宽度-深度融合时频分析的径流智能预测方法

doi:10.16182/j.issn1004731x.joss.22-1137

系统仿真学报 ›› 2024, Vol. 36 ›› Issue (2): 363-372.doi: 10.16182/j.issn1004731x.joss.22-1137

宽度-深度融合时频分析的径流智能预测方法

韩莹¹^,²(), 王乐豪¹, 王淑梅³(), 张翔³, 罗星星³

^1.南京信息工程大学自动化学院，江苏南京 210044
^2.南京信息工程大学江苏省大气环境与装备技术协同创新中心，江苏南京 210044
^3.信江饶河水文水资源检测中心，江西上饶 334000

收稿日期:2022-09-26 修回日期:2022-11-23 出版日期:2024-02-15 发布日期:2024-02-04
通讯作者: 王淑梅 E-mail:hanyingcs@163.com;Eddie3208@163.com
第一作者简介:韩莹(1978-)，女，副教授，硕士生导师，博士，研究方向为大数据处理方法及其应用。E-mail：hanyingcs@163.com
基金资助:
国家自然科学基金(62076136);教育部新农科研究与改革实践(20200251)

Runoff Intelligent Prediction Method Based on Broad-deep Fusion Time-frequency Analysis

Han Ying¹^,²(), Wang Lehao¹, Wang Shumei³(), Zhang Xiang³, Luo Xingxing³

^1.School of Automation, Nanjing University of Information Science & Technology, Nanjing 210044, China
^2.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China
^3.Xinjiang Raohe Hydrological and Water Resources Testing Center, Shangrao 334000, China

Received:2022-09-26 Revised:2022-11-23 Online:2024-02-15 Published:2024-02-04
Contact: Wang Shumei E-mail:hanyingcs@163.com;Eddie3208@163.com

摘要/Abstract

摘要：

为解决现有基于LSTM的径流预测模型易陷入局部最优的问题，提出了基于VMD-LSTM-BLS(variational mode decomposition-LSTM-broad learning system)的径流预测模型。将宽度学习系统与LSTM结合，针对径流序列多噪音特点，采用时频分析方法中的变分模态分解，将径流时间序列的一维时域信号变换到二维时频平面，减少噪声对预测结果的影响。仿真结果表明：与基线模型及现有基于LSTM的径流预测模型相比，该模型的预测精度有较为明显的提高。

关键词: 径流预测, 变分模态分解, 长短时记忆网络, 宽度学习系统, 时频分析, 智能预测

Abstract:

Broad learning system(BLS) is introduced to tackle the existed disadvantage that LSTM-based runoff prediction model is easy to fall into local optimization. To reduce the influence of noise on the prediction results, the variational mode decomposition (VMD) is adopted to transform the one-dimensional time-domain runoff signal to the two-dimensional time-frequency plane. The runoff prediction model based on VMD-LSTM-BLS is proposed. The simulation results demonstrate that the prediction accuracy of the new model is more significantly improved compared with the baseline model and the existing LSTM-based runoff prediction model.

Key words: runoff forecast, variational mode decomposition, long and short-term memory network, broad learning system, time-frequency analysis, intelligent prediction

中图分类号:

TP391.9

韩莹,王乐豪,王淑梅等 . 宽度-深度融合时频分析的径流智能预测方法[J]. 系统仿真学报, 2024, 36(2): 363-372.

Han Ying,Wang Lehao,Wang Shumei,et al . Runoff Intelligent Prediction Method Based on Broad-deep Fusion Time-frequency Analysis[J]. Journal of System Simulation, 2024, 36(2): 363-372.

