Digital filter design of frequency weighting function to measure and assess human vibration

Abstract

Frequency weighting functions that reflect the frequency characteristics of the human body must be applied in order to measure and assess the amount of human vibration. These functions, applied differently according to the part of the body exposed to vibration and vibration direction, exist in the form of frequency transfer functions and are mandated by international standards. To efficiently apply weighting functions to the measured acceleration time signal, it is essential to design digital filters for such functions. Frequency weighting functions consist of transfer functions from high orders. The z-transform process is performed for the digital filter design of frequency weighting functions. The applicable range of frequency also varies with sampling time and record length of the measured acceleration data. This study proposed a method of deriving the digital filter coefficient of z-transform based on matrix calculations. The validity of digital filters designed from applying the coefficient was assessed and applicable frequency ranges were presented in relation to the sampling time and record length.