A solder joint quality inspection equipment and inspection method based on infrared thermal imaging technology

Abstract

The present invention provides a solder joint quality detection equipment and detection method based on infrared thermal imaging technology. The equipment of the present invention includes a detection module, a mechanical module, a control module, a task module and an evaluation module. The detection module is used to detect solder joints and transmit the detection data to the evaluation module for processing; the detection module is installed, the circuit board to be tested is clamped, and the circuit board to be tested is moved and positioned; the mechanical module and the detection module are controlled by the control module; the detection tasks are formulated and executed by the task module; and the detection data obtained by the detection module are processed by the evaluation module to analyze and judge the quality of solder joints. This equipment collects the temperature-time curve of the solder joint under pulse thermal excitation, uses a solder joint defect characteristic parameter extraction algorithm to output characteristic parameters that characterize the quality of the solder joint, and further uses a solder joint defect evaluation model to output the defect area percentage of the solder joint, which has the advantages of automated operation, high detection efficiency and accuracy.

Description

Technical Field

[0001] The invention belongs to the field of solder joint quality detection of circuit boards, and in particular relates to solder joint quality detection equipment and a detection method based on infrared thermal imaging technology. Background Art

[0002] During the production and storage of electronic products for spacecraft and military equipment, PCB circuit boards may suffer from poor solder joints due to various environmental factors during the mounting process and problems such as rust, oxidation, and contamination on the pad surface. Solder joint defects may cause the entire electronic product to be scrapped or the entire satellite power module to fail, affecting the stable operation of the spacecraft.

[0003] Traditional PCB circuit board solder joint quality inspection based on infrared thermal imaging technology has the following shortcomings: directly using temperature signals as defect judgment criteria, the defect identification signal of the temperature-time curve is weak; the laser excitation is inaccurate; the planning of solder joint path one by one is time-consuming and laborious; and it has the characteristics of slow detection speed and low detection accuracy. Summary of the invention

[0004] The technical problem solved by the present invention is: to overcome the shortcomings of the prior art and provide a solder joint quality inspection equipment and inspection method based on infrared thermal imaging technology, which has the functions of laser precision excitation, automatic solder joint identification, inspection path planning and intelligent defect identification, and effectively improves the inspection efficiency and accuracy.

[0005] The technical solution of the present invention is:

[0006] A solder joint quality inspection equipment based on infrared thermal imaging technology comprises a detection module, a mechanical module, a control module and an evaluation module; the mechanical module moves and positions a circuit board to be tested under the control of the control module, and moves the solder joint to be tested on the circuit board to be tested to the detection module for inspection; the detection module applies pulse thermal excitation to the solder joint to be tested and collects infrared thermal imaging data under the control of the control module, and transmits the infrared thermal imaging data to the evaluation module, wherein the infrared thermal imaging data is a temperature-time curve of the solder joint to be tested under pulse thermal excitation; the evaluation module obtains solder joint defect characteristic parameters from the infrared thermal imaging data of the solder joint to be tested, and obtains the defect area percentage of the solder joint according to the solder joint defect characteristic parameters; the solder joint defect characteristic parameters are obtained in the following manner:

[0007] (1) extracting the excess temperature-time curve of the temperature drop section from the temperature-time curve of the solder joint to be tested;

[0008] (2) performing a fast Fourier transform on the excess temperature-time curve of the temperature drop section to obtain a phase-frequency curve, and taking the minimum value of the phase as the first characteristic parameter;

[0009] (3) The excess temperature curve in the temperature drop section is fitted with a polynomial using the least square method, the first-order derivative of the fitted polynomial is calculated, and the maximum value of the first-order derivative is used as the second characteristic parameter;

[0010] (4) taking the logarithm of the excess temperature of the excess temperature time curve of the temperature drop section to obtain the logarithmic excess temperature time curve of the temperature drop section, and on this basis, using the least square method to perform polynomial fitting, calculating the first-order derivative of the polynomial fitted, and taking the maximum value of the first-order derivative as the third characteristic parameter;

[0011] (5) According to the excess temperature-time curve of the cooling stage, the regularized apparent heat absorption coefficient is calculated and its logarithm is taken, and the minimum value of the logarithmic regularized apparent heat absorption coefficient is taken as the fourth characteristic parameter.

