Probe card replacement method and device for high-precision test socket

Abstract

The application relates to the technical field of probe cards and discloses a probe card replacement method and device of a high-precision test seat. The method comprises the following steps: S1, positioning a probe card to be replaced and controlling the probe card to be replaced to descend to a preset hovering height; S2, through four corner piezoelectric drivers, the positioned probe card is lifted and pressed to a preset first pressing force; S3, target displacement adjustment amounts of the four corners are calculated according to the first pressing force; S4, the target displacement adjustment amounts of the four corners are synchronously fed to the piezoelectric drivers for adjustment and second pressing forces of the four corners are output; and S5, a first mass center offset amount is recalculated based on the second pressing force, when the first mass center offset amount exceeds a balance threshold value and the iteration number does not reach an upper limit, the step S3 is returned, and when the first mass center offset amount is smaller than the balance threshold value or the iteration number reaches the upper limit, it is determined that the probe card is installed. The application realizes intelligent distribution of the pressing force, realizes high speed, stability and high precision of the probe card replacement process.

Description

Technical Field

[0001] This invention relates to the field of probe card technology, and in particular to a method and apparatus for replacing probe cards in a high-precision test fixture. Background Technology

[0002] In the process of replacing probe cards in high-precision test fixtures, a multi-point mechanical clamping method is used for fixation. However, due to the difficulty in precisely controlling the force distribution at each clamping point, the probe card often tilts after installation, severely affecting test accuracy and probe card lifespan. Existing technology uses a single-axis sequential adjustment strategy to adjust each clamping point one by one. However, this sequential adjustment method causes a significant force coupling effect, meaning that adjusting the clamping force at one point will affect the force state at other points. Since single-axis sequential adjustment requires multiple iterations to balance the clamping forces at each point, the entire adjustment process is time-consuming and cannot meet the requirements of rapid probe card replacement in high-throughput testing. Summary of the Invention

[0003] The main objective of this invention is to provide a method and apparatus for replacing probe cards in a high-precision test fixture. This invention achieves intelligent distribution of clamping force, and realizes high speed, stability and high precision in the probe card replacement process.

[0004] To achieve the above objectives, the present invention provides a method for replacing the probe card of a high-precision test fixture, comprising the following steps: S1. Locate the probe card to be replaced and control its descent to the preset hovering height; S2. Using the piezoelectric actuators at the four corners, the probe card that has completed positioning is raised and pressed to the preset first pressing force; S3. Calculate the target displacement adjustment amount at each corner position based on the first clamping force; S4. The target displacement adjustment amount of each corner position is synchronously sent to the piezoelectric actuator for adjustment and the second clamping force of each corner position is output. S5. Recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card installation is complete.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of locating the probe card to be replaced and controlling its descent to a preset hovering height includes: Collect edge feature points of the probe card to be replaced and reference marker points of the test seat, and extract the first coordinates of the edge feature points and the second coordinates of the reference marker points; The probe card position deviation data is calculated based on the first coordinate and the second coordinate. The probe card position deviation data includes the X-direction position deviation value, the Y-direction position deviation value, and the θ rotation angle deviation value. Based on the probe card position deviation data, the robotic arm is controlled to adjust and determine whether the probe card horizontal calibration is complete. If horizontal calibration is complete, the probe card that has completed horizontal calibration will be lowered at a preset descent speed. The descent will stop when the vertical distance between the bottom surface of the probe card and the surface of the test seat reaches the preset hovering height, thus confirming that the probe card has completed positioning.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of raising and pressing the positioned probe card to a preset first pressing force using piezoelectric actuators at the four corners includes: The piezoelectric actuators at the four corners are controlled to start synchronously from the zero position and rise synchronously at a preset rising speed. When the torque sensor at any corner detects that the contact force reaches a preset contact threshold, the piezoelectric actuators are controlled to stop rising synchronously and the initial contact position of each corner is recorded. Based on the initial contact position, the piezoelectric actuator is controlled to continue to rise synchronously by a preset displacement. During the rising process, the real-time clamping force at each corner is monitored in real time. When the real-time clamping force at each corner reaches the preset first clamping force, the piezoelectric actuator is controlled to stop rising synchronously.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the step of calculating the target displacement adjustment amount at each angle based on the first clamping force includes: The first force centroid coordinates are calculated based on the first clamping force at each corner and the position vector of each corner, and the centroid offset vector is calculated based on the first force centroid coordinates and the preset geometric center coordinates. The initial centroid offset is obtained by calculating the magnitude of the centroid offset vector. The dot product operation is performed based on the position vector of each corner and the initial centroid offset to obtain the dot product projection value of each corner. The torque compensation coefficient of each corner is calculated based on the dot product projection value of each corner and the preset compensation intensity coefficient. Calculate the average clamping force of the first clamping force at each corner position, and multiply the average clamping force by the torque compensation coefficient of each corner position to obtain the target clamping force at each corner position; The target displacement adjustment amount at each corner is obtained by dividing the difference between the target clamping force at each corner and the first clamping force at each corner by a preset stiffness coefficient.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the step of performing a dot product operation based on the position vector of each corner and the initial centroid offset to obtain the dot product projection value of each corner, and calculating the torque compensation coefficient of each corner based on the dot product projection value of each corner and a preset compensation intensity coefficient, includes: The dot product of the position vectors of each corner and the initial centroid offset is performed to obtain the dot product projection value of each corner. The squared values ​​of the X and Y components of the position vectors at each angular position are calculated by squaring them and then summed to obtain the squared modulus of each position vector. The target product is obtained by multiplying the dot product projection value of each corner position by the square of the magnitude of the position vector of each corner position and then multiplying it by the preset compensation intensity coefficient. The difference between the target product and the preset reference value is calculated to obtain the torque compensation coefficient of each corner position.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of synchronously sending the target displacement adjustment amount of each corner position to the piezoelectric actuator for adjustment and outputting the second clamping force of each corner position includes: The target displacement adjustment amount of each corner position is sent to the corresponding piezoelectric actuator. At the unified trigger time, the piezoelectric actuator is controlled to start the adjustment movement synchronously at the preset adjustment speed, and the current clamping force of each corner position is collected. The difference between the current clamping force at each corner and the target clamping force at each corner is calculated to obtain the clamping force deviation value at each corner. When the clamping force deviation at each corner is less than the preset convergence threshold, the piezoelectric actuator is controlled to stop the adjustment movement synchronously, and the second clamping force at each corner is collected after the stop.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the step of controlling the piezoelectric actuator to synchronously stop the adjustment movement when the clamping force deviation value at each corner is less than a preset convergence threshold, and collecting the second clamping force at each corner after stopping, includes: The clamping force deviation at each corner is compared with the preset convergence threshold. When the clamping force deviation at each corner is less than the preset convergence threshold, a synchronous stop signal is generated. Based on the synchronous stop signal, the piezoelectric actuator is controlled to synchronously stop the adjustment movement and maintain the current position, and the second clamping force at each corner position is collected after a delayed settlement.

