JP7109703B1 - MACHINING PROGRAM GENERATION DEVICE AND MACHINING PROGRAM GENERATION METHOD - Google Patents

Abstract

A machining program generation device (10) comprises a product shape memory unit (11), a material shape memory unit (12), a first work coordinate system including a first work origin, and a work origin/machining for selecting a first work direction. A direction storage unit (13), a machining program generation unit (17) for generating a first machining program for machining the first machining shape, and an uncut shape generation unit (18) for generating an uncut shape. . A work origin/machining direction storage unit (13) selects a second work coordinate system including a second work origin and a second machining direction for machining an uncut shape. A machining program generator (17) extracts a second machining shape that can be machined in a second machining direction, and generates a second machining program for machining the extracted second machining shape.

Description

The present disclosure relates to a machining program generation device and a machining program generation method for automatically generating a machining program for numerically controlling a machine tool.

In the field of machine tools controlled by numerical controllers, the structure of machine tools has become more complex, the number of axes controlled by machine tools has increased, and the processing of numerically controlled objects has become more complicated, so that complex shapes can be machined precisely. increased, and machining programs are becoming more complex. Moreover, in recent years, it has become common to automatically generate a machining program from three-dimensional CAD data in order to improve the efficiency of generating a machining program.

In Patent Document 1, after extracting each characteristic portion of the processed product from the three-dimensional CAD data of the product, the mounting direction of the processed product is determined to determine the cutting direction, and processing of each characteristic portion is performed based on the cutting direction. You have set up a machining process that includes conditions and toolpaths.

Japanese Patent No. 6719790

Patent Document 1 does not disclose how to generate a machining program for an uncut portion in the determined cutting direction.

The present disclosure has been made in view of the above, and aims to obtain a machining program generation device that can efficiently generate a machining program for an uncut portion in a short period of time without requiring an operator troublesome work. aim.

In order to solve the above-described problems and achieve the object, the machining program generation device in the present disclosure is a machining program generation device that generates a machining program including a plurality of cutting processes for cutting a cut product from a workpiece. is. The machining program generation device includes a product shape storage unit that stores the product shape of the cut product, a material shape storage unit that stores a first material shape that is the material shape of the cut product before machining, the first material shape and Generating a machining shape based on the product shape, selecting the first work coordinate system including the first work origin of the machined product and the first machining direction based on the generated machining shape, and storing the work origin and machining A first machining shape that can be machined in the first machining direction is extracted from the machining shape based on the direction storage unit, the selected first machining direction, the product shape, and the first material shape, and the extracted first machining shape and a machining program generation unit that generates a first machining program for machining the first material shape to generate a machining shape by removing the product shape from the first material shape , removing the first machining shape from the generated machining shape and cutting and a uncut shape generating unit that generates a uncut shape. The work origin/machining direction storage unit removes the first machining shape from the first material shape to generate a second material shape, and processes the uncut shape based on the generated second material shape and the uncut shape . A second workpiece coordinate system including a second workpiece origin and a second machining direction are selected for. The machining program generator extracts from the machining shape a second machining shape that can be machined in the second machining direction based on the selected second machining direction and the uncut shape, and machining the extracted second machining shape. to generate a second machining program.

According to the machining program generation device of the present disclosure, it is possible to efficiently generate a machining program for an uncut portion in a short period of time without giving the operator troublesome work.

1 is a block diagram showing the configuration of a numerical control device according to a first embodiment; FIG. 4 is a flowchart showing the procedure of machining program generation processing performed by the machining program generation device according to the first embodiment; 4 is a flowchart showing the procedure of learning model generation processing performed by the machine learning device according to the first embodiment; 4 is a flowchart showing the procedure of inference processing performed by the inference apparatus according to the first embodiment; 4 is a flow chart showing details of processing for storing a workpiece coordinate system including a workpiece origin and a machining direction performed by a workpiece origin/machining direction storage unit according to the first embodiment; 4 is a perspective view showing an example of a product shape corresponding to product shape data stored in a product shape storage unit according to the first embodiment; FIG. 4 is a perspective view showing an example of material shape corresponding to material shape data stored in the material shape storage unit according to the first embodiment; FIG. FIG. 4 is a perspective view illustrating candidate points of the workpiece origin displayed by the workpiece origin/machining direction storage unit according to the first embodiment; 4 is a perspective view exemplifying a machining direction vector set by a work origin/machining direction storage unit according to the first embodiment; FIG. 4 is a flowchart showing details of machining shape development processing performed by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view exemplifying a machining shape generated by the machining shape development unit according to the first embodiment; 4 is a perspective view exemplifying a visible surface of the product shape with the machining direction vector as the line of sight, which is specified by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view illustrating machining shapes that can be machined according to machining direction vectors extracted by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view illustrating machining shapes that can be machined according to machining direction vectors extracted by the machining shape development unit according to the first embodiment; FIG. FIG. 4 is a perspective view showing a plurality of surface process shapes divided by the processing shape developing portion according to the first embodiment; 4 is a flowchart showing details of processing unit allocation processing performed by the processing unit allocation unit according to the first embodiment; 4 is a flowchart showing details of machining shape division processing of a stepped surface machining shape performed by a machining shape development unit according to the first embodiment; FIG. 4 is a perspective view illustrating a stepped surface machining shape processed by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view illustrating a surface machining shape obtained by horizontal division performed by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view illustrating a surface machining shape obtained by vertical division performed by the machining shape development unit according to the first embodiment; 4 is a flowchart showing details of machining program generation processing performed by the machining program generation unit according to the first embodiment; Flowchart showing details of uncut shape generation processing performed by the uncut shape generation unit according to the first embodiment FIG. 4 is a perspective view illustrating an uncut shape generated by the uncut shape generation unit according to the first embodiment; Flowchart showing details of uncut shape generation processing performed by the machining program generation device according to the first embodiment FIG. 4 is a perspective view for explaining a work coordinate system including the work origin set by the work origin/machining direction storage unit according to the first embodiment; FIG. 4 is a perspective view showing an example of a machined shape developed in the next step with the machining direction vector as the machining direction in Embodiment 1; FIG. 4 is a perspective view showing a new uncut shape that cannot be machined in the next step in Embodiment 1; Flowchart showing processing for generating a machining program without generating an uncut shape in the first embodiment FIG. 4 is a perspective view illustrating examples of a facing shape and a hole machining shape for which a machining program has been generated up to the present stage in Embodiment 1; 4 is a perspective view showing the shape of the material in the next step in Embodiment 1. FIG. 4 is a flowchart showing the procedure of machining direction determination processing performed by the machining program generation device according to the first embodiment; 4 is a flowchart showing the procedure of learning model generation processing performed by the machine learning device according to Embodiment 1; 4 is a flow chart showing the procedure of undeveloped shape process development processing performed by the machining shape development unit according to the first embodiment; FIG. 4 is a perspective view showing a plurality of surface process shapes obtained by development processing to surface process shapes for rough machining in the machining shape development unit according to the first embodiment; Block diagram showing the hardware configuration of the second embodiment in the machining program generation device, machine learning device, and inference device.

A machining program generation device and a machining program generation method according to embodiments will be described in detail below with reference to the drawings.

Embodiment 1.
FIG. 1 is a block diagram showing the configuration of a numerical control device 100 according to Embodiment 1. As shown in FIG. The numerical controller 100 has a machining program generation device 10 , a machine learning device 20 , an inference device 30 , an instruction input section 40 , an interactive operation processing section 50 and a display section 60 . The numerical controller 100 is mounted on a machine tool (not shown) or connected to the machine tool, and numerically controls the operation of the machine tool according to a machining program. Here, the machining program is used for cutting the object to be machined from the state of the raw material to carve the product shape of the machined product. A machine tool is, for example, a machining center.

