Obstacles Avoidance in a Self Path Plannning of a Polar Robot

Abstract

The control of a robot's movement in the self path planning problem requires the balance between time response and task requirements. In the case of robotic arms, both conditions are essential to turn a task efficient enough to be included into a productive environment. The robot's movement has been researched several times trying to give them the capability of establishing a proper response under specific conditions without human interference. To achieve this goal, several strategies have been used, including spatial maps, area sweeping, iterative models and, as in this case, a neural genetic algorithm, in which the genetic algorithm builds the path based in a set of spatial positions, and the neural network learns from each of these paths to associate them with future positions and tasks. The target, is to adjust the robot's performance to a level in which it is able to self define if a new path requires a searching process, if a previous sequence can be used, or if the points that conform the movement structure need to be adjusted in accordance to the accuracy needed