Abstract:

Thermoelectric generators (TEGs) produce power in direct proportion to the temperature difference between their surfaces. The Seebeck coefficient and internal resistance of the thermoelements (TEs) that make up the TEGs change depending on the temperature change. In simulation studies, it is seen that these two values are kept constant. However, this situation prevents approaching the data of TEG in real applications. In this study, a TEG Simulink/MATLAB ® model has been developed to capture real TEG module data, which considers changing of both the Seebeck coefficient and the internal resistance depending on the temperature difference change. To achieve this aim, a commercially available TEG data used in also academic studies has been used. A boost converter with a perturb and observe (P&O) maximum power point tracker (MPPT) algorithm has been designed to maximize the TEG power. The TEG Simulink/MATLAB ® model data are compared with commercially available TEG data at different temperatures. The error between the actual values and the simulation results, and the mean absolute percent errors (MAPEs) are calculated. The open circuit voltage and short circuit current error rates of the designed TEG module are 0.125% and 0.256%, respectively. The MAPE values of the designed model are 0.5104%, 0.7837%, and 2.0952% for 30 °C, 50 °C, and 80 °C cold surface temperatures, respectively. In addition, simulations are made in order to see the effect of temperature-dependent parameters in a TEG system built using the designed model. While the simulations made with the designed model give realistic results, with the simulations made with constant coefficients, up to 2.63% more power is obtained than the capacity of the system, contrary to reality. Simulation and validated results show that this new TEG Simulink/MATLAB® model gives more realistic results.

Keywords: