Abstract: Aiming at the problem of high consumption during picking operations in the B2C e-commerce distribution center's human arrival picking system,this paper proposes a new optimization algorithm for storage location—vassal algorithm by considering the multi-location operation of a single product and the degree of product correlation.The relationship between goods was modeled as a graph based on commodity correlation.The vassal algorithm considered each node of the graph as a single piece of territory.By determining the vassal and Lord of each territory through invasion,each node in the graph was divided into at most two communities based on the vassal and Lord.The algorithm adjusted the number of nodes to the desired number of nodes.We assigned the goods to storage shelf according to the generated communities by using greedy allocation strategy.The experimental results show that when the number of commodity types increases,the optimization degree of the storage allocation scheme generated by the vassal algorithm is improved compared with that of the random strategy.With iteration of 400 rounds,this algorithm is superior to genetic algorithm,simulated annealing algorithm,artificial fish swarm optimization and particle swarm optimization by 3.00％,28.76％,22.03％and 11.42％,and its running time only takes 0.04％～3.85％of other algorithms.