Genetic Algorithms To Predict The Optimums Path For A Polar Robot

摘要

This document is a review of how to employ genetic algorithms in the self path planning process by an industrial robot with a strong linkage with traditional procedures as a direct and inverse kinematics due the Denavit and Hartemberg algorithm. The first step solves gradually the direct kinematics equations providing a great number of positions which will be our first population. It is supposed that the robot will only move through existing individuals.rnThe second step involves the creation of a virtual environment capable to replicate the machines and equipment surrounding the robot.rnFinally, due a required target point the system looks for all the possible paths or individuals with capacity of avoiding all the obstacles along the path and at same time, capable to predict the convenience to continue in that way, if the individuals generated before are not capable to find a satisfactory condition. Genetic operators are applied in order to produce more positions but now the individuals will only be produced with an elitist criteria; that is, near the possible paths, out of the obstacles and with potential to continue advancing toward the target point.