CAREER: Multivariate Quality Control of Semiconductor Manufacturing Processes via Adaptive Optimizing Controllers

本研究旨在为先进的半导体制造过程开发有效的多变量质量控制（QC）方法。特别关注的是基于优化自适应控制思想的半导体制造过程监督控制的新型多输入多输出质量控制器的开发和实现。在许多这些过程中，一批或硅晶圆的质量特性由比例积分导数（PID）类型的自动控制器调节，该控制器调节诸如压力，流量，温度等变量。所提出的优化自适应质量控制器（OAQC）通过在线估计和优化非线性传递函数模型，提供每次运行时PID设定点的配方。OAQC将与Sematech的统计方法和高级过程控制小组密切合作开发和测试。该项目的教育部分与拟议的研究相结合，反映了PI教授现代QC课程的理念，这是应用统计学和控制理论交叉的跨学科领域。研究和综合教育计划深思熟虑，产业合作良好。PI已经有一个强大的记录，并被认为是非常胜任。研究领域，质量控制，是特别重要的半导体制造行业。

9623669 Del Castillo This research aims to develop effective multivariate quality control (QC) methods for advanced semiconductor manufacturing processes. Specifically of concern is the development and implementation of a novel multiple-input-multiple-output quality controller for supervisory control of semiconductor manufacturing processes based on the idea of optimizing adaptive control. In many of these processes, the quality characteristics of a batch or run of silicon wafers are regulated with automatic controllers of the proportional-integral-derivative (PID) type, that regulates variables such as pressures, flows, temperatures, etc. The proposed optimizing adaptive quality controller (OAQC) provides a recipe of PID set points at each run by the on-line estimation and optimization of a nonlinear transfer function model. The OAQC will be developed and tested in close collaboration with Sematech's Statistical Methods and Advanced Process Control Group. The educational component of this project is integrated with the proposed research and reflects the PI's philosophy on teaching modern QC courses an interdisciplinary field on the intersection between applied statistics and control theory. The research and integrated educational plan is well thought out and the industrial collaboration excellent. The PI already has a strong track record and is felt to be extremely competent. The area of research, QC, is of particular importance to the semiconductor manufacturing industry.