Gêmeos digitais aplicado no monitoramento e controle de estufas

Abstract

This work investigates the application of digital twins in the control of greenhouses with the aim of optimizing the use of resources in agricultural production. Challenges such as climate change, the need for sustainable food production and digital twin concepts are considered. Digital twins are divided into the physical and virtual domains. For the physical domain, an embedded simulator was developed that is able to control temperature, humidity and radiation in greenhouses and provide data. Mathematical models were used to represent the dynamics of the microclimate and DSSATTools was used as a tool to interpret the behavior of the plants. With the help of the simulator, it was possible to generate the necessary data to compile a training data set for the computational model of the microclimate using RNN-LSTM. Thus, the RNN-LSTM model and the DSSATTools tool were used as digital twins. The system was validated by comparing it with traditional monitoring methods and fixed setpoint PID control, demonstrating higher energy efficiency. The digital twin achieved a gain of 18% and 61% in the initial and optimal phases compared to the higher-level controller. The results show that the use of digital twins not only improves microclimate control, but also enables more accurate predictive analysis and optimized resource management. It is concluded that technology can significantly change agricultural practices in controlled environments, offering innovative and sustainable solutions.