[0011] Optionally, in a seventh implementation of the first aspect of the present invention, the step of recalculating the first centroid offset based on the second clamping force, returning to step S3 when the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, and determining that the probe card installation is complete when the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, includes: The second force center coordinates are calculated based on the second clamping force at each corner and the position vector of each corner, and the first center offset is calculated based on the second force center coordinates and the preset geometric center coordinates. The first centroid offset is compared with a preset balance threshold, and the current iteration number is compared with a preset iteration limit. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card is installed and that the probe card is in a balanced state. The piezoelectric actuator is controlled to synchronously lock the probe card, which is in a balanced state, to a preset locking force. The standard deviation of the second centroid offset and the third clamping force at each corner is calculated. When the second centroid offset and the standard deviation of the third clamping force both meet the preset verification conditions, the probe card installation is determined to be complete.

[0012] Optionally, in an eighth implementation of the first aspect of the present invention, the control piezoelectric actuator synchronously locks the probe card, which is in a balanced state, to a preset locking force, calculates the standard deviation of the second centroid offset and the third clamping force at each corner, and determines that the probe card installation is complete when both the second centroid offset and the standard deviation of the third clamping force meet preset verification conditions, including: The piezoelectric actuator controls the probe card, which is in a balanced state, to rise synchronously at a preset locking speed. When the third clamping force at each corner increases from the current value to the preset locking force, the movement stops synchronously. The third force centroid coordinates are calculated based on the third clamping force at each corner and the position vector of each corner. The second centroid offset is calculated based on the third force centroid coordinates and the preset geometric center coordinates. After calculating the average and variance of the third clamping force at each corner, take the square root to obtain the standard deviation of the third clamping force at each corner. The standard deviations of the second centroid offset and the third clamping force are compared with preset verification conditions. When the standard deviation of the second centroid offset and the third clamping force both meet the preset verification conditions, the probe card is confirmed to be installed and a replacement success signal is output.

[0013] The present invention also provides a probe card replacement device for a high-precision test fixture, comprising: The positioning module is used to locate the probe card to be replaced and control its descent to a preset hovering height; The rising and pressing module is used to press the positioned probe card up to a preset first pressing force through the piezoelectric actuators at the four corners; The calculation module is used to calculate the target displacement adjustment amount at each corner position based on the first clamping force; The displacement adjustment module is used to synchronously send the target displacement adjustment amount of each corner position to the piezoelectric actuator for adjustment and output the second clamping force of each corner position; The iteration module is used to recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, the process returns to step S3. When the first centroid offset is less than the balance threshold or the number of iterations reaches the upper limit, the probe card installation is determined to be complete.

[0014] In summary, this invention, based on the calculation of the first force center coordinates and the analysis of the center offset vector, accurately calculates the torque compensation coefficient at each point by projecting the dot product of the position vectors at each corner and the center offset vector, thus achieving intelligent distribution of the clamping force. The unified triggering time four-axis synchronous adjustment mechanism ensures strict synchronization of the four piezoelectric actuators in the time dimension, eliminating the force coupling oscillation phenomenon caused by single-axis sequential adjustment and improving the stability of the adjustment process. The iterative optimization convergence mechanism dynamically adjusts the control strategy based on real-time collected clamping force data, achieving a minimum force center offset through multiple rounds of rapid convergence. The multi-dimensional verification system established by this invention quantitatively evaluates the replacement quality by calculating the final force center offset and the standard deviation of the clamping force at each corner, ensuring that the probe card attitude accuracy meets high-precision testing requirements, thereby achieving high-speed, stable, and high-precision probe card replacement. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the steps for replacing the probe card of a high-precision test fixture in one embodiment of the present invention; Figure 2 This is a structural block diagram of a probe card replacement device for a high-precision test fixture according to an embodiment of the present invention.

[0016] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Reference Figure 1This embodiment provides a method for replacing the probe card of a high-precision test fixture, including the following steps: S1. Locate the probe card to be replaced and control its descent to the preset hovering height; Specifically, the system acquires feature point images of the edge of the probe card to be replaced and images of preset reference marker points on the test stand using an image sensor system. A sub-pixel edge extraction algorithm is then used to calculate the first coordinates of the edge feature points and the second coordinates of the reference marker points. These coordinates are input into a visual error analysis module, where coordinate difference calculations and rotation vector decomposition are performed to calculate the spatial position deviation data of the probe card relative to the test stand. This deviation data includes translational deviations in the X and Y directions and a rotational angle deviation θ around the Z-axis. Based on this three-dimensional pose deviation information, the control system drives a multi-axis robotic arm to adjust the X, Y, and θ axes, respectively. Real-time visual feedback is used to update the adjustment parameters in a closed-loop manner, ensuring continuous convergence of the probe card's relative deviation in the horizontal plane. When the deviation values ​​in all three directions are less than preset accuracy thresholds (e.g., less than 50 μm in the X and Y directions and less than 0.05 degrees in the θ direction), the probe card's horizontal calibration is considered complete. The controller switches to the vertical Z-axis control logic, instructing the robotic arm to perform a descent operation along the Z-axis at a stable and uniform speed, and continuously monitors the relative height between the bottom surface of the probe card and the surface of the test seat. When the vertical distance gradually approaches and reaches the preset hovering height value, such as 2mm, the control system issues an immediate stop descent command to keep the robotic arm stationary, lock the three-dimensional coordinate position at this time, and determine that the probe card has completed the positioning preparation.

[0019] S2. Using the piezoelectric actuators at the four corners, the probe card that has completed positioning is raised and pressed to the preset first pressing force; Specifically, the position states of the four piezoelectric actuators are initialized to ensure they are all in a zero-position ready state. A unified synchronous start command is then issued to them. The actuators rise synchronously in the vertical direction at a preset rising speed—for example, a uniform speed of 0.5 mm / s. During this process, torque sensors installed at the four corners continuously collect vertical force data at their respective contact points at a high sampling rate (e.g., 1 kHz). When any sensor detects that the current contact force reaches the set contact threshold, for example, 5 N, it is considered that the bottom surface of the probe card at that corner has made physical contact with the surface of the test seat for the first time. The control system immediately responds and issues a global synchronous stop command, causing all piezoelectric actuators to stop moving at the same time. The axial displacement positions at the four corners at this moment are recorded as the initial contact position information in the current batch replacement process. According to the preset displacement feed strategy, the control system issues a synchronous rising command again, causing the four drives to rise from the initial contact position at the same speed for a fixed stroke, such as 0.5mm, to establish an effective contact interface between the probe card and the test seat. During the rising process, the four sensors continue to monitor the clamping force data at their corresponding positions in real time, polling and collecting the latest clamping force values ​​at the four points at a period of 1ms, and comparing them with the set target first clamping force (such as 100N). When it is determined that the real-time clamping force at all corners has reached or slightly exceeded the preset value and is within the allowable tolerance range (such as ±10N), the control system issues a synchronous stop command to the four drives, and the four axes complete the stop action at the same time and maintain the current position.

