A joint possibilitic-probabilistic uncertainty power flow model based on D-S theory is developed. Classified models are built for input variables of both probability and possibility according to different characteristics of various distributed energy and loads

and an unscented transformation(UT)method as well as α- cut set algorithm is used respectively for solving the uncertainty power flow problem. The off limits of uncertainty power flow results are discussed and an operational risk analysis system with respect to power loss is set up. By improved IEEE 33-bus power distribution system model simulation

verifications of three-phase balanced and unbalanced cases are made of the necessity of the joint uncertainty model and the timeliness of the algorithms

and risk buses and branches are picked up to realize risk perception of the active distribution network. This work is supported by National Natural Science Foundation of China(No. 51577096)

Beijing Natural Science Foundation(No. 3161002)and Basic and Prospective Program of State Grid Corporation of China(No. DZ71-15-004).