A Virtual Actuator for Advanced Individual Pitch Control (IPC)

Abstract

We present a virtual actuator concept for wind turbine control, wherein rotor force and moment references from various controllers are optimally combined to compute individual blade pitch angles. The approach aims to minimize pitch bearing wear, too. The combined references can come from known control algorithms such as speed regulation, tilt-yaw control, tower dampening and helical wake control [3]. We formulate this as an optimization problem which we solve in an MPC fashion, however, instead of the usual prediction horizon over time, we use a discretized azimuth map as our finite horizon. Serving as a unified interface for all control features utilizing individual pitching, the virtual actuator replaces the plethora of multi-blade transformations and gain-scheduling functions in traditional IPC with one coherent function. Users can directly prioritize features and input constraints on actuators and structural loads. Notably, upstream control algorithms provide rotor force or moment references rather than pitch references. Simulations showcase the virtual actuator’s ability to compute intricate pitch trajectories, surpassing the capabilities of conventional IPC methods. Our method yields novel individual pitching which optimally merges conflicting IPC objectives while minimizing actuator wear.