[0020] S3. Calculate the target displacement adjustment amount at each corner position based on the first clamping force; Specifically, based on the first clamping force at each of the four corners acquired in real time by the torque sensor, and combined with the preset spatial radius vector positions of each corner sensor in the probe card coordinate system, a four-point force model is constructed. The control system calculates the weighted average of the clamping force at each corner with its position radius vector, and obtains the coordinate values ​​of the center of mass along the X and Y axes respectively, thus obtaining the coordinates of the first center of mass under the current state. The coordinate difference between the first center of mass coordinates and the ideal geometric center of the probe card (defined as the origin) is calculated to obtain the center of mass offset vector in the two-dimensional plane. The magnitude of the center of mass offset vector is calculated to obtain the absolute value of the offset, which is recorded as the initial center of mass offset, used to determine whether there is a significant unevenness in the current force distribution. The dot product is calculated between the spatial radius vector at each of the four corners and the center of mass offset vector, obtaining the dot product projection value of each position relative to the offset direction. The sign and magnitude of the dot product result determine the response trend and influence intensity of that position to the current offset state. A preset compensation strength coefficient is introduced, and the dot product value is mapped to the torque compensation coefficient corresponding to each corner position through normalized calculation. This coefficient is used to adjust the increase or decrease of the target clamping force at each point. The arithmetic mean of the first clamping force at each of the four corner positions is taken to obtain the overall average clamping force value. The overall average clamping force value is then multiplied by the torque compensation coefficient at each corner position to generate the corrected target clamping force at each corner position. The target clamping force at each corner position is subtracted from its current first clamping force to obtain the clamping force error value. The clamping force error value is then divided by the known axial stiffness coefficient of the piezoelectric ceramic actuator to calculate the target displacement adjustment amount at each corner position.

[0021] S4. The target displacement adjustment amount of each corner position is synchronously sent to the piezoelectric actuator for adjustment and the second clamping force of each corner position is output. Specifically, the industrial control computer loads the target displacement adjustment values ​​for the four corner positions into the control registers of the four-channel drivers, and establishes a unified trigger clock so that the four drivers enter the adjustment mode at exactly the same time. The drivers are then controlled to simultaneously initiate vertical displacement at a preset adjustment speed, such as 0.2 mm / s. Simultaneously, the clamping force sensors initiate high-frequency acquisition to monitor the actual force changes at each corner position during the adjustment process in real time. In each sampling cycle, the control system receives the current clamping force values ​​from each sensor and calculates the difference between the current clamping force value and the preset target clamping force to obtain the clamping force deviation value for each corner position. The system continuously monitors whether the clamping force deviation value is within the convergence range, using the condition that the absolute value of the deviation at all corner positions is less than a preset convergence threshold (e.g., 5 N) as the criterion. If any corner position does not yet meet this condition, adjustment continues until all corner positions meet it. When the clamping force deviation values ​​of all corner positions are within the convergence range, a unified stop command is issued, causing the four drives to terminate the adjustment action at the same time and maintain the current position stability. The final clamping force data of each corner position at the end of the adjustment is collected and used as the second clamping force.

[0022] S5. Recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card installation is complete.

[0023] Specifically, by weighted summing of the second clamping force at each of the four corners with the corresponding spatial position vectors, the coordinate components of the center of mass in the X and Y axes are calculated to construct the coordinates of the second center of mass in the current state. The difference between the coordinates of the second center of mass and the coordinates of the ideal geometric center constitutes a two-dimensional center of mass offset vector. The magnitude of the two-dimensional center of mass offset vector is calculated to obtain the current first center of mass offset. The first center of mass offset is compared with a set balance threshold, which is on the order of 0.1 mm. At the same time, the current number of adjustment iterations is compared with a preset upper limit for iterations. If the first center of mass offset exceeds the balance threshold and the number of iterations has not yet reached the maximum limit (e.g., 3 times), it is determined that the attitude balance condition has not been met, and the process returns to step S3 to recalculate the target clamping force and adjust the displacement. If the offset is lower than the set threshold or the number of iterations has reached the upper limit, it is considered that the probe card installation is complete and the device is in an acceptable balance state. At this point, the control system controls the four piezoelectric actuators to perform the final synchronous locking operation, continuing to rise at a constant speed until a higher level of locking force (e.g., 150N) is reached, and an attitude verification process is performed after locking. During the verification phase, the third clamping force at all corners in the locked state is re-acquired, and the final position of the center of gravity and its second center of gravity offset from the geometric center are calculated accordingly. Simultaneously, the standard deviation of the third clamping force at the four corners is calculated through statistical analysis. When the second center of gravity offset is lower than the set offset verification threshold (e.g., 0.15mm) and the standard deviation of the clamping force is lower than the set uniformity threshold (e.g., 10N), the probe card clamping state is determined to have reached final equilibrium, and the entire installation process is successfully completed; otherwise, a warning is output, indicating an abnormal installation deviation requiring manual intervention.

[0024] In one example, locating the probe card to be replaced and controlling its descent to a preset hovering height includes: Collect edge feature points of the probe card to be replaced and reference marker points of the test seat, and extract the first coordinates of the edge feature points and the second coordinates of the reference marker points; The probe card position deviation data is calculated based on the first coordinate and the second coordinate. The probe card position deviation data includes the X-direction position deviation value, the Y-direction position deviation value, and the θ rotation angle deviation value. The robotic arm is controlled to adjust based on the probe card position deviation data, and the horizontal calibration of the probe card is determined. If horizontal calibration is complete, the probe card that has completed horizontal calibration will be lowered at a preset descent speed. The descent will stop when the vertical distance between the bottom surface of the probe card and the surface of the test seat reaches the preset hovering height, thus confirming that the probe card has completed positioning.

[0025] In this example, a high-resolution CCD camera mounted on the end effector of the robotic arm acquires images of the edge contour of the probe card to be replaced, and simultaneously obtains images of fixed reference markers on the test base surface. An image processing module then performs edge enhancement, sub-pixel extraction, and centroid analysis on these images to extract the first coordinates of the probe card's edge feature points and the second coordinates of the test base reference markers. Based on these first and second coordinates, a spatial matching operation is performed, and the probe card's attitude offset information is obtained through coordinate difference calculation. This includes the translational deviation values ​​along the X and Y directions in the horizontal plane, as well as the θ rotation angle deviation value caused by rotation around the Z-axis. These three values ​​together constitute the probe card's position offset data vector. The probe card position deviation data is used as an input signal and loaded into the robotic arm motion control module. This sequentially drives the servo actuators of the X, Y, and θ axes to execute displacement adjustment commands in the corresponding directions. Through a closed-loop feedback control strategy, the probe card's attitude in the plane gradually approaches the target reference state. After each fine adjustment, images are re-acquired in real time to update the deviation value. The horizontal calibration process of the probe card is considered complete when the deviations in the X and Y directions are less than the set position accuracy threshold (e.g., 50 μm) and the θ angle deviation is less than the set angle accuracy threshold (e.g., 0.05 degrees). At this point, the controller automatically switches to the Z-axis control mode and slowly moves the probe card vertically downward at a preset uniform descent rate. Simultaneously, non-contact height measurement devices such as laser rangefinders or capacitive sensors continuously monitor the real-time vertical distance between the bottom surface of the probe card and the surface of the test base. When the real-time vertical distance decreases and accurately reaches the set pre-hovering height (e.g., 2 mm), the system issues a stop descent command, keeping the robotic arm stationary at the pre-hovering height and locking this position as the final positioning state of the probe card.

[0026] The process includes an initial contact difference compensation step after controlling the piezoelectric actuator to synchronously stop rising: recording the contact time when the torque sensor at each corner detects a preset contact threshold, calculating the time difference between the contact time at each corner and the earliest contact time, and obtaining contact time difference data for each corner; calculating the contact position deviation of each corner relative to the earliest contact corner based on the contact time difference data and the preset rising speed, comparing the contact position deviation of each corner with the preset contact position deviation threshold, and obtaining a contact position deviation evaluation result; when the contact position deviation evaluation result shows that the contact position deviation of a corner exceeds the preset contact position deviation threshold, calculating the zero-position compensation amount for each corner based on the initial contact position and the contact position deviation amount, and then... The zero-position compensation of each corner position is superimposed on the initial contact position of each corner position to obtain the calibrated reference position of each corner position. Based on the calibrated reference position of each corner position, the piezoelectric actuator is re-controlled to rise synchronously, so that each corner position starts to apply clamping force from the calibrated reference position. When the real-time clamping force of each corner position reaches the preset first clamping force, it stops synchronously, and the first clamping force of each corner position after eliminating the influence of the initial contact difference is obtained. After the first clamping force of each corner position after eliminating the influence of the initial contact difference is subjected to a delayed settlement process, multiple sets of clamping force data are continuously collected. The average value and the difference between the maximum and minimum values ​​are calculated for the multiple sets of clamping force data. When the difference between the maximum and minimum values ​​is less than the preset clamping force uniformity threshold, the initial contact difference compensation is confirmed to be completed, and the average value is used as the input data for subsequent torque distribution calculation.

[0027] In one example, the positioned probe card is raised and pressed to a preset first clamping force using piezoelectric actuators at the four corners, including: The piezoelectric actuators at the four corners are controlled to start synchronously from the zero position and rise synchronously at a preset rising speed. When the torque sensor at any corner detects that the contact force reaches the preset contact threshold, the piezoelectric actuators are controlled to stop rising synchronously and the initial contact position of each corner is recorded. Based on the initial contact position, the piezoelectric actuator is controlled to continue to rise synchronously by a preset displacement. During the rising process, the real-time clamping force at each corner is monitored in real time. When the real-time clamping force at each corner reaches the preset first clamping force, the piezoelectric actuator is controlled to stop rising synchronously.

[0028] In this example, piezoelectric actuators positioned at the four corners are simultaneously activated from a zero-position state. All actuators enter a synchronized rising mode under a unified clock trigger and slowly advance vertically at a set uniform speed (e.g., 0.5 mm / s). Simultaneously, torque sensors installed at the corresponding positions of each piezoelectric actuator continuously sample the force state in the pressing direction at a high frequency (e.g., 1 kHz). During the actuator's ascent, the contact force data at each corner is continuously polled. When the contact force measured by any sensor reaches a set contact threshold (e.g., 5 N), a unified stop command is triggered. All four piezoelectric actuators simultaneously stop rising at the trigger moment and synchronously latch the axial displacement values ​​at the current four corners. These values ​​represent the initial contact positions during the probe card's pressing process. After completing the initial contact recording, the four piezoelectric actuators are controlled to start synchronously again and continue rising a fixed distance (e.g., 0.5 mm) from their original positions. During this process, the torque sensor continuously monitors the real-time clamping force value at each corner. The control system receives and updates the force data stream of each channel in real time with a 1ms cycle, and compares the current real-time clamping force with the set target first clamping force (e.g., 100N) to determine whether the four corners simultaneously meet the clamping force target. When the real-time clamping force of all four corners is within the set allowable range (e.g., ±10N error band) and no channel has over-pressure or under-pressure deviation, the system issues a synchronous stop command to the four piezoelectric actuators. The four actuators terminate their actions at the same time and maintain their current position locked, completing the entire process of the initial clamping stage.

[0029] In one example, the target displacement adjustment at each corner is calculated based on the first clamping force, including: The first force centroid coordinates are calculated based on the first clamping force at each corner and the position vector of each corner, and the centroid offset vector is calculated based on the first force centroid coordinates and the preset geometric center coordinates. The initial centroid offset is obtained by calculating the magnitude of the centroid offset vector. The dot product operation is performed based on the position vector of each corner and the initial centroid offset to obtain the dot product projection value of each corner. The torque compensation coefficient of each corner is calculated based on the dot product projection value of each corner and the preset compensation strength coefficient. Calculate the average clamping force of the first clamping force at each corner position, and multiply the average clamping force by the torque compensation coefficient at each corner position to obtain the target clamping force at each corner position; The target displacement adjustment amount at each corner is obtained by dividing the difference between the target clamping force at each corner and the first clamping force at each corner by the preset stiffness coefficient.

[0030] In this example, the first clamping force value at each corner is obtained, and combined with the preset spatial position vectors of each corner in the local coordinate system of the probe card, the first force centroid coordinates under the current state are calculated by weighted summation of the product of the clamping force and its vector in the X and Y axes respectively, thereby determining the overall force centroid of the probe card under the action of force. The difference between the first force centroid coordinates and the ideal geometric center position (defined as the origin of the coordinate system) constitutes the two-dimensional centroid offset vector. The Euclidean norm is calculated on the two-dimensional centroid offset vector to obtain its magnitude value, which is recorded as the initial centroid offset, used to quantitatively measure the degree of deviation between the current force state and the ideal equilibrium state. To identify the response contribution of each corner to the current centroid offset, the control system performs a dot product operation on the spatial position vector and the centroid offset vector of each corner to obtain the dot product projection value of each corner. The sign of the dot product projection value is used to determine the consistency between the current position and the offset direction, and the magnitude reflects the intensity of the influence of the position on the centroid offset. A predetermined compensation strength coefficient is introduced, and the dot product projection value is converted into a corresponding torque compensation coefficient through linear scaling. The torque compensation coefficient is used to correct the force output intensity at each corner. The arithmetic mean of the first clamping force at all corners is calculated as the benchmark for the overall clamping force level. The average clamping force is then multiplied by the torque compensation coefficient at each corner to obtain the corresponding target clamping force. The target clamping force is the expected force value after correction by the torque balance algorithm. The difference between the target clamping force at each corner and the current first clamping force is calculated, and the clamping force error value is divided by the known axial stiffness coefficient of the piezoelectric ceramic actuator to calculate the target displacement adjustment amount at each corner.

