Device and method for qualification inspection and fault diagnosis of vapor chamber type radiator

Abstract

The invention discloses a device and a method for qualification inspection and fault diagnosis of a vapor chamber type radiator, and belongs to the technical field of vapor chamber detection. In order to realize high accuracy of qualification inspection and fault diagnosis of the uniform temperature plate type radiator and simple operation, the device comprises a computer, a thermal imager and a constant temperature water tank, a first iron rod, a second iron rod and an iron net are arranged in the constant temperature water tank, and the first iron rod, the second iron rod and the iron net form a test position for fixedly placing a to-be-detected uniform temperature plate; the evaporation section of the vapor chamber to be detected is soaked in the constant-temperature water tank, and the condensation section of the vapor chamber to be detected is exposed in normal-temperature air; the computer and the thermal imager are connected through a USB connecting line. And the thermal imager is placed on the outer side of the constant-temperature water tank at a certain distance and faces the front position of the condensation section of the vapor chamber to be detected. According to the invention, the universality is good, different types of uniform temperature plates can be rapidly matched by adjusting the clamping tool, and the detection capability of the detection equipment on different types of uniform temperature plates can be rapidly expanded.

Description

Technical Field

[0001] The present invention belongs to the technical field of temperature evaporator detection, and in particular relates to a device and method for quality inspection and fault diagnosis of a temperature evaporator type radiator. Background Art

[0002] Today, electronic technology is developing rapidly, and the computing performance of various chips is rapidly improving. With the increase in the number of transistors and the operating frequency, the thermal design power consumption (TDP) of the chip is also increasing. The thermal conductivity of the heat sink supporting the chip is crucial to the stable operation and service life of the equipment. The temperature plate type heat sink is a new type of heat sink. Its basic principle is to use the evaporation-condensation of the liquid stored in the internal cavity of the heat sink for rapid heat conduction. Compared with various heat sinks made of traditional copper, aluminum and other materials, the thermal conductivity speed and heat dissipation capacity have been significantly improved.

[0003] At present, for the heat dissipation performance level of the temperature equalization board, the key performance that many electronic manufacturers and end users pay attention to is its "temperature equalization performance", that is, the relative temperature difference between the evaporation end and the condensation end of the temperature equalization board under a high load state. There are two main types of detection methods for the detection of temperature equalization performance and the judgment of qualified products: contact and non-contact. The contact detection method generally uses equipment such as thermal conductivity testers to test its thermal conductivity efficiency, uses thermal resistors and other methods to heat it, and pastes thermocouple sensors on the temperature equalization board to be tested for single-point or multi-point temperature measurement; the non-contact method mainly uses laser heating, infrared temperature measurement and other technologies to perform non-contact detection of the temperature equalization board.

[0004] The popular methods for testing the thermal conductivity of temperature equalizing plates on the market, such as contact method, single-point method with adhesive sensors, multi-point method, and non-contact infrared temperature measurement and infrared imaging methods, all have certain problems and limitations.

[0005] (1) The current temperature vapor chamber inspection technology focuses on batch inspection of newly produced products, and the inspection results are limited to whether the products are qualified or not, aiming to determine whether the products meet the predetermined quality standards, but no further fault diagnosis and judgment are performed on unqualified products. However, for some products with extremely high stability requirements, it is necessary to manage the temperature vapor chamber throughout the product life cycle and return it to the factory for regular inspection to detect defects in the product during operation as early as possible; or for some high-cost temperature vapor chamber products, it is necessary to diagnose and judge the faults of unqualified products, and determine whether they can be returned to the factory for repair to achieve the qualified state according to the type of fault, so as to achieve the purpose of cost saving. At present, the above two requirements are all achieved manually by experienced technicians, which is not only time-consuming and labor-intensive, but also has a certain risk of misjudgment.

[0006] (2) The contact detection method generally uses equipment such as a thermal conductivity tester to test its thermal conductivity efficiency, uses thermal resistors and other methods to heat it, and pastes thermocouple sensors on the temperature-averaging board to be tested for single-point or multi-point temperature measurement. Although the contact detection method has high detection accuracy, the detection speed is slow, and there are some difficult problems to solve: for example, the use of heating resistors for contact heating will be affected by the heating power of the components, the temperature rise time is slow, and it cannot meet the requirements of high-power extreme heat dissipation performance testing; when using the water bath method for heating detection, the sensor may be affected by water vapor; at the same time, when the temperature-averaging board to be tested contacts with the detection tooling and the pasted sensor, scratches, deformation or adhesive residue may occur.

