Dhivasari, Citra (2012) Identifikasi Sinyal Akustik Bawah Air Yang Diakibatkan Oleh Kavitasi Propeller. Other thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
2207100147-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (8MB) | Request a copy |
Abstract
ldentifikasi suatu gelombang akustik dapat berguna dalam bidang aplikasi yang cukup luas. Tugas akbir ini difokuskan pada pengenalan terhadap sinyal akustik bawah air akibat kavitasi propeller. Karena penghasil noise terbesar dari propeller suatu kapal adalah cavitation noise. Permasalahan di dalamnya cukup kompleks karena melibatkan banyak sinyal dengan karakteristik yang tidak stabil. Untuk mencapai tujuan dari masalah di atas, solusi yang diterapkan adalah melakukan identifikasi berdasarkan metode neural networks. Dengan pre-processing berupa FFT 512 point. Algoritma neural networks yang digunakan adalah backpropagation, yaitu sistem pembelajaran yang melakukan fase backward saat output jaringan belum memenuhi target yang diinginkan. Jaringan menggunakan 3 lapisan, yaitu lapisan input, lapisan tersembunyi, dan lapisan output. Dengan neuron pada masingmasing lapisan secara berurutan adalah 31 , 18, dan l. Hasil pengujian dari database Sinyal Kavitasi Awal, Sinyal Kavitasi Panjang, dan Sinyal Non-Kavitasi didapatkan kecocokan pola secara berturut-turut sebesar 80%, 20%, dan 95%. Yang berarti bahwa database Sinyal Non-kavitasi lebih mampu membedakan sinyal kavitasi dibandingkan database lainnya. Semua data dan parameter yang digunakan disimulasikan dan dianalisa menggunakan software MatLab
===================================================================================================================================
Identification of acoustic waves can be useful in a wide range of applications. This final project focuses on recognizing underwater acoustic signals caused by propeller cavitation. Cavitation noise is the largest noise generator from a ship's propeller. The problem is quite complex, involving multiple signals with unstable characteristics. To achieve the above objectives, the solution implemented is to perform identification based on the neural network method. Preprocessing is in the form of a 512-point FFT. The neural network algorithm used is backpropagation, a learning system that performs a backward phase when the network output does not meet the desired target. The network uses three layers: an input layer, a hidden layer, and an output layer. The number of neurons in each layer is 31, 18, and 1, respectively. Test results from the Early Cavitation Signal, Long Cavitation Signal, and Non-Cavitation Signal databases obtained pattern matches of 80%, 20%, and 95%, respectively. This means that the Non-Cavitation Signal database is better able to distinguish cavitation signals than other databases. All data and parameters used were simulated and analyzed using MatLab software
Item Type: | Thesis (Other) |
---|---|
Additional Information: | RSE 621.384 11 Dhi i-1 2012 (weeding) |
Uncontrolled Keywords: | Underwater Acoustic Noise, Propeller Noise, Cavitation Noise, Backpropagation Neural Networks |
Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7878 Electronic instruments |
Divisions: | Faculty of Industrial Technology > Electrical Engineering > 20201-(S1) Undergraduate Thesis |
Depositing User: | EKO BUDI RAHARJO |
Date Deposited: | 19 Sep 2025 09:23 |
Last Modified: | 19 Sep 2025 09:23 |
URI: | http://repository.its.ac.id/id/eprint/128325 |
Actions (login required)
![]() |
View Item |