Rapid Double-Layer Identification Method of Multi-Relevant Bad Data in Large-Scale Power Grids

Abstract

Bad data detection and identification is an important part of state estimation. When the relevant bad data appears, however, there is residual pollution and residual submerged condition in currently available methods of bad data detection and identification. In view of the above problem, this article presents a double-layer bad data detection and identification technique. At first, it is based on regularization residual detection method (Rn detection method) to identify the suspect measurement sets. And then, it presents a fast search technique of interrelated suspect measurements to search interrelated measurements in all the suspect measurements of the entire power grid and produce interrelated suspect measurement sets. Furthermore, use double-layer identification method to fast identify the bad data in interrelated suspect measurement sets, in other words, identify all the bad data in entire power grid. At last, taking IEEE39 node power grid for example, this detection method of bad data is analyzed, the accuracy and effectiveness of this method is to be verified.