基于VMD-MWOA-ELM的日前光伏功率预测  

Prediction of day-ahead photovoltaic power generation based on VMD-MWOA-ELM

作　　者：刘丽桑[1,2] 郭凯琪 陈健 郭琳[3] LIU Lisang;GUO Kaiqi;CHEN Jian;GUO Lin(School of Electronic,Electrical Engineering and Physics,Fujian University of Technology,Fuzhou 350118,China;Automation Engineering Research Center of Colleges and Universities in Fujian,Fuzhou 350118,China;Yongchun Power Supply Company,State Grid Power Supply CO.,Ltd.,Quanzhou 362600,China)

机构地区：[1]福建工程学院电子电气与物理学院,福建福州350118 [2]工业自动化福建省高校工程研究中心,福建福州350118 [3]国网福建省供电有限公司永春县供电公司,福建泉州362600

出　　处：《福建工程学院学报》2023年第3期269-276,共8页Journal of Fujian University of Technology

基　　金：福建省科技厅自然科学基金(2022J01952);福建省科技厅高校产学研合作项目(2022H6005)。

摘　　要：为了提高光伏发电功率的预测精度,提出一种结合变分模态分解、多策略改进的鲸鱼优化算法和极限学习机的光伏日前预测方法。利用变分模态分解影响光伏功率的关键气象因素,获得不同特征规律的本征模态分量,降解了数据的随机波动性,减少了噪声的影响。引入鲸鱼优化算法,利用多策略改进的鲸鱼优化算法(MWOA)对ELM模型的权重和偏置系数进行优化,获得最终的光伏功率预测结果。仿真结果验证了所提方法的有效性与优越性。In order to improve the prediction accuracy of photovoltaic power generation,a photovoltaic day-ahead prediction method combining variational mode decomposition,multi-strategy improvement whale optimization algorithm and extreme learning machine(ELM)was proposed.By utilizing variational mode decomposition to decompose key meteorological factors that affect photovoltaic power,the intrinsic mode components with different characteristic patterns were obtained,which degraded the random volatility of the data and reduced the impact of noise.The final photovoltaic power prediction results were obtained by introducing the multi-strategy improvement whale optimization algorithm(MWOA)to optimize the weights and bias coefficients of the ELM mode.The simulation results validated the effectiveness and superiority of the proposed method.

关 键 词：相关性分析 变分模态分解 多策略改进的鲸鱼优化算法 极限学习机 光伏发电功率预测 

分 类 号：TM615[电气工程—电力系统及自动化]