Mejora de la Eficiencia en los Generadores Empleados en Parques ...

ABSTRACT
This thesis introduces a new technique for efficiency optimization of variable speed
drives, with an emphasis on vector-controlled wind generators drives, working at
partial speed and load. The technique combines two distinct control methods,
namely, on-line search of the optimal operating point, with a machine model based
efficiency control. For a given operating condition, characterized by a given turbine
speed (ωT) and electric torque (Te), the search control is implemented via the
“Rosenbrock” method, which determines the level of the flux current component that
results in maximum output power. Once the optimal level of the flux current
component is found, this information is used to update the rule base of an adaptive
fuzzy controller, which plays the role of an implicit mathematical model of the system.
As the optimum points associated with the different operating conditions are
identified, the rule base is progressively updated, so that the fuzzy controller learns to
model the optimal operating conditions for the entire torque-speed plane. After every
rule base update, the Rosenbrock searcher output is reset, but it is kept active to
track possible minor deviations from the optimum point. This way, the speed for an
optimum point is guaranteed, as well as, the true optimum operation for any load
condition. Once a transient condition is detected, the higher torque requirement
inherently forces the rule base output to a higher flux level, since the optimal
condition for high torque is close to rated flux. If compared with other techniques
proposed in the scientific literature, this method shows better performance, because,
the output of fuzzy controller, provide immediately the reference of current flux
component for a maximum efficiency, for any operating point.