A method for time sequence planning of numerical control machining trajectory based on a directed hybrid graph

Abstract

The application discloses a kind of based on the time sequence planning method of numerical control machining track of guide mixed graph.The present application comprises the following steps: firstly, the profile and filling track on machining surface are classified and defined, and the heuristic algorithm using the principle of nearest neighbor is solved preliminary connection track;Then, construct guide mixed graph model, so that it can represent the time sequence travel and connection relationship of complex machining surface, and the connection order is reconstructed using depth-first algorithm and local edge exchange;Finally, the position of each profile or filling sub-region connection point is locally evolved, and the final jump connection track scheme in profile and filling process is obtained.The present application classifies different machining tracks, reconstructs transition time sequence and determines specific guide mixed graph objective function, reduces the length and number of track jumps, improves machining efficiency while reducing mechanical vibration of machining.The present application can be widely applied to numerical control forming surface track planning, and has a positive effect on improving the machining efficiency and workpiece quality of multi-gap metamaterials.

Description

TECHNICAL FIELD

[0001] The present application relates to a shaped trajectory planning method in the field of advanced manufacturing technology, in particular to a numerical control machining trajectory timing planning method based on a guided hybrid graph. BACKGROUND

[0002] In traditional numerical control machine tool machining, such as milling machine, lathe machining, etc., tool path optimization is an important field that is increasingly concerned in manufacturing industry, and is an important consideration for top-level conceptual design or scheme evolution in mechanical industry software. Tool path optimization is a comprehensive optimization technology that takes machining efficiency as the main target, and considers precision, process and resource utilization, etc. fields. At the same time, the changes of modern manufacturing industry have also been constantly promoting the direction of tool path optimization based on numerical control machine tools. For example, the emergence of complex workpieces, high-precision machining and large-scale customization requirements requires the tool to minimize the generation of jump trajectories while machining at high speed, in order to ensure precision and manufacturing efficiency at the same time.

[0003] At the same time, since the 1980s of the 19th century, 3D printing (3D Printing) or additive manufacturing (Additive Manufacturing, AM) technology has been developing, which is an advanced manufacturing and shaping process based on multi-layer accumulation manufacturing method. For commonly used additive manufacturing processes based on line scanning, such as fused deposition modeling (Fused deposition modeling, FDM), fused filament fabrication (Fused Filament Fabrication, FFF), selective laser melting (Selective laser melting, SLM), selective laser sintering (Selective laser sintering, SLS), stereolithography (Stereolithography, SLA), etc., the tool (laser or actuator) will scan and print along the pre-designed shaping trajectory and make the material solidify and shape. However, when printing complex multi-void metamaterials in the field of biomedicine, or large components in the field of aerospace, the scanning-accumulation process has a long machining time and low shaping efficiency. The switching process of non-machining jump trajectory and actual machining travel often increases the shaping time and leaves residues on the workpiece, affecting the manufacturing quality. Therefore, the layer-by-layer trajectory optimization of additive manufacturing can effectively increase the printing efficiency and printing quality.

[0004] Currently, the common NC machining trajectory optimization is mainly divided into two directions. One is to optimize the total length of the forming trajectory. Enough points or line segments are taken in the material forming area, and how to connect and traverse them is considered. The Traveling Salesman Problem model or the Rural Postman Problem model is established. The total length of the forming trajectory is taken as the objective function. After adding different constraint conditions, the heuristic algorithm is used to obtain the optimal or approximate optimal solution. This method reduces the total length of forming, but due to the huge number of points / arc solutions, it is usually necessary to balance between solving time and optimization effect. In addition, due to the combination of machining travel and non-machining jump trajectory, the random appearance of empty travel-machining travel switching and the residual position that cannot be positioned in advance may greatly affect the final manufacturing efficiency and quality. The other is to classify different forming steps and different forming regions first, and then optimize the solving process. The forming objects of each layer are classified into different task units such as contour, filling area, and non-connected area. The starting point and the connection point between different manufacturing units are controlled, and optimization is performed respectively. This strategy ensures the manufacturing quality (controllable residual) while avoiding long algorithm solving time and providing more flexible parameter adjustment space. At present, for the forming objects with multiple connected domains, a hybrid forming strategy of contour offset and filling can be used. By determining the position of the contour starting point (Shell sp (Shell sp .x,Shell sp .y), the transition timing of each contour is determined.

