PCB probe positioning method and system and storage medium

Abstract

The invention relates to the technical field of probe testing, in particular to a PCB probe positioning method and system and a storage medium, and the method comprises the steps: obtaining a three-dimensional contour of a bonding pad, screening interference nodes, evaluating a contact error in combination with a probe structure, analyzing the density of test points, dividing a compression region, and carrying out the statistics of a space interference probability. And probe positioning and contact verification are completed according to the error and probability sorting path. The method has the beneficial effects that the dot matrix height on the surface of the bonding pad is comprehensively sampled, the space contour of the bonding pad is reconstructed, and the edge point residual error is extracted, so that the identification of the interference node is more accurate. And in combination with CAD boundary identification of a probe end face structure, comparing a node set formed by three-dimensional sampling points of a bonding pad, and establishing a contact area error evaluation mechanism in normal distance standard deviation calculation to realize multi-dimensional contact adaptation judgment. And forming a density evaluation sequence by the ratio of the spatial distribution to the projection area of all point locations in the test area, mining a local abnormal compression state, and completing the calibration of the high-density interference area.

Description

Technical Field

[0001] The present invention relates to the field of probe testing technology, and in particular to a PCB probe positioning method, system and storage medium. Background Art

[0002] The field of probe testing technology encompasses methods and equipment used to verify the functionality and performance of electronic components and circuit boards. This technology focuses on verifying and measuring signal paths within circuit systems through contact or non-contact methods, primarily used in electronics manufacturing and packaging testing.

[0003] PCB probe positioning refers to a technical approach used to ensure that test probes are accurately aligned with target pads or pins during printed circuit board testing. When multiple sets of test probes are applied to complex circuit board layouts, precise spatial positioning and calibration are crucial. This involves extracting feature points based on visual image recognition to construct a reference coordinate system. This, combined with the coordinate command parameters of a multi-axis motion platform, enables precise movement and positioning of the probes relative to the target test points.

[0004] Existing technologies primarily rely on visual image recognition to construct a reference coordinate system, using a multi-axis platform to perform simple coordinate positioning. This presents a risk of misjudgment when encountering complex pad layouts or large structural variations. Since alignment is based solely on two-dimensional feature images, it is difficult to identify the three-dimensional structural fluctuations of the pad's actual contact surface, resulting in increased deviation between the probe contact point and the actual contact surface. Spatial path setting fails to consider the impact of dense pad distribution on compressed areas, potentially causing path interference or compression between adjacent probes, increasing the probability of probe contact failure. The lack of detailed error analysis and boundary conflict assessment during point selection can lead to the omission of some critical test points due to interference or improper path setting priorities. During the actual contact verification phase, closed-loop test feedback is not implemented, making it impossible to identify invalid contacts and correct path deviations in real time, compromising overall test integrity and efficiency. For example, in complex circuit boards, some pads may be located in corners or have curved, concave and convex structures. Existing positioning points often deviate from their target due to insufficient height adaptation, resulting in mismeasurements, missed measurements, or repeated tests, impacting the accuracy of finished product test data and the efficiency of subsequent repairs. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a PCB probe positioning method, system and storage medium.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a PCB probe positioning method, comprising the following steps: S1: Obtain the dot height of the sampling points on the surface of the pad contact area, perform reconstruction fitting processing on the pad structure contour, and select the edge points of the pad structure contour with fitting errors as the interference node coordinate set; S2: Obtaining the structural shape of the probe end face, comparing it with the height of the dot matrix of sampling points on the surface of the contactable area of the pad, performing three-dimensional contact area error calculation processing, and obtaining the contact adaptation error; S3: Obtain the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, perform density analysis on the projection area, and divide the abnormal compression rate point set; S4: constructing a three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculating the boundary interference probability of the two sets; S5: Obtain the boundary interference probability and the corresponding contact adaptation error of each test point, perform test path priority sorting and probe positioning, apply a constant test voltage to the corresponding pad of the positioned probe, and judge the validity of the contact point to obtain the probe positioning result.

[0007] As a further solution of the present invention, the specific steps of obtaining the lattice heights of the sampling points on the surface of the pad contactable area, performing reconstruction fitting processing on the pad structure contour, and screening the edge points of the pad structure contour with fitting errors as the interference node coordinate set are as follows: S101: Obtaining the pad surface lattice height through a three-dimensional image collector, extracting the X, Y, and Z coordinates of each point within the pad structure outline, establishing an evenly spaced grid on the XY plane, and associating each grid node with its corresponding Z-axis height value to generate the lattice height of the pad in a spatial state; S102: using the spatial coordinates of each node in the lattice height in the pad space state as input parameters, performing surface reconstruction between nodes using a B-spline surface fitting model to obtain a pad surface fitting height; S103: performing point-by-point residual calculation based on the difference between the pad surface fitting height and the original Z-direction height of the corresponding node, screening out nodes exceeding the local variation range, extracting the node XY coordinate set, and obtaining the interference node coordinate set.

[0008] As a further solution of the present invention, the specific steps of obtaining the probe end face structural shape and comparing it with the dot matrix height of the sampling points on the surface of the pad contactable area to calculate the three-dimensional contact area error and obtain the contact adaptation error are as follows: S201: Acquire the probe end face structural shape, extract the contour projection boundary of the probe bottom structure corresponding to the CAD drawing, identify the coordinates of all boundary points, and construct the probe contour boundary coordinate set based on the projected contour boundary; S202: Calling the probe contour boundary coordinate set, combining the node set in the area intersecting with the coordinate range in the lattice height of the sampling points on the surface of the pad contactable area, performing corresponding point fitting processing between the probe bottom surface and the pad surface nodes in three-dimensional space, and obtaining the contact area node pair coordinates; S203: extracting all normal distances according to the spatial normal distance between each pair of nodes in the contact area node pair coordinates and performing standard deviation calculation to obtain the contact adaptation error.

[0009] As a further solution of the present invention, the specific steps of obtaining the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, performing density analysis of the projection area, and dividing the abnormal compression rate point set are as follows: S301: Obtain the X, Y, and Z coordinates of all test points in the probe target area in a three-dimensional coordinate system, determine the two-dimensional projection boundary range of each point, calculate the corresponding two-dimensional projection boundary area, and generate a comparison result between the three-dimensional coordinates and the projection area; S302: Based on the comparison result of the three-dimensional coordinates and the projected area, the minimum bounding volume of each local point group is calculated using the coordinate set of the test point in space as input, and the volume-area ratio is calculated in combination with the corresponding projected area to obtain a point group density ratio sequence; S303: Based on the point group density ratio sequence, taking the variation range of all density ratios in the region as a judgment criterion, selecting the spatial coordinates corresponding to the point groups that exceed the variation range to obtain an abnormal compression rate point set.

