Recognition of formation characteristics based on vibration signals in shield tunneling

Abstract

Shield machines are widely used in the construction of urban subway tunnels. Clear geological conditions are the prerequisite for safe and efficient shield tunneling. The vibration signal generated by the disc cutter cutting the rock contains rich information during the shield tunneling process. This paper analyzes the basic characteristics of vibration generated by the interaction between the disc cutter and the rock through theoretical calculation and numerical simulation. By measuring the shield vibration data under different surrounding rock strength on site and extracting the main characteristic indicators, a method is established relying on the BP neural network model for identifying the surrounding rock strength of the excavation face, which is characterized by the time domain-frequency domain of the vibration signal and successfully applied in the Jinan Metro Line 6 project. The research results show that the vibration frequency of the disc cutter force is concentrated at 0–25 Hz and 40–60 Hz, and the rock strength mainly affects the vibration amplitude of the disc cutter. The vibration signal characteristics generated during shield tunneling are highly sensitive to the surrounding rock strength, which is manifested as the greater the surrounding rock strength, the more obvious the time domain characteristic response of the vibration signal, and the more concentrated the main frequency of the IMF component. The accuracy of the model established in this paper for identifying the surrounding rock strength of excavation face is 98.88 %, and good application effects have been achieved in engineering.