In the example shown in FIG. 1, the machining program generation device 10, the machine learning device 20, and the inference device 30 are installed in the numerical control device 100, but the first embodiment is not limited to such an example. For example, the machining program generation device 10, the machine learning device 20, and the inference device 30 may be devices different from the numerical control device 100. FIG. Moreover, the machining program generation device 10 may be a device different from the machine learning device 20 and the inference device 30 . Also, the machine learning device 20 may be a device different from the inference device 30 .

The machining program generation device 10 generates a machining program based on machining shape data input to the machining program generation device 10 from outside the numerical control device 100 . The processed shape data includes product shape data and material shape data. The material shape data is data that defines the material shape, which is the shape of the object to be processed before processing. The product shape data is data that defines the product shape, which is the final shape of the object after processing. The material shape data and the product shape data are CAD (Computer-Aided Design) data 1, for example. The machining program generation device 10 uses the shape development method, the machining unit, and the tool/cutting conditions, which are the inference results of the inference device 30, when generating the machining program. A machining unit is a machining unit in which continuous machining is performed using the same spindle and the same tool. The machining unit has machining data including information on machining methods, tool data including information on tools used for machining and cutting conditions, and shape sequence data including shape information defining a machining shape consisting of a single shape. .

The machine learning device 20 generates a learning model to be used when the machining program generating device 10 generates a machining program based on a plurality of machining programs 2 created in the past. That is, the machining program 2 input to the machine learning device 20 is a machining program created in the past for learning, and the machining program generated by the machining program generation device 10 is created for numerically controlling the machine tool. It is a new machining program that is used.

The inference device 30 uses the learning model generated by the machine learning device 20 to infer the shape expansion method, machining unit, and tool/cutting conditions in the machining program generated by the machining program generation device 10, and outputs the inference result. Return to the machining program generation device 10 .

The interactive operation processing unit 50 is an interface between the numerical control device 100 and the operator, and is also an interface between the machining program generation device 10, the machine learning device 20 or the inference device 30 and the operator. The interactive operation processing unit 50 transmits instruction information input by the operator via the instruction input unit 40 to the machining program generation device 10 , the machine learning device 20 or the inference device 30 . Further, the interactive operation processing unit 50 displays instruction information input by the operator via the instruction input unit 40 on the display unit 60 .

The instruction input unit 40 is composed of input devices such as a mouse and a keyboard. The instruction input unit 40 receives instruction information from the operator and transmits the instruction information to the interactive operation processing unit 50 .

The display unit 60 is a display device such as a liquid crystal monitor, and displays the CAD data 1, the machining program 2, instruction information input by the operator via the instruction input unit 40, and the like. Further, the display unit 60 can display various information related to various processes performed by the numerical control device 100 , the machining program generation device 10 , the machine learning device 20 and the inference device 30 .

The machining program generation device 10 includes a product shape storage unit 11, a material shape storage unit 12, a workpiece origin/machining direction storage unit 13, a machining shape developing unit 14, a machining unit allocation unit 15, and a machining process storage unit 16. , a machining program generator 17 , and an uncut shape generator 18 .

Product shape data composed of the CAD data 1 is input from an external device of the numerical controller 100 to the machining program generator 10 . The product shape storage unit 11 receives product shape data input from an external device and stores the product shape data.

The product shape data includes product shape information, which is the finished shape of the machined product, and material information, which indicates the material of the material. The product shape data is not limited to the CAD data 1, and may be any data that can be interpreted by the machining program generation device 10. FIG.

The material shape storage unit 12 stores material shape data, which is material shape data that is the shape of an object to be processed before processing. A material shape, which is the shape of the object to be processed before processing, corresponds to the first material shape. The material shape data is not limited to the CAD data 1, and any data that can be interpreted by the machining program generation device 10 may be used. The shape of the raw material includes, for example, a columnar shape or a rectangular parallelepiped shape that encloses the shape of the product. Also, the material shape does not necessarily include the product shape, and may be a shape in which any surface of the product shape is thickened, or a shape in which holes are removed from the product shape.

The work origin/machining direction storage unit 13 stores a work coordinate system including the coordinate values of the work origin set at an arbitrary position by the user, and the machining direction set by the user. The work origin is an origin for creating a machining program, and is set at, for example, the center or corner of the upper surface of the product shape or material shape. The work coordinate system is a coordinate system whose origin is the work origin, and is set as a world coordinate system having, for example, an X-axis, a Y-axis, and a Z-axis. The work origin does not have to be set at the center or corner of the plane, and can be set at any position by the user. The machining direction is a direction parallel to the rotation axis of the cutting tool, and is the direction in which the cutting tool cuts into the shape of the object to be machined.

The machining shape expansion unit 14 stores the product shape data stored in the product shape storage unit 11, the material shape data stored in the material shape storage unit 12, the work origin stored in the work origin/machining direction storage unit 13, A machining shape to be machined by cutting is extracted based on the work coordinate system and the machining direction. The processed shape is the shape of the portion removed from the material shape. The machining shape development unit 14 divides the machining shape so as to perform planar machining or hole machining.

The machining unit allocation unit 15 allocates a machining method, a tool, and a cutting condition to each machining shape divided by the machining shape development unit 14 .

The machining process storage unit 16 stores a plurality of machining shapes divided by the machining shape developing unit 14 and machining methods, tools, and cutting conditions assigned by the machining unit allocation unit 15 .

Based on the work coordinate system including the work origin stored in the work origin/machining direction storage unit 13 and a plurality of machining shapes, machining methods, tools, and cutting conditions stored in the machining process storage unit 16, the machining program generation unit 17 to generate a machining program.

The uncut shape generation unit 18 generates a cut shape from the product shape data stored in the product shape storage unit 11, the material shape data stored in the material shape storage unit 12, and the machining shape extracted and divided by the machining shape expansion unit 14. Generate the remaining shape.

The machine learning device 20 has a machining program storage unit 21 , a machining program analysis unit 22 , a machine learning unit 23 and a machine learning model storage unit 24 .

A machining program 2 is input to the machine learning device 20 from an external device of the numerical controller 100 . The machining program storage unit 21 receives a machining program 2 input from an external device and stores the received machining program 2 . The machining program 2 is a program created in the past, and is a computer program for numerically controlling a machine tool (not shown). The machining program 2 includes information on machining methods, tools, cutting conditions, tool trajectories, material shapes, material properties of materials, and the like.

The machining program analysis unit 22 extracts first parameters and second parameters from the machining program 2 stored in the machining program storage unit 21 . The first parameter and the second parameter are parameters used within the machining program 2 . A first parameter is a parameter to be adjusted in the machining program 2 . The second parameter is a parameter that is not subject to adjustment in the machining program 2 and is used for adjusting the first parameter. Parameter adjustment refers to determining the value of the parameter. The value of the first parameter is determined when the machining program 2 is generated.

The first parameters include parameters representing, for example, machining method, machining order, tool type, feed, cutting speed, radial depth of cut, axial depth of cut, direction of shape division, and the like. The second parameter is, for example, a parameter whose value is determined based on the material shape, material quality, processing shape, etc., 3D shape image data, 3D shape voxel data, or inside/outside determination data on a 3D grid space. Contains fixed-length data such as The second parameters include adjusted parameters. A second parameter used when adjusting the first parameter is associated with each first parameter. A first parameter is adjusted based on a second parameter corresponding to the first parameter.

The machining program analysis unit 22 determines a second parameter for each first parameter and extracts the determined second parameter. The machining program analysis unit 22 inputs the first parameter and the second parameter extracted for each machining program 2 to the machine learning unit 23 .

The machine learning unit 23 generates a learning model by learning using the extracted data set including the first parameter and the second parameter. The machine learning unit 23 generates a learning model for inferring the value of the first parameter from the second parameter of the machining program edited by the operator. In Embodiment 1, the machine learning unit 23 performs supervised learning for generating a learning model. The machine learning unit 23 inputs the generated learning model to the machine learning model storage unit 24 .

