A severe challenge in controlling wind turbines is ensuring controller performance in the presence of a stochastic and unknown wind field, relying on the response of the turbine to generate control actions. Recent technologies such as LIDAR, allow sensing of the wind field before it reaches the rotor. In this work a field-testing campaign to test LIDAR Assisted Control (LAC) has been undertaken on a 600-kW turbine using a fixed, five-beam LIDAR system. The campaign compared the performance of a baseline controller to four LACs with progressively lower levels of feedback using 35 hours of collected data. The collected data indicates that utilising measurements from multiple range gates on a pulsed LIDAR system can result in rotor averaged wind speed (RAWS) estimates with greater levels of correlation with wind speed at the rotor than using a single range gate. The LACs showed higher levels of speed control performance with significantly reduced levels of pitch activity and generally lower levels of tower excitation. Although the loading spectrum for the test turbine was dominated by responses at twice the rotor speed (2P) and the first tower fore-aft natural frequency, the reduction is likely to show greater relative significance on typical full-sized turbines, which show lower excitation levels due to harmonic clashes.
展开▼
