Determination of Output Parameters of a Thermoelectric Module using Artificial Neural Networks

摘要

The concept of the thermoelectric (TE) defines the direct transformation between thermal energy and electrical energy. This transformation is performed by the TE semiconductors most efficiently [1, 2]. TE modules are obtained by connecting tens of P and N type of TE semiconductors as electrically serial and thermally parallel [3, 4]. TE modules are working as a heat pump by pumping the heat in one surface to the other surface. Thus, while the one surface is cooling, the other surface is heating. If the heat energy that occurred in one surface is sinked by using a fun or another device, the temperature of the cooling surface can be reduced further [1]. TE modules have many advantages which are silent operation, being durable, small size and practical applicability. Nowadays, there are many application areas which are especially heating and cooling or production of electrical energy as a thermoelectric generator. Example application areas are medicine, space working and hybrid cars etc. [5, 6].