Abstract: Particle swarm algorithm is widely applied in many fields. According to weight convergence range, weight function is established and it is applied to the process of velocity evolvement. In the evolvement process, sub-swarms are adopted. So the algorithm has good global convergence ability in the initial iterative stage and has good local convergence ability in the retral iterative stage. This can realize the balance between global search and local search. Weight particle swarm algorithm is used in registration for scattered point cloud and three dimensional CAD model. It has good registration result and fast convergence when it is compared with the basic particle swarm algorithm.

Key words: computer application, weight particle swarm algorithm, model registration, optimization