A machine tool processing system based on 3D laser scanning analysis

Abstract

The present invention discloses a machine tool processing system based on three-dimensional laser scanning analysis, comprising a processing machine tool with a line laser profiler and a control system; the control system includes a laser profiler calibration module, a point cloud data processing module, and a processing feedback module; the laser profiler calibration module is used to calibrate the conversion relationship between the line laser profiler coordinate system and the processing machine tool coordinate system; the point cloud data processing module, based on the calibrated conversion relationship, converts the scanned contour line data of the processed workpiece into point cloud data in the processing machine tool; the processing feedback module, after each processing step is completed, aligns the scanned point cloud data with the theoretical numerical model in the processing drawing to update the processing path of the processing machine tool and correct processing errors. The system provided by the present invention can improve detection and processing efficiency, and significantly enhance the processing accuracy of CNC machine tools.

Description

Technical Field

[0001] The present invention belongs to the field of numerical control machine tool processing, and in particular relates to a machine tool processing system based on three-dimensional laser scanning analysis. Background Art

[0002] In response to the pain point of the current CNC machine tool processing process in which the processing quality is difficult to detect and evaluate, the machine tool processing system based on three-dimensional laser scanning analysis can reduce manual inspection time, improve inspection accuracy, and enhance production efficiency. Compared with two-dimensional low-precision detection methods such as industrial cameras and infrared cameras and contact measurement methods such as three-dimensional coordinate measuring machines, line laser profilers have higher surface measurement accuracy of processed parts and the advantage of non-contact non-destructive testing, which is of great significance for the inspection and evaluation of the processing quality of processed parts. In addition, the amount of surface contour point cloud data of processed parts obtained by actual line laser profiler scanning is very large. The processing of scan data and matching analysis with the theoretical model of parts can further improve the analysis level of part processing quality, which is of great significance for the improvement of subsequent processing quality. In summary, it is very necessary to propose a processing system that combines software and hardware to realize the inspection, evaluation and iterative optimization of part processing quality.

[0003] Patent document CN114708587A discloses a method for determining machining allowances for CNC machine tools based on image recognition. The method comprises: acquiring image data of the workpiece to be machined, performing feature recognition on the image data, determining workpiece shape features in the image data, and, based on the workpiece shape features, determining a reference 3D model corresponding to the workpiece shape features. The reference 3D model is used to reflect the shape of the workpiece after machining; acquiring laser point cloud data of the workpiece to be machined, and, based on the laser point cloud data, determining a real-time 3D model of the workpiece to be machined; and determining the machining allowance of the workpiece to be machined based on the real-time 3D model and the reference 3D model. This method fails to consider the difficulty in processing point cloud data when the workpiece is too large.

[0004] Patent CN 114290177A discloses a non-contact precision tool setting method for grinding aspheric optical components. The method involves mounting an optical component on a machine tool worktable while securing an oilstone adjacent to the component in a position that does not interfere with machining. A grinding wheel is then used to grind a pit on the component surface. A laser sensor is fixed to the machine tool spindle and measures the coordinates of the pit's lowest point in a first coordinate system. The offset distance between the probe's light spot and the grinding wheel's lowest point is calculated from this. The position of the aspheric optical component is then determined, and a spatial relationship between the component and the grinding wheel is established. The laser sensor in this method is used for equipment calibration, not for machining path generation. Summary of the Invention

[0005] The purpose of the present invention is to provide a processing system for solving the problems of difficult quality detection, feedback and improvement in CNC processing of parts in industrial production.

[0006] To achieve the purpose of the present invention, the present invention provides a machine tool processing system based on three-dimensional laser scanning analysis, including a processing machine tool with a line laser profiler and a control system;

[0007] The control system includes a laser profiler calibration module, a point cloud data processing module and a processing feedback module.

[0008] The laser profiler calibration module is used to calibrate the conversion relationship between the line laser profiler coordinate system and the processing machine tool coordinate system.

[0009] The point cloud data processing module converts the scanned contour line data of the workpiece into point cloud data in the processing machine tool based on the calibrated conversion relationship.

[0010] The processing feedback module aligns the point cloud data obtained by scanning with the theoretical digital model in the processing drawing after each processing step is completed, so as to update the processing path of the processing machine tool and correct the processing error.

[0011] Specifically, the point cloud data processing module also includes preprocessing of point cloud data, including invalid point removal and point cloud resampling.

[0012] Specifically, the processing feedback module can automatically align the theoretical digital model and the scanned point cloud when there are markers on the tooling.

[0013] Specifically, when there is no marker on the tooling, the processing feedback module can first manually perform a rough alignment between the theoretical digital model and the scanned point cloud, including translating and rotating the theoretical digital model along the XYZ axis according to the set distance and angle.

