Reinforcement Learning for the Face Support Pressure of Tunnel Boring Machines

In tunnel excavation with boring machines, the face is supported to avoid collapse and minimise settlement. This article proposes use of reinforcement learning, specifically deep Q-network algorithm, predict support pressure. The algorithm uses a neural network make decisions based on expected rewards each action. approach tested both analytically numerically. By using soil properties ahead overburden depth as input, capable predicting optimal pressure whilst minimising settlement, adapting changes in geological geometrical conditions. reaches maximum performance after 400 training episodes can be used for random settings without retraining.