Three-coordinate measuring machine re-measurement system and control method

Abstract

The invention discloses a three-coordinate measuring machine re-measurement system and a control method, and belongs to the technical field of three-coordinate measurement, the three-coordinate measuring machine re-measurement system comprises a support, a workbench, a portal frame, a detection system, a first electric telescopic rod, a second electric telescopic rod, a third electric telescopic rod, a sliding rail, an adjusting column and a gradienter, and further comprises a main control module and a control logic module, the control logic module controls the three-coordinate measuring machine to execute a measuring task through electric connection, ceramic shells are arranged on the surfaces of the portal frame, the connecting frame, the sliding clamping block and the sliding rail, and the characteristics of high rigidity and low expansion coefficient of ceramic materials are utilized, so that deformation of a traditional metal structure caused by temperature change or long-term friction is avoided. Friction loss is reduced while the contact area is reduced, and the track precision when the portal frame translates along the X axis is ensured. The rigid support of the ceramic shell enables the equipment not to depend on a strict constant temperature environment, eliminates a thermal deformation error source from the mechanical structure level, and improves the measurement stability in a complex environment.

Description

Technical Field

[0001] The present invention relates to the technical field of three-coordinate measurement, in particular to a three-coordinate measuring machine complex measurement system and a control method. Background Art

[0002] In existing technology, the support rods of three-dimensional coordinate measuring machines are basically made of metal, so they need to operate in a constant temperature environment (usually requiring temperature fluctuations to be ≤±1°C), and the workpiece needs to be kept at a constant temperature for more than 24 hours in advance, otherwise the thermal expansion and contraction of the material will cause significant errors.

[0003] In addition, equipment tables and platforms that have been placed for a long time will also undergo subtle deformation due to external factors such as temperature and humidity, causing the equipment's workbench to tilt, which in turn leads to errors in the measurement results.

[0004] A solution is proposed based on the existing technology. Summary of the Invention

[0005] The purpose of the present invention is to remedy the deficiencies of the prior art and to provide a coordinate measuring machine complex measurement system and a control method.

[0006] To achieve the above object, the present invention provides the following technical solutions: A coordinate measuring machine re-measurement system includes a bracket, a workbench, a gantry, a detection system, a first electric telescopic rod, a second electric telescopic rod, a third electric telescopic rod, a slide rail, an adjustment column, and a level. The system also includes a control logic module, which controls the coordinate measuring machine to perform measurement tasks through electrical connections. The control logic module includes: a combination generating unit, which generates, based on the multiple sets of measurement data of the detection system, a measurement combination including different probe needle combinations and displacement paths, and a combination instruction for driving the measurement combination; a prediction calculation unit, which monitors the levelness data of the workbench in real time through the level meter, predicts the tilt trend of the workbench based on a time series model, and implements pre-compensation of the levelness of the workbench according to the tilt amount in the predicted tilt trend; The displacement control unit adopts a dual closed-loop control algorithm. The outer loop uses a grating scale to feedback the real-time position of the gantry, and the inner loop uses a current loop to compensate for mechanical resonance. According to the combined instructions of the combination generation unit, the third electric telescopic rod and the sliding rail are driven to control the translation of the gantry along the X-axis. At the same time, the second electric telescopic rod is used to drive the detection system to move along the Y-axis. A detection direction control unit adjusts the vertical height of the detection head via a first electric telescopic rod. The detection head is provided with a first detection pin and a second detection pin. The detection head is rotated by a first servo motor driving a harmonic reducer to achieve pitch and rotation of the detection head. The angle between the first detection pin and the second detection pin is 90°, so that the first detection pin or the second detection pin is aligned with the measurement point. A correction value calculation unit compares the measurement data of the first probe and the second probe at the same measurement point, calculates the position offset error, and obtains the correction parameter; A selection unit, based on the correction parameters, dynamically adjusts and evaluates the error indicators of each measurement combination using weights according to the workpiece type to screen out the measurement combination with the minimum error; The three-dimensional position determination unit iteratively eliminates outliers to fit the surface contour of the workpiece according to the result of the selection unit, and outputs a three-dimensional measurement result including X-axis, Y-axis and Z-axis coordinates.

