Genetic Algorithm for Walking Robots Motion Optimization

摘要

The paper presents a method to determine the optimal solution regarding the movement done by the leg of a hexapod walking robot with three degrees of freedom, with higher positioning precision taking into account the importance of three factors (Precision, Movement and Friction). To find the coefficients that establish the order of these factors we use an Analytical Hierarchy Process. With these coefficients we build the fitness function for a genetic algorithm that is slightly modified (instead of two chromosomes crossed, we crossed three) in order to obtain a diversity in the exchanged information. The obtained results prove that, applying the genetic algorithm in correlation with the analytic hierarchy method of processing, we can optimize the precision in the robot movement with a relative small increase of the consumed energy.