The key behind the research represent in this paper is to understand the behavior of the particle swarm algorithm. This study proposes guaranteed convergence Particle Swarm Optimizer (GCPSO) with various topologies. The proposed GCPSO has evaluated the topology such as GBest, LBest, and Von Neumann Topology. It would be the most appropriate for different benchmark function such as Quadratic, Rosenbrock , Rastrigin, Griewank, Ackley, Shaffer’s f6 for guaranteed coverage using GCPSO, faster convergence and to find better local optima (local PSO) for large no of particle swarm for unimodel and multimodel functions. There are several parameters that need to be defined in order to successfully use on PSO and GCPSO to solve a given problem.