[0007] Furthermore, the measurement combination at least includes the selection of the probe, the X-axis coordinate and the Y-axis coordinate of the gantry, and the orientation angle and the Z-axis height of the probe head; The combination generation unit optimizes the measurement path through a genetic algorithm, and the optimized measurement path includes constraint conditions, which include path length, probe switching times, and obstacle avoidance parameters.

[0008] Furthermore, when the predicted tilt exceeds a set value, a pre-adjustment parameter is generated and the threaded sleeve of the adjusting column is controlled to rotate, thereby achieving pre-compensation of the workbench horizontality through the screw pair transmission.

[0009] Furthermore, the adjusting column includes a base, a support column and a support cap. The base is connected to the support column through an angular contact ball bearing. The support cap cooperates with the threaded cap and the threaded sleeve of the trapezoidal thread pair. The adjusting column is driven by a servo motor to achieve micro-displacement adjustment. The adjustment signal is sent by the prediction calculation unit to correct the horizontality of the workbench in real time.

[0010] Furthermore, a U-shaped groove is opened on the top of the sliding rail, and the U-shaped groove forms point contact with the zirconia ceramic beads in the sliding block at the bottom of the gantry; the displacement control unit drives the sliding block to slide along the U-shaped groove through the third electric telescopic rod.

[0011] Furthermore, the detection head is spherical and movably embedded in the connecting frame, and the contact force is detected by three orthogonally arranged strain gauge sensors; the detection direction control unit controls the rotation of the detection head through a first servo motor and a magnetic grating encoder to ensure that the first detection needle or the second detection needle vertically contacts the surface to be measured.

[0012] The present application also provides a control method for a three-coordinate measuring machine, which is applicable to a three-coordinate measuring machine complex measurement system, comprising the following steps: Combination generation step: Based on the workpiece features, a genetic algorithm is used to generate a measurement combination containing different probe pins and displacement paths; Pre-adjustment steps: Use level meter data to monitor levelness in real time and set dual threshold control to trigger pre-compensation of the adjustment column; Displacement control step: driving the third electric telescopic rod and the sliding rail to achieve X-axis and Y-axis positioning of the gantry and detection system based on a double closed-loop control algorithm; Double measurement steps: using a timestamp synchronization mechanism, use the first probe and the second probe to obtain measurement data at the same location in sequence; Error correction steps: Introduce iterative weighted least squares method, dynamically adjust weights according to residuals, and iterate 3-5 times to converge error indicators; Coordinate determination step: Fit the workpiece surface contour through the least squares method and output the three-dimensional coordinates.

[0013] Furthermore, in the pre-adjustment step, the prediction calculation unit establishes a time series model based on historical data to predict the tilt trend of the workbench within a future set time, and the pre-compensation parameters are updated in real time to the servo motor of the adjustment column.

[0014] Furthermore, in the combination generation step, the optimization objectives of the genetic algorithm include path smoothness and probe life, and the constraints include workpiece geometric obstacle avoidance and the maximum load of the probe.

[0015] Compared with the existing technology, the three-coordinate measuring machine complex measurement system and control method have the following beneficial effects: By installing ceramic shells (such as zirconia ceramics) on the surfaces of the gantry, connecting frame, sliding block, and sliding rail, the present invention utilizes the high rigidity and low expansion coefficient of ceramic materials to avoid deformation of traditional metal structures caused by temperature changes or long-term friction. For example, the U-shaped groove at the top of the sliding rail forms a point-contact sliding pair with the ceramic beads in the sliding block. Driven by a third electric telescopic rod, this reduces the contact area and friction loss, ensuring the trajectory accuracy of the gantry as it translates along the X-axis. The rigid support of the ceramic shell eliminates the need for a strictly constant temperature environment (breaking the ±1°C temperature limit of traditional metal structures), eliminating thermal deformation error sources at the mechanical structure level and improving measurement stability in complex environments.

[0016] A level monitor monitors the worktable's levelness in real time. A predictive calculation unit analyzes historical data based on time series models (such as ARIMA models) to predict future tilt trends. When the predicted tilt exceeds a set threshold, the threaded sleeve of the adjustment column is controlled to rotate. Through the trapezoidal thread pair (the threaded cap and sleeve fit together) and the low-friction rotation of high-precision angular contact ball bearings, the support column is driven to micro-displace and adjust the height of the worktable's four corners. A preloaded spring assembly eliminates backlash in the thread pair, and a servo motor achieves submicron precision adjustment. Compared to traditional manual leveling, this closed-loop feedback system dynamically compensates for tabletop deformation caused by ambient temperature and humidity during measurement, keeping the levelness error within 0.001° and ensuring the absolute levelness of the measurement reference plane.

