By considering the bus characteristic uncertainty after wind power integration at the original load bus

a new method based on probability statistics is proposed to learn about and model steady-state characteristics of generalized load buses. To analyze the change of the power direction after wind power integration

the bus characteristics are divided into source characteristics and load characteristics. In view of the change in the uncertainty of bus characteristics

the sample space of active power is segmented adaptively according to the past measured data. The probability distribution is obtained by probability statistics. Levenberg-Marquardt neural network is used to abstract the bus characteristics prior to the development of the unified model. The application of the new model is described by taking the risk analysis as an example. Simulation results show that the method proposed can not only accurately build the model

but also analyze the uncertainty problem by making use of probability statistics according to the different scenarios. The new method has made up the inadequacy of the traditional method in random characteristic description

which is an extension and supplement to the traditional method in uncertain scenario application useful for simulation analysis and dispatch control following wind power integration.