Model-based maximum power point tracking (MPPT) of wind generators (WGs) eliminates dead times and increases energy yield with respect to iterative MPPT techniques. However, it requires the measurement of wind speed. Under this premise, this paper describes the implementation of a high-performance virtual anemometer on a field programmable gate array (FPGA) platform. Said anemometer is based on a growing neural gas artificial neural network that learns and inverts the mechanical characteristics of the wind turbine, estimating wind speed. The use of this device in place of a conventional anemometer to perform model-based MPPT of WGs leads to higher reliability, reduced volume/weight, and lower cost. The device was conceived as a coprocessor with a slave serial peripheral interface (SPI) to communicate with the main microprocessor/digital signal processor (DSP), on which the control system of the WG was implemented. The best compromise between resource occupation and speed was achieved through suitable hardware optimizations. The resulting design is able to exchange data up to a 100 kHz rate; thus, it is suitable for high-performance control of WGs. The device was implemented on a low-cost FPGA, and its validation was performed using input profiles that were experimentally acquired during the operation of two different WGs.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited