Intelligent Algorithms and Their Application in Electrical Impedance Tomography

Abstract

EIT (Electrical impedance tomography) problem is an inverse problem of mathematical physics. In the past years, numerous scientists and researchers put their effort to this problem and a lot of traditional mathematics algorithms have been applied. The difficulty of this problem is that we do not know inside information, all we know is potential on the surface of the object, then use this known data to calculate inside structure. With the time going by, Intelligent algorithms have become more popular. In this paper, we used an intelligent algorithm Back Propagation (BP) neural network algorithm to solve this problem instead of complex traditional mathematics algorithm. Different from normal usage of BP neural network algorithm, we look each nodes in different layers as the each result in solution this time. Mend each node weight constantly by some groups of prior data, we will get the solution which is consist of each node weight. In the experiment simulation, solutions of iteration with 500times and 1000times show that this algorithm is feasible. At the end of the paper, we concluded this experiment and analysis the executable of another intelligent algorithm which will be used in this EIT problem, Genetic algorithm.