Maximum power point tracking algorithm with advanced state detection and regression method for small wind energy systems

摘要

A maximum power point (MPP) tracking algorithm that uses advanced state detection (ASD) and regression analysis (RA) is proposed in this paper. The ASD and RA allow quick and accurate extractions of the system's MPPs with minimal training and oscillation around the MPP. The ASD measures, stores, and analyses sets of the wind systems' rotational speed and power data to identify steady state operation, trends, and wind speed changes. The ASD therefore enables the algorithm to distinguish between meaningful measurements and misleading transient data. The RA utilizes a database of MPPs that is initially populated during the training phase using “perturb & observe” (P&O). The RA requires a small data set to build the regression model of the wind system's maximum power curve. Operating points are determined relative to the model and always progresses towards the MPP regardless of wind speed changes. Performance of the proposed solution is verified through simulation.