Abstract:

Presently, the tunneling control of tunnel boring machine(TBM) mainly depends on artificial experience without any type of scientific control strategy. The control characteristics of the manequipmentrock closedloop system are examined. In addition, the control rules and treatment of conventional and adverse strata are summarized and the control intention of the main TBM operator and parameter control method under various strata are described. According to the differences of control strategies in various surrounding rock grades, a multimodal control strategy and a prediction method of tunneling parameters based on the surrounding rock classification of TBM  are put forward. The proposed methods are applied to the TBM 3 section of the Songhua river water conveyance project in Jilin, China. The research results show that the accuracy of the prediction model of TBM tunneling parameters based on a multimodal control strategy is more than 88% and the prediction accuracies of cutterhead and tunneling speeds are 94% and 88%, respectively, thus meeting the engineering requirements.

Key words:

tunnel boring machine, tunneling control, surrounding rock classification, tunneling parameter prediction, driving behavior learning