基于改进DBO-BPNN的短期光伏功率预测研究  

Research on Short Term Photovoltaic Power Prediction Based on Improved DBO-BPNN

作　　者：陈嘉辉 唐林杰 CHEN Jiahui;TANG Linjie(Foshan University,Foshan 528225,China)

机构地区：[1]佛山大学,广东佛山528225

出　　处：《电工技术》2025年第1期46-48,51,共4页Electric Engineering

摘　　要：为了提高光伏功率预测的准确性,提出了一种改进的蜣螂算法优化BPNN预测模型。首先引入变螺旋搜索、莱维飞行、自适应t分布和动态选择策略来对蜣螂算法进行改进;其次采用改进的蜣螂算法来优化BPNN的权值和阈值;最后通过实验验证,与传统BPNN模型、DBO-BPNN模型的预测结果相比,TDBO-BPNN模型具有更高的预测精度,在光伏功率预测中具有良好的应用前景。In order to improve the accuracy of PV power prediction,an improved DBO algorithm is proposed to optimize BPNN prediction model.Firstly,variable spiral search,Levy flight,adaptive T-distribution and dynamic selection strategies are introduced to improve the DBO algorithm.Secondly,improved dung DBO algorithms are used to optimize the weights and thresholds of BPNN.Finally,the experimental results show that compared with the traditional BPNN model and the BDB-BPNN model,the TBDB-BPNN model has higher prediction accuracy,and has a good application prospect in the photovoltaic power prediction.

关 键 词：光伏功率预测 蜣螂算法 BPNN 

分 类 号：TM615[电气工程—电力系统及自动化]