Self-tuning controllers for turbogenerator excitation and governing systems

Authors:

Abstract
The paper describes the development of self-tuning controllers for excitation and governor control of turbogenerators. Two approaches have been adopted. In one, separate self-tuners are designed for the exciter and governor loops, whereas the other combines these functions in a multivariable controller. The algorithms are based on minimum-variance and detuned-minimum-variance techniques. The controllers have been implemented and tested on a laboratory micro-machine system. Particular attention was paid to practical aspects of implementation, and the software was structured to safeguard parameter estimation, and ensure effective control during normal and abnormal operating conditions. The test results include responses to major disturbances, and illustrate the effective operation of the controllers, together with good system performance.