[0014] Specifically, the origin of the rotation axis of the rotation operation in the manual coarse registration operation is located inside the theoretical digital model corresponding to the processing drawing.

[0015] Specifically, the registration analyzes the deviations in the XYZ directions between the point cloud data obtained by scanning and the point cloud data of the theoretical numerical model, and generates a corresponding overall deviation field of the scanning points as an evaluation basis for whether reprocessing is required.

[0016] Specifically, the deviation is expressed as follows:

[0017]

[0018] Where, d x d y d zare the deviations of each point in the scanned point cloud in the XYZ directions, E r is the deviation between the corresponding points on the scanned point cloud and the theoretical numerical model, T E When the distance between the point on the scanned point cloud and the corresponding point on the digital model is greater than a certain threshold, it can be considered that the processing quality at this point is low.

[0019] Specifically, the evaluation based on the overall deviation field of the scanning point is as follows:

[0020]

[0021] Where, E r M is the deviation between the point cloud data obtained by scanning and the corresponding point of the theoretical model, E The average value of the deviation of N point clouds within the specified area size in the scanned point cloud is greater than the set threshold T M When , it can be considered that the workpiece surface within this area needs to be reprocessed.

[0022] Specifically, the processing feedback module uniformly uses the coordinates in the machine tool coordinate system for path programming during the processing process, and the processing path of the part in the machine tool coordinate system can be updated by adjusting the registration matrix between the theoretical digital model and the scanned point cloud.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] This invention applies 3D laser scanning to CNC machining error analysis and path planning. This technology visualizes the surface contour of the workpiece being machined in software, analyzes the error between the actual machined workpiece surface and the theoretical model through registration calculations, and regenerates the machining path based on this error analysis. This continuous 3D laser scanning analysis and feedback improves inspection and machining efficiency, significantly enhancing the machining accuracy of CNC machine tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a machine tool processing system based on three-dimensional laser scanning analysis provided in this embodiment;

[0026] Figure 2 A system block diagram of the machine tool processing system provided in this embodiment;

[0027] Figure 3 This is a workflow diagram of a machine tool processing system provided in this embodiment. DETAILED DESCRIPTION

[0028] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0029] like Figure 1 As shown, a machine tool processing system based on three-dimensional laser scanning analysis includes a CNC machining machine tool with a line laser profiler and a control system.

[0030] The CNC machine tool is a gantry-type machine based on a Siemens PLC. The hardware system primarily includes the laser scanning system host computer S100, the CNC machining control system host computer S102, the CNC system PLC module S104, the CNC machine tool S106, and the line laser profiler S108. The various hardware components of the system are connected via industrial Ethernet.

[0031] The line laser profiler S108 is installed at the end of the CNC machine tool S106 actuator and is powered by a DC power supply. The line laser profiler is controlled by the profiler controller pulse trigger to collect contour point data. The line laser profiler S108 and the laser scanning processing system host computer S100 are connected to transmit data via a gigabit network cable.

[0032] After the line laser profiler S108 is installed on the CNC machine tool S106, the relative positional relationship between the two needs to be calibrated. This involves collecting the contour point data of the line laser irradiating the calibration sphere at different postures, along with the machine tool's position data at these postures. This data, i.e., the contour point coordinates in the line laser profiler coordinate system and the corresponding motion of each machine tool axis at these postures, is then converted to the corresponding hand-eye relationship between the line laser profiler coordinate system and the machine tool coordinate system using a specific calibration algorithm.

[0033] Before the inspection process begins, the network connection between the CNC system S104 and the line laser profiler S108 must be successfully established on the laser scanning system host computer S100. The line laser profiler parameters, including but not limited to the sampling period (fastest trigger period), exposure time, number of contour points, measurement range, and photosensitivity, must be set to ensure the captured laser contour is clear and free of noise. The correct axis addresses for the CNC system PLC S104 must be set on the laser scanning system host computer S100 to ensure accurate acquisition of the machine tool's position data.

[0034] like Figure 2 As shown, the control system includes a laser profiler calibration module, a profile point cloud data acquisition / processing module, and a processing feedback module.

[0035] The laser profiler calibration module collects the contour point data of the line laser on the calibration sphere under different machine tool postures and the machine tool posture data based on the line laser profiler on the fixed machine tool, and obtains the conversion relationship between the machine tool coordinate system and the line laser profiler coordinate system through a certain calibration algorithm.

[0036] The point cloud data processing module first converts the acquired scanning contour line data from the line laser profiler coordinate system to the machine tool coordinate system based on the machine tool kinematic model and the hand-eye relationship obtained by calibration. Secondly, the acquired point cloud data of the workpiece surface to be measured is pre-processed, including the removal of invalid points in the point cloud and point cloud resampling.

