Seminars of the Focus Area Complex Systems

Prof. Dr. C. Beta, Prof. Dr. K. Dethloff, Prof. Dr. R. Engbert, Prof. Dr. M. Holschneider, Prof. Dr. W. Huisinga, Prof. Dr. Ralf Metzler, Prof. Dr. A. Pikovsky, Prof. Dr. S. Reich, Prof. Dr. M. Rosenblum, Prof. Dr. G. Rüdiger, Prof. Dr. T. Scheffer, Prof. Dr. F. Scherbaum, Prof. Dr. J. Selbig, Prof. Dr. F. Spahn


Speaker: Simon Laflamme, Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, USA

Title: Self-Organizing Inputs for Black-Box Modeling of Large-Scale Nonstationary Dynamic Systems

Time: Wed, Sept 22, 2010, 14:15

Place: Building 28, Room 2.123

Black-box modeling of large-scale dynamic systems is challenged by the lack of readily available input-output data sets. This lack of data is a consequence of the large-scale nature, where pre-training of a black-box model is either technologically or economically constrained. A solution is real-time sequential training. Real-time training is necessary for control systems, where a prescribed level of performance is required while the external input is occurring. A major difficulty in real-time sequential training is proper input selection for the black-box model.

A new algorithm for real-time sequential selection of inputs, termed the Self-Organizing Inputs (SOI) algorithm, is proposed. The novelty of the algorithm is the capability to sequentially organize the inputs of a black-box model by updating the required number of inputs for the representation as well as their time delay, which allows black-box models to capture the essential dynamics of the system and to cope with nonstationarity. The SOI capability is based on chaotic time series analysis, and on the theory that dynamic systems can be topologically reconstructed using a single observation delayed in time, provided the choice of time delay and embedding dimension is appropriate.

The SOI algorithm is simulated in the context of a neurocontroller to mitigate vibrations of an existing structure located in Boston, Massachusetts. Results show that the proposed neurocontroller performs well at vibration mitigation for both strategies using limited acceleration measurements and limited knowledge about the dynamic properties of the large-scale structure.

Back to the seminar schedule

Past NLD Seminars (1994-2007) & (2008 ...)

Students' seminar: Theoretical Physics, PIK, Modeling & TSA Berlin-Potsdam-Colloquia: PhysGesellschaft Berlin, TU Berlin, Pro Physik, AIP, AEI, MPI-KGF, GFZ, HMI, PIK, AWI, Max Planck Institute for the History of Science, Mathematik, DPG Disputationen, & Vorschau UP

Udo Schwarz, Zentrum für Dynamik komplexer Systeme,
Universität Potsdam, Campus Golm Karl-Liebknecht-Str. 24, 14476 Potsdam, building 28, room 2.107
Phone: (+49-331) 977-1658, Fax : (+49-331) 977-1045

Email: Udo.Schwarz AT

DFG SFB 1294

DFG Sonderforschungsbereich 1294 Data assimilation

DFG SPP 1488

DFG Schwerpunktprogramm 1488 Planetmag

News: odeint C++ library

Check out boost::odeint – our fast and flexible C++ library for integrating differential equations!