US 12,228,907 B2
Chip processing device for machine tool and chip processing method
Masahiro Shimoike, Nara (JP)
Assigned to DMG MORI CO., LTD., Nara (JP)
Appl. No. 17/439,347
Filed by DMG MORI CO., LTD., Nara (JP)
PCT Filed Mar. 13, 2020, PCT No. PCT/JP2020/011068
§ 371(c)(1), (2) Date Sep. 14, 2021,
PCT Pub. No. WO2020/189547, PCT Pub. Date Sep. 24, 2020.
Claims priority of application No. 2019-047865 (JP), filed on Mar. 15, 2019.
Prior Publication US 2022/0179390 A1, Jun. 9, 2022
Int. Cl. G05B 19/402 (2006.01); B23Q 11/00 (2006.01); G05B 19/18 (2006.01); G05B 19/4093 (2006.01); G06N 20/00 (2019.01)
CPC G05B 19/402 (2013.01) [B23Q 11/0067 (2013.01); G05B 19/182 (2013.01); G05B 19/40938 (2013.01); G06N 20/00 (2019.01); G05B 2219/49042 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A chip processing device for a machine tool, comprising
a cleaning nozzle control unit that controls a cleaning nozzle for spraying a cleaning fluid onto chips scattered during machining so as to guide the chips to a chip collection portion,
wherein the cleaning nozzle control unit includes:
a position estimation unit that estimates an accumulation position of chips to be generated, before the chips are generated, by analyzing machining conditions of a workpiece that are conditions for machining the workpiece with a tool; and
a nozzle orientation adjustment unit that adjusts an orientation of the cleaning nozzle toward the accumulation position estimated by the position estimation unit;
wherein the position estimation unit includes a machine learning device from which an accumulation position of the chips is output when a machining parameter included in an NC program used for machining and an in-machine shape that is an internal shape of the machine tool are input to the machine learning device;
wherein the machine learning device comprises a primary learned device that has been subjected to primary learning in advance, using, as input data, the machining parameter and the in-machine shape, and using, as training data, the accumulation position of the chips calculated based on the scattering trajectory of the chips obtained by the FEM analysis based on the machining parameter and the in-machine shape;
wherein the machine learning device comprises a secondary learned device obtained by subjecting the primary learned device to secondary learning, using, as training data, an accumulation position of the chips obtained from captured images of an interior of the machine tool before and after machining.