基于狐獴优化算法的短期海上风电预测-《计算机技术与发展》

文章信息/Info

Title:: Short-term Offshore Wind Power Prediction Based on Meerkat Optimization Algorithm

Author(s):: BAI Jun-ming; ZHOU Xin-yu; JIANG Zhi-hang; LIANG Hong-tao; School of Information Science and Technology,Qingdao University of Science and Technology,Qingdao 266061,China

Keywords:: offshore wind power; meerkat optimization algorithm; local mean decomposition; power prediction; deep learning

摘要:: 海洋风力发电代表了前沿的绿色能源趋势。准确预测海上风力发电的输出对于实现大规模海上风力发电站的顺利接入电网起着关键作用,这有利于保障电网系统的平稳与安全运作,并能减少能源消耗的损耗。本项研究构建了一种混合深度学习预测模型,基于狐獴优化算法和局部均值分解,用于预测短期海上风力发电功率。本研究采用局部均值算法分解原始海上风电数据,采用卷积神经网络和门控神经单元混合预测海上风电发力功率。为了提高预测的准确度并降低不确定性因素,本研究引入狐獴优化算法,优化局部均值算法和混合模型超参数。使用中国海上风电场实际存在的风力涡轮机发电数据集进行了深入研究,对比结果显示,该模型优于其他模型,证明了该模型的可行性和有效性。

Abstract:: Offshore wind power represents a cutting-edge green energy trend. Accurate prediction of offshore wind power output plays a key role in realizing the smooth connection of large-scale offshore wind farms to the power grid,which helps to guarantee the smooth and safe operation of the power grid system and reduces the loss of energy consumption. In this study,a hybrid deep learning prediction model based on the meerkat optimization algorithm and local mean decomposition is constructed for predicting short-term offshore wind power. In this study,the local mean algorithm is used to decompose the raw offshore wind power data,and a hybrid of convolutional neural network and gated neural unit is used to predict the offshore wind power generation power. In order to improve the accuracy of the prediction and reduce the uncertainty factors,the meerkat optimization algorithm is introduced to optimize the local mean algorithm and the hybrid model hyperparameters. An in-depth study was conducted using wind turbine power generation datasets that actually exist in Chinese offshore wind farms,and the comparison results show that the proposed model outperforms other models,proving its feasibility and effectiveness.