OBSTACLE AVOIDANCE CONTROL SYSTEM DESIGN FOR AUTONOMOUS TRUCK-TRAILER WITH LANE CHANGE MANEUVER

• Main Author: Komang Agus Jony Wirawan, I
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49898
• Institution: Institut Teknologi Bandung

Autonomous vehicle can be a solution to reduce cost of delivering container from dock to yard or vice versa. One of the autonomous vehicles that can be used is Automated Guided Vehicle (AGV). In general, AGV is designed to be easily controlled but in exchange it has high procurement cost. To resolve this problem, type of AGV that can be use is manual truck-trailer that has been modified with integrated sensors and actuators so it can be controlled and operate autonomously. This study will explain about the method of object detection and obstacle avoidance in autonomous truck-trailer with control objective that focus on trailer part. Two methods of lane change used in this study are slant lane and two quarter circle with same radius that done in MATLAB. Simulation result shows that these two methods are stable and dynamic errors for head-truck and trailer convergence to zero. Furthermore, two quarter circle with same radius method gives a smaller Euclidean norm of error for head and trailer, 0.56 and 0.23 for each as well as smaller total IAE, ITAE and ISE, 2,649.87, 112,595.30 and 2,388.92, respectively. Then, two quarter circle with same radius method is applied on autonomous truck-trailer prototype in Lego Mindstorm EV3 Brick with RPLiDAR A1M8 on trailer part to detect object that block the lane of autonomous truck-trailer in 120° angle range. Implementation results on prototype represent simulation result in MATLAB before. Prototype can detect object, change its lane, and its position dynamic errors convergence to zero.