Virtual compensator for correcting the disturbing variable effect in transducers

Abstract

Transducers are exposed to different types of environments during the course of measurement. The performance of a transducer is affected adversely by variations in excitation quantities and change in ambient conditions including temperature and humidity. The effect of change in environmental conditions on the transducer and subsequent on its signal conditioning circuit is quite nonlinear in nature. Especially, change in transducer excitation causes its response characteristics highly nonlinear and complex signal processing is required invariably to obtain correct readout. In this paper, we propose an artificial neural network-based novel virtual compensator for correcting the effect of a disturbing variable in transducers. The correction is carried out by a nonlinear two-dimensional artificial neural network-based inverse model of the transducer trained with Levenberg–Marquardt learning algorithm. By training the neural model suitably, the digital readout of the applied input (measurand) is obtained which is independent of disturbing variable.