AUTHOR=Zhang Qinglong , Zhu Yanwen , Ma Rui , Du Canxun , Du Sanlin , Shao Kun , Li Qingbin TITLE=Prediction Method of TBM Tunneling Parameters Based on PSO-Bi-LSTM Model JOURNAL=Frontiers in Earth Science VOLUME=10 YEAR=2022 URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2022.854807 DOI=10.3389/feart.2022.854807 ISSN=2296-6463 ABSTRACT=

With the wide application of full-face rock tunnel boring machine (TBM) in tunnel construction, the self-adaptive adjustment of TBM tunneling parameters is of great significance for the safety and efficiency of TBM tunnelling. Aiming at the shortcomings of the current TBM data mining capability and optimization methods of tunneling parameters, this paper proposes a prediction method of TBM tunneling parameters based on particle swarm optimization-bi-directional long short-term memory (PSO-Bi-LSTM) model, which selects the complete tunneling cycle data to predict the TBM tunneling parameters, and uses a number of numerical methods such as binary state discriminant function and 3σ criterion to preprocess the operation data of TBM3 bid section of Songhua River water conveyance project. By comparing with the Bi-LSTM model and evaluating the prediction effect under different surrounding rock levels, the applicability and prediction performance of the model to different strata are verified. The results show that the prediction accuracy of the model is proportional to the surrounding rock grade. Compared with the Bi-LSTM, the overall prediction effect of the proposed PSO-Bi-LSTM model is better, which can assist the intelligent construction of TBM with similar geological conditions.