[0010] As a further solution of the present invention, the specific steps of constructing a three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set and calculating the boundary interference probability of the two sets are as follows: S401: Acquire the spatial coordinates of each target point in the abnormal compression rate point set, locate the spatial coordinates of all interference nodes in the interference node coordinate set, and sequentially calculate the minimum Euclidean distance from each target point to the corresponding interference node to obtain a minimum interference distance value set; S402: Based on the minimum interference distance value set and taking a specified spatial scale as a judgment condition, the number of nodes of each target point whose distance from the corresponding interference node within the corresponding scale meets the error segment judgment standard is counted to obtain the error segment frequency of the target point; S403: calling the target point error segment frequency, taking the number of nodes corresponding to each target point as an input parameter, performing a normalization operation according to the distribution of all frequencies in the sample space, and obtaining the boundary interference probability.

[0011] As a further solution of the present invention, the boundary interference probability and the corresponding contact adaptation error of each test point are obtained, the test path priority is sorted and the probe is positioned, a constant test voltage is applied to the pad corresponding to the positioned probe, and the validity of the contact point is determined to obtain the probe positioning result. The specific steps are as follows: S501: Obtain the boundary interference probability and contact adaptation error corresponding to each test point, sort the interference probability scores in descending order, sort the error values in descending order, and perform an intersection operation to obtain a priority coincidence point set number; S502: extracting the starting coordinate information of the path node corresponding to the priority coincidence point set number and the three-dimensional space distance and access angle between the preceding and following adjacent path nodes and combining them to establish a path segment geometric relationship data set; S503: Based on the path segment geometric relationship data group, with the minimum distance and the angle closest to the previous step path direction as the criteria, perform greedy selection sorting on the path node numbers, extract the three-dimensional coordinate information of each node in the sorted path segment as the position reference value for probe positioning, apply a constant test voltage to the pad corresponding to the positioned probe, measure the response current of the circuit connected to the pad and calculate the equivalent resistance value, judge the validity of the contact point, and reposition if the contact point is invalid to obtain the probe positioning result.

[0012] The PCB probe positioning system is used to perform the PCB probe positioning method, and the PCB probe positioning system includes: The interference node extraction module obtains the dot matrix height of the sampling points on the surface of the pad contact area, performs reconstruction fitting processing on the pad structure contour, and selects the edge points of the pad structure contour with fitting errors as the interference node coordinate set; The contact error evaluation module obtains the structural shape of the probe end face, compares it with the height of the dot matrix of the sampling points on the surface of the contactable area of the pad, and performs three-dimensional contact area error calculation processing to obtain the contact adaptation error; The compression rate determination module obtains the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, performs density analysis on the projection area, and divides the abnormal compression rate point set; The boundary interference probability module constructs a three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculates the boundary interference probability of the two sets; The path optimization positioning module obtains the boundary interference probability and the corresponding contact adaptation error of each test point, prioritizes the test path and positions the probe, applies a constant test voltage to the corresponding pad of the positioned probe, and judges the validity of the contact point to obtain the probe positioning result.

[0013] A computer-readable storage medium stores a computer program, which implements the steps of the PCB probe positioning method as described above when executed by a processor.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by comprehensively sampling the height of the dot matrix on the pad surface and reconstructing its spatial contour, the edge point residual extraction makes the identification of interference nodes more accurate. Combined with the CAD boundary recognition of the probe end face structure, the node set formed by the three-dimensional sampling points of the pad is compared, and a contact area error evaluation mechanism is established in the calculation of the normal distance standard deviation to achieve multi-dimensional contact adaptation judgment. The spatial distribution of all points in the test area and the ratio of the projected area constitute a density evaluation sequence, which mines the local abnormal compression state, completes the calibration of the high-density interference area, and improves the global perception ability of the test path. By constructing the minimum distance relationship between the interference node and the abnormal compression point, a normalized probability distribution mechanism is formed, the boundary interference effect is statistically analyzed, and a judgment factor is provided for the subsequent priority path planning to accurately control the interference risk of the contact operation. The path sorting process integrates the spatial distance and angle change factors, introduces continuous geometric relationship analysis, and strikes a balance between positioning accuracy and path sequence rationality. Finally, the electrical performance measurement results are fed back to the contact point validity judgment, and a closed-loop correction mechanism is constructed to achieve the dynamic repositioning capability of the probe. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flow chart of S1 of the present invention; Figure 3 This is a detailed flow chart of S2 of the present invention; Figure 4 This is a detailed flow chart of S3 of the present invention; Figure 5 This is a detailed flow chart of S4 of the present invention; Figure 6 This is a detailed flow chart of S5 of the present invention; Figure 7 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present 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 only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0018] See also Figure 1 The present invention provides a technical solution: a PCB probe positioning method, comprising the following steps: S1: Obtain the dot height of the sampling points on the surface of the pad contact area, perform reconstruction fitting processing on the pad structure contour, and select the edge points of the pad structure contour with fitting errors as the interference node coordinate set; S2: Obtain the structural shape of the probe end face, compare it with the height of the sampling points on the surface of the pad contactable area, perform three-dimensional contact area error calculation and processing, and obtain the contact adaptation error; S3: Obtain the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, perform density analysis on the projection area, and divide the abnormal compression rate point set; S4: Construct the three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculate the boundary interference probability of the two sets; S5: Obtain the boundary interference probability and corresponding contact adaptation error of each test point, prioritize the test paths and locate the probes, apply a constant test voltage to the pads corresponding to the located probes, and determine the validity of the contact points to obtain the probe positioning results.

