Development of Modeling and Optimization Tools for Hybrid Systems

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

ABSTRACT PI: Lorenz Biegler Proposal Number : 0457379 Institution: Carnegie Mellon University Title: Development of Modeling and Optimization Tools for Hybrid Systems Research: Optimal design under unknown information is a key task in process systems engineering. This project considers formulations that incorporate two types of unknown information, uncertainty in the modeling information and variability in the operation of the process. We will apply these formulations to the modeling and optimization of dynamic hybrid systems. Such systems occur widely in nature (particular in biological, chemical and mechanical applications) and are also essential components of economic planning models. For optimization formulations, these discontinuous decisions are represented through mixed integer nonlinear programs (MINLPs), or mathematical programs with complementarity constraints (MPCC). Our research will consider both MINLP and MPCC problems over a broad set of application domains and we intend to develop new optimization strategies. These research results will be applied to broad set of applications including systems biology and drug design, planning and scheduling models for enterprise wide optimization, applications of hybrid models in mechanical and process engineering systems. Finally, the results of this research will allow consideration of challenging applications too large to be considered with existing optimization strategies. Broader Impacts: The broader impacts of the project are: i. creation of collaboration between the Carnegie Mellon University, Koc University (Turkey), and Middle East Technical University (Turkey) ii. establishment of a bioinformatics related research activity in the PSE group at Carnegie Mellon University iii. incorporation of the modeling and optimization methods in the graduate courses at the participating universities iv. participation of undergraduate students in hybrid systems research will be accomplished through term projects in the courses (see .Educational Plan. section) v. models and optimization algorithms developed in the project and the results will be made publicly available through publications, presentations and through a web page specifically designed for this project vi. a concerted effort to include underrepresented groups and minorities into our research program will be a key component of this proposal.