Dynamic State Estimation Incorporating Innovation Graph

Detection and identification of anomalies in power system state estimation has been a challenged to the researchers. Many debugging procedures have suggested in the state estimation literature but all procedures are suffered from some kind of limitation. Therefore to overcome the limitation in the debugging procedure a new hybrid approach is proposed in this thesis.Among the various schemes, innovation graph (IG) based topology error and sudden load change discrimination scheme avoiding the influence of bad data and bad data and sudden load change discrimination scheme based on forecasting added state estimation (FASE) are the foundation of the proposed method.The proposed Hybrid Dynamic State Estimation (HDSE) scheme, which incorporates innovation graph into dynamic state estimation for the detection and diagnosis of anomalies data gathered during real time power system monitoring in an integrated way. The diagnosis about which type of errors is actually presents in the measurement data set is carried out taking into account the results of innovation graph, normalized innovation and residual test. Numerical results obtained by simulation studies using the proposed estimator are presented and discussed for various practical operating conditions in the thesis.Bad data, topology error and sudden load changes in the system operating points are taken as a cause of anomalies suspicion. Suspicion of anomalies is detected by the inconsistencies between the recent measurement and forecasted values. These inconsistencies appear in the innovation vector. Identification of these inconsistencies is achieved using the proposed method.The proposed method has two main significances. First is that it can identify various anomalies including reactive power bad data in the measurement and the second is overall time required for anomalies identification processes is reduced.