[0017] The probe head, spherical in design and embedded in the connecting frame, detects contact force in real time using three orthogonal strain gauge sensors. A first servo motor drives a harmonic reducer to achieve pitch and rotation, precisely switching between the first and second probe pins. During measurement, a timestamp synchronization mechanism is used to acquire data from the same measurement point using two probe pins in sequence. A correction value calculation unit compares the two sets of data to calculate position offset errors (such as X- and Y-axis coordinate deviations and Z-axis height difference). An iterative weighted least squares method (3-5 iterations) is used to dynamically adjust weights, eliminating outliers and generating correction parameters. This dual measurement mechanism effectively eliminates random errors such as single probe pin wear and contact angle deviation by comparing redundant data, reducing single-point measurement uncertainty to ±2μm. It is particularly suitable for high-precision surface contour measurement.

[0018] Based on workpiece features (such as geometry and material hardness), the combination generation unit utilizes a genetic algorithm to generate a multi-dimensional measurement combination, including probe selection, gantry X / Y coordinates, probe head orientation angle, and Z-axis height. The optimization process uses path smoothness (to minimize sudden stops and turns) and probe life (to ensure balanced use of different probes) as objective functions. Constraints include minimizing path length, minimizing probe switching times, and avoiding workpiece geometry obstacles. For example, for complex cavity workpieces, the algorithm automatically plans a hierarchical path that "measures the rough contour first, then the fine features," reducing idle travel time by over 30% while preventing probe-workpiece collisions.

[0019] The displacement control unit utilizes a dual closed-loop control architecture using a grating scale and current loop. The outer loop uses a grating scale (resolution ≤1nm) to provide real-time feedback on the gantry's X-axis position, forming a closed-loop position control system with a positioning accuracy of ±1μm. The inner loop uses a current loop to compensate for mechanical resonances (such as the flexible vibration of the electric telescopic rod) and suppress high-frequency noise. When driving the third electric telescopic rod in conjunction with the slide rail, a feedforward control algorithm precalculates acceleration changes to ensure smooth motion during gantry start and stop. When controlling the second electric telescopic rod's detection system along the Y-axis, a magnetic encoder (accuracy ±0.5μm) is used to precisely position the Z-axis height. This dual closed-loop system effectively eliminates mechanical drive chain backlash (such as rack and pinion backlash) and servo motor hysteresis, maintaining a three-axis coordinated positioning error of ≤±3μm, meeting nanometer-level measurement requirements.