[0019] See also Figure 2 , obtain the dot matrix height of the sampling points on the surface of the pad contact area, perform reconstruction fitting processing on the pad structure contour, and filter the edge points of the pad structure contour with fitting errors as the interference node coordinate set. The specific steps are as follows: S101: Obtaining the pad surface lattice height through a three-dimensional image collector, extracting the X, Y, and Z coordinates of each point within the pad structure outline, establishing an evenly spaced grid on the XY plane, and associating each grid node with its corresponding Z-axis height value to generate the lattice height of the pad in a spatial state; To obtain the height of the pad surface dot matrix through a 3D image collector, it is necessary to clearly select a high-precision non-contact 3D scanning device, such as a white light interferometric scanner. In a scenario where the pad size is 3mm×3mm, a complete scan is performed on the target pad surface to obtain a 3D coordinate set of sampling points. The sampling density is set to obtain one point every 0.05mm. Then, a regular grid with equal spacing is constructed in the XY plane, forming a 60×60 two-dimensional grid node array, containing a total of 3600 nodes. For each node (i, j), its Z-axis height value needs to be determined. This process uses the inverse distance weighted (IDW) interpolation method, using the n original point cloud data around the node as a reference. The formula is as follows: ; The meanings of the symbols are as follows: : The interpolated Z-axis height of the grid node (i, j) in the i-th row and j-th column (unit: mm), : Z-axis height value of the kth reference origin (unit: mm), : The Euclidean distance from the kth original point to the current grid node (i, j) on the XY plane is calculated as ; Represents the two-dimensional Euclidean distance between the kth original sampling point and the target grid node (i, j) (unit: mm). This value is used as part of the denominator of the weight factor in the inverse distance weighted method to reflect the difference in the contribution of distance to interpolation. Indicates the coordinate of the kth original sampling point in the X-axis direction (unit: mm). This is the actual point coordinate obtained from the 3D scanning collector. Indicates the coordinate of the kth original sampling point in the Y-axis direction (unit: mm). It also comes from the point cloud data obtained by the sampling device. Indicates the coordinate of the target node in the i-th row and j-th column of the two-dimensional grid in the X-axis direction (unit: mm). The coordinate is obtained by dividing the grid structure from the set boundary range. Indicates the Y-axis coordinate of the target node in the i-th row and j-th column in the two-dimensional grid (unit: mm), and Similarly, the grid node coordinates are formed by evenly dividing the XY plane. : Distance weighting index, which determines the degree of influence of distance on the result. Regarding the setting of the weighting exponent p, the value is generally 2. This setting is mainly based on two considerations: first, it conforms to the basic physical property that the closer the points are in actual spatial measurement, the greater the influence on the interpolation result; second, the exponent 2 is widely used in spatial interpolation and has strong stability and numerical convergence, which can effectively suppress the interference of distant points on the interpolation result. For example, in the actual sample, if a reference point is only 0.01mm away from the target node, and the distance to another reference point is 0.10mm, then under the setting of p=2, the weight of the former is 1 / 0.0001=10000, and the weight of the latter is 1 / 0.01=100, which is a difference of 100 times, reflecting the principle that the closer point is more dominant on the interpolation result. This setting makes the interpolation result more reflective of the local surface state, and is particularly suitable for pad surface modeling scenarios with high precision requirements.

[0020] Taking node (10, 15) as an example, assume its two-dimensional coordinates are (0.5mm, 0.75mm), and select n=3 nearest original point cloud data as follows: Point 1: Point 2: Point 3: .

[0021] First, calculate the distance between each original point and the target grid node: ; ; .

[0022] According to the setting of p=2, calculate the weight and interpolation height: molecular: ; Denominator: ; Interpolation result: .

[0023] In this way, the 3,600 nodes are assigned Z-height values one by one, and finally the height mapping of the entire pad area in the spatial state is completed. In this way, the lattice modeling of the actual pad surface is realized, ensuring that the subsequent surface reconstruction and residual analysis have a high-precision input data foundation.

[0024] S102: using the spatial coordinates of each node in the lattice height in the pad space state as input parameters, performing surface reconstruction between nodes using a B-spline surface fitting model to obtain the pad surface fitting height; The spatial coordinates of each node in the dot matrix height under the pad space state are used as input parameters. First, the three-dimensional coordinate data (X, Y, Z) of all nodes are normalized to avoid the influence of the fitting results due to the difference in the magnitude of the coordinate values. The normalization method adopts the range normalization, that is, the coordinate is subtracted from the minimum value of the corresponding dimension and then divided by the difference between the maximum and minimum values. Taking the X direction as an example, assuming Xmin=0, Xmax=3mm, the normalized value of the node X=1.5mm is Similarly, the normalized results of Y and Z can be obtained. After normalization, the data points are converted into dimensionless values within a unified scale to provide stable input for the subsequent fitting model. In the B-spline surface fitting process, the control point matrix, node vector and basis function are used as core elements. The input node matrix is set to 60×60 dimensions, that is, there are 3600 points in total. The B-spline model selects a lower-dimensional control point array to simplify the calculation, set to 11×11 control points (a total of 121), the node vector is uniformly distributed, the parameter direction uses the (u, v) coordinate system, and the definition interval is [0, 1]. The normalized (Xi, Yi) is mapped to the corresponding (u, v). For example, Xi'=0.5, Yi'=0.75, then u=0.5, v=0.75.

[0025] The fitting process uses the following B-spline surface expression: ; The symbols are explained as follows: : The height of the fitting surface under the parameter coordinates (u, v), : The i-th basis function in the u direction, with a degree of p, : The jth basis function in the v direction, with a degree of q, : The Z-direction weight value of the control point (i, j), which participates in the generation of the fitting results. : The maximum row and column index of the control point, which is 10 here, p, q: The number of fittings, usually set to 3 (cubic B-spline). The parameter coordinates (u, v) are obtained by normalizing the nodes in the X and Y directions. The normalization process uses the maximum and minimum coordinate difference of the pad in the corresponding direction as the denominator, and the current coordinate of each node minus the minimum coordinate as the numerator, so as to map the position of each node to the interval [0, 1]. The height value of the control point Pij is obtained by performing least squares fitting or interpolation on the original point cloud data, and is set to a uniform grid structure in the initial stage, and then gradually optimized and adjusted according to the fitting error. The values of the basis functions Ni(u) and Nj(v) depend on the setting of the node vector and the node vector. The node vectors are usually set to be uniformly spaced in the parameter interval of the point. Combined with the number of fitting times (for example, three times), the values of the basis functions of each order are calculated step by step in a recursive manner. The number of control points m and n is determined by the resolution of the input point array and the desired fitting accuracy. For example, selecting an 11×11 control point array under a 60×60 input grid can ensure accuracy while controlling the amount of calculation. The number of fitting times p and q is generally set to 3, which is an empirical setting and is suitable for most engineering surface fitting problems. Finally, all parameters are collaboratively determined by constrained optimization during the fitting process to determine the fitting height output of each node.

