A novel method to obtain the characteristic constant of a measurement chain

Abstract

Abstract The characteristic constant is commonly used to reconstruct the measured signal when a measurement chain with flat wideband frequency response is applied. The characteristic constant is often calibrated by a standard pulse that follows a rectangular pulse behavior. However, a pulse that is more than similar to the measured signal is more suitable theoretically than a rectangular pulse as the calibration waveform. To reduce the measurement error caused by the calibration waveform, a novel method to optimize the calibration waveform is proposed in this paper. A dataset is constructed based on the prior information of the signal to be measured. Criterions for better calibration waveforms are also discussed. Dataset construction along with this criterion makes the calibration waveform optimization a solvable mathematical problem. The calibration waveform can then be quantitively evaluation and optimized. The optimized calibration waveform will make the error caused by the fluctuations in the frequency response of a proportional sensor as small as possible statistically. As an actual application case, simulation results are also provided to intuitively explain the method. The procedure to obtain the characteristic constant based on the optimized calibration waveform is then outlined. A calibration system for the given application case is built. Finally, an experiment is designed and executed. The experimental results approved the method proposed in this paper well.