Automated post-fault diagnosis of power system disturbances

In order to automate the analysis of SCADA and digital fault recorder (DFR) data for a transmission network operator in the UK, the authors have developed an industrial strength multi-agent system entitled protection engineering diagnostic agents (PEDA). The PEDA system integrates a number of legacy intelligent systems for analyzing power system data as autonomous intelligent agents. The integration achieved through multi-agent systems technology enhances the diagnostic support offered to engineers by focusing the analysis on the most pertinent DFR data based on the results of the analysis of SCADA. Since November 2004 the PEDA system has been operating online at a UK utility. In this paper the authors focus on the underlying intelligent system techniques, i.e. rule-based expert systems, model-based reasoning and state-of-the-art multi-agent system technology, that PEDA employs and the lessons learnt through its deployment and online use