The philosophy behind the original particle swarm optimization(PSO)is to learn from individual’s own experience and the best individual’s experience in the whole swarm.Estimation of distribution algorithms(EDAs)generate new solutions from a probability model which characterizes the distribution of the current promising solutions in the search space.A novel discrete particle swarm optimization algorithm based on estimation of distribution(EDPSO)is proposed by reasonably combining the ideas of PSO and EDAs.The proposed algorithm breaks the confine of the original speed and location model

and each particle in the population have comprehensive learning ability.Therefore the proposed algorithm effectively extends the PSO to solve combinatorial optimization problems.Simulation results show that the proposed algorithm has superior performance to other discrete PSOs.