[0005] Some scholars have proposed corresponding optimization algorithms for this kind of line scanning shaping planning problem. For example, researchers at the U.S. Oak Ridge National Laboratory (Thompson B, Yoon H-S. Efficient path planning algorithm for additive manufacturing systems [J]. IEEE Transactions on components, packaging and manufacturing technology, 2014, 4(9): 1555-1563) developed a trajectory planning algorithm optimized for XY motion platforms to meet the needs of aerosol constant speed shaping, thereby minimizing material waste. Researchers at the Massachusetts Institute of Technology introduced the application of graph theory model in shaping trajectory planning, and proposed a method of single line wide grid filling shaping manufacturing based on RPP model, which optimized the shaping trajectory on the given grid. Researchers from the University of UniMoRe in Italy (Iori M, Novellani S. Optimizing the nozzle path in the 3D printing process [C] / / proceedings of the Design Tools and Methods in Industrial Engineering 2019. Modena, Italy: Springer, 2020: 912-924) used different integer linear programming and heuristic algorithms to optimize the shaping trajectory, and compared the differences in calculation time and optimization results between different algorithms.

[0006] U.S. Patent (Nomura T, Dede E M. Method of tool path generation for additive manufacturing with vector distribution [Z]. Google Patents. 2021) provides an optimization method for fiber composite part shaping manufacturing, which optimizes the model through structural mechanics algorithm and provides topologically optimized fiber composite part design, realizing anisotropic optimization trajectory.

[0007] US patent (Lewicki J, Compel W, Tortorelli D, et al. Optimal toolpath generation system and method for additively manufactured composite materials [Z]. Google Patents. 2021) uses an optimization subsystem and a level set function profile to optimize tool path for each layer of parts to control the movement of the printed assembly in the layer-by-layer additive manufacturing process, thereby maximizing the efficiency and accuracy of part manufacturing.

[0008] US patent (Kniola R, D. Method for calculating a path in additive manufacturing [Z]. Google Patents. 2019) proposes a method for Hilbert curve to calculate filling tracks and analyze profiles to obtain reasonable forming tracks.

[0009] These methods have their own advantages and application scope, but still have some limitations, especially in the planning of profile and filling / surface machining tracks, and there is no good distinction and combination between them, especially in the forming and machining of porous materials or complex hole surfaces, it is difficult to achieve better optimization results and faster calculation efficiency, and there is also a lack of numerical control machining track timing planning method supporting problem reconstruction and scheme evolution. SUMMARY

[0010] In order to solve the problems in the background art, in order to improve the machining or forming efficiency in the numerical control machining process, so as to further improve the production efficiency of the workpiece. The present application provides a numerical control machining track timing planning method based on a guide mixed graph. For the surface machining of numerical control forming and the tool movement process of additive manufacturing of porous materials, a machining track optimization algorithm oriented to reduce jump tracks is realized, which has the advantages of high efficiency and good optimization results.

[0011] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0012] Step one: after classifying the original data point set of the workpiece to be machined, the original profile data corresponding to each workpiece surface to be machined of the workpiece to be machined is obtained, and the initial workpiece surface to be machined is determined;

[0013] Step two: according to the original profile data of the current workpiece surface to be machined, the profile machining track of the current workpiece surface to be machined is generated in combination with the guide mixed graph;

[0014] Step three: based on the minimum jump trajectory of the current to-be-processed surface in profile processing and the original profile data of the current to-be-processed surface, a surface processing trajectory of the current to-be-processed surface is generated in combination with the guide hybrid graph;

[0015] Step four: profile processing is performed on the current to-be-processed surface according to the profile processing trajectory, and surface processing is performed on the to-be-processed surface after profile processing according to the surface processing trajectory, thereby completing the processing of the current to-be-processed surface.

[0016] Step five: repeating steps two to four, the remaining to-be-processed surfaces are processed, thereby completing the processing of the to-be-processed part.