[0031] In one example, a dot product operation is performed based on the position radius vector of each corner and the initial centroid offset to obtain the dot product projection value of each corner. The moment compensation coefficient for each corner is then calculated based on the dot product projection value and a preset compensation intensity coefficient, including: The dot product is performed based on the position vector of each corner and the initial centroid offset to obtain the dot product projection value of each corner. The squared values ​​of the X and Y components of the position vector at each angular position are obtained by squaring and then summing them. The target product is obtained by multiplying the dot product projection value of each corner position by the square of the magnitude of the position vector of each corner position and then multiplying it by the preset compensation strength coefficient. The difference between the target product and the preset reference value is calculated to obtain the torque compensation coefficient of each corner position.

[0032] In this example, the initial centroid offset vector caused by uneven force is obtained. A vector dot product is then performed between this initial centroid offset vector and the spatial position vector of each corner, yielding the dot product projection value for each corner. This dot product value quantifies the projected intensity of each position point in the current offset direction, while the sign of the dot product reflects the force response tendency of that corner relative to the offset direction. The X and Y components of the position vector of each corner are squared, and the two squared results are added to obtain the squared modulus of the position vector, which describes the distance intensity index of the corner relative to the geometric center. The dot product value of each corner is divided by its corresponding squared modulus to form a normalized response ratio with directional information. A predetermined compensation intensity coefficient is then introduced to linearly amplify the normalized response ratio, resulting in a target product value that includes compensation trend and intensity weight. The target product value is then compared with a fixed reference constant (e.g., set to 1) to generate the torque compensation coefficient for each corner.

[0033] In one example, the target displacement adjustment amount for each corner position is synchronously sent to the piezoelectric actuator for adjustment and outputs the second clamping force for each corner position, including: The target displacement adjustment amount of each corner position is sent to the corresponding piezoelectric actuator. At the unified trigger time, the piezoelectric actuator is controlled to start the adjustment movement synchronously at the preset adjustment speed, and the current clamping force of each corner position is collected. The difference between the current clamping force at each corner and the target clamping force at each corner is calculated to obtain the clamping force deviation value at each corner. When the clamping force deviation at each corner is less than the preset convergence threshold, the piezoelectric actuator is controlled to stop the adjustment movement synchronously, and the second clamping force at each corner is collected after the stop.

[0034] In this example, the target displacement adjustment amounts at the four corner positions are loaded into the corresponding driver control registers. After data buffering confirmation, a unified synchronous start trigger timestamp is set, ensuring that all drivers simultaneously initiate adjustment movements at the same preset adjustment speed (e.g., 0.2 mm / s) at the set time node, thereby ensuring multi-axis coordination at the physical execution level. During driver operation, high-sensitivity torque sensors installed at each corner position collect current clamping force data in real time. The latest feedback values ​​from all sensors are polled at 1 ms time steps, and the difference between the current clamping force at each corner position and its respective target clamping force is calculated in each cycle to form the clamping force deviation value for each corner position. The clamping force deviation reflects the quantitative error between the current state and the target state and is used to determine whether the adjustment action has reached the convergence condition. During the adjustment process, the absolute value of the deviation values ​​of the four channels is continuously judged. When the clamping force deviation values ​​of all corner positions are less than the set convergence threshold (e.g., 5N), it is considered that the multi-axis clamping action has been completed and the attitude has been basically balanced. At this time, the control system issues a synchronous stop command, causing the four piezoelectric actuators to terminate the current adjustment action and lock the final position at the same time. Subsequently, the actual clamping force of each corner position after the adjustment is completed is collected, and this set of data is marked as the second clamping force, which is used as the mechanical output parameter of the current attitude state.

[0035] The process includes a dynamic speed adaptive adjustment step after controlling the piezoelectric actuator to synchronously initiate the adjustment movement at a preset adjustment speed. This step involves continuously calculating the remaining clamping force difference between the current clamping force and the target clamping force at each angle during the adjustment process. The remaining adjustment amount ratio at each angle is calculated by dividing the remaining clamping force difference by the initial remaining clamping force difference, thus obtaining the adjustment progress data for each angle. Based on the adjustment progress data, the leading angle with the fastest adjustment progress and the lagging angle with the slowest adjustment progress are identified. The adjustment progress difference between the leading and lagging angles is calculated. When the adjustment progress difference exceeds a preset coordination threshold, adaptive speed adjustment is initiated. The adjustment speed of the piezoelectric actuator at the leading angle is reduced to a reference speed. The preset adjustment ratio maintains the baseline adjustment speed or moderately increases the adjustment speed of the piezoelectric actuators at lagging corners, so that the adjustment progress difference between each corner gradually converges. When the remaining clamping force difference at any corner is less than the preset early arrival threshold, it is determined that the corner has reached the area near the target clamping force ahead of schedule. The force holding mode is activated for that corner, and the clamping force at that corner is kept constant within the fluctuation range near the target clamping force through high-frequency adjustment of small displacement increments, while waiting for other corners to complete the adjustment. When the remaining clamping force difference of all corners is less than the preset convergence threshold, a four-axis synchronous arrival signal is generated, controlling all piezoelectric actuators to synchronously stop the adjustment movement and exit the force holding mode, and the second clamping force of each corner is collected.

[0036] In one example, when the clamping force deviation at each corner is less than a preset convergence threshold, the piezoelectric actuator is controlled to synchronously stop the adjustment movement, and the second clamping force at each corner after stopping is collected, including: The clamping force deviation at each corner is compared with the preset convergence threshold. When the clamping force deviation at each corner is less than the preset convergence threshold, a synchronous stop signal is generated. Based on the synchronous stop signal, the piezoelectric actuator is controlled to stop the adjustment movement synchronously and maintain the current position. After a delayed settling, the second clamping force at each corner position is collected.

