Multivariable adaptive controller for a turbogenerator

Abstract

The paper describes the design and application of an adaptive controller which overcomes the nonlinearity of the turbogenerator system. The controller tracks the operating conditions and updates an optimal controller as conditions change. Modelling of the turbogenerator employs an output prediction equation and recursive least-squares algorithm. The design uses identified input/output models, obtained over a range of operating conditions in the PQ-plane. Optimal controllers are designed offline for each zone of the PQ-plane and stored in online memory as a look-up table. The computer monitors the operating conditions of the generator and selects the corresponding optimal controller. The paper presents a practical comparison between the fixed-gain and the look-up-table controllers over a wide range of operating conditions, with various fault conditions, including transient stability boundaries of each controller. The results show that a very high quality of control may be achieved using this adaptive controller.