Abstract: Particle Swarm Optimization algorithm (PSO) is a population-based stochastic optimization techniques.In this paper, the EDA algorithm and PSO algorithm combine to form a new improved algorithm (EPSO) .The EPSO algorithms can effectively global statistics and global optimal solution used in the solution space search, the EPSO algorithm for solving the multidimensional knapsack problem.Experimental results show that solving the multidimensional knapsack problem, the EPSO is better than the traditional PSO algorithm, as well as a variety of heuristic intelligent algorithms, and at the same time, the EPSO Algorithm using fewer parameters, and therefore easy to implement, run more stable.