[0026] Taking a certain node (u=0.6, v=0.4) as an example, it is necessary to calculate the values of several basis functions in the interval in the parameter space during the fitting process. The basis function is recursively generated using the Cox-deBoor recursive formula, and the non-zero interval of the basis function in the sub-region of the corresponding control point is intercepted. For example, the interval in the u direction is [U5, U6], and the interval in the v direction is [V4, V5]. A total of 16 surrounding control points Pij are affected. The influence factor between each control point and the node (u, v) is calculated separately, and then multiplied by the Z value of each control point to obtain the cumulative effect. Fitting value: If there is a significant slope change in a certain area, such as an edge jump area, the fitting model will be refined by adjusting the control point density or increasing the order of the basis function to ensure that the surface fitting has sufficient resolution in the key area. The final fitting height matrix is 60×60 dimensions, which is consistent with the original input.

[0027] Assume that some control points and basis function values are as follows (simplified to linear representation): Control point height: ; The corresponding basis function values are: ; The fitting value of this point is: .

[0028] The above process needs to be executed on all nodes in sequence, and the surface fitting height Z_fit(i, j) is reconstructed point by point to form a complete three-dimensional surface fitting data of the pad, providing a continuous surface model reference for subsequent residual calculation.

[0029] S103: performing point-by-point residual calculation based on the difference between the pad surface fitting height and the original Z-direction height of the corresponding node, screening out nodes that exceed the local variation range, extracting the node XY coordinate set, and obtaining the interference node coordinate set; According to the difference between the pad surface fitting height and the original Z-direction height of the corresponding node, perform point-by-point residual calculation, and set the original height of the (i, j) node to be , the fitting height is , then the residual value calculation formula of this point is: ; in: : represents the residual value of the grid node in row i and column j (unit: mm), : is the Z-axis height value obtained from the original 3D image acquisition (unit: mm), : Z-axis fitting value of the node after B-spline fitting (unit: mm).

[0030] Taking the (20, 35) node as an example, if its original Z value is 0.325 mm and the fitted Z value is 0.307 mm, the residual is: .

[0031] The residual values of all nodes are calculated in sequence to form a complete residual matrix. Then the residual values are screened to determine whether they constitute interference nodes. A local variation range threshold, that is, the maximum allowable residual range, needs to be set. Set this threshold as , if the node residual It is considered an interference point. The threshold setting principle is based on manufacturing tolerances and pad flatness requirements. In high-precision electronic component placement scenarios, the recommended setting range is ±0.015mm. The reasons are as follows: Usually, the BGA or CSP package has a tolerance for pad surface flatness error between ±0.01mm and ±0.02mm. Taking into account the measurement error of the equipment and the actual process fluctuation, ±0.015mm is used as the critical value to balance reliability and recognition sensitivity. Therefore, if the residual value of a node is 0.018mm, the point is determined to be an interference point. If a node is 、 , the residual is 0.012mm, which means it does not constitute an interference point. The XY coordinates of all nodes that are judged to be interference will be extracted to form an interference node coordinate set. The specific extraction form is a two-dimensional coordinate array, such as: ; To avoid misidentification due to occasional errors, further density assessment and spatial clustering are required. This method can use a density threshold setting, whereby an interference region is considered to exist when more than n interference points exist within any radius r. For example, assuming a radius of r = 0.1mm and a threshold of n = 5, if there are seven interference points within 0.1mm around the node (1.20, 1.35), it is identified as an interference region. This density threshold setting is based on the actual pad contact area and defect distribution size, and is applicable to designs with a typical solder joint radius of approximately 0.15mm and a spacing of 0.4mm. The resulting set of interference node coordinates will be used for subsequent repair modeling or process compensation input.

[0032] See also Figure 3 , obtain the probe end face structure shape, compare it with the dot matrix height of the sampling points on the contactable area of the pad to calculate the three-dimensional contact area error, and obtain the contact adaptation error in the following specific steps: S201: Acquire the probe end face structural shape, extract the contour projection boundary of the probe bottom structure corresponding to the CAD drawing, identify the coordinates of all boundary points, and construct the probe contour boundary coordinate set based on the projected contour boundary; To obtain the structural shape of the probe end face, it is necessary to base it on the CAD drawings or actual measurement data of the probe manufacturing. First, read the two-dimensional projection contour corresponding to the bottom structure from the CAD file of the probe bottom surface. The projection is usually composed of closed polygons or curves. The contour boundary is extracted point by point using the graphics processing module. Assuming the graphics is a polyline structure, the point-edge tracking algorithm is used to extract all vertex coordinates from the starting point in sequence. During the extraction process, each edge is interpolated to form a high-density point set for the contour boundary. For example, if 20 equidistant points are generated for each edge interpolation, a polygon with 10 edges can generate 200 boundary points. For these points, their locations under the CAD plane are recorded one by one. - Coordinates, establish a unified coordinate system, remove duplicate points and construct the final probe contour boundary coordinate set. If the graph is an irregular curve contour, it is necessary to use the boundary curvature analysis method to sample boundary points at equal intervals. For example, the sampling interval is set to 0.01mm to ensure that the true shape of the contour can be fully captured. This process corresponds to the pad contact area analysis scenario. For example, if the bottom surface of a probe is elliptical, its major axis is 1.2mm, and its minor axis is 0.8mm, then the projection contour formula is ,in are the coordinates of any point on the projection contour, 0.6 and 0.4 are the lengths of the semi-major axis and semi-minor axis respectively. On this basis, the coordinate points are uniformly sampled to form a boundary set.

[0033] S202: Calling the probe contour boundary coordinate set, combining the node set in the area intersecting with the coordinate range in the lattice height of the sampling points on the surface of the pad contactable area, performing corresponding point fitting processing between the probe bottom surface and the pad surface nodes in three-dimensional space, and obtaining the coordinates of the contact area node pairs; Call the probe contour boundary coordinate set, combine the node set in the lattice height of the sampling points on the surface of the pad contact area that intersects with the coordinate range, first align the probe contour boundary coordinates with the pad surface coordinate system through coordinate transformation, the transformation parameters include rotation matrix and translation vector, according to the spatial distribution of the pad surface lattice, select the node point set that intersects with the probe bottom surface coverage area, the judgment method is to perform a set of calculations on each pad node. Determine whether it falls within the closed area formed by the probe boundary, use the ray method or the parity principle of points in the polygon to identify, add the points that meet the conditions to the matching pool, form a pad area dot matrix that may have contact with the probe bottom surface, and then project a set of candidate contact points from the probe bottom surface boundary according to the preset spacing. Use the shortest distance matching strategy to find the nearest node on the pad surface for each probe bottom surface point to form a one-to-one correspondence. If two points If the height difference is within the set distance threshold, such as 0.01mm, the point is considered to be the actual contact point, and the contact area node pair coordinates are finally formed. For example, if 300 boundary points are selected from the bottom boundary of the probe, and 182 pairs of valid contact pairing points are obtained after matching with the pad area dot matrix, then these 182 groups of node pairs are used as the basis for subsequent spatial distance calculations to ensure that all contact point pairs have a reasonable spatial structure. Represents the two-dimensional coordinates of the pad point, Indicates its corresponding height value, and the bottom point of the probe is represented by .

