Multidirectional slice type on-machine measurement method and system

Abstract

The invention discloses a multidirectional slice type on-machine measurement method and system, solves the problems that in the prior art, an on-line detection strategy of a workpiece is simple and interference is prone to occurring, and has the beneficial effects that interference is avoided, and the measurement accuracy is improved. Comprising the steps that a probe of a distance measuring sensor in the optical on-machine measuring device is detachably installed on a main shaft of a machine tool and can rotate relative to the main shaft of the machine tool, a workpiece is installed on a workbench, and the probe of the distance measuring sensor emits light beams towards the workpiece; selecting proper slicing parameters to segment the workpiece theoretical model, and extracting a feature contour; path design is carried out based on the extracted feature contour, and one or more of a contour path, a broken line path and a step path are selected as paths according to the surface curvature and shape characteristics of the workpiece; introducing a topological relation model and a bounding box modeling mechanism; filling invalid points, and extracting valid data; and carrying out coordinate data fusion.

Description

Technical Field

[0001] The present invention relates to the field of surface precision measurement, and in particular to a multi-directional slicing on-machine measurement method and system. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Traditional off-machine measurement methods require removing the part from the machine tool fixture and transporting it to the measurement room before inspection. This repeated clamping process can easily lead to secondary errors. However, on-machine measurement technology uses the machine tool hardware as a carrier and is equipped with corresponding measurement tools. This allows the workpiece to be measured directly on the processing equipment. Whether in the process or after processing, measurements can be performed while the workpiece remains in place, avoiding errors caused by repeated clamping.

[0004] On-machine measurement requires consideration of sensor acquisition characteristics and pre-set scanning paths, with the ultimate goal of converting sampled data into a point cloud model. However, during this process, the following issues still need to be addressed:

[0005] (1) The existing on-machine measurement system has an unclear scanning strategy. After the system determines the path, it randomly moves along the set path. The safety of the path cannot be guaranteed and interference with the parts may occur, resulting in large deviations in the obtained data.

[0006] (2) For parts of various shapes, the existing measurement methods use the same path selection, which not only affects the accuracy of the measurement results, but also requires complex spatial coordinate analysis to integrate it with the preset path information and construct a complete surface;

[0007] (3) The device lacks flexibility, which makes it impossible to efficiently measure the contours of parts with complex and diverse shapes, affecting the measurement accuracy and scope of application;

[0008] (4) Existing research has not fully considered the sensor acquisition characteristics in terms of invalid point processing and data segmentation of time series data, which affects the data quality and the accuracy of subsequent analysis. Summary of the Invention

[0009] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a multi-directional slicing on-machine measurement method with a clear workflow, high flexibility of the entire device, and effective guarantee of measurement accuracy, data quality and accuracy of subsequent analysis.

[0010] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0011] A multi-directional slice-type on-machine measurement method, comprising the following contents:

[0012] The probe of the distance measuring sensor in the optical on-machine measuring device is detachably mounted on the main shaft of the machine tool, the probe being rotatable relative to the main shaft of the machine tool, the workpiece is mounted on the workbench, and the probe of the distance measuring sensor emits a light beam toward the workpiece;

[0013] The workpiece is divided into several planes, and the appropriate slice parameters are selected for each plane to segment the workpiece theoretical model and extract the feature contour;

[0014] Path design is performed based on the extracted feature contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the distance sensor moves along the path to form a theoretical contour line.

[0015] Introducing the topological relationship model and bounding box modeling mechanism, integrating logical topological relationship judgment into path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position;

[0016] According to the collision detection results, fill the invalid points and extract the valid data after filling the invalid points;

[0017] Based on the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system;

[0018] Repeat the above steps to obtain the coordinates of the sampling points in all planes in the machine tool coordinate system.

[0019] In the multi-directional slicing on-machine measurement method described above, the distance measuring sensor is installed on the spindle of the machine tool through a mounting component, the mounting component includes a connecting rod that can be detachably connected to the spindle, the connecting rod is connected to a connecting block, the connecting block is rotatably connected to the mounting block, the mounting block is provided with an open groove for installing the probe of the distance measuring sensor, and the connecting block drives the probe of the distance measuring sensor to rotate.

[0020] In the multi-directional slicing on-machine measurement method described above, if the theoretical model of the workpiece is a surface model, the characteristic contour is extracted through the cross-section; if the theoretical model of the workpiece is a point cloud model, the plane characteristic points are extracted through the point cloud coordinates, and then the characteristic contour is extracted using curve interpolation;

[0021] If the theoretical model of the workpiece is a point cloud model, the slicing position is determined by using a reference plane, the screening range is expanded by using two other parallel boundary planes, and the screened point cloud is projected onto the reference plane for interpolation to obtain the characteristic contour.

