enhancing photovoltaic system performance prediction: a synergistic approach with fuzzy logic and response surface methodology

In recent years, extensive and expensive research has been conducted on solar energy systems. studying and investigating solar panel output requires considerable experimental work, increasing both time and costs. this research aims to reduce these by integrating fuzzy logic with response surface methodology (rsm). in fuzzy models, data inputs are processed through membership functions and rules based on expert knowledge or assumptions. these rules generate outputs, which are then defuzzified into actionable decisions. these outputs were used as inputs in the rsm to develop a statistical prediction model. the model developed is based on three inputs: light intensity, temperature, and humidity, with one output: power. the fuzzy model was processed assuming two levels for humidity, temperature, and light intensity. the rsm was designed using data extracted from the fuzzy system for seventeen runs, using the box-behnken design (bbd) as part of the rsm with design-expert software. the advantage of using bbd is that it avoids extreme corners in the design. the results were analyzed using analysis of variance (anova). the anova table showed significant results for the quadratic regression model. the results were compared with real data using random samples of twenty readings each for two-time intervals. the validation showed variations averaging 7.50% and 5.53%.