Any learning algorithm may be used by the machine learning unit 23 . Examples include algorithms such as neural networks and SVM (Support Vector Machine). The neural network may be multi-layered deep learning. Also, the learning algorithm used by the machine learning unit 23 may be genetic programming, inductive logic programming, support vector machine, or the like. Machine learning is the process of optimizing parameters such as weights or biases in neural networks.

The machine learning model storage unit 24 stores learning models that are the learning results of the machine learning unit 23 . The learning model indicates the relationship of the optimum first parameter to the input second parameter.

The inference device 30 has an inference unit 31 . Input data including a second parameter is input to the inference unit 31 as input data. The inference unit 31 infers the value of the first parameter from the second parameter using the learning model. The inference unit 31 inputs the inference result to the machining program generation device 10 . The inference unit 31 outputs multiple values of the first parameter as inference results.

Next, operations of the numerical controller 100 will be described. In the numerical controller 100, machining program generation processing performed by the machining program generation device 10, learning model generation processing performed by the machine learning device 20, and inference processing performed by the inference device 30 are performed.

FIG. 2 is a flowchart showing the procedure of machining program generation processing performed by the machining program generation device 10 according to the first embodiment. The machining program generation device 10 generates a machining program using the inference result of the inference device 30 using the learning model of the learning result of the machine learning device 20 .

In step S1, the product shape storage unit 11 reads product shape data composed of the CAD data 1 from a storage area (not shown) and stores the product shape data.

In step S2, the material shape storage unit 12 generates a material shape based on the product shape data stored in the product shape storage unit 11, and stores the material shape data.

In step S3, the work origin/machining direction storage unit 13 stores the work coordinate system including the work origin coordinates specified by the operator and the machining direction vector.

In step S<b>4 , the machining shape development unit 14 generates machining shapes based on the product shape data stored in the product shape storage unit 11 and the material shape data stored in the material shape storage unit 12 . Further, the machining shape expansion unit 14 converts the machining shape into a machining shape for surface machining, a machining shape for line machining, and a machining direction based on the work coordinate system including the work origin stored in the work origin/machining direction storage unit 13 and the machining direction. shape, and machining shape for hole machining. At that time, the machining shape development unit 14 generates and develops the machining shape based on the result of the shape development method inferred by the inference device 30 .

In step S<b>5 , the machining unit allocation unit 15 determines machining methods, tools, and cutting conditions for each machining shape developed by the machining shape development unit 14 based on the inference results of the inference device 30 .

In step S<b>6 , the machining process storage unit 16 stores the machining shape developed by the machining shape development unit 14 and the machining method, tool, and cutting conditions allocated by the machining unit allocation unit 15 .

In step S7, the machining program generation unit 17 stores the workpiece coordinates including the workpiece origin stored in the workpiece origin/machining direction storage unit 13, and the machining shape, machining method, tool, and cutting conditions stored in the machining process storage unit 16. A machining program is generated based on this.

In step S8, the uncut shape generating unit 18 extracts and divides the product shape data stored in the product shape storage unit 11, the material shape data stored in the material shape storage unit 12, and the processed shape developing unit 14. Generates uncut shape from machining shape.

FIG. 3 is a flowchart showing the procedure of learning model generation processing performed by the machine learning device 20 according to the first embodiment. In the learning model generation process, a learning model for generating a machining program is generated based on the machining program 2 .

In step S11, the machining program storage unit 21 reads a plurality of machining programs 2 from a storage area (not shown). The machining program storage unit 21 stores the machining program 2 .

In step S<b>12 , the machining program analysis section 22 extracts the first parameter from the machining program 2 . A machining program analysis unit 22 extracts a plurality of first parameters used in the machining program 2 .

In step S13, the machining program analysis unit 22 extracts a second parameter for each of the plurality of extracted first parameters. At this time, the machining program analysis unit 22 determines a second parameter to be extracted for each first parameter, and extracts the determined second parameter. The machining program analysis unit 22 performs the processing of steps S12 and S13 for each machining program. The machining program analysis unit 22 inputs the extracted first parameter and second parameter to the machine learning unit 23 for each machining program.

In step S14, the machine learning unit 23 performs machine learning processing using the input first and second parameters. Machine learning unit 23 generates a data set based on the first parameter and the second parameter, and performs machine learning according to the generated data set. A data set is a set of data in which a first parameter to be adjusted and a second parameter, which is a non-adjustable parameter used to determine the value of the first parameter, are associated with each other. The machine learning unit 23 generates an optimized model as a learning model using predetermined criteria. The machine learning unit 23 generates a learning model as a learning result.

In step S15, the machine learning model storage unit 24 stores the generated machine learning model. With the above, the machine learning device 20 ends the learning model generation processing according to the procedure shown in FIG.

FIG. 4 is a flowchart showing the procedure of inference processing performed by the inference device 30 according to the first embodiment. In the inference process, the second parameter extracted by the machining program generation device 10 is input, and the machine learning device 20 uses the learning model generated based on the machining program 2 to infer the first parameter.

In step S<b>21 , the second parameter selected by the machining program generation device 10 is input to the inference section 31 .

In step S22, the inference unit 31 selects a learning model stored in the machine learning device 20 in accordance with the first parameter to be inferred, and uses the selected learning model from the second parameter input in step S21. , to infer the first parameter. Thus, the inference device 30 ends the inference processing according to the procedure shown in FIG.

Next, the details of the machining program generation process performed by the machining program generation device 10 will be described. FIG. 5 is a flow chart showing details of processing for storing the work coordinate system including the work origin and the processing direction performed by the work origin/machining direction storage unit 13 according to the first embodiment. The processing in FIG. 5 corresponds to steps S1 to S3 in FIG. In the following, an example of a machining program generation process relating to milling machining in which an object to be machined is cut by moving the tool while rotating the end mill will be described.

In step S<b>31 , the work origin/machining direction storage unit 13 reads the product shape data stored in the product shape storage unit 11 and the material shape data stored in the material shape storage unit 12 .

FIG. 6 is a perspective view showing an example of the product shape SH1 corresponding to the product shape data stored in the product shape storage unit 11 according to the first embodiment. FIG. 7 is a perspective view showing an example of the material shape SH2 corresponding to the material shape data stored in the material shape storage unit 12 according to the first embodiment.

As shown in FIG. 6, the product shape SH1 is generally a rectangular parallelepiped KH having four upper corners with recesses KH1, four side faces with recesses KH2, and two vertically extending round holes KH3. A rectangular parallelepiped frame RH is placed on the rectangular parallelepiped KH3 and has round holes RH1 on four side surfaces. A space is formed inside the frame RH. Further, a stepped concave portion KH4 is formed in the rectangular parallelepiped KH.

In FIG. 7, the rectangular parallelepiped material shape SH2 is selected, and the coordinates of the eight corners of the rectangular parallelepiped material shape SH2 are displayed.

In step S<b>32 , the work origin/machining direction storage unit 13 displays the product shape SH<b>1 , the material shape SH<b>2 , and candidate points for the work origin on the display unit 60 . Candidate points for the work origin include, for example, all vertices, side midpoints, and surface midpoints in the product shape SH1 and material shape SH2. FIG. 8 is a perspective view illustrating candidate points of the workpiece origin displayed by the workpiece origin/machining direction storage unit 13 according to the first embodiment. In FIG. 8, a plurality of work origin candidate points Q are indicated by black dots. The material shape SH2 is arranged so as to include the product shape SH1.