[0020] The detection direction control unit adjusts the vertical height of the probe head via a first motorized telescopic rod, and uses a magnetic encoder to achieve ±0.1μm fine-tuning of the Z axis. A first servo motor drives the probe head through a harmonic reducer (transmission error ≤ 10 arc seconds). Three orthogonal strain gauge sensors monitor the contact force in real time (resolution 0.01N). When the contact force between the probe and the workpiece surface exceeds a threshold, the system automatically adjusts the probe head's pitch angle to ensure that either the first or second probe is perpendicular to the surface being measured. For example, when measuring inclined surfaces, the system calculates the normal direction in real time and dynamically adjusts the probe's posture to avoid cosine errors caused by angular deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the three-dimensional structure of the present invention; Figure 2 Schematic diagram of the structure of the regulating column in the present invention; Figure 3 For the present invention Figure 1 Schematic diagram of the enlarged structure at A in the middle; Figure 4 Schematic diagram of the internal structure of the detection head of the present invention; In the figure: 1- bracket, 2- workbench, 201- storage table, 202- anti-slip sticker, 203- level, 3- gantry, 301- sliding block, 302- zirconia ceramic beads, 4- detection system, 401- detection head, 402- first detection needle, 403- second detection needle, 404- connecting frame, 5- first electric telescopic rod, 6- second electric telescopic rod, 7- adjusting column, 701- base, 702- support column, 703- support cap, 704- angular contact ball bearing, 705- threaded sleeve, 706- threaded cap, 8- third electric telescopic rod, 9- sliding rail, 10- U-shaped groove. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0023] like Figure 1-4 As shown, the present invention provides a technical solution: A three-coordinate measuring machine complex measurement system includes a bracket 1, a workbench 2, a gantry 3, a detection system 4, a first electric telescopic rod 5, a second electric telescopic rod 6, a third electric telescopic rod 8, a sliding rail 9, an adjustment column 7 and a level 203, and also includes a main control module and a control logic module. The control logic module controls the three-coordinate measuring machine to perform measurement tasks through electrical connections, wherein the control logic module includes: a combination generation unit, which generates a measurement combination including different detection needle combinations and displacement paths based on multiple sets of measurement data of the detection system 4, and a combination instruction for driving the measurement combination; a prediction calculation unit, which monitors the horizontality data of the workbench 2 in real time through the level 203, predicts the tilt trend of the workbench 2 based on a time series model, and realizes pre-compensation of the horizontality of the workbench 2 according to the tilt amount in the predicted tilt trend; a displacement control unit, which adopts a double closed-loop control algorithm, wherein the outer loop feeds back the real-time position of the gantry 3 based on the grating ruler, and the inner loop compensates for mechanical resonance through the current loop; according to the combination instruction of the combination generation unit, the third electric telescopic rod 8 and the sliding rail 9 are driven. , controlling the gantry 3 to translate along the X-axis, while driving the detection system 4 to move along the Y-axis through the second electric telescopic rod 6; a detection direction control unit, adjusting the vertical height of the detection head 401 through the first electric telescopic rod 5, the detection head 401 is provided with a first detection needle 402 and a second detection needle 403, and controlling the rotation of the detection head 401 by driving the harmonic reducer through the first servo motor to realize the pitch and rotation of the detection head 401, and the angle between the first detection needle 402 and the second detection needle 403 is 90°, so that the first detection needle 402 or the second detection needle 403 is aligned with the measurement point; a correction value calculation unit, comparing the measurement data of the first detection needle 402 and the second detection needle 403 at the same measurement point, calculating the position offset error, and obtaining a correction parameter; a selection unit, based on the correction parameter, using a weight to dynamically adjust and evaluate the error index of each measurement combination according to the workpiece type to screen out the measurement combination with the smallest error; a three-dimensional position determination unit, based on the result of the selection unit, iteratively eliminates outliers to fit the surface contour of the workpiece, and outputs a three-dimensional measurement result including X-axis, Y-axis and Z-axis coordinates.

[0024] In some embodiments, the time series model can be established as follows: Step 1) Data acquisition and preprocessing: a level 203 is installed at each of the four corners of the workbench 2, and a MEMS tilt sensor (accuracy ±0.0005°) is used to collect the tilt angle data of the X / Y axis in real time at a frequency of 20 Hz, which are recorded as (θ x1 ,θ x2 ,θ y1 ,θ y2 ), forming a four-dimensional time series (t,θ x1 ,θ x2 ,θ y1 ,θ y2 ), where θ x1 / θ x2 is the X-axis tilt angle on the left and right sides of the workbench, θ y1 / θ y2 It is the Y-axis tilt angle on the front and rear sides. Among them, the MEMS tilt sensor is a sensor manufactured using Micro-Electro-Mechanical Systems (Micro-Electro-Mechanical Systems) technology, which is mainly used to sense the tilt angle of an object.

[0025] Data preprocessing: High-frequency noise is removed through sliding average filtering (window size 5s), and the Laida criterion (3σ principle) is used to identify and repair jump outliers to ensure that the input model is continuous and smooth tilt trend data.

[0026] Among them, the Laida criterion includes: for each level θ x / θ y Data (such as θ x1 ), calculate the mean (μ) and standard deviation (σ). Normal data should satisfy μ-3σ≤θ x1 ≤μ+3σ. Points outside this range are considered outliers. If there are ≤3 consecutive outliers, they are corrected using linear interpolation of the preceding and following normal data (e.g., if the nth point is an outlier, the average of points (n-1) and (n+1) is used instead). If there are >3 consecutive outliers, the segment is marked as invalid and the data for that time period is excluded from subsequent modeling.

[0027] By using the ARIMA model (Autoregressive Integrated Moving Average Model) as the basis of the time series model and performing iterative training with the above data, the time series model implemented in the levelness pre-compensation in this application can be obtained.

