A Method of Predicting the Cascading Trip Accident in Power Network Considering the Operation State

Concerning the actual status of protection in the power grid, a new strategy is proposed in this paper to evaluate the cascade tripping. Based on an operating equation of the backup relay protection, the relationship between nodal injection power and the cascade tripping is analyzed. The relationship between node injection power and cascade tripping is reconstructed by pattern recognition technology, which is employed to assess the probability of cascading failure in a power network subjected to an initial failure. A certain number of samples are obtained by simulation in IEEE39 system, which are trained and tested based on BP neural network pattern. The validity of the assessment method is verified by the calculation result.