Introduction to Adjoint Optimization
14-16 September 2015, Glasgow
IODA Kick-off meeting and Training
IODA Training Feedback
This was the first training event for the 'Industrial Optimal Design using Adjoint CFD' EU Horizon 2020 funded Project and covered a wide introduction to general optimization and continuous adjoint methods.
A general overview was provided about numerical optimization and univariate methods. Also, an extension from univariate to multi-variate optimization and an introduction for gradient computation was made. A detailed introduction to the continuous adjoint method, based on some quasi-one-dimensional flow examples took place and finally a course on parametrization of shape optimization comparing different shape optimization approaches (node-bases, lattice-based, CAD-based).
This training was a very helpful introduction into optimization which is the basic field of interest of the IODA project. The most well-established methods were introduced giving the correct directions for further research on the topic. Also, the adjoint introduction gave, in a brief and compact forms the basics of the adjoint methods giving a good overview on that field.