[0028] In some embodiments, an iterative weighted least squares error correction implementation can be established as follows: First, dual measurement data acquisition: When the probe reaches the target measurement point, the first probe 402 and the second probe 403 are synchronously triggered by a hardware timestamp to perform contact measurements in sequence, with an interval of ≤50ms to prevent small workpiece displacements from affecting data consistency. The measurement data obtained by the first probe 402 and the second probe 403 are acquired separately. A basic iterative weighted algorithm, such as RANSAC-LS (random sample consensus, LS stands for minimum square error), is used to fuse the statistics (average or median) of the multiple measurement data sets. Outliers are iteratively removed (with a threshold of 3σ), and the workpiece surface profile is fitted. The resulting three-dimensional measurement results, including X, Y, and Z coordinates, are output with a repeatability accuracy of ±0.5μm.

[0029] In some embodiments, the attitude adjustment actuator includes vertical adjustment: the first electric telescopic rod 5 has a stroke of 20 mm and an accuracy of ±0.1 μm, which is driven by a ball screw and is used to adjust the Z-axis height of the detection head so that the tip of the detection needle is aligned with the measurement point; rotation adjustment: the first servo motor (power 200 W, speed 1000 rpm) drives the detection head to rotate through a harmonic reducer (reduction ratio 100:1), with an angular resolution of 0.0036°, and cooperates with a magnetic grating encoder to achieve closed-loop control.

[0030] In some embodiments, dynamic compensation scenarios include: inclined plane measurement. When the inclination angle of the measured surface is greater than 15°, the system automatically calculates the normal direction and adjusts the probe posture to ensure that more than 90% of the contact force is along the normal direction to avoid slippage or wear of the probe. Surface scanning. During continuous scanning, the normal direction is updated every 50ms, driving the probe head to follow the changes in the curvature of the surface in real time, realizing dynamic alignment of the "probe posture-surface normal". Compared with fixed posture measurement, the surface contour error can be reduced from ±10μm to ±3μm.

[0031] In some embodiments, the measurement combination includes at least the selection of the probe needle, the X-axis coordinate and Y-axis coordinate of the gantry 3, the orientation angle and Z-axis height of the probe head 401; the combination generation unit optimizes the measurement path through a genetic algorithm, and the optimized measurement path includes constraints, and the constraints include path length, number of probe needle switching times and obstacle avoidance parameters.

[0032] In some embodiments, when the predicted tilt exceeds a set value, a pre-adjustment parameter is generated and the threaded sleeve 705 of the adjusting column 7 is controlled to rotate, and the horizontality pre-compensation of the workbench 2 is achieved through the screw pair transmission.

[0033] In some embodiments, the adjusting column 7 includes a base 701, a support column 702 and a support cap 703. The base 701 is connected to the support column 702 through an angular contact ball bearing 704. The support cap 703 cooperates with the threaded sleeve 705 through the threaded cap 706 of the trapezoidal thread pair. The clearance of the trapezoidal thread pair is eliminated by the pre-tightening spring assembly. The adjusting column 7 is driven by a servo motor to achieve micro-displacement adjustment. The adjustment signal is sent by the prediction calculation unit to correct the horizontality of the workbench 2 in real time.

[0034] In some embodiments, a U-shaped groove 10 is opened on the top of the sliding rail 9, forming point contact with the zirconia ceramic bead 302 in the sliding block 301 at the bottom of the gantry 3; the displacement control unit drives the sliding block 301 to slide along the U-shaped groove 10 through the third electric telescopic rod 8.

[0035] In some embodiments, the detection head 401 is spherical and movably embedded in the connecting frame 404, and detects the contact force through three orthogonally arranged strain gauge sensors; the detection direction control unit controls the rotation of the detection head 401 through a first servo motor and a magnetic grating encoder to ensure that the first detection needle 402 or the second detection needle 403 vertically contacts the surface to be measured.

[0036] The present invention also provides a control method for a three-coordinate measuring machine, comprising the following steps: a combination generation step: based on the workpiece features, a genetic algorithm is used to generate a measurement combination including different detection pins and displacement paths; a pre-adjustment step: real-time monitoring of the levelness through the level meter 203 data, and setting a double threshold control to trigger the adjustment column 7 for pre-compensation; a displacement control step: driving the third electric telescopic rod 8 and the sliding rail 9, and realizing the X-axis and Y-axis positioning of the gantry 3 and the detection system 4 based on a double closed-loop control algorithm; a dual measurement step: using a timestamp synchronization mechanism, using the first detection pin 402 and the second detection pin 403 in sequence to obtain measurement data at the same position; an error correction step: introducing an iterative weighted least squares method, dynamically adjusting the weight according to the residual, and iterating 3-5 times to converge the error index; a coordinate determination step: fitting the surface contour of the workpiece by the least squares method, and outputting three-dimensional coordinates.