[0017] The to-be-processed surface is specifically a processing layer in 3D printing or a processing surface in a numerical control machine tool.

[0018] In step one, the original profile data of each to-be-processed surface is composed of a set of processing outer contour points and a set of internal hole contour points of each unconnected domain.

[0019] The step two is specifically:

[0020] Step 2.1: based on the profile processing starting point Shell sp and the corresponding original profile data, the profile processing sequence of the current to-be-processed surface and the starting point and the ending point corresponding to each contour point set are preliminarily solved.

[0021] Step 2.2: based on the profile processing starting point Shell sp and the starting point of each contour point set, a first guide hybrid graph model is established, the first guide hybrid graph model is used to optimize the current profile processing sequence, and a new profile processing sequence is obtained.

[0022] Step 2.3: according to the profile processing starting point Shell sp , the new profile processing sequence and the starting point position of the front and rear contour point sets, the starting points of each contour point set are optimized, the minimum jump trajectory in profile processing is obtained and used as the profile processing trajectory of the current to-be-processed surface.

[0023] In step 2.1, a heuristic method combining the nearest neighbor principle is used to preliminarily solve the profile processing sequence of the current to-be-processed surface and the starting point and the ending point corresponding to each contour point set.

[0024] In step 2.2, the Euclidean distance between different contour point sets is used as the cost to establish an undirected graph, and a directed graph is established between the profile processing starting point and all contour point sets, thereby obtaining the first guide hybrid graph model.

[0025] The step 2.3 is specifically as follows: for each contour point set, the distance sum between each point in the contour point set and the start point of the two adjacent contour point sets before and after is calculated, and the point with the minimum distance sum in the contour point set is taken as the start point of the contour point set.

[0026] The step three is specifically as follows:

[0027] Firstly, the end point of the contour machining track is taken as the start point Fill sp of the surface machining track, then after the current surface to be machined is segmented, the end points of each segmented surface and the corresponding segmented sub-track are obtained; the start point Fill sp of the surface machining track and the end points of each segmented sub-track are used to establish a second directed hybrid graph model, then the transition timing of each segmented sub-track is optimized by using the second directed hybrid graph model, the jump track in surface machining is obtained and taken as the surface machining track of the current surface to be machined.

[0028] In the step three, the Euclidean distance between the end points of each segmented sub-track is taken as the cost to establish an undirected graph, and a directed graph is established among the start point Fill sp of the surface machining track and the end points of each segmented sub-track, so as to obtain the second directed hybrid graph model.

[0029] The present application has the beneficial effects that:

[0030] The present application adopts the idea of classifying and optimizing the machining object, reduces the tool jump track in the numerical control forming surface machining and the additive manufacturing layer deposition process, is suitable for the surface machining and overall forming manufacturing of complex porous materials, uses the heuristic gradient descent algorithm process to optimize the machining track of the manufacturing process, and improves the manufacturing and forming efficiency of products. BRIEF DESCRIPTION OF DRAWINGS

[0031] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the following drawings:

[0032] Figure 1 is a method flow diagram of the present application.

[0033] Figure 2 is a structure diagram of a V8 fuel engine cylinder body of an embodiment of the present application.

[0034] Figure 3 is a layered connected domain diagram of a workpiece to be machined of the present application.

[0035] Figure 4 is a directed hybrid graph model diagram of a 0.50 height cross-sectional view of the present application.

[0036] Figure 5is a directed hybrid graph model diagram of the 0.55 height cross-section of the present invention.

[0037] Figure 6 is a directed hybrid graph model diagram of the 0.65 height cross-section of the present invention.

[0038] Figure 7 is the preliminary solution of the contour trajectory using the heuristic algorithm of the nearest neighbor principle of the present invention.

[0039] Figure 8 is the optimization result of the contour transition timing using the directed hybrid graph model of the present invention.

[0040] Figure 9 is the iterative curve graph of the optimization using the directed hybrid graph model of the present invention.

[0041] Figure 10 is the multiple local evolution result of each contour starting point of the present invention.

[0042] Figure 11 is the iterative curve graph of the multiple local evolution result of the present invention.

[0043] Figure 12 is the combination of the contour trajectory optimization and the filling trajectory result of the present invention.

