Abstract:

Automotive crash test signal filtering is a key step in data processing. This study used vehicle body acceleration data and crash dummy head acceleration data obtained from vehicle crash tests to evaluate and compare three signal noise reduction techniques: the SAE J211 standard filter, the Moving Average(MA) method, and the Cumulative Integration. Based on the analysis of crash pulses and head acceleration data, the filtering results were directly correlated with body structure optimization and restraint system matching. The results show that the SAE filter provides precise frequency control, with a cutoff frequency of 100 Hz performing optimally in processing vehicle crash pulse acceleration data. The MA method, by obtaining the optimal window of data points, effectively removes noise while preserving key signal features. The Integration method reveals data trends through cumulative summation operations, naturally eliminating noise interference through summation. The SAE standard filter achieves a signal-to-noise ratio of 18.95 dB for crash pulse acceleration data and 30.49 dB for dummy head acceleration data; whereas the MA method based on 100 data points achieves corresponding signal-to-noise ratios of 8.59 dB and 13.82 dB, respectively. Additionally, directly integrating the acceleration signal yields smooth velocity and displacement curves. This study offers evidence-based guidance for the practical implementation of the three filtering methods under investigation.

Key words: signal filtering, SAE J211 Filter, moving average (MA) method, Cumulative Integration, crash testing data