Abstract: Power Internet of Things, which is formed by the application of Internet of Things technology to power system, undertakes the massive data streams of power system. It can promote to realize power data business through power business data and power network informationization. Nowadays, massive and fine-grained user-side data cannot be effectively applied because of the centralized data storage and processing methods based on cloud computing in power grid. In order to break the interaction barrier between the user end and the power system due to the limited computing resources, this paper proposes a user-side data application framework based on cloud-edge-user collaboration. Taking three kinds of user-side essential data applications as examples, i.e., prediction and early warning, classification and clustering, and demand response, the framework is designed according to the complementarity between edge computing and cloud computing. It aims to achieve the collaborative cycle of the bottom-top and top-bottom two-way data flow and business flow, so as to mine the value of user-side data widely and deeply. Finally, this paper analyzes and summarizes the key technologies of the framework from three aspects which are physical layer, platform layer and business layer.

Key words: edge computing, cloud-edge-user collaboration, data middle platform, big data, power Internet of Things