[0007] (3) Non-contact methods mainly use laser heating, infrared temperature measurement and other technologies to perform non-contact detection on the heat spreader. The efficiency of non-contact detection is faster than that of contact detection, but there are the following difficult-to-solve problems: First, infrared thermal imaging temperature measurement technology is easily affected by factors such as ambient temperature and humidity. The measurement error is generally ±2°C. Excessive measurement errors make it difficult to meet the inspection requirements of the heat spreader's production standards; second, the infrared imaging method is not ideal for heat spreaders with complex surface shapes; third, the use of non-contact heating, such as laser heating, is also affected by the complex surface shape and inconsistent reflectivity of the heat spreader, and the heating efficiency is difficult to measure. Summary of the invention

[0008] The problem to be solved by the present invention is to achieve high accuracy and simple operation in qualified inspection and fault diagnosis of a temperature-averaging plate type radiator, and propose a device and method for qualified inspection and fault diagnosis of a temperature-averaging plate type radiator.

[0009] To achieve the above object, the present invention is implemented through the following technical solutions:

[0010] A device for qualified inspection and fault diagnosis of a heat spreader type radiator, comprising a computer, a thermal imager, and a constant temperature water tank, wherein a first iron rod, a second iron rod, and an iron net are arranged in the constant temperature water tank, wherein the first iron rod and the second iron rod work together to immerse the evaporation section of the heat spreader to be inspected in the constant temperature water tank, and the condensation section of the heat spreader to be inspected is exposed to air at normal temperature;

[0011] The computer and the thermal imager are connected via a USB cable; the thermal imager is placed at a certain distance outside the constant temperature water tank, facing the front of the condensation section of the temperature homogenizing plate to be tested;

[0012] The computer is installed with a deep convolutional neural network model, which is used to perform qualification inspection and fault diagnosis on photos taken by the thermal imager.

[0013] Furthermore, a cooling fan is provided on the outside of the constant temperature water tank, and the cooling fan is placed on the back of the condensation section of the temperature equalizing plate to be tested, the front of the condensation section of the temperature equalizing plate to be tested is the contact surface of the chip, and the back of the condensation section of the temperature equalizing plate to be tested is the surface with the heat dissipation grooves.

[0014] Furthermore, constant temperature hot water is filled into the constant temperature water tank according to the water level line. After the evaporation section of the temperature equalizing plate to be tested is placed in the constant temperature water tank for a certain period of time, the thermal imager takes a front view of the temperature equalizing plate to be tested, and then transmits it to the computer via a USB cable.

[0015] Furthermore, the water level of the constant temperature water tank is set to immerse the evaporation section of the temperature equalizing plate to be tested, and the hot water temperature of the constant temperature water tank is 80°C.

[0016] Furthermore, after the evaporation section of the temperature homogenizing plate to be tested is placed in a constant temperature water tank for 30 seconds, a thermal imager takes a picture.

[0017] A method for quality inspection and fault diagnosis of a heat spreader plate type heat sink is implemented based on the device for quality inspection and fault diagnosis of a heat spreader plate type heat sink, and comprises the following steps:

[0018] S1. The operator takes his place, performs equipment status inspection, and conducts a comprehensive inspection of the device for qualified inspection and fault diagnosis of the temperature-averaging plate type radiator to ensure that all equipment is in normal working condition;

[0019] S2. Fill the constant temperature water tank according to the water level line and inject hot water according to the hot water temperature to provide a constant temperature water environment for the temperature equalizer to be tested;

[0020] S3. Power on all equipment, including computers, thermal imagers, constant temperature water tanks, and cooling fans;

[0021] S4. Check the temperature of the thermostatic water tank and start the test after confirming that it reaches the preset temperature of 80°C and is stable;

[0022] S5. Use manual or mechanical hands to place the temperature-averaging board to be tested at the test position in the constant temperature water tank, and use the first iron rod, the second iron rod and the iron net at the bottom of the constant temperature water tank to support the temperature-averaging board to be tested. The temperature-averaging board to be tested is immersed in the constant temperature water upright, and the evaporation section of the temperature-averaging board to be tested is consistent with the immersion depth;