[0037] The processing feedback module is mainly responsible for scanning point cloud visualization, scanning point cloud / theoretical digital model alignment, processing error analysis and processing path generation. This module mainly post-processes the collected workpiece contour point cloud and theoretical digital model, and feeds back the analysis results to the CNC equipment, thus forming a closed-loop system and gradually reducing the errors caused by processing.

[0038] like Figure 3 As shown, the working process of a machine tool processing system provided in this embodiment is as follows:

[0039] When the machine tool drives the line laser profiler to move and scans the surface contour of the workpiece to be measured, the coordinate data of the contour line and the corresponding machine tool posture data will be obtained in real time. The coordinate data of the point cloud will be converted from the line laser profiler coordinate system to the machine tool coordinate system through the machine tool kinematic model and the hand-eye relationship obtained by calibration, so as to facilitate subsequent processing quality analysis and feedback.

[0040] Secondly, the acquired point cloud data is preprocessed, including the removal of invalid points in the point cloud due to reasons such as over-range; the original point cloud is resampled. The contour points collected from the line laser profiler are very dense. When there are requirements for calculation time or computer performance limitations, the point cloud density can be appropriately adjusted to keep the point cloud density in a suitable numerical range, which can not only ensure the accuracy of the point cloud data but also reduce the calculation and display pressure. Of course, the resampling step is optional.

[0041] The processing feedback module can automatically align the theoretical digital model and the scanned point cloud when a marker is installed on the workpiece fixture. The specific method is to pre-fix a number of standard registration balls for later matching on the fixture, first perform a three-dimensional laser scan on the outline of the processed workpiece including the registration balls, and then fit the outline of the registration balls after scanning to obtain the standard position reference point of the workpiece installed on the fixture. Then, by matching the scanned fitting balls with the corresponding registration balls in the theoretical digital model (the registration process does not consider the actual surface morphology of the workpiece obtained by scanning, but only takes the registration balls as the standard), transform the theoretical digital model to the machine tool coordinate system where the scanned point cloud is located, thereby achieving the registration between the scanned point cloud and the theoretical digital model.

[0042] In the processing feedback module, if there is a large difference in the pose between the theoretical digital model and the scanned point cloud when there is no marker on the tooling, the subsequent alignment calculation may not achieve a good effect. You can manually perform a rough alignment between the theoretical digital model and the scanned point cloud first.

[0043] In manual coarse registration, the origin of each rotation axis should be located inside the theoretical digital model, such as the center of mass of the object. This can avoid excessive offset of the theoretical digital model position after rotation and also reduce errors.

[0044] The specific steps of manual coarse registration include:

[0045] 1) Import the theoretical model into the software;

[0046] 2) Select the appropriate rotation axis and direction according to the shape and characteristics of the actual object, and enter the rotation angle;

[0047] 3) According to the position and shape of the scanned point cloud, select the appropriate translation direction and enter the translation distance;

[0048] 4) According to the shape and characteristics of the scanned point cloud, continuously adjust the rotation angle and translation distance until the theoretical digital model basically matches the scanned point cloud.

[0049] Manual coarse registration is a simple and effective matching method that can quickly match the theoretical digital model with the scanned point cloud, providing a basis for subsequent fine registration and analysis.

[0050] After the coarse registration is complete, the two can be finely registered using point cloud matching algorithms such as the Iterative Closest Point (ICP) algorithm. The ICP algorithm can locally match the scanned point cloud with the theoretical model, continuously optimizing the matching results until the optimal registration effect is achieved. At the same time, various data processing techniques such as filtering, denoising, and smoothing can be used to pre-process the scanned point cloud to improve the accuracy and stability of the registration.

[0051] The processing feedback module can analyze the processing error after the scanning point cloud and the theoretical digital model are aligned. Specifically, the module can calculate the deviations of all points in the scanning point cloud and the points on the theoretical digital model closest to it in the X, Y, and Z directions point by point, thereby obtaining the deviation field of the entire scanning point cloud. By measuring and analyzing the size of the deviation, the situation of the processing error can be understood more accurately. In the deviation field, the color of the point cloud can be assigned according to the size of the deviation, so as to more intuitively display the deviation size of each area of ​​the scanning point cloud. In this way, the hot spots of the processing error can be quickly discovered, and further analyzed and optimized. In addition to visualizing the overall deviation field of the scanning point cloud, the processing quality of the workpiece in different parts is evaluated. The evaluation criteria are as follows:

[0052]

[0053] where d x d y d zare the deviations of each point in the scanned point cloud in the X, Y, and Z directions, E r is the total deviation between the corresponding points on the scanned point cloud and the theoretical model (that is, the Euclidean distance between the two), T E When the distance between the point on the scanned point cloud and the corresponding point on the digital model is greater than a certain threshold, it can be considered that the processing quality at this point is low.

