【作者】 李刚；

【导师】 宁书年； 彭苏萍；

【作者基本信息】 中国矿业大学（北京） ， 地球探测与信息技术， 2009， 博士

【摘要】 瓦斯浓度受多种地质因素的综合作用,与矿井瓦斯突出关系密切。为了预测矿井瓦斯突出危险性,以淮南潘三矿为实验矿井,研究了瓦斯浓度时间序列的基本分形特征,提出了面向瓦斯浓度时间序列的极差分形模型,应用极差分形模型研究了瓦斯突出危险性的评价方法,基于多元统计主成分分析,建立了突出危险性评价模型,对于提前发现和预报瓦斯突出有很大的参考价值。运用混沌时间序列分析方法提取瓦斯浓度的混沌特性,对传统的混沌预测方法进行了改进,提出了瓦斯浓度时间序列混沌预测模型的改进算法,提高了混沌预测算法的计算效率与预测精度。实验结果表明,混沌预测模型可以定量地预测到未来较长时间的瓦斯浓度变化,预测精度较高；极差分形模型可以定性地分析预测瓦斯浓度时间序列的动态发展变化趋势,基于瓦斯突出评价模型可以对测点的突出危险性进行定量预测。在此基础上,设计并实现了煤矿瓦斯浓度预测软件。更多还原

【Abstract】 Gas outburst, as a grave natural disaster, is a great threat to the safe production of coal mine. Inspired by the multi-geological factors, Gas density had close relations directly with gas disaster like outburst. In order to acquire the accurate and quick prediction on the risk of coalmine gas outburst, experiments had been carried out at Pansan Coalmine in Huainan, Anhui Province. Range Fractal model was proposed for the analysis of gas density series after the detailed study of basic fractal characteristics of gas density time series, which consists of gas monitor density data. Chaotic time series analysis was applied to research the chaotic properties of gas density time series. The traditional chaotic forecast algorithm for gas series was ameliorated, and for gas time series 2 reformative and improved chaotic prediction models were established with greater performance and efficiency. Experiment results verified that using range fractal model could offer a qualitative prediction to the dynamical dangerous trends of gas density time series. The quantitive prediction results of both chaotic models appeared reliable and feasible as expected. After research on the gas outburst case of Pansan coalmine, the abnormal aspects were discovered in application of analyzing the relations between gas outburst and gas density time series before outburst. And some general evolution rules appeared clearly to explanations for gas outburst. These valuable discoveries are to be testified with more gas outburst data. On the basis of Multivariate Statistics, the evaluation model was proposed for gas outburst risk. Based on above all, design of the gas density prediction software was achieved and performed smooth as expected.更多还原