A Digital twin-driven Framework for the Online Prediction and Control of Assembly Quality toward the Solid Rocket Engine

Abstract

With the rapid development of new generation technologies, the digital twin (DT) has become a core focus for the solution of assembly process in the fourth industrial evolution. So, based on DT, aiming at the unobservable, unpredictable, and uncontrollable problems during the assembly process of complex products like large solid rocket engines, we propose a digital twin-driven framework for online prediction and control method toward assembly quality for complex products. Under this framework, firstly, the high-fidelity modeling method is presented involving 3D CAD and assembly quality model based on point cloud towards unobservable problems. Secondly, online deviation traceability and quality prediction are introduced to aim at unpredictable problems. Thirdly, intelligent regulation and control mechanism for assembly quality based on DT, AR, deep learning, etc. is developed for uncontrollable problems in assembly process. Finally, the large rocket engine nozzle is taken as a case to verify the entire framework and method effectiveness.