Abstract
Earth chamber pressure is one of the key parameters for controlling ground settlement. Based on extensive real-time data collected during shield tunneling construction, this paper proposes an earth chamber pressure prediction model for Earth Pressure Balance (EPB) shield construction using the "Random Forest" algorithm, which provides reasonable setpoint values for earth chamber pressure during shield advancement to achieve settlement control. Application results from the left-line section between No. 7 ventilation shaft and Lingkong Road transition shaft of JCXSG-12 bid of Shanghai Metropolitan Railway Airport Link Line demonstrate that the setpoint values provided by the model are consistent with the actual control values, and ground settlement in the test section is well controlled. This indicates that the model can provide relatively reasonable setpoint values for earth chamber pressure during shield advancement to guide construction.
Full Text
Preamble
Application Research on Soil Chamber Pressure Prediction in Shield Tunneling Based on Random Forest
Dong Peng¹
¹ Shanghai Tunnel Engineering Co., Ltd., Shanghai 200000, China
Abstract
Soil chamber pressure represents a critical parameter for controlling ground settlement in shield tunneling operations. This study proposes a soil chamber pressure prediction model for earth pressure balance (EPB) shield construction based on the random forest algorithm and extensive real-time data collected during tunneling. The model aims to provide reasonable soil chamber pressure setpoints during shield advancement to effectively control settlement. Application results from the left line section between Ventilation Shaft No. 7 and Lingkong Road Transition Shaft of Contract JCXSG-12 on the Shanghai Metropolitan Railway Airport Link Line demonstrate that the pressure setpoints generated by the model align closely with actual control values, with satisfactory settlement control achieved in the test section. These findings indicate that the model can provide relatively reasonable soil chamber pressure setpoints to guide construction throughout the shield advancement process.
Keywords: shield tunneling construction; parameter prediction; random forest; machine learning