PENGEMBANGAN MODEL PREDIKSI NILAI RPM UNTUK MENCEGAH SLIP CLUTCH MENGGUNAKAN MACHINE LEARNING

Dany Syauqi Nazhif, - and Mahmudah Salwa Gianti, - and Diky Zakaria, - (2025) PENGEMBANGAN MODEL PREDIKSI NILAI RPM UNTUK MENCEGAH SLIP CLUTCH MENGGUNAKAN MACHINE LEARNING. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Penelitian ini mengembangkan model prediksi Revolutions Per Minute (RPM) untuk mencegah slip clutch pada kendaraan menggunakan algoritma eXtreme Gradient Boosting (XGBoost). Slip clutch merupakan kondisi di mana terjadi perbedaan signifikan antara putaran mesin dan kecepatan kendaraan, yang dapat menimbulkan penurunan performa, peningkatan konsumsi bahan bakar, dan kerusakan komponen transmisi. Pencegahan slip clutch memerlukan sistem monitoring dan prediksi RPM yang andal, sehingga potensi slip dapat diidentifikasi lebih awal dan pengemudi dapat mengambil tindakan korektif. Data diperoleh dari perangkat IoT Teltonika FMC003 pada Daihatsu Xenia 1.3R CVT 2022 melalui port OBD-II, meliputi torsi, kecepatan, horsepower, engine load, dan RPM. Proses penelitian dimulai dengan data cleaning untuk menghilangkan noise, dilanjutkan dengan analisis eksplorasi guna memahami pola hubungan antar variabel. Model XGBoost kemudian dilatih menggunakan hyperparameter tuning untuk mengoptimalkan performa prediksi. Evaluasi model dilakukan menggunakan metrik Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), serta koefisien determinasi (R^2). Hasil menunjukkan model memiliki akurasi tinggi (R^2 0,9983; RMSE 20,04; MAPE 0,98%) dan mampu memprediksi RPM dalam data 1 jam berkendara. Sistem ini berpotensi menjadi early warning untuk mencegah slip clutch serta mendukung predictive maintenance berbasis data real-time. Implementasi model ini diharapkan dapat diaplikasikan pada berbagai jenis kendaraan transmisi otomatis untuk meningkatkan keselamatan, efisiensi bahan bakar, dan umur pakai komponen transmisi. _____ This study develops a Revolutions Per Minute (RPM) prediction model to prevent slip clutch in vehicles using the eXtreme Gradient Boosting (XGBoost) algorithm. Slip clutch is a condition where there is a significant difference between engine speed and vehicle speed, which can cause decreased performance, increased fuel consumption, and damage to transmission components. Preventing clutch slippage requires a reliable RPM monitoring and prediction system, so that potential slippage can be identified early and the driver can take corrective action. Data was obtained from the Teltonika FMC003 IoT device on the 2022 Daihatsu Xenia 1.3R CVT via the OBD-II port, including torque, speed, engine load, and RPM. The research process began with data cleaning to remove noise, followed by exploratory analysis to understand the relationship patterns between variables. The XGBoost model was then trained using hyperparameter tuning to optimize predictive performance. Model evaluation was carried out using the Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and coefficient of determination (R^2) metrics. The results showed the model had high accuracy (R^2 0.9983; RMSE 20.04; MAPE 0.98%) and was able to predict RPM up to 1 hour in advance. This system has the potential to provide early warning to prevent clutch slippage and support predictive maintenance based on real-time data. This model is expected to be applicable to various types of automatic transmission vehicles to improve safety, fuel efficiency, and transmission component lifespan.

	Text S_MKB_2101034_Title.pdf Download (1MB)
	Text S_MKB_2101034_Chapter 1.pdf Download (261kB)
	Text S_MKB_2101034_Chapter 2.pdf Restricted to Staf Perpustakaan Download (730kB) | Request a copy
	Text S_MKB_2101034_Chapter 3.pdf Download (791kB)
	Text S_MKB_2101034_Chapter 4.pdf Restricted to Staf Perpustakaan Download (1MB) | Request a copy
	Text S_MKB_2101034_Chapter 5.pdf Download (329kB)
	Text S_MKB_2101034_Appendix.pdf Restricted to Staf Perpustakaan Download (2MB) | Request a copy

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?hl=en&user=RvFarwIAAAAJ ID SINTA Dosen Pembimbing Mahmudah Salwa Gianti : 6779018 Diky Zakaria : 6779007
Uncontrolled Keywords:	RPM, slip clutch, machine learning, XGBoost, IoT
Subjects:	L Education > L Education (General) Q Science > Q Science (General) T Technology > TJ Mechanical engineering and machinery T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:	UPI Kampus Purwakarta > S1 Mekatronika dan Kecerdasan Buatan
Depositing User:	Dany Syauqi Nazhif
Date Deposited:	28 Aug 2025 08:33
Last Modified:	28 Aug 2025 08:33
URI:	http://repository.upi.edu/id/eprint/136149

Actions (login required)

View Item