PetroV软件架构设计中的一些思考与实现

定位于突破国外跨国石油公司的技术壁垒,并能够被作为唯一平台、长期应用于中国石化的年度勘探部署与优选决策,基于最新的、面向不同层次的软件架构设计原则和软件重用技术,“油气资源一体化定量评价软件”(PetroV)的软件架构设计做了一些积极的尝试和探索:1以多种事件响应模型和并发控制模型为代表的、面向服务的领域基础中间件设计,充分体现数据集成、应用服务协作、分布式面向对象计算与分布式数据存储的架构要求;2基于ST-based KIDA(Spatial and Temporal-based Knowledge,Information,Data and Activity)数据建模思路而封装的领域数据模型,在充分考虑时间、空间维度信息的基础上,抽象并约定了基于组合模式的领域对象模型,有助于以元数据管理(归类、过滤)模式实现库表结构的设计、大数据量存储与数据交换;3以通用工具箱接口和远程过程调用接口为代表的不同层次二次开发接口的设计,支撑不同开发环境下对后台不同服务的调用,及不同客户端应用软件框架的快速开发和集成;4由抽象接口类、工具箱类及相互之间的应用协作组成的应用软件框架设计,通过接口反转技术,能够最大程度降低不同专业软件的开发工作量.PetroV在中国石化的深入推广从侧面证明,合理的软件架构设计方法或软件重用技术是勘探开发类专业软件平台迎合大量不确定性业务需求并确保软件品质的根本前提.

Some ideas in PetroV architecture design and development

Aimed at breaking through technical barriers of foreign international oil companies, PetroV, a quantitative evaluation system of oil and gas resources, is considered as the only platform for the long-term annual deployment of exploration and optimization decision-making for SINOPEC. Based on the latest and different level-oriented software architecture design principles and software reuse technology, we make positive attempts in PetroV software architecture design, which include the followings: A. With various types of incident response models and concurrency control models, middleware design in service-oriented field meets fully the requirements of its data integration, application service collaboration, object-oriented distributed computing and distributed data storage architecture; B. Encapsulation of domain data on ST-based KIDA model defines composite patterns on domain objects with timporal and spatial dimension information, which will be helpful for database table structure design, massive data storage, and data exchange; C. The application development interface design in different levels supports different background services calls and different client applications quick development & integration; D. Through interface inversion technology, applications design composed by abstract interface, toolbox and collaboration between them can minimize the workload of different professional software development. Current applications of PetroV to annual exploration deployment and optimization decision at SINOPEC demonstrate that its software architecture design and software reuse technology are key strength to cater for various domain requirements.