Perancangan Sistem Kontrol untuk Obstacle Avoidance pada Autonomous Vehicle dengan Kontrol Prediktif Berbasis Robot Operating System (ROS)

Arfianti, Rida Ayu (2024) Perancangan Sistem Kontrol untuk Obstacle Avoidance pada Autonomous Vehicle dengan Kontrol Prediktif Berbasis Robot Operating System (ROS). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6009221012-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (3MB) | Request a copy

Abstract

Autonomous vehicle telah meningkat pesat dalam beberapa tahun terakhir. Peningkatkan sistem keamanan saat berkendara diperlukan. Secara umum autonomous vehicle dapat melakukan self-driving dikarenakan adanya sistem persepsi dan lokalisasi, path planning, serta sistem kontrol. Sistem persepsi dan lokalisasi untuk menciptakan "kewaspadaan diri" terhadap environment. Path planning adalah merancang jalur dari posisi awal ke posisi tujuan dengan memperhatikan dari dua prespektif yaitu secara global path planning (algoritma A*) dan local path planning (Dynamic Window Approach (DWA)). Sistem kontrol yang digunakan adalah kontrol prediktif dengan receding horizon principle. Sistem kontrol yang menggunakan pemodelan matematis akan lebih baik dan tepat, ketika memprediksi aksi selanjutnya yang dilakukan autonomous vehicle dengan obstacle avoidance. Hal ini ditunjukkan dengan hasil terbaik dari sistem kontrol prediktif adalah memiliki nilai error untuk skenario pertama 0.798, skenario kedua 0.606. dan skenario ketiga 0.806. Hasil ini menunjukkan bahwa sistem kontrol prediktif yang dirancang dapat menghindari obstacle statis menyerupai kendaraan, sampai di titik tujuan, dan dapat berjalan secara autonomous
===================================================================================================================================
Autonomous vehicles have improved rapidly in recent years. Improving the safety system while driving is needed. In general, autonomous vehicles can carry out self-driving due to the presence of a perception and localization system, path planning, and control system. Perception and localization system to create "self-awareness" of the environment. Path planning is designing a path from the initial position to the destination position by taking into account two perspectives, namely global path planning (A* algorithm) and local path planning (Dynamic Window Approach (DWA)). The control system used is predictive control with the receding horizon principle. A control system that uses mathematical modeling will be better and more precise when predicting the next action carried out by an autonomous vehicle with obstacle avoidance. This is shown by the best results from the predictive control system which has an error value for the first scenario of 0.798, the second scenario is 0.606. and the third scenario 0.806. These results show that the designed predictive control system can avoid static obstacles such as vehicles, arrive at the destination point, and can run autonomously

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Kendaraan Otonom, Kontrol Prediktif, Penghindaran Rintangan Statis, ROS; Autonomous Vehicle, Model Predictive Control, Static Obstacle Avoidance, ROS
Subjects:	T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.8 Vehicles, Remotely piloted. Autonomous vehicles.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30101-(S2) Master Thesis
Depositing User:	Rida Ayu Arfianti
Date Deposited:	19 Feb 2024 03:41
Last Modified:	19 Feb 2024 03:41
URI:	http://repository.its.ac.id/id/eprint/107462

Actions (login required)

View Item