THE OPTIMIZATION OF THE ENERGY PERFORMANCES OF A PMRR BY USING NEURAL NETWORKS

摘要

In recent years, a large number of experimental and numerical studies have highlighted the potential of thernpermanent magnetic rotary refrigerators (PMRR) than those reciprocating. For a PMMR, it is well known it isrnpossible to obtain the desired performance by contemporary acting on two operational parameters: the massrnflow rate and the cycle frequency. Consequently, with the aim to improve the energy performances of an actualrnPMRR, it is necessary to experience an innumerable amount of operating conditions regarding mass flow raternand cycle frequency. The present work introduces ANNTEO (artificial neural networks technique forrnoptimization), a technique based on artificial neural networks and able to reduce the number of experimentsrnnecessary to define an optimization map for an actual PMRR. The experimental setup and test procedure arernhere reported to demonstrate the technical soundness of ANNTEO.