P02 - Heuristic, Statistical and Analytical Experimental Design

P02 - Experimental Design

Based on the foundation of the current knowledge and known uncertainties, P02 develops schedules for the establishment of micro process parameters that result in the desired material features. In order to be able to quickly traverse the large search space that results from the given use case, a combination of heuristic methods, statistical planning and analytical inverse modeling are combined.



R. Drechsler, S. Huhn, Chr. Plump: Combining Machine Learning and Formal Techniques for Small Data Applications - A Framework to Explore New Structural Materials. Euromicro Conference on Digital System Design (DSD), Portorož, Slowenien, 2020, [Link to Conference], [Link to PDF]

A. Bader, A. Toenjes, N. Wielki, A. Mändle, A.-K. Onken, A. v. Hehl, D. Meyer, W. Brannath, K. Tracht: Parameter Optimization in High-Throughput Testing for Structural Materials, Materials 2019, 12,  https://www.doi.org/10.3390/ma12203439 .

T. Czotscher, D. Otero Baguer, F. Vollertsen, I. Piotrowska-Kurczewski, P. Maass: Connection Between Shock Wave Induced Indentations And Hardness By Means Of Neural Networks, ESAFORM 2019, 2019.

A. Toenjes, H. Sonnenberg, C. Plump, R. Drechsler, A. von Hehl: Measurement and Evaluation of Calorimetric Descriptors for the Suitability for Evolutionary High-Throughput Material Development, Metals 2019, 9, 149 https://www.doi.org/10.3390/met9020149

S. Dittmer, T. Kluth, P. Maass, D. Otero Baguer: Regularization by architecture: A deep prior approach for inverse problems, arXiv:1812.03889 [cs.LG] 2018

D. Otero Baguer, I. Piotrowska-Kurczewski, P. Maass: Inverse Problems in designing new structural materials, Springer, 2018

J. Stoppe, Chr. Plump, S. Huhn, R. Drechsler: Building Fast Multi-Agent Systems using Hardware Design Languages for High-Throughput Systems. &th International Conference on Dynamics in Logistics (LDIC), Bremen, Germany, 2018 [Link to Conference] [Link to PDF]

N. Drechsler, A. Sülflow, R. Drechsler: Incorporating user preferences in many-objective optimization using relation ε-preferred. Natural Computing, 2015. 14(3): p. 469-483.