(312a) Distributed Control and Estimation of Chemical Plants: Loop Partitioning and Design

This talk will discuss the evolution of research on distributed control and estimation of chemical plants using nonlinear dynamic models. The talk will cover fundamental topics including distributed model predictive control design, partitioning of control loops for distributed control, and distributed state estimator design. In doing so, the talk will highlight the work of Professor Daoutidis with his students on the introduction of graph-theoretic concepts and methods for accomplishing optimal loop partitioning which is a key requirement for the effective implementation of distributed model predictive control and estimation methods. Examples and future research directions will be briefly discussed.