Development of the method for predicting the behavior of rock structures by combining numerical simulation and AI

Details of research

Numerical simulation is commonly used to predict and understand the behaviour of rock structures such as rock slopes and underground cavities. In numerical simulation, a numerical model of the rock structure of interest is created, and information about geological structure and various rock properties (mechanical, hydrological, and thermal properties) needs to be input into the model. However, obtaining the accurate information from surveys and experimental results is not straightforward. Consequently, methods such as back analysis have been used to estimate rock properties, but this approach is not applicable in cases of complex geological structures.

In our laboratory, we are developing a new method to predict future rock behaviour by analysing and modelling field measurements of rock displacements, such as those obtained from GPS, using AI. This method does not require knowledge of geological structures or rock properties, making it well-suited for predicting rock behaviour. Additionally, we are also developing a method that estimates rock behaviour using machine learning with numerical simulation results as training data

Defromation Measurement and Numerical Simulation in an Open Pit Mine