基於DTW的瓦斯爆炸災害風險模式識別

摘　要:以煤礦瓦斯爆炸災害防控為目的，構建了瓦斯爆炸災害風險模式識別指標體係，提出了以指標含義與值域設置風險模式類別分界點的區間線性化數據規範方法，減少了計算過程中的信息量損失.通過區間數方法對指標進行賦權後，應用動態時間彎曲算法(DTW)，以指標風險模式類別分界點為基礎形成了多個加權參考序列，並將待識別樣本轉換成待識別加權序列，通過待識別加權序列與加權參考序列的最短動態彎曲距離獲得待識別樣本的風險模式類別，建立了基於DTW的煤礦瓦斯爆炸災害風險模式識別模型.算例結果符合實際，該方法對瓦斯爆炸災害風險模式具有良好的識別能力.

關鍵詞:瓦斯爆炸 風險 模式識別 動態時間彎曲 煤礦

分類號:　TD712.7Research on patter recognition of gas explosion disaster risk based on dynamic time warping in coal minesLI Runqiu, SHI Shiliang, LUO Wenke, WU Aiyou ( 1. School of Energy and Safety Engineering, Hunan University of Science and Technology, Xiangtan 411201, China; 2. Hunan Provincial Key Laboratory of Safe Mining Techniques of Coal Mines, Hunan University of Science and Technology, Xiangtan 411201, China)Abstract:For the prevention and control of gas explosion disaster risk in the coal mine, the risk index system was established to patter recognition of gas explosion. Interval linear data normalization method was put forward to reduce the information loss in the process of computing by using the risk patter cut - off of indexes. The risk index was weighted with interval data of multiple comparison method. According to dynamic time warping (DTW) , the 5 weighted reference sequence was constructed by using the indexes risk patter cut - off which was normalized, and the weighted sequence was constructed to be identified for the unknown risk patter sample. And then, disaster risk pattern recognition model was established based on DTW which obtaining risk pattern by the shortest dynamic time warping distance of the weighted reference sequence and to be identified weighted sequence. The pattern recognition example was gave that based on 14 indicators of gas explosion disaster risk, and the risk identification results were in good agreement with the actual situation. Results show that, DTW model of risk patter recognition has good identification ability for explosion disaster risk model, which will be a new way to study the pre - control of gas explosion disaster risk in coal mines.

Keyword:gas explosion; risk; patter recognition; dynamic time warping (DTW) ; coal mine