Performance Evaluation of Membership Function on Fuzzy Logic Model for Solar PV array

Fuzzy logic is based on the observation that human decisions involve imprecise and non-numerical information. Fuzzy models have the capability of recognizing, representing, manipulating, interpreting, and utilizing data and information that are vague and lack certainty. Fuzzy logic has been applied to many fields, including control theory, management, medical diagnosis, artificial intelligence operations etc. One of the main difficulties of fuzzification of the input data is the selection of an appropriate membership function. The shape of the membership function is responsible for the effect on the particular fuzzy inference system. Different shapes of fuzzy membership function like Gaussian, Trapezoidal, Triangular, etc are available. It is presumed that the right type of membership function will yield the best results in terms of accuracy of prediction. However, the choice of the most appropriate membership function for a given fuzzy model has been subject of much uncertainty. The proposed research paper focuses on the comparative evaluation of different fuzzy membership functions to evaluate the output performance of a solar photovoltaic (PV) array panel. Three membership functions namely, triangular, trapezoidal and gaussian are applied to fuzzify data pertaining to performance of the solar PV array for three different temperature levels. The performance is in terms of the power outputs of the panel for solar radiation levels extending from 200 W/m2 to 1000 W/m2. The best choice of the membership function for the model is decided in terms of the mean square error (MSE) with reference to the working database outputs. One of the main difficulties of fuzzification of the input data is the selection of an appropriate membership function. The shape of the membership function is responsible for the effect on the particular fuzzy inference system. Different shapes of fuzzy membership function like Gaussian, Trapezoidal, Triangular, etc are available. It is presumed that the right type of membership function will yield the best results in terms of accuracy of prediction. However, the choice of the most appropriate membership function for a given fuzzy model has been subject of much uncertainty. The proposed research paper focuses on the comparative evaluation of different fuzzy membership functions to evaluate the output performance of a solar photovoltaic (PV) array panel. Three membership functions namely, triangular, trapezoidal and gaussian are applied to fuzzify data pertaining to performance of the solar PV array for three different temperature levels. The performance is in terms of the power outputs of the panel for solar radiation levels extending from 200 W/m² to 1000 W/m². The best choice of the membership function for the model is decided in terms of the mean square error (MSE) with reference to the working database outputs.