基于L2范数组合云的风电场短期风速–功率拟合方法

Short Term Wind Speed-power Fitting Method for Wind Farms Based on L2 Norm Combination Cloud Model

为解决大型风电场运行分散性引起的风速–功率拟合误差较大的问题,提出一种基于L2范数组合云的风电场短期风速–功率拟合方法。首先,针对不同风速下风电场功率概率密度曲线峰值特性不同问题,以其概率密度曲线峰值点为边界点,建立组合云拟合模型;其次,为了避免样本峰值扰动引起的组合云模型拟合误差升高,基于L2范数理论对风电场特定风速下的功率离散点进行多项式拟合,以该多项式的峰值点作为边界,对上述组合云拟合模型进行修正;最后,采用贝叶斯法对上述组合云拟合模型参数进行计算。经仿真验证:对具有不确定性的不规则单峰或多峰风速–功率概率分布,所提组合云拟合方法均能获得较高的拟合精度,拟合误差较小。

In order to solve the large wind speed-power fitting error problem caused by the dispersion and uncertainty of wind farm operation, a novel short-term fitting method based on L2 norm combination cloud model was proposed.Firstly, a combination cloud model was established with the peak point of the curve as boundary point, to solve the problem that power probability density distributions differ from each other under different wind speeds. Secondly, in order to reduce the fitting error caused by the disturbance of sample peak, the L2 norm theory was used to fit the power discrete point, and the combination cloud model was modified with L2 norm peak instead of the original peak point. Finally, the Bayesian method was used to calculate the parameters of the combination cloud fitting model. It is verified by simulation that the proposed combination cloud fitting method can get higher fitting accuracy and smaller fitting error for the wind-power curve either with single or multi-peak characteristic.