Visual servo test method and test equipment for circuit board diagnosis, computer equipment and medium

Abstract

The invention relates to a visual servo test method and test equipment for circuit board diagnosis, computer equipment and a medium. The method comprises the steps that the posture of a mechanical arm is adjusted based on teaching operation and the situation of obstacles around a measurement point of a to-be-tested circuit board; driving the mechanical arm to drive the test probe to move, and in the moving process of the test probe, circularly executing visual servo positioning operation based on the template image of the measurement point until the positioning error between the test probe and the measurement point meets a preset precision condition; when the precision condition is met, the test probe is controlled to be fed in the axis direction according to the contact pressure measurement value, collected by the pressure feedback module, of the test probe and the measurement point until the duration of the contact pressure measurement value in the preset pressure interval reaches the preset duration; and when the test probe stops feeding, the electrical measurement tool is controlled to measure electrical parameters of the measurement point. According to the invention, independent, accurate and safe electrical diagnosis of the faulted PCB can be realized without accurate geometric information and a special fixture.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a visual servo testing method, testing equipment, computer equipment, and medium for circuit board diagnostics. Background Technology

[0002] In the repair and manufacturing of electronic equipment, fault diagnosis of printed circuit boards (PCBs) is crucial. Currently, the mainstream automated testing technologies are mainly divided into bed-of-needle testing and flying probe testing.

[0003] Bed-of-needle testing achieves multi-point parallel contact through customized fixtures, which is highly efficient and suitable for mass production testing. However, its dedicated fixtures are expensive, have long production cycles, and each fixture corresponds to only one PCB layout, resulting in extremely poor flexibility and making it completely unsuitable for testing small batches and various types of faulty boards in repair workshops.

[0004] Flying probe testing uses programmable moving probes, eliminating the need for specialized fixtures and significantly improving flexibility. However, its accuracy typically relies on precise computer-aided design (CAD) files (such as Gerber files) of the PCB under test to obtain the accurate coordinates of each test point. In repair scenarios, especially for used, outdated, or modified faulty boards, this precise geometric information is often missing or unreliable, severely limiting the effectiveness of flying probe testing. Furthermore, the probe trajectory and posture of flying probe testers are usually relatively fixed, making it difficult to actively and flexibly avoid obstacles such as heat sinks, large components, or connecting wires on the PCB, potentially leading to testing failures or damage.

[0005] At the microscopic level of testing, whether it's a bed of needles or a flying probe, the contact between the probe and the PCB pads largely relies on a simple spring mechanism for cushioning. This passive method cannot sense and control the contact force in real time, which may result in excessive pressure damaging the precision pads, or insufficient pressure causing unstable contact, introducing measurement noise, and affecting diagnostic accuracy.

[0006] In recent years, machine vision has been introduced into PCB inspection to identify appearance issues such as missing components and solder joint defects. However, it cannot obtain key electrical parameters (such as current, voltage, resistance, and continuity) and cannot independently diagnose electrical faults. Some studies have attempted to use vision for guided testing, but these are mostly limited to features with regular shapes (such as vias) and have not fully considered the complex clamping errors, lighting changes, and the closed-loop integration of vision positioning, force control contact, and electrical measurement in practical applications. Summary of the Invention

[0007] In response to the above-mentioned deficiencies or disadvantages, this application provides a visual servo testing method, testing equipment, computer equipment, and medium for circuit board diagnostics.

[0008] This application provides a visual servo testing method for circuit board diagnostics according to a first aspect, the method comprising: The posture of the robotic arm equipped with the test probe is adjusted based on the operator's teaching operation and the surrounding obstacles on the circuit board under test, so that the test probe avoids obstacles as it approaches the measurement point. The robotic arm is driven to move the test probe. During the movement of the test probe, the visual servo positioning operation is performed cyclically based on the template image of the pre-stored measurement point until the positioning error between the test probe and the measurement point meets the preset accuracy condition. When the positioning error meets the accuracy requirements, the test probe is controlled to feed along its axial direction based on the contact pressure measurement value of the test probe and the measurement point collected by the pressure feedback module in the test probe until the contact pressure measurement value collected by the pressure feedback module is in the preset pressure range for a preset duration. When the test probe stops feeding, the electrical measuring tool connected to the test probe measures the electrical parameters of the measurement point.

[0009] In some embodiments, the execution process of the visual servo positioning operation includes: The camera mounted on the test probe captures images of the area containing the measurement points. The visual servoing algorithm identifies two marker points on the test probe in the regional image, and the axis of the test probe is fitted based on the coordinates of the two marker points. A template matching algorithm is used to search for the region in the region image that has the highest correlation with the template image, and the location of the measurement point in the region image is determined based on the region. The positioning error between the test probe and the measurement point is determined based on the axis of the test probe and the position of the measurement point. The positioning error is converted into a physical space error, and a running compensation command is generated based on the physical space error. The running compensation command drives the robotic arm to move, thereby reducing the positioning error.

[0010] In some embodiments, the method further includes: During the movement of the test probe, monitor whether the displacement error condition, force error condition, and iterative convergence condition are met. When the displacement error condition, force error condition, or iteration convergence condition is met, the current measurement process is terminated and an alarm is issued. The displacement error condition refers to the displacement of the tool center point of the test probe relative to the initial position at the start of visual servoing being greater than the preset error radius. Force error condition refers to the contact pressure measurement value collected by the pressure feedback module being greater than the preset force threshold; The iterative convergence condition refers to the number of iterations in the visual servo positioning operation reaching the preset maximum number of iterations.

[0011] In some embodiments, the method uses multiple sets of measuring devices, each set of measuring devices including two test probes defined as a positive probe and a negative probe, respectively; the positive probe and the negative probe are respectively connected to the positive and negative terminals of an electrical measuring tool; When the circuit board under test is divided into multiple non-overlapping test areas, a corresponding test area is assigned to each group of measuring devices, and each group of measuring devices is controlled to perform measurements within its assigned test area.

[0012] In some embodiments, adjusting the posture of the robotic arm equipped with the test probe based on operator teaching and the surrounding obstacles at the measurement point on the circuit board under test includes: In response to the operator's teaching operation, the control robot arm moves the test probe, defined as the positive probe, to a safe plane near the measurement point; Based on visual analysis of the environment surrounding the measurement point or operator input, obstacle information is identified; Adjust the end effector posture of the robotic arm based on obstacle information.

[0013] In some embodiments, the method further includes: A comprehensive test report is generated based on the electrical parameters, contact pressure curves, visual servo data, and corresponding area images of the measurement points obtained during the measurement process.

[0014] In some embodiments, the method further includes: The measured electrical parameters are compared with the pre-stored reference parameters, and abnormal measurement points are marked based on the comparison results; Fault-assisted location is performed based on the contact pressure curve characteristics and image features of abnormal measurement points.

[0015] According to a second aspect, this application provides a vision servo testing device for circuit board diagnostics, the device comprising: At least two robotic arms; The number of test probes is the same as that of the robotic arms. Each test probe is fixed to the end of a robotic arm. The test probe includes a housing, a test needle set at the front end of the housing, a vision positioning module set outside the housing, and a pressure feedback module integrated in the loading mechanism of the test needle. The vision positioning module includes two cameras. A worktable, located within the range of motion of all robotic arms, is used to hold the circuit board to be tested. Electrical measurement unit; A visualization screen is used to display the electrical parameters, contact pressure curves, and / or visual servo data acquired by the test probe during the measurement process; Control system; the control system communicates with all robotic arms, all test probes, all pressure measurement modules, and electrical measurement units.

[0016] In some embodiments, the control system is used to perform the method provided in any embodiment of the first aspect.

[0017] According to a third aspect, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the visual servo testing method for circuit board diagnostics provided in any embodiment of the first aspect.

[0018] This application provides a computer device according to a fourth aspect, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executed, implements the visual servo testing method for circuit board diagnostics provided in any embodiment of the first aspect.

[0019] This application constructs a collaborative operation process centered on real-time visual feature guidance and closed-loop control, which enables autonomous, accurate, and safe electrical diagnosis of complex faulty PCBs without the need for precise geometric information and special fixtures.

