Fault diagnosis for boilers in thermal power plant by data mining

Abstract

A new approach to diagnose faults of boilers in thermal power plants is proposed and a hybrid-intelligence data-mining framework is developed to extract hidden diagnosis information from supervisory control and data acquisition (SCADA) system. The hard core of this framework is a data mining algorithm based on rough set theory. The decision table mining from SCADA system is expressed directly by variables in its database, it is easy for engineers to understand and apply. This makes it possible to eliminate additional test or experiments for fault diagnosis which are usually expensive and involve some risks to boilers. This approach is tested in a thermal power plant; the decision accuracy is varied from 91.6 percent to 96.7 percent in different months.