【摘要】：盡管我國煤礦死亡人數(shù)逐年下降，但形勢依然嚴峻。通過物聯(lián)網(wǎng)技術(shù)和多智能體理論將煤礦井下的物封裝成含有煤礦事故預警知識庫的智能體，可以有效降低煤礦事故的發(fā)生率。但目前尚缺乏基于事故致因機理的煤礦事故預警知識庫模型方面的研究。又近些年，本體在知識表示、人工智能方面獲得了長足發(fā)展。因此，論文研究基于本體的煤礦事故預警知識庫模型構(gòu)成要素及相互關(guān)系，并實踐應用。 論文首先對危險源和隱患的概念進行了重新界定，分析了二者的關(guān)系。然后，在此基礎(chǔ)上，結(jié)合時空邏輯，提出了抽象的、作為事故預警知識庫模型理論基石的時空視角下的基于根源危險源的事故致因機理。接著，設(shè)計了適合描述具體事故發(fā)生機理的時空事故樹分析法。然后，根據(jù)根源危險源智能體事故預警流程及本體論，，構(gòu)建了基于本體的煤礦事故預警知識庫模型。模型主要包括：基于根源危險源的事故致因機理、時空邏輯、煤礦危險源庫、具體事故致因機理表示法、具體事故致因機理描述和推理機。 事故預警知識庫的時空邏輯設(shè)計是在現(xiàn)有的時空邏輯研究的基礎(chǔ)上，研究時空實體的概念層次及其相互關(guān)系，為描述事故致因機理及預警規(guī)則提供了時空支持，時間邏輯采用了點段結(jié)合的時間表示，將時間實體分為時間點和時間段，時間實體間的關(guān)系分為13個大類�？臻g邏輯采用OGC空間數(shù)據(jù)模型的子集（點、線、面）構(gòu)建根源危險源的空間表示，空間實體間的關(guān)系包括拓撲關(guān)系、方向關(guān)系、度量關(guān)系。 為了使智能體能夠理解事故預警知識庫模型并進行推理，構(gòu)建了相應的本體和推理算法。煤礦事故預警知識庫本體包括：用于描述基于根源危險源的事故致因機理和煤礦領(lǐng)域概念及關(guān)系的煤礦風險本體、描述時間實體及其關(guān)系的時間本體、描述空間實體及其關(guān)系的空間本體、用以構(gòu)建具體的事故致因機理的時空事故樹表示本體。事故智能預警推理算法設(shè)計研究基于本體的描述邏輯推理算法、時間推理算法、空間推理算法及時空事故樹推理預警算法，為事故的智能預警提供算法支持。描述邏輯推理基于Tableau算法實現(xiàn)，基于時間約束網(wǎng)絡進行時間的定性定量推理，基于組合表進行空間拓撲、方位、度量關(guān)系的推理，在現(xiàn)有的事故樹定性定量算法的基礎(chǔ)上，結(jié)合時空約束描述，實現(xiàn)了時空事故樹的定性定量推理計算。 最后，設(shè)計實現(xiàn)了四層架構(gòu)的王樓煤礦事故智能預警平臺，對構(gòu)建的基于本體的煤礦事故預警知識庫模型進行了實踐應用。應用表明：論文所建立的基于本體的煤礦事故預警知識庫模型理論合理、實踐有效，可以為煤礦事故預警知識庫的構(gòu)建提供有益的參考。
[Abstract]:Although the number of coal mine deaths in China has decreased year by year, the situation is still grim. Through the technology of internet of things and the theory of multi-agent, the objects in coal mine can be encapsulated into an agent containing the knowledge base of mine accident warning, which can effectively reduce the incidence of coal mine accidents. However, there is still a lack of research on the knowledge base model of coal mine accident warning based on accident cause mechanism. In recent years, ontology has made great progress in knowledge representation and artificial intelligence. Therefore, this paper studies the knowledge base model of coal mine accident warning based on ontology and its relationship, and applies it to practice. Firstly, the concept of hazard and hidden danger is redefined and the relationship between them is analyzed. Then, based on the spatio-temporal logic, the mechanism of accident cause based on the source of hazard is proposed, which is the theoretical cornerstone of the knowledge base model of accident early warning, which is based on the theory of time and space. Then, a spatio-temporal accident tree analysis method is designed to describe the mechanism of specific accidents. Then, according to the process and ontology of agent accident warning, the knowledge base model of coal mine accident warning based on ontology is constructed. The model mainly includes: accident cause mechanism based on source hazard source, space-time logic, coal mine hazard source database, specific accident cause mechanism representation method, specific accident cause mechanism description and inference machine. The design of spatio-temporal logic of accident warning knowledge base is based on the existing research of spatio-temporal logic, which studies the concept level of space-time entity and its relationship, which provides spatio-temporal support for describing the mechanism of accident cause and the rules of early warning. Time logic is expressed by the combination of point and segment. Time entities are divided into time points and time periods, and the relationships between time entities are divided into 13 categories. Spatial logic uses the subsets (points, lines, surfaces) of the OGC spatial data model to construct the spatial representation of the source hazard source. The relationships among spatial entities include topological relations, directional relationships, and metric relationships. In order to make the agent understand the knowledge base model of accident warning and infer, the corresponding ontology and reasoning algorithm are constructed. The knowledge base ontology of coal mine accident warning includes: coal mine risk ontology which is used to describe the mechanism of accident cause based on the source of hazard and the concept and relation of coal mine domain, and the time ontology to describe the time entity and its relation. The spatial ontology which describes the spatial entity and its relationship is used to construct the spatio-temporal accident tree representation ontology of the specific accident cause mechanism. Design of reasoning algorithm for Intelligent accident early warning; description logic reasoning algorithm based on ontology, temporal reasoning algorithm, spatial reasoning algorithm and spatio-temporal accident tree reasoning early warning algorithm, which provide support for intelligent early warning of accident. The description of logic reasoning is based on Tableau algorithm, qualitative and quantitative reasoning based on time-constrained network, spatial topology, azimuth and metric relationship reasoning based on combinatorial table, based on the existing qualitative and quantitative algorithms of accident tree. The qualitative and quantitative inference calculation of spatio-temporal accident tree is realized by using spatio-temporal constraint description. Finally, a four-story structure of Wanglou coal mine accident intelligent early warning platform is designed and implemented, and the ontology based coal mine accident warning knowledge base model is applied in practice. The application shows that the ontology based model of coal mine accident warning knowledge base is reasonable in theory and effective in practice, which can provide a useful reference for the construction of mine accident early warning knowledge base.
【學位授予單位】：中國礦業(yè)大學
【學位級別】：博士
【學位授予年份】：2014
【分類號】：TD76