[0022] In the multi-directional slicing on-machine measurement method described above, the contour path is selected for a measurement area where the curvature of the workpiece surface is less than a set value and the shape is regular;

[0023] For a measurement area with a broken line on the workpiece surface, selecting the broken line path;

[0024] For measurement areas with layered or stepped structures, the described step path is selected.

[0025] In the multi-directional slicing on-machine measurement method described above, the formation of the broken line path includes the following:

[0026] The distance sensor probe is used to scan the workpiece. During the scanning process, the first point p of the contour is s and the last dot p e The formed line segment is translated as the path, the distance between the path and the theoretical contour is calculated, and it is judged whether it is within the range. If all distances are within the range, this path is taken as the final path. If there is a part that exceeds the range, the midpoint of the theoretical contour is taken as the node p i , for the first point p s -Node p i and node p i - tail dot p e The above steps are repeated for the formed line segments until the measurement range of the ranging sensor probe on the moving path can cover the entire theoretical contour, and all nodes are connected in sequence to form a broken line path.

[0027] In the multi-directional slicing on-machine measurement method described above, the formation of the step path includes the following:

[0028] Determine the measurement direction and range of the ranging sensor, divide the theoretical contour into equally spaced parts according to this direction, and use the ranging sensor range as the threshold for segmentation to obtain a series of node p i , the coordinates are (x pi ,z pi ), and then by inserting a new node q between the split nodes j , the coordinates are (x pi+1 ,z pi ), and then generate a new point set. Finally, all the nodes in the new point set are connected in sequence to form a ladder path.

[0029] In the multi-directional slicing on-machine measurement method described above, the collision detection includes the following:

[0030] First, the light beam emission point of the ranging sensor probe is moved along a preset path, and the coordinates of the emission point at a certain moment are determined;

[0031] Determine the shape and position of the bounding box at that moment according to the bounding box generation principle;

[0032] The bounding box is converted into a surface in the measurement plane, and its interaction with the theoretical contour line is determined based on the DE-9IM model;

[0033] First, the preset path is segmented, and each segment is processed step by step, and the jth step of the i-th segment is recorded as S ij Each step detects the collision between the current bounding box and the workpiece contour. If a collision occurs in a certain step, the detection will be immediately exited and the next step will be carried out.

[0034] During the movement of the bounding box, the measurement range of the ranging sensor is calculated, and the contours are checked to see if they are all within the scanning range, and the contour parts that are not scanned are determined;

[0035] Comprehensively analyze the collision point and the unscanned contour parts, and adjust the path direction and the position of the ranging sensor probe based on the shape of the workpiece and the posture information of the ranging sensor probe.

[0036] In the multi-directional slice-type on-machine measurement method described above, extracting valid data includes the following:

[0037] Collect the continuous measurement data obtained by the ranging sensor and perform preliminary processing to construct one-dimensional time series data to form raw data;

[0038] The polynomial fitting method is used to perform trend modeling on the original data to form trend line data that characterizes the global change characteristics;

[0039] Perform statistical analysis on the residuals between the original data and the trend line data, calculate the residual value and its standard deviation to characterize the local deviation of the data, scale the standard deviation, and dynamically generate a dynamic ε value associated with the data characteristics;

[0040] The dynamic ε value is input into the RDP algorithm to simplify the data structure and generate the minimum point set containing the key turning points;

[0041] The dynamic ε value is input into the RDP algorithm to screen and simplify the data, determine the location of the data mutation point and extract the valid data.

[0042] In the multi-directional slice-type on-machine measurement method described above, the coordinate data fusion includes the following contents:

[0043] Establish plane MON and plane XOY. Plane MON is determined based on the position of the distance sensor light and the center point of the spindle end, while plane XOY is established based on the center point of the spindle end and the horizontal plane of the worktable.

[0044] In the plane MON, the coordinates of the sampling points are converted from the workpiece coordinate system to the machine tool coordinate system;

[0045] In the plane coordinate system, the positional relationship between the sampling point and the spindle reference point is obtained;

[0046] Since the reference planes MON and XOY are perpendicular to each other, the conversion from two-dimensional coordinates to three-dimensional coordinates can be realized. In plane XOY, the spindle and the distance sensor probe rotate around the origin of the workpiece coordinate system with a rotation angle of γ. In the machine tool coordinate system, the coordinates of the sampling point in the machine tool coordinate system are obtained based on the relationship between the plane coordinates and the spatial coordinates of the spindle reference point s.

