Online robust parameter design considering observable noise factors

Abstract

Variation is the root cause of product quality problems. Variations caused by noise factors often affect the quality of the product during manufacturing, resulting in products that fail to meet customer requirements. In this article, a Bayesian online robust parameter design method is proposed to address the parameter design problem with observable noise factors. The proposed method not only takes into account the online observations of noise factors through using a time-series model, but also adjusts the optimal settings of the control factors in real time to reduce the impact of the variations from noise factors on product quality by using a response surface model and an expected loss function. The advantages of the proposed method are illustrated by a numerical example and a case study. The results show that the proposed method can give more reasonable results than the existing methods.

KEYWORDS:

• Robust parameter design
• Bayesian inference
• noise factors
• multivariate time-series model

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (NSFC) [grant numbers 71771121, 71931006].