Process optimization method for complex curved surface

Abstract

The invention belongs to the technical field of complex curved surface process optimization, and discloses a process optimization method for a complex curved surface. The method comprises the steps of obtaining a curved surface three-dimensional structure diagram of a machined part; performing curved surface division on the curved surface of the processing part based on the three-dimensional model of the curved surface of the processing part to obtain N sub curved surfaces of the processing part; performing feature extraction on the N machined part sub-curved surfaces to obtain machined part sub-curved surface feature data corresponding to the N machined part sub-curved surfaces; based on the feature data of the N machined part sub-curved surfaces, performing self-adaptive adjustment and optimization on an idle stroke path of machined part curved surface machining of the N machined part sub-curved surfaces; according to the scheme, the limitation of traditional numerical control machining is broken through, idle stroke path adjustment and optimization based on an intelligent optimization algorithm are adopted, the machining sequence of the sub curved surfaces of the machined part is optimized, the optimal machining sequence of the machined part is automatically selected, and the effects of reducing non-cutting idle stroke and repeated cutting phenomena and improving the machining quality of the curved surfaces of the machined part are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of complex curved surface process optimization, and more specifically, to a process optimization method for complex curved surfaces. Background Art

[0002] CNC machining technology has been widely used due to its excellent machining accuracy and stability, and valve surface machining is an important application scenario. With the rapid development of modern industry, valves are taking on increasingly critical fluid control tasks under various complex working conditions, which has prompted the valve surface structure to continue to evolve in the direction of high complexity and multi-surface combinations. Today, the shapes of valve surfaces are becoming increasingly complex and diverse, such as irregular streamlines, multi-curvature composite surfaces, etc., which have brought unprecedented challenges to valve manufacturing, and the machining quality of valve surfaces has become one of the core factors that determine the overall performance of valves.

[0003] The Chinese patent application with the existing authorization announcement number CN109901514B proposes a CNC process optimization and adjustment method for complex parts for process reuse, which is used to solve the technical problem of low efficiency of existing CNC process optimization and adjustment methods. The technical solution is to optimize and adjust from two aspects: processing tools and cutting depth. In terms of processing tools, the combination of processing tools for complex features is optimized, and the tools for local structure processing are combined and adjusted to reduce the number of tool changes and shorten the length of empty tools. In terms of cutting depth parameters, the cutting depth is optimized by deciding on a reasonable cutting mode to obtain an optimized processing area, so that the tool processing capacity can be fully utilized, and ultimately an efficient CNC process solution is obtained to improve the processing efficiency of complex parts.

[0004] In valve surface machining, the existing technology generally uses pre-set machining paths and surface parameters for machining operations. This traditional machining mode has significant defects. Due to the lack of real-time dynamic monitoring means and adaptive adjustment mechanism for the valve surface machining process, it is difficult to take into account multiple key performance indicators during the machining process, resulting in a large number of non-cutting empty strokes and repeated cutting during the machining process, which in turn prolongs the machining time and affects the machining quality of the valve surface. Ultimately, it cannot meet the high-precision and high-efficiency requirements of modern industry for valve surface machining.

[0005] Based on this, developing a valve surface CNC machining technology with real-time monitoring and adaptive adjustment capabilities to achieve coordinated optimization of multiple key performance indicators has become an important issue that needs to be urgently addressed in the current valve manufacturing field. Summary of the invention

[0006] To overcome the above-mentioned defects of the prior art and to achieve the above object, the present invention provides the following technical solution: A process optimization method for complex curved surfaces, comprising:

[0007] Obtain the three-dimensional structure diagram of the curved surface of the workpiece;

[0008] Based on the three-dimensional model of the curved surface of the workpiece, divide the curved surface of the workpiece to obtain N sub-curved surfaces of the workpiece;

[0009] Extract the features of the N sub-curved surfaces of the workpiece to obtain the feature data of the sub-curved surfaces of the workpiece corresponding to the N sub-curved surfaces of the workpiece;

[0010] Based on the feature data of the N sub-curved surfaces of the workpiece, adaptively optimize the non-cutting path of the machining of the curved surface of the N sub-curved surfaces of the workpiece.

[0011] Further, the method for adaptively optimizing the non-cutting path of the machining of the curved surface of the N sub-curved surfaces of the workpiece based on the feature data of the N sub-curved surfaces of the workpiece includes:

[0012] Input the feature data of the N sub-curved surfaces of the workpiece into the machining path evaluation model to obtain H candidate machining paths for the workpiece; each candidate machining path for the workpiece includes the machining sequence of the N sub-curved surfaces of the workpiece; construct the H candidate machining paths for the workpiece into a machining path population;

[0013] Evaluate the fitness of the H candidate machining paths for the workpiece in the machining path population to obtain the corresponding machining path fitness values;

[0014] Sort the H machining path fitness values in descending order, and select the first Q machining path fitness values to construct an elite population of machining paths;

[0015] Perform mutation operations on the elite population of machining paths according to a preset method to obtain R mutant candidate machining paths for the workpiece and add them to the elite population of machining paths; overwrite the elite population of machining paths to the machining path population, and set H = Q + R;

[0016] Repeat the above operations. When the number of iterations reaches the preset maximum number of iterations, stop the execution; select the candidate machining path for the workpiece with the largest machining path fitness value from the machining path population as the optimal solution of the non-cutting path for the machining of the curved surface of the workpiece.

[0017] Further, the method for obtaining the R mutant candidate machining paths for the workpiece includes:

[0018] Preset path fitness threshold one and path fitness threshold two, where path fitness threshold one is less than path fitness threshold two; divide the elite population of processing paths according to path fitness threshold one and path fitness threshold two to obtain a high fitness path set, a medium fitness path set, and a low fitness path set;

[0019] Perform random crossover in the high fitness path set to obtain NUM1 candidate processing paths for the workpiece and add them to the high fitness path set;

[0020] Perform random crossover between the candidate processing paths for the workpiece in the medium fitness path set and the candidate processing paths for the workpiece in the high fitness path set to obtain NUM2 candidate processing paths for the workpiece and add them to the medium fitness path set; perform random crossover in the medium fitness path set to obtain NUM3 candidate processing paths for the workpiece and add them to the medium fitness path set;

[0021] Perform random crossover between the candidate processing paths for the workpiece in the low fitness path set and the candidate processing paths for the workpiece in the high fitness path set to obtain NUM4 candidate processing paths for the workpiece and add them to the low fitness path set; perform random crossover in the low fitness path set to obtain NUM5 candidate processing paths for the workpiece and add them to the low fitness path set;

[0022] When NUM1 + NUM2 + NUM3 + NUM4 + NUM5 = R, stop the above random crossover operation to obtain R mutant candidate processing paths for the workpiece.

[0023] Furthermore, the method for obtaining the high fitness path set, the medium fitness path set, and the low fitness path set includes:

[0024] Construct a high fitness path set from the candidate processing paths for the workpiece in the elite population of processing paths whose processing path fitness value is greater than or equal to path fitness threshold two; construct a medium fitness path set from the candidate processing paths for the workpiece in the elite population of processing paths whose processing path fitness value is greater than or equal to path fitness threshold one and less than path fitness threshold two; construct a low fitness path set from the candidate processing paths for the workpiece in the elite population of processing paths whose processing path fitness value is less than path fitness threshold one.