[0012] Preferably, the evaluation module uses a solder point defect evaluation model to obtain the defect area percentage of the solder point according to the solder point defect characteristic parameters. The solder point defect evaluation model is established by a BP neural network algorithm, specifically:

[0013] (31) Prefabricate solder joints with different defect area percentages;

[0014] (32) collecting infrared thermal imaging data of each solder joint under pulse thermal excitation to obtain a first characteristic parameter, a second characteristic parameter, a third characteristic parameter, and a fourth characteristic parameter;

[0015] (33) constructing a BP neural network model, including an input layer, a hidden layer and an output layer, wherein the input layer parameters are the first characteristic parameter, the second characteristic parameter, the third characteristic parameter and the fourth characteristic parameter, and the number of nodes is 4; the hidden layer is used to process the nonlinear relationship between input and output, and the output layer parameter is the defect area percentage of the solder joint, and the number of nodes is 1;

[0016] (34) The prefabricated solder joints are used as samples to train the BP neural network model until its fitting error meets a preset threshold. The trained BP neural network model is the solder joint defect evaluation model.

[0017] Preferably, it also includes a task module, which reads the imported GERBER design file of the circuit board to be tested, identifies all solder joints and generates position coordinate information, selects the solder joints to be tested from all the solder joints, and plans the detection path according to the position coordinate information of each solder joint to be tested.

[0018] Preferably, the task module plans the detection path according to the position coordinate information of each solder joint to be tested, specifically: the solder joints to be tested are divided into rows with a fixed row height, and the solder joints to be tested in each row are detected in turn, and each row is detected in order from small to large according to the distance from the solder joint to the rightmost side.

[0019] Preferably, the mechanical module includes a marble table, a product clamping table, an XY two-dimensional mobile platform, and a mounting frame; the marble table serves as an operating platform to ensure the overall stability of the detection equipment; the product clamping table is used to clamp the circuit board to be detected, and the product clamping table is installed on the XY two-dimensional mobile platform, and the XY two-dimensional mobile platform includes an X-axis slide and a Y-axis slide, which are used to drive the circuit board to be detected to move in the X-axis direction and the Y-axis direction respectively; the mounting frame is used to carry the detection module.

[0020] Preferably, the detection module includes a heat loading source, an infrared thermal imager and a digital microscope; the heat loading source is used to load pulse thermal excitation to the solder joint to be tested; the infrared thermal imager is used to collect the temperature-time curve of the solder joint to be tested under the pulse thermal excitation of the loading source; the digital microscope is used to collect and record the optical image of the solder joint to be tested and assist in positioning.

[0021] Preferably, the heat loading source is a laser whose minimum focal diameter is smaller than the size of the weld spot, including a first laser and a second laser respectively located on both sides of the infrared thermal imager, and the laser emission directions of the first laser and the second laser are opposite and the angle with the vertical direction is 45°.

[0022] Preferably, the control module includes a motion module controller and a detection module controller, the motion module controller includes: a motion control card, a servo motor, a driver and a grating ruler, the motion control card is used to send control instructions to the driver, the driver is used to control the servo motor to drive the X-axis slide and the Y-axis slide to move, and the grating ruler is used to perform position feedback; the detection module controller controls the heat loading source, infrared thermal imager and electron microscope.

[0023] Preferably, the digital microscope is mounted on the mounting frame via a horizontal movable shaft, the horizontal movable shaft is used to drive the digital microscope to move along the mounting frame, and the movement of the horizontal movable shaft is driven by a servo motor.

[0024] A solder joint quality detection method based on infrared thermal imaging technology comprises the following steps:

[0025] Step (1) clamping the circuit board to be tested onto the product clamping table so that two adjacent sides of the circuit board to be tested are in contact with the positioning baffles in the X direction and the Y direction;

[0026] Step (2) makes the center of the field of view of the infrared thermal imager, the center of the field of view of the digital microscope, the heating point of the heat loading source coincide with the solder joint to be tested, calibrates the position correspondence between the infrared thermal imager, the heat loading source, the digital microscope and the XY two-dimensional moving platform, and establishes a coordinate system;

[0027] Step (3) formulates the detection task, including detection parameter setting, measurement point selection and path planning.

[0028] Detection parameter setting: Set the detection parameters of the heat loading source and infrared thermal imager according to the type of solder joints on the circuit board to be tested;

[0029] Select test points: Identify all solder joints by reading the imported GERBER design file of the circuit board to be tested and generate the corresponding position coordinate information, and then generate a list of solder joints to be tested;

[0030] Path planning: Sort the solder points to be tested according to their position coordinate information, and automatically plan the detection path using a row-first search algorithm. Specifically, the solder points to be tested are divided into rows with a fixed row height, and the solder points to be tested in each row are detected in turn. Each row is detected in order from the smallest to the largest distance from the solder point to the rightmost side.

