Abstract: The shipboard integrated power system (SIPS), which is of high efficiency, reliability and controllability, is the technological frontier and key support for large ships. Digital twin technology provides a new solution for the perception, analysis, and optimization of the SIPS. Our paper explores the digital twin of SIPS, i.e. the SIPS-DT. Aiming at a highly integrated, adaptive, closed-loop, and self-optimized SIPS, we proposed a framework as well as its technical route in accordence with the trend of SIPS development. Specially, considering the dynamic process of the SIPS, a physical model and a data model are built for the various operating conditions and time-scales, taking account of the global and local characteristics, and the rules and experience in a coupling way. Combining the data-driven mode and the knowledge-driven mode, we conduct a data mining. The result leads to a novel and better solution, which enables a better application on the anomaly detection, system optimization, and life-cycle maintenance. Finally, on the basis of the development demand of the SIPS, we discuss some key problems and urgent technologies, which support the typical applications of the SIPS-DT.