Primal-Dual Asynchronous Particle Swarm Optimization (pdAPSO) Algorithm For Self-Organized Flocking of Swarm Robots

This paper proposed a hybrid PSO algorithm that combines the Primal-Dual method with APSO algorithm to address the problem of swarm robotics flocking motion. This algorithm combines the explorative ability of APSO with the exploitative capacity of the Primal Dual Interior Point Method. We hypothesize that the fusion of the two algorithms (APSO and Primal Dual) offers a robust prospect of preventing premature convergence of robots, and also make sure that the robots are not stuck in their local minimal. We did a comparison of the performances of the total iteration for pdAPSO, PSO, APSO and Primal Dual algorithms as the robots flock from the centre of the search space to the various zones (z1, z2, z3, and z4). In three (3) out of the four (4) cases, the pdAPSO proves to be more effective for the flocking of the robots than the PSO, APSO and Primal Dual algorithms for the fifty (50) simulations that was done. The results of our simulation gives a clear evidence of the efficacy of the pdAPSO algorithms. The hybrid algorithm contributed to the field of swarm robotics by providing novel algorithm that possess a flocking capability attained under suitable parameter values that is relatively robust and produces effective self-organized flocking in constrained environments.