摘要
随着风电装机容量的持续增长,风力发电的间歇性和随机性对电网造成的不利影响越来越明显。因此,有效的风电功率预测是解决大规模风电并网的关键问题之一。为此提出采用主成分分析法和果蝇优化广义神经网络(FOA-GRNN)对风电功率进行短期预测。首先采用主成分分析对样本数据进行降维处理,提取有效的主成分以降低预测模型的复杂度。然后,引入自适应步长公式,采用FOA-GRNN对处理后的样本数据进行预测。经湖南某风电厂实际运行数据验证,FOA-GRNN的平均相对误差为8. 81%,相比粒子群算法-径向基函数(PSO-RBF)、PSO-BP等预测模型,具有更高的预测精度和更快的收敛速度,为短期风电功率预测提供了一种有效方法。
As the installed capacity of wind power is continuously increasing,the adverse effects of intermittent and random wind power generation on the power grid become more and more obvious. Therefore,the effective prediction of wind power is one of the key factors to solve the problem of large-scale wind power grid connection. A short-term prediction of wind power is proposed by using principal component analysis and fruit fly optimization based generalized regression neural network( FOA-GRNN). Firstly, the principal component analysis method is adopted to reduce the dimensionality of the preprocessed sample data,and the effective principal component are extracted to lower the complexity of the prediction model. Then,the FOA-GRNN is applied to predict the short-term power based on the preprocessed sample data. According to the practical operation data from a wind power plant in Hunan,FOA-GRNN has an average relative error of 8. 81%. Compared with PSO-RBF and PSO-BP,FOA-GRNN has faster convergence rate and higher prediction accuracy,and it provides an effective prediction method for short-term wind power.
作者
王慧莹
吴亮红
梅盼盼
张红强
周少武
Wang Huiying;Wu Lianghong;Mei Panpan;Zhang Hongqiang;Zhou Shaowu(School of Information and Electrical Engineering, Hunan University of Science and Technology, Xiangtan 411201, China)
出处
《电子测量与仪器学报》
CSCD
北大核心
2019年第6期177-183,共7页
Journal of Electronic Measurement and Instrumentation
基金
国家自然科学基金(61603132,61672226)
湖南省自然科学基金(2018JJ2137,2018JJ3188)
湖南省科技创新计划(2017XK2302)资助项目
关键词
风电功率预测
主成分分析
神经网络
果蝇优化算法
wind power prediction
principal component analysis
neural network
fruit fly optimization algorithm
作者简介
王慧莹,2016年于湖南科技大学获得学士学位,现为湖南科技大学硕士研究生,主要研究方向为风电功率预测。E-mail:821343803@qq.com;吴亮红,分别在2001年于湖南科技大学获得学士学位,2007年和2011年于湖南大学获得硕士学位和博士学位,现为湖南科技大学教授,主要研究方向为计算智能及应用、多目标优化、电力系统优化调度等。E-mail:lhwu@hnust.edu.cn.
