Intelligent manufacturing method for rotary body workpiece

Abstract

The invention discloses an intelligent manufacturing method for a rotary body workpiece, belongs to the technical field of rotary body workpiece machining, is used for manufacturing the rotary body workpiece, and comprises the steps of automatic process design, entity topology analysis, process optimization closed loop and numerical control program self-adaptive correction. The automatic process design comprises geometric feature extraction, abnormal topology processing and parametric modeling; the entity topology analysis comprises entity topology decomposition, curved surface type discrimination, geometric feature parameterization and exception handling; the process optimization closed loop comprises multi-mode on-machine detection, real-time data processing, intelligent compensation modeling and closed loop control; the adaptive correction of the numerical control program comprises the steps of data analysis and preprocessing, intelligent path matching, path planning optimization and closed-loop verification. Manual intervention is greatly reduced in the aspects of mechanical design and machining, work errors are reduced, and the quality and efficiency are greatly improved; repeated design, programming and detection workloads are greatly reduced, and the design and development efficiency is greatly improved.

Description

Technical Field

[0001] The invention discloses an intelligent manufacturing method for a rotating workpiece, and belongs to the technical field of rotating workpiece processing. Background Art

[0002] With the rapid development of the global manufacturing industry, intelligent manufacturing has become an important direction for the transformation and upgrading of the manufacturing industry. As the basic parts in mechanical manufacturing, the processing quality of rotating workpieces is directly related to the performance, reliability and life of the products. With the continuous improvement of modern industry's requirements for product quality and efficiency, the traditional processing technology of rotating workpieces can no longer meet the development needs of intelligent manufacturing. Therefore, it is of great significance to study the processing technology of rotating workpieces based on intelligent manufacturing. The challenges faced by the processing technology of rotating workpieces are as follows: high processing accuracy requirements. With the improvement of the degree of industrial automation, the processing accuracy requirements for rotating workpieces are getting higher and higher; high processing efficiency requirements. Modern industry has an increasing demand for the processing efficiency of rotating workpieces to meet the production rhythm; processing cost control. Under the premise of ensuring processing quality, how to reduce production costs has become the focus of enterprises; low intelligence and automation levels. Traditional processing methods rely on manual operation, and the level of intelligence and automation is low. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent manufacturing method for a rotating workpiece, so as to solve the problem in the prior art that the intelligent manufacturing method for a rotating workpiece relies on manual labor, resulting in low efficiency.

[0004] Intelligent manufacturing methods for rotating workpieces, including automated process design, entity topology analysis, closed-loop process optimization, and adaptive correction of CNC programs;

[0005] The automated process design includes geometric feature extraction, abnormal topology processing and parametric modeling;

[0006] The geometric feature extraction includes size topology association analysis and contour boundary reconstruction;

[0007] The dimension topology association analysis includes obtaining the graphics element of the rotating workpiece, calling the 2D drawing software API to traverse the graphics element database, establishing the dimension-graphic element mapping relationship, and determining the main contour through extreme value screening;

[0008] The reconstruction of contour boundaries includes sorting the polar coordinates of the main contour rows to determine the starting edge, and using depth-first search DFS to construct a closed main contour;

[0009] The abnormal topology processing includes repeated curve disambiguation and geometric discontinuity repair;

[0010] The parametric modeling includes importing the repaired main contour into the 3D modeling software, ensuring the correct resolution of the curve through geometric verification, verifying the curve in combination with the geometric tolerance detection method, and issuing a warning and automatically repairing it in time if any deviation is found;

[0011] The entity topology analysis includes entity topology decomposition, surface type identification, geometric feature parameterization, and exception processing;

[0012] The entity topology decomposition includes generating feature sensitive sampling points using an adaptive chord difference method for the reconstructed three-dimensional features and establishing a curvature weight sampling model;

[0013] The surface type identification includes plane feature detection, rotation surface analysis, and torus feature analysis;

[0014] The process optimization closed loop includes multi-modal on-machine detection, real-time data processing, intelligent compensation modeling, and closed-loop control;

[0015] The adaptive correction of the numerical control program includes data analysis and preprocessing, intelligent path matching, path planning optimization, and closed-loop verification.

[0016] Duplicate curve disambiguation includes extracting geometric features, building an adjacency matrix to detect topological conflicts, performing redundant path pruning, and defining a curve set. ,curve From continuous points Indicates that calculating any two curves and The starting and ending distances, when the distance is less than the threshold ,determination and Overlap and delete one of them, the starting and ending distances are Euclidean distances :

[0017] ;

[0018] ;

[0019] In the formula, yes point The starting point, yes point The starting point, yes point The end point, yes point The end point, yes and The coordinates of the two starting points, yes and The coordinates of the two end points of

[0020] Geometric discontinuity repair includes quickly locating breakpoints based on the R-tree spatial index for the curves after repeated curve disambiguation, processing the breakpoints using the minimum distance bridging algorithm, and defining the open point set. Existence Points, , calculate the Euclidean distance between any two opening points, if it is less than the threshold, restore the curve by interpolation :

[0021] ;

[0022] In the formula, is the Bernstein polynomial, For the Points, It represents the total number of interpolation points. The least square method is used to fit and optimize the interpolation process to ensure that the repaired curve is smooth and continuous.

