摘要
This article studies the pod layout problem in the Kiva mobile fulfillment system which adopts the synchronized zoning strategy. An integer programming model for the pod layout problem is formulated under the premise of knowing the relationship of the pods and items. A three-stage algorithm is proposed based on the Spectral Clustering algorithm. Firstly, the pod similarity matrix and the Laplacian matrix are constructed according to the relationship of the pods and items. Secondly, the pods are clustered by the Spectral Clustering algorithm and assigned to each zone based on the cluster results. Finally, the exact locations of pods in each zone are determined by the historical retrieval frequency of items, using the real data of a large-scale Kiva mobile fulfillment system to simulate and calculate the order picking efficiency before and after the adjustment of the pod layout. The results showed that the pod layout using synchronized zoning strategy can effectively improve the picking efficiency.
This article studies the pod layout problem in the Kiva mobile fulfillment system which adopts the synchronized zoning strategy. An integer programming model for the pod layout problem is formulated under the premise of knowing the relationship of the pods and items. A three-stage algorithm is proposed based on the Spectral Clustering algorithm. Firstly, the pod similarity matrix and the Laplacian matrix are constructed according to the relationship of the pods and items. Secondly, the pods are clustered by the Spectral Clustering algorithm and assigned to each zone based on the cluster results. Finally, the exact locations of pods in each zone are determined by the historical retrieval frequency of items, using the real data of a large-scale Kiva mobile fulfillment system to simulate and calculate the order picking efficiency before and after the adjustment of the pod layout. The results showed that the pod layout using synchronized zoning strategy can effectively improve the picking efficiency.
