Implementation of wind turbine controllers

Three of the important, generic, implementation issues encountered when developing controllers for pitch-regulated constant-speed wind turbines are considered, namely, (1) accommodation of the strongly nonlinear rotor aerodynamics; (2) automatic controller start-up/shut-down; and (3) accommodation of velocity and acceleration constraints within the actuator. Both direct linearisation and feedback linearisation methods for accommodating the nonlinear aerodynamics are investigated and compared. A widely employed technique for accommodating the nonlinear aerodynamics, originally developed on the basis of physical insight, is rigorously derived and extended to cater for all wind turbine configurations. A rigorous stability analysis of controller start-up is presented for the first time and novel design guidelines are proposed which can significantly reduce the power transients at controller start-up. The relation to anti-wind-up is noted and several aspects of an existing wind-turbine controller start-up strategy are observed to be novel in the anti-wind-up context. Restrictive position, velocity and acceleration constraints may all be present in wind turbines and the dynamic behaviour of the actuator cannot be neglected. A novel, and quite general, anti-wind-up method, based on the startup strategy, is proposed which caters for all these circumstances. The separate strategies for resolving the implementation issues are combined to achieve an elegant controller realisation which accommodates all the implementation issues in an integrated manner. The importance of adopting an appropriate controller realisation is considerable and is illustrated for a 300 kW wind turbine. The implementation issues encountered in this paper are, of course, not confined to wind turbines but are of wider concern