Aiming at the problems of low visualization level, inadequate data linkage, and delayed quality assessment under dynamic conditions in traditional grinding unit monitoring systems, this paper proposes a digital-twin-based real-time monitoring system for workpiece surface roughness. Integrating digital space modeling, virtual-reality interaction, and three-dimensional visualization techniques, the proposed system leverages the “five-dimensional model” concept to establish a comprehensive 3D visualization architecture tailored specifically for grinding unit. By capturing three key parameters: grinding speed (GS), robot feed multiplier (RFM) and emery wheel compression (EWC), an Particle Swarm Optimization-Backpropagation (PSO-BP) neural network model is developed to achieve accurate surface roughness (Ra) prediction and anomaly detection. Moreover, virtual-reality interactive simulation further strengthens monitoring effectiveness and responsiveness. Experimental results demonstrate that the proposed neural network model achieves excellent predictive performance in surface quality estimation, with a root-mean-square error (RMSE) of less than 0.003 μm. These findings verify the developed system’s high accuracy and practical feasibility, providing innovative insights and approaches for applying digital twin technology within traditional manufacturing environment.
Jin et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: