Metode Supervised Machine Learning Untuk Sistem Pemantauan Emisi Prediktif Gas Buang Kapal Dalam Upaya Peningkatan Energy Efficiency Existing Ship Indeks (EEXI)

Hidayat, Ryamizard Gymnastiar (2025) Metode Supervised Machine Learning Untuk Sistem Pemantauan Emisi Prediktif Gas Buang Kapal Dalam Upaya Peningkatan Energy Efficiency Existing Ship Indeks (EEXI). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5019211104-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy

Abstract

Emisi gas buang kapal merupakan salah satu sumber polusi udara yang perlu dikendalikan untuk memenuhi regulasi efisiensi energi kapal seperti Energy Efficiency Existing Ship Index (EEXI) yang ditetapkan oleh IMO. Penelitian ini bertujuan untuk mengembangkan dan mengevaluasi pemodelan algoritma Random Forest dalam memprediksi jumlah emisi gas buang kapal, mengidentifikasi serta menganalisis tingkat pengaruh variabel yang memengaruhi emisi, serta menganalisis dampak pembatasan daya mesin kapal dalam mengevaluasi kepatuhan terhadap regulasi EEXI. Data yang digunakan bersumber dari buku harian deck, buku harian mesinn, dan data ship particulars, dengan proses feature creation yang dilakukan melalui perhitungan tahanan total, daya mesin, dan parameter lain menggunakan pendekatan model STEAM 2. Model dibangun dan dievaluasi menggunakan metrik Mean Absolute Error (MAE) dan Root Mean Square Error (RMSE) untuk mengukur tingkat akurasi prediksi. Hasil pemodelan menunjukkan nilai Mean Absolute Error (MAE) sebesar 0.14 kg/h dan Root Mean Square Error (RMSE) sebesar 0.52 kg/h. Hasil ini menunjukkan model mampu memprediksi nilai konsumsi bahan bakar dengan akurasi yang baik, di mana kecepatan kapal dan total resistance menjadi variabel dominan yang memengaruhi hasil prediksi. Selain itu, untuk mempermudah penerapan hasil model, dikembangkan pula sebuah dasbor interaktif menggunakan Streamlit yang dapat digunakan untuk memasukkan parameter operasi kapal dan menampilkan hasil prediksi konsumsi bahan bakar serta estimasi emisi secara langsung. Simulasi pembatasan daya mesin juga menunjukkan model dapat digunakan untuk mengevaluasi potensi kepatuhan kapal terhadap nilai EEXI yang disyaratkan.
=======================================================================================================================================
Ship exhaust emission are one of the main sources of air pollution that must be controlled to comply with ship energy efficiency regulations such as the Energy Efficiency Existing Ship Index (EEXI) set by IMO. This study aims to develop and evaluate a Random Forest algorithm model for predicting the amount of ship exhaust emissions, to identify and analyze the influence of variables affecting emissions, and to assess the impact of engine power limitation in evaluating compliance with the EEXI regulation. The data used in this study were obtained from deck log books, engine log books, and ship particulars, with the feature creation process conducted through the calculation of total resistance, engine power, and other parameters using the STEAM 2 model approach. The model was developed and evaluated using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) metrics to measure prediction accuracy. The modeling results indicate MAE of 0.14 kg/h and RMSE of 0.52 kg/h. These results demonstrate that the model is capable of predicting fuel oil consumption with good accuracy, where speed and total resistance are the dominant variables influencing the prediction outcomes. In addition, to facilitate the practical application of the model results, an interactive dashboard was developed using Streamlit, enabling users to input operational parameters and display fuel oil consumption predictions as well as emission estimates in real-time. The engine power limitation simulation also shows that the model can be utilized to evaluate the potential compliance of ships with the required EEXI value.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Emisi, Energy Efficiency Existing Ship Index, Konsumsi Bahan Bakar, Pembatasan Daya Mesin, Random Forest
Subjects:	V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering > VM276.A1 Fuel (Including supplies, costs, etc.)
Divisions:	Faculty of Marine Technology (MARTECH) > Marine Engineering > 36202-(S1) Undergraduate Thesis
Depositing User:	Ryamizard Gymnastiar Hidayat
Date Deposited:	04 Aug 2025 01:38
Last Modified:	20 Aug 2025 03:28
URI:	http://repository.its.ac.id/id/eprint/126477

Actions (login required)

View Item