[0034] S203: extracting all normal distances based on the spatial normal distances between each pair of nodes in the contact area node pair coordinates and performing standard deviation calculation to obtain the contact adaptation error; According to the spatial normal distance between each pair of nodes in the contact area node pair coordinates, all normal distances are extracted and the standard deviation calculation is performed. First, for each pair of paired nodes , respectively calculate the distance vector between the two in the three-dimensional coordinate system, and then combine the normal vector information of the pad surface fitting area, perform vector projection operation, extract the component of each group of points in the normal vector direction as the normal distance, and set the pad surface at the node The normal direction at , then the normal distance is a vector exist The projection length on the point is calculated to form a normal distance sequence ,in: Represents the three-dimensional coordinates of a contact point on the bottom surface of the probe , Indicates the three-dimensional coordinates of the contact point on the pad surface that is paired with it , is the unit normal vector derived from the fitting surface at the pad point, for arrive The direction vector is in the normal direction The projection value on the pad represents the actual fitting distance between the two contact surfaces at that point. After repeating this calculation operation for all contact pairs, a set of numerical results is formed. Then, the difference fluctuation analysis of this set of values is performed to calculate its overall dispersion, thereby evaluating the uniformity of contact between the pad surface and the probe surface. In actual operation, if the number of contact node pairs is 182, all The values are mainly distributed between 0.001mm and 0.022mm, with an average value of 0.012mm. On this basis, the deviation average method or existing software modules can be used to perform discrete evaluation on this set of data sets. If the maximum deviation degree does not exceed the pre-set error threshold (such as 0.007mm), it is considered that the contact error meets the target process requirements. The error measurement result is used as the basis for judging the matching accuracy between the probe and the pad.

[0035] See also Figure 4 , obtain the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, perform density analysis of the projection area, and divide the abnormal compression rate point set into the following specific steps: S301: Obtain the X, Y, and Z coordinates of all test points in the probe target area in a three-dimensional coordinate system, determine the two-dimensional projection boundary range of each point, calculate the corresponding two-dimensional projection boundary area, and generate a comparison result between the three-dimensional coordinates and the projection area; Obtain the positions of all test points in the probe scanning area in the three-dimensional coordinate system. Automated scanning operations can be performed using three-dimensional imaging equipment such as three-dimensional laser radar and structured light camera to record their spatial coordinates point by point. The three-dimensional coordinates of the test points are set as ,in: : No. The position coordinate of a point in the X-axis direction, in millimeters (mm); : No. The position coordinate of a point in the Y-axis direction, in millimeters (mm); : No. The position coordinates of each point in the Z-axis direction are in millimeters (mm). Then, a two-dimensional projection is performed on each point, and only its and value, ignoring the vertical direction , forming a planar point set , these points are then used to construct a two-dimensional boundary region. After the boundary point sequence is obtained using the QuickHull convex hull algorithm, it is connected to form a closed polygon, and the area of the region is calculated using the Shoelace formula. The formula is as follows: ; in: : the area formed by the two-dimensional projection boundary, in square millimeters (mm²); : The number of vertices that make up the boundary; : No. The coordinates of the boundary vertices.

[0036] For example, suppose the boundary vertices are three points: , then the area is calculated as: , finally, for each set of three-dimensional points Bind an area value , establish a mapping relationship and form a comparison table for processing in the next stage.

[0037] S302: Based on the comparison results of the three-dimensional coordinates and the projected area, the minimum bounding volume of each local point group is calculated using the coordinate set of the test point in space as input. The volume-area ratio is calculated in combination with the corresponding projected area to obtain a point group density ratio sequence; The three-dimensional coordinates of each test point As the basic input, the density clustering algorithm is used to automatically divide the spatial points. The commonly used DBSCAN algorithm can set the neighborhood radius , the minimum number of points is set to 4, and the clustering results in multiple point groups, each of which constitutes a local space. For each point group, its minimum circumscribed volume needs to be calculated. This operation can be achieved using axis-aligned bounding boxes (AABBs). The calculation steps are: find the maximum and minimum coordinates of all points in the point group; calculate the bounding box size , , . Calculate the volume: ,in: : spatial extension length along each axis, in millimeters; : The volume of the bounding box, in cubic millimeters (mm³). For example, the maximum and minimum coordinates of a point group are as follows: Extract the maximum and minimum values of the point group in the X direction and set them as 、 ,but: , then, extract the maximum and minimum values in the Y direction and set them as 、 ,but: Finally, extract the maximum and minimum values in the Z direction and set them as 、 ,but: . Calculate the volume: , the corresponding projected area , then the point group density ratio is: , perform this process on all point groups one by one, and after calculating all density ratios, form a series of data sets consisting of point group density results.

[0038] S303: Based on the point group density ratio sequence, taking the variation range of all density ratios in the region as a judgment criterion, selecting the spatial coordinates corresponding to the point groups that exceed the variation range to obtain an abnormal compression ratio point set; Perform statistical analysis on all point group density ratios in paragraph 2. Assume that all ratios are recorded as a set of floating-point numbers. Calculate the mean and standard deviation of this set. The formula is as follows: , ; in: : No. The density ratio of the point groups; : the average value of the density ratio of all point groups; : Standard deviation, which measures the fluctuation of density ratio; : Total number of point groups.

[0039] Sample set For example: Calculate the average of these values: ; Next, calculate the standard deviation using the formula: ; Set the judgment interval with twice the standard deviation as the upper and lower limits: , If the density ratio exceeds this range, for example, the sample median value is 12, which is critically high and can be marked as an abnormal point group.

[0040] The coordinate boundary range of the point group is as follows: The X-direction boundary is ; The Y-direction boundary is ; The Z-direction boundary is All test points in this spatial region are listed as abnormal compression rate point sets as the final output.

