<![CDATA[Vertical farming is a crop cultivation system with a tiered planting medium configuration designed to optimize accessibility, maintenance, and harvesting efficiency. The implementation of modern technology based on environmental sensors allows real-time monitoring of microclimate parameters to support precise irrigation management through estimation of evapotranspiration rates (ETo). This study aims to evaluate and estimate ETo values in vertical farming systems using a DNN architecture. Estimation is carried out through Python programming language simulations on the Google Colaboratory platform using a pre-trained DNN model (4 hidden layers) based on input data of average temperature (Tmean) and average relative humidity (RHmean) over a 4-hours duration. The implemented DNN model was validated against actual ETo data in previous studies to ensure the reliability of predictions. The results show that DNN-based evapotranspiration values are significantly influenced by temperature and relative humidity factors. Furthermore, evapotranspiration values, plant growth phases, and planting area are variables needed to calculate irrigation water requirements in the vegetative, generative, and final phases, which require 6.41 liters, 22.85 liters, and 21.73 liters, respectively. Thus, the use of the validated DNN model is proven to be a reliable predictive instrument for precisely determining crop water requirements to achieve more efficient irrigation management.]]>
Copyrights © 2026
