本文選題：多傳感器 + 數(shù)據(jù)融合��； 參考：《寧夏大學(xué)》2015年碩士論文

【摘要】：煤礦開(kāi)采過(guò)程中,由于自然環(huán)境因素復(fù)雜多變,對(duì)井下災(zāi)害進(jìn)行事前安全預(yù)警較為困難。針對(duì)該問(wèn)題,本文研究利用多傳感器數(shù)據(jù)融合技術(shù)對(duì)井下安全狀態(tài)預(yù)警的方法。論文對(duì)現(xiàn)有多傳感數(shù)據(jù)融合模型進(jìn)行分類評(píng)估,提出井下自然災(zāi)害安全預(yù)警模型的一般設(shè)計(jì)原則。按照上述設(shè)計(jì)原則,提出一種井下多傳感器數(shù)據(jù)融合預(yù)警模型的設(shè)計(jì)方案。該模型采用兩級(jí)分層融合結(jié)構(gòu),通過(guò)特征層和決策層兩層實(shí)現(xiàn)數(shù)據(jù)融合。特征層面向針對(duì)單一危險(xiǎn)源指標(biāo),將主成分分析和神經(jīng)網(wǎng)絡(luò)算法相結(jié)合,實(shí)現(xiàn)危險(xiǎn)源特征提取。選取合適的權(quán)重系數(shù),利用神經(jīng)網(wǎng)絡(luò)對(duì)樣本數(shù)據(jù)進(jìn)行訓(xùn)練,輸出預(yù)警指標(biāo)的危險(xiǎn)程度。在決策層利用基于D-S證據(jù)理論的改進(jìn)算法,構(gòu)造BPAsO函數(shù),給出單一危險(xiǎn)源的預(yù)警決策輸出。最后,論文利用MATLAB對(duì)預(yù)警模型框架進(jìn)行仿真,以劉莊煤業(yè)瓦斯數(shù)據(jù)為例,將預(yù)警結(jié)果與實(shí)測(cè)數(shù)據(jù)進(jìn)行比對(duì)研究。仿真結(jié)果表明,瓦斯單一危險(xiǎn)源經(jīng)融合后,其各種狀態(tài)的安全判斷評(píng)估與實(shí)測(cè)數(shù)據(jù)相對(duì)具有較好的一致性,結(jié)果符合實(shí)際情況。預(yù)警模型的研究對(duì)實(shí)際井下安全預(yù)警具有一定的指導(dǎo)意義。本文研究工作獲得國(guó)家自然科學(xué)基金項(xiàng)目《無(wú)線傳感器網(wǎng)絡(luò)數(shù)據(jù)匯聚傳送關(guān)鍵技術(shù)研究及在半干旱設(shè)施農(nóng)業(yè)中的應(yīng)用》(基金編號(hào)：612610001)的支持。
[Abstract]:In the process of coal mining, due to the complex and changeable natural environment factors, it is difficult to pre-warn the downhole disaster. In order to solve this problem, this paper studies the method of using multi-sensor data fusion technology to predict the underground safety state. In this paper, the existing multi-sensor data fusion model is classified and evaluated, and the general design principles of downhole natural disaster safety early warning model are proposed. According to the above design principle, a design scheme of downhole multi-sensor data fusion early warning model is proposed. The model adopts two-level hierarchical fusion structure, and implements data fusion by feature layer and decision layer. In view of the single hazard index, the feature level is extracted by combining principal component analysis and neural network algorithm. Select the appropriate weight coefficient, use neural network to train the sample data, output the danger degree of early warning index. An improved algorithm based on D-S evidence theory is used to construct the BPAsO function at the decision level, and the decision output of the single hazard source is given. Finally, the paper uses MATLAB to simulate the early warning model frame, taking Liuzhuang coal industry gas data as an example, the early warning results are compared with the measured data. The simulation results show that after the gas single hazard source is fused, the safety judgment and evaluation of its various states are relatively consistent with the measured data, and the results are in line with the actual situation. The research of the early warning model has certain guiding significance to the actual underground safety early warning. The work of this paper is supported by the National Natural Science Foundation of China "Research on key Technology of data Convergence and Transmission in Wireless Sensor Networks and its Application in Semi-arid Facility Agriculture" (Fund No.: 612610001).
【學(xué)位授予單位】：寧夏大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TD76;TP212

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 王慧敏,陳寶書(shū);煤炭行業(yè)預(yù)警指標(biāo)體系的基本框架結(jié)構(gòu)[J];中國(guó)煤炭經(jīng)濟(jì)學(xué)院學(xué)報(bào);1996年04期

2 韓德強(qiáng);鄧勇;韓崇昭;侯志強(qiáng);;基于證據(jù)距離與不確定度的證據(jù)組合方法(英文)[J];紅外與毫米波學(xué)報(bào);2011年05期

3 孔繁森,王宇,于駿一;顫振征兆早期識(shí)別的模糊信息融合法[J];機(jī)械工程學(xué)報(bào);2004年02期

4 張山鷹,潘泉,張洪才;一種新的證據(jù)推理組合規(guī)則[J];控制與決策;2000年05期

5 劉小生;薛萍;;基于神經(jīng)網(wǎng)絡(luò)的礦山安全預(yù)警專家系統(tǒng)[J];煤礦安全;2008年12期

6 黃瑛,陶云剛,周潔敏,蘇登軍;D-S證據(jù)理論在多傳感器數(shù)據(jù)融合中的應(yīng)用[J];南京航空航天大學(xué)學(xué)報(bào);1999年02期

7 潘泉;王增福;梁彥;楊峰;劉準(zhǔn)釓;;信息融合理論的基本方法與進(jìn)展(Ⅱ)[J];控制理論與應(yīng)用;2012年10期

8 趙代英;何學(xué)秋;江田漢;;我國(guó)煤礦行業(yè)安全生產(chǎn)預(yù)警指數(shù)模型研究[J];中國(guó)安全生產(chǎn)科學(xué)技術(shù);2014年01期

，

本文編號(hào)：2116009