数据挖掘在冲击地压智能预警系统中的应用与研究.pdf

分类号分类号TDTD 324324 密密 级级 公公 开开 U D CU D C 单位代码单位代码 1042410424 学学 位位 论论 文文 数据挖掘在冲击地压智能预警系统数据挖掘在冲击地压智能预警系统 中的应用与研究中的应用与研究 贾冬冬贾冬冬 申请学位级别申请学位级别硕士学位硕士学位 专业专业名名称称矿业工程矿业工程 指导教师姓名指导教师姓名朱朱 学学 军军 职职 称称副副 教教 授授 副副指导教师姓名指导教师姓名许许 志志 杰杰 职职 称称高级工程师高级工程师 山山 东东 科科 技技 大大 学学 二二〇〇一八年六一八年六月月 万方数据 论文题目论文题目 数据挖掘在冲击地压智能预警系统数据挖掘在冲击地压智能预警系统 中的应用与研究中的应用与研究 作者姓名作者姓名贾贾 冬冬 冬冬 入学时间入学时间 2015 年年 9 月月 专业名称专业名称矿业工程矿业工程 研究方向研究方向矿山压力与岩层矿山压力与岩层控制控制 指导教师指导教师朱朱 学学 军军 职职 称称副副 教教 授授 副副指导教师指导教师许许 志志 杰杰 职职 称称高高 级级 工工 程程 师师 论文提交日期论文提交日期2018 年年 4 月月 论文答辩日期论文答辩日期2018 年年 5 月月 授予学位日期授予学位日期2018 年年 6 月月 万方数据 THE APPLICATION AND RESEARCH OF DATA MINING IN THE IMPACT GROUND PRESSURE INTELLIGENT EARLY WARNING SYSTEM A Dissertation ted in fulfillment of the requirements of the degree of MASTER OF ENGINEERING from Shandong University of Science and Technology by Jia Dongdong Supervisor Associate Professor Zhu Xuejun Vice Supervisor Senior Engineer Xu Zhijie College of Natural Resources and Environmental Engineering June 2018 万方数据 万方数据 万方数据 万方数据 山东科技大学硕士学位论文 摘要 I 摘摘 要要 我国煤炭资源丰富，占世界储量的 11.60，国民经济快速发展的同时更要 对井下工作人员安全进行保护，其中冲击地压具有极大破坏性，严重威胁煤矿 安全生产，冲击地压可能诱发煤与瓦斯突出、煤层自燃发火、冒顶等群死群伤 的次生灾害，况且现在我国进入深部开采高峰期，煤矿冲击地压发生强度、频 次及区域都在逐年増加。因此，研制实时、可靠的冲击地压监测系统，建立高 效、准确的冲击地压预警系统符合我国智能矿山的战略部署。 本文首先参考国内外优秀的冲击地压监测技术的成功经验，通过冲击地压 前兆信息显现，结合日趋先进的数据挖掘技术、数据仓库技术等大数据技术， 深入研究并架构了冲击地压智能预警系统的基本框架，设计并开发了一套基于 B/S 模式的数据可视化界面。其次，由于煤矿冲击地压的复杂机理和冲击地压 监测数据的复杂性、多样性特点，采用了市场上比较流行的数据库技术和数据 仓库技术，对冲击地压监测数据进行了数据仓库的建模，使开发者能从不同角 度分析和使用冲击地压监测数据，并且，数据仓库的建立也是数据挖掘工作的 前提。然后，详细介绍了数据挖掘技术并重点探讨了其中的聚类算法，特别是 K-means 算法，针对 K-means 算法的不足点，对算法进行了改进，把改进后的 算法运用到冲击地压监测数据样本后大大提高了准确率和运行速率。最后，设 计并开发了冲击地压智能预警系统的可视化界面，通过可视化技术的直观、形 象的特点来让每个用户都可以看懂监测数据。 冲击地压智能预警系统分为软件启动、数据获取、数据挖掘、数据可视化 四个大模块组成，此系统是以 Java 语言为主，CSS、JavaScript、HTML 等语言 为辅，Tomcat 为服务器，Mysql 为冲击地压监测数据仓库的基于 B/S 模式的 JavaWeb 项目。用户可以通过 B/S 模式的 Web 可视化界面将业务请求发送至 Tomcat 服务器，服务器请求执行用户的请求后返回浏览器，用户随时随地就可 以通过浏览器实现实时查询、分析监测数据等功能，给用户提供了极大的便利 性。 数据挖掘技术在冲击地压智能预警系统中的应用，有助于改进现有的冲击 地压监测系统，增强冲击地压灾害的预报和预警能力，不仅保障了井下的人员 安全，也对我国煤炭经济的可持续发展做出了一定的贡献。 关键词关键词冲击地压；数据挖掘；聚类分析；智能预警；数据仓库 万方数据 山东科技大学硕士学位论文 摘要 II Abstract My country is rich in coal resources, accounting for 11.60 of the worlds reserves, the rapid development of national economy and to the protection of the safety of staff in the underground, the impact pressure has a great destructive, a serious threat to coal mine safety production, the impact pressure may induce, coal seam spontaneous combustion of coal and gas outburst, roof caving, such as group die group of injury of secondary disasters, and now into the peak period of deep mining in our country, the intensity, frequency of rock bursts occurred in coal mines, and area are raised year by year.Therefore, the development of real-time and reliable impact pressure monitoring system, the establishment of an efficient and accurate impact pressure warning system is in line with the strategic deployment of Chinas smart mines. This paper reference the successful experience of the excellent impact ground pressure monitoring technology at home and abroad, through the impact ground pressure precursor ination, combined with the increasingly advanced technology of data mining, data warehouse technology, big data technologies such as in-depth study and architecture of the intelligent warning system the basic framework of percussive ground pressure, design and develop a set of data visualization interface based on B/S mode.Secondly, due to the complex mechanism of impact ground pressure of coal mine and the complexity of the impact ground pressure monitoring data, the characteristics of diversity, the relatively popular in the market database and data warehouse technology, the impact ground pressure monitoring data in data warehouse modeling, enables developers to analysis from different angles and the use of impact ground pressure monitoring data.Moreover, the establishment of data warehouse is the premise of data mining.Then, the paper introduces the data mining technology and mainly discussed the clustering algorithms, especially k - means algorithm, aiming at the shortcomings of the k - means algorithm, the algorithm is improved, the improved algorithm is applied to impact ground pressure monitoring data samples after greatly improves the accuracy and speed.Finally, the design and development of the visual interface of the impact pressure intelligent early warning system is designed, and the visual and visual features of the visualization technology are used to make the monitoring 万方数据 山东科技大学硕士学位论文 摘要 III data available to every user. Impact ground pressure intelligent warning system is divided into software startup, data acquisition, data mining, data visualization of four big modules, this system is based on Java language is given priority to, CSS, JavaScript, HTML, the language is