Wind Turbine Power Performance Measurement And Extrapolation Method Research

Power performance measurements is an important part of wind turbine typecertification, it’s one of theessential indicators toevaluate the performance of a windturbine. It reflects the output power by wind turbine in different wind conditions, andestimates the annual energy production (AEP). But there’s no clearlyshow the method ofpower curve extrapolate, reducing the estimating ability of power performance. Therefore,based on the power performance measurements of wind turbine, perfecting the method ofpower curve extrapolate and estimating AEP is the main task of Power performancemeasurements.Particle swarm optimization algorithm is a kind of the theory of optimizationalgorithm based on swarm intelligence, by Kennedy and Eberhart put forward in1995, andmany scholars have extensive research, at present has been successfully applied tofunction optimization, neural network training, multi-objective optimization and fuzzycontrol system optimization field. This paper, by using particle swarm optimizationalgorithm for power characteristic curve outside push, each data point is defined as aparticle, the use of particle swarm optimization algorithm is optimized, finally obtained onthe basis of actual data outside push data point, and prove its credibility.This thesis gives a detailed study of power performance measurements of electricityproducing wind turbines, puts forward the basic measuring scheme based on IEC61400-12,According to the plan on the ground for power characteristic test, test site for siteevaluation, selected measured wind turbine and anemometer position, wind generator forpower characteristic test, using particle swarm optimization algorithm in actual measurement data on the base of the power curve for the push to gain the complete testdata, so as to work out a complete and accurate in the capacity. Using the real data and theprediction data were compared, which proves that the credibility of the push method.Based on the previous research results, this thesis improves the extrapolate part ofpower performance measurements, using the particle swarm optimization algorithm toextrapolate the limited data, getting a complete power curve, then achieve the purpose ofestimating the power performance measurements.