US 12,214,462 B2
Monitoring method and system for machine tool
Chun-Yu Tsai, New Taipei (TW); Chi-Chen Lin, Taichung (TW); Sheng-Ming Ma, Taichung (TW); and Ta-Jen Peng, Taichung (TW)
Assigned to INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE, Hsinchu (TW)
Filed by INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE, Hsinchu (TW)
Filed on Apr. 27, 2021, as Appl. No. 17/241,479.
Claims priority of application No. 109145291 (TW), filed on Dec. 21, 2020.
Prior Publication US 2022/0193852 A1, Jun. 23, 2022
Int. Cl. B23Q 17/12 (2006.01); B23Q 17/09 (2006.01); G01B 17/08 (2006.01); G05B 19/18 (2006.01); G05B 19/4063 (2006.01); G06N 20/00 (2019.01)
CPC B23Q 17/12 (2013.01) [B23Q 17/0971 (2013.01); G01B 17/08 (2013.01); G05B 19/182 (2013.01); G05B 19/4063 (2013.01); G05B 2219/35529 (2013.01); G05B 2219/37434 (2013.01); G06N 20/00 (2019.01)] 6 Claims
OG exemplary drawing
 
1. A monitoring method for a machine tool to machine a workpiece, the monitoring method comprising:
detecting a vibration signal of a spindle of the machine tool;
obtaining a vibration feature value of the vibration signal, wherein obtaining the vibration feature value of the vibration signal comprises:
acquiring a time-domain signal of the vibration signal with a sampling frequency;
converting the time-domain signal into a frequency-domain signal;
obtaining at least one vibration value corresponding to at least one frequency equal to N multiple of a fundamental frequency of the frequency-domain signal, wherein N is an integer of 1 or more;
calculating a sum of the at least one vibration value; and
determining whether an additional vibration value exists beyond the N multiple of the fundamental frequency, wherein the additional vibration value is greater than the vibration value corresponding to 1 multiple of the fundamental frequency; and
if the additional vibration value exists, the vibration feature value equals the sum plus the additional vibration value; and
if the additional vibration value does not exist, the vibration feature value equals the sum;
determining whether the vibration feature value exceeds a threshold condition determined by a training model based on a predetermined surface quality of the workpiece, wherein the training model is trained with a machining quality database through a machine learning algorithm, and the machining quality database is established according to the following steps:
detecting a sample vibration signal of the spindle of the machine tool when machining a specimen with a machining condition, wherein the machining condition comprises a spindle speed and a feed speed of the machine tool;
obtaining a sample vibration feature value of the sample vibration signal;
measuring a sample surface quality of the specimen under the machining condition; and
recording the machining condition and the sample vibration feature value and the sample surface quality under the machining condition; and
adjusting a machining parameter of the machine tool when the vibration feature value exceeds the threshold condition, wherein adjusting the machining parameter of the machine tool comprises obtaining an optimized spindle speed of the machine tool, and obtaining the optimized spindle speed of the machine tool comprises:
setting a speed interval;
iteratively executing the following steps with different spindle speeds within the speed interval:
detecting a test vibration signal of the spindle of the machine tool when trial machining a test piece with a predetermined spindle speed; and
obtaining a test vibration feature value of the test vibration signal; and
determining the predetermined spindle speed corresponding to a minimum test vibration feature value as the optimized spindle speed.