图/表 10

图1

图2

图3

图4

表1

表2

表3

图5

图6

图7

参考文献 18

1	金菊良, 魏一鸣, 丁晶. 预测日径流过程的最近邻仿真模型[J]. 系统仿真学报, 2002, 14(11): 1494-1496.
	Jin Juliang, Wei Yiming, Ding Jing. Nearest Neighbor Bootstrap Model for Predicting Daily Flow Process[J]. Journal of System Simulation, 2002, 14(11): 1494-1496.
2	Huseyin Cagan Kilinc, Yurtsever Adem. Short-term Streamflow Forecasting Using Hybrid Deep Learning Model Based on Grey Wolf Algorithm for Hydrological Time Series[J]. Sustainability, 2022, 14(6): 3352.
3	李伟, 周玉臣, 林圣琳, 等. 仿真模型验证方法综述[J]. 系统仿真学报, 2019, 31(7): 1249-1256.
	Li Wei, Zhou Yuchen, Lin Shenglin, et al. Review of Simulation Model Validation Methods[J]. Journal of System Simulation, 2019, 31(7): 1249-1256.
4	Qin Jinbo, Liang Ji, Chen Tao, et al. Simulating and Predicting of Hydrological Time Series Based on TensorFlow Deep Learning[J]. Polish Journal of Environmental Studies, 2019, 28(2): 795-802.
5	Peng Tian, Zhou Jianzhong, Zhang Chu, et al. Streamflow Forecasting Using Empirical Wavelet Transform and Artificial Neural Networks[J]. Water, 2017, 9(6): 406.
6	曹爽, 秦天玲, 石晓晴, 等. 亚马逊流域降雨径流时空变化分析[J]. 水文, 2018, 38(1): 90-96.
	Cao Shuang, Qin Tianling, Shi Xiaoqing, et al. Time-spatial Distribution of Rainfall and Runoff in Amazon Basin[J]. Journal of China Hydrology, 2018, 38(1): 90-96.
7	黄克威, 王根绪, 宋春林, 等. 基于LSTM的青藏高原冻土区典型小流域径流模拟及预测[J]. 冰川冻土, 2021, 43(4): 1144-1156.
	Huang Kewei, Wang Genxu, Song Chunlin, et al. Runoff Simulation and Prediction of a Typical Small Watershed in Permafrost Region of the Qinghai-Tibet Plateau Based on LSTM[J]. Journal of Glaciology and Geocryology, 2021, 43(4): 1144-1156.
8	夏容, 江官星. 基于LSTM的软件时间序列延迟预测仿真[J]. 计算机仿真, 2021, 38(12): 435-439.
	Xia Rong, Jiang Guanxing. Software Time Series Delay Prediction Simulation Based on LSTM[J]. Computer Simulation, 2021, 38(12): 435-439.
9	Chen Xi, Huang Jiaxu, Han Zhen, et al. The Importance of Short Lag-time in the Runoff Forecasting Model Based on Long Short-term Memory[J]. Journal of Hydrology, 2020, 589: 125359.
10	王军, 高梓勋, 朱永明. 基于CNN-LSTM模型的黄河水质预测研究[J]. 人民黄河, 2021, 43(5): 96-99, 109.
	Wang Jun, Gao Zixun, Zhu Yongming. Research on Yellow River Water Quality Prediction Based on CNN-LSTM Model[J]. Yellow River, 2021, 43(5): 96-99, 109.
11	Tan Qiaofeng, Lei Xiaohui, Wang Xu, et al. An Adaptive Middle and Long-term Runoff Forecast Model Using EEMD-ANN Hybrid Approach[J]. Journal of Hydrology, 2018, 567: 767-780.
12	Dragomiretskiy K, Zosso D. Variational Mode Decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544.
13	包苑村, 解建仓, 罗军刚. 基于VMD-CNN-LSTM模型的渭河流域月径流预测[J]. 西安理工大学学报, 2021, 37(1): 1-8.
	Bao Yuancun, Xie Jiancang, Luo Jungang. Monthly Runoff Prediction of Weihe River Basin Based on VMD-CNN-LSTM Model[J]. Journal of Xi'an University of Technology, 2021, 37(1): 1-8.
14	任长娥, 袁超, 孙彦丽, 等. 宽度学习系统研究进展[J]. 计算机应用研究, 2021, 38(8): 2258-2267.
	Ren Chang'e, Yuan Chao, Sun Yanli, et al. Research of Broad Learning System[J]. Application Research of Computers, 2021, 38(8): 2258-2267.
15	Zhou Jianzhong, Peng Tian, Zhang Chu, et al. Data Pre-analysis and Ensemble of Various Artificial Neural Networks for Monthly Streamflow Forecasting[J]. Water, 2018, 10(5): 628.
16	李彤, 胡国华, 顾庆福, 等. 近55年来降水及人类活动对资水流域径流的影响[J]. 水文, 2018, 38(6): 54-58, 88.
	Li Tong, Hu Guohua, Gu Qingfu, et al. Impact of Precipitation and Human Activities on Runoff in Zishui River in Recent 55 Years[J]. Journal of China Hydrology, 2018, 38(6): 54-58, 88.
17	胡庆芳, 曹士圯, 杨辉斌, 等. 汉江流域安康站日径流预测的LSTM模型初步研究[J]. 地理科学进展, 2020, 39(4): 636-642.
18	Seo Youngmin, Kim Sungwon, Singh V P. Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-runoff[J]. Atmosphere, 2018, 9(7): 251.

