Research On The Path Planning Of Unmanned Logistics Distribution In The Urban Area

The logistics industry has become a pillar industry for the national economy.While the logistics industry is prospering,the overall low efficiency of the industry is also becoming increasingly prominent.In particular,the distribution of urban small areas takes up a large amount of labor costs and affects the long-term development of the logistics industry.With the maturation of unmanned vehicle technology,it becomes a viable solution for urban small-area distribution to use improved unmanned logistics vehicles to identify environmental information in urban small areas and autonomously plan the route instead of the dispatcher to safely transport the cargo to the designated location.At present,many research institutes at home and abroad are carrying out unmanned logistics distribution experiments in urban areas.However,due to the incompleteness of unmanned driving,there are currently no companies that have officially launched unmanned logistics distribution in the world.The related technologies need further study,particularly as the path planning algorithm,which is a core function.The principle reference of the process,that unmanned logistics delivery vehicle starts from the delivery station and traverses the time-constrained target points to complete the path planning,can be made to the famous traveling salesman problem and traveling salesman problem with time windows.The related path combination optimization algorithm of these principles can be applied to unmanned logistics distribution,but it does not involve the complex road networks and different types of driving environments.Based on the general trend and existing problems of unmanned logistics and distribution applications in urban areas,this paper describes the hardware devices for unmanned logistics distribution and proposes a path planning algorithm combined of vector map and raster map,using genetic algorithm to solve mixed time windows,as the realization mechanism of unmanned logistics distribution path planning process.Through the introduction of hardware devices,paper explains the driving environment of unmanned logistics in small urban areas,including roads and housing estates.Therefore,in the path planning process,a vector-raster combined path planning algorithm is needed to calculate path for the vector road network and the grid map respectively;at the same time,in the actual logistics distribution,the time constraint of the customer to receive the parcel is often random and flexible,so the previous path planning algorithm with time window cannot be completely adopted.Since there are a large number of feasible paths between two distribution target points in the path planning process,the genetic algorithm with parallel operation mechanism and gradual optimization is very suitable for the unmanned logistics distribution path combination optimization process with a certain amount of data.This paper gives a detailed description of unmanned logistics distribution route planning problem and establishes a corresponding mathematical model,then proposes a vector-raster combined path planning algorithm based on genetic algorithm with mixed time window.In this paper,the conditional assumptions of the mathematical model of unmanned logistics distribution route planning are analyzed,the meaning of each parameter in the model is introduced,the objective function including the time and penalty time of the distribution process is established,the constraints of time window,start and end points,and target time are described,and the delivery plan containing each target point is given.In the process of solving the mathematical model,the principle and flow of the mixed time window layer algorithm,the vector-raster combined path planning layer algorithm,and the path combination optimization layer based on the genetic algorithm are designed.Experiments with simulation examples of real geographic coordinates show that the algorithm can get the optimal path of the vector-raster combined unmanned logistics distribution process with mixed time window,and the experimental results and algorithm performance are analyzed.Finally,the paper summarizes the full text research work,explains the innovation of the article,and points out the deficiencies of research,providing reference for subsequent research.