[1]袁俊杰,罗汝斌,廖 俊,等.3 种神经网络模型在平流层风场预测中的应用[J].控制与信息技术(原大功率变流技术),2019,(05):12-16.[doi:10.13889/j.issn.2096-5427.2019.05.003]
 YUAN Junjie,LUO Rubin,LIAO Jun,et al.Application of Three Neural Network Models in the Prediction ofStratospheric Wind Field[J].High Power Converter Technology,2019,(05):12-16.[doi:10.13889/j.issn.2096-5427.2019.05.003]
点击复制

3 种神经网络模型在平流层风场预测中的应用()
分享到:

《控制与信息技术》(原《大功率变流技术》)[ISSN:2095-3631/CN:43-1486/U]

卷:
期数:
2019年05期
页码:
12-16
栏目:
控制理论与应用
出版日期:
2019-10-05

文章信息/Info

Title:
Application of Three Neural Network Models in the Prediction ofStratospheric Wind Field
文章编号:
2096-5427(2019)05-0012-05
作者:
袁俊杰1罗汝斌2廖 俊1杨泽川1王 宁1李 珺1
(1.中南大学航空航天学院,湖南 长沙 410083;2.北京宇航系统工程研究所,北京 100076)
Author(s):
YUAN Junjie1 LUO Rubin2 LIAO Jun1 YANG Zechuan1 WANG Ning1 LI Jun1
( 1.School of Aeronautics and Astronautics, Central South University, Changsha, Hunan 410083, China;2.Beijing Institute of Astronautics System Engineering, Beijing 100076,China )
关键词:
平流层风场风速预测BP神经网络遗传算法粒子群算法浮空飞行器
Keywords:
stratospheric wind field wind speed prediction BP neural network genetic algorithm particle swarm optimization aerostat
分类号:
TM614
DOI:
10.13889/j.issn.2096-5427.2019.05.003
文献标志码:
A
摘要:
为了控制浮空飞行器的飞行轨迹,必须对平流层风场进行建模预测。传统理论模型只能用于预测未来几个小时的风速,而用BP神经网络模型则可预测未来几天的风速。为此,文章分别采用BP神经网络、遗传算法优化的BP神经网络和粒子群算法优化的BP神经网络作为风场预测模型,以历史时刻的风场数据作为神经网络的输入变量,对未来一段时间平流层底风场风速数据进行预测,并比较了这3种神经网络的预测精度。试验结果表明,神经网络模型可应用于平流层风场的预测,且经遗传算法和粒子群算法优化后的BP神经网络大大提高了BP神经网络的预测精度,在一定时间内有较好的预测效果;而粒子群算法优化的BP神经网络模型的预测精度略好于遗传算法优化的BP神经网络模型的预测精度。
Abstract:
Wind field forecast is of great significance for aerostat trajectory prediction. Traditional theoretical models can only predict wind speed in the next few hours, while BP neural network models can predict wind speeds in next few days. Therefore, in this paper, BP neural network, genetic algorithm-BP neural network, particle swam optimization-BP neural network were introduced and used as the wind field prediction model. The input variables of neural network predictive model are historical wind field data, and the model outputs the future wind speed data of stratospheric bottom. Meanwhile, the prediction accuracy of the three neural networks is also compared. The simulation results show that the neural network model can be applied to the prediction of stratospheric wind field and the BP neural networks optimized by genetic algorithm and particle swarm optimization can greatly improve the prediction accuracy of BP neural network, and the forecast wind speed is following the true value in a certain period of time. The prediction accuracy of the BP neural network optimized by particle swarm optimization is slightly better than that of the BP neural network optimized by genetic algorithm.

参考文献/References:

[1] 肖存英, 胡雄, 龚建村,等. 中国上空平流层准零风层的特征分析[J]. 空间科学学报, 2008, 28(3): 230-235.

2] 陶梦初, 何金海, 刘毅. 平流层准零风层统计特征及准两年周期振荡对其影响分析[J]. 气候与环境研究, 2012, 17(1): 92-102.
[3] TORRES J L, GARC?A A, BLAS M D, et al. Forecast of hourly average wind speed with ARMA models in Navarre (Spain)[J]. Solar Energy, 2005, 79(1): 65-77.
[4] 韩路跃, 杜行检. 基于MATLAB的时间序列建模与预测[J]. 计算机仿真, 2005, 22(4):105-107.
[5] 李明干, 孙健利, 刘沛. 基于卡尔曼滤波的电力系统短期负荷预测[J]. 电力系统保护与控制, 2004, 32(4): 9-12.[6] SFETSOS A. A comparison of various forecasting techniques applied to mean hourly wind speed time series[J]. Renewable Energy, 2000, 21(1): 23-35.
[7] 刘辉, 田红旗, 李燕飞. 基于小波分析法与滚动式时间序列法的风电场风速短期预测优化算法[J]. 中南大学学报(自然科学版), 2010, 41(1): 370-375.
[8] KANI S A P , Ardehali M M . Very short-term wind speed prediction: A new artificial neural network–Markov chain model[J]. Energy Conversion & Management, 2011, 52(1): 738-745.
[9] 王德明, 王莉, 张广明. 基于遗传BP神经网络的短期风速预测模型[J]. 浙江大学学报(工学版), 2012(5): 837-841.
[10] 方江晓. 短期风速和风电功率预测模型的研究[D].北京:北京交通大学, 2011.

备注/Memo

备注/Memo:
收稿日期:2019-05-30
作者简介:袁俊杰(1995— ),男,硕士研究生,研究方向为飞行器总体设计。
更新日期/Last Update: 2019-10-23