[0020] First, by combining a teaching template with a real-time visual servoing strategy, the contradiction between flexibility and precise positioning is resolved collaboratively. Specifically, template images are generated through operator teaching operations, thus abandoning the reliance on precise PCB CAD files and instead relying on more universal visual features as the recognition benchmark. This allows for testing of any PCB model, broadening its applicability. Considering that a rough teaching template alone cannot guarantee contact accuracy, this application also iteratively performs visual servoing positioning operations based on the template image. This operation, in conjunction with the template image, forms a dynamic observation-comparison-adjustment closed loop. The system continuously acquires real-time scene images (i.e., area images) through a camera, matches and compares them with the template image, and drives the robotic arm to compensate for deviations. This loop iteratively optimizes the operator's initial, imprecise teaching input into more precise positioning, achieving positioning accuracy comparable to traditional methods while maintaining flexibility without the need for fixtures and CAD files.

[0021] Secondly, this application collaboratively ensures the reliability of the measurement point contact and the safety of the circuit board through the sequential connection of visual precision aiming and force-controlled compliance. Building upon this, the crucial transition from positioning to reliable electrical contact is achieved by controlling the feed based on pressure feedback and determining the duration of stable contact. This method does not operate independently; rather, it starts with the precise position provided by the visual servo system, using pressure feedback to form a second closed loop, controlling the probe to complete the final contact smoothly and finely adjusting the contact force within a safe range. This sequential coordination of vision-guided initial aiming and force-controlled fine-tuning achieves stable and safe physical contact, avoiding PCB damage or measurement distortion caused by blind contact in traditional methods.

[0022] Ultimately, based on the aforementioned technical means, the electrical measuring tools are controlled to complete the electrical parameter measurement operations at the measurement points, realizing the automation and efficiency of the PCB board diagnostic process. From obstacle avoidance posture setting to visual servo precise positioning, and then to force-controlled compliant contact, the output of each step serves as a condition for initiating the next step, ultimately triggering the electrical measurement operation. This series of designs integrates the previously experience-dependent and fragmented process of observing the board, aligning the points, carefully contacting, and measuring with instruments into a coherent, autonomous, and condition-driven intelligent workflow. The operator's role is simplified to initial teaching and definition; once the process is started, the system can automatically and reliably cycle through all points for testing, thus fundamentally meeting the core needs of repair workshops for efficient, accurate, and reliable diagnosis of various types and small batches of faulty PCBs. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of a test probe in one or more embodiments of this application; the markings in the figure are explained as follows: 1-test probe base, 2-LED light source, 3-high-definition camera, 4-internal integrated piezoresistive sensing test probe, 5-test probe; Figure 2 This is a schematic diagram of a test probe mounted on a UR5e robotic arm in one or more embodiments of this application; the markings in the diagram are explained as follows: 6-UR5e robotic arm, 7-test probe; Figure 3 This is a flowchart of a visual servo testing method for circuit board diagnostics, which is shown in one or more embodiments of this application. Figure 4 This is a flowchart of a visual servo positioning operation for circuit board diagnostics in one or more embodiments of this application; Figure 5 This is a schematic diagram of the internal structure of a computer device according to one or more embodiments of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the described embodiments are merely some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0025] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0026] In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances. Furthermore, in the description of this application, unless otherwise stated, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0027] This application provides a vision servo testing device for circuit board diagnostics. In some embodiments, the vision servo testing device for circuit board diagnostics includes: (1) At least two robotic arms; (2) The same number of test probes as the robotic arm; Each test probe is fixed to the end of a robotic arm. The test probe includes a housing, a test needle located at the front end of the housing, a vision positioning module located outside the housing, and a pressure feedback module integrated into the loading mechanism of the test needle. The vision positioning module includes two cameras. (3) A worktable set within the range of motion of all robotic arms is used to fix the circuit board to be tested; (4) Electrical measurement unit; (5) A visualization screen for displaying electrical parameters, contact pressure curves and / or visual servo data acquired by the test probe during the measurement process; (6) Control system; The control system communicates with all robotic arms, all test probes, all pressure measurement modules, and electrical measurement units.

[0028] The end effector of the robotic arm is rigidly connected to the test probe, allowing the probe to move flexibly and adjust its orientation in three-dimensional space, meeting the probe access requirements of PCB measurements in different directions. A 6-DOF robotic arm, such as the UR5e model, is selected. Six degrees of freedom (DOF) enable the end effector of the robotic arm (in this embodiment, the measuring tool, i.e., the test probe), to have complete positioning and orientation capabilities. A degree of freedom refers to the number of independent movements of the end effector in three-dimensional space. Each DDF corresponds to an independent joint or axis of motion. The 6 DDFs include 3 translational degrees of freedom (translational movements along the X, Y, and Z axes perpendicular to each other in space) and 3 rotational degrees of freedom (rotational movements about the X, Y, and Z axes).

[0029] A test probe is an integrated measurement tool that is fixedly mounted at the end of a robotic arm, combining vision, force sensing, and electrical contact functions. The test probe includes a housing and several components integrated within it. These components include a test probe located at the front of the housing, a vision positioning module located outside the housing, and a pressure feedback module integrated into the loading mechanism of the test probe. The housing refers to the external mechanical structure of the test probe, which can be a slender conical or cylindrical 3D-printed modular structure used to fix and protect all internal components. The test probe is a conductive probe with a slender cylindrical design, used to form physical contact and electrical connection with the measurement point on the circuit board under test; its surface can be gold-plated to reduce contact resistance; the tail of the probe is threaded to the tip of the housing for easy replacement. The test probe contains an internal loading mechanism. A loading mechanism is a mechanical device that provides axial cushioning and feed force to the test probe. It typically includes a guide sleeve, a preload spring, and related fasteners. It allows the test probe to have a certain compression stroke (e.g., 2-3 mm) when it contacts the measurement point on the PCB (PCB board) to initially absorb impact and provide a force transmission path for the pressure feedback module integrated within the loading mechanism. Figure 1 The diagram shown is a schematic representation of an exemplary test probe. Figure 2 The diagram shown is a schematic of the test probe mounted on the UR5e robotic arm.

[0030] The pressure feedback module is a force sensing unit, such as a patch piezoresistive sensor, specifically integrated between the tail of the test probe and the spring. Its sensing surface is in contact with the end of the spring, thus enabling real-time acquisition of pressure data. The spring structure of the loading mechanism can initially buffer the impact force, while the pressure feedback module can achieve fine adjustment, compensating for the insufficiency of the fixed spring constant. The two work together to ensure the reliability of the contact between the probe and the measurement point, while also ensuring the safety of the PCB and preventing the probe from damaging the PCB.

[0031] The visual positioning module is an image sensing system integrated outside the housing, comprising two cameras and an LED light source. The optical axes of the two cameras are precisely adjusted and arranged at a specific angle (such as 90 degrees or other suitable angles) so that the optical axes of the two cameras intersect with the axis of the probe tip. This ensures that the overlapping area of ​​the two cameras' fields of view covers the tip of the test probe and the area to be measured in front of it, allowing for simultaneous image acquisition from different perspectives to achieve high-precision positioning and depth information acquisition. One or more LED light sources are used to compensate for uncertainties in ambient lighting conditions, enabling precise positioning of the measurement point and the test probe. In some embodiments, the rear end of the housing serves as the test probe base, and the visual positioning module can be integrated onto this base.

[0032] An electrical measurement unit (EPU) is an instrument used to perform electrical parameter measurements. It can be a multimeter, such as a high-precision digital multimeter. The positive and negative measurement terminals of the EPU are connected to the test probes of various test probes via cables and switching devices.

[0033] The control system is a central processing unit that coordinates all hardware components of the entire device and executes core algorithms (such as visual servoing, pressure control, and task scheduling). It can consist of an industrial control computer (PC) and corresponding motion control cards, data acquisition cards, and dedicated control software.

[0034] The control system communicates with all robotic arms, test probes, pressure measurement modules, electrical measurement units, and other hardware modules in the equipment to achieve synchronous communication among multiple modules. Communication connection refers to the physical link and logical protocol established between the control system and other hardware modules for command transmission and data exchange. For example: it communicates with the robotic arm controller via Ethernet to send motion commands and receive attitude data; it communicates with the camera via USB or GigE interface to acquire image data; it communicates with the pressure feedback module via SPI (Serial Peripheral Interface) or analog input interface to receive pressure data; it communicates with the electrical measurement unit via serial interface or GPIB (General Purpose Interface Bus) interface to control the start and stop of electrical measurements and receive electrical measurement data; and it communicates with the LED light source via digital I / O interface to adjust the LED illumination.

[0035] The core algorithms executed by the control system include the visual servoing algorithm and the termination condition algorithm.

[0036] The visual servoing algorithm mainly calculates the compensation vector for servoing through two tasks. First, it detects marker points in the area image captured in real time by the camera to determine the test needle axis, achieving bidirectional error compensation. Second, it selects a template image for each measurement point and the camera, and then matches the template image in the area image captured in real time by the camera during each servoing iteration, selecting the pixel position with the highest correlation as the target point, and then calculating the vertical pixel distance between the target point and the test needle axis. Finally, it obtains the compensation vectors in two vertical directions in the corresponding PCB top plane, which serve as the robot's motion commands.

[0037] The termination condition algorithm ensures the safety and reliability of PCB testing by defining three types of termination conditions: displacement error termination, force error termination, and iteration non-convergence termination. Displacement error termination uses a preset maximum initial error radius, establishing a virtual circular boundary centered on the initial position of the TCP (Tool Center Point) at the start of the servo operation. After each iteration, the straight-line distance between the current TCP position and the initial position is calculated to determine the displacement error. If the displacement error exceeds a preset threshold, the displacement error termination condition is met, and the servo displacement is immediately terminated. Force error termination refers to the pressure feedback module acquiring the contact force of the test probe in real time. If the pressure value detected in multiple consecutive sampling cycles (e.g., 3) exceeds a preset force threshold, the force error termination condition is met. This often occurs when the probe collides with onboard obstacles (such as a heatsink). Continuous pressure can cause the probe to bend or the PCB to be damaged. Therefore, the robotic arm is immediately driven to retract a preset distance (e.g., 2 mm), and the servo operation is terminated. Iteration non-convergence termination means that after each iteration, the physical distance of the compensation vector is calculated. If the physical distance of the compensation vector fails to reach the convergence threshold in multiple consecutive iterations, and the number of iterations reaches the preset maximum number of iterations, then it is determined that the iteration non-convergence termination condition is met, the servo is terminated, and an alarm is issued.

[0038] The worktable is used to fix the PCB board to be tested. It is equipped with positioning pins, which can prevent the PCB board from shifting due to external forces and facilitate the camera to accurately locate the pixel coordinates of feature points.

[0039] The visualization screen can be a large-size display or a touch screen, serving as the primary human-machine interface. It can also be used to receive electrical parameter data such as current, voltage, and resistance collected by the test probes and visualize it on the screen for technicians to reference and analyze. In some embodiments, the electrical measurement unit is integrated into the visualization screen.

[0040] The working principle of the visual servo testing equipment is explained below.

[0041] After the equipment is powered on and initialized, the various modules within the equipment coordinate the following processes under the unified scheduling of the control system: (1) Command issuance and data acquisition: The control system sends motion commands (including position, speed, and attitude) to the controllers of each robotic arm via Ethernet; triggers and receives synchronized image data from the dual cameras in the vision positioning modules of each test probe via USB interface; reads the pressure digital signals from the pressure feedback modules of each test probe in real time via interfaces such as SPI; and sends measurement commands (such as "measure resistance") to the electrical measurement unit via serial port and reads the returned measurement result string.

[0042] The control system can realize multiple functions such as visual servo positioning, force-controlled contact, collaborative scheduling, and integrated information display. Visual servo positioning refers to the control system running a visual servo algorithm to process dual-camera images from a specific test probe. It fits the probe axis by identifying marker points in the image, finds the measurement point through template matching, calculates the error, and generates compensation commands to drive the corresponding robotic arm movement, forming a closed loop. Force-controlled contact refers to the control system processing data from the pressure feedback module in real time, comparing it with a preset threshold, and dynamically calculating and generating new feed speed commands. The robotic arm controls the Z-axis movement of the probe, achieving closed-loop stable pressure control. Collaborative scheduling refers to the control system acting as a scheduling center when multiple test probes are present. It assigns measurement tasks to each test probe (e.g., assigning different test probes to different test areas on a PCB) and coordinates their movement sequence and resource usage (e.g., time-sharing use of electrical measurement units) to avoid interference. Integrated information display refers to the control system sending all processed data (electrical parameters, pressure curves, servo status, real-time images) to a visualization screen for comprehensive display throughout the measurement process, and automatically generating a report after the measurement is completed.

[0043] The visual servo testing equipment provided in this embodiment, through the specific hardware configuration and collaborative working method described above, can realize a visual servo testing method suitable for circuit board diagnosis (the method will be described in detail in subsequent embodiments), and can at least produce the following technical effects: (1) The equipment is based on multiple general-purpose robotic arms and interchangeable test probes. There is no need to make special bed of needle fixtures for different PCBs. Test tasks can be switched by software configuration alone, which provides a hardware foundation for small-batch, multi-variety fault PCB diagnosis. (2) By tightly integrating a visual positioning module (including dual cameras) and a pressure feedback module at the end of the test probe, the coaxiality and synchronization of "eye" and "force" in physical space are realized, providing an optimized hardware platform for the control system to achieve high bandwidth and low latency visual servo and force control fusion.

[0044] (3) The control system in the equipment integrates other hardware modules, such as distributed robotic arms, sensors, and measuring instruments, into an organic whole through standardized communication connections, and runs a unified control algorithm. This centralized architecture simplifies system complexity, improves reliability, and makes complex multi-probe task scheduling and coordination possible.

[0045] (4) The equipment includes not only test probes and electrical measurement units for performing measurements, but also a control system for data processing and a visualization screen for presenting results. This allows the entire process, from measurement to result analysis, to be completed on the same equipment, improving the continuity of the diagnostic process and the user experience.

[0046] To address the shortcomings or defects of related technologies, this application also provides a visual servo testing method for circuit board diagnosis. This method constructs a collaborative operation process with real-time visual feature guidance and closed-loop control as its core, enabling autonomous, accurate, and safe electrical diagnosis of complex faulty PCBs without the need for precise geometric information and special fixtures.

[0047] In some exemplary embodiments of this application, such as Figure 3 As shown, the method includes steps S110 to S140, and each step is described in detail below.

[0048] S110: Adjust the posture of the robotic arm equipped with the test probe based on the operator's teaching operation and the surrounding obstacles on the circuit board under test, so that the test probe avoids obstacles as it approaches the measurement point.

[0049] A test probe is an integrated measurement device that is fixedly mounted on the end of a robotic arm and integrates vision, force sensing, and electrical contact functions. It includes a test stylus for physical contact, a vision positioning module with dual cameras, a pressure feedback module (such as a piezoresistive sensor) for sensing contact force, and an LED light source for illumination.

[0050] Teaching is a process where, before testing a circuit board of a certain model for the first time, an operator manually guides a robotic arm using a robot teach pendant to roughly move the test probes above various predetermined measurement points (such as chip pins and test pads). The purpose is to allow the system to "see" and "remember" the visual features and approximate location of each measurement point, rather than providing precise geometric coordinates.

[0051] S120: Drives the robotic arm to move the test probe. During the movement of the test probe, visual servo positioning is performed cyclically based on the template image of the pre-stored measurement points until the positioning error between the test probe and the measurement points meets the preset accuracy conditions.

[0052] A template image refers to a local image region captured by dual cameras during operator teaching, when the test probe is positioned above the measurement point. This image region contains the unique visual features of the measurement point (such as pin shape, via ring, and pad texture). The control system can crop and save this image as a visual basis for identifying and locating the measurement point in subsequent automated testing.

[0053] Visual servo positioning refers to a closed-loop control process based on real-time visual feedback. It primarily involves comparing the area image of the target measurement point captured by the current camera with a pre-stored template image to calculate the positional deviation (i.e., positioning error) between the test probe and the target measurement point. Based on this, control commands are generated to drive the robotic arm to move and eliminate the deviation. This process is executed iteratively until the preset positioning accuracy is achieved.

[0054] S130: When the positioning error meets the accuracy condition, the test probe is controlled to feed along its axial direction according to the contact pressure measurement value of the test probe and the measurement point collected by the pressure feedback module in the test probe, until the contact pressure measurement value collected by the pressure feedback module is in the preset pressure range for a preset duration.

