【作者】 盛秀杰；

【导师】 梅廉夫；

【作者基本信息】 中国地质大学 ， 油气田开发工程， 2010， 博士

【摘要】 在我国,甚至世界范围内,经过数十年的高强度的勘探与开发,含油气盆地已进入勘探与开发的高成熟阶段,对油田的投入与产出有着越来越高的要求。统筹兼顾、多学科、多因素综合分析以及最快采用新技术与新方法是获取地下更准确认知、最大化提高采收率的最终有效途径。同时,油气E&P领域不同学科的专家们需要突破常规研究思维,在深入对本学科研究的同时,在跨学科的新技术引入、不同学科团队合作的有效性方面做大胆突破。多年来,围绕有利勘探目标圈定,精细油藏描述等研究课题,油气E&P领域已经在石油地质、勘探地球物理、油藏地球物理、测井解释、钻井工程等方面形成有效的跨学科研究模式,包括相对成熟的研究团队组织模式。但,目前油气E&P领域的综合研究在团队运作和技术手段上还相对落后。如,项目资料仍过多停留在纸介质层面,常见的有课题报告、离散的图件、和诸如数据表格等简单的计算机文件。项目研究则借助于通用的数据处理和图形处理软件来进行,这样的现状难以满足当下勘探与开发的高水准要求,从而也导致了一系列问题：通用数据处理和图形软件均有各自的原始数据格式和并非面向油气勘探与开发的数据处理功能,原始数据的准备一般只能以人工的方式进行,导致基础资料整理工作繁重且容易出错。研究成果和原始数据之间缺乏迅速高效的互动反馈及追踪机制,在研究成果中补充考虑新获取的信息和资料极为困难；多种来源和形式的资料难以有效的集成化管理,研究人员往往无法保证在应用全部资料的基础上得出合理的解释和判断,进而影响最终的研究成果的有效性等。以本文所在课题为例,本文所在的“鄂北杭锦旗地区油气地质综合评价与目标优选”课题对上述问题已经有“很好”的反映,也是目前典型主流研究团队的运作模式。本课题涉及到多门学科的最新研究,以便于该工区新的有利勘探目标的圈定。不管是地化方面的研究(各种地表异常圈定)、构造方面的研究(等时构造层位的建立)、地球物理方面的研究(波阻抗的反演计算)等,课题涉及的不同学科的研究基本上各自独立进行,项目关注的是最终研究结果及基于不同学科研究结果的主观判断与综合分析。显然,尽管上述学科最后分别提供了以不同的图件形式表达的研究结果,但无法实现多学科间研究成果的统一量化表达,进而更快速准确地进行空间对比和分析,同时多学科间缺乏具有实际意义的协同互动研究过程。针对上述问题,本文研究在油气E&P领域与信息科学、地理信息等其它领域结合方面做了一些探索性研究。本文研究最终落脚在信息科学、空间数据管理等最新技术发展在油气E&P领域中的应用,希望一方面把跨领域的最新研究思维和技术成果引领到当下油气E&P领域,更为重要的是以“适当的形式”能够量化、串接、组织起当下多学科的研究过程及成果,达到真正意义上的多学科综合分析效果。目前,“适当的形式”就是本文提出的油气E&P领域基础中间件。本文在领域基础中间件的定义、体系架构设计及基于领域基础中间件的相关应用软件开发等方面探讨了全新的技术路线及有意义的尝试：①在最快新技术与新方法应用方面,需要释放油气E&P领域专家们更多的时间与精力,不在被羁绊于本学科典型应用软件开发,在能够集成既有不同研究成果(也就是典型专业应用软件)基础上,更多推动本学科新算法、新技术的研究；②在推进不同学科间的有效研究流程方面,需要“映射”不同学科的研究活动,定义研究流程中的关键节点及上下文约束,使得不同学科的研究流程具有高度的可组合性及可匹配性；③在不同学科间的基本沟通方面,需要“抽象”油气E&P领域的数据、信息以及研究结果(知识的表现),一方面使得不同学科的研究具有一致的沟通与信息表达基础,同时通过对以往、现今及未来的海量领域数据的建模、存储及基于空间数据进行管理等方面有新的突破；④有了油气E&P领域不同学科的研究活动“代表”(基于通用应用软件框架的专业应用软件实现),符合不同学科的一致性“语言”定义(领域统一数据模型)以及不同学科的一致性研究“过程”定义(系列领域服务的接口表达),接下来需要搭建不同学科的实时沟通平台,形成真正有效的面向油气E&P领域的“智能化”协作平台。因此,围绕上述要点,本文主要做了以下几个方面的研究：(1)理顺了当下信息科学技术最新的软件开发理念(软件重用)及基础软件实现(通用基础中间件),强调了系列软件重用的成果(面向对象封装、不同层次的设计模式、框架等)在形成弹性软件体系结构,快速、健壮软件实现的同时,更为重要的是以系列设计模式为代表的创新性思维模式可以被直接应用到油气E&P领域数据模型设计中,突破了传统的数据建模手段。等同系列操作系统和系列商业数据库,面向分布式异构环境软件开发的基础中间件作为近几年来信息科学技术的最新技术发展,其必将在油气E&P领域信息化的过程中扮演越来越重要的角色。(2)通过深度参与油气E&P领域的典型横向课题,识别、定义领域基础中间件在多学科协作方面覆盖的范畴及需求,顺承通用基础中间件的封装与设计理念,国内第一次比较完整的阐述了面向油气E&P领域的基础中间件定义及特点。定位多学科的有效协作,在数据集成、应用软件集成、面向对象的分布式计算及面向系列领域服务的“智能化”平台四个方面,给出了清晰的技术路线阐述及关键点说明。(3)在数据集成方面,除了强调信息科学技术在软件重用方面的新思维引入到统一领域对象的建模中,给出全新的领域对象间的关系定义外,强调了统一领域数据模型建模中的其它若干关键技术点。如,基于数据、信息、知识驱动的领域实体对象的归类技术,通过“活动”的概念来真实反映现实世界中不同学科的实际研究过程及信息传递。基于海量数据管理的角度,面向油气E&P领域比较系统完整地引入空间数据对象概念,其与领域实体对象、空间索引机制等一起为领域数据的管理提供了最新的技术途径。最后,面向软件开发层面,基于模板技术等给出了部分领域对象的工具箱代码实现,展示了最新软件开发过程中的实践成果。(4)在应用软件集成方面,特别强调了通用应用软件框架的概念。其一方面定位于有助于不同学科专业软件的快速开发外,更为重要的是通过融合相关领域服务的代理类,使系列油气E&P领域的专业软件达到企业级甚至更高级别的软件集成效果。(5)在面向对象的分布式计算方面,更多强调了直接受益于当下满足CORBA规范的最新基础中间件的开源软件。一方面展示了基于CORBA规范的典型分布式应用软件的开发流程,另一方面在异步消息传递、请求与处理的并发处理等方面,给出了一些细节的实现思路。(6)面向系列领域服务的“智能化”平台概念奠定了面向油气E&P领域基础中间件的基本分层体系架构。命名服务、负载服务为代表的跨领域共性服务与基于地震属性的含油气检测为代表的领域服务即插即用于同一软总线,通过面向分布式环境的不同事件驱动等机制,与数据访问服务一起构建了支撑当下最新多学科研究深层次合作的技术方案。(7)以地震属性学的最新研究成果应用为切入点,基于领域基础中间件的设计理念及部分基础实现,在纵向上(相对单一的领域“智能化”服务平台)验证了基于领域基础中间件进行软件快速开发、软件集成、统一领域数据模型的存储与访问的验证。更为关键的是,结合杭锦旗地区的实际课题需要,给出了基于BIOT理论的含油气检测结果,为该工区的最终有利圈闭评价提供必要的佐证。最后,本文的研究过程及成果也希望被看作是对传统石油工程研究范围的一个积极突破,信息科学技术作为当下推动整体经济发展的不可缺少的原动力所在,在地学的学科分类研究中已经有很明显的体现,如地球探测与信息技术等。因此,信息科学技术为代表的其它领域的最新研究成果与油气E&P领域应该有更加明确的交叉研究方向与定义。一方面定位于利用其它领域的最新发展推动油气E&P行业的快速发展,即“资源有限,技术无限”,另一方面也为作为核心产业的油气E&P行业提供更多的民族软件。更多还原

【Abstract】 Over last decades, the oil fields of China, even the word side have been step into the mature phase of exploration and production, which is seeking for the more advanced ways to decrease the maximum ratio between investment and production. Those ways including the consideration of more factors, the combination among disciplines, the systematic analysis driven by many factors, the adoption of new techniques as soon as possible etc are benefiting the oil fields from the more accurate description of subsurface and the higher recovery ratio. In order to push research actions into deeper status and to accelerate the whole research process, it does need the specialists on behalf of varieties of E&P majors to break through the constraints of daily research and to lend more and more the latest techniques from other disciplines, even other domains like information