Mechanism and Convergence Analysis of a Multi-Robot Swarm Approach Based on Natural Selection

摘要: The Darwinian Particle Swarm Optimization (DPSO) is an evolutionary algorithm that extends the (PSO) using natural selection, or survival-of-the-fittest, to enhance ability escape from local optima. An extension of DPSO multi-robot applications has been recently proposed and denoted as Robotic PSO (RDPSO), benefiting dynamical partitioning whole population robots. Therefore, RDPSO decreases amount required information exchange among robots, scalable large populations This paper presents a stability analysis better understand relationship between parameters robot's convergence. Moreover, further extended for real robot constraints (e.g., dynamics, obstacles communication constraints) experimental assessment with physical optimal are evaluated in groups robots larger simulated mobile different target distributions within scenarios. Experimental results show able converge regardless defined attraction domain. However, more conservative parametrization significant influence on convergence time. To evaluate herein approach, compared four state-of-the-art swarm robotic alternatives under simulation. It observed provably converges solution faster accurately than other approaches.