[0023] The parametric modeling includes: after the curve is imported, according to Endpoints Sort and generate rotation bodies in order. The sorting algorithm is based on heuristic search. First, The coordinates are sorted and then The coordinates are sorted to ensure smooth connection of all curves; a three-dimensional rotation body is obtained through rotation transformation The mathematical formula is:

[0024] ;

[0025] In the formula, It is around The rotation matrix for axis rotations, is the rotation angle, and the generated solid of revolution is visualized by computer-aided design software;

[0026] After the 3D modeling is completed, the point cloud data is used to analyze the model in detail, and the plane fitting technology based on the RANSAC algorithm is used to automatically identify the feature surface from the point cloud:

[0027] ;

[0028] In the formula, is the error function, is the normal vector, is the offset, Indicates Points, It represents the total number of plane points. It extracts characteristic surfaces and boundaries through plane fitting, providing a basis for subsequent machining path generation.

[0029] The plane feature detection includes performing principal component analysis (PCA) on the reconstructed three-dimensional features to verify the consistency of normal vectors and establish a flatness error model;

[0030] The rotational surface analysis includes cylindrical surface recognition, conical surface verification and spherical surface determination for the reconstructed three-dimensional features;

[0031] The torus feature analysis includes constructing a hyperbolic feature space for the reconstructed three-dimensional features and detecting the sweeping trajectory using Hough transform.

[0032] The geometric feature parameterization includes the reconstruction of the reference axis system and the inverse calculation of the feature parameters;

[0033] The reconstruction of the reference axis system includes determining the rotation axis and establishing the axis system tolerance band by least square fitting;

[0034] The feature parameter inversion includes respectively obtaining the radius, half angle, main radius and center of the cylindrical surface, conical surface, torus and spherical surface of the three-dimensional feature;

[0035] The abnormality processing includes mixed surface discrimination and transition feature processing;

[0036] The mixed surface discrimination includes constructing a Mahalanobis distance classifier and setting a type confidence threshold to discriminate the mixed surface;

[0037] The transition feature processing includes using curvature derivatives to detect chamfers and fillets, using PointNet++ to classify and identify features of point cloud data, and combining deep learning to automatically identify key features in complex set shapes. The input of the deep learning network is Point cloud dataset , the output is the processing feature category, each layer of the network learns local features through the multi-layer perceptron MLP, and extracts global features through the global pooling operation:

[0038] ;

[0039] In the formula, is the feature of the i-th point, represent No. Points, is the weight, is a constant term, through global pooling Extract global features and finally output machining feature categories to help the system automatically identify key machining features on the workpiece;

[0040] Based on the identified key processing features, the tool path interpolation algorithm is used to automatically generate the finishing milling path to ensure that the tool moves along the optimal path to avoid processing errors. For different processing features, the system automatically adjusts the cutting parameters and optimizes the processing efficiency according to the material and geometric features of the workpiece.

[0041] After completing the exception handling, the features of the rotating workpiece are identified.

[0042] The multimodal on-machine detection includes constructing a high-precision tactile sensing unit and standardizing the machine tool coordinate system;

[0043] The construction of the high-precision tactile sensing unit includes constructing a probe dynamic compensation model according to the characteristics of the rotating workpiece, and using a quaternion normal vector solution algorithm to solve the probe dynamic compensation model;

[0044] The normalization of the machine tool coordinate system includes establishing the machine tool coordinate transformation matrix, performing path planning based on the curvature adaptive spiral scanning path, finding the optimal closed path using a global optimization algorithm, iteratively adjusting the link order based on the endpoint set, and minimizing the target parameters. :

[0045] ;

[0046] In the formula, For the closed function of the curve, the local optimal solution problem is avoided through global optimization;

[0047] The real-time data processing includes intelligent analysis of measurement data and analysis of machining allowance;

[0048] The intelligent analysis of the measurement data includes constructing an abnormal point filtering model for the features of the rotating workpiece, adding curvature constraints using the ICP algorithm, and performing point cloud matching;

[0049] The machining allowance analysis includes calculating the normal allowance and the tangential deviation.

[0050] The intelligent compensation modeling includes an inverse surface reconstruction algorithm and dynamic adjustment of process parameters;

[0051] The inverse surface reconstruction algorithm includes using NURBS surfaces for adaptive interpolation and gradient descent optimization;

[0052] The dynamic adjustment of process parameters includes constructing a cutting parameter optimization model and calculating tool compensation amounts.

