Applying Knowledge Engineering and Data Mining for Optimization of Control Monitoring of Power Plants

C. Athanasopoulou, V. Chatziathanasiou, M. Komninou, and Z. Petkani (Greece)


Power plants, knowledge engineering, data mining


As the electric power generation industry enters a new deregulated era, plants are seeking ways to improve efficiency with the minor possible cost. This paper describes research into the application of data mining algorithms for deriving new control monitoring patterns that can improve the plant’s performance. It proposes the use of a knowledge engineering methodology, CommonKADS, as a tool for completing the Knowledge Discovery in Databases phases that precede data mining. The application of data mining classification algorithms resulted in new control monitoring rules, which improve performance without demanding installation of new equipment. The derived rules can also be used to trace a possible malfunction of a measurement instrument and even more to replace the recording values with those resulting from the data mining algorithms.

Important Links:

Go Back