complementary, Tomcat as the server, Mysql for impact ground pressure monitoring data warehouse based on B/S mode JavaWeb project.Users can through the B/S mode Web visual interface to request our business to the Tomcat server, the server request cution after the users request to return the browser, the user anytime and anywhere through the browser can achieve real-time query, analysis of monitoring data and provides great convenience to the user. Data mining technology in the intelligent warning system, the application of percussive ground pressure is helpful to improve the existing impact ground pressure monitoring system, enhance the capacity of impact ground pressure disaster forecast and early warning, not only guarantee the safety of underground personnel, also for the sustainable development of Chinas coal economy has made a certain contribution. Keywords Impact ground pressure;Data mining;Cluster analysis;Intelligent warning;The data warehouse 万方数据 山东科技大学硕士学位论文 目录 IV 目目 录录 摘摘 要要 .............................................................................................................................................................................................................................. I I 目目 录录 ........................................................................................................................................................................................................................ IVIV 1 1 绪论绪论 .......................................................................................................................................................................................................................... 1 1 1.1 选题背景及研究意义 .................................................................................... 1 1.2 冲击地压监测系统与数据挖掘技术的研究现状 ........................................ 2 1.3 存在的问题与不足 ........................................................................................ 6 1.4 研究内容与技术路线 .................................................................................... 7 2 2 冲击地压智能预警系统的架构设计冲击地压智能预警系统的架构设计 .............................................................................................. 1010 2.1 设计目标 ...................................................................................................... 10 2.2 冲击地压智能预警系统设计 ....................................................................... 11 2.3 系统功能设计 .............................................................................................. 12 2.4 小结 .............................................................................................................. 15 3 3 冲击地压智能预警系统数据仓库的设计冲击地压智能预警系统数据仓库的设计 ............................................................................ 1616 3.1 冲击地压概述及孕育过程 .......................................................................... 16 3.2 冲击地压数据仓库的设计 .......................................................................... 19 3.3 小结 .............................................................................................................. 26 4 4 数据挖掘技术在冲击地压中的应用数据挖掘技术在冲击地压中的应用 ............................................................................................ 2727 4.1 数据挖掘概述 .............................................................................................. 27 4.2 聚类分析概述 .............................................................................................. 29 4.3 K-means 聚类算法在冲击地压预测中的应用 .......................................... 31 4.4 小结 .............................................................................................................. 37 5 5 冲击地压数据可视化的实现冲击地压数据可视化的实现 ........................................................................................................................ 3939 5.1 软件架构 ...................................................................................................... 39 万方数据 山东科技大学硕士学位论文 目录 V 5.2 冲击地压数据可视化的开发环境 .............................................................. 40 5.3 冲击地压智能预警系统各模块的实现 ...................................................... 41 5.4 小结 .............................................................................................................. 59 6 6 结论与展望结论与展望 ...................................................................................................................................................................................... 6060 6.1 主要结论 ...................................................................................................... 60 6.2 有待进一步研究的问题和展望 .................................................................. 60 参考文献参考文献 .............................................................................................................................................................................................................. 