A Research Of Automatic Data Collecting Technology Base On Compilation Technique

Automatic Fare Collection (AFC) system is an important part of integrated Urban Rail Transit Automation System, which is also a core system in the whole process of rail transit operation. AFC system generally includes 5 layers: ticket, equipment, station computer, line computer and clearing centre. Data Collection and Processing (DCP) is one of the core parts in station computer and line computer level which collects data from equipment, decodes data and saves them into database. These data are source of report and clearing which makes Data Collection to be very important. Its development and maintenance go through the whole life cycle of AFC development.First of all, the paper generally introduces concept, history and development of AFC system. Then it briefly depicts the topic from AFC system, central computer system to DCP itself and raises the issues on development, deployment, portability and accuracy of current DCP module. Thus, the paper introduces XDR, a standard for the description and encoding of data between diverse machines. XDR data format has been wildly used in diverse fields globally. It only has, however, a short history on rail transit. Harnessing its characteristic of portability and standardization, many AFC integrators start to use XDR. But it is demonstrated by the experience of using XDR that the change to data content and format is the origin of low efficiency of integration, disrupt of data accuracy and low efficiency of development.This research addresses an automatic solution of DCP based on compilation technique. This solution can improve reusability of the whole system development and productivity, prevent developers from handling endless change request, and reduce cost.The core of this solution is automatic analysis and decoding of AFC data described by XDR. But the paper uses object oriented analysis and design to develop an industrial version of full functional DCP. First, it gives a whole functional and non-functional requirements list based on past project experiences and requirement of the topic which has already covered the set of requirements of a general DCP. Then it breaks the whole package to sub modules and writes detailed use cases for key concepts of this paper which are parsing of XDR data structure description and decoding of XDR binary data. Afterwards, it analyses and designs the system using OOAD methodology, especially, the solution of lexical analysis and syntax analysis of XDR language. It describes implementation of major parts of this solution: XDR parser, XDR decoder, Data saving and Data verifier and designs the test platform for performance analysis at the end. The solution was implemented and validated by the test platform aforementioned and current undergoing project testing platform, so as the usability and feasibility.