[0025] Furthermore, the method for obtaining the N sub - surfaces of the workpiece includes:

[0026] S100: Randomly distribute M sampling points on the three - dimensional structure diagram of the workpiece surface;

[0027] S101: Let the initial value of m be 1, and the value range of m is from 1 to M; let the initial value of n be 1, and n is a counting variable;

[0028] S102: Obtain the machining part surface parameter coordinates of the m-th sampling point, and obtain the corresponding machining part surface parameter equation based on the machining part surface parameter coordinates; calculate the machining part surface curvature of the m-th sampling point based on the machining part surface parameter equation.

[0029] S103: Determine the corresponding machining part surface movement step and machining part sub-surface type based on the machining part surface curvature of the m-th sampling point; perform a diffusion operation according to the machining part surface movement step and machining part sub-surface type to obtain the machining part surface diffusion region of the m-th sampling point.

[0030] S104: Use the edge points of the machining part surface diffusion region of the m-th sampling point as a new diffusion starting point, and repeat S103. When there are no longer points with the same machining part sub-surface type as the m-th sampling point during the diffusion process, the diffusion stops, and the machining part sub-surface of the n-th sampling point is obtained; let n = n + 1, and execute S105.

[0031] S105: Let m = m + 1. If m is less than or equal to M, then execute S106; if m is greater than M, then end the current process, assign n to N, and obtain N machining part sub-surfaces.

[0032] S106: If the m-th sampling point is already included in the constructed machining part sub-surface, then execute S105; if the m-th sampling point is not included in the constructed machining part sub-surface, then execute S102 to S104.

[0033] Further, the method for obtaining the machining part surface diffusion region of the m-th sampling point includes:

[0034] With the m-th sampling point as the center and the machining part surface movement step as the radius, form the diffusion to-be-processed region of the m-th sampling point; calculate the machining part surface curvature of the points on the edge of the diffusion to-be-processed region, denoted as the edge surface curvature, and obtain the corresponding machining part sub-surface type according to the edge surface curvature.

[0035] Mark the points on the edge of the diffusion to-be-processed region with the same machining part sub-surface type as the m-th sampling point as the to-be-connected points of the m-th sampling point; connect all the to-be-connected points of the m-th sampling point with the m-th sampling point to form the machining part surface diffusion region of the m-th sampling point.

[0036] Further, the method for obtaining the machining part surface curvature of the m-th sampling point includes:

[0037] Take the first-order partial derivative of the machining part surface parameter equation with respect to u m to obtain the first first-order partial derivative; take the first-order partial derivative of the machining part surface parameter equation with respect to v mFind the first-order partial derivatives to obtain the second-order partial derivatives; u m and v m are the parametric coordinates of the workpiece surface at the m-th sampling point, and the unit normal vector of the m-th sampling point is calculated based on the first-order partial derivatives and the second-order partial derivatives;

[0038] Multiply the first-order partial derivatives by the first-order partial derivatives to obtain the corresponding first-order formal parameters; multiply the first-order partial derivatives by the second-order partial derivatives to obtain the corresponding second-order formal parameters; multiply the second-order partial derivatives by the second-order partial derivatives to obtain the corresponding third-order formal parameters;

[0039] The first-order partial derivatives with respect to u m Find the second-order partial derivatives to obtain the first second-order partial derivatives; the first-order partial derivatives with respect to v m Find the second-order partial derivatives to obtain the second second-order partial derivatives; the second-order partial derivatives with respect to v m Find the second-order partial derivatives to obtain the third second-order partial derivatives;

[0040] Multiply the first second-order partial derivatives by the unit normal vector to obtain the first second-order formal parameters of the m-th sampling point; multiply the second second-order partial derivatives by the unit normal vector to obtain the second second-order formal parameters of the m-th sampling point; multiply the third second-order partial derivatives by the unit normal vector to obtain the third second-order formal parameters of the m-th sampling point;

[0041] Calculate the curvature of the workpiece surface at the m-th sampling point based on the first-order formal parameters, the second-order formal parameters, the third-order formal parameters, the first second-order formal parameters, the second second-order formal parameters, and the third second-order formal parameters.

[0042] Furthermore, the method for determining the corresponding movement step of the workpiece surface and the type of workpiece sub-surface based on the curvature of the workpiece surface at the m-th sampling point includes:

[0043] Preset a first threshold for the curvature of the workpiece surface and a second threshold for the curvature of the workpiece surface, where the first threshold for the curvature of the workpiece surface is less than the second threshold for the curvature of the workpiece surface; preset a first movement step of the workpiece surface, a second movement step of the workpiece surface, and a third movement step of the workpiece surface, where the first movement step of the workpiece surface is less than the second movement step of the workpiece surface, and the second movement step of the workpiece surface is less than the third movement step of the workpiece surface;

[0044] If the curvature of the workpiece surface at the m-th sampling point is greater than or equal to the second threshold for the curvature of the workpiece surface, then the type of the workpiece sub-surface is a high-curvature region, and the movement step of the workpiece surface is the first movement step of the workpiece surface;

[0045] If the curvature of the workpiece surface at the m-th sampling point is greater than or equal to the first threshold of the workpiece surface curvature and less than the second threshold of the workpiece surface curvature, the type of the workpiece sub-surface is the medium curvature region, and the moving step of the workpiece surface is the second moving step of the workpiece surface;

[0046] If the curvature of the workpiece surface at the m-th sampling point is less than the first threshold of the workpiece surface curvature, the type of the workpiece sub-surface is the low curvature region, and the moving step of the workpiece surface is the third moving step of the workpiece surface.

[0047] Furthermore, the method for obtaining the workpiece sub-surface feature data corresponding to N workpiece sub-surfaces includes:

[0048] S200: Let the initial value of n be 1, and the value range of n is from 1 to N;

[0049] S201: Obtain the coordinate range of the workpiece surface parameters, the first first-order form parameters, the second first-order form parameters, and the third first-order form parameters of the n-th workpiece sub-surface; calculate the corresponding area of the workpiece sub-surface according to the coordinate range of the workpiece surface parameters, the first first-order form parameters, the second first-order form parameters, and the third first-order form parameters; divide the boundary of the n-th workpiece sub-surface into J boundary points, and calculate the boundary length of the n-th workpiece sub-surface according to the coordinates of the J boundary points;

[0050] Obtain the curvature of the workpiece surface of the J boundary points in the same way as the method for obtaining the curvature of the workpiece surface; calculate the surface boundary complexity of the n-th workpiece sub-surface according to the curvature of the workpiece surface of the J boundary points;

[0051] S202: Construct the corresponding workpiece sub-surface feature data from the area of the n-th workpiece sub-surface, the boundary length, and the surface boundary complexity;

[0052] S203: Let n = n + 1. If n is less than or equal to N, continue to execute S201 to S202. If n is greater than N, end the current process.

