基于云计算数据集成模式的矿井瓦斯预警研究
发布时间：2019-04-19 09:54
【摘要】：矿井瓦斯一直是我国煤矿主要的灾害形式之一，并严重困扰着煤矿的安全生产。研究矿井日常检/监测数据的有效处理及其预测预警应用，有利于拓展安全监测监控系统的功能，是提高瓦斯灾害预警能力的重要手段。本论文在分析矿井瓦斯检/监测数据特征及其集成管控模式的基础上，深入研究了基于云计算数据集成模式下的矿井瓦斯预警分析理论和方法。 研究了矿井瓦斯检/监测数据的特点及其集成管控模式。分析了瓦斯检/监测数据的特征，并对于环境、人为、管理等因素影响下存在的异常数据、数据缺失问题，针对其特征进行平滑处理，使其形成完整的监测数据序列，符合监测数据整体统计特性，并构建了云计算模式下检/监测数据集成管控模式。 研究了矿井瓦斯浓度变化趋势预测预警方法。在瓦斯监测数据预处理的基础上，基于时间序列分析的自回归滑动平均（ARMA）模型，建立了适用于实时监测数据的瓦斯浓度动态趋势预测预警分析模型，结合实时预测结果与所在时段瓦斯监测数据的统计特征实现了瓦斯浓度变化趋势的动态预警。 研究了矿井瓦斯突出危险性预测预警方法。通过分析瓦斯实时监测数据的特征，提取反映瓦斯浓度实时变化趋势的参数、瓦斯浓度变化速率的参数以及用于表达瓦斯涌出特征的参数，结合防突检测参数，，基于v-支持向量机（v-SVM）模型，构建了瓦斯突出危险性预测预警模型，结合瓦斯突出危险性预测结果与防突检测数据的统计特征，实现了瓦斯突出危险性预警。 研究了矿井瓦斯预警的云计算模型架构。基于云计算的原理，构建了应用于矿井瓦斯预警分析的云计算模式的物理架构及其云计算平台模式，并将矿井瓦斯检/监测数据处理及预测预警算法予以封装，为瓦斯预警计算的程序化服务构建了云计算模式，实现了高效的预警分析。 研究了云计算数据集成模式下瓦斯预警分析应用。基于所建立的瓦斯预警数学模型，将瓦斯检/监测数据处理的云计算模式应用于现场预警分析，经过实际检/监测数据的对照检验，表现出了良好的适用性和有效性。 本论文研究的云计算数据集成模式下瓦斯预警分析理论和方法，适用于煤矿现场的瓦斯预警分析应用，为煤矿瓦斯灾害防治提供了新的数字化平台构建方法和手段。
[Abstract]:Mine gas has always been one of the main disaster forms of coal mines in our country, and seriously perplexed the safety production of coal mines. To study the effective processing of mine daily inspection / monitoring data and its prediction and early warning application is helpful to expand the function of safety monitoring and monitoring system and is an important means to improve the ability of gas disaster early warning. On the basis of analyzing the characteristics of mine gas inspection / monitoring data and its integrated management and control mode, this paper deeply studies the theory and method of mine gas early warning analysis based on cloud computing data integration mode. The characteristics of mine gas inspection / monitoring data and its integrated control mode are studied. In this paper, the characteristics of gas inspection / monitoring data are analyzed, and the problems of abnormal data and data missing under the influence of environment, man-made, management and other factors are analyzed, and the characteristics of gas inspection / monitoring data are smoothed to form a complete monitoring data sequence. In accordance with the overall statistical characteristics of monitoring data, the integrated control model of inspection / monitoring data under cloud computing model is constructed. The prediction and early warning method of gas concentration change trend in mine is studied. Based on the pretreatment of gas monitoring data and the autoregressive moving average (ARMA) model based on time series analysis, a prediction and early warning model of gas concentration dynamic trend is established, which is suitable for real-time monitoring data. Combined with the real-time prediction results and the statistical characteristics of the gas monitoring data in the time period, the dynamic early-warning of the gas concentration change trend is realized. The prediction and early warning method of mine gas outburst risk is studied. By analyzing the characteristics of real-time gas monitoring data, the parameters reflecting the trend of real-time change of gas concentration, the parameters of gas concentration change rate and the parameters used to express the characteristics of gas emission are extracted, and combined with the detection parameters of outburst prevention. Based on the v-support vector machine (v-SVM) model, a gas outburst risk prediction and early warning model is constructed. Based on the statistical characteristics of gas outburst risk prediction results and outburst prevention detection data, the gas outburst risk early warning is realized. The cloud computing model architecture of mine gas early warning is studied. Based on the principle of cloud computing, the physical structure and cloud computing platform model of cloud computing model applied to mine gas early warning analysis are constructed, and the mine gas inspection / monitoring data processing and prediction and early warning algorithm are encapsulated. The cloud computing mode is constructed for the programmed service of gas early warning calculation, and the efficient early warning analysis is realized. The application of gas early warning analysis in cloud computing data integration mode is studied. Based on the established gas early warning mathematical model, the cloud computing model of gas detection / monitoring data processing is applied to the field early warning analysis. Through the comparison test of actual inspection / monitoring data, it shows good applicability and effectiveness. This paper studies the theory and method of gas early warning analysis under cloud computing data integration mode, which is suitable for the application of gas early warning analysis on the spot of coal mine, and provides a new method and means of constructing digital platform for the prevention and control of coal mine gas disaster.
