Abstract: The scale and difficulty of shield tunneling construction are becoming increasingly complex. The existing advance detection methods are subject to many limitations in the shield tunneling environment with extremely small exploration space and severe electromagnetic interference. In order to improve the continuity, accuracy and anti-interference of advanced detection in complex environments, based on electrical resistance tomography technology, the formation resistivity data is obtained by excitation of shield cutterhead electrodes, and the 3D distribution of resistivity in front of the tunneling working face is reconstructed by the inversion algorithm to identify the unfavorable geological body. Firstly, a simulation model of the shield cutterhead with electrodes is constructed. The shield cutterhead consists of a hollow area and a main beam area. The electrode array is arranged on the main beam panel to avoid interference from pillars such as scrapers. Through numerical simulation analysis, a "cross-shaped" array with 13 electrodes is selected; secondly, six typical tunnel models are established, and abnormal bodies of different sizes and shapes are arranged at the proximal and distal ends of the tunnel face respectively, the underground resistivity distribution is restored from the inversion measurement data by the GREIT reconstruction algorithm, and 3D reconstruction imaging of the unfavorable geological structure in front of the tunneling working face is performed; finally, an experimental platform for advanced detection during tunneling was built, which consisted of a physical model with a geometric similarity ratio of 80 and an experimental measurement unit, and two different foreign objects were arranged at 150 mm and 300 mm away from the tunnel face, respectively; the excitation and measurement strategies can be flexibly switched through the self-encoding program, and the experiment excited and measured the potential difference through the electrode array. The excitation current was 1 A; the 3D resistivity distribution of the geological body was reconstructed through the inversion algorithm, and inversion imaging was performed to identify poor geological bodies. The experimental results show that this method can effectively detect poor geological body structures within the range of 1.5 to 3.0 times the cutterhead diameter in front of the tunnel face, and can better reconstruct geological anomalies with the sizes of 1/2 and 1 times the cutterhead diameter, the average position error (LE) and shape error (SE) are 0.08 and 0.54, respectively, which provides a new idea for continuous detection of shield tunnels during tunneling.