[0053] The closed-loop control includes on-machine detection and data verification for the features of the rotating workpiece. If the verification is qualified, point cloud alignment, residual analysis, compensation modeling, tool path regeneration, secondary processing, and return to on-machine detection are performed. If the verification fails, an alarm is issued and the machine is shut down.

[0054] The data analysis and preprocessing include building an efficient coordinate extraction engine and building a spatial index;

[0055] The construction of the efficient coordinate extraction engine includes using a buffer stream to accelerate the file reading of the features of the rotating workpiece through closed-loop control, and constructing an abnormal data filtering model to perform data filtering;

[0056] The construction of the spatial index includes performing a fast search of point clouds based on Octree; The intelligent path matching includes calling a dynamic distance analysis model and optimizing a projection algorithm; The calling of the dynamic distance analysis model includes calling a hybrid distance calculation strategy and setting an adaptive threshold to perform dynamic distance analysis;

[0057] The optimized projection algorithm includes using LM algorithm to accelerate the surface projection and introducing normal constraint projection correction;

[0058] The path planning optimization includes three types of corrections, namely linear movement, arc path and free-form surface, which are adjusted by minimum energy interpolation method, curvature matching correction and parameterized remapping, respectively, and acceleration continuity, continuity maintenance and bow height error not exceeding 0.001 mm are respectively applied as constraints;

[0059] In path planning optimization, program segments are managed based on a bidirectional linked list, and an incremental program update strategy is called.

[0060] The closed-loop verification includes performing point cloud matching analysis on the features of the acquired rotating workpiece and the manufactured rotating workpiece. If the deviation does not exceed the threshold, it is directly output; if the deviation exceeds the threshold, surface projection optimization, path parameter recalculation, tool compensation update, verification path generation, virtual processing verification, and returning to the point cloud matching analysis.

[0061] Compared with the existing technology, the present invention has the following beneficial effects: the present invention greatly reduces manual intervention in mechanical design and processing, reduces work errors, and has greatly improved quality and efficiency; greatly reduces the workload of repeated design, programming, and testing; realizes the automation of products from scratch, greatly improving the efficiency of design and development; through the inheritance of two-dimensional curve repair, three-dimensional conversion and deep learning feature recognition, it solves the efficiency and accuracy problems of traditional methods in complex geometric shape processing. In practical applications, the repair and modeling time is shortened to 70% of the traditional method, effectively controlling the processing error, and is suitable for high-precision processing needs in the fields of aerospace, automotive molds, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a closed-loop control flow chart;

[0063] Figure 2 It is a closed-loop verification flow chart;

[0064] Figure 3 A schematic diagram of a rotating workpiece to be processed;

[0065] Figure 4 For Figure 3 A schematic diagram of feature recognition on the first side;

[0066] Figure 5 For Figure 3 A schematic diagram of feature recognition on the second side;

[0067] Figure 6 This is a schematic diagram of spline interpolation fitting;

[0068] Reference numerals include, 1-big end, 2-small end. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0070] Intelligent manufacturing methods for rotating workpieces, including automated process design, entity topology analysis, closed-loop process optimization, and adaptive correction of CNC programs;

[0071] The automated process design includes geometric feature extraction, abnormal topology processing and parametric modeling;

[0072] The geometric feature extraction includes size topology association analysis and contour boundary reconstruction;

[0073] The dimension topology association analysis includes obtaining the graphics element of the rotating workpiece, calling the 2D drawing software API to traverse the graphics element database, establishing the dimension-graphic element mapping relationship, and determining the main contour through extreme value screening;

[0074] The reconstruction of contour boundaries includes sorting the polar coordinates of the main contour rows to determine the starting edge, and using depth-first search DFS to construct a closed main contour;

[0075] The abnormal topology processing includes repeated curve disambiguation and geometric discontinuity repair;

[0076] The parametric modeling includes importing the repaired main contour into the 3D modeling software, ensuring the correct resolution of the curve through geometric verification, verifying the curve in combination with the geometric tolerance detection method, and issuing a warning and automatically repairing it in time if any deviation is found;

[0077] The entity topology analysis includes entity topology decomposition, surface type identification, geometric feature parameterization, and exception processing;

[0078] The entity topology decomposition includes generating feature sensitive sampling points using an adaptive chord difference method for the reconstructed three-dimensional features and establishing a curvature weight sampling model;

[0079] The surface type identification includes plane feature detection, rotation surface analysis, and torus feature analysis;

[0080] The process optimization closed loop includes multi-modal on-machine detection, real-time data processing, intelligent compensation modeling, and closed-loop control;

[0081] The adaptive correction of the numerical control program includes data analysis and preprocessing, intelligent path matching, path planning optimization, and closed-loop verification.

