ABSTRACT
As shown in Figure 1, the wind-power points scattered around the power curve, usually wind turbine has its own design parameters, such as cut-in wind speed vIn , cut-out wind speed vOut and rated wind speed vN . Since the turbine has different control strategy at different speeds, and the power satisfies different distributions, here we partition the whole wind speed range by minimal interval Δv (like 0.2m/s), which is much less than the value of v vOut In , to obtain different operation conditions. We take the average speed of each minimal interval, vi , as the speed of all points in the interval. Then several common distributions (beta, exponential, gamma, generalized extreme value, logistic, normal, Rayleigh, Weibull, Poisson and so on) are applied to fit the power distribution in each minimal interval of wind speed by maximum likelihood estimation, and the best distribution is selected by the principles such as negative of the log likelihood (NLogL), Bayesian Information Criterion (BIC),
1 INTRODUCTION
With the increasing of wind power’s installed capacity in recent years, China has become the world’s largest country in this field. According to the wind power development plan in the “13th Five-Year” period formulated by the National Energy Administration, the installed capacity of wind power in 2020 will reach to 200 million kilowatts, which is almost twice as much as the current installed capacity, and the market prospects and scale are very large (Xiao Q, 2014). The increase in single turbine capacity and wind farm turning gradually from land to sea are the two major trends in wind power field in the future (Guo S.Q. 2016). At the same time, the maintenance costs are increasing, according to the literature statistics. The maintenance cost of onshore wind turbine accounts for 10% of the total cost, while the maintenance cost of offshore wind turbine is up to 30% (M.I. Blanco, 2009). Therefore, research on the assessment of wind turbine generation performance is of great significance to reduce costs by optimizing operation and maintenance.
