基于均生函数的长短期记忆神经网络径流建模研究与应用
Long Short-Term Memory Neural Networks Based on Mean Generating Function for Runoff Prediction Modeling Research and Application
DOI: 10.12677/MOS.2020.92018, PDF,  被引量    国家自然科学基金支持
作者: 吴建生*, 谢永盛：广西科技师范学院，数学与计算机科学学院, 广西 柳州；金 龙：广西气象局，广西 南宁
关键词: 均生函数；信息熵；循环神经网络；长短期记忆神经网络；径流；Mean Generating Function； Information Entropy； Recurrent Neural Network； Long Short-Term Memory Neural Networks； Runoff
摘要: 准确的径流建模预测是水文领域中一项具有挑战性的任务，也是减灾防灾的重要研究课题，精确可操作的径流预报模型对于水资源规划及综合开发利用，水利枢纽运行管理等重大决策问题提供基本决策依据，对国民经济健康发展具有十分重要的意义。针对径流预测建模中难以选择建模因子和难以确定非线性径流预测模型的问题，本文利用均生函数生成延拓因子矩阵，并提取延拓矩阵中信息熵高矩阵生成建模因子，进一步利用长短期记忆神经网络建立柳江径流预测模型，该模型通过分析原始降水序列特性，提取径流系统不同振荡周期特征形成建模因子，并结合长短期记忆神经网络对在非线性时间序列记忆特征和非线性时序动力系统控制的优势，以此建立柳江径流模型。该方法充分利用均生函数提取系统序列前后依赖和周期关系特性，并结合长短期记忆神经网络对非线性时间序列记忆特性，此方法能有效地提高预报长度，并能获得较高的建模及预报精度。柳江径流实例进行验证，结果表明了方法预测的精度高，稳定性好，为径流预测分析提供了一种可靠、有效的方法。
Abstract: Aiming at the problem that it is difficult to determine the modeling factors and the non-linear runoff prediction modeling in runoff prediction modeling, this paper uses a homogeneous function to generate a continuation factor matrix, and extracts a high information entropy matrix in the continuation matrix to generate a modeling factor. A long-term and short-term memory neural network is used to establish a Liujiang runoff prediction model. This model analyzes the charac-teristics of the original precipitation sequence and extracts different oscillation cycle characteristics of the runoff system to form modeling factors. The advantage of time series dynamic system control is to establish a Liujiang runoff model. This method makes full use of the homogeneous function to extract the sequence dependence and periodic relationship characteristics of the system sequence, and combines the long-term and short-term memory neural network memory characteristics of the non-linear time series to establish a Liujiang runoff simulation prediction model. The experimental results show that the method has good prediction accuracy and good stability, and provides a reliable and effective method for runoff prediction analysis.
文章引用：吴建生, 谢永盛, 金龙. 基于均生函数的长短期记忆神经网络径流建模研究与应用[J]. 建模与仿真, 2020, 9(2): 163-177. https://doi.org/10.12677/MOS.2020.92018

参考文献

[1] Wang, W., Chau, K., Xu, D., et al. (2015) Improving Forecasting Accuracy of Annual Runoff Time Series Using ARIMA Based on EEMD Decomposition. Water Resources Management, 29, 2655-2675. https://xs.scihub.ltd/ [Google Scholar] [CrossRef] 
[2] Farajzadeh, J., Fard, A.F. and Lotfi, S. (2014) Modeling of Monthly Rainfall and Runoff of Urmia Lake Basin Using “Feed-forward Neural Network” and “Time Se-ries Analysis” Model. Water Resources and Industry, 7, 38-48. [Google Scholar] [CrossRef] 
[3] 黄强, 赵雪花, 等. 河川径流时间序列分析预测理论与方法[M]. 郑州: 黄河水利出版社, 2008: 2-12.
[4] 金龙著. 神经网络气象预报建模理论方法与应用[M]. 北京: 气象出版社, 2004: 14-15.
[5] Wu, J.S. and Xie, Y.S. (2019) Hybrid Support Vector Regression with Parallel Co-Evolution Algorithm Based on GA and PSO for Forecasting Monthly Rainfall. Journal of Software Engineering and Applications, 12, 524-539. [Google Scholar] [CrossRef] 
[6] 李彦彬, 黄强, 徐建新, 等. 河川径流混沌特征及预测理论与实践[M]. 北京: 中国水利水电出版社, 2011: 10-12.
