In this study, a novel adaptive solar dryer with a coating of polymer-dispersed liquid crystals (PDLC) was designed, built, and characterized to evaluate its optical properties and thermal performance by modulating the transmission of solar radiation toward the drying chamber. PDLC films can electrically switch between opaque and transparent states, making them ideal for controlling solar radiation. Unlike conventional direct dryers, the proposed solar drying system takes advantage of direct solar radiation, known for its rapid drying process, while minimizing the impact on product properties. The results show that the PDLC coating exhibited a significant difference in transmittance between the “on” and “off” states under ambient conditions. At 400 nm, the transmittance varied by 20 %, increasing to 30 % at 450 nm. The solar dryer with dynamic control of solar irradiance constantly maintained a temperature range of 52.19 °C–57 °C. Lower relative humidity values were achieved compared to the uncontrolled solar dryer, resulting in a uniform irradiance distribution within the drying cabinet. The solar dryer with dynamic control of solar irradiance (acrylic and PDLC film) presents a total thermal resistance (0.02817 mK/W) exhibiting better heat permanence inside the drying cabinet compared to polycarbonate (one of the materials most used for drying systems). The system's thermal efficiency ensures reliable operation regardless of weather conditions, efficiently capturing and converting solar energy into heat.

This research established an experimental database characterizing the solar drying behavior of seven distinct agricultural products (including fruits, vegetables, and meat) in a tunnel-type solar dryer. Based on this experimental data, a dual-model architecture using multilayer perceptron (MLP) neural network was developed to predict both the internal microclimate and the drying kinetics. The first network (ANN 1) mapped meteorological variables to the dryer’s internal conditions, while the second (ANN 2) predicted moisture loss. The results demonstrate distinct predictive capabilities for each physical phenomenon: the thermodynamic model (ANN 1) captured stochastic weather fluctuations, with an R² of 0.9878 and an MAPE of 4.64%. The kinetic model (ANN 2) achieved near-perfect linearity with an R² of 0.9997 and an MAPE of 0.49%, significantly outperforming baseline linear regression models (R² approx 0.78). These findings confirm the system’s capacity to generalize across diverse food types and variable weather conditions, providing a robust tool for future Model Predictive Control (MPC) applications.