Distributed Power System Fault Diagnosis Based on Bayesian Network and Dempster-Shafer Evidence Theory
Sun Mingwei · Dianli xitong zidonghua · 2011
A novel distributed fault diagnosis model based on the Bayesian network and Dempster-Shafer(D-S) evidence theory is proposed.Firstly,a real-time wiring analysis method is used to determine the fault zones to narrow the diagnosis scope.Secondly,two kinds of segmentation method with butterfly and leaf are respectively adopted to reasonably divide the power grid.The concept of coincidence degree is introduced.The geometric mean of coincidence degree and fault coefficient of sub-grid is constructed for the fashioning of D-S evidence theory.Some examples are given for the centralized diagnosis,distributed diagnosis with two kinds of partitioning and two types of D-S evidence fashioning methods.The experimental results illustrate that the distributed fault diagnosis model with butterfly segmentation and geometric mean method is more accurate and reasonable.