Peramalan Seri Waktu Permukaan Laut Menggunakan Deep Learning Rnn, Lstm, Dan Bilstm, Studi Kasus Di Teluk Jakarta, Indonesia
DOI:
https://doi.org/10.34818/eoe.v7i2.12760Abstract
Abstrak Peramalan permukaan laut sangat penting dalam aktivitas di daerah pesisir, seperti pada bidang enginee- ring dan juga dapat digunakan untuk menentukan navigasi pelayaran kapal. Selain itu, dapat digunakan untuk membuat rancangan pembangunan dan perencanaan pada daerah pesisir di masa depan, dan juga untuk mengurangi resiko yang ditimbulkan akibat pasang surut air laut. Peramalan permukaan laut de- ngan metode tradisional, seperti tidal harmonic analysis, tidak dapat mempertimbangkan kontribusi kom- ponen non-tidal dalam peramalan permukaan laut. Dalam tugas akhir ini, penulis menggunakan metode deep learning untuk meramalkan permukaan laut. Penulis menggunakan tiga metode deep learning yaitu Recurrent Neural Network (RNN), Long Short Term Memory (LSTM), dan Bidirectional Long Short Term Memory (BiLSTM). Ketiga metode deep learning ini dibandingkan untuk melihat performanya dalam me- ramalkan permukaan laut selama 48 jam, 72 jam dan 168 jam ke depan. Data yang digunakan pada tugas akhir ini yaitu data permukaan laut yang diperoleh dari observasi di Pelabuhan Jakarta, Indonesia, sebagai studi kasusnya. Hasil percobaan menunjukkan bahwa metode BiLSTM memberikan hasil yang lebih baik dibandingkan dengan RNN dan LSTM. Meningkatnya kinerja BiLSTM dari pada dua metode lainnya ka- rena memungkinkan pelatihan tambahan dengan melintasi data input dua kali (yaitu, 1) dari kiri ke kanan dan 2) dari kanan ke kiri. Kata kunci : Permukaan laut, Peramalan, RNN, LSTM, BiLSTM Abstract Sea level forecasting is vital in coastal activities, such as engineering and naval navigation. Moreover, it can be used for making strategies for future coastal development and planning, and also for mitigating its serious consequences. Traditional sea level forecasting, such as tidal harmonic analysis, do not consider a non-tidal component contribution in the sea level forecasting. In this study, we use a deep learning approach to forecast sea level. We use three deep learning methods: the Recurrent Neural Network (RNN), the Long Short Term Memory (LSTM), and the Bidirectional Long Short Term Memory (BiLSTM). These three methods of deep learning are compared to show their performances to forecast the sea level for 48 h, 72 h and 168 h ahead. We use the sea level data obtained from observation at Jakarta Port, Indonesia, as our study case. The results of the numerical experiment show that the BiLSTM method gives better performance than the RNN and the LSTM. The BiLSTM improve performance than the other two methods due enables additional training by traversing the input data twice (i.e., 1) left to right and 2) right to left. Keywords: Sea Level, Forecasting, RNN, LSTM, BiLSTM.
Downloads
References
[1] N. Anggraini, B. Trisakti, and T. E. B. Soesilo. Application of sattelite data to analyze inundation potential and the impact of sea level rise. Jurnal Penginderaan Jauh dan Pengolahan Data Citra Digital, 9(2), 2012.
[2] Q. Cao, B. T. Ewing, and M. A. Thompson. Forecasting wind speed with recurrent neural networks. European Journal of Operational Research, 221(1):148–154, 2012.
[3] E. Chaussard, F. Amelung, H. Abidin, and S.-H. Hong. Sinking cities in indonesia: Alos palsar detects rapid subsidence due to groundwater and gas extraction. Remote sensing of environment, 128:150–161, 2013.
[4] R. Djaja, J. Rais, H. Z. Abidin, and K. Wedyanto. Land subsidence of jakarta metropolitan area. In Procee- dings of the 3rd FIG Regional Conference for Asia and the Pacific, pages 3–7, 2004.
[5] T. Firman, I. M. Surbakti, I. C. Idroes, and H. A. Simarmata. Potential climate-change related vulnerabilities in jakarta: Challenges and current status. Habitat International, 35(2):372–378, 2011.
[6] M. A. Ghorbani, R. Khatibi, A. Aytek, O. Makarynskyy, and J. Shiri. Sea water level forecasting using genetic programming and comparing the performance with artificial neural networks. Computers & Geosciences, 36(5):620–627, 2010.
[7] A. Graves and J. Schmidhuber. Framewise phoneme classification with bidirectional lstm and other neural network architectures. Neural networks, 18(5-6):602–610, 2005.
[8] S. Hochreiter and J. Schmidhuber. Long short-term memory. Neural computation, 9(8):1735–1780, 1997.
[9] S. Karimi, O. Kisi, J. Shiri, and O. Makarynskyy. Neuro-fuzzy and neural network techniques for forecasting sea level in darwin harbor, australia. Computers & Geosciences, 52:50–59, 2013.
[10] W. Kong, Z. Y. Dong, Y. Jia, D. J. Hill, Y. Xu, and Y. Zhang. Short-term residential load forecasting based on lstm recurrent neural network. IEEE Transactions on Smart Grid, 10(1):841–851, 2017.
[11] G. A. Meehl, W. M. Washington, W. D. Collins, J. M. Arblaster, A. Hu, L. E. Buja, W. G. Strand, and H. Teng. How much more global warming and sea level rise? Science, 307(5716):1769–1772, 2005.
[12] X. Qing and Y. Niu. Hourly day-ahead solar irradiance prediction using weather forecasts by lstm. Energy, 148:461–468, 2018.
[13] A. Rahman, V. Srikumar, and A. D. Smith. Predicting electricity consumption for commercial and residential buildings using deep recurrent neural networks. Applied energy, 212:372–385, 2018.
[14] M. Ramdhan, S. N. Amri, and D. G. Priyambodo. Identification survey of sinking island in jakarta bay.
Jurnal Riset Jakarta, 12(1):1–6, 2019.
[15] A. Saxena and T. Sukumar. Predicting bitcoin price using lstm and compare its predictability with arima model. Int. J. Pure Appl. Math, 119(17):2591–2600, 2018.
[16] P. K. Srivastava, T. Islam, S. K. Singh, G. P. Petropoulos, M. Gupta, and Q. Dai. Forecasting arabian sea level rise using exponential smoothing state space models and arima from topex and jason satellite radar altimeter data. Meteorological Applications, 23(4):633–639, 2016.
[17] Y.-T. Tsai, Y.-R. Zeng, and Y.-S. Chang. Air pollution forecasting using rnn with lstm. In 2018 IEEE 16th Intl Conf on Dependable, Autonomic and Secure Computing, 16th Intl Conf on Pervasive Intelligence and Computing, 4th Intl Conf on Big Data Intelligence and Computing and Cyber Science and Technology Congress (DASC/PiCom/DataCom/CyberSciTech), pages 1074–1079. IEEE, 2018.
[18] Z. Zhao, W. Chen, X. Wu, P. C. Chen, and J. Liu. Lstm network: a deep learning approach for short-term traffic forecast. IET Intelligent Transport Systems, 11(2):68–75, 2017.