In step S<b>33 , the work origin/machining direction storage unit 13 selects an arbitrary work origin from a plurality of work origin candidate points Q displayed on the display unit 60 based on the operator's instruction via the instruction input unit 40 . Select and set. Coordinates other than the candidate point Q may be set as the work origin. In FIG. 8, the coordinate position (0.0, 0.0, 0.0) in the world coordinate system is selected as the work origin Q0, and the work coordinate system (X, Y, Z) with the work origin Q0 as the origin is selected. A work coordinate system (X, Y, Z) having the work origin Q0 as its origin corresponds to the first work coordinate system including the first work origin.

In step S34, the work origin/machining direction storage unit 13 sets a machining direction vector based on the plane selected by the operator. If a plane is selected, the reverse direction vector of the normal vector of the plane is determined as the machining direction vector. When the cylindrical surface is selected, the direction selected by the operator is determined as the machining direction vector from the two positive and negative axial direction vectors of the cylindrical surface. FIG. 9 is a perspective view illustrating machining direction vectors set by the workpiece origin/machining direction storage unit 13 according to the first embodiment. In FIG. 9, the uppermost surface K1 of the material shape SH2 is selected by the operator, and the machining direction vector V1 is set by the normal vector of the uppermost surface K1. Machining direction vector V1=(0.0, 0.0, -1.0). A machining direction vector V1 corresponds to the first machining direction.

FIG. 10 is a flowchart showing details of machining shape development processing performed by the machining shape development unit 14 according to the first embodiment. The processing in FIG. 10 corresponds to the processing in step S4 in FIG.

In step S41, the machining shape expansion unit 14, based on the product shape data stored in the product shape storage unit 11 and the material shape data stored in the material shape storage unit 12, removes the portion to be removed from the material shape SH2. Generate a machining shape that is the shape of The processed shape can be generated by removing the product shape SH1 from the material shape SH2. FIG. 11 is a perspective view illustrating a machining shape SH3 generated by the machining shape development unit 14 according to the first embodiment. A processing shape SH3 shown in FIG. 11 indicates a difference area between a material shape SH2 indicated by a solid line and a product shape SH1 indicated by a broken line.

In step S42, the machining shape development unit 14 extracts a machining shape that can be processed based on the machining direction vector V1 stored in the work origin/machining direction storage unit 13 from among the machining shapes generated in step S41. Specifically, first, the visible surface of the product shape SH1 is identified with a position sufficiently distant from the machining shape SH3 in the opposite direction of the machining direction vector V1 as a viewpoint and with the machining direction vector V1 as the line of sight. Next, by specifying and extracting the portion formed by the visible surface of the product shape SH1 from the machining shape SH3, a machining shape that can be machined according to the machining direction vector V1 is extracted. FIG. 12 is a perspective view illustrating a visible surface of the product shape SH1 with the machining direction vector V1 as the line of sight, which is specified by the machining shape development unit 14 according to the first embodiment. FIG. 13 is a perspective view illustrating machining shapes that can be machined according to the machining direction vector V1 extracted by the machining shape development unit 14 according to the first embodiment. FIG. 12 shows a plurality of visible surfaces C of the product shape SH1 with the machining direction vector V1 shown in FIG. 9 as the viewing direction. FIG. 13 shows a machining shape SH4 that is made up of a plurality of visible surfaces C shown in FIG. 12 and that can be machined according to the machining direction vector V1. Machining shapes that cannot be machined according to the machining direction vector V1 are excluded. It is A machining shape SH4 that can be machined according to the machining direction vector V1 corresponds to the first machining shape.

In step S43, the machining shape developing unit 14 extracts a surface machining shape for cutting a surface and a hole machining shape for cutting a hole from the machining shape SH4 extracted in step S42. Specifically, a cylindrical surface and a conical surface are extracted from the extracted machining shape SH4. Next, of the extracted cylindrical and conical surfaces, the cylindrical and conical surfaces whose axial vectors are parallel to the machining direction vector V1 are extracted, and the extracted cylindrical and conical surfaces are generated as a hole machining shape. The surface machining shape is extracted by excluding the generated hole machining shape from the machining shape SH4 extracted in step S42. FIG. 14 is a perspective view exemplifying the facing shape and the drilling shape extracted by the machining shape development unit 14 according to the first embodiment. FIG. 14 shows the extracted facing shape SH5 and two drilling shapes SH6.

In step S44, the machining shape development unit 14 divides the surface machining shape SH5 extracted in step S43 into a plurality of surface process shapes according to the planes to be machined. Specifically, a plane perpendicular to the machining direction vector V1 is extracted from the facing shape SH5, and the extracted plane divides the facing shape SH5 in the horizontal direction or the vertical direction. FIG. 15 is a perspective view showing a plurality of surface process shapes divided by the processing shape expanding section 14 according to the first embodiment. FIG. 15 shows a surface process shape SH7, a surface process shape SH8, a surface process shape SH9, a surface process shape SH10, and a surface process shape SH11 divided from the surface processing shape SH5. 6 and 7, the surface process shape SH7 corresponds to the processing shape from the upper surface of the material shape SH2 to the upper surface of the frame RH. The surface processed shape SH8 corresponds to the processed shape of the outside of the side surface of the frame RH. The surface processing shape SH9 corresponds to the processing shape of the outside of the side surface of the rectangular parallelepiped KH. The surface processed shape SH10 corresponds to the processed shape inside the frame RH. The surface process shape SH11 corresponds to the recess KH2 of the rectangular parallelepiped KH.

FIG. 16 is a flowchart showing details of processing unit allocation processing performed by the processing unit allocation unit 15 according to the first embodiment. The processing in FIG. 16 corresponds to the processing in step S5 in FIG.

In step S51, the machining unit allocation section 15 reads the surface process shapes SH7 to SH11 and the hole machining shape SH6 generated by the machining shape developing section .

In step S52, the machining unit allocation section 15 selects an end mill surface unit, an end mill crest unit, a pocket mill unit, a pocket crest unit, or a pocket trough unit for each of the surface process shapes SH7 to SH11 according to the shape feature. Assign any facing unit.

For the end mill face unit, the entire contour of the defined shape is machined. When processing, the inner contour of the defined shape is left and processed.

In the end mill crest unit, the outer shape is pond-shaped and the inner shape is chevron-shaped. Pond shapes are machined with the tool protruding, and mountain shapes are not protruded.

For the pocket mill unit, an end mill is used to machine a defined shape to be the pocket.

For the pocket ridge unit, an end mill is used to machine a shape defined to form a pocket while leaving the contour of the inner shape out of the defined shapes. The outer shape is pond-shaped, and the inner shape is mountain-shaped. The tool does not protrude from the pond shape and mountain shape.

For the pocket valley unit, an end mill is used to machine a shape defined to form a pocket, leaving the contour of the inner shape of the defined shape. The outer shape is a pond shape, and the inner shape is a valley shape. The tool does not protrude for the pond shape, but the tool protrudes for the valley shape.

The surface process shape SH7 in FIG. 15 is assigned an end mill surface unit. The surface process shape SH8 in FIG. 15 is assigned an end mill crest unit. The surface process shapes SH10 and SH11 in FIG. 15 are assigned to the pocket mill unit. The surface process shape SH9 in FIG. 15 is an unassigned shape because there are portions that cannot be finished by planarization. Unassigned shapes are also called undeployed shapes.

Further, the machining unit allocation unit 15 allocates the drilling shape SH6 to one of the drill unit and the seated hole unit according to the shape feature. In the drill unit, simple holes are machined to defined locations. In the seated hole unit, stepped holes are machined for defined positions. The two drilling shapes SH6 shown on the right side of FIG. 14 are assigned to drill units.

In step S53, the machining unit allocation section 15 sets shape data corresponding to each machining unit. For the surface machining unit, the contour shape to be machined, the machining position, and the machining depth are set. For the hole machining unit, the position to be machined, the hole diameter, and the hole depth are set.

