Autonomous adaptive control of manufacturing parameters based on local regression modeling

Abstract

AbstractThe demand for the autonomous adaptive control of manufacturing lines has been growing to realize productivity improvement and carbon neutrality. We propose a methodology to realize the autonomous control of product quality under the existence of effects from non-measurable parameters. By local linear regression modeling with temporal neighborhood data, a single manufacturing parameter is selected by the obtained regression coefficients. Simulation results demonstrated that the straightforward multiple regression modeling often resulted in unstable control behavior with vibrations in product quality. We tested three approaches (scaling the amount of control, introduction of control interval, and semiparametric regression modeling for regression) to cope with the unstable behavior. The semiparametric regression model exhibited the best performance in realizing the stable control by correctly selecting the control target parameter with the additional non-linear term, that compensates the time-dependent non-measurable parameters.