Research On Operation Performance Evaluation Method For Wind Farm

Nowadays,wind power has become the third largest power supply in China,ranking behind the thermal power and hydropower.Objective evaluation for wind farm operation performance is urgently needed.Performance efficiency has always been quantified by annual energy output and annual utilization hours.However,the influence from wind resource,wind farm design,wind turbine operation performance,wind farm operation and maintenance skills has been ignored.It is unable to either measure the operation efficiency scientifically or provide reasonable suggestions to improve the annual energy output.In this paper,we focused on the analysis of wind farm operation performance.Firstly,factors influencing the wind farm power generation efficiency were analyzed;secondly,computational methods for wind farm designed power generation and theoretical power generation were proposed;finally,the evaluation indexes were build and the evaluation system was achieved.The main work is as follows:1)Analyze the influence factors of wind farm power generation efficiency.Analyze the factors which may cause wind farm energy loss in life-cycle,including the wind resource,wind farm design,wind turbine performance,wind farm operation and maintenance.Based on these analysis results,the definition of the designed wind power generation and the theoretical power gener ation are proposed,which lays the foundation for constructing operation performance evaluation system for wind farm.2)Based on the density-based clustering of application with Noise algorithm,the wind turbine operation condition identifying method was built.Wind turbine operation data are complex and the parameters are nonlinear.Based on the above characteristics,density-based clustering of application with Noise algorithm was proposed to identify the similar operation conditions among the large number of data.DBSCAN model was built to describe and quantify the similarity of wind turbine operation data.The results show that this method can effectively identify the normal operation data among the dataset,and also can distinguish the power-limited data and the under-power operation data caused by turbine fault or other reasons.3)Research on computation method of wind turbine designed power production and theoretical power production.According to the classification results in the above part,the designed wind turbine power production model and theoretical power production model were built.Besides,energy in-loss model was also proposed.The designed wind turbine power production model is based on air density and turbulence intensity correction.In this part,two kinds of theoretical power production model were set up: theoretical power production model based on the theoretical power curve,and theoretical power production model based on segmented support vector machine.The results show that both of this two theoretical power production modesl can calculate the theoretical power production,and the accuracy is 97.5% and 97.1% respectively.4)Research on operation performance system for wind farmIn this part,operation performance evaluation system for wind farm was built,including operation performance evaluation system for wind turbine: wind turbine power curve,wind turbine availability and wind turbine reliability;operation performance evaluation system for wind farm: optimized operation capability for wind farm,maintenance capability and manage capability for wind farm.Compared to the existing operation performance evaluation system,the operation performance can be investigated more comprehensive in the proposed evaluation system.