适用于风速的三参数威布尔分布的参数估计

发布时间：2018-06-18 12:18

本文选题：风速概率分布 + 威布尔分布　；参考：《华北电力大学(北京)》2017年硕士论文

【摘要】：风能受到风速变化的影响,具有很大的随机性、间歇性和不可控性,因此,研究风速分布的概率、选择合适的风速概率分布模型、准确估计模型参数对风资源的合理利用有重大意义。本文围绕风速的概率分布模型和模型参数的估计算法来展开研究。作为一种常用的风速概率分布模型,两参数威布尔分布在概率密度较低的风速带的估计精度不高,因此,本文提出将三参数威布尔分布作为一种风速概率分布的新模型;运用灰色估计法、极大似然估计法、线性回归估计法和矩估计法对三参数威布尔分布的模型参数进行估计,比较四种算法在不同样本下的优劣,确定合适的算法。针对云南两个风电场不同季度的风速数据,将三参数威布尔分布作为当地风速概率分布的模型,运用上述四种算法对模型的参数进行估计,绘制风速概率分布拟合图,计算模型的拟合优度;针对两个风电场的全年风速的实测数据,分别将两参数威布尔分布和三参数威布尔分布作为当地风速概率分布的模型,估计出两种模型的参数,比较两种模型在不同风速带的估计精度;通过对两个风电场实测风速数据的验证,本文得到以下结论:1、当采用灰色估计法和极大似然估计法时,三参数威布尔分布拟合实测风速概率分布是合理的;2、在概率较低的风速带,三参数威布尔分布比两参数威布尔分布的估计精度更高,能更好地反映当地的风资源分布情况。
[Abstract]:Wind energy is influenced by wind speed change, which is random, intermittent and uncontrollable. Therefore, the probability of wind speed distribution is studied and the appropriate wind speed probability distribution model is selected. Accurate estimation of model parameters is of great significance to the rational utilization of wind resources. This paper focuses on the probability distribution model of wind speed and the estimation algorithm of model parameters. As a commonly used wind speed probability distribution model, the estimation accuracy of two-parameter Weibull distribution in the wind speed band with low probability density is not high. Therefore, the three-parameter Weibull distribution is proposed as a new model of wind speed probability distribution. The grey estimation method, maximum likelihood estimation method, linear regression estimation method and moment estimation method are used to estimate the model parameters of the three-parameter Weibull distribution. The advantages and disadvantages of the four algorithms in different samples are compared and the appropriate algorithm is determined. According to the wind speed data of two wind farms in different seasons in Yunnan, the three-parameter Weibull distribution is taken as the model of local wind speed probability distribution, and the parameters of the model are estimated by the above four algorithms, and the fitting map of wind speed probability distribution is drawn. The two parameter Weibull distribution and three parameter Weibull distribution are used as the models of local wind speed probability distribution, respectively, and the parameters of the two models are estimated according to the measured wind speed data of two wind farms. The estimation accuracy of the two models in different wind speed bands is compared, and by verifying the wind speed data measured in two wind farms, this paper obtains the following conclusion: 1, when the grey estimation method and the maximum likelihood estimation method are used, It is reasonable to fit the measured wind speed probability distribution with three parameter Weibull distribution. In the wind speed belt with lower probability, the three parameter Weibull distribution is more accurate than the two parameter Weibull distribution, which can better reflect the local wind resource distribution.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM614

【参考文献】