基于季节波动序列的PSO-FNSGM(1,1,k)模型及其应用
PSO-FNSGM(1,1,k)Model Based on Seasonal Fluctuation Sequence and Its Application
张怡萱 1胡坰煌 1李钒 1熊昕 1胡曦1
作者信息
- 1. 江汉大学 人工智能学院,湖北 武汉 430056
- 折叠
摘要
针对具有年度波动特征和季节波动特征的复杂序列,使用季节因子、粒子群算法以及傅里叶级数优化构建灰色预测模型,以实现对季节波动序列的精确预测.通过年度作用系数的改变列出了 3种季节因子,并对 3种季节因子进行了比较.为减少时间变化对序列的干扰,在模型中加入线性修正项以及使用粒子群算法寻找最优参数来提高模型精度.考虑到序列受季节变化影响较大,采用傅里叶级数对模型预测残差序列进行修正.将模型用于中国水力净发电量的模拟与预测,误差仅为1.22%,表明该模型针对波动序列具有较高的预测精度.
Abstract
For the complex sequence with the characteristics of annual fluctuation and sea-sonal fluctuation,a grey prediction model based on seasonal factors,particle swarm optimi-zation(PSO),and Fourier optimization was used in this paper to achieve accurate prediction of seasonal fluctuation series.Firstly,this prediction model proposed three seasonal factors by changing the annual effect coefficient and then compared these factors.Secondly,to re-duce the interference of time variation on the sequence,this paper added linear correction terms to the prediction model and used the PSO algorithm to find the optimal parameters to improve the prediction accuracy of the model.Finally,by considering the influence of season-al variation on the sequence,the Fourier series was used to fit the residual series of the mod-el.In this paper,the model was applied to the simulation and prediction of the net hydroelec-tric power generation in China,and the final error was 1.22% .The research shows that the model has higher prediction accuracy for fluctuation sequences.
关键词
季节因子/周期序列/灰色模型/粒子群优化算法/傅里叶级数Key words
seasonal factor/periodic sequence/grey model/particle swarm optimization/Fourier series引用本文复制引用
基金项目
国家自然科学基金资助项目(61901298)
江汉大学省部共建精细爆破国家重点实验室自主研究课题(PBSKL2022303)
江汉大学省级大学生创新训练项目(2022zd106)
江汉大学校级科研项目(2021yb057)
湖北省教育厅科学研究计划指导性项目(B2022280)
出版年
2024
国家科技期刊平台
NSTL
万方数据