Adaptive and supplementary intelligent control of power system stabilizers

Abstract

The presentation is devoted to the methodology of design of an adaptive control algorithm combined with knowledge-based rules to improve the robustness of power-voltage control under variable load conditions and with a variable number of turbogenerators, which are connected to the tie line system.

This presentation is centered around a control-research project in the Netherlands. The power-generator system is modeled and field tests are carried out for verification. The purpose of the research project is to design and to implement an adaptive algorithm, based upon the PSS-approach, combined with observer filters.

Because of the fact that already a power system consisting of a cluster of generators with variable load conditions, operating in an isolated area, has several well-known boundary conditions, related to different operating points, the MRAC adaptive control algorithm has to be supervised by a set of rules, governed by a knowledge base. This expert-oriented knowledge is based upon the normal “intelligent” control actions as they are carried out by the unit operators. These rules are non-analytic and can be described by fuzzy membership functions.

In the presentation this fuzzy-set approach, in this project used to improve the adaptive-controlled behaviour, will be explained.

Experimental results of the total system will be presented.