[0037] In some embodiments, in the pre-adjustment step, the prediction calculation unit establishes a time series model based on historical data to predict the tilt trend of the workbench within a future set time, and the pre-compensation parameters are updated in real time to the servo motor of the adjustment column 7.

[0038] In some embodiments, in the combination generation step, the optimization objectives of the genetic algorithm include path smoothness and probe life, and the constraints include workpiece geometric obstacle avoidance and probe maximum load.

[0039] The present invention uses the first detection needle 402, the second detection needle 403, and the spherical detection head 401 to enable two repeated measurements using different detection needles during the measurement process, thereby ensuring the accuracy of the measurement results.

[0040] The following describes the specific steps of blade measurement in detail, using blade measurement as an example. The first step is to initialize the equipment and perform level pre-compensation. The blade is placed on the storage platform 201 in the center of the workbench 2 and fixed with anti-slip stickers 202 to prevent the workpiece from moving during measurement. The level 203 (MEMS tilt sensor, accuracy ±0.0005°) at the four corners of the workbench 2 collects X / Y axis tilt angle data (θ x1 ,θ x2 ,θ y1 ,θ y2 ), after outliers are corrected using the Laida criterion and sliding average filtering, the data is input into the prediction calculation unit. This unit analyzes historical data based on the ARIMA model and predicts the worktable's tilt trend within the next five minutes. If the predicted tilt exceeds a set threshold (e.g., 0.005°), pre-adjustment parameters are immediately generated, driving the servo motor of adjustment column 7 to rotate threaded sleeve 705. Adjustment column 7 achieves micro-displacement adjustment through a trapezoidal thread pair (threaded cap 706 and threaded sleeve 705) and high-precision angular contact ball bearing 704. A preload spring assembly eliminates thread play, ensuring that the worktable's horizontality error is controlled within 0.001°.

[0041] The second step is to analyze the workpiece features. The 3D CAD model of the blade is imported through the main control module to identify the geometric features of the blade (such as the blade surface, leading edge / trailing edge radius, blade root tenon size, etc.) and determine the key measurement points (such as discrete points on the blade surface, edge plate transition fillet, etc.).

[0042] The third step is to generate a multivariate measurement combination. The combination generation unit uses a genetic algorithm based on blade characteristics to generate a measurement combination with the following parameters: Probe selection: To meet the required blade surface accuracy, the first probe 402 (tip radius 0.5mm) is preferred for coarse measurement, while the second probe 403 (tip radius 0.2mm) is preferred for fine measurement. Gantry coordinates: The X-axis covers the blade span (0-200mm), the Y-axis is positioned at the blade chord center (e.g., Y = 50mm), and the Z-axis height is dynamically adjusted based on blade thickness (initial value Z = 30mm). Probe posture: The initial pitch angle of probe 401 is set to 0°, and the rotation angle is pre-adjusted based on the blade twist angle (e.g., if the blade tip twist angle is 20°, it is pre-rotated by -20°). Path optimization and constraints: Optimization objectives: path smoothness (to minimize sudden stops and turns) and probe life (to ensure balanced use of both probes). Constraints: path length ≤ 300mm, probe switching ≤ 5, and obstacle avoidance parameters (maintaining a 5mm safety distance from the blade edge). Output result: Generate a hierarchical measurement path of "roughly measure the contour - accurately measure the features - and then measure the blind spots", reducing the idle travel time by about 30%.

[0043] The fourth step involves displacement control and probe posture adjustment, including positioning the gantry's X / Y axes. The displacement control unit utilizes a dual closed-loop control algorithm: a grating scale (resolution ≤ 1 nm) provides real-time feedback on the X-axis position of the gantry 3, driving the third electric telescopic rod 8 and the slide rail 9 to control the gantry's X-axis translation to the target coordinate (with an accuracy of ±1 μm). The inner loop uses a current loop to compensate for the mechanical resonance of the third electric telescopic rod 8, suppressing high-frequency vibrations and ensuring smooth motion. Simultaneously, the second electric telescopic rod 6 drives the detection system 4 along the Y axis, using a magnetic encoder (with an accuracy of ±0.5 μm) to achieve Y-axis positioning.

