Real-time, non-linear optimization for industrial applications
(Actualized on 20-12-2012)
Non-linear Model Predictive Control directly adresses many aspects of complex industrial plant dynamics, like strong non-linearities, inverse responses, variable dead times and complex couplings between process vars. The controler uses a non-linear plant model to make predictions of future plant behaviour and optimize closed loop plant dynamics in real-time. The computational cost of such optimizations is high, hovever with modern computers and algorithms it is now applicable to a wider range of systems.
The aim of the project is to develop a real-time non-linear optimization platform for wide range of industrial control applications. This framework is based on a novel, low computational cost formulation of a Non-linear Programming Problem with inequality constraints and one step Newton-type method. It has been originally developed for application of the Nonlinear Model Predictive Control (NMPC) for the Superfluid Helium Cryogenic Circuit at the Large Hadron Collider (LHC) Arcs, see the homepage of that project at www.predictive-control.com/lhc.
 R Noga. "PhD Thesis description: Non-linear Model based Predictive Control (NMPC)of the Large Hydron Collider's (LHC) Super fluid Helium Cryogenic Circuit". University of Valladolid, CERN. 2008 (Download)