[0037] In this example, the control system continuously reads the current clamping force data fed back by the four corner torque sensors at a period of 1ms, and performs a difference calculation with each target clamping force to obtain the current clamping force deviation value for each channel. These four clamping force deviation values ​​are then compared with a preset convergence threshold, set to 5N, to define the allowable range of small adjustment errors. When the absolute values ​​of the deviations of all channels are less than the convergence threshold, i.e., |deviation1|<5N, |deviation2|<5N, |deviation3|<5N, |deviation4|<5N, the controller generates a synchronous stop signal to instruct all piezoelectric actuators to simultaneously terminate the current adjustment process. After the synchronous stop signal is distributed to the four-channel actuator control module, each actuator enters position holding mode and locks its current position, and the probe remains unchanged in its current attitude state. Because minor stress relaxation or structural springback still exists within the mechanical structure and materials after the actuator stops adjusting, a short delay window (e.g., 50ms) is set as the settlement time to avoid transient disturbances affecting the accuracy of the force data. During this period, data acquisition and control response are paused. After the delay is complete, the torque sensor restarts synchronous data acquisition, obtains the latest clamping force measurements at all angles, and uses this set of data as the second clamping force.

[0038] In one example, the first centroid offset is recalculated based on the second clamping force. If the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, the process returns to step S3. If the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, the probe card installation is determined to be complete, including: The second force center coordinates are calculated based on the second clamping force at each corner and the position vector at each corner, and the first center offset is calculated based on the second force center coordinates and the preset geometric center coordinates. The first centroid offset is compared with a preset balance threshold, and the current iteration number is compared with a preset iteration limit. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card is installed and that the probe card is in a balanced state. The piezoelectric actuator is controlled to synchronously lock the probe card, which is in a balanced state, to a preset locking force. The standard deviation of the second centroid offset and the third clamping force at each corner is calculated. When the standard deviation of the second centroid offset and the third clamping force both meet the preset verification conditions, the probe card installation is determined to be complete.

[0039] In this example, based on the second clamping force value at each corner and the corresponding position vector data, a torque center calculation operation is performed. By weighted summing of the clamping force at each corner and its spatial position vector in the coordinate system, the coordinates of the second force center of mass are obtained in the X and Y axes respectively, and the current actual force center position vector is constructed accordingly. The second force center of mass coordinates are then differentially calculated with the preset ideal geometric center coordinates to obtain the first center of mass offset reflecting the probe card's attitude deviation. The magnitude of the first center of mass offset represents the degree of deviation of the current force state from the ideal symmetrical state. The first centroid offset is compared with a set balance threshold, which is set at the 0.1mm level. Simultaneously, the number of adjustment iterations already performed is compared with the system's maximum allowed iteration limit (e.g., 3 times). If the current centroid offset is still greater than the balance threshold and the number of iterations has not yet reached the limit, the attitude is determined not to have converged to the target state, and the process returns to step S3 to recalculate the target clamping force and adjust the displacement. If the centroid offset is less than the set threshold or the number of iterations has been reached, even with minor residual deviations, the current state is considered an acceptable balance state, and the probe card installation stage is confirmed to be complete. Afterward, the piezoelectric actuator enters the locking process. The four-axis actuator continues to advance upward at a preset locking speed under synchronous start triggering until the clamping force at each corner increases from the current value to the set final locking force (e.g., 150N), completing the mechanical constraint solidification action. After locking, the latest clamping force values ​​for all corners are acquired again as the third clamping force. Based on this data, the coordinates of the center of gravity are recalculated to obtain the second center of gravity offset under locking conditions. Simultaneously, the standard deviation of the third clamping force is calculated to assess the overall uniformity of the force distribution. The second center of gravity offset is compared with a preset offset verification threshold. The standard deviation of the third clamping force is also compared with the clamping force uniformity control threshold (e.g., 10N) in parallel. If both indicators meet the set acceptance conditions, the probe card is finally determined to be in a stable, stress-balanced, and consistent contact installation state. If either indicator fails to meet the standard, an alarm prompts that manual verification is required.

[0040] The process includes a multi-index fusion iterative judgment step after calculating the first centroid offset: First, based on the second clamping force at each corner, calculate the uniformity index of the clamping force distribution in the current iteration, including the standard deviation of the clamping force at each corner, the difference between the maximum and minimum clamping forces, and the dispersion coefficient of the clamping force at each corner relative to the average clamping force, to obtain the clamping force distribution characteristic data for the current iteration. Second, retrieve the historical value of the centroid offset from the previous iteration, calculate the difference between the first centroid offset of the current iteration and the historical value of the centroid offset, and calculate the centroid offset change rate based on the difference divided by the historical value of the centroid offset. When the absolute value of the centroid offset change rate is less than a preset convergence rate threshold, it is determined as a slowdown in convergence, obtaining the iterative convergence trend evaluation result. Third, calculate the iterative optimization efficiency index by weighted summing of the first centroid offset, the standard deviation in the clamping force distribution characteristic data, and the centroid offset change rate in the current iteration. When the absolute value of the iterative optimization efficiency index is less than a preset convergence rate threshold, it is determined as a slowdown in convergence, obtaining the iterative convergence trend evaluation result. When the value is less than the preset optimization efficiency threshold, it is determined that the marginal benefit of iterative optimization is decreasing, and the iterative efficiency evaluation result is obtained. Based on the comparison result of the first centroid offset and the balance threshold, the iterative convergence trend evaluation result, the iterative efficiency evaluation result, and the current iteration number, a multi-condition logic judgment is performed. When the first centroid offset is less than the balance threshold, or the iterative convergence trend evaluation result shows that the convergence is slowing down and the iterative efficiency evaluation result shows that the marginal benefit is decreasing, an iteration termination command is generated and the termination reason identifier is recorded. The termination reason identifier distinguishes between two types: balance achievement termination and system error limitation termination. When the termination reason identifier is balance achievement termination, it is directly determined that the probe card is in a balanced state. When the termination reason identifier is system error limitation termination, it is evaluated whether the current clamping force distribution characteristic data meets the acceptable conditions. If it does, it is determined that the probe card is in an acceptable balanced state. If it does not, a replacement failure warning signal is output, requiring the probe card to be replaced again.

