Abstract: On-line state estimation can provide real-time, reliable, consistent and complete network state information for the optimal operation and security analysis of the multi-energy network, which is the basic function of the integrated energy management system. To realize the on-line state estimation, there are three main challenges: the universality of different multi-energy flow system scenarios, the difference among different energy network time scales, and the low redundancy of multi-energy network measurement configuration. Given the above challenges, a multi-scene adaptive on-line state estimation functional framework of the multi-energy network was designed. Several practical technologies, such as comprehensive observable analysis under multi-source measurement fusion, modular modeling, steady-state estimation model, dynamic state estimation model, and real-time energy storage state estimation, were proposed. This function is applied in several real tests, which proves the generality and practicability.