PSO-based Elman neural network model for predictive control of air chamber pressure in slurry shield tunneling under Yangtze River

Abstract

Abstract The excavation face stability is crucial for safety and risk management in slurry shield tunneling, especially for the river-crossing tunnel. To avoid face collapse or blow-out, shield operators need to keep air chamber pressure balanced using their own experience, which would be difficult, discontinuous and less reliable in the process of construction. Considering the disadvantage of the manual control process, this paper presents a predictive control system for air chamber pressure in slurry shield tunneling using Elman neural network (ENN) model. It mainly contains a theoretical model, an ENN predictor and an ENN controller to set optimal control parameters automatically tracking the desired air chamber pressure. Moreover, to improve the learning capability of ENN model, a particle swarm optimization (PSO) algorithm is implemented. This system has been tested with collected data of slurry shield operation parameters in the Yangtze riverbed metro tunnel project in Wuhan, China. Analysis revealed that the predictive control system using PSO-based Elman neural network model in this paper has the potential for enhancing face stability in slurry shield tunneling.