Research On Service Network Optimization For China Railway Express Based On Customers’ Selection Behaviours

As an effective complement and strong competitor to maritime shipping,China Railway Express(CR Express)is playing an important role in promoting economic and trade cooperation between China and countries along the belt and road,promoting the implementation of the BRI strategy,and ensuring the stability of the global supply chain to adapt to the changes of domestic and international policy and economy.CR Express has developed rapidly recently but is also faced with a series of troubles at the same time,including vicious internal competition,extensive transport operations,and weak awareness of logistics services.As an important method to improve the operation level of CR Express,it is significant to both theory and practice,when it comes to service network optimization.It is beneficial to CR Express to improve the logistics service quality,reduce operation costs and increase efficiency.Based on the above background,this dissertation establishes a logistics services network between China and Europe,the skeleton of which consists of three railway corridors(the Eastern,Central,and Western corridors of CR Express),and two maritime shipping lines(the East and South China Sea Silk Routes).It also includes the railway,highway,and inland waterway services for the “first mile” and “last mile” transport.This dissertation refines the service selection determination to a specific departure,and then works on the service network optimization based on customers’ selection behaviours.(1)Construction of China-Europe international logistics service network under the BRITo make a clear production plan,customers usually specify logistics service selections to specific departures according to investigations.Thus,departure selections are introduced into logistics services selections,where logistics services are distinguished by route,transport cost,departure date,service capacity,carbon emission,etc.,making up an alternative logistics services set.Considering the situation where flows use the same services and services share the common transport resource,this dissertation establishes a time-space network where one physical arc corresponds to multiple services and multiple flows,laying the studying foundation for the full dissertation.(2)Research on customers’ selection behaviours in the China-Europe international logistics service marketFrom the aspect of customers,this dissertation proposed a Capacitated Multi-Commodity Flow Service Selection Problem.It takes the individual operation situation of CR Express and local transport market competitions into consideration.A mixed-integer programming model is established to minimize the sum of transport cost,in-transit inventory cost,and carbon emission cost.Then,this model is solved by CPLEX for multiple tests.The collection zone for CR Express and maritime shipping is obtained,as well as a boundary belt.Finally,more interesting results are found in the sensitivity analyses of the due date,unit in-transit inventory cost and carbon rate.It is worthwhile for customers to adjust due dates to save logistics costs.When the in-transit inventory cost and carbon rate fluctuate,it is also a good choice to change the departure date or nodes besides changing transport mode.The service plan that chooses CR Express as the main haul could generate less carbon dioxide than maritime.(3)Operation plan optimization of CR Express based on customers’ selection behavioursTo optimize the monthly operation plans of regional CR Express trains,this dissertation establishes a bi-level programming model that takes the outline schedule of China State Railway Group as the guide.It considers the impact of customers’ selection behaviours and the demand for trains’ departure concentration(on some days of a week)comprehensively,and also allows to split flows.Its upper-level objective is maximizing the operation profits of regional railway companies,and the lower-level objective is minimizing the total logistics cost of customers.Then,a modified label correcting algorithm and a simplification algorithm are proposed to generate alternative service plans for each order.The bi-level model is reformulated to a single-level model by the KKT condition and then solved by CPLEX.Finally,a real case of CR Express(Wuhan-Hamburg)is studied.It finds that it could be better for customers to select other departures at current nodes than to move to other departure nodes when a capacity shortage happens.It helps to increase profits for railway operators when optimizing the departure dates of CR Express trains,extending collection time appropriately,and allowing customers to split orders.(4)Service network optimization of CR Express based on trains’ departure coordination of multiple regionsTo optimize the outline schedule of CR Express from a nationwide coordination aspect,this dissertation puts forward a method to calculate the utilization equilibrium of transport capacity based on the operation plan optimization results of CR Express that consider customers’ selection behaviours.This dissertation also studies the dual impacts of loading work at border stations and the trunk-branching pattern on service network design.A monthly outline schedule of CR Express is obtained by solving the bi-level programming model,the uppler-level objective of which is maximizing the operation profits,and the lower-level objective of which is minimizing the total logistics cost of customers.An adaptive genetic algorithm(AGA)is proposed and works with CPLEX to solve the upper and lower model interactively.The AGA introduces the operators’ weight updating method of Adaptive Large Neighborhood Search Algorithm to Genetic Algorithm,and considers the substitution of similar services when adjusting outline schedules.Finally,this dissertation takes the Western Corridor as an example to verify the effectiveness of this model and algorithm.It is proved that customers usually select nearby departure stations or ports.The common section is the bottleneck of the CR Express service network,and China State Railway Group will give priority to long-haul trains to increase profit.