中文摘要：

利用数据挖掘技术,根据风光出力特性采用不同场景数据划分方法对风光资源丰富地区的实际风电、光电出力数据进行场景划分,在划分的典型场景数据下对风光出力互补耦合特性进行分析,研究合成出力跟踪系统负荷的机制特性及提高出力预测精度问题,提出耦合度和跟踪负荷度计算方法.研究表明,风光互补合成出力耦合特性在一定程度上减小了出力的波动性,合成出力对系统负荷跟踪度达到了12.2％,同时风光合成出力预测的耦合所得误差比单独的风、光电出力预测系统所得误差小.

英文摘要：

Based on the data mining technology,the actual wind and photovoltaic power outputs of rich- resource regions are classified by different scenario data classification methods according to their characteristics. Based on the typical scenario data classified, the coupling characteristics of the complementary wind and photovohaic power outputs are analyzed. The mechanism of the coupled power output following the system load and the improvement of power output prediction precision are researched, and a method is proposed to calculate the coupling degree and load-following degree. Research shows that, the complementary coupling of wind and photovohaic power outputs reduces the fluctuation of total power output in a certain degree,its load-following degree reaches 12.2%,and its prediction error is less than that of either wind or photovohaic power output prediction.