[0023] S6. Keep the temperature plate to be tested in the constant temperature water tank for 30 seconds to achieve heat dissipation equilibrium;

[0024] S7. Use a thermal imager to photograph the temperature balancing board to be tested to obtain an infrared thermal imaging picture;

[0025] S8. The infrared thermal imaging image is transferred to the computer via a USB cable, and the computer runs a pre-trained deep convolutional neural network model to extract the temperature distribution characteristics of the temperature equalizing plate to be tested, and obtain the output result of the deep convolutional neural network model;

[0026] S9. Based on the output result of the deep convolutional neural network model, determine whether the temperature equalizing board to be tested is a qualified product. If it is judged to be a qualified product, it will be processed as a qualified product. If it is not, it will be processed as an unqualified product. Then, the next temperature equalizing board to be tested will be tested until all the temperature equalizing boards to be tested have been tested.

[0027] Furthermore, the qualified products are processed by outputting a test certificate, which is then taken out manually or by a robot and transferred to a bellows for cooling and air drying, and then the product and the certificate are packed together into a qualified product box.

[0028] Furthermore, the defective product processing is that the deep convolutional neural network model outputs the fault type, and the labels of the faulty products include over-filling of working fluid, under-filling of working fluid, poor machining accuracy, deformation tolerance, rust, and unqualified sintered structure. The products are then taken out manually or by a robot and transferred to a bellows for cooling and air drying. After that, the products and fault reports are put into a defective product box for subsequent processing.

[0029] Furthermore, the deep convolutional neural network model is a CNN model.

[0030] Beneficial effects of the present invention:

[0031] The device for qualified inspection and fault diagnosis of a temperature-averaging plate type radiator described in the present invention provides a water bath heating method. Due to the large specific heat capacity of water, compared with thermal resistors, laser heating and other forms, the temperature-averaging plate to be tested can be quickly brought into a heat dissipation equilibrium state, thereby effectively speeding up the detection speed. At the same time, by setting reasonable positioning tooling, equipping a low-air-volume cooling fan and other means, the influence of interfering factors such as water vapor is reduced, and the immersion depth of the evaporation section of the temperature-averaging plate is consistent, so that the temperature-averaging plate can be heated in a standard and uniform manner, and the detection results are more reliable.

[0032] The device for qualified inspection and fault diagnosis of a heat spreader type radiator described in the present invention has good versatility. First, different types of heat spreaders can be quickly matched by adjusting the positioning tooling; second, the deep neural network of the present invention has good generalization ability. By expanding the corresponding data set and after training, the detection capability of the detection equipment for different types of heat spreaders can be quickly expanded.

[0033] The device for qualified inspection and fault diagnosis of a heat spreader type radiator described in the present invention can quickly and accurately inspect heat spreader products. The present invention uses deep neural network technology to change the determination of single-point or multi-point temperature of the heat spreader by traditional detection technology to extracting the features of the entire infrared imaging picture and making a determination, thereby avoiding the disadvantage of low precision of infrared imaging technology. Therefore, the present invention combines the advantages of high accuracy of contact detection technology and high speed of non-contact detection technology. At the same time, for some high-value heat spreader products or heat spreader products that require full life cycle monitoring, the fault classification function provided by the present invention can accurately judge the faulty products, effectively solving the problem that manual fault diagnosis relies heavily on work experience and is time-consuming and labor-intensive. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a structural schematic diagram of a device for quality inspection and fault diagnosis of a temperature-homing plate type heat sink according to the present invention;

[0035] Figure 2 It is a flow chart of a method for quality inspection and fault diagnosis of a temperature-vaporizing plate type heat sink according to the present invention;

[0036] Figure 3 This is a flow chart of the deep convolutional neural network model training of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention more clear, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the specific embodiments described are only part of the embodiments of the present invention, rather than all of the specific embodiments. The components of the specific embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations, and the present invention can also have other embodiments.

