摘要
针对光伏输出功率的点预测精度不足、传统支持向量机(SVM)在参数优化方面存在的固有缺陷等问题,根据模糊信息粒化理论和纵横交叉算法,提出一种基于CSO-SVM和模糊信息粒化理论的光伏出力模糊区间预测方案。结果表明:该方案可以得到较好的点预测值及其置信区间,某种程度上解决了传统点预测结果的信息局限性。
Aiming at the problems that the point prediction accuracy of photovoltaic output power is insufficient and the inherent defects of traditional support vector machine(SVM)in terms of parameter optimization,puts forward a scheme of fuzzy interval prediction of photovoltaic output based on crisscross optimization algorithm(CSO)-SVM and fuzzy information granulation theory.The results show that the proposed scheme can obtain better prediction value and its confidence interval,and in a certain degree reduce the information limitation of the traditional point prediction.
作者
陈云龙
殷豪
黄强
周亚武
CHEN Yunlong;YIN Hao;HUANG Qiang;ZHOU Yawu(School of Automation, Guangdong University of Technology, Guangzhou Guangdong 510006, China)
出处
《宁夏电力》
2017年第5期39-44,共6页
Ningxia Electric Power
基金
广东省科技计划项目(2016A010104016)
广东电网公司科技项目(GDKLQQ20152066)
关键词
区间预测
支持向量机(SVM)
模糊信息粒化
置信区间
interval prediction
support vector machine (SVM)
fuzzy information granularization
confidence interval