[0053] Furthermore, the training method of the machining path evaluation model includes:

[0054] Pre-construct a machining path evaluation data set, where the machining path evaluation data set includes Y groups of machining path evaluation data and H machining workpiece candidate paths corresponding to the Y groups of machining path evaluation data. Y is a positive integer greater than 0. The machining path evaluation data includes N workpiece sub-surface feature data; divide the machining path evaluation data set into a machining path evaluation data training set and a machining path evaluation data verification set, where the machining path evaluation data training set is used for parameter learning of the machining path evaluation model, and the machining path evaluation data verification set is used for real-time evaluation of the generalization ability of the machining path evaluation model;

[0055] During the training process of the machining path evaluation model, a deep neural network structure based on a multi-layer perceptron is adopted. The machining path evaluation data is converted into feature vectors as inputs. Non-linear features in the data are extracted through multiple hidden layers, and finally, the softmax activation function is used in the output layer to generate the probability distribution of H candidate machining paths for the workpieces. The H candidate machining paths corresponding to the maximum probability are output as the final prediction result. The training process aims to minimize the cross-entropy loss function. At the same time, an early stopping strategy is introduced to monitor the performance of the machining path evaluation data validation set. When the prediction accuracy on the machining path evaluation data validation set reaches the preset threshold, it is considered that the machining path evaluation model has converged, and the training stops immediately.

[0056] Compared with the prior art, the technical effects and advantages of the process optimization method for complex surfaces of the present invention are as follows:

[0057] Based on the three-dimensional structure model of the workpiece surface, this solution divides and extracts features from the workpiece surface, divides the complex workpiece surface into N sub-workpiece surfaces, and ensures that the machining strategy is optimized individually for different regions. The non-cutting path is optimized based on an intelligent optimization algorithm, and the machining order of the sub-workpiece surfaces of the workpiece is optimized to automatically select the optimal machining order of the workpiece, reduce the ineffective movement of the tool between the sub-workpiece surfaces of the workpiece, and make the machining process more intelligent and adaptive.

[0058] In addition, considering that when the tool for machining the workpiece surface switches from one sub-workpiece surface to another, problems such as trajectory mutation or sharp tool turning may occur, this solution further optimizes the smoothness of the surface machining path for switching across sub-workpiece surfaces, and adaptively selects the optimal tool movement path based on the smoothness of the surface machining path of different candidate paths.

[0059] In summary, this solution breaks through the limitations of traditional numerical control machining. It can not only optimize the path within the sub-surface but also intelligently optimize the switching across sub-surfaces, ensuring continuous trajectories and smooth transitions during tool switching, and ultimately achieving the effects of reducing non-cutting non-cutting strokes and repeated cutting phenomena and improving the machining quality of the workpiece surface. Brief Description of the Drawings

[0060] Figure 1 Schematic diagram of a process optimization system for complex surfaces according to Embodiment 1 of the present invention;

[0061] Figure 2 Flowchart of a process optimization method for complex surfaces according to Embodiment 3 of the present invention;

[0062] Figure 3Schematic diagram of a process optimization system for complex curved surfaces according to Embodiment 2 of the present invention;

[0063] Figure 4 It is a flowchart of a method for adaptively optimizing the non-cutting path of machining the curved surface of a workpiece. Specific embodiments

[0064] Next, the technical solutions in the embodiments of the present invention will be described in detail, clearly and completely with reference to the accompanying drawings in the embodiments of the present invention. It should be particularly noted that the specific embodiments described below are only used to better illustrate and explain the technical solutions of the present invention, aiming to enable those skilled in the art to better understand and implement the present invention, and should not be construed as limiting the protection scope of the present invention. Without departing from the spirit and essence of the present invention, those skilled in the art can modify, adjust or make equivalent substitutions according to the content disclosed in the present invention, and these should all be regarded as the protection scope of the present invention.

[0065] Embodiment 1

[0066] Please refer to Figure 1 As shown, this embodiment discloses a process optimization system for complex curved surfaces, including a first acquisition module, a first processing module, a second processing module, and a first optimization module. Each module is connected by wire and / or wirelessly to achieve data transmission.

[0067] The first acquisition module is used to obtain the three-dimensional structure diagram of the curved surface of the workpiece.

[0068] It should be noted that the three-dimensional structure diagram of the curved surface of the workpiece is obtained from the three-dimensional model database of the curved surface of the workpiece in the workpiece manufacturing enterprise. In the initial stage of product development, the workpiece manufacturing enterprise will use a computer-aided design (CAD) system to model the overall structure and its curved surface of the workpiece in detail, so as to obtain the three-dimensional structure diagrams of the curved surfaces of different workpieces. The workpiece includes valves.

[0069] The first processing module divides the curved surface of the workpiece based on the three-dimensional model of the curved surface of the workpiece to obtain N sub-curved surfaces of the workpiece.

[0070] The method for obtaining the N sub-curved surfaces of the workpiece includes:

[0071] S100: Randomly distribute M sampling points on the three-dimensional structure diagram of the curved surface of the workpiece;

[0072] S101: Let the initial value of m be 1, and the value range of m is from 1 to M; let the initial value of n be 1, and n is a counting variable used to record the current number of sub-curved surfaces of the workpiece that have been identified;

[0073] S102: Obtain the machining part surface parameter coordinates of the m-th sampling point, and based on the machining part surface parameter coordinates, obtain the corresponding machining part surface parameter equation; calculate the machining part surface curvature of the m-th sampling point based on the machining part surface parameter equation;

[0074] S103: Determine the corresponding machining part surface movement step and machining part sub-surface type based on the machining part surface curvature of the m-th sampling point; perform a diffusion operation according to the machining part surface movement step and machining part sub-surface type to obtain the machining part surface diffusion region of the m-th sampling point;

[0075] S104: Use the edge points of the machining part surface diffusion region of the m-th sampling point as the new diffusion starting point, and repeat the execution of S103. When there are no longer points of the same machining part sub-surface type as the m-th sampling point during the diffusion process, the diffusion stops, and the machining part sub-surface of the n-th sampling point is obtained; let n = n + 1, and execute S105;

[0076] S105: Let m = m + 1. If m is less than or equal to M, then execute S106; if m is greater than M, then end the current process, assign n to N, and obtain N machining part sub-surfaces;

[0077] S106: If the m-th sampling point is already included in the constructed machining part sub-surface, then execute S105; if the m-th sampling point is not included in the constructed machining part sub-surface, then execute S102 to S104.

[0078] The method for obtaining the machining part surface diffusion region of the m-th sampling point includes:

[0079] Taking the m-th sampling point as the center and the machining part surface movement step as the radius, form the diffusion to-be-processed region of the m-th sampling point; calculate the machining part surface curvature of the points on the edge of the diffusion to-be-processed region, denoted as the edge surface curvature, and obtain the corresponding machining part sub-surface type according to the edge surface curvature;

[0080] Mark the points on the edge of the diffusion to-be-processed region with the same machining part sub-surface type as the m-th sampling point as the to-be-connected points of the m-th sampling point; connect all the to-be-connected points of the m-th sampling point with the m-th sampling point to form the machining part surface diffusion region of the m-th sampling point.