[7] 高新波, 张军平. 机器学习及其应用2015[M]. 北京: 清华大学出版社, 2015: 8-9.
[8] Wu, J.S., Jin, L. and Liu, M. (2010) A Hybrid Support Vector Regression Approach for Rainfall Forecasting Using Particle Swarm Optimization and Projection Pursuit Technology. International Journal of Computational Intelligence and Applications, 9, 87-104. [Google Scholar] [CrossRef] 
[9] Makungo, R., Odiyo, J.O. and Mwak, B. (2010) Rain-fall-Runoff Modelling Approach for Ungauged Catchments: A Case Study of Nzhelele River Sub-Quaternary Catchment. Physics and Chemistry of the Earth, 35, 45-62. [Google Scholar] [CrossRef] 
[10] Wu. J.S. (2011) An Effective Hybrid Semi-Parametric Regression Strategy for Rainfall Forecasting Combining Linear and Nonlinear Regression. International Journal of Applied Evo-lutionary Computation, 2, 50-65. [Google Scholar] [CrossRef] 
[11] Refona, J., Lakshmi, M., Dhamodaran, S., et al. (2019) Machine Learning Techniques for Rainfall Prediction Using Neural Network. Journal of Computational and Theoretical Nano-science, 16, 3319-3323. [Google Scholar] [CrossRef] 
[12] Mosavi, A., Ozturk, P. and Chau, K. (2018) Flood Prediction Using Machine Learning Models: Literature Review. Water, 10, 1536. [Google Scholar] [CrossRef] 
[13] Mishra, N., Soni, H.K., Sharma, S., et al. (2018) Development and Analysis of Artificial Neural Network Models for Rainfall Prediction by Using Time-Series Data. International Journal of Intelligent Systems and Applications, 11, 16. [Google Scholar] [CrossRef] 
[14] Liu, Q., Zhou, Y., Liu, X. and Nigel, L. (2019) A Survey on Rainfall Forecasting Using Artificial Neural Network. International Journal of Embedded Systems, 11, 240-249. [Google Scholar] [CrossRef] 
[15] Yaseen, Z.M., Ghareb, M.I., Ebtehaj, I., et al. (2018) Rainfall Pattern Forecasting Using Novel Hybrid Intelligent Model Based ANFIS-FFA. Water Resources Management, 32, 105-122. [Google Scholar] [CrossRef] 
[16] Graves, A., Wayne, G., Reynolds, M., et al. (2016) Hybrid Computing Using a Neural Network with Dynamic External Memory. Nature, 538, 471-476. [Google Scholar] [CrossRef] [PubMed]
[17] Van, G.M. and Bohte, S. (2017) Artificial Neural Networks as Models of Neural Information Processing. Frontiers in Computational Neuroscience, 11, 114. [Google Scholar] [CrossRef] [PubMed]
[18] Niu, W., Feng, Z., Cheng, C., et al. (2018) Forecasting Daily Runoff by Extreme Learning Machine Based on Quantum-Behaved Particle Swarm Optimization. Journal of Hydrologic Engineering, 23, Article ID: 04018002. [Google Scholar] [CrossRef] 
[19] Feng, Z., Niu, W., Tang, Z., et al. (2020) Monthly Runoff Time Series Prediction by Variational Mode Decomposition and Support Vector Machine Based on Quan-tum-Behaved Particle Swarm Optimization. Journal of Hydrology, 583, Article ID: 124627. [Google Scholar] [CrossRef] 
[20] 刘炳春, 齐鑫, 王庆山. 北京城市代谢预测研究-基于长短期记忆神经网络模型[J]. 地理科学进展, 2019, 38(6): 851-860.
[21] Ayvaz, E., Kaplan, K. and Kuncan, M. (2020) An Integrated LSTM Neural Networks Approach to Sustainable Balanced Scorecard-Based Early Warning System. IEEE Access, 8, 37958-37966. [Google Scholar] [CrossRef] 
[22] 毛雪娇, 谈健, 姚颖蓓, 等. 基于长短期记忆神经网络的饱和负荷预测方法及应用[J]. 水电能源科学, 2019(6): 192-195.