[0044] Figure 13 is the comparison and presentation of the results of the multiple optimization processes of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with the accompanying drawings and examples.

[0046] As shown in Figure 1 , the present invention comprises the following steps:

[0047] Step one: after classifying the original data point set of the workpiece to be processed, the original contour data corresponding to each work surface to be processed of the workpiece to be processed is obtained, and the initial work surface to be processed of the workpiece to be processed is determined;

[0048] The work surface to be processed is specifically a processing layer in 3D printing or a processing surface in a numerical control machine tool. In 3D printing, the original contour data corresponding to the work surface to be processed is recorded as a slice contour; in a numerical control machine tool, the original contour data corresponding to the work surface to be processed is recorded as a processing surface contour. In specific implementation, when the workpiece to be processed is a 3D printed part, the 3D printed part needs to be sliced by a manifold layer cutting method before classification.

[0049] In step one, the original contour data of each work surface to be processed is composed of the processing outer contour point set and the internal hole contour point set of each unconnected domain, recorded as {NCP1 1 ,..., NCP1 j, ..., NCP1 n}, NCP i j Let j represent the set of contour points of surface i to be processed. Determine the number n of the original contours for this layer. lc And construct the original contour set LC = {LC1, LC2, ..., LC} n} and the j-th contour LC j vertex set

[0050] like Figure 2 Front view (a) and Figure 2 The rear view (b) shows the non-mesh model of the embodiment of the present invention, a V8 fuel engine cylinder block model. The V-type cylinder has a 90° angle layout, a side-mounted valve train, and a reduced main cylinder block height. The crankshaft axis is denoted as the longitudinal x-direction (shown by the dotted line). The length-width-height (xyz) ratio is 2.3978:1.7654:1, the total specific surface area is 0.2334 (mme-1), and the center of gravity is located at (45.0187%, 49.7875%, 48.3774%) of the length, width, and height.

[0051] like Figure 3 The diagram shown is a layered connected domain diagram of the workpiece to be processed according to the present invention, wherein... Figure 3 Figure (a) shows the connected domain of the 3D manifold model at a height of 0.50, with a fill rate of 72.0833% and an area of ​​7.5910 per unit length. The centroid is located at (83.1869, 101.2570) (indicated by *), accounting for (49.4850%), 44.3519%. The eccentricity of the largest ellipse is 0.7073, indicating that the angle between the cylindrical cylinder bore and the horizontal plane is 45.0160°. Figure 3 (b) shows the connected domain of the 3D manifold model at a height of 0.55, with a fill rate of 59.3807% and an area of ​​4.7154 per unit length. The centroid is located at (83.0697, 102.4304) (indicated by *), accounting for (49.3983%, 44.8658%). Figure 3 Figure (c) shows the connected components of the 3D manifold model at a height of 0.65, with a fill rate of 43.0666% and an area of ​​4.4529 per unit length. The centroid is located at (82.3325, 107.0219) (indicated by *), occupying a percentage of (48.8005%, 48.5743%). Figure 3Taking (b) as an example with a height of 0.55, the optimization and solution of the layered CNC trajectory are performed. The area of ​​the outer contour is 30716.1698, and the areas of the remaining 54 inner contours, from largest to smallest, are 1339.83895598, 1334.95451000, 850.52647045, 850.52239938, 850.51699750, 850.49532283, 850.49378837, 850.48210431, 850.48099106, 850.48030525, and 124.593284. 13, 124.44135408, 124.44122068, 124.44120261, 124.44102726, 121.76299335, 120.33500230, 120.33493886, 120.33492584, 120.33492553, 120.33492357, 120.33491751, 120.33489819, 120.33486264, 120.334 85629, 120.33484792, 120.33484698, 120.33484415, 120.33484178, 120.33479282, 120.33478904, 120.33340061, 61.75691647, 49.05927185, 13.53738096, 13.53737696, 13.53737283, 13.53736211, 13.537359 08, 13.53735822, 13.53732560, 13.53731096, 13.53730163, 13.53729835, 13.53729487, 13.53729036, 13.53728371, 13.53728217, 13.53727024, 13.53725516, 0.25310612, 0.25310603, 0.25309502, 0.25309075. Note: Area units are the square of the length unit, the same applies below.

