RECENT DEVELOPMENTS IN THE DYNAMIC OPTIMISATION PACKAGE DYNO

LESAGE, F.; FIKAR, M.; LATIFI, R.

Abstract

DYNO (DYNamic Optimisation) is a dynamic optimisation package which have been developing since almost a decade in the “Laboratoire des Sciences du Génie Chimique” (Nancy, France) in collaboration with the “Department of Information Engineering and Process Control” (Bratislava, Slovakia). Basically, the objective is to determine the optimal profiles of decision variables (temperature, pressure, flow, heat, ...) and/or optimal parameter values of the dynamic model which optimise (minimise, maximise) a given performance index (batch period, yield, energy consumption,...), over a time horizon, under specified constraints (environment, process physical limits,...). The processes concerned are dynamic processes that are described by a detailed mathematic models, typically with non-linear and differential equations. Control vector parameterisation (CVP) is the optimisation method used. This sequential and feasible type approach consists in the approximation of the control variables by piece-wise (mainly constant) functions and leaving the state variables in their continuous form. The original dynamic optimisation problem is then converted into a non linear programming (NLP) problem that is solved using gradient-based optimisation methods (e.g., SQP methods). The recent developments of DYNO mainly deal with its implementation within Matlab environment.

Coresponding author e-mail: Francois[dot]Lesage[at]ensic[dot]inpl-nancy[dot]fr

Session: Dynamic and Global Optimisation of Processes (Invited Session, M. Fikar)