【作者】 王文铭；

【导师】 宋守志； 孙豁然；

【作者基本信息】 东北大学 ， 采矿工程， 2000， 博士

【摘要】 矿山企业信息系统以提高矿山企业生产经营管理水平和经济效益为目标,运用系统科学方法,为矿山企业的数据处理和不同层次的辅助管理与决策提供服务,充分利用矿山生产经营信息,判断矿山生产的运行状态,发现矿山生产经营管理中存在的问题,找出生产经营失调的原因,从而调整和控制矿山的生产行为,保证矿山生产正常高效运行。矿山企业信息系统与矿山工程、矿山管理、计算机技术密切相关。本文结合计算机网络技术,在矿山企业中引入数据仓库技术、知识发现技术来研究和解决矿山企业信息系统建设与应用中的一些问题,进一步发挥信息系统在矿山生产管理中的重要作用。针对矿山生产管理所需基础数据的收集与组织,本文提出以数据为中心建立矿山企业内部网,重点分析了矿山企业内部网的结构和实现技术,并且较好地解决了矿山现有计算机应用系统与应用数据库向矿山企业内部网无缝迁移的方法,最大限度地保护了矿山原有信息系统的投资。实际应用表明,矿山企业内部网可以作为行之有效的矿山企业信息系统框架,为矿山生产管理提供及时、方便、完整的信息服务。随着计算机应用的不断深入,矿山企业已经积累了大量的矿山地质勘探、采掘工程、矿山生产管理等方面的数据,由于不具备对这些数据进行集中存储和管理的能力,目前这些数据资源并没有发挥出其应有的作用。因此,本文提出应在矿山企业信息系统中建立数据仓库,采用多层体系结构,以现有应用系统的数据库为基础,通过相应的矿山主题数据,对数据进行集中抽取、转换,“自下而上”建立数据集市,然后再集成为矿山企业数据仓库,利用数据仓库就容易做到为矿山企业生产经营管理提供丰富、有效的数据支持。在数据仓库的基础上,采用在线分析处理技术对矿山的生产经营管理数据进行多方面和多角度的分析和处理,从而了解矿山生产运营状况。本文分析了在线分析处理技术的结构特点和矿山生产管理数据的多维特性,并在矿山生产实际中加以运用。数据仓库的建立为矿山企业智能诊断提供了有力的数据支持,知识发现可以对数据仓库中的矿山生产经营数据进行更高层次的分析,自动地、智能地将数据转化为有用的信息和知识,这是实现矿山企业智能化诊断的有效途径。本文探讨了利用知识发现技术进行矿山企业智能诊断的方式,提出了集数据仓库、在线分析处理、知识发现、模型库、专家系统于一体的智能诊断系统结构,用于矿山企业的智能诊断,可以有效实现数据、诊断模型、诊断知识的统一表达。针对矿山企业诊断对象常具有不确定性特点,从实用角度出发,初步研究了诊断知识的表示,采用多维、多层次的知识表示方法,引入模糊逻辑、神经网络等多种知识表示形式进行有机集成。在矿山企业智能诊断中,通过数据仓库、人工神经网络技术和专家系统的集成,利用神经网络极强的学习能力,进行诊断知识的获取和推理,从大量数据中找出矿山企业生产经营中存在的知识模式,模拟专家发现企业生产中的影响因素。此外,本文引入演化计算方法,利用自然界遗传规律对矿山企业的生产经营数据进行分析,从而判断企业的运行状态。初步研究表明,演化计算既可以直接诊断矿山生产活动,又可以显著提高诊断效率。更多还原

【Abstract】 The goal of mine enterprise information system is to improve the management standard and economic benefit. It provides service for data processing, different levels of subsidiary management and policy making in mine enterprise by using systemic science methods. This system can modulate and control mine production behaviors, guarantee the optimum efficiency of mine production through making full use of management information of production, judging the running situation of production, finding out the problems existing in the management, locating the reasons for maladjustment of mine production. The management information system of mine enterprise has close relationship with mining engineering, mine management and computer technology. This paper studies and solves the problems involving in the construction and application of mine-enterprise information system by not only combining computer network technology but also applying data warehouse technology and KDD to mine enterprise. The study can strengthen importance of information system in the management of mine production. In accordance with the collection and organization of the basic data needed by mine production, this paper propounds the idea that mine enterprise Intranet must be set up based upon the data. It also analyzes emphatically the structure and implement technology of mine enterprise Intranet, solves the problems involving in the seamless transference technology from computer application system and database to the mine-enterprise net, conserves to the utmost the investment in the former mine information management system. Practical application shows that mine-enterprise net is an effective structure of mine-enterprise information system, and can provide convenient, integrated and timely information service for mine production management. With the development of computer application, mine enterprise has collected a lot of data during the production such as mine geological exploration, extraction development engineering and mine management. But most of the data has never been used effectively because there is no ability in mine enterprise to store and manage these data synthetically. Based on this situation, the paper renders that the data warehouse should be set up in the mine-enterprise information system. The data warehouse should be one which has multi-dimensional structure based on the present application system database, and which through corresponding subject data, abstracts and transfers data synthetically, sets up data marts from the bottom to the top, then integrates the data marts into mine enterprise data warehouse by which solid data support for the management of mine enterprise is guaranteed easily. On the basis of data warehouse, the OLAP technology applied in this paper analyzes and processes the management data of mine production from different aspects in order to understand the situation of mine production. This paper analyzes the construction characteristics of OLAP and the multi-dimension characteristics of management data during mine producing, both of which are applied to practical mine production. The foundation of data warehouse provides strong data support for the intelligent diagnosis for mine enterprise. KDD technology can give more advantaged analysis to the management data of mine production existing in data warehouse and change the data into useful information and knowledge automatically and intelligently, which are effective way to realize the intelligent diagnosis for mine enterprise .By using KDD technology, the paper discusses the way to implement intelligent diagnosis for mine enterprise, propounds the structure of diagnosis system integrating data warehouse, OLAP, KDD , model database and expert system. These technologies, if applied to intelligent diagnosis for mine enterprise, are able to achieve effectively the uniform expression of data, diagnosis model and diagnosis knowledge. According to the undetermined characteristics of the diagnoses objective in mine enterprise, for the purpose of practical application, this paper studies the expression of diagnosis knowledge, in which the knowledge is represented by multidimensional and multi-arrangement methods and then integrated synthetically by many knowledge expressions such as fuzzy logic, neutral network, etc.. The intelligent diagnosis for mine enterprise can not only accumulate and reason diagnosis knowledge by using data warehouse, artificial neutral network technology and expert system integrating, but also simulate experts to find any factors to affect mine production through discovering knowledge models existing in the management production of mine enterprise form a great deal of data. Besides this, the paper applies evolution computation and natural inheritance law to analyze the management production data in mine enterprise and judge the production situation of mine enterprise .It indicates that evolution computation can either diagnose mine production activities directly or improve diagnosis efficiency greatly. The study result of this paper shows that the application of mine enterprise Intranet can construct a open, unified, convenient management window to manage information resource synthetically. The establishment of data warehouse lays the sound application foundation for developing the potential function of mine information system. Based upon the achievement, each kind of application such as analysis and process of mine enterprise data, intelligent diagnosis and decision support can improve management of mine production significantly and thus enhance mine economic benefit. The practical application of this system in mine enterprise shows that what has been achieved in this paper can be used for reference in the construction of mine information system and also has important application and popularization significance.更多还原