[0052] like Figure 4 The diagram shown is a guide hybrid diagram model of the 0.50 height cross-sectional view of the present invention, with coordinates [125, 180] as its CNC machining starting point. The points on the contour are bidirectionally connected, while the points on the contour are unidirectionally connected to the machining starting point. Taking this diagram as an example, the generated guide hybrid diagram contains 28 nodes, 28 unidirectional edges, and 378 bidirectional edges (the bidirectional edges are somewhat omitted in the diagram for clarity).

[0053] like Figure 5The diagram shown is a guide hybrid diagram model of the middle part (53≤x≤113 and 10≤y≤185) in the cross-sectional view of the 0.55 height of this invention. The coordinates [120, 100] are taken as its CNC machining starting point. The points of the contour are bidirectionally connected, while the points of the contour are unidirectionally connected to the machining starting point. Taking this diagram as an example, the generated guide hybrid diagram contains 20 nodes, 20 unidirectional edges, and 190 bidirectional edges.

[0054] like Figure 6 The diagram shown is a guide hybrid diagram model of the middle part of the 0.65 height cross-sectional view of the present invention. The coordinates [75, 75] are taken as its CNC machining starting point. The points of the contour are bidirectionally connected, while the points of the contour are unidirectionally connected to the machining starting point. Taking this diagram as an example, the generated guide hybrid diagram contains 13 nodes, 13 unidirectional edges, and 78 bidirectional edges.

[0055] Step 2: Based on the original contour data of the surface to be processed, generate the contour processing trajectory of the surface to be processed by combining the guide blending map;

[0056] Step two is as follows:

[0057] Step 2.1: Based on the contour machining start point of the current surface to be machined (Shell) sp And the corresponding original contour data, using a heuristic algorithm based on the nearest neighbor principle, initially calculates the contour processing order of the current surface to be processed and the start and end points corresponding to each contour point set; subsequently, each contour point set is renumbered to conform to the determined contour processing order, such as... Figure 7 As shown. The total trajectory length is 944.3957, the maximum value is 158.5792, the minimum value is 2.3474, the average value is 17.1708, the standard deviation is 20.6354, the variance is 425.8178, and the second-order central moment is 418.0756. The number of trajectory transition segments is 55. The trajectory with the largest jump accounts for 16.79% of the total trajectory, indicating that no global optimization was performed.

[0058] In step 2.1, the contour processing sequence of the current surface to be processed and the starting point and ending point of each contour point set are initially calculated using a heuristic method combining the nearest neighbor principle.

[0059] Specifically:

[0060] First, calculate the distance from the starting point of the contour processing in the original contour data of the surface to be processed. sp (Shell sp .x,Shell sp .y) The nearest vertex CP i j And named it the starting point NCP1 1The contour is named as NLC1.

[0061] Next, the obtained contour NLC1 is removed from the original contour data, and the end point of the contour NLC1 (i.e. the start point NCP1 1 ) is taken as a new start point to search for the nearest neighbor points in the updated contour data until the original contour set is completely solved, and a new contour set and the vertex set with the start point as the first item of each contour are constructed

[0062] Step 2.2: Shell based on contour processing start point sp A first directed mixed graph model is established based on the start point corresponding to each contour point set, and the first directed mixed graph model is used to optimize the current contour processing sequence to obtain a new contour processing sequence.

[0063] In step 2.2, the Euclidean distance between different contour point sets is taken as the cost to establish an undirected graph, and a directed graph is established based on the contour processing start point and all contour point sets, so as to obtain the first directed mixed graph model. The total length of the processing sequence and the jump trajectory is optimized by using a depth-first algorithm and local edge exchange:

[0064] Step 1: the printing start point Shell sp of the layer and the known contour start point set {NCP1 1 ,..., NCP1 j ,..., NCP1 n} are taken as nodes to define the algorithm input, a mixed graph data structure G = (V, E) and a cost matrix (Euclidean distance) are established, wherein the distance cost from each contour start point to the processing start point is 0, while the cost from the processing start point to each contour start point is still the Euclidean distance, and the iteration number K and the parameter ε for controlling the reception of suboptimal solutions are defined.