[0047] In a second aspect, the present invention further discloses a multi-directional slicing on-machine measurement system, comprising a distance sensor and a computing device, wherein a probe of the distance sensor is mounted on a spindle of a machine tool via a mounting component, the mounting component being rotatable relative to the spindle of the machine tool, a workpiece being mounted on a workbench, and the probe of the distance sensor emitting a light beam toward the workpiece;

[0048] The computing device is configured to:

[0049] Select appropriate slicing parameters to segment the workpiece theoretical model and extract feature contours;

[0050] Path design is performed based on the extracted feature contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the distance sensor moves along the path to form a theoretical contour line.

[0051] Introducing the topological relationship model and bounding box modeling mechanism, integrating logical topological relationship judgment into path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position;

[0052] Fill the invalid points according to the collision detection results;

[0053] Extract valid data after filling invalid points;

[0054] According to the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system.

[0055] The beneficial effects of the present invention are as follows:

[0056] 1) The process of the measurement method in the present invention is relatively clear. First, the feature contour is extracted, and then the path design is performed to determine the movement path of the probe. The probe moves along the path to form a theoretical contour line, and collision detection is performed to ensure the safety of the path and avoid interference between the probe and the workpiece. According to the collision detection results, invalid points are filled and valid data is extracted. Finally, coordinate fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system, which can obtain accurate workpiece surface data and has high measurement accuracy.

[0057] 2) In the present invention, one or more paths are selected from contour paths, broken line paths, and step paths according to the surface curvature and shape characteristics of the workpiece, and a single path selection is no longer adopted. Because the path fits the shape of the workpiece, complex spatial coordinate analysis is no longer required subsequently, which not only ensures the accuracy of the measurement results, but also effectively avoids interference between the probe and the workpiece.

[0058] 3) The probe of the distance measuring sensor in the present invention is installed on the main shaft of the machine tool. The main shaft can drive the probe to move in multiple directions, and the probe of the distance measuring sensor can also rotate relative to the main shaft, thus ensuring the motion range of the distance measuring sensor probe.

[0059] 4) In the present invention, by introducing the DE-9IM topological relationship model and bounding box modeling mechanism, logical topological relationship judgment is integrated into the path feasibility analysis, which effectively avoids the interference between the probe and the workpiece surface during movement. It is suitable for on-machine measurement tasks of various types of complex structure workpieces and is conducive to fast and accurate measurement.

[0060] 5) The present invention extracts valid data by simplifying the data structure based on trend line modeling and the adaptive RDP algorithm. This allows for effective data extraction, effectively identifying the official scanning starting point of the ranging sensor probe, automatically eliminating non-measurement segment data, and extracting valid contour data with engineering significance, significantly improving the accuracy of subsequent modeling and evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0062] Figure 1 The present invention is a schematic diagram of a distance measuring sensor probe installed on a machine tool in a multi-directional slicing on-machine measurement method according to one or more embodiments.

[0063] FIG2( a ) is a schematic diagram of a feature contour extraction method when a theoretical model of a workpiece is a curved surface model in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0064] FIG2( b ) is a schematic diagram of a feature contour extraction method when the theoretical model of a workpiece is a point cloud model in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0065] FIG3( a ) is a schematic diagram of a contour path in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0066] FIG3( b ) is a schematic diagram of a broken line path in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0067] FIG3( c ) is a schematic diagram of a step path in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0068] Figure 4 It is a schematic diagram of a collision detection method in a multi-directional slicing on-machine measurement method according to one or more embodiments of the present invention.

[0069] Figure 5 The present invention is a flowchart of effective data extraction in a multi-directional slicing on-machine measurement method according to one or more embodiments.

[0070] Figure 6 The present invention is a schematic diagram of coordinate system conversion based on a ranging sensor in a multi-directional slicing on-machine measurement method according to one or more embodiments.

[0071] Figure 7 The present invention is a flowchart of on-machine measurement of a standard spherical surface in a multi-directional slicing on-machine measurement method according to one or more embodiments.

[0072] Figure 8 It is the average deviation of the profile measured in a multi-directional slice-type on-machine measurement method according to one or more embodiments of the present invention.

[0073] In the figure: the distances or sizes between parts are exaggerated to show the positions of various parts, and the schematic diagram is for reference only.

