本文選題：信息融合　切入點(diǎn)：數(shù)據(jù)挖掘　出處：《中國礦業(yè)大學(xué)（北京）》2013年博士論文

【摘要】：煤炭資源一直以來作為重要的能源來源，在生活和工業(yè)中的需求日益增加，而煤礦的開采因?yàn)槠涮厥獾沫h(huán)境因素而有一定的危險(xiǎn)性，尤其是煤礦瓦斯災(zāi)害作為煤礦首要的惡性事故，其發(fā)生頻率高，破壞性廣，社會(huì)影響大。長期以來，人們一直是以瓦斯的檢測為主，無法提前預(yù)測瓦斯災(zāi)害的發(fā)生。本文綜合闡述了煤礦瓦斯監(jiān)測的主要問題所在，討論了預(yù)測災(zāi)害發(fā)生的重要性和可能性。通過對(duì)監(jiān)測數(shù)據(jù)資源進(jìn)行數(shù)據(jù)挖掘、分析，采用支持向量機(jī)、模糊集等理論建立多源、多平臺(tái)、多傳感器煤礦瓦斯災(zāi)害在線辨識(shí)、隱患判別和決策模型；攻克了瓦斯監(jiān)測系統(tǒng)不能及時(shí)發(fā)現(xiàn)重大瓦斯災(zāi)害隱患的關(guān)鍵技術(shù)。為了解決瓦斯災(zāi)害在線辨識(shí)中的不確定性和不精確性的問題，分別建立了基于Bayes network和D S證據(jù)理論的煤礦瓦斯災(zāi)害特征級(jí)融合模型及算法。并對(duì)這兩種模型進(jìn)行實(shí)驗(yàn)研究，檢驗(yàn)這兩種模型的有效性。提出利用多種監(jiān)測信息進(jìn)行“數(shù)據(jù)挖掘”、分析、處理、融合和綜合判斷瓦斯災(zāi)害危險(xiǎn)性的先進(jìn)理論和方法，為解決預(yù)測問題提供了理論依據(jù)和科學(xué)方法，具有重要的實(shí)際意義。
[Abstract]:As an important source of energy, coal resources have been increasing demand in daily life and industry, and coal mining is dangerous because of its special environmental factors. Especially, the gas disaster in coal mine is the most serious accident in coal mine, its occurrence frequency is high, destructive is wide, social influence is big. For a long time, people always take gas detection as the main factor. It is impossible to predict the occurrence of gas disaster in advance. This paper comprehensively expounds the main problems of gas monitoring in coal mine, and discusses the importance and possibility of predicting disaster. Based on support vector machine and fuzzy set theory, the online identification, hidden trouble discrimination and decision model of gas disaster in multi-source, multi-platform and multi-sensor coal mine are established. In order to solve the problem of uncertainty and inaccuracy in online gas disaster identification, the key technology that gas monitoring system can not find the hidden danger of major gas disaster in time is overcome. The characteristic level fusion model and algorithm of coal mine gas disaster based on Bayes network and DS evidence theory are established, and the experimental research on these two models is carried out. To test the validity of these two models, the advanced theory and method of "data mining", analyzing, processing, merging and synthetically judging the hazard of gas disaster using various monitoring information are put forward. It provides theoretical basis and scientific method for solving prediction problem and has important practical significance.
【學(xué)位授予單位】：中國礦業(yè)大學(xué)（北京）
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2013
【分類號(hào)】：TD712;TP202

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 郭獻(xiàn)文,汪延,周立民;瓦斯爆炸:煤礦礦難頭號(hào)殺手[J];w，

本文編號(hào)：1655111