Research On Optimization Of Logistics Distribution Route For Manned-automated Hybrid Fleet

With the implementation of national-level strategies such as my country’s new urbanization and urban agglomerations,the urbanization rate of my country’s permanent population continues to grow,and it has broken through by the end of 2020.Urban logistics is not only the "last mile" of the logistics supply chain,but also a basic industry that supports urban economic development,and an important project to ensure the daily life of urban residents.Promoting the healthy and high-quality development of urban logistics is important to improve urban competitiveness and economic quality.way.At present,my country’s urban logistics is still a labor-intensive industry.The increasing labor cost has greatly reduced the growth space of urban logistics and distribution enterprises.The continuous breakthroughs and the maturity of L4 level unmanned vehicles have a broad field of urban logistics.Application prospects.As an important part of urban logistics,distribution route planning plays an important role in reducing distribution costs and improving distribution efficiency.Therefore,this article will explore the scenario where L4 level unmanned delivery vehicles can provide unmanned logistics and distribution services in some urban areas permitted by national laws and supported by digital infrastructure,consisting of manned delivery vehicles and L4 level unmanned delivery vehicles.The distribution route optimization problem of mixed fleet,the main research content is:(1) Considering the characteristics of AVs that can only provide unmanned logistics and distribution services in the AV area,this paper proposes a two-level distribution model for manned-unmanned hybrid fleets.This mode requires the establishment of transfer stations in the AV area.MVs are mainly responsible for the delivery of goods from The central transportation to the transfer station serves as a bridge and complements the provision of logistics and distribution services to customer demand points outside the AV area.AVs are responsible for the logistics and distribution services within the AV area.A two-level distribution route optimization model for an automatic-manual driving hybrid fleet with vehicle capacity constraints is established.The model takes the minimization of the total distribution time of MVs and AVs as the optimization goal.Aiming at this mixed integer programming model,an adaptive neighborhood search algorithm including neighborhood exchange operations,neighborhood insertion operations,and neighborhood reversal operations are designed.Finally,based on Solomon’s R101 standard calculation example data,cases of 30 customer demand points and 60 customer demand points were constructed for data analysis.Experimental results show that the adaptive neighborhood search algorithm designed in this paper can effectively solve the manned-unmanned hybrid fleet distribution routing problem;in the two-level distribution mode,AVs can greatly reduce the working time of MVs drivers,thereby greatly reducing logistics Distribution costs,improve the efficiency of logistics and distribution.(2)Taking into account the tidal characteristics of the distribution requirements of some AV areas in time distribution,the two-level distribution mode requires at least one AVs in each AV area.AVs in different AV areas cannot be used together,which is likely to cause a waste of AVs capacity.Therefore,this article proposes a manned-unmanned hybrid fleet collaborative distribution model,which requires MVs as the pilot vehicle and the following vehicle as the AVs to start from the distribution center and provide the customer demand points outside the AV area in a formation.Logistics and distribution services.After reaching the first customer demand point in the AV area,MVs and AVs are released from the formation,and the AVs separately provide logistics and distribution services to the customer demand points in the AV area,and then re-formation at the last customer demand point in the AV area and continue.Logistics distribution service,and finally back to the distribution center.Aiming at the problem of coordinated delivery route optimization for manned-unmanned hybrid fleets with capacity constraints,a mixed integer programming model with the shortest working time of delivery personnel as the goal is established.In terms of algorithm design,in view of the large number of variables and complex constraints of the manned-unmanned hybrid fleet collaborative delivery route optimization problem,the difficulty of solving the problem has increased sharply.In this paper,the k-means clustering method is embedded in the adaptive neighborhood search algorithm,and a The large-scale manned-unmanned hybrid fleet coordinated delivery route problem is transformed into the k manned-unmanned hybrid fleet coordinated delivery TSP route problem.The neighborhood structure includes exchange neighborhood operations,reverse neighborhood operations,intra-cluster neighborhood exchange operations,and connection point update operations.The experimental results of the calculation example prove that the adaptive clustering neighborhood search algorithm designed in this paper can effectively solve the problem of manned-unmanned hybrid vehicle fleet cooperative distribution,and the algorithm convergence speed is significantly better than the adaptive neighborhood search algorithm;manned-unmanned The collaborative delivery model of driving a mixed fleet can greatly reduce the working time of the delivery staff when the number of AVs is small.