Analysis of dynamic measurements and determination of time-dependent measurement uncertainty using a second-order model

摘要

The task of analysing dynamic measurements is considered for the case that the input-output behaviour of the employed sensor can be described in terms of a second-order dynamic model. It is assumed that estimates of the parameters of this model are available which have been determined by the analysis of previous dynamic calibration measurements. The goal is to estimate the unknown input of the sensor given its discrete-time output signal. It is proposed to apply an FIR filter which is constructed according to the underlying dynamic model. The uncertainties associated with the obtained estimates of the time-dependent measurand are then determined by taking into account the uncertainty associated with the estimates of the parameters of the dynamic model, the uncertainty of the sensor's output signal and the applied signal processing steps. The proposed procedure allows real-time calculations including the real-time calculation of uncertainties, and it could therefore be directly integrated into the measurement chain. Utilizing dynamic calibration measurements of accelerometers, the proposed analysis is demonstrated by its application to acceleration measurements upon shock excitations.