[0082] Duplicate curve disambiguation includes extracting geometric features, building an adjacency matrix to detect topological conflicts, performing redundant path pruning, and defining a curve set. ,curve From continuous points Indicates that, calculate any two curves and The starting and ending distances, when the distance is less than the threshold ,determination and Overlap and delete one of them, the starting and ending distances are Euclidean distances :

[0083] ;

[0084] ;

[0085] In the formula, yes point The starting point, yes point The starting point, yes point The end point, yes point The end point, yes and The coordinates of the two starting points, yes and The coordinates of the two end points of

[0086] Geometric discontinuity repair includes quickly locating breakpoints based on the R-tree spatial index for the curves after repeated curve disambiguation, processing the breakpoints using the minimum distance bridging algorithm, and defining the open point set. Existence Points, , calculate the Euclidean distance between any two opening points, if it is less than the threshold, restore the curve by interpolation :

[0087] ;

[0088] In the formula, is the Bernstein polynomial, For the Points, Represents the total number of interpolation points, and the interpolation process is optimized by fitting using the least squares method, such as Figure 6 As shown, the repaired curve is guaranteed to be smooth and continuous.

[0089] The parametric modeling includes: after the curve is imported, according to Endpoints Sort and generate rotation bodies in order. The sorting algorithm is based on heuristic search. First, The coordinates are sorted and then The coordinates are sorted to ensure smooth connection of all curves; a three-dimensional rotation body is obtained through rotation transformation The mathematical formula is:

[0090] ;

[0091] In the formula, It is around The rotation matrix for axis rotations, is the rotation angle, and the generated solid of revolution is visualized by computer-aided design software;

[0092] After the 3D modeling is completed, the point cloud data is used to analyze the model in detail, and the plane fitting technology based on the RANSAC algorithm is used to automatically identify the feature surface from the point cloud:

[0093] ;

[0094] In the formula, is the error function, is the normal vector, is the offset, Indicates Points, It represents the total number of plane points. It extracts characteristic surfaces and boundaries through plane fitting, providing a basis for subsequent machining path generation.

[0095] The plane feature detection includes performing principal component analysis (PCA) on the reconstructed three-dimensional features to verify the consistency of normal vectors and establish a flatness error model;

[0096] The rotational surface analysis includes cylindrical surface recognition, conical surface verification and spherical surface determination for the reconstructed three-dimensional features;

[0097] The torus feature analysis includes constructing a hyperbolic feature space for the reconstructed three-dimensional features and detecting the sweeping trajectory using Hough transform.

[0098] The geometric feature parameterization includes the reconstruction of the reference axis system and the inverse calculation of the feature parameters;

[0099] The reconstruction of the reference axis system includes determining the rotation axis and establishing the axis system tolerance band by least square fitting;

[0100] The feature parameter inversion includes respectively obtaining the radius, half angle, main radius and center of the cylindrical surface, conical surface, torus and spherical surface of the three-dimensional feature;

[0101] The abnormality processing includes mixed surface discrimination and transition feature processing;

[0102] The mixed surface discrimination includes constructing a Mahalanobis distance classifier and setting a type confidence threshold to discriminate the mixed surface;

[0103] The transition feature processing includes using curvature derivatives to detect chamfers and fillets, using PointNet++ to classify and identify features of point cloud data, and combining deep learning to automatically identify key features in complex set shapes. The input of the deep learning network is Point cloud dataset , the output is the processing feature category, each layer of the network learns local features through the multi-layer perceptron MLP, and extracts global features through the global pooling operation:

[0104] ;

[0105] In the formula, is the feature of the i-th point, represent No. Points, is the weight, is a constant term, through global pooling Extract global features and finally output machining feature categories to help the system automatically identify key machining features on the workpiece;

[0106] Based on the identified key processing features, the tool path interpolation algorithm is used to automatically generate the finishing milling path to ensure that the tool moves along the optimal path to avoid processing errors. For different processing features, the system automatically adjusts the cutting parameters and optimizes the processing efficiency according to the material and geometric features of the workpiece.

[0107] After the abnormality processing is completed, the characteristics of the rotating workpiece are identified. In the present invention, the rotating workpiece to be processed is as follows Figure 3 As shown, Figure 3 Feature recognition is performed on both sides of Figure 4 and Figure 5 shown.

[0108] The multimodal on-machine detection includes constructing a high-precision tactile sensing unit and standardizing the machine tool coordinate system;

[0109] The construction of the high-precision tactile sensing unit includes constructing a probe dynamic compensation model according to the characteristics of the rotating workpiece, and using a quaternion normal vector solution algorithm to solve the probe dynamic compensation model;

[0110] The normalization of the machine tool coordinate system includes establishing the machine tool coordinate transformation matrix, performing path planning based on the curvature adaptive spiral scanning path, finding the optimal closed path using a global optimization algorithm, iteratively adjusting the link order based on the endpoint set, and minimizing the target parameters. :

[0111] ;

[0112] In the formula, For the closed function of the curve, the local optimal solution problem is avoided through global optimization;

[0113] The real-time data processing includes intelligent analysis of measurement data and analysis of machining allowance;

[0114] The intelligent analysis of the measurement data includes constructing an abnormal point filtering model for the features of the rotating workpiece, adding curvature constraints using the ICP algorithm, and performing point cloud matching;

[0115] The machining allowance analysis includes calculating the normal allowance and the tangential deviation.

