Probability prediction method of branch crossing risk based on gaussian mixture model

In this paper, a probability prediction method based on gaussian mixture model is proposed to predict the risk of branch exceeding the limit. Firstly, according to the minimum and maximum value of active power, each aggregate load in the regional power grid is divided into several sections, i.e. several states of aggregate load. Based on the high-order Markov chain technology, the probability distribution of the state vector of active power at the next time of aggregate load in the regional power grid is given. On this basis, the Gaussian mixture model of each aggregate load is established, and each state of active power of each aggregate load is regarded as a probability distribution in the Gaussian mixture model, thus the gaussian mixture model of active power of each aggregate load is established. Based on the road matrix, the probability distribution function of the active power of each branch of the regional power grid is given, and then the probability distribution function of the active power of the branch at the next time is obtained, which provides the probability information for judging the next time limit of the branch. The effectiveness of the proposed method is verified by an example.