Abstract: An optimization model of picking location allocation is established for an automatic picking system. The picking time model is used as the fitness function, and the genetic algorithm is improved by adding a penalty function to judge the Heming distance. The simulation results show that the total picking time is obviously reduced by 26.25% after the optimization of the picking location. The results show that the distribution of goods picking position has a great influence on the efficiency of automatic picking system. In order to reduce the total picking time and improve the efficiency of automatic picking system, not only the location distribution of individual goods, but also the overall direction should be considered. This research has strong practical significance and provides theoretical support for improving the efficiency of the automatic picking system in distribution centers.