Extended individual memory based multi-objective particle swarm optimization

To deal with the problem of diversity distribution of the solutions in Multi-objective Particle Swarm Optimization (MOPSO), a diversity pbest based multi-objective particle swarm optimization algorithm (dp-MOPSO) is proposed. In dp-MOPSO, an individual memory is allocated to each particle for saving the non-dominated pbest set which is found in the searching process, avoiding the loss of the information. An external archive is used to save all the Pareto optimal solutions, and the dynamic neighborhood strategy is introduced to select the global optimal solution from the external archive. Several multi-objective benchmark functions are tested for comparing the performance of dp-MOPSO with two famous multi-objective evolutionary algorithm m-DNPSO and SPEA2. The results show that all the proximate Pareto optimal solutions produced by dp-MOPSO converge to the true Pareto front more closely, and also are well-distributed.