[0041] See also Figure 5 , the specific steps of constructing the three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set and calculating the boundary interference probability of the two sets are as follows: S401: Obtain the spatial coordinates of each target point in the abnormal compression rate point set, locate the spatial coordinates of all interference nodes in the interference node coordinate set, and sequentially calculate the minimum Euclidean distance from each target point to the corresponding interference node to obtain a minimum interference distance value set; To obtain the spatial coordinates of each target point within the abnormal compression rate point set, first read the boundary coordinate range of the point group identified as abnormal in the previous article. For example, it contains multiple three-dimensional coordinate points, such as point A (10.0, 20.0, 5.0), point B (11.2, 19.5, 5.3), and point C (9.8, 21.1, 4.9). The X, Y, and Z three-dimensional position values are recorded point by point. The interference node coordinate set is derived from the design model or structural model and may include component connection points, clamping device nodes, and adjacent tooling support points. For example, node N1 is (9.5, 19.5, 5.5), node N2 is (12.0, 22.0, 6.5), and node N3 is (11.0, 20.5, 5.0). The spatial coordinates of all interference nodes are unified and organized into a searchable data structure. Taking target point A as an example, we calculate its distances to points N1, N2, and N3, yielding the following values: 0.87 mm from N1, 3.04 mm from N2, and 1.12 mm from N3. The smallest of these values, 0.87 mm, is selected as the minimum interference distance for point A. This process is repeated for all target points, extracting the minimum spatial distance between each point and the nearest interfering node. Ultimately, a set of minimum interference distance values is generated, such as [0.87, 1.45, 0.62, 1.11, 0.93] mm, for subsequent statistical analysis.

[0042] S402: Based on the minimum interference distance value set and taking the specified spatial scale as the judgment condition, the number of nodes of each target point within the corresponding scale whose distance from the corresponding interference node meets the error segment judgment standard is counted to obtain the error segment frequency of the target point; Based on the set of minimum distance values between each target point and its nearest interfering node, a spatial scale is set for distance determination. For example, if the detection scale is set to 1.0 mm and the error range is ±0.2 mm, then a valid interference determination event is considered if the actual distance between the target point and the interfering node falls between 0.8 mm and 1.2 mm. For each target point, the distance values calculated from all the interfering nodes around it are screened, and the number of distance values falling within this error range is counted. This number is used as the error range matching frequency for the current target point. For example, the distances of target point B to multiple interfering nodes are 0.75 mm, 1.05 mm, 1.30 mm, 0.90 mm, and 1.22 mm, respectively. Three of these distances fall within the error range of 0.8 to 1.2 mm, resulting in a frequency of 3. The distances of target point C are 0.60 mm, 0.77 mm, 1.50 mm, and 1.18 mm. Only 1.18 mm meets the error range criteria, resulting in a frequency of 1. Through this judgment process, each target point generates an integer value, representing the number of nodes that meet the interference criteria within the error range. The frequencies of all target points are combined into a set of statistical results, such as [3, 1, 2, 4, 0, 5], which provide the input parameters for calculating the boundary interference probability.

[0043] S403: Call the target point error segment frequency, take the number of nodes corresponding to each target point as an input parameter, perform a normalization operation based on the distribution of all frequencies in the sample space, and obtain the boundary interference probability; After calling the target point error segment frequency data, the frequency value corresponding to each target point needs to be used as an input variable. Based on the overall distribution of the frequencies of all target points in the sample space, normalization is performed to finally calculate the boundary interference probability. This normalization operation should not use stacking correction coefficients or undefined constants, but should be based on the actual physical meaning and quantity of the statistical distribution to ensure dimensional consistency and data logic stability. The specific operation process is as follows: First, let the error segment frequencies of all target points constitute a sample set, recorded as ,in Indicates the The number of interference nodes matched by the target point, in units of "". In order to fully express the distribution characteristics of the frequency in the sample space, a linear normalization method based on standard deviation is used, and the minimum value of the minimum interference distance value set is calculated in combination with the above With the maximum value , in order to establish the relative distance sensitivity constraint factor of the probability output, and construct the mapping function through the relationship between frequency intensity and spatial interference extreme value. The normalization process adopts the following process: ; The parameters are defined as follows: : No. The boundary interference probability of target points is dimensionless and normalized to interval; : No. The interference matching frequency of target points, in units of pieces; : The minimum and maximum frequencies in the sample set, in units of pieces; : No. The minimum interference distance between target points, in millimeters (mm); : The minimum and maximum interference distances among all target points, in millimeters (mm).

[0044] The formula structure consists of two parts: 1. The first part is frequency normalization, which directly measures the relative position of the point in the entire sample frequency distribution; 2. The second part is the "interference proximity penalty term", which is used to suppress high-frequency points at long distances, so that the weight of points with close interference is increased, reflecting the risk weight of the physical spatial compression effect. For example, a target point , the frequency interval of the whole sample is , then the normalized value of the first part is: , assuming that the interference distance is , the interference distance interval of all target points is , then the distance penalty part is: , and the final boundary interference probability is: , if converted to percentage, it is The above calculation process is performed sequentially on all target points to form a list of boundary interference probabilities, ensuring that each target point corresponds to a normalized risk probability value. This value not only reflects the abnormal degree of its error segment frequency, but also combines its relative proximity to the interference source in the actual physical space, and has complete measurement significance. If you need to further establish a risk classification mechanism for different probability levels, you can set the segmentation as follows: : Low risk, : Medium risk, High risk. This method quantitatively models the interference situation at the target point and can be directly used in subsequent boundary behavior modeling or tolerance verification processes.

[0045] See also Figure 6 , obtain the boundary interference probability and corresponding contact adaptation error of each test point, prioritize the test paths and locate the probes, apply a constant test voltage to the pads corresponding to the located probes, and judge the validity of the contact points. The specific steps to obtain the probe positioning results are as follows: S501: Obtain the boundary interference probability and contact adaptation error corresponding to each test point, sort the interference probability scores in descending order, sort the error values in descending order, and perform an intersection operation to obtain a priority coincidence point set number; Obtain the boundary interference probability and contact adaptation error corresponding to each test point, and integrate these two data items from different detection modules. The boundary interference probability data is derived from the previously normalized probability value set, for example: [0.92, 0.78, 0.61, 0.45, 0.33]. The contact adaptation error is the deviation index measured at each point during the contact test, for example: [0.35, 0.48, 0.20, 0.75, 0.30]. Sort the interference probability data in descending order to generate a prioritized test point numbering sequence. Then, sort the contact adaptation error values in descending order to generate a list of points with the largest errors. Perform an intersection operation on the two sorted results to find the point numbers that rank at the top of both sequences, forming a prioritized point set. For example, if points 1, 3, and 4 rank in the top three in both sequences, output the point set numbered [1, 3, 4], which serves as the target position set for subsequent path node optimization and probe testing.