[0116] The intelligent compensation modeling includes an inverse surface reconstruction algorithm and dynamic adjustment of process parameters;

[0117] The inverse surface reconstruction algorithm includes using NURBS surfaces for adaptive interpolation and gradient descent optimization;

[0118] The dynamic adjustment of process parameters includes constructing a cutting parameter optimization model and calculating tool compensation amounts.

[0119] The closed-loop control Figure 1 As shown, it includes on-machine detection and data verification for the features of the rotating workpiece. If the verification is qualified, point cloud alignment, margin analysis, compensation modeling, tool path regeneration, secondary processing, and return to on-machine detection are performed. If the verification fails, an alarm is issued and the machine is shut down.

[0120] The data analysis and preprocessing include building an efficient coordinate extraction engine and building a spatial index;

[0121] The construction of the efficient coordinate extraction engine includes using a buffer stream to accelerate the file reading of the features of the rotating workpiece through closed-loop control, and constructing an abnormal data filtering model to perform data filtering;

[0122] The construction of the spatial index includes performing a fast search of point clouds based on Octree; The intelligent path matching includes calling a dynamic distance analysis model and optimizing a projection algorithm; The calling of the dynamic distance analysis model includes calling a hybrid distance calculation strategy and setting an adaptive threshold to perform dynamic distance analysis;

[0123] The optimized projection algorithm includes using LM algorithm to accelerate the surface projection and introducing normal constraint projection correction;

[0124] The path planning optimization includes three types of corrections, namely linear movement, arc path and free-form surface, which are adjusted by minimum energy interpolation method, curvature matching correction and parameterized remapping, respectively, and acceleration continuity, continuity maintenance and bow height error not exceeding 0.001 mm are respectively applied as constraints;

[0125] In path planning optimization, program segments are managed based on a bidirectional linked list, and an incremental program update strategy is called.

[0126] The closed-loop verification is as follows Figure 2 As shown, it includes performing point cloud matching analysis on the features of the acquired rotating workpiece and the manufactured rotating workpiece. If the deviation does not exceed the threshold, it is directly output. If the deviation exceeds the threshold, surface projection optimization, path parameter recalculation, tool compensation update, verification path generation, virtual processing verification, and returning to point cloud matching analysis.

[0127] Starting from the background of intelligent manufacturing, this paper analyzes the challenges faced by the machining technology of rotating workpieces, explores the development status of the machining technology of rotating workpieces based on intelligent manufacturing, and proposes corresponding application strategies to provide a reference for the development of the intelligent manufacturing industry. In the specific implementation, this paper adopts three parts: intelligent geometric reconstruction, feature-driven process planning, and adaptive machining compensation system.

[0128] In intelligent geometric reconstruction, a model-driven automated modeling method is adopted, including geometric feature analysis and mean difference reconstruction, parametric modeling engine, topology optimization and verification.

[0129] Geometric feature analysis and mean difference reconstruction include analyzing the geometric topological relationship of the input engineering drawing through the constraint solver (D-Cubed component), extracting the cross-sectional feature contours of the revolved parts (including straight lines / arcs / splines and other primitives), and performing mean difference model reconstruction based on tolerance zones.

[0130] The parametric modeling engine includes the conversion of 2D contours to 3D entities based on the ACIS geometry kernel, axial stretching to generate basic solids of revolution, polar coordinate transformation to achieve circumferential uniform distribution of feature arrays, and Boolean operations to perform union / difference operations on geometric bodies to construct complex cavity structures.

[0131] Topology optimization and verification include applying B-rep based geometry validity checks, ensuring the manifold properties of the solid model, performing minimum wall thickness detection (Ray-casting algorithm) and process accessibility analysis.

[0132] In feature-driven process planning, an intelligent programming method based on model feature recognition is adopted, including feature extraction and characterization, similarity-driven strategy matching, and incremental feature library optimization.

[0133] The feature extraction and characterization includes geometric topological analysis based on a mean difference model (MID), using multi-scale feature descriptors (including Fourier descriptors and shape context features) to extract processing features, constructing a processing feature knowledge graph, and defining a feature-process mapping relationship.

[0134] The similarity-driven strategy matching includes applying a nearest neighbor search algorithm (k-NN) to perform similarity matching in a processing feature library, and automatically loading the associated processing strategy and parameter combination when the similarity is greater than or equal to a preset threshold of 0.85.

