Fusion BASED AGV Robot Navigation Solution Comparative Analysis and Vrep Simulation

Abstract

Today, Automated Guided Vehicle (AGV) robots are integral to many factories. One of the basic problems of these robots is accurate navigation, which, in addition to creating security in performing tasks, it helps the robot to manage the battery power and energy and move on a predetermined path. Over the various method, the visions are based on good performance, recently been widely used. In this research, a simple but effective fusion method as the combination of vision(camera) and infrared (IR) sensors with the minimum number of sensors is proposed and implemented. The proposed method has been simulated and evaluated using the Vrep simulator with real dimensions of our previous design AGV system named Hongma and the Python API. In the simulation, the proposed method was carried out. In the experiment, five paths named Circle, Elliptical, Spiral,8 shapes, and Special path, different paths with different complexity were tested, and the experiment aimed to find the maximum speed at which the proposed algorithm and the vision sensor (camera sensors) can track the path with a 100% success rate. The results obtained in the experiment show the fusion method’s effectiveness over the five mentioned scenarios in tracking the planned path compared to the routine vision method.