DESIGN AND IMPLEMENTATION OF TRAJECTORY GENERATION AND TRACKING SYSTEM FOR FLOCKING IN NONHOLONOMIC ROBOT WITH OBSTACLE AND COLLISION AVOIDANCE

• Main Author: Julian Damanik (13314004), Dewi Retno Andriani (13314054) , Joshua
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/28324
• Institution: Institut Teknologi Bandung

<p align="justify">The behavior of clustering and congregating in living creatures is a natural phenomenon that gives many advantages. This behavior is also known as flocking. Each member in a swarm or flock is called an agent. The algorithm of flocking is being developed in multi-agents control system engineering field. The behaviors of multiagents, which are able to work fast and well coordinated, could benefit industries, which is needed in high mobility products or raw materials distribution, to increase productivity. <br /> <br /> The multi-agents system has two controllers, flocking controller and tracking controller. The objects, which are used in this research, are three nonholonomic robots with differential wheels type from Lego Mindstorm EV3. The flocking controller could be divided into three criterias such as the interactions between main flocking agents, the interactions of main flocking agents with obstacles, and the interactions of main flocking agents with virtual leader. To construct tracking control system, it is needed to derive dynamic model of nonholonomic robot in a cartesian coordinate to make a stable control’s input model. <br /> <br /> The implementation of the flocking controller and the tracking controller on the nonholonomic robots requires some constants to satisfy the success criteria of the nonholonomic robot movement. The flocking controller constants, ka, ko, and ky with the value of 0.1, 2, and 0.15 respectively, and the tracking controller constants, k1 and k2 with the value of 1 and 0.5 respectively are able to satisfy the criterias.<p align="justify">