基于贝叶斯网络的软岩巷道大变形影响因素分析

随着煤炭资源开采的深度和强度的大幅度增加,软岩条件巷道的维护出现了一系列新的控制难题,岩石性质、地应力类型、采场采动、含水层、巷道形状尺寸以及支护方案参数的选择等诸多因素耦合作用影响着软岩的位移量及巷道围岩应力场分布规律。为了实现对巷道围岩状态的实时定量化控制,将巷道监测数据通过全断面扫描仪等传感设备实时传入集控中心,利用贝叶斯神经网络方法进行软岩大变形相关影响因素参数分析,将巷道分段确定导致大变形的最高权重因素,并针对巷道断面形状、支护参数、支护方案以及回采工艺参数四个方面给出及时调整的最优方案。形成监测传感器-设备列车-集控中心-贝叶斯神经网络分析-方案定制-加强支护控制围岩的闭环机电智能化系统。可为类似条件下软岩巷道大变形机理与控制技术研究提供参考。

Study on large deformation mechanism and control system of soft rock roadway based on Bayesian method

With the increase of the depth and intensity of coal resources mining, a series of new problems have emerged in the maintenance of roadway under soft rock conditions, such as rock properties, in-situ stress types, stope mining, aquifer, roadway shape and size, and the selection of support scheme parameters. Many factors affect the displacement of soft rock, and the distribution law of stress field in roadway surrounding rock has new characteristics. In order to realize the real-time quantitative control of the surrounding rock state of the roadway, it is bound to transmit the roadway monitoring data to the centralized control center in real time through the full section scanner and other sensing equipment, analyze the influencing factors and parameters related to the large deformation of soft rock by using the neural network method, determine the highest weight factor leading to the large deformation in sections of the roadway, and according to the roadway section shape, support parameters The optimal scheme for timely adjustment is given from four aspects: support scheme and mining process parameters. Form a closed-loop electromechanical intelligent system of monitoring sensor - equipment train - centralized control center - Bayesian neural network analysis - scheme customization - strengthening support and controlling surrounding rock. It can provide reference for the study of large deformation mechanism and control technology of soft rock roadway under similar conditions.