As new power systems develop, the uncertainty and complexity of distribution network operation are increasing. Using digital twin technology to establish the mapping and interaction between virtual spaces and physical power grids is an effective way to reveal the complex operation of distribution networks. A method of constructing distribution network operation portraits based on digital twins is proposed. First, an index that reflects the operations of supply districts and feeders is established, and a statistical ranking strategy is used to transform the characteristic index into a fuzzy label template. Then, the day operation index is extracted as a unit of acquired samples and as a cloud drop of the cloud model. Fuzzy expressions of the index are obtained using a reverse cloud generator. The fuzzy distances between the index and the labels are weighted and summed to obtain the score of the operation. A distribution network operation portrait with clear physical meaning and objective evaluation basis is formed. The distribution network operation portrait also provides a technical means for comprehensively depicting the operational characteristics of the new digital distribution network. This work is supported by the Key Research and Development Program of Hebei Province (No. 21312102D).