[0031] Step (4) inspecting each solder joint to be tested in turn according to the established inspection task, specifically: the XY two-dimensional mobile platform moves the solder joint to be tested to the inspection position, turns on the heat loading source for heating, the infrared thermal imager records the infrared thermal image data, and the digital microscope records the optical image;

[0032] Step (5) The evaluation module processes the detection data and outputs the defect area percentage of the solder joint.

[0033] The advantages of the present invention compared with the prior art are:

[0034] (1) The solder joint quality detection equipment and detection method based on infrared thermal imaging technology of the present invention determines the characteristic parameters that can characterize the quality of the solder joint by analyzing and studying the temperature-time curve of the solder joint under the excitation of a pulse heat source. By prefabricating samples with defects of different degrees as samples and using BP neural network for training and modeling, a solder joint defect judgment model is obtained, which effectively improves the solder joint detection accuracy;

[0035] (2) The solder joint quality inspection equipment and inspection method based on infrared thermal imaging technology of the present invention automatically identifies solder joints and generates position coordinate information by importing the GERBER design file of the circuit board, and at the same time comprehensively plans all solder joints on the circuit board to achieve optimal path planning, and inspects according to the planned path. A single inspection task can inspect 1,000 solder joints, achieving fast and efficient solder joint quality inspection, which is convenient for batch circuit board inspection;

[0036] (3) The solder joint quality inspection equipment and inspection method based on infrared thermal imaging technology of the present invention adopts an automated two-dimensional mobile platform combined with a grating ruler, and has the characteristics of automated inspection, fast speed, and high positioning accuracy, solving the problem of inaccurate positioning and inaccurate excitation of tiny solder joints;

[0037] (4) The solder joint quality inspection equipment and inspection method based on infrared thermal imaging technology of the present invention adopts dual-focusing lasers working in coordination, and can realize the inspection of solder joints on both sides of components without changing the direction of the lasers. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a structural schematic diagram of the solder joint quality detection equipment based on infrared thermal imaging technology of the present invention;

[0039] Figure 2 This is a schematic diagram of the shell and base structure of the solder joint quality inspection equipment based on infrared thermal imaging technology of the present invention; DETAILED DESCRIPTION

[0040] The present invention is further described below in conjunction with examples and drawings, but the protection scope of the present invention should not be limited thereto.

[0041] The solder joint quality inspection equipment based on infrared thermal imaging technology of the present invention comprises five parts: inspection module, mechanical module, control module, task module and evaluation module. The inspection module is used to inspect solder joints and transmit the inspection data to the evaluation module for processing; the mechanical module is used to install the inspection module, clamp the circuit board to be tested and realize the movement and positioning of the circuit board to be tested; the control module is used to control the mechanical module and the inspection module; the task module is used to formulate and execute inspection tasks; the evaluation module is used to process the inspection data obtained by the inspection module to analyze and judge the quality of solder joints.

[0042] The mechanical module includes: a marble table 1, a product clamping table 2, an XY two-dimensional moving platform 3, and a mounting frame 4.

[0043] Specifically, the marble table 1 serves as an operating platform to ensure the overall stability of the equipment system, effectively improving the shock resistance and stability of the equipment; the product clamping table 2 is used to clamp the circuit board to be tested, and the product clamping table is installed on the XY two-dimensional moving platform 3; the XY two-dimensional moving platform 3 is installed on the marble table 1, including an X-axis slide and a Y-axis slide, the X-axis slide is used to drive the circuit board to be tested to move in the X-axis direction, and the Y-axis slide is used to drive the circuit board to be tested to move in the Y-axis direction; the mounting frame 4 is installed on the marble table 1, and the mounting frame 4 is used to install the detection module, which can ensure the effective loading of the heat loading source 6 and the stability and repeatability of the signal acquisition of the infrared thermal imager 5.

[0044] In a specific embodiment, the mounting frame 4 is made of aluminum profile by machining.

[0045] To be more specific, an X-direction positioning baffle and a Y-direction positioning baffle are provided at one corner of the product clamping platform 2, which are parallel to the X-axis and the Y-axis respectively. When clamping, the adjacent sides of the circuit board to be tested are attached to the X-direction and Y-direction positioning baffles, and are installed on the product clamping platform 2 by magnetic screws. The accurate position of the circuit board to be tested in the entire mechanical coordinate system can be determined, so as to locate the designated solder joint position conveniently and quickly; the clamping platform 2 can adapt to the detection requirements of different product sizes.

