DESIGN AND IMPLEMENTATION OF BACKSTEPPING CONTROL SYSTEM FOR PLATOONING FORMATION IN DOWN SCALE AUTONOMOUS TRUCK TRAILER

• Main Author: Muhamadinah, Yana
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/43372
• Institution: Institut Teknologi Bandung

Indonesia as an archipelago holds a large potential in the shipping sector. Trade between countries using containers in Indonesia continues to increase with an average of 2.39% per year with a total flow of containers reaching 14.92 million TEUs. Port automation can use autonomous truck trailers to reduce dwelling time at ports. Dense port activity requires space to load autonomous truck trailers so that the platooning system is designed with an objective in controlling the distance between robots. Platooning system is one type of the vehicle formation system consists of leader and follower(s) which the distance between the vehicle(s) is constant at a desired value. Research in this field is likely to use single rigid body assumption, such as in a car and heavy duty vehicle. In this research, platooning system is implemented in the scaled truck trailer where we use multi rigid bodies assumption. We design a platooning system which control the longitudinal velocity and keep the position in a desired value using integrator backstepping method. The objectives of this control are to make the velocity error and position error stable and converge to zero (the origin of the eror). Simulations show that the objectives control platoon achieved. When the leader’s velocity accelerates or decelerates, follower can track the leader’s velocity and stable again. The platoon control implemented on the scaled autonomous truck trailer using Lego Mindstorm EV3. Platoon coefficient used is kp1 = 0.5 and kp2 = 0.5, with the set point of distance between the robots, ds = 0.3 meters. From the results of the implementation, obtained an average position error value of ep = -0.3 meters and an average speed error of evp = 3.78 m / s. The ep and evp values are slightly oscillating but quite good with convergent values leading to zero.