A bio-inspired approach to enhancing wind power conversion

The primary objective of this investigation is to develop a new control method for wind turbines that can increase their wind energy capture efficiency. Since the resultant wind turbine system dynamics are profoundly nonlinear and coupled with significant uncertainties, traditional model-based control is found to be not only structurally complex but also computationally expensive. Here, we explore two sets of control algorithms to enhance wind to electrical energy conversion. The first accounts for system nonlinearities and external disturbances by integrating variable structure control with adaptive control. The second accommodates the nonlinearities arising from rotor aerodynamics and pitch (actuation) dynamics, as well as external disturbances, through a method inspired by a 1st order human memory/learning model. The second method allows direct maximum power coefficient tracking for winds under the rated speed and ensures rated power output for winds over the rated speed. Basically, it uses the system current and most recent memorized responses, together with past control experience, to generate new control actions. Both rotor dynamics and actuation (pitch) dynamics are reflected indirectly through the observed/measured system response at each instant, and are embedded within the control mechanism. Thus, there is no need for detailed information on the system model or system parameters in the control's design and implementation. The efficacy of both proposed approaches is analyzed through numerical simulations.