[0046] In a specific embodiment, the product clamping table 2 is made of aluminum alloy by machining.

[0047] More specifically, the motion range of the X-axis slide or the Y-axis slide is ±250mm, with left and right limits, the 0 point position is at the center of the X-axis slide and the Y-axis slide, and the effective detection area is 500mm×500mm; the X-axis slide and the Y-axis slide can move simultaneously for fast and accurate positioning, the repeatability of the X-axis slide or the Y-axis slide is 0.02mm, and the movement speed can reach 150mm / s.

[0048] In a specific embodiment, if Figure 2 As shown, the mechanical module also includes a base 9 and a shell 10. The shell 10 is a closed box structure so that the temperature change of the circuit board solder joint during detection is not affected by the outside world; the base 9 serves as an operating platform for the detection equipment.

[0049] Specifically, the shell 10 is made of a steel frame and tempered glass through mechanical processing. The closed shell design can protect the internal operating mechanism and instruments. The tempered glass window is convenient for loading and unloading circuit boards. After closing, the detection system is in a closed dark box environment. The shell 10 is equipped with a high-brightness LED to provide the required lighting for the inside of the detection system and monitor the detection operation status; the base 9 is made of angle steel, and the four-corner support adjustment design ensures the levelness of the platform and the stability of operation.

[0050] The detection module includes a heat loading source 6 , an infrared thermal imager 5 and a digital microscope 7 .

[0051] Specifically, the heat loading source 6 is used to load heat flow to the solder joint to be tested; the infrared thermal imager 5 is used to collect infrared thermal imaging data of the solder joint to be tested, and the installation direction of the infrared thermal imager 5 is perpendicular to the marble table 1. The infrared thermal imager 5 is equipped with a height adjustment device to quickly adjust the distance; the digital microscope 7 is used to collect and record optical images of the solder joint and assist in positioning.

[0052] More specifically, the heat loading source 6 is a laser, which is focused through transmission to achieve energy focusing and focal reduction, and includes a first laser and a second laser, the first laser and the second laser are respectively located on both sides of the infrared thermal imager 5, the laser emission directions of the first laser and the second laser are opposite and the angle with the vertical direction is 45°; one of the lasers is turned on during detection, and when the laser blocks the heating of the solder joint to be tested, the other laser can be switched for heating, and the solder joints on both sides of the components can be detected without changing the direction of the laser, solving the problem of partial blocking of the measuring points under the condition of a single laser, which requires adjustment of the circuit board or direction; the minimum focal diameter of the laser is smaller than the solder joint size.

[0053] More specifically, the digital microscope 7 is mounted on a horizontal H-axis 8, which is mounted on the mounting frame 4 and is used to move the digital microscope 7 along the mounting frame 4. The zero point of the horizontal H-axis 8 is at the rightmost side of the mounting frame 4, and the positioning point is at the 0-point position of the mechanical module. When setting solder joint detection tasks, observing solder joints, and collecting images, the digital microscope 7 can be moved to the positioning point, and can be moved to the origin during solder joint detection, which can effectively avoid obstruction of data collection by the infrared thermal imager 5. The magnification of the digital microscope 7 is greater than or equal to 50 times.

[0054] In a specific embodiment, the horizontal H-axis 8 is formed by machining a lead screw.

[0055] The control module includes a motion module controller 12, a detection module controller 13 and a computer 11. The motion module controller 12 includes: a motion control card, a servo motor, a driver and a grating ruler. The motion control card is used to send control instructions to the driver, the driver is used to control the servo motor, the servo motor is used to drive the X-axis slide, the Y-axis slide and the horizontal H-axis 8 to move, and the grating ruler is used to perform position feedback; the detection module controller 13 controls the heat loading source 6, the infrared thermal imager 5 and the electron microscope 7; the computer 11 controls the motion module controller 12 and the detection module controller 13 through a control program.

[0056] Specifically, the servo motor is a high-power AC servo motor to ensure the power for the system operation; the grating ruler is an incremental grating ruler that can quickly feedback the position information of the XY two-dimensional mobile platform 3.

[0057] After the mechanical system is powered on and reset to zero, the zero point of the mechanical coordinate system can be obtained. After the artificially assisted infrared thermal imager 5 and the digital microscope 7 determine the coordinates of the feature points, a complete two-dimensional coordinate system can be established, and each solder point position on the circuit board can be completely matched with the coordinate system, providing position feedback for the control module, task module and evaluation module.