K	IMF0	IMF1	IMF2	IMF3	IMF4	IMF5	IMF6
3	0.359	0.185	0.025
4	0.467	0.347	0.183	0.025
5	0.467	0.348	0.182	0.026	0.006
6	0.467	0.348	0.271	0.182	0.026	0.006
7	0.486	0.467	0.348	0.271	0.182	0.026	0.006

层	参数	值
Input	时间步长	2
ARIMA	{ p, d, q}	{10, 1, 0}
SVM	C	100 000
	缓冲大小	32
	γ	0.000 1
VMD	α	10
VMD	K	5
CNN	卷积核个数	512
	卷积核大小	5
	填充	same
	激活函数	ReLU
Maxpooling	池化核大小	2
Maxpooling	填充	valid
LSTM	层数	3
	神经元个数	{256, 128, 128}
	优化器	Adam
	批大小	20
	训练次数	80
	损失函数	MAE
Dropout	P	0.2
Dense	层数	2
	神经元个数	{10, 1}
	激活函数	ReLU
BLS	s	0.8
	C	0.000 1
	N₁	10
	N₂	30
	N₃	280

模型	E_RMS	E_MA	E_MAP	S_N
ARIMA	0.921	0.532	1.659	0.133
SVR	0.914	0.513	1.619	0.145
LSTM	0.858	0.494	1.422	0.247
CNN-LSTM	0.823	0.482	1.302	0.308
LSTM-BLS	0.795	0.453	1.025	0.353
VMD-LSTM	0.381	0.286	0.688	0.515
VMD-CNN-LSTM	0.282	0.266	0.174	0.877
本文	0.004	0.003	0.025	0.985

[1]	陈静, 张昭冲, 王琳凯, 安脉, 王伟. 基于卷积长短时记忆网络的短时公交客流量预测[J]. 系统仿真学报, 2024, 36(2): 476-486.
[2]	张凯, 卢海鹏, 韩莹, 张龄允, 丁昱杰. 融合变分模态分解的时空卷积短时车速预测[J]. 系统仿真学报, 2023, 35(8): 1651-1660.
[3]	杨静, 陆铭华, 胡星辰, 吴金平. 基于不确定时间序列的来袭目标解算方法[J]. 系统仿真学报, 2023, 35(6): 1245-1259.
[4]	林硕, 安磊, 高治军, 单丹, 尚文利. 结合栈式自编码及长短时记忆的入侵检测研究[J]. 系统仿真学报, 2021, 33(6): 1288-1296.
[5]	同晓荣. 基于短时傅里叶时频分析的声信号盲分选仿真系统研究[J]. 系统仿真学报, 2019, 31(2): 353-359.
[6]	许杏花, 潘庭龙, 吴定会. 基于变分模态分解的风电场风速降噪研究[J]. 系统仿真学报, 2019, 31(1): 87-93.
[7]	贾亚飞, 朱永利, 王刘旺. 基于VMD和Wigner-Ville分布的局放信号时频分析[J]. 系统仿真学报, 2018, 30(2): 569-578.
[8]	王田田, 王艳, 纪志成. 基于改进极限学习机的滚动轴承故障诊断[J]. 系统仿真学报, 2018, 30(11): 4413-4420.
[9]	迟玉伦, 李郝林, 岳泰. 修整金刚笔磨损智能预测方法研究[J]. 系统仿真学报, 2017, 29(6): 1210-1217.

宽度-深度融合时频分析的径流智能预测方法

Runoff Intelligent Prediction Method Based on Broad-deep Fusion Time-frequency Analysis

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 18

相关文章 9

编辑推荐

Metrics

本文评价