[0044] The first motorized telescopic rod 5 (20mm travel, ±0.1μm accuracy) adjusts the vertical height of the probe head 401, aligning the tip of the first probe needle 402 with the measurement point on the blade surface. A first servo motor drives the probe head through a harmonic reducer (100:1 reduction ratio). This servo motor, coupled with a magnetic encoder (0.0036° angular resolution) and three orthogonal strain gauge sensors (0.01N contact force resolution), adjusts the probe head's pitch and rotation angles in real time to ensure perpendicular contact with the probe needle. For example, when measuring a twisted surface in the middle of a blade, the system automatically calculates the normal direction and adjusts the probe head's pitch angle to 15° to avoid cosine error.

[0045] Step 5: Dual measurement and error correction. First, dual measurement is performed with time stamp synchronization. When the probe reaches the target point, the hardware timestamp synchronization mechanism triggers the following operations: First probe 402 performs a coarse measurement: It contacts the blade surface at a speed of 2 mm / s to acquire initial coordinate data (X1, Y1, Z1), with a contact force controlled between 0.1 and 0.3 N. Second probe 403 performs a fine measurement: After a 50 ms interval, the second probe is switched to the second probe and repeatedly measures the same point at a speed of 0.5 mm / s to acquire fine measurement data (X2, Y2, Z2).

[0046] Secondly, iterative weighted least squares error correction is performed. The correction value calculation unit compares the two sets of data and calculates the position offset error (e.g., ΔX = X² - X¹, ΔZ = Z² - Z¹). Iterative weighted least squares (iteration 3) is introduced to dynamically adjust the weights and remove outliers (threshold 3σ). For example, if the Z-axis deviation between two measurements of a point exceeds ±5μm, it is identified as an outlier and corrected using neighboring point interpolation to generate the final correction parameters.

[0047] Step 6: 3D coordinate fitting and result output: The 3D position determination unit iteratively removes outliers (such as burr interference points on the blade surface) based on the minimum error measurement combination screened by the selection unit, and uses the least squares method to fit the blade surface contour to generate 3D point cloud data containing X, Y, and Z axis coordinates.

[0048] Result Output and Analysis: Compare point cloud data with the CAD model, calculate geometric tolerances (such as blade profile and leading edge curvature radius error), and generate a measurement report. Repeatability Verification: Repeat the measurement three times on the same blade, with a single-point coordinate deviation of ≤±2μm, meeting high-precision measurement requirements.

Claims

1. A three-coordinate measuring machine complex measurement system, comprising a bracket (1), a workbench (2), a gantry (3), a detection system (4), a first electric telescopic rod (5), a second electric telescopic rod (6), a third electric telescopic rod (8), a sliding rail (9), an adjustment column (7) and a level (203), characterized in that: The system also includes a control logic module, which controls the coordinate measuring machine to perform measurement tasks through electrical connections, wherein the control logic module includes: A combination generating unit, which generates a measurement combination including different probe needle combinations and displacement paths, and a combination instruction for driving the measurement combination based on multiple sets of measurement data of the detection system (4); A prediction calculation unit monitors the levelness data of the workbench (2) in real time through the level meter (203), predicts the tilt trend of the workbench (2) based on a time series model, and implements pre-compensation of the levelness of the workbench (2) according to the tilt amount in the predicted tilt trend; The displacement control unit adopts a double closed-loop control algorithm, wherein the outer loop is based on the grating ruler to feedback the real-time position of the gantry (3), and the inner loop compensates for mechanical resonance through the current loop; according to the combination instruction of the combination generation unit, the third electric telescopic rod (8) and the sliding rail (9) are driven to control the gantry (3) to move along the X-axis, and at the same time, the detection system (4) is driven to move along the Y-axis through the second electric telescopic rod (6); A detection direction control unit adjusts the vertical height of a detection head (401) through a first electric telescopic rod (5), wherein a first detection needle (402) and a second detection needle (403) are provided on the detection head (401), and controls the rotation of the detection head (401) by driving a harmonic reducer through a first servo motor, thereby achieving pitch and rotation of the detection head (401), wherein the angle between the first detection needle (402) and the second detection needle (403) is 90°, so that the first detection needle (402) or the second detection needle (403) is aligned with a measurement point; A correction value calculation unit compares the measurement data of the first detection needle (402) and the second detection needle (403) at the same measurement point, calculates the position offset error, and obtains a correction parameter; A selection unit, based on the correction parameters, dynamically adjusts and evaluates the error indicators of each measurement combination using weights according to the workpiece type to screen out the measurement combination with the minimum error; The three-dimensional position determination unit iteratively eliminates outliers to fit the surface contour of the workpiece according to the result of the selection unit, and outputs a three-dimensional measurement result including X-axis, Y-axis and Z-axis coordinates.

