自动售检票ZLC系统实时计算中间件的设计与实现

Design and Construction of Real-Time Computing Middleware for an Automatic Fare Collection Zone Line Center System

针对AFC系统中的实时通信处理系统,结合国内各城市AFC线网标准的差异性,探讨区域线路中心(ZLC)系统实时计算中间件Panda-Adapter系统的设计与实现。Panda-Adapter系统将容器技术、并行处理、内存计算技术、内存数据库存储技术(Redis)有效结合,通过语义识别Adapter计算引擎解决区域中心处理高并发、实时计算及存储的技术瓶颈,其可持久化的技术特点保障了业务的连续性,热插拔的设计模式降低模块运行时的故障干扰,保证系统的稳定性。高并发、高可靠及可跨平台的软件架构设计满足国内不同AFC标准体系的运营需求,为后续多线路中心AFC通信系统改造、升级和扩容提供一种新的设计思路。

This study focuses on the design and construction of real-time computing middleware(Panda-Adapter)for zone line centers(ZLCs),given the real-time processing communication system in the Automatic Fare Collection(AFC)system and the differences in the AFC standards in different cities in China.A Panda-Adapter system effectively combines container,memory computing,and memory database storage technologies with parallel processing.It solves the technical bottleneck issues that ZLCs have,processing high concurrent,real-time computing and managing storage,using a semantic recognition engine(adapter calculating engine).Its technical persistence ensures continuity of operation while the hot-swap design mode reduces the fault interference when the module runs and guarantees the stability of the system.The high concurrency,high reliability,and cross-platform software architecture design meets the day-to・day operational requirements of different AFC standards in China,and provides a new design idea for the future improvement of the AFC communications system towards the multi-line center.