Reconstruction of selected operating parameters of a thermoelectric device

Abstract

This paper presents preliminary research aimed at recognizing some selected operating parameters of a thermoelectric device. The inverse problem was formulated, for the solution of which a population heuristics (Ant Colony Optimization) was used. In the inverse task, selected parameters important for the cell operation were reconstructed based on relatively easy to obtain temperature measurements within heat exchangers and appropriate measurements of electrical quantities. The heuristics used, reconstructs the estimated variables, minimizing the differences between data from the measurements and data calculated in the model for their determined values. Since inverse tasks, as ill-conditioned problems, are characterized by high sensitivity to measurement errors, the tests began with calculations based on numerically generated data in order to fully maintain control of their disturbances.