[0135] Incremental feature library optimization includes establishing a feature library expansion mechanism based on online learning, initiating the process parameter migration process for new features whose similarity is less than a preset threshold, realizing knowledge reuse across manufacturing scenarios through transfer learning, and supporting the adaptive generalization of process parameters.

[0136] In the adaptive machining compensation system, a closed-loop machining control system based on digital twins is constructed, including multimodal data fusion acquisition, intelligent diagnosis of machining deviations, and a dynamic path optimization engine.

[0137] The multimodal data fusion acquisition includes an integrated on-machine measurement (OMM) system, which uses a contact probe (Renishaw MP700) for measurement, generates three-dimensional point cloud data in real time, and achieves measurement noise suppression (signal-to-noise ratio improvement ≥35dB) through a Kalman filter.

[0138] The intelligent diagnosis of processing deviation includes establishing a three-dimensional registration model based on the ICP algorithm and calculating a characteristic dimension error vector.

[0139] The dynamic path optimization engine includes the use of PID control algorithm to correct the processing trajectory in real time, the control parameters are adaptively adjusted through the particle swarm optimization (PSO) algorithm, and it supports multi-axis linkage compensation (RTCP function), with a maximum correction speed of 120mm / s.

[0140] The intelligent manufacturing of rotating workpieces is divided into three parts: automatic modeling conversion, feature programming and on-machine detection. Automatic modeling is to use software to identify and read the contour of the rotating body in the customer's drawing, adjust the difference, and then use software development to realize the conversion from 2D to 3D drawings. Then generate the final processing model through stretching, array, Boolean and other operations. Feature programming is to identify the features in the difference model, and then match them in the processing feature library, identify the same or similar features, and automatically match the processing strategy and processing parameters. This system will automatically fill the feature library during the user's use. When the feature matching degree is lower than the preset threshold, it will automatically learn the processing method and processing parameters used by the user, and continuously improve the feature library. The detection system obtains real-time processing data through on-machine measurement, compares the theoretical data, and automatically adjusts the processing path.

[0141] The present invention proposes an intelligent processing flow optimization technology based on two-dimensional drawings and three-dimensional modeling, which aims to extract the contour curves required for processing from two-dimensional drawings through automated means, and generate accurate three-dimensional models and processing paths. The core technologies in the scheme include automatic contour curve recognition and repair, three-dimensional modeling and topological analysis, deep learning-assisted feature recognition, processing path optimization and adaptive control, etc. By combining point cloud technology, deep learning technology and real-time feedback system, the scheme can not only effectively improve the processing accuracy, but also reduce manual intervention, adapt to complex processing tasks, and improve production efficiency. In the processing process, the accurate extraction and repair of contour curves are the key links to ensure processing accuracy. Especially in complex geometric shapes, the repair and extraction of contour curves are crucial.

[0142] The present invention obtains the graphics and size information exported by the two-dimensional software by reading the text file, and offsets the surface that needs to be adjusted in the middle difference by calling the offset surface method of the three-dimensional software in the automatic modeling mode, and adjusts the corresponding size to the middle difference state. The identified surface features are matched with the corresponding processing process, processing strategy, processing parameters, processing tools, processing machine tools, and process equipment from the formed feature database. The general matching process is to identify the processing features starting from the surface with the largest X value, identify the feature image by machine learning, and then match the processing parameters from the feature library.

[0143] Taking the blank size of Ø152×78 and the material of #45 stainless steel as an example, the processing process includes 4 processing processes, namely rough turning big end 1, fine turning big end 1, rough turning small end 2, and fine turning small end 2. The processing machine tool adopts Mazak QTE-300L horizontal lathe, and the clamping method adopts three-jaw chuck positioning clamping. The processing parameters and processing strategies are automatically matched from the SQL SERVER parameter library. The matching content is to match the most suitable speed feed parameters from the parameter library according to the processing material, machine tool type, processing procedure (rough turning, semi-finishing turning, fine turning), and roughness requirements. If there are no parameters suitable for the current working conditions in the parameter library, the software will automatically record the currently set parameters and automatically update the parameter library.

[0144] The section inspection line is discretized into a large number of points by using the function of discretizing the curve into points. The UV direction and normal vector of the specified point on the surface can be obtained by obtaining the vector of the specified point on the surface. Finally, the curve is generated according to the obtained point and vector direction. The generated curve can be added to the processing strategy by adding the curve to the processing strategy in the 3D software to automatically generate the program. In the process of generating the program, due to the continuous changes in the processing conditions, the processing parameter library may not be able to meet the current situation, resulting in the inability to generate the program. In order to solve this problem, the system has added process iteration and learning functions. When encountering unmatched conditions, the system will prompt the user to manually operate unknown processing features. When generating the program, the system will record the process and processing parameters provided by the user, and automatically recommend them when encountering similar conditions next time.

