Data-driven digital twin model for predicting grinding force

Abstract

Digital twin gives a new approach for predictive and monitoring of manufacturing machines which can consider the influence of working condition on grinding wheel and application of prediction. In this article, we a develop methodology for grinding force prediction using digital twin approach, with the vertical double side grinding machine performing the required work while connecting the PLC program. The proposed approach integrates the information obtained from sensor data, physic models, and operational of system to establish the grinding machine model. Data driven modelling and quantification of the model form uncertainties associated with the resulting reduced order models. Simulation results show the proper connection between models and communication. The digital model was programmed to exactly match the operation of the physical system.