[0054] The processing feedback module can determine whether each area of ​​the workpiece surface needs to be reprocessed based on the calculated scanning point cloud error field. If the surface of the workpiece does not meet the requirements compared with the theoretical digital model, it will be reprocessed. The criteria for reprocessing are as follows:

[0055]

[0056] Among them E r M is the total deviation between the scanned point cloud and the corresponding point of the theoretical model, E It is the average value of the deviation of N point clouds within the specified area size in the scanned point cloud. The number of point clouds N is related to the area of ​​the selected judgment area and the point cloud density. When the average deviation is greater than the set threshold T M When , it can be considered that the surface quality of the workpiece in this area is not within the standard range of qualified processing.

[0057] The machining feedback module uses the transformation matrix calculated through registration to transform all points in the theoretical model to the machine coordinate system of the scanned point cloud. This allows for quality analysis and subsequent machining of the workpiece within the actual machine coordinate system. Once the transformation relationship from the theoretical model to the scanned point cloud is established, any areas identified as non-compliant can be directly located within the machine coordinate system, eliminating the need for manual tool setting.

[0058] During machining, the machining feedback module uniformly uses coordinates in the machine tool coordinate system for path programming. By adjusting the transformation matrix between the theoretical model and the scanned point cloud, the machining path of the part in the machine tool coordinate system can be directly updated. After registration, the coordinate system of the theoretical model has been converted to the machine tool coordinate system. The machining path coordinates regenerated for the theoretical model are now also in the machine tool coordinate system and can be directly used for subsequent machining.

[0059] After each machining process is completed, the workpiece surface can be scanned in three dimensions by laser to analyze and compensate for the current machining errors.

[0060] The principles and implementation methods of the present invention are described above using specific examples. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A machine tool processing system based on three-dimensional laser scanning analysis, characterized in that: Includes processing machine tools and control systems with line laser profilers; The control system includes a laser profiler calibration module, a point cloud data processing module and a processing feedback module; The laser profiler calibration module is used to calibrate the conversion relationship between the line laser profiler coordinate system and the processing machine tool coordinate system; The point cloud data processing module converts the scanned contour line data of the workpiece into point cloud data in the processing machine tool based on the calibrated conversion relationship; The processing feedback module, after each processing step is completed, aligns the scanned point cloud data with the theoretical digital model in the processing diagram to update the processing path of the processing machine tool and correct the processing error. The alignment analyzes the deviation in the XYZ direction between the scanned point cloud data and the point cloud data of the theoretical digital model to generate the corresponding overall deviation field of the scanned points as a basis for evaluating whether reprocessing is needed; The expression of the deviation is as follows: Where, d x d y d z are the deviations of each point in the scanned point cloud in the XYZ directions, E r is the deviation between the corresponding points on the scanned point cloud and the theoretical numerical model, T E is the set threshold. When the distance between the point on the scanned point cloud and the corresponding point on the digital model is greater than a certain threshold, it can be considered that the processing quality at this point is low. The evaluation of the overall deviation field based on the scanning points is as follows: Where, E r M is the deviation between the point cloud data obtained by scanning and the corresponding point of the theoretical model, E The average value of the deviation of N point clouds within the specified area size in the scanned point cloud is greater than the set threshold T M When , it can be considered that the workpiece surface within the area needs to be reprocessed; The processing feedback module uniformly uses the coordinates in the machine tool coordinate system for path programming during the processing process. By adjusting the registration matrix between the theoretical digital model and the scanned point cloud, the processing path of the part in the machine tool coordinate system can be updated.

2. The machine tool processing system based on three-dimensional laser scanning analysis according to claim 1, characterized in that: The point cloud data processing module also includes preprocessing of point cloud data, including invalid point removal and point cloud resampling.

3. The machine tool processing system based on three-dimensional laser scanning analysis according to claim 1, characterized in that: The processing feedback module can automatically align the theoretical digital model and the scanned point cloud when there are markers on the tooling.

4. The machine tool processing system based on three-dimensional laser scanning analysis according to claim 1, characterized in that: In the absence of markers on the tooling, the processing feedback module can first manually perform a rough registration of the theoretical digital model and the scanned point cloud, including translating and rotating the theoretical digital model along the XYZ axis according to the set distance and angle.

5. The machine tool processing system based on three-dimensional laser scanning analysis according to claim 3, characterized in that: The origin of the rotation axis of the rotation operation in the manual coarse registration operation is located inside the theoretical digital model corresponding to the processing drawing.

Images