The aim of this paper is twofold. First to present multivariable frequ
ency domain techniques as a tool for controller design and dynamic ana
lysis of an autonomous wind-diesel power system. Secondly to study how
robust model based controllers can be designed for such systems. Dyna
mic system analyses using multivariable frequency domain techniques ar
e verified against detailed nonlinear simulation studies. The results
are encouraging in the sense that the main conclusions in terms of rob
ust stability and performance agree very well with the simulation resu
lts. It is also shown that improved performance of the system can be a
chieved using simple model based controllers.