【參考文獻】

相關(guān)期刊論文 前10條

1 鄭晶;;礦山安全預警信息系統(tǒng)的研究[J];長春工程學院學報(自然科學版);2011年01期

2 王保云;;物聯(lián)網(wǎng)技術(shù)研究綜述[J];電子測量與儀器學報;2009年12期

3 劉小生;孫群;;礦山安全預警專家系統(tǒng)知識庫的研究[J];礦業(yè)安全與環(huán)保;2008年03期

4 劉永強;劉志輝;;新疆融雪洪水預警決策支持系統(tǒng)研究[J];干旱區(qū)資源與環(huán)境;2007年02期

5 沈建新;;物聯(lián)網(wǎng)技術(shù)及其在監(jiān)管場所中的應用[J];警察技術(shù);2010年05期

6 李善平,尹奇椺,胡玉杰,郭鳴,付相君;本體論研究綜述[J];計算機研究與發(fā)展;2004年07期

7 王琦,黃繼鴻;基于MAS的企業(yè)預警支持系統(tǒng)設(shè)計[J];計算機工程與應用;2003年35期

8 翟東升;黃焱;;預警指標體系的本體建模及其應用[J];計算機工程;2008年21期

9 鞠可一;周德群;吳君民;;石油價格波動對經(jīng)濟影響的預警知識庫構(gòu)造[J];能源技術(shù)與管理;2011年05期

10 蔣國瑞;李慧;顧傳龍;;汽車環(huán)境污染預警機制研究[J];計算機應用研究;2010年01期

相關(guān)博士學位論文 前3條

1 戴小鵬;知識網(wǎng)格及其在農(nóng)業(yè)生物災害預警中關(guān)鍵技術(shù)研究[D];湖南農(nóng)業(yè)大學;2010年

2 云紅艷;設(shè)備功能視點下的海洋生態(tài)本體構(gòu)建及應用研究[D];中國海洋大學;2012年

3 劉年平;煤礦安全生產(chǎn)風險預警研究[D];重慶大學;2012年

本文編號：2302546