In step S54, the machining unit allocation section 15 sets a machineable tool and cutting conditions for each machining shape. The machining unit set in step S52 and the shape data set in step S53 are extracted as second parameters, and the inference device 30 performs inference using the tool type to be used and the tool nominal diameter to be used as the first parameters, Decide which tools to use. Next, the inferred tool to be used, the machining unit set in step S52, and the shape data set in step S53 are extracted as second parameters, and the amount of cut in the axial direction, the amount of cut in the depth direction, the convergence, and the feed are extracted. , respectively as the first parameter, by the inference device 30 to set the tool and the cutting conditions.

Here, it is assumed that the machining unit allocation unit 15 designates "tool material", "tool shape", and "tool nominal diameter" as the first parameters to be inferred. The machining unit allocation unit 15 uses the material, rough machining, upper surface coordinates "110.0, 110.0, 0.0" of the surface process shape SH7, surface The bottom surface coordinates of the process shape SH7 are "0.0, 0.0, -5.0", the X-axis dimension of the surface process shape SH7 is "110.0", and the Y-axis dimension of the surface process shape SH7 is "110.0". , the Z-axis dimension of the surface process shape SH7 "5.0", the cross-sectional area "12,100", the shape adjacent to the surface process shape SH7 in the horizontal direction "None", the mountain shape included in the surface process shape SH7 "None" ”, and the valley shape “None” included in the surface process shape SH7 (see FIG. 15). The machining unit allocation unit 15 inputs input data including these second parameters to the inference device 30 .

Note that the machining unit allocation unit 15 acquires the material quality set when the material shape is generated as a second parameter. Further, the machining unit allocation unit 15 acquires the coordinates and dimension values of the surface process shape SH7 by analyzing the surface process shape SH7. The machining unit allocation unit 15 may acquire the worker name and the type of machine as the second parameters.

The processing unit allocation unit 15 acquires from the inference device 30 a plurality of values of the first parameter as a result of inference by the inference device 30 and a probability of use for each of the plurality of values. That is, the machining unit allocation unit 15 acquires a set of multiple values and probabilities of the first parameter.

Here, the machining unit allocation unit 15, for example, selects the first parameter "tool material": carbide, "tool shape": square end mill, and "tool nominal diameter": 20 as the most probable inference result. Set the tool by obtaining Next, the machining unit allocation unit 15 inputs the set tool as one of the second parameters, and sets the cutting conditions such as radial depth of cut, axial depth of cut, peripheral speed, feed, and tool path pattern. Can be set.

In step S55, the machining unit allocation unit 15 determines whether or not there is a machining shape to which machining units are not assigned, and if there is a machining shape to which machining units are not assigned, the process of step S52 is performed. If there is no machining shape to which machining units have not been assigned, the machining unit assignment unit 15 ends the machining unit assignment process.

FIG. 17 is a flow chart showing the details of the machining shape dividing process for the stepped surface machining shape performed by the machining shape expansion unit 14 according to the first embodiment.

As described above, the machining shape expansion unit 14 stores the product shape data stored in the product shape storage unit 11, the material shape data stored in the material shape storage unit 12, and the work origin/machining direction storage unit 13. A machining shape to be machined by cutting is extracted based on the work coordinate system including the work origin and the machining direction. When dividing a facing shape in the extracted machining shape into a plurality of facing process shapes, if a stepped facing shape exists, the following processing is performed. FIG. 18 is a perspective view illustrating a stepped surface machining shape SH12 processed by the machining shape development unit 14 according to the first embodiment. The stepped surface processing shape SH12 corresponds to the stepped recess KH4 formed in the rectangular parallelepiped KH shown in FIG. In FIG. 18, the upper surface coordinates are "75.0, 75.0, 0.0", the lower surface coordinates are "25.0, 25.0, -20.0", the X-axis direction dimension is "50.0", and the Y-axis direction Dimension "50.0", Z-axis direction dimension "20.0", upper cross-sectional area "24785.4", lower cross-sectional area "12285.4", upper depth dimension "10.0", lower A step facing shape SH12 having a depth dimension of "10.0" is shown.

In step S61, the machining shape development unit 14 sets the upper surface coordinates of "75.0, 75.0, 0.0" of the stepped surface machining shape SH12 and the lower surface coordinates of "75.0, 75.0, 0.0" of the stepped surface machining shape SH12 as the second parameters. 25.0, 25.0, -20.0", X-axis dimension of stepped facing shape SH12 "50.0", Y-axis dimension of stepped facing shape SH12 "50.0", stepped The Z-axis direction dimension of the facing shape SH12 is "20.0", the upper cross-sectional area of the stepped facing shape SH12 is "24785.4", the lower cross-sectional area is "12285.4", and the upper depth dimension is "10. 0” and depth dimension “10.0” in the lower row.

In step S62, the machining shape expansion unit 14 designates the division method of the stepped surface machining shape SH12 as the first parameter, and inputs the input data including the second parameter generated in step S61 to the inference device 30. . The inference device 30 infers the division direction of the machining shape as a first parameter, and outputs the inference result to the machining shape expansion unit 14 . Here, there are two methods for dividing the stepped surface machining shape SH12: "horizontal division" for dividing the shape in the horizontal direction and "vertical division" for dividing the shape in the vertical direction.

In step S63, the machining shape expansion unit 14 divides the stepped surface machining shape SH12 into either horizontal division or vertical division according to the inference result. FIG. 19 is a perspective view exemplifying the surface machining shape obtained by the horizontal division performed by the machining shape development unit 14 according to the first embodiment. FIG. 20 is a perspective view exemplifying the surface machining shape obtained by the vertical division performed by the machining shape development unit 14 according to the first embodiment. In FIG. 19, by horizontally dividing the stepped surface machining shape SH12, a surface process shape SH13 and a surface process shape SH14 are obtained. In FIG. 20, by vertically dividing the stepped surface processing shape SH12, a surface process shape SH15 and a surface process shape SH16 are obtained.

FIG. 21 is a flowchart showing details of machining program generation processing performed by the machining program generation unit 17 according to the first embodiment. The processing in FIG. 21 corresponds to the processing in steps S6 and S7 in FIG.

In step S71, the work origin/machining direction storage unit 13 reads the stored work origin and work coordinate system.

In step S72, the machining process storage unit 16 generates shape data based on the workpiece origin and the workpiece coordinate system that have been read, and the machining shape. Shape data is generated from the machining shape by generating an outline shape by XY-projecting the machining shape onto the work coordinate system and converting the generated outline shape into straight line and circular arc data.

In step S73, the machining process storage unit 16 generates parameters necessary for the machining program such as the machining depth of the machining shape and the machining allowance in the Z direction based on the read work coordinate system including the work origin and the machining shape. Then, a machining program is generated based on the generated parameters and the shape data generated in step S72.

In step S74, the machining program generator 17 determines whether or not there is a machining shape for which a machining program has not yet been generated. If there is no machining shape for which the machining program has not yet been generated, the machining program generation process is terminated.

FIG. 22 is a flowchart showing details of uncut shape generation processing performed by the uncut shape generation unit 18 according to the first embodiment. The processing in FIG. 22 corresponds to step S8 in FIG.

In step S<b>81 , the uncut shape generator 18 reads the product shape data stored in the product shape storage unit 11 .

In step S<b>82 , the uncut shape generation unit 18 reads the material shape data stored in the material shape storage unit 12 .

In step S<b>83 , the uncut shape generation unit 18 reads the facing shape and the hole processing shape stored in the machining process storage unit 16 .

In step S84, the uncut shape generating unit 18 generates a uncut shape from the product shape data, the material shape data, the facing shape, and the drilling shape. The uncut shape is generated by removing the product shape from the material shape to generate the machining shape, and excluding the facing shape and the hole machining shape from the generated machining shape. FIG. 23 is a perspective view illustrating an uncut shape generated by the uncut shape generation unit 18 according to the first embodiment. The uncut shape SH20 shown in FIG. 23 is obtained by removing the product shape SH1 from the material shape SH2 to generate the machining shape SH3, and excluding the facing shape SH5 and the hole machining shape SH6 from the generated machining shape SH3. generated.