【学位授予单位】：西安科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TD712.7
本文编号：2460839
[Abstract]:Mine gas has always been one of the main disaster forms of coal mines in our country, and seriously perplexed the safety production of coal mines. To study the effective processing of mine daily inspection / monitoring data and its prediction and early warning application is helpful to expand the function of safety monitoring and monitoring system and is an important means to improve the ability of gas disaster early warning. On the basis of analyzing the characteristics of mine gas inspection / monitoring data and its integrated management and control mode, this paper deeply studies the theory and method of mine gas early warning analysis based on cloud computing data integration mode. The characteristics of mine gas inspection / monitoring data and its integrated control mode are studied. In this paper, the characteristics of gas inspection / monitoring data are analyzed, and the problems of abnormal data and data missing under the influence of environment, man-made, management and other factors are analyzed, and the characteristics of gas inspection / monitoring data are smoothed to form a complete monitoring data sequence. In accordance with the overall statistical characteristics of monitoring data, the integrated control model of inspection / monitoring data under cloud computing model is constructed. The prediction and early warning method of gas concentration change trend in mine is studied. Based on the pretreatment of gas monitoring data and the autoregressive moving average (ARMA) model based on time series analysis, a prediction and early warning model of gas concentration dynamic trend is established, which is suitable for real-time monitoring data. Combined with the real-time prediction results and the statistical characteristics of the gas monitoring data in the time period, the dynamic early-warning of the gas concentration change trend is realized. The prediction and early warning method of mine gas outburst risk is studied. By analyzing the characteristics of real-time gas monitoring data, the parameters reflecting the trend of real-time change of gas concentration, the parameters of gas concentration change rate and the parameters used to express the characteristics of gas emission are extracted, and combined with the detection parameters of outburst prevention. Based on the v-support vector machine (v-SVM) model, a gas outburst risk prediction and early warning model is constructed. Based on the statistical characteristics of gas outburst risk prediction results and outburst prevention detection data, the gas outburst risk early warning is realized. The cloud computing model architecture of mine gas early warning is studied. Based on the principle of cloud computing, the physical structure and cloud computing platform model of cloud computing model applied to mine gas early warning analysis are constructed, and the mine gas inspection / monitoring data processing and prediction and early warning algorithm are encapsulated. The cloud computing mode is constructed for the programmed service of gas early warning calculation, and the efficient early warning analysis is realized. The application of gas early warning analysis in cloud computing data integration mode is studied. Based on the established gas early warning mathematical model, the cloud computing model of gas detection / monitoring data processing is applied to the field early warning analysis. Through the comparison test of actual inspection / monitoring data, it shows good applicability and effectiveness. This paper studies the theory and method of gas early warning analysis under cloud computing data integration mode, which is suitable for the application of gas early warning analysis on the spot of coal mine, and provides a new method and means of constructing digital platform for the prevention and control of coal mine gas disaster.
【学位授予单位】：西安科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TD712.7
【参考文献】
相关期刊论文 前10条
1 郭德勇;郑茂杰;鞠传磊;郝相龙;;采煤工作面瓦斯涌出量预测逐步回归方法[J];北京科技大学学报;2009年09期
2 汪云甲;杨敏;张克;;数字矿山与煤矿瓦斯监测及预警[J];地理信息世界;2008年05期
3 ;Discussion of Some Problems About Nonlinear Time Series Prediction Using v-Support Vector Machine[J];Communications in Theoretical Physics;2007年07期
4 李益国;沈炯;;基于ν-支持向量机的边际电价预测及置信区间估计[J];东南大学学报(自然科学版);2007年01期
5 周邦全;;煤矿安全监测监控系统的发展历程和趋势[J];矿业安全与环保;2007年S1期
6 崔小彦;赵国强;;混沌预测在井下瓦斯浓度监测的应用[J];辽宁工程技术大学学报(自然科学版);2008年S1期
7 刘永波;代君;;云计算环境下的信息资源整合技术研究[J];计算机光盘软件与应用;2012年08期
8 刘源骏;袁梅;马科伟;;煤矿安全监测监控系统探讨[J];矿业工程;2010年02期
9 李桂银;;基于云计算的信息共享空间资源建设[J];福建图书馆理论与实践;2011年02期
10 刘婷婷;史久根;韩江洪;;基于SVM的瓦斯体积分数混沌时间序列预测[J];合肥工业大学学报(自然科学版);2009年08期
相关博士学位论文 前3条
1 陈祖云;煤与瓦斯突出前兆的非线性预测及支持向量机识别研究[D];中国矿业大学;2009年
2 王汉斌;煤与瓦斯突出的分形预测理论及应用[D];太原理工大学;2009年
3 黄为勇;基于支持向量机数据融合的矿井瓦斯预警技术研究[D];中国矿业大学;2009年
本文编号：2460839
本文链接：https://www.wllwen.com/kejilunwen/anquangongcheng/2460839.html
