Berlin 2008 – scientific programme

Parts | Days | Selection | Search | Downloads | Help

DY: Fachverband Dynamik und Statistische Physik

DY 21: Nonlinear dynamics, synchronization and chaos III

DY 21.5: Talk

Wednesday, February 27, 2008, 17:45–18:00, MA 001

Multivariate Phase Rectified Signal Averaging - A tool for studying complex interrelated time series — •Aicko Y. Schumann, Jan W. Kantelhardt, and Fabian Gans — Institut für Physik Theorie, Martin-Luther University Halle-Wittenberg

Many natural systems generate periodicities on different time scales because some of their components form closed regulation loops in addition to causal linear control chains, e.g., cardio-respiratory rhythms in physiology or the El-Niño phenomenon in geophysics. In most cases non-stationary and noisy data from several simultaneously recorded signals is available for a multivariate analysis. In order to understand the underlying control chains and loops an time series analysis tool capable of identifying periodicities and the direction of causal relations in the presence of non-stationarities and noise is needed. In a previous work [1,2,3] we have therefor introduced the (monovariate) phase-rectified signal averaging technique (PRSA), which can distinguish effects caused by acceleration and deceleration of a signal in the signal itself. Further developing this approach for the study of two related signals, we propose the multivariate-PRSA method (MPRSA), which is capable of detecting and quantifying related quasi-periodic oscillations as well as causal interrelations in two signals masked by non-stationarities and noise.

[1] Bauer, A. et. al. Physica A, 2006,364, 423–434

[2] Bauer, A. et. al. The Lancet, 2006 367, 1674

[3] Kantelhardt, J.W. et. al. CHAOS, 2007 17(1), 015112