[0074] Among them: 1. Machine tool, 2. Probe, 3. Spindle, 4. Connecting rod, 5. Connecting block, 6. Mounting block, 7. Tool holder, 8. Data cable, 9. Controller, 10. Computer. DETAILED DESCRIPTION

[0075] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0076] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless otherwise clearly indicated in the present invention, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "include" and / or "comprising" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations;

[0077] As introduced in the background art, the online detection strategy of workpieces in the prior art is relatively simple and prone to interference. In order to solve the above technical problems, the present invention proposes a multi-directional slicing on-machine measurement method.

[0078] Example 1

[0079] In a typical embodiment of the present invention, referring to Figure 1 As shown, a multi-directional slice-type on-machine measurement method includes the following contents:

[0080] 1) A probe 2 of a distance measuring sensor in an optical on-machine measuring device is detachably mounted on a spindle 3 of a machine tool 1, the probe being rotatable relative to the spindle 3 of the machine tool 1, a workpiece being mounted on a workbench, and the probe of the distance measuring sensor emitting a light beam toward the workpiece;

[0081] 2) Divide the workpiece into several planes, select appropriate slicing parameters for each plane to segment the workpiece theoretical model, and extract the feature contour;

[0082] 3) Path design is performed based on the extracted characteristic contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the ranging sensor moves along the path to form a theoretical contour line.

[0083] 4) Introducing the topological relationship model and bounding box modeling mechanism, integrating the logical topological relationship judgment into the path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position;

[0084] 5) Fill the invalid points according to the collision detection results, and extract valid data after filling the invalid points;

[0085] 6) Based on the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system;

[0086] 7) Repeat the above steps to obtain the coordinates of the sampling points in all planes in the machine tool coordinate system.

[0087] The measurement method provided in this embodiment first extracts the feature contour, then performs path design to determine the movement path of the probe. The probe moves along the path to form a theoretical contour line, and collision detection is performed to ensure the safety of the path and avoid interference between the probe and the workpiece. Invalid points are filled in according to the collision detection results, and valid data is extracted. Finally, coordinate fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system, which can obtain accurate workpiece surface data and has high measurement accuracy.

[0088] In step 1), the ranging sensor is a laser ranging sensor, which can emit laser. The probe of the ranging sensor is installed on the spindle 3 of the machine tool 1 through a mounting component. The mounting component includes a connecting rod 4 that can be detachably connected to the tool holder 7 in the spindle 3 (the tool holder 7 is installed at the spindle 3). The connecting rod 4 is connected to the connecting block 5, and the connecting block 5 is rotatably connected to the mounting block 6 (specifically connected through a pin shaft). The mounting block 6 is provided with an open groove to install the probe 2 of the ranging sensor. The probe 2 can be inserted into the open groove of the mounting block 6 by snapping. The connecting block 5 drives the probe of the ranging sensor to rotate. In this way, the probe 2 can realize X, Y, and Z axis movement through the movement of the spindle 3, and can rotate in the C direction under the drive of the mounting block 6, so that the laser spot falls on any point on the surface of the workpiece, which is conducive to scanning the surface of the workpiece from different perspectives and orientations.

[0089] In step 2), if the theoretical model is a surface model, the feature contour is extracted using a cross-section, as shown in Figure 2(a). If the theoretical model is a point cloud model, the plane feature points are extracted according to the point cloud coordinates, and then the feature contour is extracted using curve interpolation. Due to the discontinuity of the point cloud, using only one plane to filter the point cloud is prone to information loss, which is not conducive to the subsequent design of the scanning path. The slicing position is determined by a reference plane, and the screening range is expanded by two other parallel boundary planes. The filtered point cloud is then projected onto the reference plane for interpolation to obtain the feature contour, as shown in Figure 2(b).

[0090] The slice position is controlled by the slice plane equation and slice interval. The base plane equation is:

[0091] P:Ax+By+Cz+D i =0

[0092] Among them, A, B, C are the three components of the plane normal vector, D i is the constant term of the plane equation.

[0093] Slice interval d p Can be achieved through D i With D i+1 Expressed as

[0094]

[0095] Use two boundary planes to increase the number of point clouds, and their plane equations satisfy

[0096]

[0097] Among them, d thick is the thickness of the slice, ΔD represents the offset of the two parallel planes relative to the original plane P, P1 is the plane translated relative to P along the positive direction of the normal vector, and P2 is the plane translated relative to P along the negative direction of the normal vector.

