Multi-Point Locational Wind Speed Estimation from Meso-Scale Wind Speeds for Wind Farm Applications

This paper proposes a novel methodology to translate meso-scale wind speeds to wind turbine wind speeds at a large wind farm. This is accomplished by modeling wind speed dispersion using a statistical distribution, coupled with a Monte Carlo model approach which employs random sampling and optimized seeding to arrive at a deterministic solution. The optimization procedure returns different spatial distributions partitioned seasonally, as well as by wind flow conditions for improved accuracy. Results are presented for a utility size wind farm and it is shown that the proposed method performs well, with a Mean Absolute Error of approximately 0.8 ms
^-1
for any given season. This methodology has potential merit for specialized applications which require wind turbine wind speeds such as High Wind Speed Shutdown (HWSS).