K-Means based path optimization for smart warehouse system

Abstract

In order to improve picking efficiency in the "Part-to-picker" warehouse system, the path optimization for smart warehousing of small commodity is investigated by combing with item classification in this paper. First of all, a mathematical model of the global environment for a small commodity warehouse with a robot that can pick multiple items at one time is established by using grid method. Secondly, the order picking strategy is studied from the perspective of items classification based on K-Means clustering algorithm, and the rule of cluster items is defined. Then, the ant colony optimization is used to plan the optimal path for picking up goods based on the item classification. Finally, the effectiveness of the proposed approach was demonstrated through simulations, compared with the traditional order-based sequential picking strategy, the modified picking strategy which fusing item clustering and path planning could improve the system's outbound efficiency about 28.1%.