基于CEEMDAN二次分解的风速预测  

Wind Speed Prediction Based on the QuadraticSecond Decomposition of CEEMDAN

作　　者：李颖智 王维庆[1] 王海云[1] LI Ying-zhi;WANG Wei-qing;WANG Hai-yun(Engineering Research Center of Education Ministry for Renewable Energy Power Generation and Grid Control,Xinjiang University,Urumqi Xinjiang 830047,China)

机构地区：[1]新疆大学可再生能源发电与并网控制教育部工程技术研究中心,新疆乌鲁木齐830047

出　　处：《计算机仿真》2023年第2期89-93,384,共6页Computer Simulation

基　　金：国家自然科学基金资助项目(52067020);自治区重点研发计划(2020B02001)。

摘　　要：随着分散式风机装机量日益增加,对分散式风机风速预测便于合理规划其并网及就地消纳容量。于是提出基于CEEMDAN二次分解的蝴蝶优化算法改进最小二乘支持向量机的风速预测模型。利用自适应白噪声的完备总体经验模态分解(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)对风速历史数据进行处理,获取模态分量(intrinsic mode function,IMF),计算各IMF的排列熵;采用CEEMDAN对随机程度高的IMF进行第二次分解,改善数据的随机程度;最后,采用蝴蝶优化算法改进最小二乘支持向量机的模型对重构的IMF进行预测,将各IMF的预测风速相加,求出预测值。通过分析风速预测数据与风速实测数据的误差,验证基于CEEMDAN二次分解的蝴蝶优化算法改进最小二乘支持向量机的风速预测模型的有效性。With the increasing installed capacity of decentralized fans,the wind speed forecast of decentralized fans facilitates reasonable planning of their grid connection and local consumption capacity.Therefore,a butterfly optimization algorithm based on CEEMDAN quadratic decomposition is proposed to improve the wind speed prediction model of least squares support vector machine.First,use the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)was used to process the wind speed history data to obtain the modal component(intrinsic mode function,IMF),and calculate the arrangement of each IMF Entropy.Secondly,CEEMDAN wais used to decompose the highly random IMF for the second time to improve the randomness of the data.Finally,the butterfly optimization algorithm wais used to improve the least squares support vector machine model to predict the reconstructed IMF,and each IMF Add added the predicted wind speeds to obtain the predicted value.By analyzing the errors between the wind speed prediction data and the measured wind speed data,the effectiveness of the butterfly optimization algorithm based on CEEMDAN quadratic decomposition to improve the wind speed prediction model of the least square support vector machine is was verified.

关 键 词：风速预测 完备总体经验模态分解 二次分解 蝴蝶优化算法 排列熵 

分 类 号：TM614[电气工程—电力系统及自动化]