Fault Risk Prevention Model of Distribution Network based on Hidden Markov

Because of the factors such as the unknown or accidental fault causes and the few factors such as the number of records before and after the distribution network fault, in view of the difficulty of the fault analysis of the distribution network, the risk prediction is inaccurate, the risk exclusion or fault isolation scheme can not be formed systematically and scientifically, so it is difficult to improve the emergency handling capacity of the distribution network. A risk control model of distribution network fault based on the non deterministic finite state machine is proposed, and the migration path of each state in the risk process of the distribution network is analyzed. The non deterministic finite state machine and the hidden Markov algorithm are combined to calculate the state transfer probability through the sample training, and the various input causes are considered comprehensively and comprehensively. In order to minimize the impact of data on the risk prediction results, a series of control schemes are formed, and the correctness of the model is verified by using a more perfect fault data set of typical distribution network in large city.