Abstract: The accessing of a number of intermittent distributed generations (DG) to distribution network may cause the operation risk of voltage and power limit violation. Active distribution network with the capability of active management is an effective way to control the risk. This paper proposes an operation risk assessment method of active distribution network based on probabilistic power flow. By considering the uncertainty of wind power, photovoltaic power and load, probabilistic power flow is solved by using Monte Carlo simulation. The operation risk index of distribution network is then established. In addition, active management strategy is applied to control the violation risk by combining such optimized regulation methods as reactive power compensation, DG power factor and DG active power curtailment. In the end, improved harmony search algorithm is applied to solve the problem of DG active power output curtailment. Through case analysis, the operation risk of distribution network is studied in three scenes, and the correctness of the proposed model and algorithm are verified.