Towards a Computationally Efficient Recursive Model Reduction and Controller Design Approach for Spatially Distributed Processes

针对空间分布式过程的计算高效的递归模型简化和控制器设计方法

• 批准号: 1300322
• 负责人: Antonios Armaou
• 金额: $ 33.95万
• 依托单位: Pennsylvania State Univ University Park
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2013
• 资助国家: 美国
• 起止时间: 2013-09-01 至 2017-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1300322&HistoricalAwards=false
• 关键词: Towards; Computationally; Efficient; Recursive; Model

The objective of this research project is to relax the requirements of data-based nonlinear order reduction methods and extend their applicability towards the control and optimization of dissipative partial differential equation (PDE) systems arising in the context of spatially distributed processes. To achieve this objective, the research will initially focus on creating a computationally efficient data-driven algorithm for: a) the derivation of nonlinear low-order, approximate models for dissipative PDE systems that are specifically tailored for control and optimization purposes, and b) characterization of the error between the low-order model and PDE system solutions. Subsequently, the research will focus on the synthesis of practically implementable feedback control structures that can deal with the issues of nonlinearity, model uncertainty, constrains and limited measurement availability. Concurrently, computational issues of optimization/optimal operation policies for spatially distributed processes will be resolved. This will be achieved via the derivation of a systematic scheme for the formulation of computationally efficient dynamic optimization problems that are amenable to standard search algorithms. The research results will be transferred into the industrial sector through the development and dissemination of software with a transparent user-machine interaction interface. Analyzing, optimizing and tightly controlling transport-reaction processes will benefit key processes of a wide range of industries such as lithographic reactors for microelectronics and photovoltaics fabrication and advanced catalytic reactors and industrial glass furnaces. Moreover, numerous activities will be pursued to integrate the research with education including incorporation of research results in optimization and control courses, undergraduate student participation in research through the honors program, and the development of educational tools such as matlab add-ons, java applets and wikis.

本研究项目的目标是放宽基于数据的非线性降阶方法的要求，并将其适用性扩展到空间分布过程背景下产生的耗散偏微分方程（PDE）系统的控制和优化。为了实现这一目标，该研究最初将重点关注创建一种计算高效的数据驱动算法，用于：a）推导专门针对控制和优化目的定制的耗散偏微分方程系统的非线性低阶近似模型，以及b）表征低阶模型和偏微分方程系统解之间的误差。B）低阶模型和偏微分方程系统解之间的误差。随后，研究将集中在合成的实际可实现的反馈控制结构，可以处理的问题的非线性，模型的不确定性，约束和有限的测量可用性。同时，空间分布过程的优化/最优操作策略的计算问题将得到解决。这将是通过推导出一个系统的计划，制定计算效率的动态优化问题，是服从标准的搜索算法。将通过开发和传播具有透明的用户-机器交互界面的软件，将研究成果转移到工业部门。分析、优化和严格控制传输-反应过程将使各种行业的关键过程受益，例如用于微电子和光电子制造的光刻反应器以及先进的催化反应器和工业玻璃熔炉。 此外，许多活动将追求研究与教育相结合，包括在优化和控制课程的研究成果，本科生通过荣誉计划参与研究，并开发教育工具，如matlab插件，java applets和wiki。