ITR/AP COLLABORATIVE RESEARCH: Real Time Optimization for Data Assimilation and Control of Large Scale Dynamic Simulations

ITR/AP 合作研究：大规模动态模拟数据同化和控制的实时优化

• 批准号: 0121360
• 负责人: Matthias Heinkenschloss
• 金额: $ 55万
• 依托单位: William Marsh Rice University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-10-01 至 2007-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0121360&HistoricalAwards=false
• 关键词: ITR; AP; COLLABORATIVE; RESEARCH; Real

This project will create and apply algorithms and software tools for on-line simulations that continuously (1) assimilate sensor data from dynamic physical processes, and (2) generate optimal strategies for their control. A number of critical industrial, scientific, and societal problems stand to benefit from this research such as aerodynamics, energy, geophysics, infrastructure, manufacturing, medicine, chemical process and environmental applications; two of these will be the focus of the current research. In these and many other cases, the underlying models have become capable of sufficient fidelity to yield meaningful predictions, provided unknown parameters (typically initial/boundary conditions, material coefficients, sources, or geometry) can be estimated appropriately using observational data.The critical step is the solution of a large-scale nonlinear optimization problem that is constrained by the simulation equations, typically PDEs or their reduced order models. A data assimilation phase will seek to minimize the mismatch between sensor data and model-based predictions by adjusting unknown parameters of the PDE simulation, and the optimal control phase will find an optimal control strategy based on the updated model.Despite advances in hardware, networks, parallel PDE solvers, large-scale optimization algorithms, and real-time ODE optimization, significant algorithmic and software challenges must be overcome before the ultimate goal of real-time PDE data assimilation and optimal control can be realized. Needed are fundamentally new PDE optimization algorithms that must: (1) run sufficiently quickly to permit decision-making at time scales of interest; (2) scale to the large numbers of variables and constraints that characterize PDE optimization and processors that characterize high-end systems; (3) adjust to different solution accuracy requirements; (4) target time-dependent objectives and constraints; (5) tolerate incomplete, uncertain, or errant data; (6) be capable of bootstrapping current solutions; (7) yield meaningful results when terminated prematurely; and (8) be robust in the face of ill-posedness.To create, apply, and disseminate the enabling technologies for real-time PDE data assimilation and optimal control, the project will: (1) Develop algorithms and tools for real-time data assimilation and optimal control that meet the above specifications for a class of important applications. (2) Implement and publicly distribute these algorithms within an object-oriented framework that incorporates problem structure, interfaces easily with high performance PDE solver libraries fosters applicability of our tools to a broad range of real-time data assimilation and optimal control problems, and enables extension of the algorithms without interfering with applications. (3) Apply these algorithms and tools to two critical environmental and industrial problems: modeling and control of chemical vapor deposition (CVD) reactors and of wildland firespread. (4) Interact and work with other user communities to ensure that the algorithms and software we produce are useful across a broad range of applications.

该项目将创建和应用用于在线模拟的算法和软件工具，这些算法和软件工具将持续(1)吸收来自动态物理过程的传感器数据，并(2)为其控制生成最佳策略。一些关键的工业、科学和社会问题将从这项研究中受益，如空气动力学、能源、地球物理、基础设施、制造、医学、化学工艺和环境应用；其中两个将是当前研究的重点。在这些和许多其他情况下，如果未知参数(通常是初始/边界条件、材料系数、源或几何)可以使用观测数据适当地估计，则基础模型已经能够足够的保真度来产生有意义的预测。关键步骤是求解受模拟方程约束的大规模非线性优化问题，通常是偏微分方程组或其降阶模型。数据同化阶段将寻求通过调整PDE模拟的未知参数来最小化传感器数据与基于模型的预测之间的失配，而最优控制阶段将基于更新的模型找到最优控制策略。尽管在硬件、网络、并行PDE求解器、大规模优化算法和实时常量优化方面取得了进展，但在实现实时PDE数据同化和最优控制的最终目标之前，必须克服重大的算法和软件挑战。需要新的PDE优化算法，这些算法必须：(1)运行足够快，以允许在感兴趣的时间尺度上进行决策；(2)适应表征PDE优化的大量变量和约束；(3)适应不同的解精度要求；(4)目标依赖于时间的目标和约束；(5)容忍不完整、不确定或错误的数据；(6)能够引导当前解；(7)在过早终止时产生有意义的结果；为了创造、应用和传播实时PDE数据同化和最优控制的使能技术，该项目将：(1)为一类重要的应用开发满足上述规范的实时数据同化和最优控制的算法和工具。(2)在面向对象的框架内实现并公开分发这些算法，该框架包含问题结构，易于与高性能PDE求解器库接口，促进了我们的工具对广泛的实时数据同化和最优控制问题的适用性，并使算法能够在不干扰应用程序的情况下进行扩展。(3)将这些算法和工具应用于两个关键的环境和工业问题：化学气相沉积(CVD)反应堆和荒地火焰蔓延的建模和控制。(4)与其他用户社区互动和合作，以确保我们生产的算法和软件适用于广泛的应用程序。