Research On Logstics Path Planning And Frequent Path Mining Based On Internet Of Things

As the popularization of Internet of Things (IoT), IoT is widely used in many fields, one of them is logistics. In the field of logistics, path planning influences the expense and time cost of logistics directly. As the things in logistics containing real time information of spatial, it makes real time path planning and mining frequent path from logistics data possible. This paper works mainly in two aspects.First, for the costs of paths of logistics networks always varying with time, this paper proposes a time-dependent networks model, and then studies the real time logistics path planning problem based on it. For the predict accuracy decreases with time in reality, and when the predict level is low it is hard to get a good path planning result, we propose a minimum-time path algorithm with arc cost predict level parameter(SWPL) and a real time solution with step by step strategy based on SWPL. This algorithm takes predict accuracy into consideration, introduces an arc cost compute method based on the traditional Dijkstra algorithm. The experimental result shows that when the predict level is high or when the predict level is low but adapted the step by step solution, we can get a good path planning result.Second, logistics based on IoT brings out a lot of moving data which contains spatial information. And a lot of knowledge which contains in these data can improve the scientific management of logistics. And now, data mining is the main method to get some knowledge from these data. Frequent path as one of the important knowledge can provide important reference information for optimizing logistics path planning and studying the variation of logistics and so on. The frequent paths are gained by frequent sequence mining, so according to the feature of logistics and logistics networks, this paper provides a frequent path sequence mining algorithm PMWTI, which takes the topological information of logistics networks into consideration. A cost tolerable degree pruning method used for the deep pruning of candidate path sequences is introduced in this algorithm. This method discards some candidate path sequences which couldn’t be frequent path sequences, so it can downscale the candidate path sequences. The experimental result shows that compared with the same algorithm which not adopts this pruning method, PMWTI has better mining efficiency.The work of this paper can support the scientific logistics management, the method provided can be used in real time logistics path planning, logistics routing optimizing, logistics variation discovering and so on.