2. The coordinate measuring machine complex measurement system according to claim 1, characterized in that: The measurement combination at least includes the selection of the probe, the X-axis coordinate and the Y-axis coordinate of the gantry (3), and the orientation angle and the Z-axis height of the probe head (401); The combination generation unit optimizes the measurement path through a genetic algorithm, and the optimized measurement path includes constraint conditions, which include path length, probe switching times, and obstacle avoidance parameters.

3. The coordinate measuring machine complex measurement system according to claim 1, characterized in that: When the predicted tilt exceeds a set value, a pre-adjustment parameter is generated and the threaded sleeve (705) of the adjustment column (7) is controlled to rotate, thereby achieving pre-compensation of the horizontality of the workbench (2) through the screw pair transmission.

4. The coordinate measuring machine complex measurement system according to claim 1, characterized in that: The adjusting column (7) comprises a base (701), a supporting column (702) and a supporting cap (703); the base (701) is connected to the supporting column (702) via an angular contact ball bearing (704); the supporting cap (703) cooperates with a threaded sleeve (705) via a threaded cap (706) of a trapezoidal thread pair; the adjusting column (7) is driven by a servo motor to achieve micro-displacement adjustment; an adjustment signal is sent by the prediction calculation unit to correct the horizontality of the workbench (2) in real time.

5. The coordinate measuring machine complex measurement system according to claim 1, characterized in that: A U-shaped groove (10) is provided on the top of the sliding rail (9), and the U-shaped groove (10) forms point contact with a zirconia ceramic bead (302) in a sliding block (301) at the bottom of the gantry (3); the displacement control unit drives the sliding block (301) to slide along the U-shaped groove (10) via a third electric telescopic rod (8).

6. The coordinate measuring machine complex measurement system according to claim 1, characterized in that: The detection head (401) is spherical and movably embedded in the connecting frame (404), and detects the contact force through three orthogonally arranged strain gauge sensors; the detection direction control unit controls the rotation of the detection head (401) through a first servo motor in conjunction with a magnetic grating encoder to ensure that the first detection needle (402) or the second detection needle (403) vertically contacts the surface to be measured.

7. A control method for a three-coordinate measuring machine, applicable to the three-coordinate measuring machine complex measurement system according to any one of claims 1 to 6, characterized in that: The following steps are involved: Combination generation step: Based on the workpiece features, a genetic algorithm is used to generate a measurement combination containing different probe pins and displacement paths; Pre-adjustment step: real-time monitoring of the levelness through the level meter (203) data, setting a double threshold control to trigger the adjustment column (7) for pre-compensation; Displacement control step: driving the third electric telescopic rod (8) and the sliding rail (9) to achieve X-axis and Y-axis positioning of the gantry (3) and the detection system (4) based on a double closed-loop control algorithm; Double measurement step: using a timestamp synchronization mechanism, sequentially using a first detection needle (402) and a second detection needle (403) to obtain measurement data of the same position; Error correction steps: Introduce iterative weighted least squares method, dynamically adjust weights according to residuals, and iterate 3-5 times to converge error indicators; Coordinate determination step: Fit the workpiece surface contour through the least squares method and output the three-dimensional coordinates.

8. The control method of a three-dimensional coordinate measuring machine according to claim 7, characterized in that: In the pre-adjustment step, the prediction calculation unit establishes a time series model based on historical data to predict the tilt trend of the workbench (2) within a future set time, and the pre-compensation parameters are updated in real time to the servo motor of the adjustment column (7).

9. The control method of a three-dimensional coordinate measuring machine according to claim 7, characterized in that: In the combination generation step, the optimization objectives of the genetic algorithm include path smoothness and probe life, and the constraints include workpiece geometric obstacle avoidance and the maximum load of the probe.

Images