Rancang Bangun Sistem Kontrol Kecepatan dan Kemudi pada Mobil Otonom dengan Model Predictive Control Berbasis Robot Operating System

Farizi, Atras Yonda (2024) Rancang Bangun Sistem Kontrol Kecepatan dan Kemudi pada Mobil Otonom dengan Model Predictive Control Berbasis Robot Operating System. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009201127-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Pengembangan mobil otonom yang dapat diandalkan perlu diiringi dengan perancangan sistem kontrol kecepatan dan kemudi yang mampu menyelesaikan persoalan model kendaraan yang non-linier serta aspek kontrol longitudinal-lateral yang kompleks agar dapat berjalan mengikuti lintasan yang ditentukan oleh local planner. Sistem kontrol kecepatan dan kemudi yang dirancang pada penelitian ini menggunakan model predictive control untuk mengendalikan aspek kinematika gerak mobil otonom. Model kinematika gerak mobil yang non-linier didekati dengan model sistem linear time-varying yang berubah terhadap kondisi operasi disetiap waktu. Algoritma sistem kontrol diterapkan pada prototipe mobil otonom yang dirancang menggunakan single board computer Jetson Nano serta mikrokontroler STM32 sebagai perangkat pengendali motor DC dan servo kemudi. Pada penelitian ini, performansi kontroler MPC dievaluasi berdasarkan nilai root mean square error dan beban komputasi dalam mengeksekusi program. Beban komputasi kontroler MPC yang didapatkan pada saat simulasi dengan Gazebo ROS serta pengujian hardware skenario 1 dan 2 masing-masing sebesar 0,2456, 0,8689, dan 0,9064 detik setiap iterasi. Hal tersebut menunjukkan adanya peningkatan beban komputasi antara simulasi dan pengujian hardware yang mempengaruhi kinerja kontroler MPC untuk dapat bekerja secara real-time. Pada saat simulasi dan pengujian dengan hardware, nilai root mean square error masing-masing state memiliki nilai dibawah 0,6 dan 0,8. Secara keseluruhan, hasil pengujian tersebut menunjukkan bahwa sistem kontrol kecepatan dan kemudi mobil otonom dengan model predictive control yang dirancang membuat mobil dapat berjalan mengikuti setpoin lintasan yang diberikan oleh local planner.
============================================================
The development of a reliable autonomous car requires the design of a speed and steering control system capable of addressing the non-linear car model and complex longitudinal-lateral control aspects to ensure it follows the trajectory provided by local planner. The speed and steering control system designed in this research employs model predictive control to manage the kinematic aspects of the autonomous car's motion. The non-linear motion kinematics model of the car is approximated by a linear time-varying system model that changes according to different operational conditions over time. The control system algorithm is implemented on an autonomous car prototype designed using a Jetson Nano single-board computer and an STM32 microcontroller to control the DC motor and steering servo. In this research, the performance of the MPC controller is evaluated based on the root mean square error and computational load in executing the program. The computational loads of the MPC controller obtained during simulation with Gazebo ROS and hardware testing for scenarios 1 and 2 are 0.2456, 0.8689, and 0.9064 seconds per iteration, respectively. This indicates an increase in computational load between simulation and hardware testing, which affects the real-time performance of the MPC controller. During simulation and hardware testing, the root mean square error values for each state were below 0.6 and 0.8, respectively. Overall, these test results indicate that the speed and steering control system of the autonomous car using model predictive control designed in this study enables the car to follow the trajectory setpoint provided by the local planner.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Mobil Otonom, MPC, ROS, Sistem Kontrol, Autonomous Car, Control System, MPC, ROS
Subjects:	T Technology > T Technology (General) T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.8 Vehicles, Remotely piloted. Autonomous vehicles. T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL521.3 Automatic Control
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Atras Yonda Farizi
Date Deposited:	02 Aug 2024 08:43
Last Modified:	02 Aug 2024 08:43
URI:	http://repository.its.ac.id/id/eprint/109976

Actions (login required)

View Item