[0081] The method for obtaining the m-th machining part surface parameter equation includes:

[0082] S m (u m , v m ) = (x m (u m , v m ), y m (um , v m ), z m (u m , v m ));

[0083] Among them, S m (u m , v m ) is the parametric equation of the workpiece surface at the m-th sampling point, where u m and v m are the parametric coordinates of the workpiece surface at the m-th sampling point, x m (u m , v m ) is the parametric equation of the horizontal axis at the m-th sampling point formed by u m and v m ; y m (u m , v m ) is the parametric equation of the vertical axis at the m-th sampling point formed by u m and v m ; z m (u m , v m ) is the parametric equation of the vertical axis at the m-th sampling point formed by u m and v m .

[0084] For example, if the workpiece surface is a rotational ellipsoid, the parametric equation of the workpiece surface is:

[0085] S m (u m , v m ) = (a × cos u m × sin v m , b × sin u m × sin v m , c × cos v m );

[0086] Among them, a is the semi-axis length of the rotational ellipsoid along the X-axis, b is the semi-axis length of the rotational ellipsoid along the Y-axis, and c is the semi-axis length of the rotational ellipsoid along the Z-axis.

[0087] The method for obtaining the curvature of the workpiece surface at the m-th sampling point includes:

[0088] Take the first-order partial derivative of the parametric equation of the workpiece surface with respect to u m to obtain the first first-order partial derivative; take the first-order partial derivative of the parametric equation of the workpiece surface with respect to v m to obtain the second first-order partial derivative;

[0089] Calculate the unit normal vector of the m-th sampling point based on the first first-order partial derivative and the second first-order partial derivative;

[0090] Multiply the first first-order partial derivative by the first first-order partial derivative to obtain the corresponding first first-order formal parameter; multiply the first first-order partial derivative by the second first-order partial derivative to obtain the corresponding second first-order formal parameter; multiply the second first-order partial derivative by the second first-order partial derivative to obtain the corresponding third first-order formal parameter;

[0091] The first first-order partial derivative with respect to u m Find the second-order partial derivative to obtain the first second-order partial derivative; the first first-order partial derivative with respect to v m Find the second-order partial derivative to obtain the second second-order partial derivative; the second first-order partial derivative with respect to v m Find the second-order partial derivative to obtain the third second-order partial derivative;

[0092] Multiply the first second-order partial derivative by the unit normal vector to obtain the first second-order formal parameter of the m-th sampling point; multiply the second second-order partial derivative by the unit normal vector to obtain the second second-order formal parameter of the m-th sampling point; multiply the third second-order partial derivative by the unit normal vector to obtain the third second-order formal parameter of the m-th sampling point;

[0093] Calculate the curvature of the workpiece surface at the m-th sampling point based on the first first-order formal parameter, the second first-order formal parameter, the third first-order formal parameter, the first second-order formal parameter, the second second-order formal parameter, and the third second-order formal parameter.

[0094] The first first-order partial derivative is:

[0095]

[0096] Where, is the first first-order partial derivative, is the first-order partial derivative of the horizontal axis parametric equation of the m-th sampling point with respect to u m , is the first-order partial derivative of the vertical axis parametric equation of the m-th sampling point with respect to u m , is the first-order partial derivative of the vertical axis parametric equation of the m-th sampling point with respect to u m ;

[0097] The second first-order partial derivative is:

[0098]

[0099] Where, is the second first-order partial derivative, is the first-order partial derivative of the horizontal axis parametric equation of the m-th sampling point with respect to v m ; is the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m , is the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m .

[0100] The method for calculating the unit normal vector of the m-th sampling point based on the first first-order partial derivative and the second first-order partial derivative includes:

[0101]

[0102] wherein, is the unit normal vector of the m-th sampling point. is the normal vector of the m-th sampling point, is the modulus of the normal vector of the m-th sampling point.

[0103] The first first-order form parameter is:

[0104]

[0105] wherein, DYYJ m is the first first-order form parameter of the m-th sampling point.

[0106] The second first-order form parameter is:

[0107]

[0108] wherein, DEYJ m is the second first-order form parameter of the m-th sampling point.

[0109] The third first-order form parameter is:

[0110]

[0111] wherein, DSYJ m is the third first-order form parameter of the m-th sampling point.

[0112] The first second-order partial derivative is:

[0113]

[0114] wherein, is the first second-order partial derivative, is the second-order partial derivative of the first-order partial derivative of the horizontal-axis parametric equation of the m-th sampling point with respect to u m , is the second-order partial derivative of the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to u m , is the second-order partial derivative of the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to um The second-order partial derivative.

[0115] The second second-order partial derivative is:

[0116]

[0117] Wherein, is the second second-order partial derivative, is the second-order partial derivative of the first-order partial derivative of the horizontal-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative. is the second-order partial derivative of the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative. is the second-order partial derivative of the first-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative.

[0118] The third second-order partial derivative is:

[0119]

[0120] Wherein, is the third second-order partial derivative, is the second-order partial derivative of the second-order partial derivative of the horizontal-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative. is the second-order partial derivative of the second-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative. is the second-order partial derivative of the second-order partial derivative of the vertical-axis parametric equation of the m-th sampling point with respect to v m The second-order partial derivative.

[0121] The first second-order form parameter is:

[0122]

[0123] Wherein, DYEJ m is the first second-order form parameter of the m-th sampling point.

[0124] The second second-order form parameter is:

[0125]

[0126] Wherein, DEEJ m is the second second-order form parameter of the m-th sampling point.

[0127] The third second-order form parameter is:

[0128]

[0129] Among them, DSEJ m is the third second - order form parameter of the m - th sampling point.

[0130] The method for obtaining the surface curvature of the workpiece at the m - th sampling point includes:

[0131]

[0132] Among them, FMQL m is the surface curvature of the workpiece at the m - th sampling point.

[0133] The method for determining the corresponding moving step of the workpiece surface and the type of sub - workpiece surface based on the surface curvature of the workpiece at the m - th sampling point includes:

[0134] Preset a first threshold of the workpiece surface curvature and a second threshold of the workpiece surface curvature, where the first threshold of the workpiece surface curvature is less than the second threshold of the workpiece surface curvature; preset a first moving step of the workpiece surface, a second moving step of the workpiece surface, and a third moving step of the workpiece surface, where the first moving step of the workpiece surface is less than the second moving step of the workpiece surface, and the second moving step of the workpiece surface is less than the third moving step of the workpiece surface;

[0135] If the surface curvature of the workpiece at the m - th sampling point is greater than or equal to the second threshold of the workpiece surface curvature, then the corresponding type of sub - workpiece surface is the high - curvature region, and the corresponding moving step of the workpiece surface is the first moving step of the workpiece surface;

[0136] If the surface curvature of the workpiece at the m - th sampling point is greater than or equal to the first threshold of the workpiece surface curvature and less than the second threshold of the workpiece surface curvature, then the corresponding type of sub - workpiece surface is the medium - curvature region, and the corresponding moving step of the workpiece surface is the second moving step of the workpiece surface;

[0137] If the surface curvature of the workpiece at the m - th sampling point is less than the first threshold of the workpiece surface curvature, then the corresponding type of sub - workpiece surface is the low - curvature region, and the corresponding moving step of the workpiece surface is the third moving step of the workpiece surface.

