Bad Data Detection And Identification Study In Three-phase State Estimation

ABSTRACTThe problem about bad data detection and identification in distribution system state estimation is focused in this thesis.Firstly, the development of distribution system automation and the function of distribution system state estimation are discussed. Present status of the distribution system state estimation, and methods of bad data detection and identification are surveyed here.Secondly, measuring arrangement in three-phase distribution systems, decomposing Jacobian matrix algorithms and how to choose pseudo-measurement weight are discussed.After this, two kinds of algorithms are proposed in this thesis. Both of them are based on a decomposing Jacobian matrix algorithm.The first algorithm uses an iterative self-organizing data analysis technique and fuzzy clustering analysis theory. It is fast, simple and easy for programming, but more suitable for small system. The second one is a recursive algorithm. This one is complicated than the former. By using sparsity matrix technique and recursion, it can also be used in the larger system.The simulation system is constituted with MATLAB and the programs are formed with C language. The results of simulation denote that they are efficient and reliable.Lastly, several conclusions on bad data detection and identification for distribution system are given.