Assessment of Dynamic Pile Driving Using Machine Learning

本项目旨在通过严格的数值模拟、室内试验和基于人工智能的计算模型，开发一种在打桩过程中实时确定地基性质和地基承载力的新技术。虽然冲击打桩通常用于现场打桩，但目前还没有技术可以准确和实时地解释收集到的数据，以提供实时反馈和优化施工过程。这项研究将提供新的机器学习模型来评估施工过程中的地基和地基特征，并将增加澳大利亚基础设施投资的确定性，特别是昂贵的交通资产和基础设施。

This project aims at developing new technology to determine ground properties and foundation capacity in real-time during pile installation by adopting rigorous numerical simulation, laboratory experiments and artificial intelligence-based computational model. Although impact driving is used commonly to install piles on site, there is no technology currently available to interpret collected data accurately and in real-time to provide live feedback and optimise construction processes. This research will provide new machine learning model to assess the ground and foundation characteristics during construction, and will increase certainty in infrastructure investment in Australia particularly for costly transport assets and infrastructure.