Feasibility of surface acoustic wave (SAW) sensor array processing with formal neural networks

Abstract

The poor selectivity of sensors, used to detect different species mixed in a real gas mixture, requires efficient signal processing devices to improve it. We have tried to apply neural techniques to this problem in order to obtain some ‘electronic’ selectivity. This work is the first step of a feasibility study of an electronic device using SAW gas sensors. Using linear approximation of experimental and static results of SAW sensors given in previous data, we also attempted to apply neural networks for multisensor array signal processing. In order to extend the operating range, the saturation effects yield by chemical layers was considered. To optimize the neural network, several activation functions were tested. A design close to radial basis functions networks was successfully applied. A new network connectivity led to increased interpolation capability. For real time processing, it is important to escape from slow kinetic adsorption of chemical layers. In this paper, the theoretical possibility of response time compensation has been shown by using a deconvolution process.