NEURAL-NETWORK FOR PATTERN ASSOCIATION IN ELECTRICAL CAPACITANCE TOMOGRAPHY

The paper describes the basic principles of an artificial neuron, the multilayer perceptron network and the back-propagation training algorithm, applied to electrical capacitance tomography systems for real-time, noninvasive imaging and measurement of multicomponent flows such as gas/oil and water/oil. Particular attention is given to the problems of distortion of the field ('soft field' error) and limitations on spatial resolution (imposed by the number of electrodes) in conventional image reconstruction algorithms for current systems. In addressing these issues, for the first time, an artificial neural network is employed to replace conventional image reconstruction algorithms. The system consists of a simulation program for a single layer multiple output network, using a variant of the back-propagation training algorithm with the principle of pattern association. The input vector consists of preprocessed capacitance measurements, and the output of the network directly corresponds to the spatial image. Two similar networks are trained for gas/oil flow (small difference in permittivity) and water/oil flow (large difference in permittivity) with results compared. The results illustrate the feasibility of such networks for relatively accurate image reconstruction, without the difficulties associated with conventional algorithms.