Study on special problems such as high and steep slope stability, landslide prediction and ground settlement.

(1) At present, the search for mineral resources is from shallow to deep, and in mine development, open-pit mining has turned to deep-buried deposits. At the same time, some old open-pit mines with long mining time have gradually turned to deep-concave mines, so the stability of high and steep slopes has become a prominent problem in deep-concave open-pit mining. Usually, a high and steep slope refers to a slope with a height of more than 500m and a slope of more than 55. China has made many theoretical achievements and engineering practice in the study of slope stability in open pit mines. Based on the practice in Jinchuan Mine in 1970s, the principle and method of slope stability criterion are established, that is, taking the structural characteristics of rock mass and geological structure control as the key link, stereographic projection and stereo proportional projection are used as simple action analysis methods of slope stability. The geological model of rock slope deformation and failure was established in 1980s, and the importance of engineering geological work and comprehensive evaluation of various analysis and evaluation methods were emphasized in 1990s. Breakthrough progress has been made in high and steep slope investigation (core directional drilling), model establishment, physical and numerical simulation (such as geomechanics physical simulation and finite element, boundary element and discrete element coupling analysis), slope reinforcement and so on. In recent years, fractal theory has been applied to the study of slope stability, which has practical significance for the analysis of slope deformation and failure mechanism and the quantitative description of geological model.

The stability of high and steep slope has its particularity: the rock group and rock mass structure of the slope are complex, and the stress state of the slope is complex due to the influence of geostress and groundwater. Therefore, in order to further study the stability of high and steep slopes, we should (1) infer the characteristics and distribution of underground structures, especially weak structural planes, from the relationship and laws between the surface and small and medium-sized underground structures; ② Study the combination of jointed rock mass and weak structural plane, and establish appropriate geological model and mechanical model; ③ Environmental factors of rock mass mainly refer to the influence of ground stress, groundwater and human engineering activities (excavation and mining); ④ Stress state of slope rock mass; ⑤ In-depth study on five key issues, such as slope deformation and failure mechanism analysis, made a breakthrough (Bing Xu et al., 1992).

(2) The prediction of landslide occurrence time, landslide activity intensity and landslide risk, just like earthquake prediction, is also very difficult because of the complexity of inducing factors and the randomness and uncertainty of changes. So far, there are still many key problems that need to be explored and improved. At present, successful landslide prediction methods are mainly based on geological analysis and empirical judgment, as well as quantitative trend prediction based on monitoring data. For example, the phenomenon prediction of landslide deformation precursor, the trend judgment of displacement-time curve and the prediction method with strain rate as the basic parameter. These methods are effective for predicting impending slip. Statistical mathematical simulation method, golden section method developed in 1980s and nonlinear dynamic model prediction method proposed in 1990s are mainly used for inversion fitting prediction of landslides that have occurred, and the results have been well verified afterwards. For example, using fractal theory, various fractal dimensions are obtained, and it is found that the corresponding relationship between information dimension, capacity dimension and correlation dimension and landslide evolution stage has certain laws. Therefore, "based on geological research, it is still the goal of landslide sliding time prediction research in the future to explore the macro-geometric law, micro-physical and chemical law and mathematical statistics law of landslide deformation." There are two methods to predict landslide activity intensity, including sliding speed and sliding distance, particle kinematics and the combination of landslide motion mechanism research and block kinematics, but they are not mature enough. "How to apply the physical theory of particle motion and the law of conservation of energy conversion to the study of landslide motion characteristics more reasonably is the main research trend in the future" (Wen Baoping, 1996).

(3) The research on land subsidence at home and abroad began in 1950s and 1960s, and four academic seminars on land subsidence have been held so far. China has reached a high level in studying the mechanism and controlling land subsidence, and is at the forefront of the world. At present, Shanghai and Tianjin have entered the stage of dynamic micro-settlement control (the annual settlement is about 1 ~ 2 cm). In China, the geological causes of land subsidence can be divided into five kinds of subsidence geological models and investigated according to their different characteristics. The research on settlement mechanism mainly focuses on the restrictive relationship between the quantitative characteristics of microstructure and main engineering geological properties of cohesive soil and the migration of pore water in cohesive soil. Changchun Institute of Geology put forward that the essence of pore water permeability law of cohesive soil is "involving the mutual transformation of gravity water and bound water in two pore channels". The important progress in the calculation and prediction of land subsidence is to consider the occurrence of overflow (the water level model is established under the assumption that there is no overflow recharge), and to establish a mathematical model by coupling the three-dimensional water flow model with the one-dimensional consolidation settlement model under the condition that the water released by cohesive soil recharges the aquifer. The practical application effect is good.

Experts believe that the development trend of land subsidence research in the future is to further introduce nonlinear science, focus on the migration law of pore water in cohesive soil, and explore the relationship between soil structure changes, pore water migration and physical and mechanical properties changes, so as to obtain the best groundwater pumping and irrigation scheme to maximize the rational exploitation of groundwater resources, and make the research and control of land subsidence in China reach a higher level (Yan Shijun et al., 1996).