[0046] S502: extracting the starting coordinate information of the path node corresponding to the priority coincidence point set number and the three-dimensional space distance and access angle between the preceding and following adjacent path nodes and combining them to establish a path segment geometric relationship data set; Extract the starting spatial coordinate position of each point number in the detection path in the priority coincidence point set, retrieve the previous and next nodes corresponding to the point in the path sequence, and read their three-dimensional coordinates. Calculate the spatial distance and access angle between the node and the two adjacent nodes, where the distance can be judged based on the straight line length of the three points, and the angle can be determined by the difference in the direction of the angle formed by the nodes. For example, a point is numbered P5, and the corresponding spatial coordinates are (10, 15, 5). The front and rear nodes are (9, 14, 5) and (11, 15.5, 5.5), respectively. The distances are 1.73 mm and 1.80 mm, respectively, and the angle direction changes by 5 degrees. Combine the three pieces of information of each point: the starting coordinates, the distance between the front and rear nodes, and the difference in the angle direction, to construct a path segment geometric relationship data group, which is recorded as a structured data table for subsequent path sorting.

[0047] S503: Based on the path segment geometric relationship data set, a greedy selection sort is performed on the path node numbers, with the minimum distance and the angle closest to the previous path direction as the criteria. The three-dimensional coordinate information of each node in the sorted path segment is extracted as a position reference value for probe positioning. A constant test voltage is applied to the pad corresponding to the positioned probe, and the response current of the circuit connected to the pad is measured and the equivalent resistance value is calculated to determine the validity of the contact point. If the contact point is invalid, the probe is repositioned to obtain the probe positioning result. Based on the geometric relationship data set of the path segments, a greedy selection algorithm is used during the path planning process to sort the path node numbers. The sorting goal is: starting from the starting node, the node with the closest distance and the smallest direction change among the nodes that have not been visited yet is selected as the next path node, realizing a dynamic selection mechanism for point-to-point path construction. Each selection is determined by a minimum evaluation function that also considers the three-dimensional distance between the current node and the candidate node. and direction change , the specific evaluation function is as follows: ; in: : No. The evaluation value of each candidate node, in millimeters (mm). The smaller the evaluation value, the better the node. : The Euclidean space distance between the current path end node and the candidate node, in millimeters (mm); : The angle between the candidate node's incident direction and the current path direction, in radians; : Pi, used as the denominator for angle normalization, ensuring that the angle scale value is [0,1].

[0048] In the formula, "directional deviation" is added to the distance as a product term, reflecting that the greater the deviation in direction, the greater the evaluation value, and the less likely the system is to select that node. If the distance between a node and the current node is 1.5 mm, and the directional deviation angle is 18°, or approximately 0.314 radians, then the calculation is: This value is compared with the results of other candidate nodes. The system automatically selects the node with the lowest evaluation value as the next path node, and proceeds backward in sequence until all priority coincident point sets are sorted, completing the path node sequence. After sorting, the three-dimensional coordinates are read point by point in this order and used as spatial positioning reference values for the probe control system, controlling the robotic arm or micro-drive assembly to position the probe to the corresponding pad surface. Subsequently, a set test voltage (e.g., 3.3 volts) is applied to the target pad, and the response current generated by its connected circuit is read (e.g., 8 mA), corresponding to a calculated equivalent resistance of 0.4125 ohms). A determination is made as to whether this resistance is within the allowable range. If not, the current node is marked as contact failure, and the drive system automatically retracts and adjusts its angle or position for repositioning. The test voltage is reapplied, the current is collected, and the judgment process is repeated. Ultimately, a complete set of probe positioning and contact validity detection records for the path sorted according to the greedy strategy is generated for subsequent data analysis or structural modification.

[0049] See also Figure 7 The PCB probe positioning system is used to perform the PCB probe positioning method. The PCB probe positioning system includes: The interference node extraction module obtains the dot matrix height of the sampling points on the surface of the pad contact area, performs reconstruction fitting processing on the pad structure contour, and selects the edge points of the pad structure contour with fitting errors as the interference node coordinate set; The contact error evaluation module obtains the structural shape of the probe end face, compares it with the height of the sampling points on the surface of the pad contactable area, and calculates the three-dimensional contact area error to obtain the contact adaptation error; The compression rate determination module obtains the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, performs density analysis on the projection area, and divides the abnormal compression rate point set; The boundary interference probability module constructs the three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculates the boundary interference probability of the two sets; The path optimization positioning module obtains the boundary interference probability and corresponding contact adaptation error of each test point, prioritizes the integrated test path, and obtains the probe positioning result.

[0050] A computer-readable storage medium stores a computer program, which implements the steps of the above-mentioned PCB probe positioning method when executed by a processor.

[0051] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. PCB probe positioning method, characterized in that, The following steps are involved: S1: Obtain the dot height of the sampling points on the surface of the pad contact area, perform reconstruction fitting processing on the pad structure contour, and select the edge points of the pad structure contour with fitting errors as the interference node coordinate set; S2: Obtaining the structural shape of the probe end face, comparing it with the height of the dot matrix of sampling points on the surface of the contactable area of the pad, performing three-dimensional contact area error calculation processing, and obtaining the contact adaptation error; S3: Obtain the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, perform density analysis on the projection area, and divide the abnormal compression rate point set; S4: constructing a three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculating the boundary interference probability of the two sets; S5: Obtain the boundary interference probability and the corresponding contact adaptation error of each test point, perform test path priority sorting and probe positioning, apply a constant test voltage to the corresponding pad of the positioned probe, and judge the validity of the contact point to obtain the probe positioning result.

