Artificial neural network-based hysteresis estimation of capacitive pressure sensor

The output of the Capacitive Pressure Sensor (CPS) depends both on the input pressure and on its recent history, an effect called hysteresis. A Switched Capacitor Interface (SCI) converts and amplifies either the change in capacitance of pressure-sensor into an equivalent voltage and the difference between the input pressure due to the hysteresis effect. A new approach to CPS modeling based on artificial neural network (ANN) is proposed taken into account the dynamic environment in which the temperature variation is quite large and the response characteristics of a CPS are highly nonlinear and temperature dependent. The first model is used to simulate the behaviour of the CPS. The second model is based on inverse modeling, which can be used to compensate the hysteresis effect the temperature drift and the nonlinearity. (C) 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.