[23] Sagheer, A. and Kotb, M. (2019) Time Series Forecasting of Petroleum Production Using Deep LSTM Recurrent Networks. Neurocomputing, 323, 203-213. [Google Scholar] [CrossRef] 
[24] Karevan, Z. and Suykens, J.A.K. (2020) Transudative LSTM for Time-Series Prediction: An Application to Weather Forecasting. Neural Networks, 125, 1-9. [Google Scholar] [CrossRef] [PubMed]
[25] Mandic, D.P. and Chambers, J. (2001) Recurrent Neural Net-works for Prediction: Learning Algorithms, Architectures and Stability. John Wiley & Sons, Inc., New York.
[26] Rumelhart, D.E., Hinton, G.E. and Williams, R.J. (1986) Learning Representations by Back-Propagating Errors. Nature, 323, 533-536. [Google Scholar] [CrossRef] 
[27] Werbos, P.J. (1988) General-ization of Back-Propagation with Application to a Recurrent Gas Market Model. Neural Networks, 1, 339-356. [Google Scholar] [CrossRef] 
[28] Elman, J. (1990) Finding Structure in Time. Cognitive Science, Cognitive Science, 14, 179-211. [Google Scholar] [CrossRef] 
[29] 邱锡鹏. 神经网络与深度学习[M]//第六章 循环神经网络.
[30] https://baike.baidu.com/item/%E5%BE%AA%E7%8E%AF%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C/23199490?fr=aladdin#reference-[4]-23754076-wrap
[31] Bengio, Y., Simard, P. and Frasconi, P. (1994) Learning Long-Term Dependencies with Gradient Descent Is Difficult. IEEE Transactions on Neural Networks, 5, 157-166. [Google Scholar] [CrossRef] [PubMed]
[32] Hochreiter, S. and Schmidhuber, J. (1997) Long Short-Term Memory. Neural Computation, 9, 1735-1780. [Google Scholar] [CrossRef] [PubMed]
[33] Zia, T. and Zahid, U. (2019) Long Short-Term Memory Recurrent Neural Network Architectures for Urdu Acoustic Modeling. International Journal of Speech Technology, 22, 21-30. [Google Scholar] [CrossRef] 
[34] Kong, W., Dong, Z.Y., Jia, Y., et al. (2017) Short-Term Residential Load Forecasting Based on LSTM Recurrent Neural Network. IEEE Transactions on Smart Grid, 10, 841-851. [Google Scholar] [CrossRef] 
[35] Kim, H.Y. and Won, C.H. (2018) Forecasting The Volatility of Stock Price Index: A Hybrid Model Integrating LSTM with Multiple GARCH-Type Models. Expert Sys-tems with Applications, 103, 25-37. [Google Scholar] [CrossRef] 
[36] Oh, S.L., Ng, E.Y.K., San, T.R., et al. (2018) Automated Diag-nosis of Arrhythmia Using Combination of CNN and LSTM Techniques with Variable Length Heart Beats. Computers in Biology and Medicine, 102, 278-287. [Google Scholar] [CrossRef] [PubMed]
[37] 魏凤英, 曹鸿兴. 长期预测的数学模型及其应用[M]. 北京: 气象出版社, 1990. 9-12.
[38] 魏生生, 曹鸿兴. 均生函数模型预报检验[J]. 气象, 1998, 24(5): 42-46.
[39] 李文辉, 陈妍, 梁嘉雯, 等. 时间序列均生函数模型在降水量预测中的应用[J]. 广东气象, 2019, 41(4): 11-14.
[40] 刘婷婷, 李月瑶. 时间序列均生函数模型在降水量预测中的应用[J]. 吉林水利, 2013, 371(4): 32-34.
[41] http://blog.csdn.net/pipisorry/article/details/78361778
[42] http://colah.github.io/posts/2015-08-Understanding-LSTMs/
[43] Abdel-Nasser, M. and Mahmoud, K. (2019) Accurate Photovoltaic Power Forecasting Models Using Deep LSTM-RNN. Neural Computing & Applications, 31, 2727-2740. [Google Scholar] [CrossRef] 
[44] Zhao, J.C., Deng, F., Cai, Y.Y., et al. (2019) Long Short-Term Memory-Fully Connected (LSTM-FC) Neural Network for PM2.5 Concentration Prediction. Chemosphere, 220, 486-492. [Google Scholar] [CrossRef] [PubMed]
[45] Gers, F.A. and Schmidhuber, E. (2001) LSTM Recurrent Networks Learn Simple Context-free and Context-sensitive Languages. IEEE Transactions on Neural Net-works, 12, 1333-1340. [Google Scholar] [CrossRef] [PubMed]

为你推荐