Research Of Rail Transit Automatic Fare Collection System Based On Private Cloud Platform

The urban rail transit automatic fare collection(AFC)system is an important part of the urban rail transit which realizes ticket selling,ticket checking,fare change,clearing settlement,operation management and so on.Using the emerging cloud computing technology to optimize AFC system and construct a rail transit AFC system based on private cloud platform in the operation of networks contributes to solving massive data computing and storage.Above all,the assumption of integrating cloud platform and AFC system was proposed in this paper according to the current research on cloud computing technology and AFC system optimization both at home and abroad.Based on a clear knowledge related to the definition,characteristics,key technology and deployment model of cloud platform,the functional and non-functional needs of AFC system based on cloud platform were also analyzed.Private cloud platform deployment model was chosen for AFC system through synthetical consideration.A more concise AFC system architecture was discussed from the logical architecture,physical topology and software architecture.Private cloud platform which consisted of infrastructure layer,cloud platform middleware,software layer and cloud security mechanism was used to replace ACC and LC,and resources for data storage and computation in SC will not be set.Afterwards,the virtualization,data management and resource scheduling technology for cloud platform were explored.There are two methods of creating virtual machine based on hardware and software.Data management technology involves file management,data storage and distributed programming model.Load balancing algorithm is the key of resource scheduling.The algorithm based on polling policy and greedy policy were studied for the aim of task number and execution time separately.On this basis,an algorithm based on dynamic balancing policy was proposed considering the server resource utilization.In this algorithm,the greedy policy was used to accomplish the mapping between the task and the virtual machine in order to form the initial solution.Then the CPU,memory and bandwidth utilization of nodes were periodically monitored to implement dynamic migration of virtual machines.A simulation cloud platform for algorithm testing was established with CloudSim software,and the performance of algorithm was evaluated according to the task execution time and load balance degree.The result shows that the execution time with dynamic algorithm(dynamic balancing policy)is shorter,and the load of the virtual machine is more balanced.Finally,the system implementation scheme was elaborated so that in order to improve the efficiency of social capital reduce maintenance costs and improve the efficiency of operation management.The research provides a reference for the design,optimization and adjustment of urban rail transit AFC system in our country.