A fuzzy-logic rule based speed and load controller for combustion turbines in power generation

This paper is concerned with the implementation of a fuzzy logic controller (FLC) in the control system belonging to a Combustion Turbine (CT) applied to power generation. The selected CT is a W501b type installed in two combined cycle power plants in Mexico. The original control system was a hybrid control (analog and early digital) that was upgraded by an all-digital micro-processor based control system in 1993. The speed and load PID control substitution was made with two FLCs, one for the speed control and the other for the load control. We used our experience and process knowledge to develop the control rules. Laboratory simulation tests were carried on a dynamic mathematical model of the CT, and compared against known test cases to validate its results. Initial results were not as good as desired and changes were made to the values of the FLC's scaling factors and to the mappings encodes in the rules, in a so called manual tuning process. After this tuning process the results obtained during the Combustion Turbine "startup" and "picking load" operation stages showed better results that improved those obtained with the PID control.