FIG. 24 is a flowchart showing details of the uncut shape generation process performed by the machining program generation device 10 according to the first embodiment. In step S91, the workpiece origin/machining direction storage unit 13 performs the processes of steps S31 to S34 described above, and performs the machining of the facing shape SH5 and the hole machining shape SH6. A work coordinate system including the work origin and a machining direction vector are set based on SH21 and the uncut shape SH20. The material shape SH21 is a shape obtained by removing the facing shape SH5 and the drilling shape SH6 from the material shape SH2. The material shape SH21 corresponds to the second material shape. FIG. 25 is a perspective view for explaining a work coordinate system including the work origin set by the work origin/machining direction storage unit 13 according to the first embodiment. FIG. 25 shows the material shape SH21 in the next process, the uncut shape SH20, a new coordinate system (X, Y, Z) with the new work origin Q1 as the origin, and a new machining direction vector V2. It is shown. Machining direction vector V2=(0.0, -1.0, 0.0). The work origin Q1 is set at a position of (5.0, 5.0, -45.0) in the world coordinate system. A machining direction vector V2 corresponds to the second machining direction. A new coordinate system (X, Y, Z) with the work origin Q1 as the origin corresponds to the second coordinate system including the second work origin.

Subsequently, the processing of step S4 described above, more specifically, the processing of steps S41 to S44 in FIG. It is developed into machining shapes for machining, machining shapes for line machining, and machining shapes for hole machining.

Subsequently, the processing of step S5 described above, more specifically, the processing of steps S51 to S55 in FIG. , tools and cutting conditions are determined.

Subsequently, the processing of steps S6 and S7 described above, more specifically, the processing of steps S71 to S74 in FIG. A machining program is generated based on.

Subsequently, the processing of step S8 described above, more specifically, the processing of steps S81 to S84 in FIG. 22 is executed, and a new uncut shape that cannot be machined even in the next step with the machining direction vector V2 as the machining direction is generated. generated.

FIG. 26 is a perspective view illustrating an example of a machined shape developed in the next step with the machining direction vector V2 as the machining direction in the first embodiment. FIG. 26 shows a pocket face unit machining process shape SH22, a pocket face unit machining process shape SH23, a pocket face unit machining process shape SH24, and a drill unit machining process shape SH25. Machining step shapes SH22 to SH25 shown in FIG. 26 correspond to the second machining shape.

27 is a perspective view showing a new uncut shape that cannot be machined in the next step in Embodiment 1. FIG. FIG. 27 shows a new uncut shape SH27 that cannot be machined even by machining based on the coordinate system (X, Y, Z) whose origin is the work origin Q1 shown in FIG. 25 and the machining direction vector V2. .

A new coordinate system (X, Y, Z) with the new workpiece origin as the origin and a new machining direction vector are set, although the following description of the machining of the uncut shape SH27 is omitted. A machining program for the uncut shape SH27 is generated by repeating the processing shown in FIG.

In the processing of FIGS. 22 and 24 described above, the uncut shape is generated, and a machining program for the uncut shape portion is generated using the generated uncut shape. It is also possible to generate a machining program for the shape portion. 28 is a flowchart showing processing for generating a machining program without generating an uncut shape in Embodiment 1. FIG.

In step S101, the material shape storage unit 12 stores the surface machining shape SH5 and the hole machining shape SH6 generated based on the work coordinate system (X, Y, Z) having the first work origin Q0 as the origin and the machining direction vector V1. to get FIG. 29 is a perspective view illustrating examples of a facing shape and a hole machining shape for which machining programs have been generated up to the present stage in Embodiment 1. FIG. FIG. 29 shows surface processing shapes SH7, SH8, SH10, and SH11 developed from the surface processing shape SH5, and two hole processing shapes SH6.

In step S102, the material shape storage unit 12 removes the obtained surface process shapes SH7, SH8, SH10, and SH11 and the two drilling shapes SH6 from the material shape SH2, thereby obtaining the next process material shape SH21. to generate 30 is a perspective view showing the material shape SH21 in the next step in Embodiment 1. FIG. The material shape SH21 shown in FIG. 30 and the material shape SH21 shown in FIG. 25 are the same.

In step S103, the work origin/machining direction storage unit 13 stores the work including the work origin in the next process based on the material shape SH21 and the product shape SH1 in the next process according to the processing of steps S31 to S34 in FIG. Set the coordinate system and machining direction vector.

Subsequently, the machining shape developing section 14 performs machining shape development processing based on the flowchart of FIG. 10, and the machining unit allocation section 15 performs machining unit allocation based on the flowchart of FIG. 16 to generate a machining program. In this way, a machining program can be generated without generating an uncut shape.

FIG. 31 is a flow chart showing the procedure of machining direction determination processing performed by the machining program generation device 10 of the first embodiment. Although the operator may manually set the machining direction, FIG. 31 shows the machining direction determination process when the machining direction is automatically determined. As described above, in the machining direction determination process, the machining direction is determined based on the product shape, material shape, and uncut shape.

In step S<b>111 , the workpiece origin/machining direction storage unit 13 stores the product shape data stored in the product shape storage unit 11 , the material shape data stored in the material shape storage unit 12 , and the uncut shape generation unit 18 . Reads the uncut shape generated by .

In step S112, the work origin/machining direction storage unit 13 extracts the machining direction based on the product shape read in step S111. Machinable directions are directions that can be machined by flat machining and hole machining. For the directions that can be machined by hole machining, the cylindrical surface and conical surface that form the hole shape are extracted from the product shape, and the axial direction vectors of the cylindrical surface and the conical surface are extracted. For the directions that can be machined by surface machining, a plane that can be the bottom surface of the surface machining is extracted from the product shape. In order to extract a plane that can be the bottom surface of flat processing from the product shape, it is necessary to extract a concave shape in which the wall surface is perpendicular to the bottom surface and the bottom surface is flat, and a concave shape that has no bottom surface and penetrates through the product shape. . If there is a plane that serves as the bottom surface, the opposite direction of the normal vector of the plane becomes the processing direction. In the case of a penetrating concave shape without a flat surface serving as a bottom surface, the working direction is the positive direction and the opposite direction of the cross product vector of the normal vectors of the wall surfaces of a plurality of flat surfaces. In the case of a wall surface of a cylindrical surface, the positive direction and the opposite direction of the axial direction vector of the cylindrical surface are the processing directions. Furthermore, not only the concave shape but also the portion where the uppermost surface of the convex shape is flat is extracted. Also in this case, the direction opposite to the normal vector of the plane is the machining direction.

In step S113, the workpiece origin/machining direction storage unit 13 calculates the removal volume of the machining shape of the next step based on the machining shape of the next step for each machining direction extracted in step S112. For the removal volume of the machining shape for a specific machining direction, a plane perpendicular to the specific machining direction is extracted from the machining shape of the next step. Next, the highest position of the machining shape in the next process is obtained with respect to a specific machining direction, and the volume of the shape obtained by sweeping the extracted plane to the highest position of the machining shape is obtained. should be calculated. If there are a plurality of machining directions, the machining direction that allows cutting with the largest removal volume (cutting volume) is determined as the machining direction for the next step. Alternatively, the processing direction for the next step may be determined based on the area of the extracted plane without sweeping.

Alternatively, the order of the machining directions may be determined in advance without extracting the machining directions in step S112, and the machining directions may be determined in the order of the machining directions. For example, the machining direction is (0,0,-1), (1,0,0), (0,1,0), (-1,0,0), (0,-1,0), (0 , 0, 1), and the order of these six machining directions may be determined in advance.

