AI-Driven Intelligent Decision Support System for Digital Twin Industrial Operation

In response to the poor accuracy of industrial operation intelligent decisions and the difficulty of synchronously updating equipment model data with real-time data, this paper took hot stamping production enterprises as the object, and combined the Gated Recurrent Unit (GRU) model and Digital Twin (DT) technology to construct an industrial operation intelligent decision support system. First, based on the equipment model of the actual hot stamping production enterprise, a digital twin model was constructed using digital twin technology. Combined with the five-dimensional four-perspective strategy and Virtual Reality (VR) technology, the digital twin model was formally expressed to ensure the visual display and interactive operation of the digital twin model. Then, the GRU model was utilized to predict and optimize the production process state of hot stamping production enterprise, and a composite material element information entropy model was used to quantitatively analyze the workshop operational state and reliability. Finally, based on the digital twin model of the hot stamping production enterprise, the digital twin system of the hot stamping production enterprise was constructed. The experiment was based on three months of data from a hot stamping production enterprise of a certain company. Combined with the Gated Recurrent Unit–Digital Twin (GRU-DT) method, the decision accuracy reached 96.5%, 18.2% higher than traditional rule-based and standardized decisions. The data update speed reached 5558 pieces per hour, and the data response time was only 0.48[Formula: see text]s. The digital twin hot stamping production enterprise operation intelligent decision system driven by Artificial Intelligence (AI) has improved decision accuracy and achieved synchronous hot stamping production enterprise equipment model data and real-time data updates.