CAREER: Optimal Operation and Control of Multiscale Process Systems

摘要：Antonios Armaou机构：Pennsylvania State UniversityProposal Number：0644519 Title：CAREER：Optimal Operation and Control of Multiscale Process SystemsIntellectual merit精确调控关键工业产品的微观结构，如具有特定表面形态的薄膜和具有特定形状的生物晶体，是目前一个重要但尚未解决的研究问题，具有广泛的工业意义。 在严格限制内的微观产品特性的监管需要在制造特定产品的过程上开发和实施基于计算效率的模型的最优操作和控制策略。 将微观产品特性与可以自动调节的宏观过程变量联系起来的基本数学模型通常被称为多尺度过程系统，并且它们的解具有非常高的计算要求。 用于计算高性能操作和控制策略的计算高效技术应该利用多尺度过程系统的高精度但计算上易处理的近似。 出于这一动机，本研究的目的是解决与发展计算效率高的技术计算多尺度过程系统的最佳操作和控制策略，这将允许在一定的严格的置信限度内设置产品的微观结构相关的基本问题。 为实现这一目标，计划开展一些具体项目，包括：a）开发用于使用各种微观/多尺度目标的多尺度过程系统的优化和最佳操作的计算有效算法，B）开发多尺度过程系统的非线性、低阶、近似模型，以及构造能够处理非线性问题的实际可实现的反馈控制系统，模型的不确定性和限制，c）将最佳操作和控制方法应用于微电子制造工艺（特别是广泛用于生产绿色/蓝色激光器的GaN薄膜的金属有机气相外延（MOVPE）），以及d）将研究成果融入教育。 除了为多尺度系统的优化操作和控制提供计算效率高的方法外，该研究还将为多尺度过程系统的优化和控制问题的本质提供基本见解，并解决实际实施问题。有广泛的工业工艺和产品将受益于研究结果，包括微电子制造/先进材料工艺，如气相外延，化学气相沉积和蚀刻，其中膜微结构是重要的产品质量变量。 为了将拟议研究的结果和见解转移到工业部门并追求未来的实验实施，PI还将积极寻求与宾夕法尼亚州立大学的工业和实验教师的合作。 最后，计划开展一些活动，将拟议的研究与教育相结合，包括将研究成果纳入先进的运输现象和控制课程，通过荣誉计划参与研究的本科生，以及开发aLEGO Mindstorms实验室，以促进向代表性不足的群体的高中和大学生推广。

ABSTRACTPI: Antonios Armaou Institution: Pennsylvania State UniversityProposal Number: 0644519Title: CAREER: Optimal Operation and Control of Multiscale Process SystemsIntellectual meritPrecise regulation of the microstructure of key industrial products, such as thin films with specific surface morphology and biological crystals with specific shape, is currently an important yet unresolved research problem with broad industrial implications. Regulation of microscopic product properties within tight limits requires the development and implementation of computationally-efficient model-based optimal operation and control policies on the processes that make the specific product. Fundamental mathematical models that link microscopic product properties to macroscopic process variables that can be automatically regulated are usually referred to as multiscale process systems and their solution has very high computational requirements. Computationally-efficient techniques for calculating high-performance operation and control policies should utilize highly-accurate, yet computationally tractable approximations of multiscale process systems. Motivated by this, the objective of this research is to resolve the fundamental issues associated with the development of computationally-efficient techniques for computing optimal operation and control policies for multiscale process systems which will allow setting product microstructure within certain stringent confidence limits. To achieve this objective, a number of specific projects are planned including: a) development of computationally-efficient algorithms for optimization and optimal operation of multiscale process systems using a variety of microscopic/multiscale objectives, b) development of nonlinear, low-order, approximate models of multiscale process systems and construction of practically-implementable feedback control systems that can deal with the issues of nonlinearity, model uncertainty and constraints, c) application of the optimal operation and control methods to micro- electronics manufacturing processes (in particular, metal-organic vapor phase epitaxy (MOVPE) of GaN thin films widely used for the production of green/blue lasers), and d) integration of the= research results into education. In addition to computationally-efficient methods for optimal operation and control of multiscale systems, the research will provide fundamental insight into the nature of the optimization and control problems for multiscale process systems and address practical implementation issues.Broader Impact:There is a wide range of industrial processes and products that will benefit from the results of theresearch including microelectronics manufacturing/advanced materials processes such as vapor phase epitaxy, chemical vapor deposition and etching, where film microstructure is an important product quality variable. To transfer the results and insight of the proposed research to the industrial sector and pursue future experimental implementation, the PI will also actively seek collaborations with industry and experimental faculty at Penn State. Finally, a number of activities are planned to integrate the proposed research with education, including incorporation of research results in the advanced transport phenomena and control courses, undergraduate student participation in the research through the honors program, and the development of aLEGO Mindstorms laboratory to facilitate outreach to high-school and college students from under-represented groups.