Automatic track and balance of wind turbine rotors

This paper presents a new method that, by using the harmonic response of the rotor in the fixed frame, first detects imbalances due to pitch misalignment and then automatically corrects them. Since a balanced rotor acts like a filter, quantities measured in a fixed frame of reference only contain frequencies equal to multiples of the number of blades. Therefore, a 1 per revolution (1P) frequency in the fixed frame is a symptom of rotor imbalance. The phase of the 1P signal is uniquely associated with the misaligned blade, while its amplitude relates to the severity of the misalignment. This allows for the formulation of a simple linear model relating 1P harmonic components to pitch misalignment angles. Given the linearity of the model, if a persistent 1P frequency is detected, it is only necessary to further misalign the faulty blade once to allow for the identification of the model. When this has been obtained, it is possible to readily compute the pitch angle that will rebalance the rotor if applied to the faulty blade. The performance of the novel track and balance approach is demonstrated in the paper and particular attention is given to its robustness in the face of uncertainties in the model and the operating conditions.