Extracting knowledge from substations for decision support

The growth in substation data generated by microprocessor-based IEDs has far outstripped our capability for interpretation. In competitive energy markets, the success and survival of power utilities hinge upon their ability to skillfully locate, analyze and make use of the information. Hence, it is now necessary to consider moving beyond data acquisition to knowledge discovery. To accomplish this task, we require a new information processing and data analysis tool that can intelligently process a large volume of data. The paper describes the data explosion problem and illustrates how a new computational intelligence approach can be used within a substation to group objects of interest into classes indiscernible with respect to some or all of their features. This reduces superfluous and irrelevant data, which then improves the performance of the data analysis tools and helps speed up the knowledge acquisition process. It also provides a condensed report summary that can be used by the operator to react to the emergency.