[0058] The task module includes a computer 11. The computer 11 uses a task formulation algorithm. The task formulation algorithm includes: a welding point recognition algorithm and a path planning algorithm.

[0059] The solder joint identification algorithm reads the imported GERBER design file of the circuit board to be tested, draws a multi-layer overlay view of the circuit board, and identifies all solder joints and generates position coordinate information by analyzing the structures of the routing layer, pad layer, silk screen layer, etc.

[0060] The path planning algorithm sorts the solder joints to be tested according to the solder joint position coordinate information, and automatically plans the detection path using a row-first search algorithm, specifically: the solder joints to be tested are divided into rows with a row height of 50 mm, and each row is sorted in ascending order according to the distance from the solder joint to the rightmost side.

[0061] The specific steps of developing a testing task are: setting testing parameters, selecting measuring points and planning testing paths.

[0062] The detection parameter setting is specifically as follows: setting the detection parameters of the heat loading source 6 and the infrared thermal imager 5 according to the type of solder joints of the circuit board to be tested;

[0063] The specific steps of selecting measuring points are as follows: all solder joints are identified through solder joint recognition algorithm and position coordinate information is generated; all solder joints are selected by default; the identified solder joints are selected and deleted as needed to generate a list of solder joints to be detected; the solder joints are further moved to the inspection position through manual control and added to the list of solder joints to be detected; the task module can support 32 layers and 1,000 solder joints can be displayed simultaneously.

[0064] Furthermore, the task module can also read the planned inspection list file, generate a list of solder joints to be inspected based on the solder joint information provided in the file, and perform editing, selection, and deletion operations.

[0065] In a specific embodiment, the inspection list file format is a csv file, and the welding point information includes a measuring point number, a measuring point X coordinate, a measuring point Y coordinate, a measuring point width, and a measuring point height.

[0066] Planning the inspection path specifically includes: using a path planning algorithm to form a solder joint inspection plan.

[0067] The task module performs the inspection task specifically as follows: execute the inspection plan, control the XY two-dimensional translation stage 3 to drive the circuit board to the specified position according to the established measurement point list, turn on the laser 6 for heating, the infrared thermal imager 5 records the thermal image sequence, the digital microscope 7 records the optical image, and transmits the measurement data to the evaluation module.

[0068] Furthermore, anchor point calibration is required during the inspection. The anchor point is the first measuring point in the measuring point list. The center area of ​​the solder joint is controlled to be at the center point 0 of the field of view of the digital microscope 7, and the laser 6 is turned on to accurately excite the laser 6 to the center position of the solder joint, thereby realizing the calibration of the anchor point positioning information. During the measurement, the center of the field of view of the thermal imager, the laser heating point and the solder joint to be measured coincide with each other.

[0069] The evaluation module includes a computer 11, and the computer 11 adopts a solder joint quality evaluation algorithm.

[0070] The solder joint quality evaluation algorithm evaluates the solder joint quality based on the infrared thermal image data of the solder joint.

[0071] The infrared thermal imaging data of the solder joint is processed, characteristic parameters characterizing the quality of the solder joint are calculated, the characteristic parameters are input into the intelligent evaluation model of the solder joint quality, and the solder joint quality evaluation result is output, and the solder joint quality evaluation result is the defect area percentage of the solder joint.

[0072] Specifically, the characteristic parameters characterizing the quality of solder joints include the phase of the first low frequency of the pulse phase method, the maximum absolute value of the seventh-order polynomial fitting derivative of the trend analysis method, the maximum absolute value of the third-order polynomial fitting derivative of the trend analysis method, and the minimum logarithmic regularized apparent heat absorption coefficient.

[0073] Further, the phase of the first low frequency of the pulse phase method is specifically:

[0074] Extract the excess temperature-time curve of the cooling stage, perform FFT transformation, and obtain the phase-frequency curve. The minimum value of the phase is the phase of the first low frequency of the pulse phase method.

