Developed since the beginning of 1990s, intelligent rock mechanics methodology has been in good progress and taken important role in solving the complicated rock engineering problems. This paper starts with summary of basic ideas of intelligent rock mechanics methodology which is integration of artificial intelligent methods, empirical methods and mechanical analysis methods. The progress of intelligent rock mechanics methodology to rock engineering problems is mainly introduced. It includes the intelligent stability analysis, failure mode recognition, excavation optimization and dynamic design for large rock slope, deep tunnels and cavern group. Three examples related to dynamic feedback stability analysis and design optimization are given. The further development for intelligent rock mechanics methodology is discussed.
In order to overcome difficulties in rock mechanics modeling, rock engineering design, and stability analysis of large rock engineering under complicated geological conditions, new ideas of intelligent rock mechanics has been proposed since beginning of 1990s [1–4], which is to integrate artificial intelligence methods to perform rock mechanics and engineering modeling and analysis. It starts with development of expert systems and neural network models for rock mechanics and engineering problems. Now, it has been extended to methods for
Intelligent bask analysis of rock mechanical parameters,
Intelligent recognition of nonlinear mechanical models of rock masses with integration of intelligent methods and numerical analysis,
Global optimization of excavation procedure and support design of tunnels, cavern group and slopes,
Intelligent stability analysis considering multi-step excavation influence on damage evolution of rock masses,
Dynamic feedback analysis and dynamic optimal design of larger rock engineering projects under complicated geological conditions,
Intelligent recognition of failure modes of tunnels, cavern group and slopes,
The deformation warning classifications during rock engineering excavation of multi-steps, and
The modern rock engineering design.
Some progress of intelligent rock mechanics method is discussed in reference . Furthermore, this paper initially gives review of basic ideas of intelligent rock mechanics methodology and discusses progress in applying intelligent rock mechanics methodology to solve complicated rock engineering problems. It covers mainly stability analysis and dynamic optimal design of large rock engineering at complicated geological conditions.
There are some basic ideas for intelligent rock mechanics methodology as follows:
Self learning and representation of nonlinear relationship and models.
In case of difficult understanding on mechanism of deformation and failure of rocks and rock masses, the structures of nonlinear models are sometimes hard to be determined by using deterministic mechanical analysis and represented by using mathematical equations. There is a lot of empirical knowledge which are not easily represented by qualitative models. Three different ways were developed for recognition of nonlinear rock mechanics models.
Artificial intelligent models, such as neural networks and support vector machines, are attractive for the learning and representation of these nonlinear models.
For examples, neural network models for rock classification, tunnel support design, recognition of slope angles and nonlinear subsidence time series [6–8], etc. were established respectively.