FIG. 32 is a flowchart showing the procedure of learning model generation processing performed by the machine learning device 20 according to the first embodiment. In the learning model generation process, a learning model for generating a machining program is generated based on the machining program 2 . The machining program 2 is, for example, a machining program created in the past by the operator, as described above.

In step S121, the machining program storage unit 21 reads and stores a plurality of machining programs 2 from a storage area (not shown). Also, if there is shape data associated with the machining program 2, the shape data is read and stored.

In step S<b>122 , the machining program analysis unit 22 extracts the first parameter from the machining program 2 . A machining program analysis unit 22 extracts a plurality of first parameters used in the machining program 2 .

In step S123, the machining program analysis unit 22 extracts a second parameter for each of the plurality of extracted first parameters. At this time, the machining program analysis unit 22 determines the second parameter to be extracted from the machining program 2 or the shape data for each first parameter, and extracts the determined second parameter. The machining program analysis unit 22 performs the processing of steps S122 and S123 for each machining program 2 . The machining program analysis unit 22 inputs the extracted first parameter and second parameter to the machine learning unit 23 for each machining program 2 .

In step S124, the machine learning unit 23 performs machine learning processing using the input first and second parameters. Machine learning unit 23 generates a data set based on the first parameter and the second parameter, and performs machine learning according to the generated data set. A data set is a set of data in which a first parameter to be adjusted and a second parameter, which is a non-adjustable parameter used to determine the value of the first parameter, are associated with each other. The machine learning unit 23 generates an optimized model as a learning model using predetermined criteria. The machine learning unit 23 generates a learning model as a learning result. The machine learning model storage unit 24 stores the generated learning model. With the above, the machine learning device 20 ends the learning model generation process.

FIG. 33 is a flow chart showing the procedure of undeveloped shape process development processing performed by the machining shape development unit 14 according to the first embodiment. In the process of FIG. 33, machining shape development is executed for the surface process shape SH9 (FIG. 15) that has been set as an unassigned undeveloped shape in the process of FIG.

In steps S51 and S52 in FIG. 16, the machining shape expansion unit 14 extracts the product shape data stored in the product shape storage unit 11, the material shape data stored in the material shape storage unit 12, the workpiece origin and machining Based on the work coordinate system including the work origin and the machining direction stored in the direction storage unit 13, the machining shape to be machined by cutting is extracted, and the machining shape is divided as shapes for plane machining and hole machining, and plane machining is performed. I assigned the process and the drilling process. In step S131 of FIG. 33, the machining shape development unit 14 reads the surface process shape SH9 (FIG. 15), which remains as an undeveloped shape without being developed, as an undeveloped shape.

In step S132, the machining shape developing section 14 generates a surface machining shape for rough machining from the undeveloped shape SH9. To remove the curved surface from the undeployed shape SH9, the shape may be divided at the top of the curved surface in the processing direction, or the curved surface may be divided into a plurality of pieces, and the shape may be divided at the top of the divided curved surface in the processing direction. FIG. 34 is a perspective view showing a plurality of surface process shapes obtained by the processing for developing surface process shapes for rough machining in the machining shape developing unit 14 according to the first embodiment. 34, the plane process shape SH9 of FIG. 15 is divided into a plane process shape SH31, a plane process shape SH32, a plane process shape SH33, and four curved process shapes SH34.

Next, the machining unit allocation section 15 allocates a machining unit to each process shape developed by the machining shape development section 14 . For example, the end mill crest unit is assigned to the planar process shape SH31, the planar process shape SH32, and the planar process shape SH33 in FIG. The curved surface process shape SH34 is an undeveloped shape. The undeployed shape allows processing in other processing directions.

As described above, according to Embodiment 1, a machining shape is generated based on the product shape and the material shape, and the first machining shape for machining the first machining shape that can be machined in the first machining direction selected from the machining shape. 1 generating a machining program, generating an uncut shape that cannot be machined by the machining program in the first machining direction, selecting a second machining direction for machining the uncut shape, selecting the selected second machining direction and cutting A second machining program for machining a second machining shape that can be machined in the second machining direction is generated based on the remaining shape. Therefore, it is possible to efficiently generate a machining program for the uncut portion in a short period of time without giving the operator troublesome work.

At this time, the uncut shape may be derived directly by actual calculation as shown in FIG. It may be derived indirectly based on the processed shape.

Further, according to Embodiment 1, when generating machining programs for a plurality of machining directions, the machining program is generated after resetting the material shape from the machining program for the machining direction up to the last time. Therefore, it is possible to generate a machining program that does not overlap with the immediately preceding machining program, and efficiently generate a machining program.

Further, according to Embodiment 1, machining programs for a plurality of machining directions are generated. The operator can generate a machining program for multiple machining directions simply by inputting the product shape, setting the material shape, and setting a workpiece coordinate system including multiple workpiece origins and multiple machining directions. Therefore, it is possible to reduce the labor and time required to generate machining programs for a plurality of machining directions.

Further, according to Embodiment 1, the learning model generated by the machine learning device 20 is used to infer the first parameter. The operator can easily set the values of the plurality of first parameters using the inference result. The machining program generation device 10 can generate a machining program by utilizing a plurality of knowledge and experiences accumulated in the machining program created in the past by the operator. Therefore, the machining program generation device 10 can easily generate a high-quality machining program desired by the operator.

Although the machining program generation device 10, the machine learning device 20 and the reasoning device 30 are incorporated in the same numerical control device 100 in the first embodiment, the present invention is not limited to this example. The machining program generation device 10 , the machine learning device 20 and the inference device 30 may be provided independently outside the numerical control device 100 . In the first embodiment, the machining program for the case where the numerically controlled machine tool is a machining center has been described as an example, but the numerically controlled machine tool is not limited to the machining center, and may be another machine tool. .

Embodiment 2.
In Embodiment 2, hardware configurations of the machining program generation device 10, the machine learning device 20, and the inference device 30 will be described. FIG. 35 is a block diagram showing the hardware configuration of the machining program generation device 10, the machine learning device 20, and the inference device 30 according to the second embodiment. Each functional unit shown in FIG. 35 includes a processor 71, a memory 72 used as a work area by the processor 71, a storage device 73 storing a computer program describing each function of the numerical controller 100, and an operator. It comprises an input device 74 as an input interface, a display device 75 as an output device for displaying information to the operator, and a communication device 76 having a function of communicating with controlled devices or other numerical control devices. Processor 71 , memory 72 , storage device 73 , input device 74 , display device 75 and communication device 76 are interconnected by data bus 77 .

The processor 71 is a processing device, an arithmetic device, a microprocessor, a microcomputer, a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or the like. The memory 72 is a nonvolatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable ROM), or EEPROM (Registered Trademark) (Electrically EPROM); They include magnetic discs, flexible discs, optical discs, compact discs, mini discs, and DVDs (Digital Versatile Discs).

The product shape storage unit 11, the material shape storage unit 12, the workpiece origin/machining direction storage unit 13, the machining shape expansion unit 14, the machining unit allocation unit 15, and the machining process storage unit 16 store computer programs stored in the memory 72. 71 reads and executes it. Also, the above functions may be realized by cooperation of a plurality of processors 71 and a plurality of memories 72 . Product shape storage unit 11, material shape storage unit 12, work origin/machining direction storage unit 13, machining shape expansion unit 14, machining unit allocation unit 15 and machining process storage unit 16, machining program generation unit 17, uncut shape generation unit Some of the 18 functions may be implemented as electronic circuits, and other parts may be implemented using the processor 71 and memory 72 . A processor 71 and a memory 72 for realizing the functions of the product shape storage unit 11, the material shape storage unit 12, the workpiece origin/machining direction storage unit 13, the machining shape development unit 14, the machining unit allocation unit 15, and the machining process storage unit 16. may be the same as the processor 71 and the memory 72 for realizing the machine learning unit 23 and the inference unit 31, or may be different from the processor 71 and the memory 72 for realizing the machine learning unit 23 and the inference unit 31 A processor 71 and memory 72 may be used. The function of the product shape storage section 11 is implemented by the communication device 76 . The functions of the product shape memory unit 11 , the material shape memory unit 12 and the workpiece origin/machining direction memory unit 13 are implemented by the storage device 73 .

