Grid Frequency Estimation Using Multiple Model with Harmonic Regressor: Robustness Enhancement with Stepwise Splitting Method
Paper in proceedings, 2017
Reduction of inertia in electricity networks due to high penetration level of renewable energy sources will require wind turbines to participate in frequency regulation via Active Power Control. The performance of frequency regulation and protection system depends strongly on
the performance of network frequency estimation. Fast frequency variations and uncertainties associated with unknown harmonics and measurement noise in the network signals are the main obstacles to performance improvement of frequency estimation with classical zero crossing
method, which is widely used in industry. The same uncertainties introduce challenges in model based frequency estimation. These challenges are addressed in this paper within the framework of multiple model with harmonic regressor.
Additional challenges associated with computational complexity of matrix inversion algorithms and accuracy of inversion of ill-conditioned matrices in the multiple model are also discussed in the paper. New high order algorithms with reduced computational complexity are presented.
Instability mechanism is discovered in Newton-Schulz and Neumann matrix inversion techniques in finite precision implementation environment. A new stepwise splitting method is proposed for elimination of instability and for performance improvement of matrix inversion algorithms
in the multiple model. All the results are confirmed by simulations.
High Order Algorithms
Accurate Frequency Tracking in Electricity Networks