Performance prediction and sensitivity analysis of tunnel boring machine in various geological conditions using an ensemble extreme learning machine

Abstract

The selection of data modelling methods in the data-driven performance prediction of tunnel boring machines is a challenge since each method has its own advantages and disadvantages compared with each other. Extreme learning machine (ELM) exhibits the benefits of fast learning speed, better scalability, and generalization performance, and is easy to convert between neural networks-based and kernel function-based methods. Thus, this paper proposes an ensemble extreme learning machine model for the performance prediction of tunnel boring machines, aiming to take respective advantage of different ELM models. The proposed model is validated through six in-situ datasets of a tunnel boring machine with different geological conditions, showing that it can produce accurate dynamic and statistical performance prediction results (average error of 3.12 %). The sensitivity analysis results show that the sensitivities are mainly distributed on the parameters of driving system and chamber system when the excavation face is occupied by a single geological layer.