Method and device for generating tool paths

Patent metadata

Field	Value
Publication number	US-2021003992-A1
Application number	US-201916976931-A
Country	US
Kind code	A1
Filing date	Feb 14, 2019
Priority date	Mar 1, 2018
Publication date	Jan 7, 2021
Grant date	—

Abstract

The step for performing machine learning includes acquiring shape data; acquiring geometric information for each of a plurality of machining faces; acquiring a tool path pattern selected for the machining faces from among a plurality of tool path patterns; and performing machine learning by using the geometric data for known workpieces and the tool path patterns wherein the input is the geometric information for the machining faces and the output is the tool path pattern for the machining faces. The step for generating a new tool path includes: acquiring shape data for the workpiece; acquiring geometric information for each of the plurality of machining faces of the workpiece to be machined; and generating a tool path pattern for each of the plurality of machining faces on the workpiece on the basis of the results of the machine learning using the geometric information of the workpiece to be machined.

First claim

1 . A method for generating a tool path in NC machining, the method comprising the steps of:
performing machine learning based on information of a plurality of known workpieces having already generated tool paths, and generating a new tool path for a target workpiece based on results of the machine learning, wherein each of the plurality of known workpieces and the target workpiece has a plurality of machining surfaces, the step in which machine learning is performed includes:
for each of the plurality of known workpieces, obtaining shape data,
for each of the plurality of known workpieces, obtaining geometric information of each of the plurality of machining surfaces,
for each of the plurality of known workpieces, obtaining a tool path pattern selected for each of the plurality of machining surfaces from among a plurality of tool path patterns, and
performing machine learning in which input is geometric information of a machining surface and output is a tool path pattern of the machining surface, using the geometric information and the tool path patterns of the plurality of known workpieces, and
the step in which a new tool path pattern is generated includes:
obtaining shape data of the target workpiece,
obtaining geometric information of each of the plurality of machining surfaces of the target workpiece, and
generating a tool path pattern for each of the plurality of machining surfaces of the target workpiece based on results of the machine learning, using the geometric information of the target workpiece.

2 . The method according to claim 1 , wherein the plurality of tool path patterns include at least a contour path, a scanning line path, and a surface path.

3 . The method according to claim 1 , wherein the shape data of the plurality of known workpieces and the shape data of the target workpiece are defined in an XYZ coordinate system which is a three-dimensional cartesian coordinate system, and
the geometric information includes at least one of a machining surface type, a ratio of a maximum length in an X-axis direction to a maximum length in a Z-axis direction of each machining surface, a ratio of a maximum length in a Y-axis direction to a maximum length in the Z-axis direction of each machining surface, a ratio of a Z-axis direction maximum length of the entirety of the plurality of machining surfaces to a Z-axis direction maximum length of each machining surface, a ratio of a surface area of the entirety of the plurality of machining surfaces to a surface area of each machining surface, a long radius of a machining surface, a short radius of a machining surface, a Z component of a normal vector at a center of gravity of a machining surface, a maximum curvature of a machining surface, and a minimum curvature of a machining surface.

4 . The method according to claim 1 , wherein a neural network is used in the machine learning.

5 . A device for generating a tool path in NC machining, the device comprising:
a processor, and a display unit, wherein the processor is configured so as to:
perform machine learning based on information of a plurality of known workpieces having already generated tool paths, and
generate a new tool path for a target workpiece based on results of the machine learning, wherein
each of the plurality of known workpieces and the target workpiece has a plurality of machining surfaces, the step in which machine learning is performed includes:
for each of the plurality of known workpieces, obtaining shape data,
for each of the plurality of known workpieces, obtaining geometric information of each of the plurality of machining surfaces,
for each of the plurality of known workpieces, obtaining a tool path pattern selected for each of the plurality of machining surfaces from among a plurality of tool path patterns, and
performing machine learning in which input is geometric information of a machining surface and output is a tool path pattern of the machining surface, using the geometric information and the tool path patterns of the plurality of known workpieces,
the step in which a new tool path pattern is generated includes:
obtaining shape data of the target workpiece,
obtaining geometric information of each of the plurality of machining surfaces of the target workpiece, and
generating a tool path pattern for each of the plurality of machining surfaces of the target workpiece based on results of the machine learning, using the geometric information of the target workpiece
each of the plurality of tool path patterns is assigned a predetermined feature which can be visually distinguished, the processor recognizes the plurality of tool path patterns as the predetermined feature, and the display unit displays each of the machining surfaces along with the predetermined feature corresponding to a generated tool path pattern.

6 . The device according to claim 5 , wherein the processor is configured so as to calculate a confidence factor for the tool path pattern generated for each of the plurality of machining surfaces of the target workpiece, and
the display unit emphasizes a corresponding machining surface when the confidence factor is less than a predetermined threshold.

Assignees

• Makino Milling Machine

Inventors

• Nakamoto Keiichi
• Hashimoto Mayu
• Kono Kazumasa
• Takei Katsuhiko
• Igari Shinji

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/084

  Backpropagation, e.g. using gradient descent · CPC title

• G05B2219/35097

  Generation of cutter path, offset curve · CPC title

• G05B2219/33027

  Artificial neural network controller · CPC title