Muhammad, Nobel Danial (2021) Pemodelan Lalu Lintas Kendaraan Di Jalan Tol Ruas Jombang-Mojokerto Menggunakan Metode Machine Learning Sebagai Prediksi Dan Analisa Faktor Kecepatan Kendaraan. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
09211950055004-Master_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Tol Jombang-Mojokerto merupakan salah satu ruas tol bagian dari Tol Trans Jawa (Banyuwangi-Merak) yang terintegrasi dengan empat ruas tol lainnya dalam Cluster 3 (Surabaya-Semarang). Adanya ruas jalan tol ini membuat pengguna jalan dapat berkendara dengan aman, nyaman dan lancar. Namun seringkali pengguna jalan masih melaju dengan kecepatan yang melewati batas sesuai peraturan pemerintah mengenai jalan tol yaitu minimal 60 km/jam dan maksimal 100 km/jam (luar kota), hal ini sangat berpotensi terjadinya kecelakaan lalu lintas. Penelitian ini bertujuan melakuan pemodelan data lalu lintas kendaraan yang melewati jalan tol dengan menggunakan metode Machine Learning Random Forest yang mampu memodelkan data lalu lintas kendaraan yang melebihi batas kecepatan dengan hasil akurasi prediksi yang baik. Penelitian ini mengambil data lalu lintas kendaraan yang melewati ruas tol jombang-mojokerto yang diambil dari data transaksi kendaraan yang keluar di Gerbang Tol Bandar. Proses penelitian ini dimulai dari pengambilan & pemrosesan data, pengembangan model classifier Random Forest, pengujian data dan analisa hasil pengujian. Hasil dari penelitian ini didapatkan ketepatan klasifikasi yang cukup baik dengan akurasi 78% dan nilai AUC 0,75, serta diketahui variabel-variabel penting/signifikan dengan urutan empat besar nilai gini importance tertinggi yaitu jam masuk tol, jam keluar tol, gerbang asal, dan golongan kendaraan. Dengan demikian hasil penelitian ini dapat menjadi alat bantu keputusan bagi operator jalan tol maupun pihak lain yang berkepentingan dalam melakukan strategi tindakan preventif maupun kuratif kepada pengguna jalan yang berkendara melebihi batas aturan jalan Tol Jombang-Mojokerto dengan dasar varibel penting yang sudah diketahui.
======================================================================================================
The Jombang-Mojokerto toll road is a part of the Trans Java (Banyuwangi-Merak)
toll road which is integrated with four other toll roads in Cluster 3 (Surabaya�Semarang). The existence of this toll road allows road users to drive safely,
comfortably and smoothly. However, road users often still drive at speeds that
exceed the limits according to government regulations regarding toll roads, which
are a minimum of 60 km/hour and a maximum of 100 km/hour (outside the city),
this has the potential to cause traffic accidents. This study aims to model vehicle
traffic data that passes through toll roads using the Machine Learning Random
Forest method which is able to model vehicle traffic data that exceeds the speed
limit with good prediction accuracy results. This study takes vehicle traffic data that
passes through the Jombang-Mojokerto toll road which is taken from vehicle
transaction data that comes out at the Bandar Toll Gate. The research process starts
from data retrieval & processing, development of a random forest classifier model,
data testing and analysis of test results. The results of this study obtained a fairly
good classification accuracy with an accuracy of 78% and an AUC value of 0.75,
and it is known that the important/significant variables in the order of the top four
values of the highest Gini Importance are toll entry hours, toll exit hours, origin
gates, and vehicle class. Thus the results of this study can be a decision aid for toll
road operators and other interested parties in carrying out preventive and curative
action strategies for road users who drive beyond the limits of the Jombang�Mojokerto toll road rules on the basis of known important variables.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Machine Learning, Random Forest, Toll Road, Vehicle Speed, Jalan Tol, Kecepatan Kendaraan, Machine Learning, Random Forest. |
| Subjects: | H Social Sciences > HE Transportation and Communications H Social Sciences > HE Transportation and Communications > HE355.4 Toll roads Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. |
| Divisions: | Faculty of Creative Design and Digital Business (CREABIZ) > Technology Management > 61101-(S2) Master Thesis |
| Depositing User: | Nobel Danial Muhammad |
| Date Deposited: | 30 Jul 2021 06:17 |
| Last Modified: | 19 Sep 2024 07:47 |
| URI: | http://repository.its.ac.id/id/eprint/84601 |
Actions (login required)
 |
View Item |