[0145] The present invention uses PC on-machine measurement software, controls the numerical control system by measuring the NC system, the numerical control system controls the machine tool servo system and then controls the turntable, and also controls the probe through the spindle, the probe measures the equipment on the turntable and then transmits the data to the numerical control system, and the numerical control system feeds back the measurement results to the PC on-machine measurement software.

[0146] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features may be replaced by equivalents, and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent manufacturing of a rotating workpiece, characterized in that: Including automated process design, physical topology analysis, closed-loop process optimization, and adaptive correction of CNC programs; The automated process design includes geometric feature extraction, abnormal topology processing and parametric modeling; The geometric feature extraction includes size topology association analysis and contour boundary reconstruction; The dimension topology association analysis includes obtaining the graphics element of the rotating workpiece, calling the 2D drawing software API to traverse the graphics element database, establishing the dimension-graphic element mapping relationship, and determining the main contour through extreme value screening; The reconstruction of contour boundaries includes sorting the polar coordinates of the main contour rows to determine the starting edge, and using depth-first search DFS to construct a closed main contour; The abnormal topology processing includes repeated curve disambiguation and geometric discontinuity repair; The parametric modeling includes importing the repaired main contour into the 3D modeling software, ensuring the correct resolution of the curve through geometric verification, verifying the curve in combination with the geometric tolerance detection method, and issuing a warning and automatically repairing it in time if any deviation is found; The entity topology analysis includes entity topology decomposition, surface type identification, geometric feature parameterization, and exception processing; The entity topology decomposition includes generating feature sensitive sampling points using an adaptive chord difference method for the reconstructed three-dimensional features and establishing a curvature weight sampling model; The surface type identification includes plane feature detection, rotation surface analysis, and torus feature analysis; The process optimization closed loop includes multi-modal on-machine detection, real-time data processing, intelligent compensation modeling, and closed-loop control; The adaptive correction of the numerical control program includes data analysis and preprocessing, intelligent path matching, path planning optimization, and closed-loop verification.

2. The intelligent manufacturing method for a rotating workpiece according to claim 1, characterized in that: Duplicate curve disambiguation includes extracting geometric features, building an adjacency matrix to detect topological conflicts, performing redundant path pruning, and defining a curve set. ,curve From continuous points Indicates that, calculate any two curves and The starting and ending distances, when the distance is less than the threshold ,determination and Overlap and delete one of them, the starting and ending distances are Euclidean distances : ； ； In the formula, yes point The starting point, yes point The starting point, yes point The end point, yes point The end point, yes and The coordinates of the two starting points, yes and The coordinates of the two end points of Geometric discontinuity repair includes quickly locating breakpoints based on the R-tree spatial index for the curves after repeated curve disambiguation, processing the breakpoints using the minimum distance bridging algorithm, and defining the open point set. Existence Points, , calculate the Euclidean distance between any two opening points, if it is less than the threshold, restore the curve by interpolation : ； In the formula, is the Bernstein polynomial, For the Points, It represents the total number of interpolation points. The least square method is used to fit and optimize the interpolation process to ensure that the repaired curve is smooth and continuous.

3. The intelligent manufacturing method for a rotating workpiece according to claim 2, characterized in that: The parametric modeling includes: after the curve is imported, according to Endpoints Sort and generate rotation bodies in order. The sorting algorithm is based on heuristic search. First, The coordinates are sorted and then The coordinates are sorted to ensure smooth connection of all curves; Obtaining a three-dimensional solid of revolution by rotation transformation The mathematical formula is: ； In the formula, It is around The rotation matrix for axis rotations, is the rotation angle, and the generated solid of revolution is visualized by computer-aided design software; After the 3D modeling is completed, the point cloud data is used to analyze the model in detail, and the plane fitting technology based on the RANSAC algorithm is used to automatically identify the feature surface from the point cloud: ； In the formula, is the error function, is the normal vector, is the offset, Indicates Points, It represents the total number of plane points. It extracts characteristic surfaces and boundaries through plane fitting, providing a basis for subsequent machining path generation.

4. The intelligent manufacturing method for a rotating workpiece according to claim 3 is characterized in that: The plane feature detection includes performing principal component analysis (PCA) on the reconstructed three-dimensional features to verify the consistency of normal vectors and establish a flatness error model; The rotational surface analysis includes cylindrical surface recognition, conical surface verification and spherical surface determination for the reconstructed three-dimensional features; The torus feature analysis includes constructing a hyperbolic feature space for the reconstructed three-dimensional features and detecting the sweeping trajectory using Hough transform.