[0038] Therefore, the following detailed description of the specific embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents the selected specific embodiments of the present invention. Based on the specific embodiments of the present invention, all other specific embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0039] In order to further understand the content, features and effects of the present invention, the following specific implementation methods are given as examples, and the attached Figure 1 -Attached Figure 3 The detailed instructions are as follows:

[0040] Embodiment 1:

[0041] A device for qualified inspection and fault diagnosis of a temperature-averaging plate type radiator, comprising a computer 1, a thermal imager 3, and a constant temperature water tank 4, wherein a first iron rod 6-1, a second iron rod 6-2, and an iron net 7 are arranged in the constant temperature water tank 4, wherein the first iron rod 6-1, the second iron rod 6-2, and the iron net 7 work together to immerse the evaporation section of the temperature-averaging plate 5 to be inspected in the constant temperature water tank 4, and the condensation section of the temperature-averaging plate 5 to be inspected is exposed to air at normal temperature;

[0042] The computer 1 and the thermal imager 3 are connected via a USB cable 2; the thermal imager 3 is placed at a certain distance outside the constant temperature water tank 4, facing the front position of the condensation section of the temperature equalizing plate 5 to be detected;

[0043] The computer 1 is installed with a deep convolutional neural network model for judging the quality of the photos taken by the thermal imager and the fault diagnosis;

[0044] Furthermore, a cooling fan 8 is provided on the outside of the constant temperature water tank 4, and the cooling fan 8 is placed on the back side of the condensation section of the temperature equalizing plate 5 to be tested. The front side of the condensation section of the temperature equalizing plate 5 to be tested is the contact surface of the chip, and the back side of the condensation section of the temperature equalizing plate 5 to be tested is the surface with the heat dissipation grooves.

[0045] Furthermore, constant temperature hot water is filled into the constant temperature water tank 4 according to the water level line. After the evaporation section of the temperature equalizing plate 5 to be tested is placed in the constant temperature water tank 4 for a certain period of time, the thermal imager 3 takes a front view of the temperature equalizing plate 5 to be tested, and then transmits it to the computer 1 via the USB connection cable 2.

[0046] Furthermore, the water level of the constant temperature water tank 4 is set to immerse the evaporation section of the temperature equalizing plate 5 to be tested, and the hot water temperature of the constant temperature water tank 4 is 80°C.

[0047] Furthermore, after the evaporation section of the temperature homogenizing plate 5 to be tested is placed in the constant temperature water tank 4 for 30 seconds, the thermal imager 3 takes a picture.

[0048] Furthermore, the fan has a wind speed of about 3 m / s, and is placed 30 cm away from the measured temperature averaging board. The fan is always turned on during the experiment.

[0049] Embodiment 2:

[0050] A method for quality inspection and fault diagnosis of a heat spreader type heat sink is implemented based on the device for quality inspection and fault diagnosis of a heat spreader type heat sink described in Example 1, and includes the following steps:

[0051] S1. The operator takes his place, performs equipment status inspection, and conducts a comprehensive inspection of the device for qualified inspection and fault diagnosis of the temperature-averaging plate type radiator to ensure that all equipment is in normal working condition;

[0052] S2. Fill the constant temperature water tank according to the water level line and inject hot water according to the hot water temperature to provide a constant temperature water environment for the temperature equalizer to be tested;

[0053] S3. Power on all equipment, including computers, thermal imagers, constant temperature water tanks, and cooling fans;

[0054] S4. Check the temperature of the thermostatic water tank and start the test after confirming that it reaches the preset temperature of 80°C and is stable;

[0055] S5. Use manual or mechanical hands to place the temperature-averaging board to be tested at the test position in the constant temperature water tank, and use the first iron rod, the second iron rod and the iron net at the bottom of the constant temperature water tank to support the temperature-averaging board to be tested. The temperature-averaging board to be tested is immersed in the constant temperature water upright, and the evaporation section of the temperature-averaging board to be tested is consistent with the immersion depth;

[0056] S6. Keep the temperature plate to be tested in the constant temperature water tank for 30 seconds to achieve heat dissipation equilibrium;

[0057] S7. Use a thermal imager to photograph the temperature balancing board to be tested to obtain an infrared thermal imaging picture;

[0058] S8. The infrared thermal imaging image is transferred to the computer via a USB cable, and the computer runs a pre-trained deep convolutional neural network model to extract the temperature distribution characteristics of the temperature equalizing plate to be tested, and obtain the output result of the deep convolutional neural network model;

[0059] S9. Based on the output result of the deep convolutional neural network model, determine whether the temperature equalizing board to be tested is a qualified product. If it is judged to be a qualified product, it will be processed as a qualified product. If it is not, it will be processed as an unqualified product. Then, the next temperature equalizing board to be tested will be tested until all the temperature equalizing boards to be tested have been tested.

