Knowledge extraction within distribution substation using rough set approach

Multifunction microprocessor-based protection relays and intelligent electronic devices (IEDs) have created new opportunities for the measurement of power system signals and their diagnosis. The quantity and complexities of the information has been enhanced to the extent that it is well beyond what is required for the immediate operational issues. A significant amount of data is now continually available and during a serious operational incident or fault, it may become excessive and impossible for a substation operator to handle. A technique used to extract concise information from the data received from microprocessor-based relays will be described in this paper. The data analysis technique is based on rough set theory, which has already been used in many areas of artificial intelligence including machine learning and knowledge-based system. A 33/11kV distribution network was modeled under different types of fault scenarios using EMTDC. The resulting event database was uploaded into the Rough Set Data Analysis Module (RSDAM), which eliminates superfluous data whilst keeping all the essential information.