|
|||||||
Advanced Topics in Process Systems EngineeringTechnical University of DenmarkGeneral course objectives: This course is aimed at introducing special topics from the current state of the art in process systems engineering. The objective is to give the participants a quick and thorough introduction to relevant advanced topics in process systems engineering, which otherwise would not be possible due to lack of formal courses on such topics. Examples of such advanced topics includes among others machine learning techniques (metamodeling, surrogate modelling), Monte Carlo and Quasi Monte Carlo optimization, advanced/nonlinear control and optimization, computer aided product design, among others. At the same time, the course will help the participants in getting a broader view on the art of systematic analysis for efficient, robust and innovative solution strategies. Depending on the defined/selected topic by the course participants and the instructors, one or more of the below learning objectives apply: Learning objectives: A student who has met the objectives of the course will be able to:
Contents: Efficient, reliable and robust solution of engineering problems require solution strategies that have been developed through a systematic analysis of the needs of the problems being solved and the development and use of methods and tools that can match these needs. In process and product design, the solution strategies also need to provide innovative and new solutions. A systems approach allows the development of the necessary solution strategies through a systematic analysis of all aspects of the specified problem or a set of problems. Because of the potentially large range of topics that can be covered, a selection of topics will be considered each year, based on the background of the participant(s) and the current state of the art. Examples of topics are as follows: systematic computer aided product formulation; computer aided molecular design; systematic model identification & design of experiments; a systems approach to development of sustainable process technologies; machine learning including design of data and model identification techniques (using Radial Basis Function, Artificial Neural Network, Gaussian Process Modeling, Kriging), Global sensitivity analysis based on metamodeling (using polynomial chaos expansion, modified adaptive response splines (MARS), among others), Stochastic programming and Monte Carlo based optimization algorithms, Advanced and nonlinear optimization and control techniques, etc. Each topic will be analyzed in terms of the generic problem definition and solution strategy, verification of the proposed systems approach, the advantages/disadvantages of the approach, identification and testing of possible improvements. Finally, a systems approach to solving of process and product engineering problems will be highlighted together with the latest developments in terms of computer aided methods & tools. The selected topic will be applied to a selected case studies defined and prepared by the participants. |
|
||||||