[0138] It should be noted that during the machining process of the workpiece surface, since the workpiece surface is usually composed of multiple regions with different local curvatures, if the entire workpiece surface is directly optimized uniformly, it will lead to insufficiently fine path optimization in local regions, or even over - optimization, affecting the machining quality and efficiency. Therefore, by reasonably dividing the workpiece surface into regions, it can be ensured that each sub - workpiece surface adopts the most suitable machining strategy for its characteristics, thereby improving the accuracy and adaptability of the machining path optimization.

[0139] If the sub-region division of the surface of the workpiece to be processed is not carried out, the tool will frequently jump between high-curvature regions and low-curvature regions during the processing, resulting in an increase in the number of tool lifting operations and excessive idle strokes, thereby reducing the overall processing efficiency. By dividing the sub-surfaces of the workpiece, the tool can move continuously within regions of the same type, which not only reduces the switching of the tool path but also reduces the ineffective movements during the processing, thus improving the processing efficiency.

[0140] In addition, the division of the sub-surfaces of the workpiece also takes into account the smooth transition path between adjacent sub-surfaces, making the transition of the tool between different regions more natural, thereby avoiding abrupt path changes and improving the smoothness and stability of the overall processing path. In this way, not only can the processing accuracy be optimized, but also the errors during the processing can be effectively reduced, the processing quality of the surface of the workpiece can be improved, and the uniformity and consistency of the final processing effect can be ensured.

[0141] The second processing module is used to extract features from the N sub-surfaces of the workpiece to obtain the workpiece sub-surface feature data corresponding to the N sub-surfaces of the workpiece.

[0142] The method for obtaining the workpiece sub-surface feature data corresponding to the N sub-surfaces of the workpiece includes:

[0143] S200: Let the initial value of n be 1, and the value range of n is from 1 to N;

[0144] S201: Obtain the range of the workpiece surface parameter coordinates, the first first-order form parameter, the second first-order form parameter, and the third first-order form parameter of the nth sub-surface of the workpiece; calculate the corresponding area of the nth sub-surface of the workpiece according to the range of the workpiece surface parameter coordinates, the first first-order form parameter, the second first-order form parameter, and the third first-order form parameter; divide the boundary of the nth sub-surface of the workpiece into J boundary points, and calculate the boundary length of the nth sub-surface of the workpiece according to the coordinates of the J boundary points;

[0145] Obtain the workpiece surface curvature of the J boundary points by the same method as the workpiece surface curvature acquisition method; calculate the surface boundary complexity of the nth sub-surface of the workpiece according to the workpiece surface curvature of the J boundary points;

[0146] S202: Construct the corresponding workpiece sub-surface feature data from the area, boundary length, and surface boundary complexity of the nth sub-surface of the workpiece;

[0147] S203: Let n = n + 1. If n is less than or equal to N, continue to execute S201 to S202. If n is greater than N, end the current process.

[0148] The calculation method of the area of the sub-surface of the workpiece includes:

[0149]

[0150] Among them, QMMJ n is the sub-surface area of the workpiece for the nth boundary point, u n，min and v n，min are the lower limits of the workpiece surface parameter coordinates of the nth sub-surface of the workpiece, u n，max and v n，max are the upper limits of the workpiece surface parameter coordinates of the nth sub-surface of the workpiece; DYYJ n is the first-order form parameter of the nth boundary point, DEYJ n is the second-order form parameter of the nth boundary point, DSYJ n is the third-order form parameter of the nth boundary point.

[0151] The calculation method for the boundary length of the workpiece sub-surface includes:

[0152]

[0153] Among them, BJCD n is the boundary length of the nth sub-surface of the workpiece, x j+1 is the abscissa of the (j + 1)th boundary point, x j is the abscissa of the jth boundary point, y j+1 is the ordinate of the (j + 1)th boundary point, y j is the ordinate of the jth boundary point, z j+1 is the vertical coordinate of the (j + 1)th boundary point, y j is the vertical coordinate of the jth boundary point. What is calculated is the distance between the last point and the first point on the boundary of the workpiece sub-surface.

[0154] The calculation method for the complexity of the surface boundary includes:

[0155]

[0156] Among them, FZD n is the complexity of the surface boundary of the nth sub-surface of the workpiece, λ j is the coefficient of the jth boundary point, FMQL j is the curvature of the workpiece surface of the jth boundary point, and J is the number of boundary points.

[0157] The first tuning module adaptively tunes the idle stroke path of the workpiece surface machining for N workpiece sub-surface feature data.

[0158] It should be noted that optimizing the non-cutting movement of the tool for machining the surface of the workpiece during the air travel path, that is, reducing the non-cutting movement of the tool for machining the surface of the workpiece between each sub-surface of the workpiece, is the key to improving machining efficiency.

[0159] As Figure 4 shown, the method for adaptively optimizing the air travel path of the workpiece surface machining for N sub-surfaces of the workpiece based on the N sub-surface feature data of the workpiece includes:

[0160] Input the N sub-surface feature data of the workpiece into the machining path evaluation model to obtain H machining candidate paths for the workpiece; each machining candidate path for the workpiece includes the machining sequence of the N sub-surfaces of the workpiece; construct the H machining candidate paths for the workpiece into a machining path population;

[0161] For example, there are 5 sub-surfaces of the workpiece, and the corresponding numbers of the 5 sub-surfaces of the workpiece are 1, 2, 3, 4, and 5 respectively. There are 3 machining candidate paths for the door, which are 1→2→4→3→5, 2→1→5→4→3, and 5→1→4→3→2.

[0162] Evaluate the fitness of the H machining candidate paths for the workpiece in the machining path population to obtain the corresponding machining path fitness values;

[0163] Sort the H machining path fitness values in descending order, and select the top Q machining path fitness values to construct an elite population of machining paths;

[0164] Perform a mutation operation on the elite population of machining paths according to a preset method to obtain R mutated machining candidate paths for the workpiece and add them to the elite population of machining paths; overwrite the elite population of machining paths to the machining path population, and let H = Q + R;

[0165] Repeat the above operations. When the number of iterations reaches the preset maximum number of iterations, stop executing; select the machining candidate path with the largest machining path fitness value from the machining path population as the optimal solution for the air travel path of the workpiece surface machining.

[0166] It should be noted that the preset method refers to the specific method for performing a mutation operation on the elite population of machining paths. The following "method for obtaining R mutated machining candidate paths for the workpiece" details the specific implementation steps of the preset method.

[0167] For example, taking a ball valve as an example, a ball valve is a processed part that controls the opening and closing of a fluid. The core components of a ball valve include a rotatable sphere with a channel on it, and the fluid flow or blockage can be controlled by rotating the sphere by 90°. The processed part surfaces of the ball valve include multiple complex free-form surfaces such as the outer surface of the sphere, the seat sealing surface, and the flow path connection area. Due to the structural characteristics of the ball valve sphere, during numerical control machining, the cutting tool needs to move between multiple sub-surfaces with different curvatures. Especially when machining the contact area between the sphere and the seat, the cutting tool switching path needs to be optimized to reduce the idle stroke and improve the machining quality and efficiency.

