In the process of power system restoration after a large power outage

the power support of wind farms is taken into account that it can speed up the restoration of the system. But the output of wind farms is uncertain

which poses a new challenge for the formulation of system restoration plans. A method of power network reconfiguration restoration is present considering the uncertain output of the wind farm. Firstly

the prediction error uncertainty model is used to describe the uncertain output of the wind farm

and then output strategy is limited by the wind farm to eliminate the negative impact of wind power volatility on grid reconfiguration. Meanwhile

the indicator which reflects the success rate of network restoration is defined to depict that the uncertainty of the wind farm forecast output may lead to the risk of restoration failure. After that

the stochastic dependent-chance programming model is established under the condition that the success rate of node restoration is not lower than a certain confidence level. By maximizing the success rate of network restoration and minimizing the restoration time of grid is regarded as the optimal objective

the model is solved by the hybrid intelligent algorithm composed of discrete particle swarm optimization algorithm and stochastic simulation techniques. Finally

the comparison and analysis of the proposed method and the traditional restoration method is made based on the example of IEEE 39-bus system. The result shows that the restoration strategy proposed in this paper can speed up the process of network reconstruction and reduce the restoration time of the grid while effectively coping with the uncertain output of the wind farm.