[0075] The maximum absolute value of the seventh-order polynomial fitting derivative of the trend analysis method is specifically:

[0076] Extract the excess temperature-time curve of the cooling stage, fit it with a seventh-order polynomial, calculate the first-order derivative of the fitted polynomial, and the maximum value of the first-order derivative is the maximum absolute value of the seventh-order polynomial fitting derivative of the trend analysis method;

[0077] The maximum absolute value of the derivative of the third-order polynomial fitting of the trend analysis method is:

[0078] The excess temperature-time curve of the cooling stage is extracted, the logarithm of the excess temperature is taken to obtain the logarithmic excess temperature-time curve, a third-order polynomial fitting is performed on it based on this coordinate system, and the first-order derivative of the fitted polynomial is calculated. The maximum value of the first-order derivative is the maximum absolute value of the derivative of the third-order polynomial fitting of the trend analysis method;

[0079] The minimum value of the logarithmic regularized apparent heat absorption coefficient is:

[0080] The excess temperature-time curve of the cooling section is extracted. According to the calculation formula of the regularized apparent heat absorption coefficient, the regularized apparent heat absorption coefficient-time curve is obtained. The logarithm of the horizontal and vertical coordinates is taken at the same time. The logarithmic regularized apparent heat absorption coefficient-logarithmic time curve is taken, and the minimum value of the logarithmic regularized apparent heat absorption coefficient is taken as the characteristic value.

[0081] The definition and calculation method of the regularized apparent heat absorption coefficient refer to CN109581201 B "Method for determining cold solder joints based on regularized apparent heat absorption coefficient".

[0082] The above four characteristic parameters are determined through a large number of simulations and experiments, and can effectively characterize the degree of cold solder joints.

[0083] A batch of solder joints with different degrees of cold solder joint defects were prefabricated as training samples, the above four characteristic parameters were used as input parameters, and the defect area percentage of the solder joints was used as output parameters to construct a BP neural network model.

[0084] The BP neural network includes an input layer, a hidden layer and an output layer. The input layer is an input neuron, which is used to receive various characteristic parameters representing the degree of cold soldering. The hidden layer is a hidden neuron, which is used to process the nonlinear relationship between input and output. The output layer is an output neuron, which is used to output the degree of cold soldering of solder joints.

[0085] The number of nodes in the input layer of the present invention is 4, the number of nodes in the output layer is 1, and the reference formula for determining the optimal number of hidden layer nodes is: Among them: q is the number of nodes in the hidden layer, m is the number of nodes in the input layer, n is the number of nodes in the output layer, and a is any constant from 1 to 10. According to the above formula and multiple experiments, the number of nodes in the hidden layer is preferably 10. Therefore, the topological structure of the BP neural network model is determined to be 4-10-1.

[0086] In order to ensure a faster adjustment speed, the initial weights are randomly selected in the interval [-1,1]; in order to make the system have good stability, the learning rate is selected in the range of 0.01 to 0.7; based on multiple training results, the learning rate is set to 0.01.

[0087] The activation function used from the input layer to the hidden layer is the hyperbolic tangent S-type function (Tan-Sigmoid), and its formula is:

[0088]

[0089] When the Tan-Sigmoid function is far away from the domain, the function curve trend is relatively gentle and the function value changes slowly, so the sample data in these saturation areas will result in a lower learning rate. In addition, the input parameters of the BP neural network of this patent are of different sizes and are not completely in the same order of magnitude, which is not conducive to the training of the neural network. Therefore, it is necessary to preprocess the sample data so that it falls within the interval of [-1,1], thereby unifying the dimensions, avoiding neuron saturation, and accelerating the learning speed of the BP neural network.

[0090] The specific processing methods are:

[0091]

[0092]

[0093] in, is the preprocessed data, X max and X min are the maximum and minimum values ​​of the sample input parameters respectively.

[0094] The samples are trained to obtain the intelligent recognition model of solder joint cold soldering defects, the model is verified using a test set, and the model is optimized and iterated based on the verification results until the model accuracy meets the requirements.

[0095] A solder joint quality detection method based on infrared thermal imaging technology, the steps are as follows:

[0096] Step (1), power on self-test, check the connection status between the mechanical module and the detection module and the control module;

[0097] Step (2), clamping the circuit board to be tested onto the product clamping table 2, so that two adjacent sides of the circuit board to be tested are in contact with the X-direction and Y-direction positioning baffles;

[0098] Step (3), standard point calibration, so that the center of the field of view of the thermal imager 5, the center of the field of view of the digital microscope 7, the heating point of the laser 6 and the measured welding point coincide, calibrate the position correspondence between the infrared thermal imager 5, the laser 6, the digital microscope 7 and the XY two-dimensional mobile platform 3, and establish a coordinate system;

[0099] Step (4), formulating the inspection task, specifically including setting the inspection parameters, selecting the inspection points and planning the inspection path;

[0100] Detection parameter setting: setting the detection parameters of the laser 6 and the infrared thermal imager 5 according to the type of solder joints on the circuit board to be tested;