The machining program storage unit 21 can be realized by the processor 71 reading and executing a computer program stored in the memory 72 . The machining program analysis unit 22, the machine learning unit 23, and the inference unit 31 can be realized by the processor 71 reading and executing a computer program stored in the memory 72, for example. Also, the above functions may be realized by cooperating with a plurality of processors 71 and a plurality of memories 72 . Also, some of the functions of the machine learning unit 23 and the inference unit 31 may be implemented as an electronic circuit, and the processor 71 and the memory 72 may be used for other parts to realize the above functions. Functions of the machining program storage unit 21 are implemented by the communication device 76 . The function of the machine learning model storage unit 24 is implemented by the storage device 73 .

The configuration shown in the above embodiment shows an example of the content of the present disclosure, and can be combined with another known technology. It is also possible to omit or change the part.

1 CAD data, 2 machining program, 10 machining program generation device, 11 product shape storage unit, 12 material shape storage unit, 13 workpiece origin/machining direction storage unit, 14 machining shape expansion unit, 15 machining unit allocation unit, 16 machining process Storage unit 17 Machining program generation unit 18 Uncut shape generation unit 20 Machine learning device 21 Machining program storage unit 22 Machining program analysis unit 23 Machine learning unit 24 Machine learning model storage unit 30 Inference device 31 Inference unit 40 Instruction input unit 50 Dialogue operation processing unit 60 Display unit 71 Processor 72 Memory 73 Storage device 74 Input device 75 Display device 76 Communication device 77 Data bus 100 Numerical control device C Visible surface, K1 Top surface, KH Cuboid, KH1, KH2 Concave portion, KH3, RH1 Round hole, KH4 Stepped concave portion, Q Candidate point, Q0, Q1 Work origin, RH Frame body, SH1 Product shape, SH2, SH21 Material shape, SH3, SH4 processing shape, SH5 surface processing shape, SH6 hole processing shape, SH7, SH8 to SH10, SH11, SH13 to SH16 surface processing shape, SH9 undeveloped shape (surface processing shape), SH12 stepped surface processing shape, SH20, SH27 Uncut shape, SH22 to SH25 machining process shape, SH31 to SH33 planar process shape, SH34 curved surface process shape, V1, V2 machining direction vector.

Claims (6)

A machining program generation device for generating a machining program including a plurality of cutting processes for cutting a machined product from an object to be machined,
a product shape storage unit that stores the product shape of the cut product;
a material shape storage unit that stores a first material shape, which is a material shape before machining of the cut product;
A machining shape is generated based on the first material shape and the product shape, and a first work coordinate system including a first work origin of the cut product and a first machining direction are selected based on the generated machining shape. a workpiece origin/machining direction storage unit to be stored;
A first machining shape that can be machined in the first machining direction is extracted from the machining shape based on the selected first machining direction, the product shape, and the first material shape, and the extracted first machining shape A machining program generation unit that generates a first machining program for machining the
an uncut shape generating unit that removes the product shape from the first material shape to generate the machining shape , and removes the first machining shape from the generated machining shape to generate an uncut shape;
with
The work origin/machining direction storage unit removes the first machining shape from the first material shape to generate a second material shape, and based on the generated second material shape and the uncut shape, the selecting a second workpiece coordinate system including a second workpiece origin and a second machining direction for machining the uncut shape;
The machining program generation unit extracts from the machining shape a second machining shape that can be machined in the second machining direction based on the selected second machining direction and the uncut shape, and extracts the extracted second machining A machining program generation device for generating a second machining program for machining a shape. A machining program generation device for generating a machining program including a plurality of cutting processes for cutting a machined product from an object to be machined,
a product shape storage unit that stores the product shape of the cut product;
a material shape storage unit that stores a first material shape, which is a material shape before machining of the cut product;
A machining shape is generated based on the first material shape and the product shape, and a first work coordinate system including a first work origin of the cut product and a first machining direction are selected based on the generated machining shape. a workpiece origin/machining direction storage unit to be stored;
A first machining shape that can be machined in the first machining direction is extracted from the machining shape based on the selected first machining direction, the product shape, and the first material shape, and the extracted first machining shape A machining program generation unit that generates a first machining program for machining the
with
The material shape storage unit removes the first processed shape from the first material shape to generate a second material shape,
The workpiece origin/machining direction storage unit stores a second workpiece coordinate system and a second machining direction that include a second workpiece origin for machining an uncut shape based on the generated second material shape and product shape. to select
The machining program generation unit extracts a second machining shape that can be machined in the second machining direction based on the selected second machining direction, the second material shape, and the product shape, and extracts the extracted second machining shape. 2. A machining program generation device for generating a second machining program for machining two machining shapes. 2. When there are a plurality of second machining directions, the workpiece origin/machining direction storage unit determines a machining direction in which cutting with the largest cutting volume is possible as the second machining direction. 3. Or the machining program generation device according to 2 . The machining program generation unit is
By analyzing the first machining program, the first parameter to be adjusted in the generation of the first machining program, and the first parameter not to be adjusted in the generation of the first machining program A machining program analysis unit that extracts from the first machining program a second parameter used for adjusting the
A learning model for inferring the value of the first parameter from the second parameter of the first machining program by learning using a data set containing the extracted first parameter and the second parameter a machine learning unit that generates
The machining program generation device according to claim 1 or 2 , characterized by comprising: A machining program generation method for generating a machining program including a plurality of cutting processes for cutting out a machined product from an object to be machined,
a step of storing a product shape of the machined product;
a step of storing a first material shape, which is a material shape before machining of the cut product;
A machining shape is generated based on the first material shape and the product shape, and a first work coordinate system including a first work origin of the cut product and a first machining direction are selected based on the generated machining shape. and storing
A first machining shape that can be machined in the first machining direction is extracted from the machining shape based on the selected first machining direction, the product shape, and the first material shape, and the extracted first machining shape generating a first machining program for machining the
removing the product shape from the first material shape to generate the machining shape , and removing the first machining shape from the generated machining shape to generate an uncut shape;
a second material shape for removing the first machining shape from the first material shape to generate a second material shape, and machining the uncut shape based on the generated second material shape and the uncut shape; selecting a second work coordinate system including a work origin and a second machining direction;
A second machining shape that can be machined in the second machining direction is extracted from the machining shape based on the selected second machining direction and the uncut shape, and a second machining shape for machining the extracted second machining shape 2 generating a machining program;
A machining program generation method comprising: A machining program generation method for generating a machining program including a plurality of cutting processes for cutting out a machined product from an object to be machined,
a step of storing a product shape of the machined product;
a step of storing a first material shape, which is a material shape before machining of the cut product;
A machining shape is generated based on the first material shape and the product shape, and a first work coordinate system including a first work origin of the cut product and a first machining direction are selected based on the generated machining shape. and storing
A first machining shape that can be machined in the first machining direction is extracted from the machining shape based on the selected first machining direction, the product shape, and the first material shape, and the extracted first machining shape generating a first machining program for machining the
removing the first machining shape from the first material shape to generate a second material shape;
selecting a second work coordinate system including a second work origin and a second machining direction for machining the uncut shape, based on the generated second material shape and product shape;
To extract a second machining shape that can be machined in the second machining direction based on the selected second machining direction, the second material shape, and the product shape, and to machine the extracted second machining shape a step of generating a second machining program of
A machining program generation method comprising:

Images