Process synthesis, design, and control: Optimization with dynamic models and discrete decisions
The optimization approach to process synthesis involves the representation of process design alternatives through a process superstructure and the mathematical modeling of this superstructure. Applying this approach to problems in engineering often requires the use of differential and algebraic equa...
Saved in:
| Main Author: | |
|---|---|
| Format: | Dissertation |
| Language: | English |
| Published: |
ProQuest Dissertations & Theses
01.01.1999
|
| Subjects: | |
| ISBN: | 9780599343986, 0599343982 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The optimization approach to process synthesis involves the representation of process design alternatives through a process superstructure and the mathematical modeling of this superstructure. Applying this approach to problems in engineering often requires the use of differential and algebraic equations to adequately represent dynamic models. In addition, integer variables are necessary to model discrete decisions such as the existence of process units in the flowsheet. This mathematical modeling results in a formulation classified as a mixed-integer optimal control problem (MIOCP). This type of problem arises in analyzing the interaction of design and control and in reactor network synthesis. In the interaction of design and control, the controllability aspects of the process are considered at the early stages of design by incorporating dynamic models and process control issues into the process synthesis framework. The tradeoffs between the controllability and the economic design of the process are addressed through a multi-objective framework. The reactor network synthesis problem deals with determining the optimal flow structure through network of tubular and tank reactors. The formulation involves dynamic models due to the modeling of the tubular reactors in the superstructure. The solution framework for a dressing the MIOCP formulation extends the concepts of mixed-integer nonlinear programming algorithms so as to handle dynamic systems. The MIOCP algorithm decomposes the problem into an optimal control primal problem which provides an upper bound on the solution of the problem and a mixed-integer linear program master problem which provides a lower bound. The optimal control problem is solved using a control parameterization technique where the dynamic system is integrated as a function of time invariant parameters. The algorithm is implemented in the framework MINOPT which is used as a computational tool for the solution of process synthesis problems. The proposed approach is applied to process synthesis examples in the interaction of design and control and reactor network synthesis to illustrate the features of the methodology. |
|---|---|
| Bibliography: | SourceType-Dissertations & Theses-1 ObjectType-Dissertation/Thesis-1 content type line 12 |
| ISBN: | 9780599343986 0599343982 |