[0055] The pressure feedback module is a force sensing unit (such as a miniature piezoresistive force sensor) integrated inside the test probe, used to measure the contact pressure between the test probe and the measurement point on the circuit board in real time and with high precision as the test probe is fed along its axial direction.

[0056] The preset pressure range and preset duration are two key parameters set in this embodiment to protect the circuit board from damage and ensure the reliability of electrical contact between the test probe and the measurement point. The preset pressure range defines the safe upper and lower limits of the contact force (e.g., 50-150 g / L). The preset duration (e.g., 100 ms) is used to filter out transient fluctuations, requiring the contact pressure to be within the preset pressure range and its duration to reach the preset duration for electrical measurement operations to be performed.

[0057] S140: When the test probe stops feeding, control the electrical measuring tool connected to the test probe to measure the electrical parameters of the measuring point.

[0058] For example, the visual servoing test method of this embodiment is automatically executed according to the following process: The initial teaching and system preparation phase begins. In this phase, the operator places and secures the faulty PCB to be tested on the workbench. For the first test of this PCB model, the operator must perform a teaching demonstration: for each electrical measurement point requiring diagnosis (such as a suspected open circuit pin or a test point requiring voltage verification), the operator manually manipulates the robotic arm to move the test probe to a safe plane above the measurement point, for example, at a safe height of approximately 5-10 mm. At this position, the system controls the LED light source on the test probe to illuminate, and dual cameras simultaneously capture clear local images. The operator selects the characteristic area of ​​the measurement point in the image (such as the end of a rectangular chip pin) in the software interface and saves it as a template image. Simultaneously, the software records the position of the robotic arm's tool center point (TCP) as a rough reference coordinate and sets servo parameters, safe pressure thresholds, etc., for this point. This process is repeated for all measurement points, ultimately generating a configuration file containing visual templates, reference positions, and parameters for all points. It is worth noting that this configuration file is reusable; for subsequent tests of the same PCB model, this configuration file can be directly loaded without requiring the operator to repeat the teaching demonstration.

[0059] Next, the automated measurement execution phase begins. In this phase, after the control system loads the configuration file corresponding to the circuit board under test, it starts a fully automated test cycle and performs the following operations for each measurement point: (1) The control system can plan a path based on the reference coordinates recorded in the configuration file and move the robotic arm to the approximate area of ​​the current measurement point. At the same time, based on the pre-input obstacle information or real-time analysis of the environment through the camera, the system calculates and adjusts the posture of the robotic arm's end effector. For example, if there is a tall heat sink next to the measurement point, the control probe will approach from the side at an angle (e.g., 30°) instead of falling vertically, thereby physically avoiding the obstacle and ensuring a collision-free movement path. This operation enables intelligent obstacle avoidance approach.

[0060] (2) After the probe reaches its initial position in a safe posture, it enters a high-precision vision guidance stage to achieve precise positioning based on vision servoing. The control system will initiate a closed loop: a. Dual cameras acquire real-time images. The vision processing unit identifies two optical markers on the probe to fit the precise axis of the test needle. On the other hand, it searches the image for the region that best matches the current point template image, thereby simultaneously determining the probe's own orientation and the precise pixel position of the target measurement point.

[0061] b. Calculate the vertical pixel distance from the measurement point to the probe axis, and use the calibration parameters of the dual cameras to convert it into physical spatial errors in the X and Y axes on the PCB surface plane.

[0062] c. Generate motion compensation commands based on the physical space error calculated in the previous step, and drive the robotic arm to perform micro-motion compensation in a plane parallel to the PCB.

[0063] Each time the test probe completes a movement, steps a, b, and c above are immediately repeated to obtain new errors and compensate again. This closed loop iterates continuously, causing the positioning error to converge rapidly until it meets the preset accuracy condition (e.g., positioning error ≤ 0.05 mm).

[0064] (3) Once the vision servo system confirms that the positioning accuracy has met the standard, the control system controls the robotic arm to slowly feed along the axis of the test probe (usually the Z-axis). At this time, the pressure feedback module is activated, transmitting the contact pressure data back to the control system in real time.

[0065] The control system compares the pressure value with the preset pressure range and uses a dynamic adjustment algorithm. For example, if the pressure is lower than the lower limit of the range, the feed rate is increased appropriately; if the pressure is close to or reaches the upper limit, the rate is reduced significantly or the feed is stopped.

[0066] The aforementioned dynamic adjustment method allows the contact pressure to be gently guided and stabilized within a safe range. The system continuously monitors the pressure, and only when the pressure value remains within the range for a preset duration (e.g., 100 milliseconds) does it ultimately determine that the test probe's probe has established a stable and reliable electrical contact with the measurement point. The entire feed contact process is precisely controlled by a force feedback closed loop, perfectly avoiding impact, overpressure, or poor contact.

[0067] (4) Once the system determines that the contact between the test probe and the measurement point has reached a stable state, the control system immediately triggers a command to switch a shared electrical measuring tool (such as a high-precision digital multimeter) to the test probe of the current test probe via a multiplexer, and performs the predetermined electrical measurement operation to collect electrical parameters (such as DC voltage, resistance, and continuity). The measurement results are automatically recorded and stored in association with the current measurement point's identifier, pressure curve, positioning accuracy data, etc. After the measurement is completed, the control probe is raised back to the safe plane, and the robotic arm then moves to the next measurement point in the configuration file, repeating steps S120-S140 until all points have been tested.

[0068] This embodiment can bring the following technical effects: First, the mechanism for generating visual templates based on teaching operations provided in this embodiment can eliminate the dependence on dedicated physical fixtures and precise PCB CAD files. A single simple teaching operation is sufficient to establish testing capabilities for any PCB model, making it particularly suitable for fault board diagnosis scenarios in repair workshops involving multiple varieties, small batches, and missing drawings. This solves the core pain points of traditional methods, such as poor flexibility and high preparation costs.

[0069] Secondly, this embodiment combines visual template matching with closed-loop visual servo operation to form a powerful adaptive positioning system. This system can not only compensate for positional deviations caused by PCB clamping and mechanical errors, but also adapt to minor individual differences between PCBs of the same model. Combined with the active attitude obstacle avoidance in step S110, it enables precision measurement capabilities to be applied to complex real circuit boards filled with components, breaking through the physical limitations of traditional testing methods.

[0070] Furthermore, this embodiment introduces a dual mechanism of pressure feedback closed-loop control and pressure stabilization time determination, transforming the probe contact process from open-loop blind pressure to closed-loop compliant force control. This fundamentally prevents PCB or component damage caused by excessive pressure and eliminates unstable contact resistance caused by insufficient pressure, ensuring that data from each electrical measurement is obtained under optimal physical contact conditions, significantly improving test safety and data reliability.

[0071] Finally, this embodiment seamlessly integrates the three traditionally separate processes of visually guided positioning, force-controlled contact, and electrical parameter measurement into an automated condition-triggered chain. The system can autonomously complete the entire process from finding the location, precise alignment, reliable contact to accurate measurement, reducing manual intervention and freeing operators from tedious and focused manual testing, thus significantly improving the efficiency and automation of fault diagnosis.

[0072] In some embodiments, the execution process of the visual servo positioning operation includes, for example: Figure 4 Steps S121 to S125 are shown below, and each step is explained in detail below.

[0073] S121: Acquire images of the area containing the measurement points using a camera mounted on the test probe.

[0074] Once the robotic arm moves the test probe to a new position, the control system sends a synchronization trigger signal. The dual cameras integrated on the test probe and the LED light source activate simultaneously, capturing digital images from two slightly different perspectives, covering both the foreground of the current test probe tip and the PCB area where the target measurement point is located. These two images together constitute the raw data source for this iterative processing.

[0075] S122: Identify two marker points on the test probe in the region image using a visual servoing algorithm, and fit the axis of the test probe based on the coordinates of the two marker points.

[0076] The two marker points refer to two high-contrast, easily visually identifiable optical feature points (e.g., circular solid black dots or reflective marks) pre-positioned at specific locations on the surface of the test probe housing. These two points are located in known physical positions parallel to the axis of the test probe and are used to uniquely determine the spatial orientation of the probe in the image.

