A physics-informed Run-to-Run control framework for semiconductor manufacturing

作者：

Highlights：

• A new R2R control framework based on Dynamic Bayesian Network is proposed.

• Physic-informed DBN is constructed based on equipment sensor data and prior knowledge.

• A causal structure enables analyzing the underlying process interactions.

• Validation results with the real data confirm the usefulness of the proposed model.

摘要

•A new R2R control framework based on Dynamic Bayesian Network is proposed.•Physic-informed DBN is constructed based on equipment sensor data and prior knowledge.•A causal structure enables analyzing the underlying process interactions.•Validation results with the real data confirm the usefulness of the proposed model.

论文关键词：Advanced Process Control (APC),Chemical-Mechanical Polishing (CMP),Dynamic Bayesian Network (DBN),Fault Detection and Classification (FDC),Physics-informed,Run-to-Run (R2R) control

论文评审过程：Received 27 May 2019, Revised 12 March 2020, Accepted 30 March 2020, Available online 28 April 2020, Version of Record 7 May 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113424