[0041] After obtaining the target displacement adjustment amount for each corner, the process includes an adaptive grouping adjustment step based on the clamping force distribution characteristics: Based on the current first clamping force at each corner, spatial characteristic parameters of the clamping force distribution at the four corners are calculated, including the diagonal clamping force difference and the adjacent-side clamping force difference. The diagonal clamping force difference includes the difference between the clamping forces at the two corners of the first diagonal and the two corners of the second diagonal. The adjacent-side clamping force difference includes the difference between the clamping forces at the two adjacent corners of each of the four sides. The absolute values ​​of the diagonal clamping force difference and the adjacent-side clamping force difference are calculated and compared to identify the dominant direction of the clamping force imbalance as either a diagonal imbalance mode or an adjacent-side imbalance mode, thus obtaining an imbalance mode classification result. When imbalance... When the pattern classification result is a diagonal imbalance pattern, the four corners are divided into a first diagonal group and a second diagonal group according to their diagonal relationship. The average clamping force deviation of the two corners in the first diagonal group and the average clamping force deviation of the two corners in the second diagonal group are calculated. The diagonal group with the larger absolute value of the average clamping force deviation is marked as the main adjustment group, and the diagonal group with the smaller absolute value of the average clamping force deviation is marked as the auxiliary adjustment group. When the imbalance pattern classification result is an adjacent-side imbalance pattern, the side with the largest absolute value of the adjacent-side clamping force difference is identified as the main imbalance side. The two corners of the main imbalance side are marked as the main adjustment group, and the other two corners are marked as the auxiliary adjustment group, thus obtaining the corner grouping result. Based on the corner grouping result, the adjustment gain coefficient of each corner is calculated. A larger first adjustment gain coefficient is assigned to each corner position of the main adjustment group, while a smaller second adjustment gain coefficient is assigned to each corner position of the auxiliary adjustment group. The target displacement adjustment amount of each corner position is multiplied by the corresponding adjustment gain coefficient to obtain the weighted target displacement adjustment amount of each corner position. A force coupling prediction model based on the mechanical stiffness of the probe card is established. The force coupling prediction model calculates the coupling influence of the clamping force on other corner positions when a displacement adjustment amount is applied to a certain corner position based on the elastic modulus and geometric parameters of the probe card. By inputting the weighted target displacement adjustment amount of each corner position of the main adjustment group into the force coupling prediction model, the predicted value of the coupling clamping force change generated by the main adjustment group on each corner position of the auxiliary adjustment group during adjustment is calculated. Based on the predicted value of the coupling clamping force change... The weighted target displacement adjustment amount of each corner position of the auxiliary adjustment group is corrected by measuring the value. The target clamping force of each corner position of the auxiliary adjustment group is subtracted from the predicted value of the coupling clamping force change to obtain the corrected target clamping force. Based on the difference between the corrected target clamping force and the current clamping force, the corrected target displacement adjustment amount of each corner position of the auxiliary adjustment group is recalculated. At the unified triggering time, the weighted target displacement adjustment amount of each corner position of the main adjustment group and the corrected target displacement adjustment amount of each corner position of the auxiliary adjustment group are synchronously sent to the corresponding piezoelectric actuator and the adjustment is started synchronously. During the adjustment process, the current clamping force of each corner position is monitored in real time. When the clamping force deviation value of each corner position is less than the preset convergence threshold, the adjustment is stopped synchronously, and the second clamping force of each corner position after group optimization adjustment is obtained.

[0042] In one example, the piezoelectric actuator is controlled to synchronously lock the probe card, which is in a balanced state, to a preset locking force. The standard deviation of the second centroid offset and the third clamping force at each corner is calculated. When both the second centroid offset and the standard deviation of the third clamping force meet preset verification conditions, the probe card installation is determined to be complete, including: The piezoelectric actuator controls the probe card, which is in a balanced state, to rise synchronously at a preset locking speed. When the third clamping force at each corner increases from the current value to the preset locking force, the movement stops synchronously. The third force centroid coordinates are calculated based on the third clamping force at each corner and the position vector of each corner. The second centroid offset is calculated based on the third force centroid coordinates and the preset geometric center coordinates. After calculating the average and variance of the third clamping force at each corner, take the square root to obtain the standard deviation of the third clamping force at each corner. The standard deviations of the second centroid offset and the third clamping force are compared with the preset verification conditions. When the standard deviations of the second centroid offset and the third clamping force both meet the preset verification conditions, the probe card installation is confirmed to be complete and a replacement success signal is output.

[0043] In this example, the piezoelectric actuators at the four corners are controlled to synchronously perform a locking and lifting operation in a unified locking mode. A desired locking speed parameter (e.g., 0.3 mm / s) is set and applied to all piezoelectric actuator channels. At a unified trigger moment, all four actuators initiate the lifting action from their current positions at the same rate, while continuously monitoring the clamping force changes at each corner. When the real-time clamping force value of each channel is detected to gradually increase from the current second clamping force to the set locking target value (e.g., 150 N), the system issues a synchronous stop command. The four actuators terminate their movement at the same moment and maintain the final locking displacement unchanged, completing the stable fixation between the probe card and the test base. The third clamping force data at each corner is retrieved and combined with the known spatial position radius coordinates to perform analytical calculation of the force centroid. The third force centroid coordinates are obtained by weighting the clamping force and its coordinates at each point. The difference between this and the set geometric center position is then calculated to obtain the current second centroid offset, which serves as a geometric indicator for evaluating attitude symmetry. Simultaneously, statistical analysis is performed on the third clamping force data at the four corners, calculating their average value. Then, the sum of the squared differences of the clamping forces at the four points relative to the average value is calculated, and their average is taken. Subsequently, the square root is performed to obtain the standard deviation of the third clamping force, which is used to characterize the uniformity of the clamping force distribution. The second centroid offset is compared with the set centroid offset tolerance threshold. At the same time, the standard deviation of the third clamping force is compared with the clamping uniformity verification threshold. When both judgment conditions are met, that is, the current centroid position error does not exceed the allowable error range of the structure, and there is no significant deviation in the distribution of the four clamping forces, it is determined that the probe card replacement operation has met the verification requirements in both the attitude control and mechanical distribution dimensions. Thus, the probe card installation operation is determined to be complete, the control system automatically outputs a replacement success signal and records the replacement time point and related parameters, ending the entire replacement process.

[0044] Reference Figure 2 This embodiment provides a probe card replacement device for a high-precision test fixture, including: Positioning module 1 is used to locate the probe card to be replaced and control its descent to a preset hovering height; The lifting and clamping module 2 is used to lift and clamp the positioned probe card to a preset first clamping force through the piezoelectric actuators at the four corners; Calculation module 3 is used to calculate the target displacement adjustment amount at each corner position based on the first clamping force; Displacement adjustment module 4 is used to synchronously send the target displacement adjustment amount of each corner position to the piezoelectric actuator for adjustment and output the second clamping force of each corner position; Iteration module 5 is used to recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, it returns to step S3. When the first centroid offset is less than the balance threshold or the number of iterations reaches the upper limit, it is determined that the probe card installation is complete.

[0045] In this embodiment, the specific implementation of each unit in the above device embodiment is described in the above method embodiment, and will not be repeated here.