2. The PCB probe positioning method according to claim 1, wherein: The specific steps of obtaining the dot height of the sampling points on the surface of the pad contact area, performing reconstruction fitting processing on the pad structure contour, and screening the edge points of the pad structure contour with fitting errors as the interference node coordinate set are as follows: S101: Obtaining the pad surface lattice height through a three-dimensional image collector, extracting the X, Y, and Z coordinates of each point within the pad structure outline, establishing an evenly spaced grid on the XY plane, and associating each grid node with its corresponding Z-axis height value to generate the lattice height of the pad in a spatial state; S102: using the spatial coordinates of each node in the lattice height in the pad space state as input parameters, performing surface reconstruction between nodes using a B-spline surface fitting model to obtain a pad surface fitting height; S103: performing point-by-point residual calculation based on the difference between the pad surface fitting height and the original Z-direction height of the corresponding node, screening out nodes exceeding the local variation range, extracting the node XY coordinate set, and obtaining the interference node coordinate set.

3. The PCB probe positioning method according to claim 1, wherein: The specific steps of obtaining the probe end face structural shape and comparing it with the dot matrix height of the sampling points on the contactable area of the pad to calculate the three-dimensional contact area error and obtain the contact adaptation error are as follows: S201: Acquire the probe end face structural shape, extract the contour projection boundary of the probe bottom structure corresponding to the CAD drawing, identify the coordinates of all boundary points, and construct the probe contour boundary coordinate set based on the projected contour boundary; S202: Calling the probe contour boundary coordinate set, combining the node set in the area intersecting with the coordinate range in the lattice height of the sampling points on the surface of the pad contactable area, performing corresponding point fitting processing between the probe bottom surface and the pad surface nodes in three-dimensional space, and obtaining the contact area node pair coordinates; S203: extracting all normal distances according to the spatial normal distance between each pair of nodes in the contact area node pair coordinates and performing standard deviation calculation to obtain the contact adaptation error.

4. The PCB probe positioning method according to claim 1, wherein: The specific steps for obtaining the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, performing density analysis on the projection area, and dividing the abnormal compression rate point set are as follows: S301: Obtain the X, Y, and Z coordinates of all test points in the probe target area in a three-dimensional coordinate system, determine the two-dimensional projection boundary range of each point, calculate the corresponding two-dimensional projection boundary area, and generate a comparison result between the three-dimensional coordinates and the projection area; S302: Based on the comparison result of the three-dimensional coordinates and the projected area, the minimum bounding volume of each local point group is calculated using the coordinate set of the test point in space as input, and the volume-area ratio is calculated in combination with the corresponding projected area to obtain a point group density ratio sequence; S303: Based on the point group density ratio sequence, taking the variation range of all density ratios in the region as a judgment criterion, selecting the spatial coordinates corresponding to the point groups that exceed the variation range to obtain an abnormal compression rate point set.

5. The PCB probe positioning method according to claim 4, characterized in that: For calculating the corresponding two-dimensional projected boundary area , using the formula: ； in: is the number of vertices that make up the boundary; It is The X coordinate of the boundary point, in millimeters, It is The Y coordinate of each boundary point, in millimeters, and Represents the X and Y coordinates of the next vertex respectively.

6. The PCB probe positioning method according to claim 1, wherein: The specific steps of constructing the three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set and calculating the boundary interference probability of the two sets are as follows: S401: Acquire the spatial coordinates of each target point in the abnormal compression rate point set, locate the spatial coordinates of all interference nodes in the interference node coordinate set, and sequentially calculate the minimum Euclidean distance from each target point to the corresponding interference node to obtain a minimum interference distance value set; S402: Based on the minimum interference distance value set and taking a specified spatial scale as a judgment condition, the number of nodes of each target point whose distance from the corresponding interference node within the corresponding scale meets the error segment judgment standard is counted to obtain the error segment frequency of the target point; S403: calling the target point error segment frequency, taking the number of nodes corresponding to each target point as an input parameter, performing a normalization operation according to the distribution of all frequencies in the sample space, and obtaining the boundary interference probability.

7. The PCB probe positioning method according to claim 6, characterized in that: For the boundary interference probability, the formula is used: ； in, It is The boundary interference probability of target points, It is The interference matching frequency of target points, are the minimum and maximum frequencies in the sample set, It is The minimum interference distance of target points, are the minimum and maximum interference distances among all target points.

8. The PCB probe positioning method according to claim 1, wherein: The specific steps of obtaining the boundary interference probability and the corresponding contact adaptation error of each test point, prioritizing the test paths and positioning the probes, applying a constant test voltage to the pads corresponding to the positioned probes, and judging the validity of the contact points to obtain the probe positioning results are as follows: S501: Obtain the boundary interference probability and contact adaptation error corresponding to each test point, sort the interference probability scores in descending order, sort the error values in descending order, and perform an intersection operation to obtain a priority coincidence point set number; S502: extracting the starting coordinate information of the path node corresponding to the priority coincidence point set number and the three-dimensional space distance and access angle between the preceding and following adjacent path nodes and combining them to establish a path segment geometric relationship data set; S503: Based on the path segment geometric relationship data group, with the minimum distance and the angle closest to the previous step path direction as the criteria, perform greedy selection sorting on the path node numbers, extract the three-dimensional coordinate information of each node in the sorted path segment as the position reference value for probe positioning, apply a constant test voltage to the pad corresponding to the positioned probe, measure the response current of the circuit connected to the pad and calculate the equivalent resistance value, judge the validity of the contact point, and reposition if the contact point is invalid to obtain the probe positioning result. 9.PCB probe positioning system, characterized in that, The PCB probe positioning system is used to perform the PCB probe positioning method according to any one of claims 1 to 8, and the PCB probe positioning system includes: The interference node extraction module obtains the dot matrix height of the sampling points on the surface of the pad contact area, performs reconstruction fitting processing on the pad structure contour, and selects the edge points of the pad structure contour with fitting errors as the interference node coordinate set; The contact error evaluation module obtains the structural shape of the probe end face, compares it with the height of the dot matrix of the sampling points on the surface of the contactable area of the pad, and performs three-dimensional contact area error calculation processing to obtain the contact adaptation error; The compression rate determination module obtains the spatial distribution coordinate set of all test points in the probe target area and the corresponding XY projection area, performs density analysis on the projection area, and divides the abnormal compression rate point set; The boundary interference probability module constructs a three-dimensional spatial relationship between the interference node coordinate set and the abnormal compression rate point set, and calculates the boundary interference probability of the two sets; The path optimization positioning module obtains the boundary interference probability and the corresponding contact adaptation error of each test point, prioritizes the test path and positions the probe, applies a constant test voltage to the corresponding pad of the positioned probe, and judges the validity of the contact point to obtain the probe positioning result.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the PCB probe positioning method according to any one of claims 1 to 8 are implemented.

Images