from 2014-12-04

Every year, NAMUR confers awards for outstanding academic dissertations or theses (bachelor/master/doctorate) in the field of "intelligent process control".

During the NAMUR General Meeting 2014, three excellent theses were honored with a NAMUR AWARD:

Ms. Ilona Sonnevend received the NAMUR AWARD for her master's thesis "Quantitative Performance Evaluation of Steady-State and Dynamic Two-Layer Real-Time Optimization Achitectures".

The thesis was prepared at RWTH Aachen under the supervision of Professors Steffen Leonhardt and Wolfgang Marquardt.

In her work, Ms. Sonnevend considered an extended two-layer architecture consisting of dynamic real-time optimization with a secondary nonlinear model-predictive control in comparison to the established two-layer architecture with steady-state real-time optimization.

The first of the two architectures named above all too often fails to establish itself in practice due to insufficient acceptance, as it is difficult to see what additional benefits this would bring compared to the conventional two-layer control architecture. These benefits would have to be significant enough to warrant the added expense for implementation and maintenance.

Ms. Sonnevend has formulated heuristic evaluation criteria that allow prior estimation of the operational added value.

The practicability of these criteria was impressively demonstrated by different means, including an industrial case study.

Dr. Niko Rossner received a NAMUR AWARD for his doctoral thesis "Robust process control based on Gaussian mixture density modeling, as applied to the Bray-Liebhafsky reaction and the autotrophic cultivation of Ralstonia eutropha H16".

The thesis was written at the Technical University of Berlin and supervised by Prof. Rudibert King:

The starting point for Mr. Rossner's dissertation is the description of the variables in a process with the help of statistical parameters such as the mean and covariance.

By applying Gaussian mixture densities, it is possible to predict, even in nonlinear process models, how the statistically described input variables will affect the statistics of the output variables.

Using this statistical approach, Mr. Rossner applied nonlinear model-predictive control methods in practice to a chemical and a biological system and compared them using classical methods. He was able to show that this approach was capable of enhancing space-time yields and increasing the throughput and product concentrations.

Dr. Haiyang Hao gained his NAMUR AWARD for his dissertation "Key performance monitoring and diagnostics in industrial automation processes".

The work was undertaken at the University of Duisburg-Essen under the supervision of Prof. Steven Ding.

Mr. Hao's dissertation addresses the topic of process monitoring and fault diagnosis based on estimation methods using KPIs.

The method he has developed in great breadth and depth will help to substantially increase the performance of process monitoring and diagnostics. His work covers, in particular, solutions

  • for data-based monitoring of processes with locally distributed parameters
  • and for detection of multiplicative errors.

Within the scope of his dissertation, the method was successfully tested on three benchmark processes which are of relevance to NAMUR.

Mr. Hao was unfortunately unable to attend the awards ceremony for personal reasons.

« Back