technology, geography information system etc. In fact, domain experts have been work together to define the exploration potential reservoirs then get the detailed description of those reservoirs with cooperative way.However, the current operation and communication ways of kinds of research teams are blocking the cooperative effects. For example, many experts share its idea in paper media such as project report、series of diagrams and excels etc under the help of general word and image process software. Obviously, those different simple and general auxiliary tools don’t easily set up the uniform data format for communication and aren’t exclusively ready for the usage of E&P domains. So, from the preparation of raw data to the last project report there are no any interrupt ways and information tracking mechanisms for effective cooperation among different research activities hold by different domain experts. In addition, researchers lack the ability to make effective and deeper evaluation based on information from other majors since there are no’ "platform" to load all data, information in massive way etc while harness those experts with synthesis analysis.Experienced by ongoing project-"The Integrated Evaluation and Reservoir Location of Hang Jinqi Distract in Hubei Province", this paper makes well conclusions which reflect the status of a typical research team. This project involves many majors in order to break through the current understanding of this district such as petrochemical (for surface abnormity)、structure (for isochronous horizons)、geophysics(for the inversion of lithology)、the attribute analysis of seismic data(for direct indication of petroleum) etc. During the process of those project, different majors are almost being hold standalone and there are no any process interruption between two majors except for the collection of last research results in "paper " formats. So, the integration of different majors often occurs directly in brains not on the world which should set up uniform data expression and operation rules for deeper information integration and research constraints.This paper presents some discusses about the domain combination of petroleum engineering, information technology, geography information system etc in order to set up an effective and practical work flow for accurate description. The ultimate target of this paper is to apply the leading technologies such as software reuse, spatial data model etc into E&P domain. Not only does this paper introduce the latest research methodology and fresh technique into E&P domain study, but also it defines a "mediator" which is enable to abstract integrate the research processes of different majors and lets discipline research activities melt more for solid and intelligent analysis. More, the "mediator" here means the domain middleware specified by this paper. This paper focus on topics such as the definition, the system architecture of E&P domain middleware and the technique road map how to develop domain-specified application based on domain middleware in order to①quickly apply IT techniques and free E&P domain experts on domain research and enable the deep integration of different major software;②effectively order different major research processes, better map between those discipline research activities and its corresponding software, define key point of current research process then set prev-and post-constraints specified by other majors and let the work flow in reasonable and flexible way;③"abstract" the data, information and knowledge of E&P domain in the uniform data model, communicate each other in the same baseline. In addition, the massive information should be modeled and managed in the leading way such as spatial data model and index ways;④set up real time running platform which is plugged by different domain-specific services and common services composing an intelligent environment. Namely, the similar definition of research processes, uniform data model expression and the instantiations of the common domain-specific software framework and those domain-specific services are key points of this cooperative