Premature convergence in particle swarm optimization (PSO) algorithm usually leads to gaining local optimum and preventing from surveying those regions of solution space which have optimal points in. In this paper, by applying special mechanisms, suitable regions were detected and then swarm was guided to them by dispersing part of particles on proper times. This process is called dynamic swarm dispersion in PSO (DSDPSO) algorithm. In order to specify the proper times and to rein the evolutionary process alternating between exploring and exploiting behaviors, we used a diversity measuring approach and implemented the dispersion mechanism. To promote the performance of DSDPSO algorithm, three different policies including particle relocation, velocity settings of dispersed particles and parameters setting were applied. We compared the promoted algorithm with similar new approaches and according to the numerical results, the proposed algorithm outperformed the basic GPSO, LPSO, DMS-PSO, CLPSO and APSO in most of the 12 standard benchmark problems with different properties taken in this study.