This paper reviews the literature on the mechanisms for adapting parameters of particle swarm optimization (PSO) algorithm. The discussion focuses on the mechanisms for adaptively setting such parameters as inertia weight, acceleration constants, number of particles and number of iterations. Two mechanisms are proposed and tested. The velocity index pattern is proposed for adapting the inertia weight while the acceleration constants are adapted via the use of relative gaps between various learning terms and the best objective function values. The mechanisms are demonstrated by modifying GLNPSO for a specific optimization problem, namely, the vehicle routing problem (VRP). The preliminary experiment indicates that the addition of the proposed adaptive mechanisms can provide good algorithm performance in terms of solution quality with a slightly slower computational time.