[0098] In step 3), referring to Figure 3, for the contour path, the path is designed according to the theoretical contour of the workpiece, so that the measuring light sweeps the workpiece generatrix as continuously and evenly as possible. This path can scan continuously close to the theoretical contour and is suitable for measurement areas with small surface curvature and regular shape, ensuring the uniformity of the scanning point distribution and the integrity of the data.

[0099] For broken line paths, the threshold value is set by the range and angle of the distance sensor, and the key turning points of the curve are extracted based on the recursive segmentation algorithm to gradually simplify the complex path. s and the last dot p e The formed line segment is translated as the path, the distance between the path and the theoretical contour is calculated, and it is judged whether it is within the range. If all distances are within the range, this path is taken as the final path. If there is a part that exceeds the range, the midpoint of the theoretical contour is taken as the node p i , for the first point p s -Node p i and node p i - tail dot p e The above steps are repeated for the formed line segments until the measurement range of the ranging sensor probe on the moving path can cover the entire theoretical contour. All nodes are connected in sequence to form a broken line path. This path design method introduces a recursive segmentation strategy and integrates the sensor range constraint conditions for path judgment. It has good adaptability and can dynamically adjust the path structure according to the complexity of the workpiece contour, avoiding measurement blind spots caused by contour mutations, and realizing the unity of path compression and effective measurement coverage.

[0100] For the step path, a fixed threshold is set based on the range of the ranging sensor, and the nearest nodes are extracted at equal intervals in a certain direction of the curve to generate a simplified path. For the same workpiece, taking a standard ball as an example, the measurement direction and range of the sensor are first determined, and the theoretical contour is divided into equally spaced parts according to this direction. Taking the Z direction as an example, the range of the ranging sensor is used as the threshold for segmentation to obtain a series of nodes p i , the coordinates are (x pi ,z pi ). Then by inserting a new node q between the split nodes j , the coordinates are (x pi+1 ,z pi ), thereby generating a new point set. Finally, all nodes in the new point set are sequentially connected to form a stepped path. This method is suitable for workpieces with layered or stepped structures, such as stepped shafts and stepped balls. It can quickly generate a path with clear nodes and a clear structure, facilitating measurement process control and data processing.

[0101] In step 4), the probe light emission point is used as the starting point for creating the bounding box, and different vertices are defined by vectors and lengths to form a closed polygon. By introducing the DE-9IM (Dimensionally Extended Nine-Intersection Model) topological relationship model and the bounding box modeling mechanism, the logical topological relationship judgment is integrated into the path feasibility analysis, which is suitable for on-machine measurement tasks of various types of complex structure workpieces. First, the light beam emission point is moved along the preset path, and the coordinates of the point at a certain moment are determined. Then, the shape and position of the bounding box at that moment are determined according to the bounding box generation principle. Finally, the bounding box is converted into a surface in the measurement plane, and its interaction relationship with the theoretical contour line is judged based on the DE-9IM model.

[0102] Due to the difficulty of large computational load in continuous simulation, segmented and discrete detection methods are used to improve algorithm flexibility and simulation authenticity. Figure 4 As shown, the preset path is first segmented, and each segment is processed step by step, and the jth step of the i-th segment is recorded as S ij Then, each step detects the collision between the current bounding box and the workpiece contour. If a collision occurs in a certain step, the detection section is immediately exited and the next detection section is performed.

[0103] Furthermore, as the bounding box moves, the ranging sensor's measurement range is calculated, the contours are checked to ensure they are within the scanning range, and any unscanned contour areas are identified. Once all simulation steps are complete, the system comprehensively analyzes collision points and unmeasured areas. Combining workpiece geometry with sensor posture information, the system manually adjusts the path direction and sensor position, achieving iterative optimization of the local path.

[0104] Regarding invalid data filling in step 5), according to the factory settings of the ranging sensor, the values ​​corresponding to invalid points are typically -99.999 or other specified marker values. Therefore, the invalid point marker values ​​are identified by masking, their sequence positions are determined, and the indices and values ​​of the k surrounding valid points are extracted. Linear interpolation or other interpolation methods are then used to fill in the invalid points, ensuring consistency and rationality of the filled data.

[0105] The extraction of valid data includes the following:

[0106] Extraction process as Figure 5 As shown in the figure, continuous measurement data from sensors is first collected and initially organized to construct one-dimensional time series data. Subsequently, a polynomial fitting method is used to model the trend of the raw data, generating trend line data that characterizes global variation. This fitted trend line not only represents the overall trend of the data but also serves as an important reference for subsequent residual calculations and outlier identification.

