| With the rapid growth of import and export trade,the Rail-river intermodal transport has become the mainstream of the international trade mode of transportation in China.As the carrier and coordinate system of the information of the Rail-river intermodal transport,the level of construction of the combined transport information platform not only directly affects the efficiency and security of the joint logistics transportation,but also determines the quality and service level of the combined transport.In recent years,China's major large-scale port of the Rail-river intermodal transport business information has been basically achieved.However,due to the uneven development of transport information in different regions,the system framework is rigid,degree of information sharing is not high,and lack of data management analysis and other issues,to a large extent restricted the overall development level of intermodal informatization.With the gradual maturity of cloud computing technology,with its advantages such as flexibility,extensibility,resource sharing and high reliability,it provides a new direction to solve these bottlenecks technically and realize the comprehensive transformation of the informatization of the Rail-river intermodal transport.In this thesis,the key problems in the construction,application management,information sharing and big data processing of the combined transport information platform in the cloud environment are mainly carried out in the following aspects:1.The technical framework of the Rail-river intermodal transport cloud platform is analyzed and constructed.In view of the shortcomings of China's current rail-river intermodal transportation information system construction model,the cloud computing technology is introduced into the information construction of the Rail-river intermodal transportation.Based on virtual resource management,hierarchical fusion cloud architecture(HFCA)is proposed,which is divided into three layers: resources,business and data.In the resource layer adopt virtualization technology for pooling resources and providing the business layer IaaS and PaaS fusion virtual resource facilities.In the business layer,the intermodal application is divided into the functional resource pool,and the application reconstruction based on cloud native and the SaaS application service mode based on maturity are established.The data layer is based on the DOA to centralize,store and retrieve the data of the combined transport business of fragmentation,and form a data resource pool,and to excavate and re-use the massive business data through the establishment of scalable big data applications.In this way,the model of "chimney-type" system is broken through the isolation of the port as the center.In addition,by combing through the needs of the combined transport business,the information flow of freight import and export in the cloud environment is optimized.2.An integrated intermodal application management system(IAMS)based on DevOps in cloud environment is established.The problem is that the system is heterogeneous and inefficient.By abstracting the business application into "intermodal transport application unit(ITAPP)" to shield its hardware and software differences,and DevOps is used as the information integration management mode in the cloud environment to unify the delivery and operation and maintenance of the transport application.First of all,the continuous integration build model is designed with the virtual image as the carrier,which avoids the inefficient manual delivery process;secondly,based on the abstract runtime environment(ARE)of the resource support system,an automatic deployment algorithm and a high availability cluster management model based on the "transport application unit" are proposed to improve the operation and maintenance efficiency and service reliability;then,a unified application access control strategy based on SSO and RBAC is established,which realizes the centralized access control management of the transport application pool in multi-tenant environment;finally,the performance benefits of the virtual environment are verified by application deployment testing.3.Based on MSOA,this paper presents a model of intermodal transport information sharing and the self-scaling mechanism of data exchange system.In view of the high cost,heavy structure and low performance of the mainstream information integration technology in the transport industry,this thesis puts forward some solutions.First of all,a lightweight MSA is used instead of the bus-type information sharing architecture to construct a two-layer shared model for micro-service.And using the "intermodal transport service unit(ITSU)" to virtualize the role,interaction and process of information sharing,a decentralized intermodal transport integration unit(ITIU)is established.In this way,multimodal transportation application management and information sharing are combined to realize the sharing of low-cost information on demand.There is one more point,the EDI system is reconstructed by micro service and distributed queue,and the distributed packet concurrency processing system and POD horizontal scalable model can be dynamically extended in K8 S environment.The adaptive component scaling algorithm based on multi-indicator load set and queuing theory is proposed.The POD replica set is automatically quantified by the scaling threshold of real-time load,while taking into account the dynamic load,resource quota and demand difference of the POD,using the improved DRF algorithm to optimize the allocation of POD replica resources,and the throughput of the message data is improved by using the on-demand feature of virtual resources.4.An intelligent order matching system(IOMS)model based on the combined transport big data is established,and the model is solved by using the collaborative filtering algorithm of artificial fish swarm optimization.In the process of the current combined operation,the problem of unreasonable and inefficient matching between shippers and co-operators is unreasonable and inefficiencies.Based on the Spark and Hadoop distributed computing framework,the existing off-line recommendation model is extended,and a three-layer IOMS model based on batch order characteristics,historical business data and behavior feedback data is established.What's more,the collaborative filtering algorithm was optimized by using artificial fish to find the solution to the model.Combined with the order characteristics,carrying capacity constraints and the owner of the behavior of the feedback data on the recommended list of continuous closed-loop adjustment,in the near-line time for the different preferences of the owner to identify a number of orders to meet the needs of the higher acceptable intermodal transport operator.In this thesis,the latest technology in the field of cloud computing is introduced into the construction of molten metal intermodal information.At the different levels of the construction of the Rail-river intermodal transport cloud platform,combined with its business characteristics,the key technologies such as application management,business information sharing and large data processing are respectively technical issues were explored.The feasibility of the construction of the rail-river intermodal transport cloud platform is verified by the experimental analysis of the IOT cloud platform environment of the author's work unit,which can provide theoretical and practical support for the development of China's rail-river intermodal transport information. |
PDF Full Text Request