[0168] The training method of the processing path evaluation model includes:

[0169] Pre-construct a processing path evaluation data set, which includes Y sets of processing path evaluation data and H candidate processing paths for the processed parts corresponding to the Y sets of processing path evaluation data. Y is a positive integer greater than 0. The processing path evaluation data includes N processed part sub-surface feature data; divide the processing path evaluation data set into a processing path evaluation data training set and a processing path evaluation data validation set, where the processing path evaluation data training set is used for parameter learning of the processing path evaluation model, and the processing path evaluation data validation set is used to evaluate the generalization ability of the processing path evaluation model in real time;

[0170] During the training process of the processing path evaluation model, a deep neural network structure based on a multi-layer perceptron is adopted. The processing path evaluation data is converted into a feature vector as the input, and the non-linear features in the data are extracted through multiple hidden layers. Finally, the softmax activation function is used in the output layer to generate the probability distribution of the H candidate processing paths for the processed parts, and the H candidate processing paths corresponding to the maximum probability are output as the final prediction result; the training process aims to minimize the cross-entropy loss function, and at the same time, an early stopping strategy is introduced to monitor the performance of the processing path evaluation data validation set. When the prediction accuracy rate on the processing path evaluation data validation set reaches the preset threshold, it is considered that the processing path evaluation model has converged, and the training stops immediately.

[0171] The method for obtaining R mutant candidate processing paths for the processed parts includes:

[0172] Preset a path fitness threshold one and a path fitness threshold two, where the path fitness threshold one is less than the path fitness threshold two; divide the processing path elite population according to the path fitness threshold one and the path fitness threshold two to obtain a high fitness path set, a medium fitness path set, and a low fitness path set;

[0173] Randomly cross with each other in the high fitness path set to obtain NUM1 candidate processing paths for the processed parts and add them to the high fitness path set;

[0174] Randomly cross the machining candidate paths of the workpieces in the medium fitness path set with the machining candidate paths of the workpieces in the high fitness path set to obtain NUM2 machining candidate paths of workpieces and add them to the medium fitness path set; perform random cross among the machining candidate paths in the medium fitness path set to obtain NUM3 machining candidate paths of workpieces and add them to the medium fitness path set;

[0175] Randomly cross the machining candidate paths of the workpieces in the low fitness path set with the machining candidate paths of the workpieces in the high fitness path set to obtain NUM4 machining candidate paths of workpieces and add them to the low fitness path set; perform random cross among the machining candidate paths in the low fitness path set to obtain NUM5 machining candidate paths of workpieces and add them to the low fitness path set;

[0176] When NUM1 + NUM2 + NUM3 + NUM4 + NUM5 = R, stop the above random cross operation to obtain R mutant machining candidate paths of workpieces.

[0177] The methods for obtaining the high fitness path set, medium fitness path set, and low fitness path set include:

[0178] Construct the high fitness path set with the machining candidate paths of workpieces in the machining path elite population whose machining path fitness values are greater than or equal to the second path fitness threshold; construct the medium fitness path set with the machining candidate paths of workpieces in the machining path elite population whose machining path fitness values are greater than or equal to the first path fitness threshold and less than the second path fitness threshold; construct the low fitness path set with the machining candidate paths of workpieces in the machining path elite population whose machining path fitness values are less than the first path fitness threshold.

[0179] The method for obtaining the machining path fitness value includes:

[0180] Input H machining candidate paths of workpieces and the corresponding workpiece sub - surface feature data into the fitness evaluation model to obtain the corresponding machining path fitness values.

[0181] The training method of the fitness evaluation model includes:

[0182] Pre - collect a fitness evaluation data set, the fitness evaluation data set includes P groups of fitness evaluation data and the machining path fitness values corresponding to the P groups of fitness evaluation data, P is a positive integer greater than 0, and the fitness evaluation data includes machining candidate paths of workpieces and workpiece sub - surface feature data; divide the fitness evaluation data set into a training set and a validation set, where the training set is used to train the fitness evaluation model, and the validation set is used to evaluate the generalization performance of the fitness evaluation model;

[0183] During the training process of the fitness evaluation model, minimizing the cross-entropy loss function is used as the optimization objective. The early stopping strategy is used to monitor the performance of the validation set. By continuously adjusting the network parameters, the model performance is optimized. When the prediction accuracy on the validation set reaches the expected accuracy, it is considered that the fitness evaluation model has converged and the training is stopped. The fitness evaluation model is trained using a deep neural network based on a multi-layer perceptron.

[0184] The fitness evaluation data is converted into feature vectors. The input layer of the fitness evaluation model receives the feature vectors, extracts the non-linear relationships in the data through several hidden layers. Finally, the output layer of the fitness evaluation model calculates the probability distribution of the machining path fitness values through the softmax activation function, and outputs the machining path fitness value corresponding to the maximum probability as the final prediction result.

[0185] Embodiment 2

[0186] Please refer to Figure 3 As shown, this embodiment provides a process optimization system for complex curved surfaces, further including:

[0187] A second tuning module adaptively tunes the smoothness of the curved surface machining paths of N sub-curved surfaces of the workpieces based on the empty travel path after adaptive tuning.

[0188] The method for adaptively tuning the smoothness of the curved surface machining paths of N sub-curved surfaces of the workpieces based on the empty travel path after adaptive tuning includes:

[0189] S300: Obtain the machining order of the sub-curved surfaces of the workpieces corresponding to the empty travel path after adaptive tuning; let the initial value of sx be 1, and the value range of sx is from 1 to N;

[0190] S301: If sx is less than or equal to N - 1, obtain the boundary region between the sx-th and the (sx + 1)-th sub-curved surfaces of the workpieces from the machining order of the sub-curved surfaces of the workpieces; if sx is equal to N, obtain the boundary region between the sx-th and the 1st sub-curved surfaces of the workpieces from the machining order of the sub-curved surfaces of the workpieces;

[0191] Set the initial position and the end position of the tool for the boundary region; let the number of moves from the initial position of the tool to the end position of the tool be T times; initialize the DDYZ tool full - travel paths, calculate the smoothness of the surface machining paths corresponding to each tool full - travel path, and record the smoothness of each single - tool move at the same time. Add the tool single - move paths with the smoothness of a single - tool move greater than the single - tool move smoothness threshold to the single - move path set. Iteratively optimize the single - move paths in the single - move path set and the single - move paths not in the single - move path set according to the preset number of iterations to obtain the preset number of tool move paths, and calculate the corresponding smoothness of the surface machining paths;

[0192] Select the tool move path with the maximum smoothness of the surface machining path as the corresponding self - adaptive tuning tool move path;

[0193] S302: Let sx = sx + 1. If sx is less than or equal to SX, continue to execute S301. If sx is greater than SX, end the current process.

