Sonya, Windari Afrita (2017) Perancangan Sistem Prediktor Ketinggian Gelombang Berbasis Thiessen Polygon dan Jaringan Saraf Tiruan di Perairan Dangkal Jawa Timur. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2413100061-undergraduate_thesis.pdf - Published Version Download (2MB) | Preview |
Abstract
Kasus kecelakaan transportasi laut yang terjadi sebagian besar diakibatkan oleh faktor cuaca. Informasi data cuaca seperti ketinggian gelombang sangat penting untuk kegiatan transportasi laut agar tidak membahayakan keselamatan. Namun, peramalan cuaca oleh BMKG hanya untuk titik koordinat tertentu saja. Penelitian pada tugas akhir ini dilakukan perancangan sistem prediktor tinggi gelombang yang mengintegrasikan metode thiessen polygon dan jaringan saraf tiruan. Data yang digunakan yaitu data angin tahun 2012 sampai 2016. Objek pengambilan data diantaranya koordinat titik 6.874824oS-112.747800oE, koordinat 7.1449330S -114.106900 Edan koordinat 3.540425oS-114.484300E. Data angin tersebut digunakan untuk menghitung tinggi gelombang signifikan dengan metode Sverdrup Munk Bretchsneider . Hasil tinggi gelombang signifikandigunakan untuk peramalan tinggi gelombang dengan metode thiessen polygon koordinat yang tidak memiliki data tinggi gelombang yaitu koordinat p 5.5780290S-113.770440E.Perancangan prediktor tinggi gelombang koordinat 5.5780290S-113.770440Et+1 dilakukan dengan menggunakan jaringan saraf tiruan dengan arsitektur terbaik 36-10-12 dan nilai learning rate 0.3. Hasil validasi terhadap tinggi gelombang ramalan metode thiessen polygon diperoleh nilai MAPE 12.59%dari sumber data aktual yaitu situs www.buoyweather.com. Validasi data pengujian dari perancangan sistem prediktor jaringan saraf tiruan didapatkan nilai MAPE 18,86%. Hasil ketepatan peramalan prediksi tinggi gelombang tahun 2017 dari sistem prediktor jaringan saraf tiruan diperoleh MAPE sebesar 15,95%.
=====================================================================================================
Cases of marine transport accidents that occur largely due to weather factors. Weather data information such as wave height is essential for marine transportation activities in order not to endanger safety. However, weather forecasting by BMKG is only for certain coordinate points only. The research on this final project is designing high wave predictor system which integrates thiessen polygon and neural network method. The data used are wind data from 2012 to 2016. Objects of data retrieval are coordinates point 6.874824oS-112.747800oE, coordinates 7.1449330S -114.106900 E and coordinates 3.540425oS-114.484300E. The wind data is used to calculate significant wave heights by the Sverdrup Munk Bretchsneider method. Significant wave height results are used for wave height forecasting with thiessen polygon coordinate method which has no wave height data ie p 5.5780290S-113.770440E coordinates. The design of high-wave predictor coordinate 5.5780290S-113.770440E t + 1 is done by using artificial neural network with the best architecture 36-10-12 and the value of learning rate 0.3. The result of validation of the wave height forecast of thiessen polygon method obtained MAPE value of 12.59% from actual data source ie website www.buoyweather.com. Validation of test data from design of predictor system of artificial neural network got MAPE value 18,86%. The result of accurate prediction of wave height prediction of 2017 from artificial neural network predictor system obtained by MAPE 15,95%.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | jaringan saraf tiruan, SMB, thiessen polygon, tinggi gelombang signifikan |
| Subjects: | Q Science > QC Physics T Technology > T Technology (General) T Technology > T Technology (General) > T58.62 Decision support systems |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Afrita Sonya Windari |
| Date Deposited: | 18 Aug 2017 04:04 |
| Last Modified: | 18 Aug 2017 04:04 |
| URI: | http://repository.its.ac.id/id/eprint/43342 |
Actions (login required)
 |
View Item |