Optimal Transducer Arrangement for Temperature Distribution Measurement in Arbitrary Convex-Shaped Space by Acoustic Computerized Tomography

Abstract

We study the optimal transducer arrangement for measuring temperature distribution in an arbitrary convex-shaped space by acoustic computerized tomography (A-CT) using flexibly arranged transducers. The transducer arrangement is optimized by the combined use of two methods, namely, real-coded genetic algorithm (RCGA) and simulated annealing (SA). By RCGA, the optimized arrangement is globally searched and then locally searched by SA. A fitness function of these methods is defined to evaluate the distribution of projection data in θ– r space. By use of this function, we aim to obtain better projection data sets by two-dimensionally interpolating the projection data. By numerical simulations, we confirmed the adequacy of our method and the optimized transducer arrangement.