风速时间序列混沌判定方法比较研究

Comparative Study on Chaos Identification Methods of Wind Speed Time Series

混沌识别是对非线性时间序列进行混沌预测的前提。针对时间序列风速确定性与随机性相结合的复杂非线性特征,研究了不同的混沌识别方法,并对风速时间序列进行混沌特征识别。应用随机噪声、周期运动及经典混沌系统的时间序列对所选方法进行可靠性验证。对美国国家风能研究中心M2测风塔实测时间序列风速数据进行非线性混沌特征识别。结果表明：风速时间序列具有明显的混沌特征;各风速时间序列表现出不同程度的混沌特征;各混沌识别方法对风速时间序列混沌特征的表达形式不同,互为补充,相互验证。

Chaotic identification is a prerequisite for chaos prediction of nonlinear time series. Aiming at the complex nonlinear characteristics of time series wind speed which is deterministic and stochastic, different chaos identification methods are studied, and chaotic characteristics of wind speed time series are identified. First,the reliabilities of the proposed methods are verified by using the time series of random noise, periodic motion and classical chaotic system. Secondly, the nonlinear chaotic characteristics of the wind speed data from NWTC M2 tower are studied. The results show that the wind speed time series have clear chaotic characteristics. Different wind speed time series show different degrees of chaotic characteristics. Different chaos recognition methods are different from each other and complement each other to verify the chaotic characteristics of wind speed time series.