Aiming at the premature convergence problem in discrete particle swarm optimization algorithm，a novel cellular particle swarm optimization algorithm is proposed，which is based on the principles of cellular automata and discrete particle swarm optimization algorithm． Cellular and its neighbor are introduced into the algorithm to maintain the swarm’s diversity and the algorithm uses evolutionary rule of cellular in local optimization to avoid local optima． Simulated tests of multi-dimensional knapsack problem and comparisons with other algorithms show the algorithm is feasible and effective and the algorithm has strong global optimization ability