Publication

Abstract : The convergence of Digital Twin (DT), Internet of Things (IoT), and Artificial Intelligence (AI) presents a transformative opportunity for precision agriculture. Although promising, the application of virtual replicas of digital twins that enable bidirectional data flow between physical and digital entities remains nascent in this domain. This work proposes a novel DT framework for smart irrigation that uses a cloudbased EnergyPlus architecture and the Decision Support System for Agricultural Technology Transfer (DSSAT). The core of this framework is an advanced predictive model that utilizes embedded real-time sensor data (soil, weather, and crop) from agricultural fields. The model performs two critical functions: univariate time series forecasting of rainfall and multivariate forecasting of quarterly crop yield. A hybrid AI architecture, combining Gated Recurrent Units (GRU) with Bidirectional Long Short-Term Memory (BiLSTM) networks (GRU-BiLSTM), is developed to achieve high fidelity predictions. Empirical evaluation confirms the superiority of the model over existing benchmarks, achieving an accuracy of 98%, a precision of 97%, a recall of 99%, and an F1 score of 99%. This research validates the efficacy of integrating a sophisticated AI forecasting model within a DT environment, offering a powerful decision support tool for farmers to optimize irrigation strategies and improve agricultural productivity.