Skip to Main Content
This paper introduces an efficient and accurate stabilization method for the design of linear optimal controllers which can stabilize a directly- coupled multimachine system via excitation control. A systematic modeling method is first developed to formulate a directly-coupled multimachine system into state space form with directly-measurable state variables. With this form, the stabilization of the multimachine system is converted into a linear-quadratic- Gaussian (LQG) problem. Next, a model reduction algorithm which can deal with any system with real and/or complex poles is developed. Incorporating this algorithm with Solheim's pole-assignment technique as an example provides a new efficient and numerically accurate stabilization method. With this method, a feedback controller able to improve the dynamic stability of a directly-coupled multimachine system under wide-range operation via excitation control is developed. The design is verified by simulated results. The use of directly-measurable state variables also renders a physically implementable feedback controller design.
Date of Publication: Feb. 1984