[0077] The axis of the fitting test probe refers to the straight line determined by extracting the image pixel coordinates of two marker points through image processing technology (such as circle center detection). This straight line represents the extension direction of the test probe in the image coordinate system and is the benchmark for calculating its relative positional relationship with the measurement point.

[0078] After receiving images from the dual cameras, the vision processing unit performs the following processing: First, in one (or two merged) images, a feature detection algorithm (such as Blob analysis) is run to accurately identify two pre-set marker points on the probe housing and extract their center pixel coordinates, which can be denoted as (x_1, y_1) and (x_2, y_2). Then, based on these two coordinates, a mathematical expression for the probe's axis in the image is fitted using a two-point linear equation. This axis direction reflects the probe's current orientation in real time.

[0079] S123: Use a template matching algorithm to search for the region in the region image that has the highest correlation with the template image, and determine the location of the measurement point in the region image based on the region.

[0080] Template matching is a digital image processing method used to locate the region in a larger image (a real-time acquired regional image) that is most similar to a smaller image (a pre-stored template image). In this embodiment, methods such as normalized cross-correlation matching can be used to calculate a similarity score through a sliding window, and the region with the highest score is determined as the location of the measurement point.

[0081] In another (or the same) image, the vision processing unit can access a template image pre-stored in a configuration file corresponding to the current measurement point. A template matching algorithm (e.g., normalized cross-correlation matching) is used, which treats the template image as a sliding window, traversing the real-time acquired region image and calculating the similarity coefficient at each location. The system finds and locates the image region where the similarity coefficient reaches its peak (i.e., the highest correlation). The center of this region or the coordinates of a predefined feature point are then determined as the precise pixel position of the measurement point in the image during the current iteration, denoted as (x_t, y_t).

[0082] S124: Determine the positioning error between the test probe and the measurement point based on the axis of the test probe and the position of the measurement point.

[0083] Under a unified image coordinate system, the system can calculate the vertical distance from the measurement point to the axis based on the probe axis equation fitted in step S122 and the measurement point position (x_t, y_t) located in step S123. This vertical distance (in pixels) is the positioning error in the image space during this iteration. The positioning error directly reflects the offset of the probe tip from the target measurement point in the direction perpendicular to the probe axis.

[0084] S125: Converts positioning error into physical space error, generates operation compensation command based on physical space error, and drives the robotic arm to move based on operation compensation command to reduce positioning error.

[0085] Physical space error refers to the positioning error (in pixels) calculated in the image pixel coordinate system, which is then converted into an error vector measured in length units (such as millimeters) in the actual three-dimensional world, specifically on the plane of the PCB surface, through the pre-calibrated internal parameters (such as focal length and distortion) and external parameters (such as relative position) of the dual cameras in the visual positioning module. This error is usually decomposed into components in two mutually perpendicular directions (such as the X direction).

[0086] The system calculates the image pixel error and combines it with the pre-calibrated parameter model of the dual cameras in the visual positioning module to perform coordinate transformation. This transformation process utilizes the parallax information of the two cameras to solve the two-dimensional image error into a physical spatial error in the actual plane on the PCB board surface along two orthogonal directions (e.g., the X and Y directions parallel to the board edge), which can be denoted as (ΔX, ΔY), and the unit is usually millimeters.

[0087] The control system then generates corresponding motion compensation commands based on the physical space error (ΔX, ΔY). These commands are sent to the controller of the robotic arm to drive the end effector (i.e., the test probe) to perform a translational motion of (-ΔX, -ΔY) in a plane parallel to the PCB board surface, thereby actively compensating for the currently observed positioning error.

[0088] After this movement is completed, the process immediately returns to step S121 to begin the next iteration, which involves: acquiring images at the new location, recalculating the axis, matching templates, calculating new errors, and compensating for them. This process is repeated continuously, forming a high-speed observation-calculation-adjustment closed loop, until the calculated physical spatial positioning error meets the preset accuracy conditions. At this point, the visual servo positioning operation successfully terminates and enters the next stage.

[0089] The detailed process of visual servo positioning operation described in this embodiment, through a series of synergistic technical means, has produced at least the following technical effects: First, this embodiment, by combining probe axis fitting with template matching, can simultaneously and with high precision determine the positions of the tool (test probe's probe stylus) and the target (measurement point), achieving sub-pixel accuracy in the calculated image spatial positioning error. Then, a precise dual-camera calibration model is used to convert this error into physical space error, providing the robotic arm with extremely accurate motion correction data. This is the fundamental guarantee for achieving a final contact accuracy of ≤0.05mm.

[0090] Secondly, in traditional methods, the absolute positioning accuracy of the robotic arm, PCB clamping error, and tool calibration error accumulate with each other. However, this embodiment employs the principle of relative measurement. The vision servo system directly observes and calculates the relative positional deviation between the probe and the measurement point, and drives the robotic arm to eliminate this deviation. This makes the system accuracy independent of the absolute positioning accuracy of the robotic arm, effectively isolating and compensating for systematic errors introduced by clamping, calibration, and other processes, thus improving the overall robustness and adaptability of the system.

[0091] Furthermore, this embodiment performs recognition directly based on the image features themselves, rather than relying on a specific geometric shape model. Therefore, regardless of whether the measurement point is a regular circular via, a square pin, an irregular solder joint, or a damaged pad, as long as it has distinguishable visual texture features, a valid template can be generated through teaching and successfully matched. This greatly expands the applicable scenarios of the method.

[0092] Finally, even if new minor errors occur during the servo process due to factors such as mechanical vibration or slight slippage, they can be immediately detected and compensated in the next iteration cycle, ensuring the stability of the entire approximation process and the accuracy of the final position.

[0093] In some embodiments, the method further includes: (1) During the movement of the test probe, monitor whether the displacement error condition, force error condition and iterative convergence condition are met.

[0094] (2) When the displacement error condition, force error condition or iterative convergence condition is met, terminate the current measurement process and issue an alarm.

[0095] Among them, the displacement error condition refers to the displacement of the tool center point of the test probe relative to the initial position at the start of visual servoing being greater than the preset error radius; the force error condition refers to the contact pressure measurement value collected by the pressure feedback module being greater than the preset force threshold; and the iterative convergence condition refers to the number of loops of the visual servoing positioning operation reaching the preset maximum number of iterations.

[0096] The tool center point (TCP) mentioned above is a predefined three-dimensional spatial point on the end effector of the robotic arm (i.e., the test probe), representing the tool's operational reference point. During system calibration, it is ensured that this point precisely coincides with the physical tip of the test probe. All robotic arm motion control is planned and calculated based on the tool center point.

[0097] The error radius is a preset safety tolerance threshold used to limit the maximum allowable drift range of the tool center point during visual servoing positioning. Its value is typically set based on the complexity of the measurement point area and the potential risk of template mismatch (e.g., set to 5 mm). It defines a virtual spherical safety boundary with the TCP initial position at the start of the visual servoing cycle as its center and this threshold as its radius.

[0098] The force threshold is a preset upper limit of contact pressure, which is much higher than the upper limit of the preset pressure range required for normal and stable contact. This threshold is used to determine whether an abnormal collision or severe overload has occurred (e.g., the probe accidentally hits the heat sink instead of the measurement point). Its specific value is determined based on the mechanical strength of the test probe and the withstand capability of the weakest component on the PCB (e.g., set to 500 grams of force).

[0099] The maximum number of iterations is an upper limit set for the maximum number of executions of the visual servoing positioning loop (i.e., steps S121-S125 in claim 2). This parameter is used to prevent the visual servoing algorithm from getting stuck in an infinite loop or invalid oscillation due to feature loss, severe lighting changes, or other irreversible disturbances.

[0100] This embodiment provides a security monitoring mechanism that can be executed synchronously and concurrently with the main test process (i.e., the test process corresponding to steps S110 to S140). Its specific operation flow is as follows: (1) When the control system initiates the test process for a measurement point, especially when entering the visual servo positioning stage (operation corresponding to S120) and the subsequent contact stage (operation corresponding to S130), the safety monitoring thread is activated. This thread will continuously obtain real-time information from three independent data sources: a. Read the current real-time tool center point (TCP) 3D coordinates from the robotic arm controller; b. Read the real-time contact pressure measurement value from the pressure feedback module; c. Read the number of loops that the current visual servo positioning operation has been executed from the visual servo control unit.

