Estimating Seebeck Coefficient of a p-Type High Temperature Thermoelectric Material Using Bee Algorithm Multi-layer Perception
Küçük Resim Yok
Tarih
2017
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Thermoelectric generators (TEGs) convert heat into electrical energy. These energy-conversion systems do not involve any moving parts and are made of thermoelectric (TE) elements connected electrically in a series and thermally in parallel; however, they are currently not suitable for use in regular operations due to their low efficiency levels. In order to produce high-efficiency TEGs, there is a need for highly heat-resistant thermoelectric materials (TEMs) with an improved figure of merit (ZT). Production and test methods used for TEMs today are highly expensive. This study attempts to estimate the Seebeck coefficient of TEMs by using the values of existing materials in the literature. The estimation is made within an artificial neural network (ANN) based on the amount of doping and production methods. Results of the estimations show that the Seebeck coefficient can approximate the real values with an average accuracy of 94.4%. In addition, ANN has detected that any change in production methods is followed by a change in the Seebeck coefficient.
Açıklama
Anahtar Kelimeler
Thermoelectric material, thermoelectric generator, Seebeck coefficient, artificial neural network, back propagation, bee algorithm
Kaynak
Journal of Electronic Materials
WoS Q Değeri
Q3
Scopus Q Değeri
Q3
Cilt
46
Sayı
8
