Extraction of information from noisy 3-year-old ATD response signals in static out-of-position airbag tests.

This paper presents an approach to analyse experimental data contaminated with noise from Anthropomorphic Test Devices (ATDs). This approach is based on information extraction procedures and they are illustrated through an analysis of Hybrid III 3-year-old and Q3 ATDs test data. The methodology used for extracting information and ATD test data analysis includes optimised filtering, spectral coherence, auto- and cross-correlation analysis, and Kalman filtering. This work investigates promising techniques of extracting information from noisy ATD signals that are not commonly used in the automotive industry.