unqualified; calculate the standard deviation of all data in the result set based on the average thrust value of the result set, the standard deviation is the dispersion of the data in the result set, set the dispersion threshold, if the obtained standard deviation is higher than the threshold, the copper substrate to be tested has a local difference in the copper substrate bonding force in this area that is too large, and the copper substrate to be tested is unqualified; Data with an average value higher than the copper substrate bonding strength threshold and a standard deviation lower than the dispersion threshold is regarded as valid data. The data is recorded in the database together with the measured ambient temperature, humidity and air pressure, and marked as the copper substrate bonding strength of the copper substrate to be tested under the current ambient temperature, humidity and air pressure conditions. This enables the evaluation of the copper substrate bonding strength under the current external environmental conditions and can be used to compare the bonding strength of the copper substrate under other external environmental conditions to optimize the process.

7. An intelligent detection system for copper substrate bonding strength based on the Internet of Things, using the intelligent detection method for copper substrate bonding strength based on the Internet of Things according to any one of claims 1 to 6, characterized in that: The system includes a plastic coating module, a laser cutting module, a thrust test acquisition module, and a cloud analysis module; The plastic coating module is used to evenly coat the plastic coating material on the surface of the copper substrate to be tested, and form a plastic coating layer by heating and curing; The laser cutting module is used to use laser to process regularly arranged plastic packaging material columns on the cured plastic packaging layer for thrust test collection; The thrust test acquisition module is used to use a push knife device to apply thrust to the columnar body of the plastic packaging material, and record the data during the test in real time through the high-precision mechanical sensors, industrial cameras, temperature and humidity sensors, air pressure sensors, and time sensors installed. The data during the test is time-stamped and then transmitted to the cloud server for analysis; The cloud analysis module receives the data transmitted by the thrust test acquisition module, performs preliminary processing on the collected raw data through the edge computing platform, uploads the processed data to the cloud, and further analyzes the image through the convolutional neural network model in combination with timestamp, temperature, humidity, and air pressure. The thrust data of a certain number of columns are statistically analyzed, and the average value and dispersion of the copper substrate bonding force are calculated to realize the evaluation of the copper substrate bonding force under different external environmental conditions.

8. The intelligent detection system for copper substrate bonding strength based on Internet of Things according to claim 7, characterized in that: The thrust test acquisition module includes a blade pushing device unit, a sensor unit, and a data transmission unit; The thrust sensor accuracy of the pusher device unit is ≤0.1N, the pushing speed is 0.1-1mm / s, and the end face of the pusher is completely in contact with the top surface of the columnar body; The sensor unit is used to obtain and record the thrust value, copper substrate interface microstructure image, fracture surface image, ambient temperature and humidity, air pressure, and current time in real time during the test process; The data transmission unit is used to transmit the data acquired by the sensor unit to a cloud server.

9. The intelligent detection system for copper substrate bonding strength based on Internet of Things according to claim 7, characterized in that: The cloud analysis module includes an edge computing platform unit and a convolutional neural network model unit; The edge computing platform unit is used to perform preliminary processing on the collected raw data; The convolutional neural network model unit is used to further analyze the collected images.

Images