Systematic Space-time Route Planning Approach Coordinated With Task Assignment For Autonomous Delivery Vehicle Fleet

Autonomous delivery vehicle(ADV)has the functions of environment awareness,precise positioning,route planning and control,which can fulfill the delivery tasks of given service requirements.However,due to the limitation of capacity and range,single ADV cannot fulfill the huge task demand.At this time,it is necessary to form a fleet of multiple ADVs that cooperate with each other to complete the distribution task together.Under the premise of satisfying a certain successful service rate,the goods will be delivered to the designated place at the specified time,and the total driving distance of the ADV fleet will be the shortest.This paper takes express terminal distribution as the research scene,and takes the realization of ADV fleet spatial and temporal route optimization as the research objective.The paper discusses the relationship between the number of vehicles in the fleet,the task allocation of the fleet and the spatial and temporal planning of the fleet and their roles in the optimization of the spatial and temporal routes of the fleet,and puts forward the optimization methods in turn.At build Load-Space-Time networks,on the basis of the three-dimensional network of time and space,put forward the team number of vehicles,the team task allocation and route planning model,and specific design to balance the efficiency and accuracy of the algorithm to solve the model,the final design simulation experiment on the basis of the principle of the traffic system simulation ADV the team in a real scenario the optimal time and space distribution of route.The number of vehicles in the fleet,the task allocation of the fleet and the route planning of the fleet are the three necessary stages for the distribution center to introduce ADV and complete the distribution task.The number of vehicles in the fleet is the premise,which needs to be configured according to the relatively stable daily task requirements.Task allocation of the fleet is the basis,which needs to be carried out within the range of the number of vehicles in the fleet and provide solutions for route planning.Fleet route planning is the goal.According to the task assignment plan,the fleet route is planned and the route is found to make the ADV fleet travel distance the shortest.Terminal distribution has timeliness,and the setting of service length,waiting time and time window affects the result of route planning of the fleet.The service time is set according to the walking time of customers,and used as the probability of successful service;The waiting time should be moderate,too long will increase the fleet size,too short will affect the overall distribution efficiency of the ADV fleet;Time window can be set by the method of equal length of time.The settings of the three should be adapted to the service scope and the number of task requirements of the distribution center.The approximate algorithm and tabu algorithm designed in this paper can be applied to different scenarios.When the number of task requirements is too large,tabu algorithm is considered,which is efficient.When the number of task demands is moderate,approximate algorithm is considered to alleviate the contradiction between efficiency and accuracy.If the scale is very small,an accurate algorithm can be used.