Uncertainty design optimization of the main bearing in tunnel boring machine based on the Kriging model with partial least squares

Abstract

During the tunnel boring machine work, the main bearing failure can cause great economic losses and safety hazards. However, the multiple uncertainties in the manufacturing, assembly and operation stage of the main bearing lead to the difficulty of its stable and reliable design, and the load-carrying capacity is hardly guaranteed. For that, an efficient uncertainty design optimization strategy for the main bearing is proposed by combining the Kriging model with partial least squares and a genetic algorithm. A five-degree-of-freedom static analysis model of the main bearing is established using the vector method to provide training samples and verified by the relative displacement test of the rings. The main influencing factors are screened based on sensitivity analysis. Considering uncertainties such as the structural dimensions, material properties and operating loads of the main bearing, a surrogate model is constructed to achieve an example study and compared with the initial design and deterministic optimization strategy. The results show that the fatigue life of the main bearing has increased by 43.01% compared to the initial design. The design robustness is improved and the design reliability is ensured compared to the deterministic optimization strategy. The method realizes the rapid acquisition of the optimal stable and reliable structure for the main bearing, which is an important reference value for the uncertainty design optimization for other types of large slewing bearings.