Rogue Wave Prediction Based on Four Combined Long Short-Term Memory Neural Network Models

doi:10.16183/j.cnki.jsjtu.2021.088

Abstract

Abstract:

In order to improve the prediction accuracy of rogue waves of the long short-term memory (LSTM) neural network, prediction methods of LSTM with convolution neural networks (CNN), empirical mode decomposition (EMD), auto-aggressive integrated moving averagel (ARIMA) model, and Kalman filtering (KF) were studied. Based on the experimental data of the rogue waves of two single-peak and one three combined peaks, prediction models were established and predicted by data normalization, model parameter optimization and error evaluation. The results show that the prediction accuracy of the four combined models is significantly improved in all the three studied conditions, and the combination with the convolutional neural network has the highest prediction accuracy. The combined models provide a feasible scheme for improving the prediction accuracy of freak waves.

Key words: rogue wave, long short-term memory (LSTM), convolutional neural network (CNN), empirical mode decomposition (EMD), auto-aggressive integrated average (ARIMA), Kalman filtering (KF)

CLC Number:

U661.1

ZHAO Yong, SU Dan. Rogue Wave Prediction Based on Four Combined Long Short-Term Memory Neural Network Models[J]. Journal of Shanghai Jiao Tong University, 2022, 56(4): 516-522.

