摘要
受到光照强度、温度、湿度和风速等气象因素的影响,光伏发电系统出力具有波动性、间接性和不可控性等特点,光伏发电量预测精度较低。对此,本文采用模糊理论结合支持向量机的方法预测光伏发电量。首先通过模糊C均值聚类算法计算模糊隶属度,然后对原始样本进行聚类,生成模糊样本,再采用支持向量机对模糊样本进行训练,最后利用预测模型预测未来几天的光伏发电量。MATLAB仿真实验预测显示:相对于经典的BP神经网络模型和支持向量机模型,模糊支持向量机预测模型更稳定且预测结果误差更小;该模型克服了传统光伏预测方法中存在的极易陷入极小值以及不确定气象因素影响等缺陷,提高了系统预测精度。
Influenced by meteorological factors such as light intensity, temperature, humidity and wind speed,the output of photovoltaic power generation system has the characteristics of fluctuation, indirect- ness and uncontrollable, the forecast results of photovoltaic power prediction has low accuracy. To solve this problem,this paper combines the fuzzy theory with support vector machine （SVM） method to predict the photovoltaic power generation. Firstly,the fuzzy membership degree value is calculated by the fuzzy C- means clustering algorithm. Then, the original sample is clustered to generate the fuzzy sample. Secondly, the support vector machine is applied to train the fuzzy sample. Finally, the FSVM forecast model is em- ployed to predict the next few days＂ photovoltaic power capacity. The MATLAB experimental results show that,the accuracy of the model is higher compared with the BP and SVM model. The FSVM model can overcome the conventional troubles such as easily going into the minimum problem and uncertain meteroro- logical factors.
出处
《热力发电》
CAS
北大核心
2017年第1期116-120,共5页
Thermal Power Generation
基金
广西省教育厅重点项目(ZD2014064)~~
关键词
模糊理论
支持向量机
光伏发电
发电量预测
MATLAB软件
模糊隶属度
BP神经网络
预测误差
fuzzy theory, support vector machine,photovoltaic power generation, power generation capacityprediction, MATLAB, fuzzy membership degree, BP neural network, prediction error