[0060] Furthermore, the qualified products are processed by outputting a test certificate, which is then taken out manually or by a robot and transferred to a bellows for cooling and air drying, and then the product and the certificate are packed together into a qualified product box.

[0061] Furthermore, the defective product processing is that the deep convolutional neural network model outputs the fault type, and the labels of the faulty products include over-filling of working fluid, under-filling of working fluid, poor machining accuracy, deformation tolerance, rust, and unqualified sintered structure. The products are then taken out manually or by a robot and transferred to a bellows for cooling and air drying. After that, the products and fault reports are put into a defective product box for subsequent processing.

[0062] Furthermore, the deep convolutional neural network model is a CNN model.

[0063] Furthermore, the training method of the deep convolutional neural network model as a CNN model comprises the following steps:

[0064] Step 1. Preparation of infrared imaging data set: Collect infrared imaging data, including more than 1,000 labeled infrared imaging sample images. The data set images include labels of qualified products and unqualified products with failure causes. The label of qualified products is "1", and the labels of failed products include: too much working fluid filling "2", too little working fluid filling "3", poor machining accuracy: "4", deformation tolerance "5", rust "6" and unqualified sintered structure "7", and build an infrared imaging data set;

[0065] Step 2. Data preprocessing: Preprocess the infrared imaging data in the infrared imaging dataset, including image denoising, normalization, and resizing operations, to ensure data quality and improve the training efficiency and accuracy of the model;

[0066] Furthermore, infrared imaging data was accumulated through previous experiments using the same hardware equipment, and was classified and labeled. The total number of images is 1,600, of which 1,000 are normal data and the remaining 100 are each type of fault data.

[0067] Step 3. Data segmentation: The preprocessed infrared imaging dataset is divided into two subsets, namely the training set and the test set. The training set is used to build and train the CNN model, while the test set is used to evaluate the generalization ability of the model.

[0068] Step 4. CNN model construction: According to the characteristics of the training set, a convolutional neural network model is designed and constructed. The model includes multiple convolutional layers, pooling layers, and fully connected layers to extract features from infrared images.

[0069] Step 5. Feature extraction: extract features of the infrared images in the training set through convolutional layers and pooling layers;

[0070] Step 6. Model training: Using the extracted features, the CNN model is trained through the back propagation algorithm to optimize the model parameters until the model can effectively recognize the required features from the input image.

[0071] Step 7. Model testing and deployment: After reaching the predetermined training rounds, use the test set to perform performance testing on the trained CNN model;

[0072] Step 8. Detection rate evaluation: Evaluate the detection rate of the model on the test set to determine whether the model meets the qualification standard;

[0073] Step 9. Model deployment: If the detection rate of the model on the test set is qualified, the CNN model is deployed to the actual application scenario to achieve effective analysis and processing of infrared imaging data.

[0074] Through the above training process, the present invention provides an efficient and accurate deep neural network suitable for infrared imaging data analysis and fault judgment.

[0075] It should be noted that relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0076] Although the present application has been described above with reference to specific embodiments, various modifications may be made thereto and parts thereof may be replaced with equivalents without departing from the scope of the present application. In particular, as long as there is no structural conflict, the various features in the specific embodiments disclosed in the present application may be used in combination with each other in any manner, and the fact that these combinations are not exhaustively described in this specification is only for the sake of omitting space and saving resources. Therefore, the present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A device for quality inspection and fault diagnosis of a heat spreader, characterized in that: The invention comprises a computer (1), a thermal imager (3), and a constant temperature water tank (4); a first iron rod (6-1), a second iron rod (6-2), and an iron net (7) are arranged in the constant temperature water tank (4); the first iron rod (6-1), the second iron rod (6-2), and the iron net (7) work together to immerse the evaporation section of the temperature equalizing plate (5) to be tested in the constant temperature water tank (4), and the condensation section of the temperature equalizing plate (5) to be tested is exposed to air at normal temperature; The computer (1) and the thermal imager (3) are connected via a USB cable (2); the thermal imager (3) is placed at a certain distance outside the constant temperature water tank (4) and facing the front of the condensation section of the temperature equalizing plate (5) to be tested; The computer (1) is installed with a deep convolutional neural network model for performing qualification inspection and fault diagnosis on photos taken by the thermal imager.