[0194] The method for obtaining the smoothness of the surface machining path includes:

[0195]

[0196] Among them, PHX ((sx，sx+1)，ddyz) is the smoothness of the ddyz - th surface machining path corresponding to the boundary region between the sx - th and the (sx + 1) - th sub - surface of the workpiece. The value range of ddyz is from 1 to DDYZ. The larger PHX ((sx，sx+1)，ddyz) , the better the smoothness of the surface machining path, and the more suitable the corresponding tool move path; FMQL (ddyz，t+1) is the curvature of the workpiece surface corresponding to the (t + 1) - th tool move position of the ddyz - th tool move path, FMQL (ddyz，t) is the curvature of the workpiece surface corresponding to the t - th tool move position of the ddyz - th tool move path, YDJL (ddyz，t) is the distance of the t - th tool move of the ddyz - th tool move path, DJJD (ddyz，t) is the angle of the t - th tool move of the ddyz - th tool move path.

[0197] It should be noted that when the tool for machining the workpiece surface switches from one sub - surface of the workpiece to another sub - surface of the workpiece, problems such as trajectory mutation or sharp tool turning will occur. Therefore, it is necessary to perform self - adaptive tuning on the smoothness of the surface machining path to ensure the smooth connection of the surface machining path, avoid sudden acceleration or deceleration, reduce the idle stroke, and improve the machining quality of the workpiece surface.

[0198] Embodiment 3

[0199] Please refer to Figure 2 As shown, this embodiment provides a process optimization method for complex surfaces, which is implemented based on a process optimization system for complex surfaces, including:

[0200] Obtain the three-dimensional structure diagram of the surface of the workpiece to be machined;

[0201] Divide the surface of the workpiece to be machined based on the three-dimensional model of the surface of the workpiece to be machined, and obtain N sub-surfaces of the workpiece to be machined;

[0202] Extract features from the N sub-surfaces of the workpiece to be machined to obtain the machining feature data of the sub-surfaces of the workpiece corresponding to the N sub-surfaces of the workpiece to be machined;

[0203] Based on the machining feature data of the N sub-surfaces of the workpiece to be machined, adaptively optimize the idle stroke path of the machining of the N sub-surfaces of the workpiece to be machined.

Claims

1. A process optimization method for complex curved surfaces, characterized in that, Including: Obtain the three-dimensional structure diagram of the surface of the workpiece to be machined; Based on the three-dimensional model of the surface of the workpiece to be machined, divide the surface of the workpiece to be machined to obtain N sub-surfaces of the workpiece to be machined; Extract features from the N sub-surfaces of the workpiece to be machined to obtain the feature data of the sub-surfaces of the workpiece to be machined corresponding to the N sub-surfaces of the workpiece to be machined; Based on the feature data of the N sub-surfaces of the workpiece to be machined, adaptively optimize the idle travel path of the machining of the surface of the workpiece of the N sub-surfaces of the workpiece to be machined.

2. The process optimization method for complex curved surfaces according to claim 1, wherein The method for adaptively optimizing the idle travel path of the machining of the surface of the workpiece of the N sub-surfaces of the workpiece to be machined based on the feature data of the N sub-surfaces of the workpiece to be machined includes: Input the feature data of the N sub-surfaces of the workpiece to be machined into the machining path evaluation model to obtain H candidate machining paths for the workpiece; each candidate machining path for the workpiece includes the machining sequence of the N sub-surfaces of the workpiece; construct the H candidate machining paths for the workpiece into a machining path population; Evaluate the fitness of the H candidate machining paths for the workpiece in the machining path population to obtain the corresponding machining path fitness values; Sort the H machining path fitness values in descending order, and select the first Q machining path fitness values to construct an elite machining path population; Perform mutation operations on the elite machining path population according to a preset method to obtain R candidate machining paths for the mutated workpiece and add them to the elite machining path population; overwrite the machining path population with the elite machining path population, and set H = Q + R; Repeat the above operations. When the number of iterations reaches the preset maximum number of iterations, stop the execution; select the candidate machining path for the workpiece with the maximum machining path fitness value from the machining path population as the optimal solution of the idle travel path for the machining of the surface of the workpiece.

3. A process optimization method for complex curved surfaces according to claim 2, characterized in that, The method for obtaining the R candidate machining paths for the mutated workpiece includes: Preset a first path fitness threshold and a second path fitness threshold, where the first path fitness threshold is less than the second path fitness threshold; divide the elite machining path population according to the first path fitness threshold and the second path fitness threshold to obtain a high fitness path set, a medium fitness path set, and a low fitness path set; Perform random crossover in the high fitness path set to obtain NUM1 candidate machining paths for the workpiece and add them to the high fitness path set; Perform random crossover between the candidate machining paths for the workpiece in the medium fitness path set and the candidate machining paths for the workpiece in the high fitness path set to obtain NUM2 candidate machining paths for the workpiece and add them to the medium fitness path set; perform random crossover in the medium fitness path set to obtain NUM3 candidate machining paths for the workpiece and add them to the medium fitness path set; Perform random crossover between the candidate machining paths for the workpiece in the low fitness path set and the candidate machining paths for the workpiece in the high fitness path set to obtain NUM4 candidate machining paths for the workpiece and add them to the low fitness path set; perform random crossover in the low fitness path set to obtain NUM5 candidate machining paths for the workpiece and add them to the low fitness path set; When NUM1 + NUM2 + NUM3 + NUM4 + NUM5 = R, stop the above random crossover operation to obtain R candidate machining paths for the mutated workpiece.

4. A process optimization method for complex curved surfaces according to claim 3, characterized in that The methods for obtaining the high fitness path set, medium fitness path set, and low fitness path set include: Construct the high fitness path set with the machining candidate paths of the workpieces in the machining path elite population whose machining path fitness values are greater than or equal to the second path fitness threshold; construct the medium fitness path set with the machining candidate paths of the workpieces in the machining path elite population whose machining path fitness values are greater than or equal to the first path fitness threshold and less than the second path fitness threshold; construct the low fitness path set with the machining candidate paths of the workpieces in the machining path elite population whose machining path fitness values are less than the first path fitness threshold.

5. The process optimization method for complex curved surfaces according to claim 4, characterized in that, The methods for obtaining N sub-surfaces of the workpieces include: S100: Randomly distribute M sampling points on the three-dimensional structure diagram of the workpiece surface; S101: Let the initial value of m be 1, and the value range of m is from 1 to M; let the initial value of n be 1, and n is a counting variable; S102: Obtain the workpiece surface parameter coordinates of the m-th sampling point, and obtain the corresponding workpiece surface parameter equation based on the workpiece surface parameter coordinates; calculate the workpiece surface curvature of the m-th sampling point based on the workpiece surface parameter equation; S103: Determine the corresponding workpiece surface movement step and workpiece sub-surface type based on the workpiece surface curvature of the m-th sampling point; perform a diffusion operation according to the workpiece surface movement step and workpiece sub-surface type to obtain the workpiece surface diffusion region of the m-th sampling point; S104: Use the edge points of the workpiece surface diffusion region of the m-th sampling point as the new diffusion starting point, and repeat S103. When there are no longer points with the same workpiece sub-surface type as the m-th sampling point during the diffusion process, the diffusion stops, and the workpiece sub-surface of the n-th sampling point is obtained; let n = n + 1, and execute S105; S105: Let m = m + 1. If m is less than or equal to M, then execute S106; if m is greater than M, then end the current process, assign n to N, and obtain N workpiece sub-surfaces; S106: If the m-th sampling point is already included in the constructed workpiece sub-surfaces, then execute S105; if the m-th sampling point is not included in the constructed workpiece sub-surfaces, then execute S102 to S104.

