Research and application of a real-time operation risk assessment method for power grid based on random forest algorithm

Abstract

Real-time operation risk assessment of power grid is not only an important content of power grid dispatching and operation control, but also an important reference factor in power grid accident disposal. It is difficult for the traditional evaluation methods to meet power system online operation control requirements, because of their complex process and long calculation time. To this end, a real-time power grid operation risk assessment method based on random forest intelligent algorithm is proposed in this paper, The basic concepts and implementation points of random forest algorithm is introduced. The implementation framework of real-time operation risk assessment based on random forest is proposed taking the need of power grid operation risk assessment into consideration. Compared with the traditional evaluation method, this method can calculate the power grid operation risk level directly according to some key operation indexes without complex power grid analysis, such as N-1 security check. Finally, the two-years practical application in a certain provincial power grid in China shows that the method has high evaluation accuracy and computational efficiency, which can match the traditional evaluation method and improve the calculation efficiency and accuracy of risk assessment.