摘要
风电功率预测为电网规划提供重要的依据,研究风电功率预测方法对确保电网在安全稳定运行下接纳更多的风电具有重要的意义。针对极限学习机(ELM)回归模型预测结果受输入参数影响的问题,现将粒子群优化算法(PSO)应用于ELM中,提出了一种基于粒子群优化极限学习机的风功率预测方法。该方法首先将数值天气预报信息(NWP)数据进行数据预处理,并构建出训练样本集,随后建立ELM模型,利用粒子群算法优化ELM中的输入权值和阈值,从而建立起基于NWP和PSO-ELM风功率预测模型。对华东地区3个不同装机容量的风场NWP数据进行实验。结果表明:该方法的预测精度高且稳定性能好,能够为风电场功率预测以及风电并网安全可靠性提供科学有效的参考依据。
Wind power prediction provides an important basis for power grid planning.It is of great significance to study the method of wind power prediction to ensure that more wind power is accepted under the safe and stable operation of the power grid.Aiming at solving the problem that the prediction results of the extreme learning machine(ELM)regression model are affected by the input parameters,a particle swarm optimization(PSO)algorithm is applied to the ELM,and a wind power prediction method based on the numerical weather prediction(NWP)and PSO-ELM is proposed.Firstly,the NWP data are preprocessed,and the training sample set is constructed.Then,the ELM model is established.The PSO is used to optimize the input weights and thresholds in the ELM,and the wind power prediction model based on the NWP and PSO-ELM is established.The NWP data of the wind fields with different installed capacity in East China are simulated.The results show that the method has high prediction accuracy and good stability and can provide a scientific and effective reference for wind farm power prediction and wind power grid-connected security and reliability.
作者
赵睿智
丁云飞
ZHAO Ruizhi;DING Yunfei(School of Electrical Engineering,Shanghai Dianji University,Shanghai 201306,China)
出处
《上海电机学院学报》
2019年第4期187-192,共6页
Journal of Shanghai Dianji University
基金
国家自然科学基金项目资助(11302123)
上海市浦江人才计划项目资助(15PJ1402500)
作者简介
赵睿智(1993—),男,硕士生,主要研究方向为智能算法与风电系统,E-mail:845003063@qq.comx;丁云飞(1976—),女,教授,博士,主要研究方向为模式识别、智能控制、故障诊断、数据挖掘,E-mail:dingyf@sdju.edu.cn.
