Studies of Global Optimization in Process Systems Engineering

点击翻译按钮获取中文摘要

Floudas - Abstract - 9528150 Significant effort has been expended toward theoretical and algorithmic studies of optimization algorithms and their computational testing in applications that arise in process synthesis, design and control. Recent interest in global optimization and its application to chemical engineering problems is due to the fact that: (1) a large number of process synthesis, design, and control problems are global optimization problems; (2) existing local nonlinear optimization approaches may either fail to obtain a feasible solution or are trapped to a local optimum solution which may differ in value from the global solution; and (3) the global optimum solution may have a very different physical interpretation when it is compared to local solutions (e.g. in phase equilibrium a local solution may provide incorrect prediction of types of phases at equilibrium, as well as the components' composition in each phase). The primary objective of this research is to develop theoretical, algorithmic and computational techniques for global optimization problems that arise in a variety of chemical engineering design and synthesis problems, as well as molecular conformation problems that arise in computational chemistry and biology. These advances will complement efforts in the rigorous mathematical modeling of chemical process systems and aid the search in the design of molecules, peptides, and proteins. The PI plans to investigate (i) the extension of his recently developed global optimization approach, aBB, to general constrained nonlinear optimization problems and mixed-integer nonlinear problems which are employed in process synthesis; (ii) the development of a new global optimization based approach for the determination of all solutions of certain classes of nonlinearly constrained systems of equations which has important applications in process systems that exhibit multiple steady states (e.g., reactors, distillation columns); (iii) the parameter estimation problems of nonlin ear models; (iv) the bilevel nonlinear optimization problems that arise in the analysis of process systems under uncertainty: (v) the three-dimensional structure prediction of oligopeptides, peptides, and proteins; and (vi) the development of a distributed computing framework for the aBB global optimization approach and the development of computational tools for (i) through (v).