摘要
为提高新能源电力系统(NEPS)风/光功率预测的精确度以降低新能源并网对电网稳定性的影响,考虑风/光功率影响因素的相关性、NEPS分布式电源的特殊性以及模型预测误差的自适应优化能力,提出一种考虑相关性的新能源电力系统风光功率短期联合预测模型。通过自适应噪声的完全集合经验模态分解(CEEMDAN)对原始数据进行处理,利用基于混沌鲸鱼算法(CWOA)优化的双向长短期记忆神经网络(BILSTM)模型对风/光功率初步预测获取预测误差,将分解后的风/光功率预测误差与原始输入特征融合,对光/风功率交叉联合预测。通过华东地区某新能源场站实际数据进行试验验证,结果表明,与传统预测模型相比,该文所提联合预测模型对NEPS风/光功率预测精度均有所提升。
To reduce the impact of new energy grid integration on grid stability,by improving the accuracy of power prediction of wind and photovoltaic(PV)power generation in new energy power system(NEPS),this paper proposes a joint short-term power prediction model for wind and PV in NEPS considering correlation between the influencing factors of wind and PV power generation,the specificity of NEPS distributed power,and the adaptive optimization ability of the model according to its own prediction error.The raw data are first processed by complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN),and then the wind and PV power are initially predicted using the bidirectional long short term memory(BILSTM)model optimized based on the chaotic whale optimization algorithm(CWOA)to obtain the prediction errors.Then the prediction errors of wind and PV power after CEEMDAN decomposition are merged together into the input features for the joint prediction of PV and wind power crossover.The experimental results of a NEPS station in eastern China show that the accuracy of both wind and PV power prediction is improved by the proposed model in this paper compared with the traditional prediction model.
作者
沈赋
刘思蕊
蔡子龙
王哲
杨光兵
翟苏巍
Shen Fu;Liu Sirui;Cai Zilong;Wang Zhe;Yang Guangbing;Zhai Suwei(Faculty of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650500,China;Electric Power Research Institute,Yunnan Power Grid Co.,LTD.,Kunming 650521,China)
出处
《太阳能学报》
北大核心
2025年第5期203-212,共10页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(52107097)
云南省兴滇英才支持计划(KKRD202204021)
云南省应用基础研究计划(202101BE070001-061,202201AU070111)。
关键词
风力发电
光伏发电
功率预测
双向长短期记忆神经网络
新能源电力系统
自适应噪声的完全集合经验模态分解
混沌鲸鱼算法
wind power generation
PV power generation
power prediction
bidirectional long short term memory(BILSTM)
new energy power system(NEPS)
complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)
chaotic whale optimization algorithm(CWOA)
