The Identification Of Pallet Flow Closed Loop Which Based On Data Mining Algorithm

With the continuous development of logistics industry in China, the optimization of reusable resources represented by pallet pooling is a trend for logistics industry. The establishment of pallet pooling system can save much social logistics cost and contributes a lot to our nature environment. During the promotion of pallet pooling system, the related data base stores a large quantity of pooling information from which much knowledge could be extracted and be used to mine the concrete properties of pallet pooling. The sorted information could be used to support strategies for decision makers.By using relevant rules, the frequent relevant rules which are frequent items set could be found through pallet pooling data. And these data imply pallet information of closed circle trend. By processing these data, the closed circle information could be identified. This kind of information can serve as basis of the optimization of pallet storage and placement of service node. As a consequent, the efficiency of the operation of pallet pooling system increases and the cost decreases.This paper firstly introduced the current research condition for pallet polling and data mining. Then, the concept of closed circle of pallet pooling was proposed. And the Apriori algorithm is defined as the basic algorithm for this paper by comparison with other methods.On considering the low efficiency of the Apriori algorithm, improvement suggestions were proposed as follows:A branch operation is added between the generation of (k-1)th set and the generation of kth set in order to decline the number of the element of the kth set. As a result, the efficiency of the algorithm increases.This paper proposed an identification model of closed circle based on Apriori algorithm and proposed related hypothesis for this model. As a whole, the solution of this model can be divided into three steps:first, mining the second frequent set of closed circle by relevant rules of Apriori; second, identifying the flow direction of pallet of second frequent set; third, identifying the closed circuit by establishing matrix.After the determination of this model, this paper successfully identified two convective closed circles and three circulative closed circles through experiment and proved the efficiency of the improved algorithm.