Solar-Based Thermoelectric Generator and its ANFIS Model

Abstract

In this work, it is aimed to construct an Adaptive Neuro Fuzzy Inference System (ANFIS) model using the experimental values of our previous work on solar heating with wind chimney thermoelectric generator and to predict the generated open circuit voltage of experimental system under variable conditions. The ANFIS model constructed makes use of input parameters such as local radiation intensity on solar collector tube (W), ambiance temperature oC and average wind velocity in the chimney (m/s). Open circuit voltage (V) is denoted as output. Selected experimental data sets are used in training and testing procedures to accomplish the model required. Assessment of the outcomes of the study reveals that the proposed modeling by ANFIS is consistent and validated by the experimental results. Promising results show that ANFIS model can be used to estimate the output parameter of solar-based generator (the open circuit voltage) correctly and this result can use enhancing of presented system. Employment of artificial neural networks on renewable energy systems is a rather new area of study. Hence, continuing work via neural network structures will be related to the optimization and improvement of these generators for useful energy producing.