[0065] Step 2: initialize the algorithm, let m = 1, take the transition timing and start point position Shell sp specified in the second step as the initial trajectory, and calculate its cost (cost);

[0066] Step 3: select two random nodes (i, j) as candidate exchange edges, divide the trajectory into two parts according to the exchange edge, reconnect and merge the two parts of the trajectory into a new trajectory;

[0067] Step 4: calculate the cost of the new trajectory, if the cost is better (the cost is smaller), accept it, otherwise reject it;

[0068] Step 5: let m = m + 1, repeat the process of steps 3 to 4 until the convergence condition ΔCost (P) = Cost (P m-t)-Cost(P m If ε ≤ K or the maximum number of iterations m ≥ K is reached, repeat steps 3 to 4 until the convergence condition is met or the maximum number of iterations is reached.

[0069] Step 6: Output the optimized contour point sequence, i.e. the optimized contour processing trajectory.

[0070] like Figure 8 As shown, this invention uses a guided hybrid graph model to optimize the transition timing of the contour. The total trajectory length is 865.2124, the maximum trajectory value is 61.1432, the minimum value is 2.3474, the average value is 15.7311, the standard deviation is 9.2828, the variance is 86.1696, and the second-order central moment is 84.6028. The number of trajectory transition segments is 55. The trajectory with the largest jump accounts for 7.07% of the total trajectory, and the jump trajectories are evenly distributed.

[0071] Step 2.3: Process the starting point Shell based on the contour. sp The new contour processing sequence and the starting positions of the preceding and following contour point sets are optimized to obtain the minimum jump trajectory during contour processing and use it as the contour processing trajectory of the current surface to be processed.

[0072] In step 2.3, for each contour point set, the sum of the distances between each point in the set and the starting points of the two adjacent contour point sets is calculated. The point with the smallest sum of distances in the set is taken as the starting point of the contour point set. The starting point of the first contour point set is determined by the contour processing starting point Shell of the current surface to be processed. sp Once the starting point of the second contour point set is determined, the starting point of the last contour point set does not need to be adjusted. Repeat this step multiple times to obtain the best optimization result.

[0073] like Figure 9 The figure shows the iterative curve of contour connection sequence optimization using the guided hybrid graph model of this invention. The horizontal axis represents the number of iterations, and the vertical axis represents the cost after each iteration. The maximum value is 1005.3776 (represented by ▲), located at 1.0000%, and the minimum value is 12.0000 (represented by ■), located at 12.0000%. The minimum point is reached after 12 steps, demonstrating the efficiency of the guided hybrid graph solution. The maximum slope is 0.0000, the minimum is -14.2972, the average is -0.5597, and the average slope angle is 150.7642°.

[0074] like Figure 10The diagram shows the results of multiple local evolutions of each contour starting point according to this invention. The solid lines represent the final results, while the other lines represent the results during the optimization process. In the initial optimization step, the connection between two points was initially optimized. The total trajectory length was 853.7841, the maximum value was 74.3992, the minimum value was 2.3474, the average value was 15.5233, the standard deviation was 10.4441, the variance was 109.0797, and the second-order central moment was 107.0964. The number of trajectory transition segments was 55. In the second optimization step, the local evolution yielded the final result. The total trajectory length was 775.9585, the maximum value was 65.8100, the minimum value was 1.2726, the average value was 14.1083, the standard deviation was 9.6605, the variance was 93.3256, and the second-order central moment was 91.6288. The number of trajectory transition segments was 55. With the contour connection order unchanged, the total trajectory length is significantly optimized.

[0075] like Figure 11 The figure shows the iterative curves of multiple local evolutions in this invention. The horizontal axis represents the number of iterations, and the vertical axis represents the cost of trajectory time-series planning. The maximum value is 865.2124 (indicated by ▲), located at 0.0125%, and the minimum value is 7.879138e+02 (indicated by ■), located at 100%. The maximum slope is 100.8554, located at iteration step 1755, and the minimum value is -180.1617, located at the initial step. The average value is -0.3251, and the average slope angle is 161.9893°. The number of local minima is 444, shown as 845.6396 in the figure, located at iteration step 33; the number of local maxima is 449, shown as 853.7687 in the figure, located at iteration step 56.

