Development of curve substraction and ANFIS based MPPT for thermoelectric generators

For a sustainable development, energy must be produced from renewable sources and with high efficiency environmental technologies. In this context, energy conversion losses can be reduced and systems can be reduced to simpler structures by developing systems that directly convert energy. Thermoelectric generators (TEJ) contribute to energy efficiency by generating electrical energy from waste heat. In order for the power produced by the TEJ to be transferred to the load at the maximum level, the internal resistance of the TEJ and the load resistance must be equal. Since this equation is not always possible, various maximum power point tracker (MPPT) methods are applied to ensure the match between the TEJ internal resistance and the load. In this study, the ANFIS-based curve extraction MPPT algorithm (EÇ – ANFIS), developed over the traditional Perturb-and-Observe (P&O) method, is proposed to detect and track the maximum power point (MPP) of TEJ. The performance of the proposed MPPT design is verified by MATLAB simulation study under different thermal and electrical operating conditions, and the results are compared with a conventional P&O based MPP tracker. In the proposed method, in order to find the MPP and maintain a stable output around the MPP, the MPP is directly found through subtracting mathematical function of the power-voltage curve of the TEJ and tracked by changing the duty cycle of the DC – DC converter via ANFIS. The MPP of the TEJ and the distance (error) of the operating point to the MPP were determined quickly and clearly through the power – voltage function specific to the instantaneous operating condition. The width of the perturbation step was adjusted by ANFIS according to the size of the error and the rate of change. Variable perturbation step size enabled rapid capture of MPP and elimination of steady state oscillation. In the experiments, it was observed that the proposed system reached the MPP quickly between 2 and 5 steps, and the steady state oscillation was reset after reaching the MPP. The proposed method offered increased MPPT capability with only software modification, without requiring any additional hardware compared to the traditional P&O method. The demands from the MPPT technique to track fast and to minimize steady-state error were met.