[0101] (2) The system uses the acquired real-time data to perform the following three judgments in parallel: a. Displacement error condition judgment, namely: calculate the three-dimensional Euclidean distance between the current TCP position and the initial TCP position recorded at the beginning of this visual servoing cycle, and continuously judge whether the displacement distance is greater than the preset error radius. If it is greater, it is determined that the displacement error condition is met.

[0102] b. Force error condition judgment, that is: continuously judge whether the real-time contact pressure measurement value obtained from the pressure feedback module is greater than the preset force threshold. If it is greater, it is determined that the force error condition is met.

[0103] c. Iterative convergence condition judgment, that is: continuously judge whether the number of times the visual servo positioning loop has been executed has reached the preset maximum number of iterations. If it has, it is determined that the iterative convergence condition is met.

[0104] (3) If any of the above three conditions is met at any time, the security monitoring thread will immediately trigger the highest priority exception handling process, which may include a. Terminate the process, that is: immediately send a forced stop command to the main control thread to stop all current motion control commands (especially feed commands) and measurement commands.

[0105] b. Safety retraction: Control the robotic arm to quickly retract a safe distance (e.g., 3-5 mm) in the opposite direction of the current probe axis, so that the test probe is separated from the PCB.

[0106] c. Issue an alarm: A prominent warning message pops up on the visual screen via the human-machine interface, simultaneously activating an audible and visual alarm device to clearly indicate the type of anomaly (e.g., "displacement exceeding limits," "pressure collision," "iteration exceeding limits"). Furthermore, relevant error timestamps and the last sensor data are recorded in the log.

[0107] (4) If none of the above three conditions are triggered before the entire test point measurement is completed (i.e., S140 is completed), the safety monitoring thread will automatically go silent after the test at that point is completed, and will be activated again when the next measurement point test begins.

[0108] The real-time multi-condition security monitoring and termination mechanism introduced in this embodiment provides a crucial proactive security protection layer for the entire automated testing system, and can produce at least the following technical effects: First, by monitoring displacement error conditions, this embodiment can promptly detect situations where the visual servo system misidentifies a target due to sudden changes in illumination, interference from template feature similarity, or other reasons. Once the TCP movement exceeds a reasonable expected range, it terminates immediately, preventing the robotic arm from colliding with the worktable or other equipment due to misguided movement, thus protecting the robotic arm itself and the test probe.

[0109] Secondly, by setting a strict force threshold and monitoring force error conditions, this embodiment enables the system to react instantly when the probe accidentally comes into contact with a high-strength obstacle (such as a metal heat sink casing), stopping and retracting before the pressure reaches a destructive level. This directly protects the expensive test probe (from bending or damage) and the faulty PCB under test (from scratches or crushing of components), significantly reducing the risk of secondary damage during the testing process.

[0110] Finally, this embodiment sets a maximum number of iterations and monitors the iteration convergence conditions, thus providing a "timeout" safeguard for the visual servoing loop. When the localization fails to converge due to complete occlusion of target features or other reasons, the system can automatically exit the loop, avoiding meaningless idle spinning, promptly reporting errors and prompting the operator to intervene and check, thereby ensuring the overall efficiency of the testing process.

[0111] In some embodiments, this method uses multiple sets of measuring devices, each set including two test probes defined as a positive probe and a negative probe, respectively; the positive probe and the negative probe are connected to the positive and negative terminals of the electrical measuring tool, respectively. When the circuit board under test is divided into multiple non-overlapping test areas, a corresponding test area is assigned to each set of measuring devices, and each set of measuring devices is controlled to perform measurements within its assigned test area.

[0112] In this embodiment, a set of measuring devices constitutes the smallest functional unit capable of independently executing the basic process of a visual servo testing method. Specifically, the set of measuring devices consists of two functionally paired test probes and their respective mounted robotic arms. One probe is defined as the positive probe, and the other as the negative probe. They work together to complete an electrical measurement task requiring two physical contact points. Specifically, the positive and negative probes are functional definitions of the two test probes in the set of measuring devices, corresponding to the high-potential and low-potential ends (or current outflow and inflow ends) in the electrical measurement circuit, respectively. The positive probe is connected to the positive terminal of an electrical measuring tool (such as a multimeter) via a wire, and the negative probe is connected to the negative terminal. This definition and connection method ensures the correct polarity of the electrical circuit when performing voltage or resistance measurements between two points.

[0113] A test area is a continuous, non-overlapping region defined on a two-dimensional plane of the circuit board under test, based on certain partitioning principles (such as spatial location clustering, robotic arm workspace constraints, and functional correlation of measurement points). Each test area contains a certain number of measurement points, and each test area is assigned to a set of measurement devices responsible for completing the measurement tasks of all measurement points within the test area.

[0114] Based on this, this embodiment provides a method for multi-probe zoned collaborative measurement, the process of which is as follows: (1) Enter the first stage, which is responsible for system configuration and task planning.

[0115] Since the system comprises multiple robotic arms and test probes, during initial configuration, the operator or control system software pairs the robotic arms and their test probes together, logically defining them as a group of measuring devices and specifying which probe within the group is the positive probe and which is the negative probe. All positive probes are connected to the positive terminal of the electrical measuring tool through a common channel group of a multiplexer switch, and all negative probes are connected to the negative terminal of the electrical measuring tool through another common channel group of the same switch.

[0116] After loading the measurement profile for a specific PCB, the control system automatically or with operator assistance divides the PCB into multiple test zones based on the spatial coordinate distribution of all measurement points in the file and the working range of each robotic arm. The division principle can be to ensure that the test zones do not overlap spatially, and that the number and complexity of measurement points in each test zone match the expected working efficiency of the measurement unit responsible for that zone.

[0117] The control system assigns a unique set of measuring devices to each test area. In the task list, all points that need to be measured by the positive and negative probes of the set of devices are clearly recorded. For points that need to be measured in pairs (such as resistance measurement), the contact point A of the positive probe and the contact point B of the negative probe are clearly specified.

[0118] (2) Enter the second stage, in which the collaborative measurement execution operation is carried out.

[0119] The control system, according to the task plan, schedules multiple sets of measuring devices to work within their respective assigned test areas. The coordination method is as follows: a. For a test area assigned to a group of measuring devices, the positive and negative test probes within that group sequentially complete the testing of each measurement point (or the testing of paired points) within that test area according to the task list. Different groups of measuring devices can execute non-interfering process segments in parallel within different test areas, such as simultaneously performing visual servo positioning and motion obstacle avoidance.

[0120] b. When multiple groups of measuring devices need to perform final electrical measurement operations simultaneously, the control system arbitrates and schedules operations through a multiplexer because they share the same electrical measuring tool. For example, when group A measuring devices need to perform electrical measurements, the control system switches the switch to the positive and negative test probes connected to group A measuring devices, triggering the measurement operation and reading the measurement data; after completion, it immediately switches to the ready group B measuring devices. This time-sharing multiplexing mechanism ensures the efficient utilization of critical measurement resources.

[0121] c. For measurement items requiring simultaneous contact of both positive and negative test probes (such as resistance), the two test probes within the control system strictly synchronize the execution of steps S110 to S130 in the above embodiments. They approach each other with their respective obstacle avoidance postures, independently perform visual servo positioning to their respective accurate target positions, and then synchronously perform force-controlled compliant contact. Only when both probes report stable contact (meeting their respective pressure stability conditions) does the control system trigger the electrical measuring tool to perform measurements, thereby obtaining accurate parameters between the two points.

[0122] (3) Enter the third stage, which is responsible for data collection and synchronization. After all the measurement tasks in the test areas are completed, the control system will align the measurement data from different groups of measuring devices and different test areas with timestamps, integrate them according to the original PCB layout, and generate a unified comprehensive test report covering the entire circuit board.

[0123] The multi-probe partitioned collaborative measurement strategy adopted in this embodiment, through system-level resource organization and task scheduling, can produce at least the following technical effects: (1) By dividing the PCB into multiple test areas and assigning independent groups of measurement devices to work in parallel, the traditional single-probe serial scanning mode is changed to a multi-probe parallel scanning mode. Although the core electrical measurement resources may need to be time-division multiplexed, time-consuming processes such as visual positioning and motion alignment can be parallelized, thereby significantly shortening the total time required to complete the whole board test, which is especially suitable for large-size PCBs with many measurement points.