6. The process optimization method for complex curved surfaces according to claim 5, wherein The method for obtaining the workpiece surface diffusion region of the m-th sampling point includes: With the m-th sampling point as the center and the workpiece surface movement step as the radius, form the diffusion region to be processed of the m-th sampling point; calculate the workpiece surface curvature of the points on the edge of the diffusion region to be processed, denoted as the edge surface curvature, and obtain the corresponding workpiece sub-surface type according to the edge surface curvature; Mark the points on the edge of the diffusion region to be processed with the same workpiece sub-surface type as the m-th sampling point as the connection points to be connected of the m-th sampling point; connect all the connection points to be connected of the m-th sampling point with the m-th sampling point to form the workpiece surface diffusion region of the m-th sampling point.

7. The process optimization method for complex curved surfaces according to claim 6, wherein, The method for obtaining the workpiece surface curvature of the m-th sampling point includes: The parametric equation of the surface of the workpiece with respect to u m Take the first-order partial derivative to obtain the first first-order partial derivative; the parametric equation of the surface of the workpiece with respect to v m Take the first-order partial derivative to obtain the second first-order partial derivative; u m and v m are the parametric coordinates of the surface of the workpiece at the m-th sampling point, and the unit normal vector at the m-th sampling point is calculated based on the first first-order partial derivative and the second first-order partial derivative; Multiply the first first-order partial derivative by the first first-order partial derivative to obtain the corresponding first first-order formal parameter; multiply the first first-order partial derivative by the second first-order partial derivative to obtain the corresponding second first-order formal parameter; multiply the second first-order partial derivative by the second first-order partial derivative to obtain the corresponding third first-order formal parameter; The first first-order partial derivative with respect to u m Find the second-order partial derivative to obtain the first second-order partial derivative; the first first-order partial derivative with respect to v m Find the second-order partial derivative to obtain the second second-order partial derivative; the second first-order partial derivative with respect to v m Find the second-order partial derivative to obtain the third second-order partial derivative; Multiply the first second-order partial derivative by the unit normal vector to obtain the first second-order formal parameter of the m-th sampling point; multiply the second second-order partial derivative by the unit normal vector to obtain the second second-order formal parameter of the m-th sampling point; multiply the third second-order partial derivative by the unit normal vector to obtain the third second-order formal parameter of the m-th sampling point; Calculate the curvature of the workpiece surface at the m-th sampling point based on the first first-order formal parameter, the second first-order formal parameter, the third first-order formal parameter, the first second-order formal parameter, the second second-order formal parameter, and the third second-order formal parameter.

8. A process optimization method for complex curved surfaces according to claim 7, characterized in that The method for determining the corresponding movement step of the workpiece surface and the type of workpiece sub-surface based on the curvature of the workpiece surface at the m-th sampling point includes: Preset a first threshold value of the curvature of the workpiece surface and a second threshold value of the curvature of the workpiece surface, where the first threshold value of the curvature of the workpiece surface is less than the second threshold value of the curvature of the workpiece surface; preset a first movement step of the workpiece surface, a second movement step of the workpiece surface, and a third movement step of the workpiece surface, where the first movement step of the workpiece surface is less than the second movement step of the workpiece surface, and the second movement step of the workpiece surface is less than the third movement step of the workpiece surface; If the curvature of the workpiece surface at the m-th sampling point is greater than or equal to the second threshold value of the curvature of the workpiece surface, then the type of the workpiece sub-surface is a high-curvature region, and the movement step of the workpiece surface is the first movement step of the workpiece surface; If the curvature of the workpiece surface at the m-th sampling point is greater than or equal to the first threshold value of the curvature of the workpiece surface and less than the second threshold value of the curvature of the workpiece surface, then the type of the workpiece sub-surface is a medium-curvature region, and the movement step of the workpiece surface is the second movement step of the workpiece surface; If the curvature of the workpiece surface at the m-th sampling point is less than the first threshold value of the curvature of the workpiece surface, then the type of the workpiece sub-surface is a low-curvature region, and the movement step of the workpiece surface is the third movement step of the workpiece surface.

9. The process optimization method for complex curved surfaces according to claim 8, characterized in that The method for obtaining the workpiece sub-surface feature data corresponding to N workpiece sub-surfaces includes: S200: Let the initial value of n be 1, and the value range of n is from 1 to N; S201: Obtain the coordinate range of the workpiece surface parameters, the first first-order formal parameter, the second first-order formal parameter, and the third first-order formal parameter of the n-th workpiece sub-surface; calculate the corresponding area of the n-th workpiece sub-surface according to the coordinate range of the workpiece surface parameters, the first first-order formal parameter, the second first-order formal parameter, and the third first-order formal parameter; divide the boundary of the n-th workpiece sub-surface into J boundary points, and calculate the boundary length of the n-th workpiece sub-surface according to the coordinates of the J boundary points; Obtain the curvature of the workpiece surface of the J boundary points by the same method as the method for obtaining the curvature of the workpiece surface; calculate the surface boundary complexity of the n-th workpiece sub-surface according to the curvature of the workpiece surface of the J boundary points; S202: Construct the workpiece sub - surface feature data corresponding to the workpiece sub - surface area, boundary length, and surface boundary complexity of the nth workpiece sub - surface; S203: Let n = n + 1. If n is less than or equal to N, continue to execute S201 to S202. If n is greater than N, end the current process.

10. A process optimization method for complex curved surfaces according to claim 9, characterized in that, The training method of the machining path evaluation model includes: Pre - construct a machining path evaluation data set, which includes Y sets of machining path evaluation data and H machining candidate paths corresponding to the Y sets of machining path evaluation data. Y is a positive integer greater than 0. The machining path evaluation data includes N workpiece sub - surface feature data; divide the machining path evaluation data set into a machining path evaluation data training set and a machining path evaluation data validation set. The machining path evaluation data training set is used for parameter learning of the machining path evaluation model, and the machining path evaluation data validation set is used to evaluate the generalization ability of the machining path evaluation model in real - time; During the training process of the machining path evaluation model, adopt a deep neural network structure based on a multi - layer perceptron. Convert the machining path evaluation data into feature vectors as inputs, extract non - linear features in the data through multiple hidden layers, and finally generate the probability distribution of H machining candidate paths of the workpiece using the softmax activation function in the output layer. Output the H machining candidate paths corresponding to the maximum probability as the final prediction result; the training process aims to minimize the cross - entropy loss function, and at the same time introduce an early - stopping strategy to monitor the performance of the machining path evaluation data validation set. When the prediction accuracy on the machining path evaluation data validation set reaches the preset threshold, it is considered that the machining path evaluation model has converged, and the training stops immediately.

Images