论文部分内容阅读
矿井瓦斯监测数据变化趋势对于分析瓦斯是否异常至关重要,通过回顾和整理国内外研究成果,系统总结了矿井瓦斯时间序列自身所具的趋势性、相关性、周期性及异常性4种特性。归纳出煤与瓦斯突出、炮后瓦斯、局部停风、传感器探头校验以及瓦斯探头故障5种瓦斯异常模式,并综述了灰色系统法、神经网络法、支持向量机以及其他方法在矿井瓦斯预测方面的研究成果。最后总结了矿井瓦斯趋势研究存在的问题和不足,并展望了K线理论的应用前景。
The trend of mine gas monitoring data is very important for the analysis of whether gas is abnormal or not. By reviewing and finishing the research results at home and abroad, systematically summarizes the four characteristics of trend, correlation, periodicity and abnormality of mine gas time series. Five kinds of gas anomalous modes such as coal and gas outburst, post-mortem gas, partial stop wind, sensor probe verification and gas probe fault are summarized. The gas system prediction of mine gas including gray system method, neural network method, support vector machine and other methods The research results. Finally, the problems and deficiencies in mine gas trend research are summarized, and the application foreground of K-line theory is prospected.