[0107] Then, a statistical analysis is performed based on the residuals between the original data and the trend line data. The residual values ​​and their standard deviations are calculated to characterize the degree of local deviation in the data. The standard deviation is scaled using a system-defined sensitivity factor (adjustable between 1.0 and 2.5), dynamically generating an ε (Epsilon) value associated with the data characteristics. This ε value serves as a control parameter for the Ramer–Douglas–Peucker (RDP) simplification algorithm, determining the accuracy of the final data simplification and the sensitivity of retaining feature points.

[0108] This dynamic ε value is then fed into the RDP algorithm to simplify the data structure and generate a minimal set of points containing key turning points. This process significantly reduces the data volume and eliminates irrelevant points, while retaining key feature locations (such as mutation points and inflection points), enabling efficient extraction of unstructured data.

[0109] Finally, the dynamic ε value is input into the RDP algorithm to screen the simplified data, determine the location of the data mutation point and extract the valid data. On the one hand, the segmentation problem of the start and end parts of the measurement can be converted into a mutation point detection problem of one-dimensional ordered data. Since the first point after RDP simplification is usually the starting point of the original data, in order to avoid misjudgment, the system automatically selects the sequence position corresponding to the second simplified point as the mark point where the probe actually starts to enter the measurement state. On the other hand, for the data segmentation of the middle part, it is only necessary to determine the data sequence corresponding to the starting point of the probe scan, and then determine the segmentation point of the middle part according to the scanning path and sampling parameters.

[0110] num p =num s +l path / vt

[0111] In the formula, num p is the sequence corresponding to the split point, num s is the sampling sequence corresponding to the scanning starting point, l path is the path length, v is the moving speed of the ranging sensor, and t is the sampling period of the ranging sensor.

[0112] By combining the location of the mutation point, the path segment division information and the effective sampling range, valid measurement data with engineering semantics can be extracted, realizing automatic identification of the start and end points and accurate extraction of the effective measurement segments.

[0113] The effective data extraction method proposed in this embodiment is a measurement data extraction method based on trend line modeling and adaptive RDP simplification. This method can effectively identify the official scanning starting point of the sensor probe, automatically eliminate non-measurement segment data, extract effective contour data with engineering significance, and significantly improve the accuracy of subsequent modeling and evaluation.

[0114] In step 6), coordinate data fusion is performed as follows:

[0115] While ensuring measurement accuracy, the various components of the system must be appropriately simplified. The outline of the primary analysis object is simplified to a simple, symmetrical geometric shape, allowing light analysis within a two-dimensional plane. The MON plane is defined based on the position of the sensor light and the spindle end center point, while the XOY plane is based on the spindle end center point and the horizontal plane of the worktable. Furthermore, the XYZ coordinate systems correspond to the machine tool's coordinate axes in their orientation and scale.

[0116] In plane MON, the coordinates of the sampling points are transformed from the workpiece coordinate system to the machine tool coordinate system. Figure 6 In the figure, from top to bottom, the components are the tool holder (pink), connecting rod (gray), sensor probe (blue), and workpiece (yellow). The spindle end center is point s, the contact point between the tool holder 7 and connecting rod 4 is point p, the light emission point of the ranging sensor is point l, and the sampling point is point w.

[0117] In the plane coordinate system, according to the geometric structure of the tool handle 7, connecting rod 4, connecting block 5 and mounting block 6, the sampling point (m w ,n w ) and the spindle reference point (m s ,n s ) is derived as

[0118]

[0119] Where, l t is the handle length, l r is the connecting rod length, l p is the probe length of the ranging sensor, L is the distance between the emission point (point l) and the sampling point (point w), α is the angle between the ranging sensor light and the M axis (the M axis in the plane MON), and γ is the rotation angle.

[0120] Since the reference planes MON and XOY are perpendicular to each other, the conversion from two-dimensional coordinates to three-dimensional coordinates can be realized. In plane XOY, the spindle and sensor rotate around the origin of the workpiece coordinate system with a rotation angle of γ, see Figure 6 In the XOY reference plane. In the machine tool coordinate system O u -X u Y u Z u In the figure, the relationship between the plane coordinates and the space coordinates of the spindle reference point s can be expressed as

[0121]

[0122] In summary, the coordinates of the sampling point w in the machine tool coordinate system are obtained as

[0123]

[0124] Repeat the above steps to obtain the coordinates of the sampling points in all planes in the machine tool coordinate system, thereby obtaining the full range of contour information of the workpiece.

