Fuzzy logic technology in reactive power control

This paper presents results of research aimed at improving efficiency of coordinated on-line control of reactive power on the basis of developing models, methods, and a control system, integrating traditional numerical techniques with fuzzy logic technology. An approach associated with experimental design is considered to construct diverse types of sensitivity indices, which serve as the basis for multi-attribute evaluating efficiency of control actions and, therefore, for forming rules included in a knowledge base. Procedures related to classifying power system states are presented. Questions of the construction and tuning of linguistic variables as well as rational implementation of fuzzy logic procedures are discussed.