A Product Design Optimization Approach For Railway Door-to-Door Freight In Cloud Environment

Recent economic development has led to profound changes in the structure of transport demand,primarily reflected in the dramatic decrease in bulk cargo demand.The increasing proportion and higher status of high value-added goods in the freight market,as well as the diversity in customers' demand has new requirements for the freight rail service.The logistics strategy is comprehensively promoted in railway systems,so it is important to expand the demand market,exploit the product diversity and promote the service quality,especially offer services through the whole transition process.Under such circumstances,the railway door-to-door express frieght service is born.Railway door-to-door freight products are the important carrier and main form of the railway door-to-door express frieght service,which will attract cargoes and lead a promotion in the freight market shares,thus enhancing the influence of railway transportation.In the new era of "Internet Plus",cloud computing and big data technology are applied in our lives and work,which have a profound impact on a variety of industries.While such techniques are introduced to the conventional freight rail operation,the efficiency will highly be increased,and the traditional modes will be shifted.Therefore,this dissertation combines cloud computing and big data technology,and studies the design method for railway door-to-door express frieght products,which are mainly as follows:(1)The design theory of railway door-to-door express frieght products is expounded in this dissertation,background,principle and process of railway door-to-door freight products based on the cargo transportation planning process are also defined.Two kinds of the railway door-to-door freight products organization modes based on cloud platform are presented,and the economic benefits of these two modes are analyzed.The application of cloud technology,the operational modes,the function module,the framework and operating income of railway door-to-door freight products are studied too.(2)A method of constructing railway door-to-door freight products service network based on cloud clustering algorithm and "Hub-and-Spoke" network model is presented in this dissertation.The evaluation method of the railway door-to-door freight products service network is designed based on support vector regression machine(SVM).The traditional "Hub-and-Spoke" model is improved as considering the time restriction of delivery,so that a mathematical optimization model of the railway door-to-door freight products service network is proposed,in which the calculation parameters of big data in train performance operation are used to replace the parameter setting in traditional optimization method.(3)The flow assignment problem of the railway door-to-door freight products service network is studied in this dissertation.The method of node splitting is used to describe the operation links between railway trunk transportation,normal railway and road transportation,which is helpful to refine the research on the whole process of railway door-to-door transportation.The mathematical optimization model of the traffic distribution of the door-to-door freight product service network is constructed based on the improvement of the traditional network flow assignment model.A K-shortest path algorithm on the basis of the space-time service network is designed to generate a candidate set of the freight flow,and a solution framework by cloud computing is proposed.With the help of cloud resources and the optimization algorithm solver in IBM ILOG Cplex,the problems are solved by IBM DOcplexcloud software.(4)The strategy of generating a candidate set of railway door-to-door freight products is put forward in this dissertation.The constraints of railway trunk line transportation service and highway branch transportation service has been considered,the mathematical optimization model of railway door-to-door freight products design is constructed.Five steps for the framework of railway door-to-door freight products are established,including data preparation stage.data analysis stage,service network construction stage and candidate set generation stage.The running time of the arc section of the service network is calculated by big data technique according to the real-world case,which is the input for the model.Finally,the model can be solved by IBM DOcplexcloud software.(5)A case study based on the section from Wuhan to Jiangcun in the southern section of the Beijing-Guangzhou Railway Line is analyzed.The practical value and significance of the research are explained.