[0125] The experiment was conducted using a standard ball with a diameter of 25 mm and an accuracy of 0.5 μm. Figure 7 As shown in the figure, 36 characteristic contour lines are extracted based on the central axis of rotation of the standard sphere. Since the characteristic contours are regular arcs, a contour path is designed as the scanning path. The ranging sensor scans the characteristic contours along the preset path. After preprocessing, the valid portion of each contour line is extracted, including 3000 sampling points. The valid data of the 36 characteristic contours is converted to coordinates to generate point cloud data.

[0126] refer to Figure 8 As shown in the figure, the experimental structure shows that the average deviation of the same part is between 130μm and 150μm, and the range of the average deviation does not exceed 30μm. That is to say, the entire measurement method can obtain accurate part surface data and has a certain stability.

[0127] Example 2

[0128] This embodiment provides a multi-directional slice-type on-machine measurement system, including a distance sensor and a computing device. The distance sensor probe is mounted on the spindle of a machine tool via a mounting component. The mounting component is rotatable relative to the spindle of the machine tool. A workpiece is mounted on a worktable, and the distance sensor probe emits a light beam toward the workpiece.

[0129] The computing device is configured to:

[0130] Select appropriate slicing parameters to segment the workpiece theoretical model and extract feature contours;

[0131] Path design is performed based on the extracted feature contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the distance sensor moves along the path to form a theoretical contour line.

[0132] Introducing the topological relationship model and bounding box modeling mechanism, integrating logical topological relationship judgment into path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position;

[0133] Fill the invalid points according to the collision detection results;

[0134] Extract valid data after filling invalid points;

[0135] According to the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system.

[0136] The probe transmits the collected data to the computing device through the data line 8 and the controller 9. The controller 9 is a PLC controller or other types of controllers. The computing device selects the computer 10, and the data processing steps are performed in the computer 10.

[0137] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A multi-directional slicing on-machine measurement method, characterized in that: Includes the following: The probe of the distance measuring sensor in the optical on-machine measuring device is detachably mounted on the main shaft of the machine tool, the probe being rotatable relative to the main shaft of the machine tool, the workpiece is mounted on the workbench, and the probe of the distance measuring sensor emits a light beam toward the workpiece; The workpiece is divided into several planes, and the appropriate slice parameters are selected for each plane to segment the workpiece theoretical model and extract the feature contour; Path design is performed based on the extracted feature contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the distance sensor moves along the path to form a theoretical contour line. Introducing the topological relationship model and bounding box modeling mechanism, integrating logical topological relationship judgment into path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position; According to the collision detection results, fill the invalid points and extract the valid data after filling the invalid points; Based on the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system; Repeat the above steps to obtain the coordinates of the sampling points in all planes in the machine tool coordinate system.

2. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The distance measuring sensor is installed on the main shaft of the machine tool through an installation component. The installation component includes a connecting rod that can be detachably connected to the main shaft. The connecting rod is connected to the connecting block. The connecting block is rotatably connected to the mounting block. The mounting block is provided with an open groove for installing the probe of the distance measuring sensor. The connecting block drives the probe of the distance measuring sensor to rotate.

3. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: If the workpiece theoretical model is a surface model, the characteristic contour is extracted through the cross-section diagram; if the workpiece theoretical model is a point cloud model, the plane characteristic points are extracted through the point cloud coordinates, and then the characteristic contour is extracted using curve interpolation; If the theoretical model of the workpiece is a point cloud model, the slicing position is determined by using a reference plane, the screening range is expanded by using two other parallel boundary planes, and the screened point cloud is projected onto the reference plane for interpolation to obtain the characteristic contour.

4. The multi-directional slice-type on-machine measurement method according to claim 3, characterized in that: For a measurement area where the surface curvature of the workpiece is less than a set value and the shape is regular, the contour path is selected; For a measurement area with a broken line on the workpiece surface, selecting the broken line path; For measurement areas with layered or stepped structures, the described step path is selected.

5. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The formation of the broken line path includes the following: The distance sensor probe is used to scan the workpiece. During the scanning process, the first point p of the contour is s and the last dot p e The formed line segment is translated as the path, the distance between the path and the theoretical contour is calculated, and it is judged whether it is within the range. If all distances are within the range, this path is taken as the final path. If there is a part that exceeds the range, the midpoint of the theoretical contour is taken as the node p i , for the first point p s -Node p i and node p i - tail dot p e The above steps are repeated for the formed line segments until the measurement range of the ranging sensor probe on the moving path can cover the entire theoretical contour, and all nodes are connected in sequence to form a broken line path.

6. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The formation of the step path includes the following: Determine the measurement direction and range of the ranging sensor, divide the theoretical contour into equally spaced parts according to this direction, and use the ranging sensor range as the threshold for segmentation to obtain a series of node p i , the coordinates are (x pi ,z pi ), and then by inserting a new node q between the split nodes j , the coordinates are (x pi+1 ,z pi ), and then generate a new point set. Finally, all the nodes in the new point set are connected in sequence to form a ladder path.

7. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The collision detection includes the following: First, the light beam emission point of the ranging sensor probe is moved along a preset path, and the coordinates of the emission point at a certain moment are determined; Determine the shape and position of the bounding box at that moment according to the bounding box generation principle; The bounding box is converted into a surface in the measurement plane, and its interaction with the theoretical contour line is determined based on the DE-9IM model; First, the preset path is segmented, and each segment is processed step by step, and the jth step of the i-th segment is recorded as S ij Each step detects the collision between the current bounding box and the workpiece contour. If a collision occurs in a certain step, the detection will be immediately exited and the next step will be carried out. During the movement of the bounding box, the measurement range of the ranging sensor is calculated, and the contours are checked to see if they are all within the scanning range, and the contour parts that are not scanned are determined; Comprehensively analyze the collision point and the unscanned contour parts, and adjust the path direction and the position of the ranging sensor probe based on the shape of the workpiece and the posture information of the ranging sensor probe.

8. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The extraction of valid data includes the following: Collect the continuous measurement data obtained by the ranging sensor and perform preliminary processing to construct one-dimensional time series data to form raw data; The polynomial fitting method is used to perform trend modeling on the original data to form trend line data that characterizes the global change characteristics; Perform statistical analysis on the residuals between the original data and the trend line data, calculate the residual value and its standard deviation to characterize the local deviation of the data, scale the standard deviation, and dynamically generate a dynamic ε value associated with the data characteristics; The dynamic ε value is input into the RDP algorithm to simplify the data structure and generate the minimum point set containing the key turning points; The dynamic ε value is input into the RDP algorithm to screen and simplify the data, determine the location of the data mutation point and extract the valid data.

9. The multi-directional slice-type on-machine measurement method according to claim 1, characterized in that: The coordinate data fusion includes the following contents: Establish plane MON and plane XOY. Plane MON is determined based on the position of the distance sensor light and the center point of the spindle end, while plane XOY is established based on the center point of the spindle end and the horizontal plane of the worktable. In the plane MON, the coordinates of the sampling points are converted from the workpiece coordinate system to the machine tool coordinate system; In the plane coordinate system, the positional relationship between the sampling point and the spindle reference point is obtained; Since the reference planes MON and XOY are perpendicular to each other, the conversion from two-dimensional coordinates to three-dimensional coordinates can be realized. In plane XOY, the spindle and the distance sensor probe rotate around the origin of the workpiece coordinate system with a rotation angle of γ. In the machine tool coordinate system, the coordinates of the sampling point in the machine tool coordinate system are obtained based on the relationship between the plane coordinates and the spatial coordinates of the spindle reference point s.

10. A multi-directional slice-type on-machine measurement system, characterized in that: The device comprises a distance measuring sensor and a computing device, wherein a probe of the distance measuring sensor is mounted on a main shaft of a machine tool via a mounting component, the mounting component being rotatable relative to the main shaft of the machine tool, a workpiece being mounted on a workbench, and the probe of the distance measuring sensor emitting a light beam toward the workpiece; The computing device is configured to: Select appropriate slicing parameters to segment the workpiece theoretical model and extract feature contours; Path design is performed based on the extracted feature contour. According to the surface curvature and shape characteristics of the workpiece, one or more of the contour path, broken line path, and step path are selected as the path. The probe of the distance sensor moves along the path to form a theoretical contour line. Introducing the topological relationship model and bounding box modeling mechanism, integrating logical topological relationship judgment into path feasibility analysis to perform collision detection and adjust the path direction and ranging sensor probe position; Fill the invalid points according to the collision detection results; Extract valid data after filling invalid points; According to the extracted valid data, coordinate data fusion is performed to obtain the coordinates of the sampling points on the workpiece surface in the machine tool coordinate system.

Images