Jump label

Service navigation

Main navigation

You are here:

Main content

Computers and Chemical Engineering Best Paper 2016

Weihua Gao, Simon Wenzel, and Sebastian Engell receive the Best Paper Award 2016 of the journal Computers and Chemical Engineering with their contribution from the field of Real-time Optimization under uncertainty.
How can a chemical plant be operated optimally despite the fact that the model does not match the reality exactly? This challenge can be addressed by algorithms that work nicely in theory and are proven to converge in an iterative procedure to the real optimum of a plant. Key to a successful application of these methods are precise measurements, which in reality however are affected by measurement noise and other disturbances. The authors worked on this problem and proposed a novel approach based on quadratic approximations that finds the real optimum. The work with the title “A reliable modifier adaptation strategy for real-time optimization” has recently been awarded with the Best Paper Award 2016 of the journal Computers and Chemical Engineering.

CACE-1

The efficiency of the operation of a chemical plant can be increased significantly by optimizing its operating point. However, to achieve this a sufficiently precise (generally a rigorous and non-linear) plant model and suitable optimization techniques are required. The optimization problem is subject to various constraints that result from the technical limitations of the equipment, quality specifications of the products, and regulations of emissions. Such a mathematical optimization steers the plant towards the optimum of the model, which does not necessarily correspond to the true plant optimum, because there may be a mismatch between the prediction of the model and the real plant.

The foundation of the published work is a methodology that modifies the prediction of the model by correction terms, which result from observations of the plant, i.e., the gradient of the cost function predicted by the model is corrected by the gradient that is estimated based on past plant measurements. This scheme is known as Modifier Adaptation (MA).

The difficulty in the application Modifier Adaptation is the computation of sufficiently accurate plant gradients, which are sensitive towards disturbances. To cope with this situation the authors developed a novel approach that combines elements of derivative free optimization (DFO) with the iterative optimization by Modifier Adaptation. A quadratic surrogate model is fitted to the measurement of the plants and the gradient is determined using this approximation. Based on the theory of derivative free optimization different strategies to select suitable operating points for the surrogate model have been proposed that lead to a fast and robust optimization. As shown in the article, which is in parts based on the Master Thesis of Simon Wenzel, the new method can be used to solve even complex examples efficiently.

The article is published as an open access article and is free for download here.

Gao

Dr.-Ing. Weihua Gao received his Bachelor degree in Mechanical Engineering in 1999 and his Master degree in Mechatronics in 2002 from the Xi'an Jiaotong University in China and his Dr.-Ing. degree in Chemical Engineering in 2005 from TU Dortmund, Germany. He worked with General Electric as a Lead Research Engineer from 2006-2010 and with the Fangyuan Group Co. Ltd. as a Managing Director from 2010-2013. Since 2013 he has been working in the Process Dynamics and Operations Group at TU Dortmund as a PostDoc. His research is focused on real-time optimization of plants under uncertainty.

Wenzel Simon Wenzel, M.Sc., studied Chemical Engineering in a sandwich course at Krefeld University of Applied Science in corporation with Borgers AG in Bocholt, Germany. In 2012, he completed his B.Eng. with a Thesis in the Process and Engineering department at P&G Crailsheim, Germany. Afterwards, he studied Process Systems Engineering (PSE) at TU Dortmund and received his M.Sc. in November 2014. The topic of the Master Thesis in the Process Dynamics and Operations Group was "Combination of DFO and IGMO in Iterative Optimizing Control". Currently, he is working on market-based distributed optimization in the European Project "Improved energy and resource efficiency by better coordination of production in the process industries" (CoPro).
Engell Prof. Dr.-Ing. Sebastian Engell received a Dipl.-Ing degree in Electrical Engineering from Ruhr-Universität Bochum, Germany in 1978 and the Dr.-Ing. Degree and the venia legendi in Automatic Control from Universität Duisburg in 1981 and 1987. 1984/1985 he spent a year as a PostDoc at McGill University, Montréal, P.Q. 1986-1990 he was the head of an R&D group at the Fraunhofer Institut IITB in Karlsruhe, Germany. 1990 he was appointed to his present position as a Full Professor of Process Dynamics and Operations in the Department of Chemical Engineering at TU Dortmund. 2008 he was a Distinguished Visiting Professor at Carnegie Mellon University, Pittsburgh, USA. He was Department Chairman 1996-1999 and 2012-2014 and Vice-Rector for Research of TU Dortmund 2002-2006.

Prof. Dr.-Ing. Sebastian Engell received an IFAC Journal of Process Control Best Paper Award, a Best Paper Award of the IEEE Congress on Evolutionary Computation 2010 with Thomas Tometzki on risk-conscious planning and the PSE Model-based Innovation Prize with Ala Eldin Bouaswaig. He gave the Bayer Lecture in Process Systems Engineering at Carnegie Mellon University in 2008 and the Roger Sargent Lecture at Imperial College, London, in 2012. He has published about 150 Papers in scientific journals, more than 40 papers in edited volumes and more than 300 conference papers with peer review and full papers in proceedings. In 2012, he was awarded a European Advanced Investigator Grant for the Project MOBOCON - Model-based optimizing control from a vision to industrial reality. He has led a number of EU funded projects and currently is the coordinator of the project CoPro – Improvement of Energy and Resource Efficiency by Better Coordination of Production in the Process Industries that is funded under the EU SPIRE program.