A fuzzy logic-based temperature prediction model for indirect solar dryers using Mamdani inference under variable weather conditions

The drying process in indirect solar dryers, which is strongly affected by rapidly changing ambient conditions, represents a highly nonlinear and dynamic system. Accurate modeling is essential for performance evaluation, process optimization, and reliable prediction of drying chamber temperature, which is crucial for ensuring efficient moisture removal while maintaining the nutritional and sensory quality of dried products. In this study, a fuzzy logic-based modeling approach using the Mamdani inference system was developed to predict the drying chamber temperature under a wide range of operating conditions. Experimental measurements were performed with solar radiation varying from 400 to 950 W/m² and the ambient temperature ranging from 20 to 50 °C, covering both static and dynamic system responses. The fuzzy model inputs consisted of solar radiation and ambient temperature, represented by five triangular membership functions (“very low,” “low,” “medium,” “high,” and “very high”) for solar radiation, and three triangular membership functions (“cold,” “warm,” and “hot”) for ambient temperature. The output variable (drying chamber temperature) was modeled with five triangular membership functions (T1–T5). The Mamdani system employed 15 “if–then” rules, and the centroid method was used for defuzzification. Model validation was conducted across the full range of operating conditions, showing strong agreement between predicted and experimental data. For instance, at 700 W/m² and 46 °C, the predicted temperature was 50.9 °C versus a measured temperature of 51.0 °C, while at 750 W/m² and 50 °C, the prediction (52.0 °C) closely matched the experimental value (51.8 °C). Statistical evaluation yielded RMSE = 0.38 °C, MAE = 0.29 °C, and R² = 0.997, confirming high accuracy and robustness. These results demonstrate that Mamdani fuzzy logic can effectively model the thermal behavior of solar dryers under diverse climatic conditions, and they provide a solid basis for developing real-time intelligent control strategies to optimize energy efficiency and product quality.