[0124] (2) By defining the positive and negative probes and pre-connecting them to the correct polarity terminals of the measuring tool in the hardware, out-of-the-box support is provided for measurements of resistance, impedance, etc., that require two test points. The synchronous and coordinated control of the dual probes within the group ensures that the two contact points reach a stable state at the same time, making the measurement results more reflective of the true circuit characteristics and avoiding errors introduced by time-division contact.

[0125] (3) The division of the test area and task allocation can be optimized based on the working space range, motion performance and spatial density of the test points of each robotic arm. This helps to balance the workload of each group of measuring devices, avoid the excessive range of motion or excessive task of a single robotic arm, thereby optimizing the overall resource utilization of the system and potentially extending the equipment life.

[0126] (4) When dealing with larger or more complex PCBs, the measurement capability of the system can be improved by increasing the number of measurement devices. At the same time, the partitioning strategy enables the system to flexibly respond to the differences in testing requirements of different areas on the PCB (such as some areas requiring dense measurement, while others require simple measurement), enhancing the engineering applicability and flexibility of the testing method.

[0127] In some embodiments, adjusting the posture of the robotic arm equipped with the test probe based on operator teaching and the surrounding obstacles at the measurement point on the circuit board under test includes: (1) In response to the operator's teaching operation, control the robotic arm to move the test probe, which is defined as the positive probe, to a safe plane near the measurement point; (2) Identify obstacle information based on visual analysis of the environment surrounding the measurement point or operator input; (3) Adjust the end effector posture of the robotic arm according to the obstacle information.

[0128] A safety plane is a virtual two-dimensional plane that is parallel to the surface of the circuit board under test (PCB) and located at a certain safe height (e.g., 5-10 mm) above it. When the test probe is positioned in this plane, there is still sufficient non-contact clearance between its tip and the highest component or obstacle on the PCB, allowing the probe to move quickly and safely laterally within this plane. This is the reference plane for area localization and initial attitude planning.

[0129] Obstacle information refers to descriptive information about physical objects that exist near the target measurement point and may prevent the test probe from directly reaching and contacting the measurement point in a vertical or default orientation. This includes at least the presence of the obstacle, its approximate outline, height, and relative orientation to the measurement point.

[0130] In this embodiment, the specific operation procedure for planning and setting a safe approach posture for a single measurement point is as follows: a. The operator manually controls the robotic arm using a handheld robot teach pendant, moving the test probe (using the positive probe as an example) mounted at its end to approximately directly above the target measurement point. The operator then stops the probe on a predefined safety plane. The control system records the probe's two-dimensional coordinates (X_s, Y_s) on the safety plane at this point as the teaching reference position for that measurement point. This process is coarse and does not require precise alignment; its purpose is to provide the automated system with a reliable, collision-free search starting point.

[0131] b. After the probe is positioned in a safe plane, the system acquires obstacle information through one or a combination of the following methods: The visual analysis method (automatic) involves the control system instructing the dual cameras on the test probe to capture or scan the surrounding area of ​​the measurement point from multiple angles. Image processing algorithms (such as edge detection, contour extraction, and stereo depth estimation) automatically identify protruding components such as heat sinks, electrolytic capacitors, and connectors, estimating their contours and relative heights, and generating a simple local obstacle map within the control system.

[0132] The operator input method (semi-automatic) involves the operator viewing the real-time image displayed on the teaching software interface when the test probe is positioned on a safe plane. The operator observes the image and marks obstacle areas near the measurement point by directly drawing or clicking on the screen with the mouse. The system then converts this manually marked information into spatial constraint information.

[0133] c. The control system integrates the precise location of the measurement point (which can be determined from the template image), the current position of the probe on the safe plane (X_s, Y_s), and the obstacle information obtained in step b, and performs motion planning calculations.

[0134] First, the system determines whether there is an obstacle on the line connecting the current position of the test probe to the target measurement point (projected onto the safe plane). If an obstacle exists, a collision-free two-dimensional path from the current position to the target point is re-planned on the safe plane using an obstacle avoidance algorithm (such as the artificial potential field method or sampling method), and the final horizontal direction of approach is determined. Next, based on the height of the obstacle and the position of the measurement point, the system calculates the minimum tilt angle (i.e., the angle between the probe axis and the PCB normal) required for the test probe to avoid the side of the obstacle. For example, if the obstacle is north of the measurement point, the system may plan the probe to approach obliquely from the south at a 30-degree tilt angle. Then, the control system generates corresponding motion commands to drive the robotic arm to adjust the attitude of its end effector. After adjustment, the axis of the test probe will precisely point to the calculated approach vector direction, while keeping its tip on the safe plane. At this point, the test probe is in an optimal initial attitude, ready to perform subsequent precise positioning.

[0135] The intelligent attitude adjustment process described in detail in this embodiment lays a safe and feasible foundation for subsequent high-precision operations by introducing environmental perception and planning steps, and can produce at least the following technical effects: First, by combining the teaching reference position with real-time obstacle information, this embodiment enables the system to dynamically calculate a unique, collision-free approach posture for each measurement point. This completely changes the traditional working mode of fixed vertical downward pressure or simple preset path of testing equipment, allowing this method to adaptively cope with ever-changing actual PCB layouts and greatly expanding the application boundaries.

[0136] Secondly, the process in this embodiment supports two methods for acquiring obstacle information: purely visual automatic recognition and manual annotation by the operator. This provides operational flexibility; in well-organized environments, automatic recognition improves efficiency, while in extremely complex environments or when automatic recognition is uncertain, the operator's experience can be used for rapid confirmation and intervention. This human-machine collaborative approach improves the system's practicality and reliability while ensuring intelligence.

[0137] Third, this embodiment directly solves the industry problem of being unable to test due to obstruction by large onboard components through active obstacle avoidance posture planning. It significantly reduces the probability of the probe colliding with valuable components on the PCB during testing, protecting not only the testing equipment (test probe and robotic arm) but also the circuit board to be diagnosed.

[0138] In some embodiments, the method further includes at least one of the following operations: (1) Generate a comprehensive test report based on the electrical parameters, contact pressure curves, visual servo data and corresponding area images of the measurement points obtained during the measurement process.

[0139] (2) Compare the measured electrical parameters with the pre-stored reference parameters, and mark the abnormal measurement points based on the comparison results; perform fault-assisted location based on the contact pressure curve characteristics and image characteristics of the abnormal measurement points.

[0140] A comprehensive test report is a structured, readable electronic document (such as PDF, HTML, or a report in a specific software format). Its content is not a simple list of data, but an integrated display of multi-dimensional test information that is synchronous in time and logically related to the same measurement point.

[0141] Reference parameters refer to the baseline values ​​or allowable ranges used to determine whether electrical measurement results are normal. These parameters can be derived from theoretical values ​​in circuit design, statistically measured values ​​from similar fault-free boards, or tolerance ranges specified in relevant technical standards. These parameters can be pre-stored in the system's database before testing and associated with specific measurement point identifiers.

[0142] Fault-assisted localization refers to the system, after marking the measurement points with abnormal electrical parameters (i.e., abnormal measurement points), further utilizing the characteristics of non-electrical process data (such as contact pressure curves and area images) at those points, to automatically analyze and infer the potential physical causes of electrical abnormalities (such as solder joint detachment, pin cold solder joint, via cracking, foreign object short circuit, etc.), and provide maintenance personnel with preliminary diagnostic clues or probability ranking.

