为实现TBM施工隧道的实时地质预报，以TBM刀盘滚刀破岩震动作为震源的HSP法地质预报技术为基础，通过分析TBM施工工艺与机身结构特点，对HSP系统硬件和软件进行优化设计，使其小型化、自动化和智能化，并搭载于TBM上进行智能控制。在TBM掘进过程中，首先，通过隧道轮廓位置的检波器连续或高频次地采集地震反射波信号； 然后，经系统软件对数据的自动处理，实时获取掌子面前方地层反射特征参数图谱； 其次，通过智能识别技术完成对不良地质反射界面的有效拾取与判识； 最后，实现不良地质的探测。通过对适于TBM施工的HSP法实时预报技术进行优化设计，实时探查前方地层不良地质(体)位置、规模与性质，为TBM的高效施工起到一定的推助作用。

Based on the HSP geological prediction technology which takes the rock breaking vibration of TBM cutterhead disc cutter as the source, the hardware and software of HSP system are optimized for miniaturization, automation and intelligence by analyzing the construction technology and structure characteristics of TBM. The system is carried on TBM for intelligent control. During the TBM tunneling process, firstly, seismic reflection signals are collected continuously or frequently by geophones located in tunnel contours; secondly, the reflection characteristic parameters of stratum ahead of the tunnel face are acquired in real time through automatic processing; thirdly, the effective pickup and identification of unfavorable geological reflection interfaces are completed by intelligent recognition technology; finally, the prediction of unfavorable geological conditions is realized. By optimizing the HSP realtime prediction technology suitable for TBM construction, the position, scale and nature of the unfavorable geology ahead of the tunnel face are investigated in real time, which plays a certain role in promoting the efficient construction of TBM.