A*-Based Co-Evolutionary Approach for Multi-Robot Path Planning with Collision Avoidance
Morteza Kiadi, Enol García González, José R. Villar, Qing Tan
Abstract
In this research, a coevolutionary collision free multi-robot path planning that makes use of A* is proposed. To find collision-free paths for all robots, we generate a route for each of robot using A* path finding but introducing restrictions for each collision found. Afterward, a co-evolutionary optimization process is implemented for introducing changes in the initial paths to find a combination of routes that is collision-free. The approach has been tested in mazes with increasing the number of robots, showing a robust performance although at high time expenses. Nevertheless, several enhancements are proposed to tackle this issue.
Topics & Concepts
CollisionCollision avoidanceRobotPath (computing)Motion planningComputer scienceProcess (computing)Mathematical optimizationArtificial intelligenceSimulationMathematicsComputer securityOperating systemProgramming languageRobotic Path Planning AlgorithmsMetaheuristic Optimization Algorithms ResearchEvolutionary Algorithms and Applications