[0046] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0047] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A probe card replacement method for a high-precision test socket, characterized by, include: S1. Locate the probe card to be replaced and control its descent to the preset hovering height; S2. Using the piezoelectric actuators at the four corners, the probe card that has completed positioning is raised and pressed to the preset first pressing force; S3. Calculate the target displacement adjustment amount at each corner position based on the first clamping force; Specifically, this includes: calculating the coordinates of the first force-bearing centroid based on the first clamping force at each corner and the position vector at each corner; calculating the centroid offset vector based on the first force-bearing centroid coordinates and the preset geometric center coordinates; calculating the magnitude of the centroid offset vector to obtain the initial centroid offset; performing a dot product operation based on the position vector at each corner and the initial centroid offset to obtain the dot product projection value at each corner; squaring the X-direction component and the Y-direction component of the position vector at each corner and then adding them together to obtain the magnitude squared value of the position vector at each corner; based on the... The dot product projection value of each corner position is divided by the squared modulus of the position vector of each corner position and then multiplied by a preset compensation strength coefficient to obtain the target product. The difference between the target product and the preset reference value is calculated to obtain the torque compensation coefficient of each corner position. The average clamping force of the first clamping force of each corner position is calculated, and the average clamping force is multiplied by the torque compensation coefficient of each corner position to obtain the target clamping force of each corner position. The target displacement adjustment amount of each corner position is obtained by dividing the difference between the target clamping force of each corner position and the first clamping force of each corner position by a preset stiffness coefficient. S4. The target displacement adjustment amount of each corner position is synchronously sent to the piezoelectric actuator for adjustment and the second clamping force of each corner position is output. S5. Recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card installation is complete.

2. The probe card replacement method for the high-precision test fixture according to claim 1, characterized in that, The process of locating the probe card to be replaced and controlling its descent to a preset hovering height includes: Collect edge feature points of the probe card to be replaced and reference marker points of the test seat, and extract the first coordinates of the edge feature points and the second coordinates of the reference marker points; The probe card position deviation data is calculated based on the first coordinate and the second coordinate. The probe card position deviation data includes the X-direction position deviation value, the Y-direction position deviation value, and the θ rotation angle deviation value. Based on the probe card position deviation data, the robotic arm is controlled to adjust and determine whether the probe card horizontal calibration is complete. If horizontal calibration is complete, the probe card that has completed horizontal calibration will be lowered at a preset descent speed. The descent will stop when the vertical distance between the bottom surface of the probe card and the surface of the test seat reaches the preset hovering height, thus confirming that the probe card has completed positioning.

3. The probe card replacement method for the high-precision test fixture according to claim 1, characterized in that, The process of using piezoelectric actuators at four corners to raise and press the positioned probe card to a preset first pressing force includes: The piezoelectric actuators at the four corners are controlled to start synchronously from the zero position and rise synchronously at a preset rising speed. When the torque sensor at any corner detects that the contact force reaches a preset contact threshold, the piezoelectric actuators are controlled to stop rising synchronously and the initial contact position of each corner is recorded. Based on the initial contact position, the piezoelectric actuator is controlled to continue to rise synchronously by a preset displacement. During the rising process, the real-time clamping force at each corner is monitored in real time. When the real-time clamping force at each corner reaches the preset first clamping force, the piezoelectric actuator is controlled to stop rising synchronously.

4. The probe card replacement method for the high-precision test fixture according to claim 1, characterized in that, The step of synchronously sending the target displacement adjustment amount of each corner position to the piezoelectric actuator for adjustment and outputting the second clamping force of each corner position includes: The target displacement adjustment amount of each corner position is sent to the corresponding piezoelectric actuator. At the unified trigger time, the piezoelectric actuator is controlled to start the adjustment movement synchronously at the preset adjustment speed, and the current clamping force of each corner position is collected. The difference between the current clamping force at each corner and the target clamping force at each corner is calculated to obtain the clamping force deviation value at each corner. When the clamping force deviation at each corner is less than the preset convergence threshold, the piezoelectric actuator is controlled to stop the adjustment movement synchronously, and the second clamping force at each corner is collected after the stop.

5. The probe card replacement method for the high-precision test fixture according to claim 4, characterized in that, When the clamping force deviation at each corner is less than a preset convergence threshold, the piezoelectric actuator is controlled to synchronously stop the adjustment movement, and the second clamping force at each corner after stopping is collected, including: The clamping force deviation at each corner is compared with the preset convergence threshold. When the clamping force deviation at each corner is less than the preset convergence threshold, a synchronous stop signal is generated. Based on the synchronous stop signal, the piezoelectric actuator is controlled to synchronously stop the adjustment movement and maintain the current position, and the second clamping force at each corner position is collected after a delayed settlement.

6. The probe card replacement method for the high-precision test fixture according to claim 5, characterized in that, The step of recalculating the first centroid offset based on the second clamping force, returning to step S3 when the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, and determining that the probe card installation is complete when the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, includes: The second force center coordinates are calculated based on the second clamping force at each corner and the position vector of each corner, and the first center offset is calculated based on the second force center coordinates and the preset geometric center coordinates. The first centroid offset is compared with a preset balance threshold, and the current iteration number is compared with a preset iteration limit. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, return to step S3. When the first centroid offset is less than the balance threshold or the number of iterations has reached the upper limit, determine that the probe card is installed and that the probe card is in a balanced state. The piezoelectric actuator is controlled to synchronously lock the probe card, which is in a balanced state, to a preset locking force. The standard deviation of the second centroid offset and the third clamping force at each corner is calculated. When the second centroid offset and the standard deviation of the third clamping force both meet the preset verification conditions, the probe card installation is determined to be complete.

7. The probe card replacement method for the high-precision test fixture according to claim 6, characterized in that, The control piezoelectric actuator synchronously locks the probe card, which is in a balanced state, to a preset locking force. It calculates the standard deviation of the second centroid offset and the third locking force at each corner. When both the second centroid offset and the standard deviation of the third locking force meet preset verification conditions, the probe card installation is determined to be complete, including: The piezoelectric actuator controls the probe card, which is in a balanced state, to rise synchronously at a preset locking speed. When the third clamping force at each corner increases from the current value to the preset locking force, the movement stops synchronously. The third force centroid coordinates are calculated based on the third clamping force at each corner and the position vector of each corner. The second centroid offset is calculated based on the third force centroid coordinates and the preset geometric center coordinates. After calculating the average and variance of the third clamping force at each corner, take the square root to obtain the standard deviation of the third clamping force at each corner. The standard deviations of the second centroid offset and the third clamping force are compared with preset verification conditions. When the standard deviation of the second centroid offset and the third clamping force both meet the preset verification conditions, the probe card is confirmed to be installed and a replacement success signal is output.

8. A probe card replacement device for a high-precision test fixture, characterized in that, The steps for implementing the probe card replacement method for the high-precision test fixture according to any one of claims 1 to 7 include: The positioning module is used to locate the probe card to be replaced and control its descent to a preset hovering height; The rising and pressing module is used to press the positioned probe card up to a preset first pressing force through the piezoelectric actuators at the four corners; The calculation module is used to calculate the target displacement adjustment amount at each corner position based on the first clamping force; The displacement adjustment module is used to synchronously send the target displacement adjustment amount of each corner position to the piezoelectric actuator for adjustment and output the second clamping force of each corner position; The iteration module is used to recalculate the first centroid offset based on the second clamping force. When the first centroid offset exceeds the balance threshold and the number of iterations has not reached the upper limit, the process returns to step S3. When the first centroid offset is less than the balance threshold or the number of iterations reaches the upper limit, the probe card installation is determined to be complete.

Images