In 1993, the author proposed the idea for the development of intelligent rock mechanics, after his realization of the possibility of integration of artificial intelligent computing science and system science with rock mechanics and engineering. In this paper, the author tries to summarize developments and applications of the intelligent rock mechanics that have been realized by him and his research team recently. They include:
fundamentals of the intelligent rock mechanics and engineering;
expert systems for rockburst risk assessment, rock mass classification, rock mass support design, stability evaluation of slopes and underground caverns/tunnels;
neural network models and support vector machines for stability analysis of slope and underground caverns;
intelligent back analysis algorithms for recognizing rock mass mechanical and creep parameters;
strategy and algorithms on intelligent recognition of rock mass models and mechanical parameters;
intelligently and dynamically optimum design of excavation procedure and support system for slope, tunnel, cavern and underground powerhouse;
applications of the intelligent rock mechanics methods to several typical large rock engineering projects in China. At the end, the author points out the further development of the intelligent rock mechanics in the coming years.
Due to lack of land and requirements of resources and infrastructure, more and more large rock engineering projects are or will be constructed in complicated geological conditions with in high geo-stress, high water pressure and tectonic zones. They bring challenges and opportunities for researchers and engineers to develop suitable analysis methods and tools in rock mechanics. Artificial intelligence, with integration of other multi-disciplines, is capable to provide quantitative solutions to the challenging problems. Its foundations are self-learning and intelligent recognition of models, dynamic representation of nonlinear relationship, global optimization with their integration of mechanical analysis. A realization of the challenges and the modern high technology based possible solutions enabled the author to initiate and develop an innovative approach named as intelligent rock mechanics. The author reported this innovative idea in China in 1993, then internationally in 1997 [1] and gave further discussions in 2000 [2]. Since then, the intelligent methods have become one of the majority at the non 1:1 mapping level and also integrated with other methods at the 1:1 mapping level (see Fig.1) [3,4]. With time, development of the intelligent rock mechanics has covered the following four main parts. The first part is the application of expert systems to the field of rock mechanics and engineering starting in 1980s, which is very useful for utilization of expertise and rock engineering experiences. Some typical expert systems for rock classification, support design and risk assessment for slope and tunnels have been developed. The second is the learning and representation of nonlinear relationship for rock mechanics problems based on neural networks. Some typical neural network models have been developed for rock slope stability analysis, rockburst risk assessment, and mine subsidence prediction.