[0076] like Figure 12 The diagram shows the combination of contour trajectory optimization and fill trajectory results under 50% fill in this invention. A partial enlarged view of the contour machining starting point shows the fill relationship between the contour trajectory and the surrounding fill trajectory at the starting point coordinates [160, 95]. The maximum radius of curvature of the fill trajectory is 48.5020, the minimum is 3.3739e-5, and the average is 2.3647. The trajectory, through high-precision interpolation, can be used for servo control of in-service equipment. A partial enlarged view of the fill machining starting point shows the process of directly transitioning from the contour machining endpoint to the fill machining.

[0077] Step 3: Based on the minimum jump trajectory of the current surface to be processed during contour processing and the original contour data of the current surface to be processed, generate the surface processing trajectory of the current surface to be processed by combining the guide blending map;

[0078] Step three specifically involves:

[0079] First, the endpoint of the contour machining trajectory is used as the starting point of the surface machining trajectory. spFor 3D printing, the surface machining trajectory is a fill trajectory. After the current surface to be machined is segmented, the endpoints of each segmented surface and corresponding segmented sub-trajectory are obtained; the start point Fill sp of the surface machining trajectory and the endpoints of each segmented sub-trajectory are used to establish a second directed mixed graph model, and then the transition timing of each segmented sub-trajectory is optimized using the second directed mixed graph model to obtain the jump trajectory during surface machining and use it as the surface machining trajectory of the current surface to be machined, and the endpoint of the jump trajectory during surface machining is used as the contour machining start point of the next surface to be machined.

[0080] In step three, the Euclidean distance between the endpoints of each segmented sub-trajectory is used as the cost to establish an undirected graph, and a directed graph is established between the start point Fill sp of the surface machining trajectory and the endpoints of each segmented sub-trajectory to obtain a second directed mixed graph model. The transition timing of each sub-region is calculated using a depth-first and local edge exchange algorithm:

[0081] Step 1: Define the algorithm input, including the mixed graph data structure G=(V,E) and the cost matrix (Euclidean distance), where for any two sub-trajectory endpoints P i (P i .x,P i .y), if they belong to the same sub-trajectory, the distance cost D(P i ,P j ) is 0, otherwise it is the Euclidean distance between the two plus a penalty value M penalty , where the cost from each sub-trajectory endpoint to the start point is 0, and the cost from the start point to each sub-trajectory endpoint is still the Euclidean distance, and the iteration number K and the parameter ε that controls the acceptance of suboptimal solutions are defined;

[0082] Step 2: Initialize the algorithm, set m=1, define the starting point position Fill sp , and use a heuristic algorithm to create an initial trajectory and calculate its cost;

[0083] Step 3: Select two random endpoints (i,j) as candidate exchange edges, and divide the trajectory into two parts according to the exchange edge, reconnect and merge the two parts into a new trajectory;

[0084] Step 4: Calculate the cost of the new trajectory. If the objective function Cost(P) decreases, accept it, otherwise reject it;

[0085] Step 5: Set m=m+1, repeat steps 3 to 4 until the convergence condition ΔCost(P) = Cost(P m-t )-Cost(P m )≤ε or the maximum iteration number m≥K is reached.

[0086] Step 6: output the optimized sub-trajectory endpoint transition timing, i.e. the optimized surface machining trajectory.

[0087] Step four: contour machining the current to-be-machined surface according to the contour machining trajectory, and then surface machining the to-be-machined surface which has been contoured machined according to the surface machining trajectory, complete the machining of the current to-be-machined surface, and take the endpoint of the surface machining trajectory as the starting point of the contour machining of the next to-be-machined surface;

[0088] As Figure 13 shown, it is the result comparison and presentation of the multiple optimization process of the application. Among them, the total length of the machining jump trajectory is reconstructed and evolved by the timing planning scheme, and finally compared with the initial trajectory total length, it is decreased by 17.53%, which proves that the trajectory planning method improves the machining efficiency. At the same time, the variance and kurtosis of the jump trajectory are decreased by 78.12% and 75.23% respectively, which proves that after optimization, each trajectory has the characteristics of low dispersion, high data aggregation and relative stability, which reduces the tool vibration during high-speed machining and micro-machining, and improves the machining quality of the parts.