[0101] Select measuring points: Identify all solder joints by reading the imported GERBER design file of the circuit board to be tested and generate position coordinate information or select measuring points by manual control;

[0102] Path planning: Sort the solder joints to be tested according to the position coordinate information of the selected test points, and automatically plan the detection path using the row-first search algorithm. Specifically, the solder joints to be tested are divided into rows with a row height of 50 mm, and each row is sorted in order from small to large according to the distance from the solder joint to the far right, and a solder joint detection plan is generated;

[0103] Step (5), executing the inspection task, inspecting each solder joint to be inspected in turn according to the formulated inspection task and the solder joint inspection plan, specifically: the XY two-dimensional mobile platform 3 moves the solder joint to be inspected to the inspection position, turns on the laser 6 for heating, the infrared thermal imager 5 records the infrared thermal image data, and the digital microscope 7 records the optical image;

[0104] Step (6), solder joint quality evaluation: the evaluation module processes the test data and outputs the solder joint quality test results.

[0105] The equipment of the present invention can realize automatic detection and evaluation of cold soldering defects when the inspected product is not powered on and the temperature rise of the soldering point does not exceed 80° C. The single-point automatic detection time is less than 5 seconds.

[0106] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

Claims

1. A solder joint quality inspection equipment based on infrared thermal imaging technology, characterized in that: It includes a detection module, a mechanical module, a control module and an evaluation module; the mechanical module moves and positions the circuit board to be tested under the control of the control module, and moves the solder joints to be tested on the circuit board to the detection module for detection; the detection module applies pulse thermal excitation to the solder joints to be tested and collects infrared thermal imaging data under the control of the control module, and transmits the infrared thermal imaging data to the evaluation module, wherein the infrared thermal imaging data is a temperature-time curve of the solder joints to be tested under the pulse thermal excitation; The evaluation module obtains solder joint defect characteristic parameters from the infrared thermal imaging data of the solder joint to be tested, and obtains the defect area percentage of the solder joint according to the solder joint defect characteristic parameters; the solder joint defect characteristic parameters are obtained in the following manner: (1) extracting the excess temperature-time curve of the temperature drop section from the temperature-time curve of the solder joint to be tested; (2) performing a fast Fourier transform on the excess temperature-time curve of the temperature drop section to obtain a phase-frequency curve, and taking the minimum value of the phase as the first characteristic parameter; (3) The excess temperature curve in the temperature drop section is fitted with a polynomial using the least square method, the first-order derivative of the fitted polynomial is calculated, and the maximum value of the first-order derivative is used as the second characteristic parameter; (4) taking the logarithm of the excess temperature of the excess temperature time curve of the temperature drop section to obtain the logarithmic excess temperature time curve of the temperature drop section, and on this basis, using the least square method to perform polynomial fitting, calculating the first-order derivative of the polynomial fitted, and taking the maximum value of the first-order derivative as the third characteristic parameter; (5) According to the excess temperature-time curve of the cooling stage, the regularized apparent heat absorption coefficient is calculated and its logarithm is taken, and the minimum value of the logarithmic regularized apparent heat absorption coefficient is taken as the fourth characteristic parameter; The evaluation module uses a solder point defect evaluation model to obtain the defect area percentage of the solder point according to the solder point defect characteristic parameters. The solder point defect evaluation model is established by a BP neural network algorithm, specifically: (31) Prefabricate solder joints with different defect area percentages; (32) collecting infrared thermal imaging data of each solder joint under pulse thermal excitation to obtain a first characteristic parameter, a second characteristic parameter, a third characteristic parameter, and a fourth characteristic parameter; (33) constructing a BP neural network model, including an input layer, a hidden layer and an output layer, wherein the input layer parameters are the first characteristic parameter, the second characteristic parameter, the third characteristic parameter and the fourth characteristic parameter, and the number of nodes is 4; the hidden layer is used to process the nonlinear relationship between the input layer parameters and the output layer parameters, and the output layer parameters are the defect area percentage of the solder joint, and the number of nodes is 1; (34) The prefabricated solder joints are used as samples to train the BP neural network model until its fitting error meets a preset threshold. The trained BP neural network model is the solder joint defect evaluation model.

2. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 1 is characterized in that: It also includes a task module, which reads the imported GERBER design file of the circuit board to be tested, identifies all solder joints and generates position coordinate information, selects the solder joints to be tested from all the solder joints, and plans the detection path according to the position coordinate information of each solder joint to be tested.

3. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 2 is characterized in that: The task module plans the detection path according to the position coordinate information of each solder joint to be tested, specifically: the solder joints to be tested are divided into rows with a fixed row height, and the solder joints to be tested in each row are detected in turn, and each row is detected in order from small to large according to the distance from the solder joint to the rightmost side.

4. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 1 is characterized in that: The mechanical module comprises a marble table (1), a product clamping table (2), an XY two-dimensional mobile platform (3), and a mounting frame (4); the marble table (1) serves as a running platform to ensure the overall stability of the detection equipment; the product clamping table (2) is used to clamp the circuit board to be detected, and the product clamping table (2) is installed on the XY two-dimensional mobile platform (3), and the XY two-dimensional mobile platform (3) comprises an X-axis slide and a Y-axis slide, which are respectively used to drive the circuit board to be detected to move in the X-axis direction and the Y-axis direction; an X-direction positioning baffle and a Y-direction positioning baffle are provided at one corner of the product clamping table (2), which are parallel to the X-axis and the Y-axis respectively; the mounting frame (4) is used to carry the detection module.

5. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 4 is characterized in that: The detection module comprises a heat loading source (6), an infrared thermal imager (5) and a digital microscope (7); the heat loading source (6) is used to load pulse thermal excitation to the solder joint to be tested; the infrared thermal imager (5) is used to collect the temperature-time curve of the solder joint to be tested under the pulse thermal excitation of the loading source (6); and the digital microscope (7) is used to collect and record the optical image of the solder joint to be tested and to assist in positioning.

6. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 5 is characterized in that: The heat loading source (6) is a laser with a minimum focal diameter smaller than the size of the welding spot, comprising a first laser and a second laser respectively located on both sides of the infrared thermal imager (5), wherein the laser emission directions of the first laser and the second laser are opposite and the angle with the vertical direction is 45°.

7. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 6 is characterized in that: The control module comprises a motion module controller (12) and a detection module controller, wherein the motion module controller (12) comprises: a motion control card, a servo motor, a driver and a grating ruler, wherein the motion control card is used to send a control instruction to the driver, the driver is used to control the servo motor to drive the X-axis slide and the Y-axis slide to move, and the grating ruler is used to provide position feedback; the detection module controller (13) controls the heat loading source (6), the infrared thermal imager (5) and the electron microscope (7).

8. The solder joint quality inspection equipment based on infrared thermal imaging technology according to claim 7 is characterized in that: The digital microscope (7) is mounted on the mounting frame (4) via a horizontal movable shaft (8); the horizontal movable shaft (8) is used to drive the digital microscope (7) to move along the mounting frame (4); and the movement of the horizontal movable shaft (8) is driven by a servo motor.

9. A solder joint quality inspection method based on infrared thermal imaging technology using the solder joint quality inspection equipment based on infrared thermal imaging technology as claimed in claim 8, characterized in that: The steps include: Step (1) clamping the circuit board to be tested onto the product clamping table (2) so that two adjacent sides of the circuit board to be tested are in contact with the positioning baffles in the X direction and the Y direction; Step (2) makes the center of the field of view of the infrared thermal imager (5), the center of the field of view of the digital microscope (7), the heating point of the heat loading source (6) coincide with the solder joint to be tested, calibrates the position correspondence between the infrared thermal imager (5), the heat loading source (6), the digital microscope (7) and the XY two-dimensional moving platform (3), and establishes a coordinate system; Step (3) formulates the detection task, including detection parameter setting, measurement point selection and path planning. Setting detection parameters: setting detection parameters of the heat loading source (6) and the infrared thermal imager (5) according to the type of solder joints of the circuit board to be tested; Select test points: Identify all solder joints by reading the imported GERBER design file of the circuit board to be tested and generate the corresponding position coordinate information, and then generate a list of solder joints to be tested; Path planning: Sort the solder points to be tested according to their position coordinate information, and automatically plan the detection path using a row-first search algorithm. Specifically, the solder points to be tested are divided into rows with a fixed row height, and the solder points to be tested in each row are detected in turn. Each row is detected in order from the smallest to the largest distance from the solder point to the rightmost side. Step (4) inspects each solder joint to be inspected in turn according to the established inspection task, specifically: the XY two-dimensional moving platform (3) moves the solder joint to be inspected to the inspection position, the heat loading source (6) is turned on for heating, the infrared thermal imager (5) records the infrared thermal image data, and the digital microscope (7) records the optical image; Step (5) The evaluation module processes the detection data and outputs the defect area percentage of the solder joint.

Images