5. The intelligent manufacturing method for a rotating workpiece according to claim 4 is characterized in that: The geometric feature parameterization includes the reconstruction of the reference axis system and the inverse calculation of the feature parameters; The reconstruction of the reference axis system includes determining the rotation axis and establishing the axis system tolerance band by least square fitting; The feature parameter inversion includes respectively obtaining the radius, half angle, main radius and center of the cylindrical surface, conical surface, torus and spherical surface of the three-dimensional feature; The abnormality processing includes mixed surface discrimination and transition feature processing; The mixed surface discrimination includes constructing a Mahalanobis distance classifier and setting a type confidence threshold to discriminate the mixed surface; The transition feature processing includes using curvature derivatives to detect chamfers and fillets, using PointNet++ to classify and identify features of point cloud data, and combining deep learning to automatically identify key features in complex set shapes. The input of the deep learning network is Point cloud dataset , the output is the processing feature category, each layer of the network learns local features through the multi-layer perceptron MLP, and extracts global features through the global pooling operation: ； In the formula, is the feature of the i-th point, represent No. Points, is the weight, is a constant term, through global pooling Extract global features and finally output machining feature categories to help the system automatically identify key machining features on the workpiece; Based on the identified key processing features, the tool path interpolation algorithm is used to automatically generate the finishing milling path to ensure that the tool moves along the optimal path to avoid processing errors. For different processing features, the system automatically adjusts the cutting parameters and optimizes the processing efficiency according to the material and geometric features of the workpiece. After completing the exception handling, the features of the rotating workpiece are identified.

6. The intelligent manufacturing method of a rotating workpiece according to claim 5, characterized in that: The multimodal on-machine detection includes constructing a high-precision tactile sensing unit and standardizing the machine tool coordinate system; The construction of the high-precision tactile sensing unit includes constructing a probe dynamic compensation model according to the characteristics of the rotating workpiece, and using a quaternion normal vector solution algorithm to solve the probe dynamic compensation model; The normalization of the machine tool coordinate system includes establishing the machine tool coordinate transformation matrix, performing path planning based on the curvature adaptive spiral scanning path, finding the optimal closed path using a global optimization algorithm, iteratively adjusting the link order based on the endpoint set, and minimizing the target parameters. : ； In the formula, For the closed function of the curve, the local optimal solution problem is avoided through global optimization; The real-time data processing includes intelligent analysis of measurement data and analysis of machining allowance; The intelligent analysis of the measurement data includes constructing an abnormal point filtering model for the features of the rotating workpiece, adding curvature constraints using the ICP algorithm, and performing point cloud matching; The machining allowance analysis includes calculating the normal allowance and the tangential deviation.

7. The intelligent manufacturing method of a rotating workpiece according to claim 6, characterized in that: The intelligent compensation modeling includes an inverse surface reconstruction algorithm and dynamic adjustment of process parameters; The inverse surface reconstruction algorithm includes using NURBS surfaces for adaptive interpolation and gradient descent optimization; The dynamic adjustment of process parameters includes constructing a cutting parameter optimization model and calculating tool compensation amounts.

8. The intelligent manufacturing method of a rotating workpiece according to claim 7, characterized in that: The closed-loop control includes on-machine detection and data verification for the features of the rotating workpiece. If the verification is qualified, point cloud alignment, residual analysis, compensation modeling, tool path regeneration, secondary processing, and return to on-machine detection are performed. If the verification fails, an alarm is issued and the machine is shut down.

9. The intelligent manufacturing method of a rotating workpiece according to claim 8, characterized in that: The data analysis and preprocessing include building an efficient coordinate extraction engine and building a spatial index; The construction of the efficient coordinate extraction engine includes using a buffer stream to accelerate the file reading of the features of the rotating workpiece through closed-loop control, and constructing an abnormal data filtering model to perform data filtering; The construction of the spatial index includes performing a fast search of point clouds based on Octree; The intelligent path matching includes calling a dynamic distance analysis model and optimizing a projection algorithm; The calling of the dynamic distance analysis model includes calling a hybrid distance calculation strategy and setting an adaptive threshold to perform dynamic distance analysis; The optimized projection algorithm includes using LM algorithm to accelerate the surface projection and introducing normal constraint projection correction; The path planning optimization includes three types of corrections, namely linear movement, arc path and free-form surface, which are adjusted by minimum energy interpolation method, curvature matching correction and parameterized remapping, respectively, and acceleration continuity, continuity maintenance and bow height error not exceeding 0.001 mm are respectively applied as constraints; In path planning optimization, program segments are managed based on a bidirectional linked list, and an incremental program update strategy is called.

10. The intelligent manufacturing method of a rotating workpiece according to claim 9, characterized in that: The closed-loop verification includes performing point cloud matching analysis on the features of the acquired rotating workpiece and the manufactured rotating workpiece. If the deviation does not exceed the threshold, it is directly output; if the deviation exceeds the threshold, surface projection optimization, path parameter recalculation, tool compensation update, verification path generation, virtual processing verification, and returning to the point cloud matching analysis.

Images