Figures/Tables 11

Tab.1

Fig.1

Fig.2

Fig.3

Fig.4

Tab.2

Fig.5

Fig.6

Fig.7

Fig.8

Tab.3

References 16

[1]	王淑华. 基于ARIMA模型的冯家山水库水位预测研究[J]. 陕西水利, 2019(8): 45-47.
	WANG Shuhua. Research on water level prediction of Fengjiashan reservoir based on ARIMA model[J]. Shaanxi Water Resources, 2019(8): 45-47.
[2]	孙逸群, 包为民, 江鹏, 等. 基于无迹卡尔曼滤波的新安江模型实时校正方法[J]. 湖泊科学, 2018, 30(2): 488-496.
	SUN Yiqun, BAO Weimin, JIANG Peng, et al. Real-time updating of XAJ model by using unscented Kalman filter[J]. Journal of Lake Sciences, 2018, 30(2): 488-496. doi: 10.18307/2018.0220 URL
[3]	阚世宜, 于婷, 刘莉. 基于EMD分解的海浪有效波高短期预测研究[J]. 海洋科学前沿, 2019, 6(2): 51-63.
	KAN Shiyi, YU Ting, LIU Li. Short-term prediction of effective wave height based on EMD decomposition[J]. Frontier of Marine science, 2019, 6(2): 51-63.
[4]	DOONG D J, PENG J P, CHEN Y C. Development of a warning model for coastal freak wave occurrences using an artificial neural network[J]. Ocean Engineering, 2018, 169: 270-280. doi: 10.1016/j.oceaneng.2018.09.029 URL
[5]	赵勇, 苏丹, 邹丽, 等. 基于LSTM神经网络的畸形波预测[J]. 华中科技大学学报(自然科学版), 2020, 48(7): 47-51.
	ZHAO Yong, SU Dan, ZOU Li, et al. Rogue wave prediction based on LSTM neural network[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2020, 48(7): 47-51.
[6]	陆继翔, 张琪培, 杨志宏, 等. 基于CNN-LSTM混合神经网络模型的短期负荷预测方法[J]. 电力系统自动化, 2019, 43(8): 131-137.
	LU Jixiang, ZHANG Qipei, YANG Zhihong, et al. Short-term load forecasting method based on CNN-LSTM hybrid neural network model[J]. Automation of Electric Power Systems, 2019, 43(8): 131-137.
[7]	晏臻, 于重重, 韩璐, 等. 基于CNN+LSTM的短时交通流量预测方法[J]. 计算机工程与设计, 2019, 40(9): 2620-2624.
	YAN Zhen, YU Chongchong, HAN Lu, et al. Short-term traffic flow forecasting method based on CNN+LSTM[J]. Computer Engineering and Design, 2019, 40(9): 2620-2624.
[8]	魏骜, 茅大钧, 韩万里, 等. 基于EMD和长短期记忆网络的短期电力负荷预测研究[J]. 热能动力工程, 2020, 35(4): 203-209.
	WEI Ao, MAO Dajun, HAN Wanli, et al. Short-term load forecasting based on EMD and long short-term memory neural networks[J]. Journal of Engineering for Thermal Energy and Power, 2020, 35(4): 203-209.
[9]	SRIVASTAVA T, VEDANSHU, TRIPATHI M M. Predictive analysis of RNN, GBM and LSTM network for short-term wind power forecasting[J]. Journal of Statistics and Management Systems, 2020, 23(1): 33-47. doi: 10.1080/09720510.2020.1723224 URL
[10]	翟静, 曹俊. 基于时间序列ARIMA与BP神经网络的组合预测模型[J]. 统计与决策, 2016(4): 29-32.
	ZHAI J, CAO J. Combination forecasting model based on time series ARIMA and BP neural network[J]. Statistics and Decision, 2016(4): 29-32.
[11]	JAIN A, KUMAR A M. Hybrid neural network models for hydrologic time series forecasting[J]. Applied Soft Computing, 2007, 7(2): 585-592. doi: 10.1016/j.asoc.2006.03.002 URL
[12]	KHARIF C, PELINOVSKY E, SLUNYAEV A. Conclusion[M]//Rogue Waves in the Ocean. Berlin, Heidelberg: Springer Berlin Heidelberg, 2009: 211-212.
[13]	邹丽, 王爱民, 宗智, 等. 岛礁地形畸形波演化过程的试验及小波谱分析[J]. 哈尔滨工程大学学报, 2017, 38(3): 344-350.
	ZOU Li, WANG Aimin, ZONG Zhi, et al. Experiment and wavelet analysis of the evolution process of freak waves around reefs[J]. Journal of Harbin Engineering University, 2017, 38(3): 344-350.
[14]	LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86: 2278-2324. doi: 10.1109/5.726791 URL
[15]	魏志勇. 多时间序列上挖掘框架的研究[D]. 沈阳: 东北大学, 2009.
	WEI Zhiyong. Research on the framework of mining on multiple time series[D]. Shenyang: Northeastern University, 2009.
[16]	修春波, 任晓, 李艳晴, 等. 基于卡尔曼滤波的风速序列短期预测方法[J]. 电工技术学报, 2014, 29(2): 253-259.
	XIU Chunbo, REN Xiao, LI Yanqing, et al. Short-term prediction method of wind speed series based on Kalman filtering fusion[J]. Transactions of China Electrotechnical Society, 2014, 29(2): 253-259.

工况	有义波高/cm	最大波高/cm	波高比	峰值时间/s	峰值/cm
1	4.59	9.63	2.11	77.34	5.75
2	2.89	6.33	2.19	45.96	3.69
3	2.89	6.36	2.20	53.04	3.98

参数类别	模型	数量设置
隐藏层层数	LSTM	1
	LSTM-CNN	3
	LSTM-EMD	2
	LSTM-ARIMA	1
	LSTM-Kalman	1
隐藏层节点数	LSTM	128
	LSTM-CNN	300
	LSTM-EMD	20
	LSTM-ARIMA	3
	LSTM-Kalman	4
批尺寸	LSTM	32
	LSTM-CNN	16
	LSTM-EMD	20
	LSTM-ARIMA	1
	LSTM-Kalman	1
训练次数	LSTM	100
	LSTM-CNN	1000
	LSTM-EMD	100
	LSTM-ARIMA	100
	LSTM-Kalman	100

工况	e_m
工况	LSTM	LSTM- CNN	LSTM- ARIMA	LSTM- EMD	LSTM- Kalman
1	0.2838	0.1557	0.1621	0.1987	0.2047
2	1.0637	0.2685	0.3240	0.5965	0.6789
3	0.2664	0.1307	0.1413	0.2079	0.2366