[0089] Step five: repeat steps two to four to machine the remaining to-be-machined surface, thereby completing the machining of the to-be-machined part.

[0090] The above specific embodiments are used to explain and illustrate the application, rather than limit the application, any modifications and changes made to the application within the spirit and protection scope of the claims of the application, fall into the protection scope of the application.

Claims

1. A timing planning method for CNC machining trajectory based on a guide hybrid graph, characterized in that, Includes the following steps: Step 1: After classifying the original data point set of the workpiece, obtain the original contour data corresponding to each surface to be processed of the workpiece, and determine the initial surface to be processed of the workpiece. Step 2: Based on the original contour data of the surface to be processed, generate the contour processing trajectory of the surface to be processed by combining the guide blending map; Step two specifically involves: Step 2.1: Start point for contour machining based on the current surface to be machined. And the corresponding original contour data, to preliminarily calculate the contour processing sequence of the current surface to be processed and the start and end points of each contour point set; Step 2.2: Based on the contour processing start point A first guide hybrid graph model is established with the starting point corresponding to each contour point set. The current contour processing order is optimized using the first guide hybrid graph model to obtain a new contour processing order. In step 2.2, an undirected graph is established using the Euclidean distance between different contour point sets as the cost, and a directed graph is established at the contour processing starting point and in all contour point sets, thereby obtaining the first guided hybrid graph model. Step 2.3: Process the starting point based on the contour. The new contour processing sequence and the starting positions of the front and rear contour point sets are optimized to obtain the minimum jump trajectory during contour processing and use it as the contour processing trajectory of the current surface to be processed. Step 3: Based on the minimum jump trajectory of the current surface to be processed during contour processing and the original contour data of the current surface to be processed, generate the surface processing trajectory of the current surface to be processed by combining the guide blending map; Step three specifically involves: First, the endpoint of the contour machining trajectory is taken as the starting point of the surface machining trajectory. Next, the surface to be processed is divided into segments to obtain each segmented surface and the endpoints of the corresponding segmented sub-trajectories; then the starting point of the surface processing trajectory is determined. A second guiding hybrid graph model is established with the endpoints of each segmented sub-trajectory. Then, the transition timing of each segmented sub-trajectory is optimized using the second guiding hybrid graph model to obtain the jump trajectory during surface processing and use it as the surface processing trajectory of the current surface to be processed. In step three, an undirected graph is established using the Euclidean distance between the endpoints of each segmented sub-trajectory as the cost, and the starting point of the surface-processed trajectory is then established. A directed graph is established at the endpoints of each segmented sub-trajectory to obtain the second guided hybrid graph model; Step 4: Perform contour machining on the current surface to be machined according to the contour machining trajectory, and then perform surface machining on the surface to be machined after the contour has been machined according to the surface machining trajectory to complete the machining of the current surface to be machined; Step 5: Repeat steps 2-4 to process the remaining surfaces to be processed, thereby completing the processing of the workpiece.

2. The CNC machining trajectory timing planning method based on guide hybrid graph as described in claim 1, characterized in that, The surface to be processed is specifically a processing layer in 3D printing or a processing surface in a CNC machine tool.

3. The CNC machining trajectory timing planning method based on guide hybrid graph according to claim 1, characterized in that, In step one, the original contour data of each surface to be processed consists of the processing outer contour point set and the internal hole contour point set of each disconnected domain.

4. The CNC machining trajectory timing planning method based on guide hybrid graph as described in claim 1, characterized in that, In step 2.1, the contour processing sequence of the current surface to be processed and the starting point and ending point of each contour point set are initially calculated using a heuristic method combining the nearest neighbor principle.

5. The CNC machining trajectory timing planning method based on guide hybrid graph according to claim 1, characterized in that, In step 2.3, for each contour point set, the distance between each point in the contour point set and the starting point of the two adjacent contour point sets is calculated, and the point with the smallest distance sum in the contour point set is taken as the starting point of the contour point set.

Images