隧道磁共振测深超前探测模型参数不确定度分析方法研究

地面磁共振方法探测地下水趋于成熟.随着研究的深入，磁共振技术在隧道超前探测方面也开展了应用.然而，由于隧道空间特殊环境限制，获取的磁共振信号信噪比极低，解释结果中各参数的确定性值得深思.基于这一问题，本文提出隧道磁共振测深超前探测模型参数不确定度分析方案，实际工作前，根据不同探测目标要求及环境噪声水平，优化仪器装置参数设计，提高解释结果准确性.本文首先在地面磁共振探测理论基础上，推导了考虑天线铺设角度影响的隧道矩形线圈激发场计算表达式，模拟了隧道准全空间磁共振测深正演响应.其次，基于后验模型协方差矩阵，计算模型参数标准偏差因子，划分参数不确定度等级.最后，构建三层含水模型，将第二层含水体作为观测目标.在仿真合成数据的基础上，分别探讨了电阻率、含水量、水体厚度、线圈边长、匝数、线圈旋转角度以及噪声水平等参数对目标含水体测定的影响.通过对比分析，得到如下结论：当探测目标前方地层的电阻率小于10 Ωm时，目标含水体的不确定度随着该电阻率的增大而降低；当该电阻率大于10 Ωm时，其不影响目标含水体的不确定度；目标体前方地层含水量的增大能够明显增加目标含水体的不确定度；目标层电阻率以及含水量对该层含水体的不确定度几乎不造成影响；目标层厚度越大其含水体的确定度越高；线圈边长和匝数的增大都能在很大程度上降低含水体的不确定度；线圈的偏转角度不影响目标体的不确定度；磁共振信号中噪声的幅度越大，含水体参数的不确定度越大.本文的研究结论有助于提高隧道磁共振探测数据反演参数的准确性，同时能够为实际探测提供预先优化参数的分析方案.

Study of model parameter uncertainty for underground magnetic resonance sounding in tunnel detection

Surface magnetic resonance sounding (MRS) for the detection of groundwater is becoming mature. With the deepening of the research, MRS technique has been applied in advanced detection in tunnel. However, due to the special environmental constraints of the tunnel space, the acquired signal-to-noise ratio of the signal is extremely low. Moreover, it is essential to assess the uncertainty of the parameters in the detection model. Therefore, it is necessary to analyze the uncertainty of the parameters of the MRS model. Based on this problem, we presented a parameter uncertainty analysis scheme for tunnel advanced detection model. Before the actual work, according to different detection target requirements and environmental noise level, optimizing the instrument device parameter design can improve the accuracy of interpretation results. Based on the surface MRS detection theory, we derived the calculation formula of the tunnel rectangular coil excitation field considering the influence of the antenna laying angle, and simulates the tunnel quasi-whole space MRS forward response. Secondly, we calculated the standard deviation factor of the model parameters based on the posterior model covariance matrix, and classified the uncertainty levels of the parameters. Finally, we show the results for different set of three-layer models, in which the second layer ahead of tunnel-driving was used as the target layer. Based on the synthetic simulated data, the effects of parameters such as resistivity, water content, layer thickness, coil side length, number of coil turns, coil rotation angle and noise level on the determination of water body in target layer were considered. Through comparative analysis, we conclude that, in general, the uncertainty of the water body in target layer decreases with the increase of the resistivity of the strata in front of the target layer when it is less than 10 Ωm. However, the uncertainty of the water body in target layer is not affected by the resistivity of the strata in front of the target layer when it is greater than 10 Ωm. The uncertainty of target water body can be obviously increased by increasing the water content of the layer in front of the target. The resistivity and water content of the target layer have little influence on the uncertainty of the target. The greater the thickness of the target layer, the higher the certainty of the target water body. The uncertainty of water body can be reduced greatly with the increase of coil side length and turns. The rotation angle of the coil has little effect on the uncertainty of the target water body. The increase of noise amplitude in signal will increase the uncertainty of water body parameter. The conclusion of this paper is helpful to improve the accuracy of inversion parameters of tunnel MRS data, and can provide an analysis scheme of pre-optimized parameters for practical detection.