[0143] The data processing and diagnostic workflow in this embodiment is automatically triggered after all measurement points have been tested, and can be executed as follows: For operation (1), the control system first retrieves all data packets generated by each measurement point within the complete test cycle from the storage unit in the order of measurement points. Each data packet includes electrical parameters (referring to the final measured current value, voltage value, resistance value, etc.), contact pressure curve (referring to the pressure-time sampling sequence of the entire process from the start of feeding to contact stabilization and then lifting), and visual servo data (including information such as final positioning error, number of servo iterations, and error change sequence of each iteration). The area image of the measurement point refers to one or more frames of images captured by the camera of the test probe when the visual servo positioning is completed or before the test probe's test needle contacts the measurement point. The system can use the unique identifier and timestamp of each measurement point as an index to strictly align and associate the above four types of data in time logic, thereby obtaining a complete state snapshot. The system can automatically generate a comprehensive test report according to a preset template. The report may include an overview page, a detailed data page, and a process traceability page. The overview page displays an overall image of the PCB and marks the pass / fail status of each measurement point with different colors (such as green / red). The detailed data page displays electrical parameters (with comparisons to reference values), key pressure curve screenshots (highlighting stable contact sections), final positioning accuracy, and characteristic images of the point in the form of tables and charts. The process traceability page provides interactive links, allowing engineers to click on anomalies to view the complete servo error convergence process curve or pressure control details.

[0144] For operation (2), the system first performs an initial screening operation for electrical anomalies, that is, it automatically compares the electrical parameters obtained at each measurement point with the reference parameters corresponding to that point pre-stored in the database. For example, if the measured resistance is infinite (open circuit), while the reference value is 10Ω; or the measured voltage to ground is 0V, while the reference value should be 3.3V. According to preset rules (such as exceeding ±10% tolerance), the system automatically marks the points that do not meet the conditions as abnormal measurement points and highlights them in the report and interface. Then, for each abnormal measurement point, the system starts a diagnostic analysis routine, which does not rely solely on electrical parameters, but also analyzes the associated process characteristics. Among them, contact pressure curve characteristic analysis can be performed to analyze the shape of the pressure curve at that point. For example, if the curve shows a sudden drop in pressure after a rapid increase at the moment of contact, followed by a stabilization but with an overall low pressure value, it may indicate that the solder joint has detached or cracked, causing a feeling of missing the target during probe contact. If the curve shows an unusually long time required to reach stable pressure, or exhibits continuous micro-vibrations, it may indicate poor soldering or oxidation of the pins, leading to unstable contact. If the curve shape is completely normal, it may rule out contact issues and point to a purely electrical circuit fault. The system can also perform image feature analysis, that is, call up the area image of the measurement point and use image recognition algorithms or visual inspection by engineers to look for clues to physical defects. For example, the image may show damaged solder pads, black circles around vias (possibly burnt out), or the presence of unwanted solder bridges (short circuits).

[0145] Finally, the system will generate diagnostic suggestions. Specifically, it can logically correlate electrical anomalies, pressure curve characteristics, and image features to generate preliminary fault location assistance suggestions. For example, the system output diagnostic suggestions could be: "Measurement point A23, open circuit in resistance. The pressure curve shows momentary contact collapse, and the image shows missing solder pads. Highly suspect solder pad detachment." or "Measurement point B45, short circuit to ground. The pressure curve is normal, and the image shows solder dross between the two pins. Suspected foreign object bridging short circuit." These diagnostic suggestions, along with the raw data, are presented in the report, providing engineers with strong support for repair decision-making.

[0146] It should be noted that, regarding the various steps included in the visual servo testing method for circuit board diagnostics provided in any of the above embodiments, unless explicitly stated herein, there is no strict order restriction on the execution of these steps; they can be executed in other orders. Furthermore, at least some of these steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0147] This application also provides a computer device. In some embodiments, the computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it can implement the visual servo testing method for circuit board diagnosis provided in any of the above embodiments.

[0148] In some embodiments, the internal structure diagram of a computer device may be as follows: Figure 5 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores template images, configuration files, and other data; specific data stored may also be defined in the above method embodiments. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a visual servo testing method for circuit board diagnostics.

[0149] Those skilled in the art will understand that Figure 5The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0150] This application also provides a computer-readable storage medium, in some embodiments of which a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, it implements the visual servo testing method for circuit board diagnosis provided in any of the above embodiments.

[0151] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0152] Those skilled in the art will understand that implementing all or part of the processes in the above method embodiments can be accomplished by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink, DRAM (SLDRAM), memory bus, direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0153] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A visual servoing testing method for circuit board diagnosis, characterized by, The method includes: The posture of the robotic arm equipped with the test probe is adjusted based on the operator's teaching operation and the surrounding obstacles on the circuit board under test, so that the test probe avoids obstacles as it approaches the measurement point. The robotic arm is driven to move the test probe. During the movement of the test probe, visual servo positioning is performed cyclically based on the template image of the pre-stored measurement point until the positioning error between the test probe and the measurement point meets the preset accuracy condition. When the positioning error meets the accuracy condition, the test probe is controlled to feed along its axial direction according to the contact pressure measurement value of the test probe and the measurement point collected by the pressure feedback module in the test probe, until the contact pressure measurement value collected by the pressure feedback module is in the preset pressure range for a preset duration. When the test probe stops feeding, the electrical measuring tool connected to the test probe is controlled to measure the electrical parameters of the measuring point.

2. The method according to claim 1, characterized in that, The execution process of the visual servo positioning operation includes: The camera mounted on the test probe acquires an image of the area containing the measurement point. Two marker points on the test probe are identified in the region image using a visual servoing algorithm, and the axis of the test probe is fitted based on the coordinates of the two marker points. A template matching algorithm is used to search for the region in the region image that has the highest correlation with the template image, and the location of the measurement point in the region image is determined based on the region. The positioning error between the test probe and the measurement point is determined based on the axis of the test probe and the position of the measurement point. The positioning error is converted into a physical space error, and a running compensation command is generated based on the physical space error. The running compensation command is used to drive the robotic arm to move, thereby reducing the positioning error.

3. The method according to claim 1, characterized in that, The method further includes: During the movement of the test probe, it is monitored whether the displacement error condition, force error condition, and iterative convergence condition are satisfied. When the displacement error condition, the force error condition, or the iterative convergence condition is met, the current measurement process is terminated and an alarm is issued. The displacement error condition refers to the displacement of the tool center point of the test probe relative to the initial position at the start of visual servoing being greater than the preset error radius. The force error condition refers to the contact pressure measurement value collected by the pressure feedback module being greater than the preset force threshold. The iterative convergence condition refers to the number of iterations of the visual servo positioning operation reaching a preset maximum number of iterations.

4. The method according to claim 1, characterized in that, The method uses multiple sets of measuring devices, each set of measuring devices including two test probes defined as a positive probe and a negative probe, respectively; the positive probe and the negative probe are respectively connected to the positive and negative terminals of the electrical measuring tool; When the circuit board under test is divided into multiple non-overlapping test areas, a corresponding test area is assigned to each group of measuring devices, and each group of measuring devices is controlled to perform measurements in its assigned test area.

5. The method according to claim 4, characterized in that, The adjustment of the posture of the robotic arm equipped with the test probe based on operator instruction and the surrounding obstacles at the measurement points on the circuit board under test includes: In response to the operator's teaching operation, the robotic arm is controlled to move the test probe, defined as the positive probe, to a safe plane near the measurement point; Obstacle information is identified based on visual analysis of the environment surrounding the measurement point or operator input; The end effector posture of the robotic arm is adjusted based on the obstacle information.

6. The method according to claim 1, characterized in that, The method further includes: A comprehensive test report is generated based on the electrical parameters, contact pressure curves, visual servo data, and corresponding area images of the measurement points obtained during the measurement process.

7. The method according to claim 6, characterized in that, The method further includes: The measured electrical parameters are compared with the pre-stored reference parameters, and abnormal measurement points are marked based on the comparison results; Fault-assisted localization is performed based on the contact pressure curve characteristics and image characteristics of the abnormal measurement points.

8. A visual servo probe testing device for circuit board diagnostics, characterized in that, The device includes: At least two robotic arms; The same number of test probes as the robotic arms, each test probe being fixed to the end of one of the robotic arms, each test probe including a housing, a test pin disposed at the front end of the housing, a vision positioning module disposed outside the housing, and a pressure feedback module integrated in the loading mechanism of the test pin; the vision positioning module including two cameras; A worktable, located within the range of motion of all the robotic arms, is used to fix the circuit board to be tested. Electrical measurement unit; A visualization screen is used to display the electrical parameters, contact pressure curves, and / or visual servo data acquired by the test probe during the measurement process; The control system is communicatively connected to all the robotic arms, all the test probes, all the pressure measurement modules, and the electrical measurement unit.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.

Images