Framework for Designing Flexible Steady State and Dynamic Chemical Processes

摘要/ abstract摘要：作者：Luke E. Achenie机构：University of connecticut项目编号：0097936本研究旨在开发一种新的数学算法和相关软件，用于系统地确定化工过程（CP）的最佳设计余量。这些设计余量保证了CP在操作阶段的灵活性。尽管在CP设计过程中使用了不准确的数学过程模型，但灵活的CP能够可靠地运行。从经济、安全和环境的角度来看，CP的灵活性在化学过程设计中是一个非常重要的问题。影响灵活性的模型不确定性主要有三种：(a)模型结构不确定性，(b)参数不确定性（在存在测量不确定性的情况下确定模型参数而引入），以及(c)操作不确定性。在这项工作中，pi与(b)和(c)有关。传统上，灵活性问题是通过使用过度设计因素来考虑的，这是基于经验和经验的规则。过度设计因素本质上是经验性的，经常导致资源的昂贵和低效使用。本项目更广泛的影响是开发方法，减少在工业设计中使用经验过度设计因素的需要。具体而言，将解决以下问题：(i) CP灵活性的估计（即，即使外部和内部因素在某些已知范围内以不可预测的方式变化，CP维持可接受性能水平的能力），以及（ii）设计阶段不确定性下的CP优化。

ABSTRACTPI: Luke E. Achenie Institution: University of ConnecticutProposal Number: 0097936This research aims at developing novel mathematical algorithms and associated software for systematic determination of optimal design margins for chemical processes (CP). These design margins guarantee that flexibility of the CP at the operation stage. A flexible CP is able to operate reliably in spite of the use of inaccurate mathematical process models during the design of the CP. The flexibility of a CP is a very important concern in the design of chemical processes, from economic, safety and environmental points of view. Model uncertainties that impact flexibility are primarily of three kinds: (a) model structure uncertainty, (b) parametric uncertainty (introduced as a result of determining model parameters in the presence of measurement uncertainty), and (c) operational uncertainty. In this work, the PIs are concerned with (b) and (c).Flexibility issues have traditionally been considered through the use of over-design factors, which are based on rules of thumb and experience. Over-design factors are inherently empirical and often lead to costly and inefficient use of resources. The broader impact of this project is to develop methodologies that will reduce the need to use empirical over-design factors in industrial design. Specifically, the following problems will be addressed: (i) estimation of the flexibility of the CP (i.e. the CP's ability to sustain an acceptable level of performance even when external and internal factors change in an unpredictable fashion within certain known bounds), and (ii) CP optimization under uncertainty at the design stage.