Optimal Control For Shield Machine Based On Multipoint Earth Pressure Balance

The earth pressure balance (EPB) shield is a large engineering machine especially used in underground construction, and it is the most advanced shield, widely used in metro tunnel, highway tunnel, hydraulic engineering and so on. The imbalance of earth pressure in front of the tunnel face can lead to serious accidents, such as ground subsidence and heave. Thus, the stability of the tunnel face is a critical factor to ensure the safety and efficiency of an excavating process. The random diversity of geology condition and the multi-field coupling make the physical and mechanical properties of the excavation face very complex and unpredictable, so in practical tunneling constructions the EPB control mainly depends on human experience. Thus, the automatic control of EPB for shield chamber is an important research field of shield technology. Based on the physical mechanism analysis, with advanced control and optimal method, this thesis studied the multi-points EPB control for shield chamber. The main research work is as follows:(1) Review the research background, history and the worldwide development status of shield technology and present the value of this study. Briefly introduce the structure and working principle of EPB shield and explain the steady mechanism of the tunnel face during construction process.(2) According to the mechanism analysis of the tunneling process, considering the coupling influence of the total thrust, the advance speed, the rotating speed of screw conveyor on the tunnel face steady status, a prediction model of multi-point earth pressure based on Adaptive Neuro-fuzzy Inference System (ANFIS) is established for an EPB shield with four earth pressure sensors on its chamber’s clapboard. This model takes the total thrust, the advance speed, the rotating speed of screw conveyor and multi-point real earth pressure as the input variables, and takes the next time multi-point earth pressures as the output variables. Based on practical data, a simulation analysis is made to illustrate that the multi-point earth pressure model can effectively predict the earth pressure.(3) By minimizing the absolute error between the prediction earth pressure and the set value of earth pressure as the optimization objective, the Ant Colony System (ACS) algorithm is employed to get the optimal screw conveyor speed to control the earth pressure balance during tunneling process. Practical data of two types of EPB shields with 4 pressure sensors and 5 pressure sensors respectively are used for simulation analysis. The simulation results show that the root mean square error between earth pressure and the corresponding set values are decreased by about ten times after optimization and validate the effectiveness of the multi-point EPB optimal control strategy.