6262 附附 录录 ...................................................................................................................................................................................................................... 6767 致致 谢谢 ...................................................................................................................................................................................................................... 8080 学习经历简介学习经历简介 .............................................................................................................................................................................................. 8181 攻读硕士期间主要成果攻读硕士期间主要成果........................................................................................................................................................ 8282 万方数据 山东科技大学硕士学位论文 目录 VI Contents Abstract .......................................................................................................................................... I Catalog ........................................................................................................................................ IV 1 Introduction ............................................................................................................................ 1 1.1 background and significance of research ........................................................................... 1 1.2 the research status of rock burst monitoring system and data mining technology .......... 2 1.3 existing problems and shortcomings .................................................................................. 6 1.4 research content and technical route .................................................................................. 7 2 architecture design of intelligent early-warning system for impact ground pressure .10 2.1 design goal ....................................................................................................................... 10 2.2 design of intelligent early-warning system for impact ground pressure .......................... 11 2.3 system function design ..................................................................................................... 12 2.4 summary ........................................................................................................................... 15 3 design of data warehouse for intelligent early warning system of rock burst ..............16 3.1 overview of the impact ground pressure and the process of inoculation ......................... 16 3.2 design of rock burst data warehouse ................................................................................ 19 3.3 summary ........................................................................................................................... 26 4 the application of data mining technology in rock burst .................................................27 4.1 data mining overview ....................................................................................................... 27 4.2 cluster Analysis Overview ............................................................................................... 29 4.3 application of K-means clustering algorithm to prediction of rock burst ........................ 31 4.4 summary ........................................................................................................................... 37 5 realization of data visualization of impact ground pressure ...........................................39 5.1 software architecture ........................................................................................................ 39 5.2 development environment for data visualization of impact ground pressure ................... 40 5.3 the realization of each module of the intelligent early warning system for rock burst .... 41 万方数据 山东科技大学硕士学位论文 目录 VII 5.4 summary ........................................................................................................................... 59 6 summary and outlook ..........................................................................................................60 6.1 full text summary ............................................................................................................. 60 6.2 prospects and problems .................................................................................................... 60 Reference......................................................................................................................................62 Appendix ......................................................................................................................................67 Thanks ....................................