Research Of Digital Twin For Hull Structure Based On Model Order Reduction Technology

In recent years,due to rapid development of information technology,a new round of industrial revolution has sprung up,aiming to complete the transformation from traditional manufacturing to intelligent manufacturing.As the best solution to realize the two-way interaction between physical space and digital space,digital twin can provide reliable guarantee for high-quality sustainable development,and has gradually become the heating topic in various fields in the world.Within shipbuilding industry,the interdisciplinary research on this technology has not been fully developed,and the practical application of this technology is still under the exploratory stage.In order to realize the rapid calculation of hull structure response and fill the gaps existed in the studying areas of ship digital twin,this paper discusses the technical connotation and application direction of digital twin,and determines a method to construct a reduced order model of the hull structure based on model order reduction technology.It can greatly reduce the calculation time while maintaining the high precision,and achieve the real-time performance in the digital twin,which is of great significance to realize the digital twin within shipping field.The main works of this paper are as follow:(1)Based on the five-dimensional model,the digital twin architecture of the hull structure is proposed,and the functional meaning and collaborative relationship of each dimension is explained in detail.According to the characteristics of the whole life cycle of digital twin,the application direction and purpose of the three stages are analyzed.The multi-dimensional modeling and simulation technology that in the key technology of digital twin is selected for specific research,and a model order reduction method combining proper orthogonal decomposition and Kriging surrogate model is determined.The optimal base is intercepted by proper orthogonal decomposition,and the full-order space of the results solved by finite element method is projected into the vector space to achieve order reduction.Then,the relationship between the base coefficient and the system input is established through the kriging surrogate model,and the predicted response value of the original system is obtained.(2)The structural static and dynamic reduced order models were constructed by the model order reduction method combining the proper orthogonal decomposition and the kriging surrogate model.The typical plate panel was taking as the research object,its applicability under the situation of Bi-axial thrust and lateral pressure was verified.The results show that the relative errors of displacement and stress response in each state that generated by the reduced order model of square plate are mostly less than 1%,and the calculation efficiency is improved by 3 orders of magnitude.It lays the foundation for the rapid analysis of the hull structure response in the following.(3)The model order reduction method in this paper was used in the complex hull structure.The static and dynamic responses were evaluated by the conditions of the four-point bending and the hull bottom subjected to periodic loads separately,and the accuracy and efficiency of the reduced-order model of the hull structure were verified.Based on this,the structural dynamic response prediction method reusing the known base modes is studied,which further improves the calculation efficiency and application scope of the original method.Finally,the real-time calculation of the dynamic response of the hull structure under different working conditions is realized,which provides technical reference and scheme guidance for the realization of digital twin in the field of shipping.