Physics-based machine learning method and the application to energy consumption prediction in tunneling construction. (August 2022)

• Record Type: Journal Article
• Authors: Zhou, Siyang; Liu, Shanglin; Kang, Yilan; Cai, Jie; Xie, Haimei; Zhang, Qian
• Is Part Of: Advanced engineering informatics. Volume 53(2022)
• Publication Date: 2022-08
• Subjects: Physics-based machine learning -- Domain knowledge -- Causality -- Energy consumption -- Tunneling machine; Computer-aided engineering -- Periodicals; Engineering -- Data processing -- Periodicals; 620.00285
• DOI: 10.1016/j.aei.2022.101642
• Languages: English
• ISSNs: 1474-0346

Abstract

Highlights: Method representing causality in machine learning for intelligent control is given. Physical mapping relation is encoded into learning process by dimensional analysis. Traceable model with physical insight for control parameter prediction is realized. It achieves higher accuracy while alleviating the training sample size dependency. It is verified on the energy consumption prediction during tunneling construction. Abstract: Representing causality in machine learning to predict control parameters is state-of-the-art research in intelligent control. This study presents a physics-based machine learning method providing a prediction model that guarantees enhanced interpretability conforming to physical laws. The proposed approach encodes physical knowledge as mapping relationships between variables in engineering dataset into the learning procedure through dimensional analysis. This derives causal relationships between the control parameter and its influencing factors. The proposed machine learning method's objective function is further improved by the penalty term in the regularization strategy. Verifications on the energy consumption prediction of tunnel boring machine prove that, the established model accords with basic principles in this field. Moreover, the proposed approach traces the impact of three major factors (structure, operation, and geology) along the construction section, offering each component's contribution rates to energy consumption. Compared with several commonly used machine learning algorithms, the proposed method reduces the need for large amounts of training data and demonstrates higher accuracy. The results indicate that the revealed causality and enhanced prediction performance of the proposed method advance the applicability of machine learning methods to intelligent control during construction.