Design And Implementation Of Model Driven Software Process Big Data Acquisition And Management System

A large amount of valuable data will be produced in the process of software development,but it is difficult to store and use.At the same time,the process data generated by the development of historical projects is of great significance to the development of related software in the future.Therefore,the quality and reliability of software can be greatly improved by using an effective method to collect process data such as software requirement design,development design and development test,and combining with specific model analysis methods.But there is a lack of effective systematic software process methods and tools to obtain process big data in the software life cycle.In view of the model-driven software of high reliability software such as aerospace in China In the background that the model-driven software development method has been greatly recognized by academia and industry,the graphical model-driven software development has gradually become the mainstream tool of high-reliable software quality assurance.And based on this technology can effectively collect software process data,further combined with cloud computing technology and software quality measurement method can greatly support the evolution of CMMI model.This topic carries on the research work from the following aspects:Firstly,the paper makes a comprehensive analysis of the related support system and technology,and gives the characteristics of the interface analysis of the related unit structure of the system extension design.Secondly,according to the stage of the software life cycle supported by the model-driven engineering tools developed by the research group and the graphic system features contained in each stage,three kinds of software process big data are analyzed,summarized and refined and mapped into various attributes and function sets supported by the model tools.Then,the paper designs the attributes and contents of the software engineering document template and the document,and designs the cloud storage format and upload mode of the data in the document combined with the modeling data.According to the requirement of automatic generation of model-based code in the system,the conversion function of graphic model to XML document is designed.In addition,the model-like big data is analyzed and tested based on the given sample data using all kinds of reliability prediction models,the effectiveness of the system data set,data transmission,data storage and data prediction continuous process is gradually demonstrated.Finally,the requirements of a model-driven software process big data acquisition and analysis system and the detailed design process include the extraction function of operation class,business class and process management class data document,model description document generation function,related document upload and download function,modeler evaluation function and reliability prediction function and finally realized.The validity and correctness of the proposed method,algorithm and system function are proved by case test.