Modeling and investigation of smart capacitive pressure sensor using artificial neural networks

In this paper, a new capacitive pressure sensor (CPS) is investigated and modeled by means of neural approach. The sensing principle in our pressure sensor is based on the determination of the change in the capacity induced by the applied pressure. A ring oscillator is used to convert the capacity variation of the pressure sensor to an output frequency. A multilayer perceptron neural network is used to predict the applied pressure which causes a variation of the capacity including the temperature effects. This model is implemented as an electronic device into PSPICE simulator library, where the device should reproduce faithfully the pressure sensor behavior. Moreover, a new inverse model called smart sensor has been developed, in order to remove the nonlinearity behavior of sensor response. The obtained results make the proposed smart sensor as a potential alternative for high performances pressure sensing applications.