Research On Vehicle Route Planning In Unmanned Logistics

The development of Internet technology has promoted the development of ecommerce and logistics.With the rapid growth of business volume,the logistics industry has undergone tremendous changes.In recent years,domestic logistics companies have continued to expand their scale and increase investment in basic equipment and manpower.With the surge in business volume,some problems in the logistics process have also emerged.Especially for the distribution of small areas such as communities and closed industrial park,due to the scattered location of customers,the delivery staff needs to go to the designated location to deliver or pick up the order,which not only inefficient delivery but also waste human resources.With the development of self-driving technology,many companies at home and abroad have begun to use unmanned equipment for logistics distribution.Although initial results have been achieved,related technologies need to be further improved.The use of unmanned equipment instead of delivery staff for delivery services can not only improve delivery efficiency,but also save logistics costs.This paper mainly studies the problem of vehicle route planning in unmanned logistics.Path planning as the core technology in unmanned logistics is the main method to solve the problem of logistics transportation efficiency.An effective path planning strategy can give full play to the intelligent and efficient characteristics of unmanned vehicles.The main work of this article includes:(1)To study the existing path planning algorithms,the heuristic algorithm currently commonly used in path planning is difficult to solve large-scale path optimization problems in real time.This kind of algorithm needs to solve every instance of the problem,and the algorithm needs to be redesigned when the scene changes.Considering the realtime and high efficiency of unmanned logistics path planning,the deep learning method is studied.(2)Adopt deep learning model for path planning.In this paper,the problem instance is solved by the network model of encoder-decoder based on attention mechanism.The structure of the encoder-decoder network model has been improved.Compared with the traditional heuristic algorithm that can only solve a single instance and takes a long time,the trained network model can quickly give a path planning solution.(3)Constructed a small area logistics scenario,and carried out simulation experiments based on the model in this paper.The model is tested by different training methods and different parameters,and the experimental results of this model are compared with commonly used path planning algorithms.Through comparison,it is found that the path planning algorithm in this paper can quickly give a better solution,which can improve the real-time nature of path planning and reduce logistics consumption.The method proposed in this paper can improve the real-time and high efficiency of path planning,and has practical significance and practical value for vehicle path planning and scheduling in unmanned logistics.