Prediction method of important nodes and transmission lines in power system transactive management

In the operation of electrical power system, there are a few nodes and transmission lines that may cause cascading failures. Nowadays, the algorithms for finding important nodes and transmission lines can only be used in real-time calculation, but does not have the ability to predict important nodes and transmission lines. In order to further prevent cascading failures of power system and provide early warning time for decision-makers, the K-means -Markov chain (K-M), a predicting method for the important nodes and transmission lines in electrical power system, is proposed in this paper, which consists of the historical data mining part based on K-means and the prediction part based on the Markov chain. Moreover, the initial clustering center determination mechanism of K-means also has been improved. The comparative simulation results of Back Propagation (BP) neural network method as well as the Single moving average method based on the IEEE 39-bus system and IEEE 118-bus system proved that the K-M method is reasonable and effective in forecasting important nodes and transmission lines of power system.

Automated summary

This paper proposes a method for predicting important nodes and transmission lines in an electrical power system based on K-means and Markov chain. The method uses historical data mining and prediction to forecast important nodes and transmission lines in the power system. Simulation results prove the effectiveness and rationality of this method.