Virtual-Real Interaction and Status Monitoring of WK35 Bucket Long-Distance Laser Heating System Based on Digital Twin

Remote laser heating of WK35 buckets faces several constraints, including late-state detection, empirical parameter settings, and inadequate synchronization between the physical and virtual spaces, which affect the quality and stability of the heating process. This research aims to develop a monitoring and control system for the heating process by applying the digital twin concept, including multi-source sensing, two-way physical and virtual interaction, and intelligent prediction. The first stage includes the sensing of temperature, laser power, and movement trajectories and their synchronization with the dynamic virtual model. Meanwhile, the second stage includes the return of the results of digital twin operations to the physical world and adaptive adjustment of heating parameters depending on the current state. The states of the equipment and faults are detected with the help of closed-loop control and Transformer-BiLSTM spatiotemporal feature fusion algorithms. As a result of the experimental study, it was found that the reduction of synchronization error in comparison with the classical PID control system, no power overshoots were detected below 0.05 kW, and stabilization was reached within 0.8 seconds. The temperature of the main heating area was rather constant and equal to 965°C and the AUC of the anomaly detection algorithm was 0.96.