2. A device for quality inspection and fault diagnosis of a heat spreader according to claim 1, characterized in that: A heat dissipation fan (8) is also provided on the outside of the constant temperature water tank (4), and the heat dissipation fan (8) is placed on the back side of the condensation section of the temperature equalizing plate (5) to be tested, the front side of the condensation section of the temperature equalizing plate (5) to be tested is the contact surface of the chip, and the back side of the condensation section of the temperature equalizing plate (5) to be tested is the surface with the heat dissipation grooves.

3. A device for quality inspection and fault diagnosis of a heat spreader according to claim 2, characterized in that: Constant temperature hot water is added to the constant temperature water tank (4) according to the water level line. After the evaporation section of the temperature equalizing plate (5) to be tested is placed in the constant temperature water tank (4) for a certain period of time, the thermal imager (3) takes a front view of the temperature equalizing plate (5) to be tested, and then transmits the view to the computer (1) via the USB connection cable (2).

4. The device for quality inspection and fault diagnosis of a heat spreader according to claim 3, characterized in that: The water level of the constant temperature water tank (4) is set to immerse the evaporation section of the temperature equalizing plate (5) to be tested, and the hot water temperature of the constant temperature water tank (4) is 80°C.

5. The device for quality inspection and fault diagnosis of a heat spreader according to claim 4, characterized in that: After the evaporation section of the temperature-averaging plate (5) to be tested is placed in the constant-temperature water tank (4) for 30 seconds, the thermal imager (3) takes a picture.

6. A method for quality inspection and fault diagnosis of a heat spreader, implemented by a device for quality inspection and fault diagnosis of a heat spreader according to any one of claims 1 to 5, characterized in that: The steps include: S1. The operator takes his place, performs equipment status inspection, and conducts a comprehensive inspection of the device for qualified inspection and fault diagnosis of the temperature-averaging plate type radiator to ensure that all equipment is in normal working condition; S2. Fill the constant temperature water tank with water according to the water level line and inject hot water according to the hot water temperature to provide a constant temperature water environment for the temperature equalizer to be tested; S3. Power on all equipment, including computers, thermal imagers, constant temperature water tanks, and cooling fans; S4. Check the temperature of the thermostatic water tank and start the test after confirming that it reaches the preset temperature of 80°C and is stable; S5. Use manual or mechanical hands to place the temperature-averaging board to be tested at the test position in the constant temperature water tank, and use the first iron rod, the second iron rod and the iron net at the bottom of the constant temperature water tank to support the temperature-averaging board to be tested. The temperature-averaging board to be tested is immersed in the constant temperature water upright, and the evaporation section of the temperature-averaging board to be tested is consistent with the immersion depth; S6. Keep the temperature plate to be tested in the constant temperature water tank for 30 seconds to achieve heat dissipation equilibrium; S7. Use a thermal imager to photograph the temperature balancing board to be tested to obtain an infrared thermal imaging picture; S8. The infrared thermal imaging image is transferred to the computer via a USB cable, and the computer runs a pre-trained deep convolutional neural network model to extract the temperature distribution characteristics of the temperature equalizing plate to be tested, and obtain the output result of the deep convolutional neural network model; S9. Based on the output result of the deep convolutional neural network model, determine whether the temperature equalizing board to be tested is a qualified product. If it is judged to be a qualified product, it will be processed as a qualified product. If it is not, it will be processed as an unqualified product. Then, the next temperature equalizing board to be tested will be tested until all the temperature equalizing boards to be tested have been tested.

7. A method for quality inspection and fault diagnosis of a vapor chamber heat sink according to claim 6, characterized in that: The qualified product processing is to output the test qualified certificate, which is then taken out manually or by a robot and transferred to a wind box for cooling and air drying, and then the product and the certificate are packed together into a qualified product box.

8. A method for quality inspection and fault diagnosis of a vapor chamber heat sink according to claim 7, characterized in that: The defective product processing is that the deep convolutional neural network model outputs the fault type. The labels of the faulty products include excessive working fluid filling, insufficient working fluid filling, poor machining accuracy, excessive deformation, rust, and unqualified sintered structure. The products are then taken out manually or by a robot and transferred to a bellows for cooling and air drying. The products and fault reports are then placed in a defective product box for subsequent processing.

9. A method for quality inspection and fault diagnosis of a vapor chamber heat sink according to claim 8, characterized in that: The deep convolutional neural network model is a CNN model.

Images