【作者】 杨守国；

【导师】 唐建新； 文光才；

【作者基本信息】 重庆大学 ， 采矿工程， 2011， 博士

【摘要】 我国煤矿发生特别重大事故中,绝大多数是瓦斯爆炸事故,且不少重特大瓦斯爆炸事故都是由局部瓦斯事故在一定时间后引发瓦斯爆炸而形成的。本文针对这一情况,基于人工神经网络建立了瓦斯涌出异常类型识别模型,采用多传感器联合判识方法实现了突出地点判识和灾变影响范围解算;对灾害气体在通风系统中扩散运移规律及非稳态下通风网络仿真进行了研究,并将二者进行结合,建立了非稳态条件下风网及瓦斯分布解算模型;开发了一套矿井煤与瓦斯突出在线判识及应急辅助决策软件,能快速识别并判断瓦斯涌出异常事件发生地点、类型,自动分析、预测灾害气体波及范围,并提供应急辅助决策建议,实现了在线预警和可视化显示,可为灾变时期应急提供科学、快捷的辅助决策手段,具有创新性。论文取得的主要研究成果如下:①对各类瓦斯异常涌出特征进行分析,并从浓度曲线峰值、浓度上升阶段变化、浓度衰减阶段变化及瓦斯异常涌出持续时间,对各类瓦斯异常涌出进行了对比,得出了各类瓦斯涌出异常浓度曲线特征。②利用人工神经网络理论,建立了以浓度峰值、上升速度(%/min)、衰减速度(%/min)、衰减时是否有负指数特征作为判识指标的瓦斯涌出异常类型判识模型。③采用多传感器联合判识方法,实现了煤与瓦斯突出地点、事故扩散范围判识与计算。④将非稳态条件下通风网络解算模型和一维纵向移流扩散有限差分模型进行结合,建立了统一的计算模型,实现了煤与瓦斯突出过程中灾害气体运移情况动态仿真。⑤利用图论相关理论,建立了灾变已波及范围和将波及范围计算方法;给出了基于波及范围判定的自动断电策略和人员伤亡评估方法。⑥开发了一套矿井煤与瓦斯突出在线判识及应急辅助决策软件,能快速识别并判断瓦斯涌出异常事件发生地点、类型,自动分析、预测灾害气体波及范围,实现了在线预警和可视化显示,为矿井发生煤与瓦斯突出灾变后应急提供科学、快捷的辅助决策手段。更多还原

【Abstract】 Major accidents in coal mines in China are caused mostly by gas explosion, among which a number of serious ones result from local gas accidents after a certain amount of time. Under this situation,abnormal gas emission type recognition model has been established based on artificial neural network, while using multiple sensors to identify jointly,thus emission locations can be identified,and disaster effect range can be resolved; The diffusion and migration law of disaster gases in ventilation system and ventilation network simulation under unsteady condition has also been researched, through combination of which, solution model for ventilation network and gas distribution under unsteady condition has been established;In this way, a set of software on online identification of coal and gas emission and decision support under emergency has been developed, which can recognize and judge fastly the location and type of abnormal gas emission, analyze and forecast automatically the spread range of disaster gases, With the aid of advice and recommendation for decision support under emergency, online warning and visual of display can be implemented, providing a scientific and fast method for decision support under period of disaster emergency which has great innovation.Main researches are achieved in this essay as follows:①to analyze characteristics of various abnormal gas emission,comparing with peak values on the concentration curve, changes of concentration increased stage and concentration decay stage, duration of abnormal gas emission,thus to come up with features of curves for abnormal gas emission concentration.②using the theory of artificial neural network,to establish an identification model for abnormal gas emission with indications of peak value of concentration, ascent speed(%/min), decaying speed(%/min), whether there are negative indexes character while decaying.③adopting multiple sensors identification methods,to realize the identification and calculation for locations of coal and gas outburst, accident spread range.④combining ventilation network calculation model under unsteady conditions and finite - difference model of one dimensional vertical flow and migration, to establish a uniform calculation model, realizing dynamic simulation of disastrous gas migration in coal and gas outburst process. ⑤using graph theory, to establish calculation method for ranges of catastrophic gases spread and to be spread, thus to develop an automatic power shutoff policy and casualty assessment method based on identification of spread ranges.⑥develope a set of software on online identification of coal and gas emission and decision support under emergency, thus to recognize and judge fastly the location and type of abnormal gas emission, analyze and forecast automatically the spread range of disaster gases,realizing online early warning and visual display, and providing a scientific and fast method for decision support under period of disaster emergency which has great innovation更多还原