and intelligent environment. For those key points, this paper probes into those topics as below:(1) Based on the leading methodology of software development, namely software reuse and the base implementation such as common distribution middleware, this paper emphasizes the importance of series of software reuse technique such as object-oriented, design patterns in different levels, framework and component, which will make the adaptive architecture and let the development process in fast, robust way. Especially, the inventive thoughts about varieties of design patterns can be directly to analyze the relationships among E&P domain-specific entities and share more heuristic modeling ideas. The same as operating system and common database, common distribution middleware are playing the important role for the development of domain-specific application software.(2) With participation of a typical E&P research project, this paper recognizes and specifies the requirements and research extent of E&P domain-specific base middleware. Inherited from the encapsulation and design ideas of layered structure of common base middleware, this paper contributes the definition of E&P domain-specific base middleware in the first time and concludes its obvious characters. In order to harness the process of cooperative work among different E&P majors, the road map and key introductions of domain-specific middleware cover the topics such as data integration, application software integration, object-oriented distributed computation and service-oriented architecture.(3) This paper not only emphasize the application of heuristic thoughts of software reuse into the definition of uniform data model, but also introduce other key points to fix the issues about data integration. For example, the categorization of domain-specific entities driven by KID model, the "activity" abstraction of different research processes and its information flow. Importantly, it is creatively graft domain-specific entities on spatial data model, which embodies many advanced data management ways of massive data and deeper spatial analysis. Lastly, for the better implementation of domain-specific software, this paper examples part of domain-specific class library, which are designed and implemented by policy-driven design ideas.(4) This paper introduces the concept of common domain-specific application framework for the better integration of application software. Except for the harness of software development, the instantiation common framework easily communicate other instanced ones with adoption of proxy classes of corresponding service which are plugged in the same communication channel.(5) Benefited from the open source which is subject to the latest CORBA specification, this paper examples the general develop process of CORBA application. More, this paper abstracts some detailed implementations of concurrent request and its deal, asynchronous messages and so on.(6) The concept of service-oriented intelligent platform decides the layered structure of domain-specific base middleware. Namely, common services such as naming service, balanced services and domain-specific services such as seismic attribute analysis will be plugged on the same real time communication "software bus" and support the deeper cooperative solution among different majors under the help of asynchronous event-driven mechanism etc.(7) With application of the fresh results of seismic attribute analysis, this paper instances the common software framework driven by domain-specific middleware and verifies the reasonableness of application of domain-specific middleware whether it enhances the development process, get better integration of applications and the persistence ways of uniform data model etc. More importantly, this software as one of participants of typical research team gives its petroleum test indication based on Biot theories.Lastly, the study process and its results of this paper may be taken as the active breakthrough of the traditional petroleum engineering major since IT should stand a place and the cross research should be give clear definition.更多还原