Analysis Of The Performance According To Object Density In Static Environments of GA and PSO Algorithms Used In Mobile Robot Path Planning

Abstract

In the movement of autonomous mobile robots in static or dynamic environments, one of the important issues sought for a solution is to reach the target with the shortest and safest path without collision. For this purpose, there are many algorithms based on probabilistic, potential field and artificial intelligence. The solutions brought by these algorithms differ according to the dynamics of the environment. However, as is known, the real world environment is complex. As the environment gets more complex, more environment knowledge is required for the performance of the algorithms. Complex mobile robotic systems equipped with sensors are required to obtain environmental information. This causes more energy consumption, processing load and the formation of heavy structures. In order to solve these problems, there are algorithms that perform path planning without the need for all environment information. Two of these are the Genetic Algorithm (GA) and the Particle Swarm Optimization (PSO) algorithm. In this study, the performance of both algorithms was compared according to the object density in the environment. Objects in the environment were detected according to the image information. Distance, time, curvature, and processing speed analyzes were performed in Matlab / Simulink environment according to different density environment scenarios.