基于概率统计的风电出力时间序列生成方法
发布时间：2018-08-09 09:07
【摘要】：风电出力时间序列的生成是研究含风电场的电力系统工作的前提。基于历史风电出力波动特性,提出一种构造风电出力时间序列的方法,首先,将历史风电出力序列做小波滤波处理,得到较为光滑的风电出力曲线,该曲线可看作由大小不同的波动组成,进一步分析不同波动类别下小波滤波误差的概率分布;其次,根据波峰值将波动划分成不同类别,并以高斯函数拟合,统计各月份不同类别波动的高斯拟合函数参数的多维概率分布及波峰间隔的一维概率分布,以马尔科夫转移矩阵描述不同波动类别之间的转换概率;最后,根据统计结果序贯随机抽样波动类别及各统计参数,模拟生成各月风电出力序列。通过对比历史风电出力序列与模拟风电出力序列的评价指标,验证了该文方法的合理性。
[Abstract]:The generation of the time series of wind power output is the premise to study the work of the power system with wind electric field. Based on the characteristics of the historical wind power output, a method of constructing the time series of the wind power output is proposed. First, the historical wind power sequence is processed by the wavelet filtering to get the more smooth wind power output curve, which can be regarded as the size of the wind power. The probability distribution of the wavelet filtering error under different fluctuation categories is further analyzed. Secondly, according to the wave peak value, the fluctuation is divided into different categories, and the Gauss function is used to fit the multidimensional probability distribution of the Gauss fitting function parameters of different types of fluctuation in each month and the one dimension probability distribution of the wave peak interval. Kopf transfer matrix describes the conversion probability between different types of fluctuation. Finally, according to the statistical results of random sampling and the statistical parameters, the wind power sequence of each month is simulated, and the rationality of the method is verified by comparing the historical wind power sequence and the evaluation index of the simulated wind power output sequence.
【作者单位】： 华北电力大学电气与电子工程学院;
【基金】：国家电网公司科技项目(5230JY14003R)
【分类号】：TM614
本文编号：2173609
[Abstract]:The generation of the time series of wind power output is the premise to study the work of the power system with wind electric field. Based on the characteristics of the historical wind power output, a method of constructing the time series of the wind power output is proposed. First, the historical wind power sequence is processed by the wavelet filtering to get the more smooth wind power output curve, which can be regarded as the size of the wind power. The probability distribution of the wavelet filtering error under different fluctuation categories is further analyzed. Secondly, according to the wave peak value, the fluctuation is divided into different categories, and the Gauss function is used to fit the multidimensional probability distribution of the Gauss fitting function parameters of different types of fluctuation in each month and the one dimension probability distribution of the wave peak interval. Kopf transfer matrix describes the conversion probability between different types of fluctuation. Finally, according to the statistical results of random sampling and the statistical parameters, the wind power sequence of each month is simulated, and the rationality of the method is verified by comparing the historical wind power sequence and the evaluation index of the simulated wind power output sequence.
【作者单位】： 华北电力大学电气与电子工程学院;
【基金】：国家电网公司科技项目(5230JY14003R)
【分类号】：TM614
【相似文献】
相关博士学位论文 前1条
1 荣爽;促进供暖期风电消纳的多热源容量规划与协调调度策略[D];哈尔滨工业大学;2016年
，本文编号：2173609
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2173609.html
