多场多属性信息融合精细定位煤矿导水通道技术研究

Multi-field and multi-attribute technical information fusion for accurate positioning of coal mine water diversion channel

在煤矿发生重大安全事故中,水害事故造成的人员伤亡和经济损失较大。煤矿水害防治工作的关键是防治导水通道,如何提高陷落柱等导水通道探查和识别的准确性也是目前煤矿安全生产研究的热点。以黄玉川煤矿X6疑似陷落柱探查为例,通过提取三维地震和瞬变电磁探测中对导水陷落柱敏感的属性特征,利用粒子群算法优化的BP神经网络进行多场多属性信息融合技术精细定位导水通道,结合地质资料及钻孔工程探测,验证了该方法探查陷落柱的准确性,解决了陷落柱等导水通道精细定位的问题,为煤矿安全高效开采提供了重要保障。

Casualties and economic losses due to water inrush accidents are greater than commonly recognized in large safety accidents of coal mines.Water hazard prevention and control in coal mines is necessary to establish prevent and control water diversion channel.Improving the accuracy of detection and identification of water diversion channel such as collapse column is also the essential of coal mine safety production research at present.Taking the exploration of X6 suspected collapse column in Huangyuchuan Coal Mine as an example,by extracting the attribute characteristics sensitive to water diversion collapse column in 3D seismic and transient electromagnetic detection,technical information fusion was used,based on the BP neural network optimized by particle swarm optimization algorithm,to accurately position the water diversion channel.Combining the geological data with borehole radar detection,the accuracy of positioning technology was verified.The water diversion channels such as collapse column were accurately positioned.