By applying the Wavelet Relevance Vector Machine (WRVM) method, this research proposes the loose zone of roadway surrounding rock prediction. Based on the theory of relevance vector machine (RVM), the wavelet function is introduced to replace the original Gauss function as the model kernel function to form the WRVM. Five factors affecting the loose zone of roadway surrounding rock are selected as the model input, and the prediction model of the loose zone of roadway surrounding rock based on WRVM is established. By using cross-validation method, the kernel parameters of three kinds of wavelet relevance vector machines (RVMs) are calculated. By comparing and analyzing the root mean square (RMS) error of the test results of each predictive model, the advantages and accuracy of the model are verified. In practical engineering applications, the average relative prediction errors of the Mexican relevance vector machine, the Morlet relevance vector machine and the difference of Gaussian (DOG) relevance vector machine models are accordingly 4.581%, 4.586% and 4.575%. The square correlation coefficient of the predicted samples is 0.95 > 0.9, which further verifies the accuracy and reliability of the proposed method.

To combat the uncertainty of the multiple factors affecting roadway surrounding rock stability, five initial indexes are selected for reduction according to concept lattice theory: rock quality designation (RQD), uniaxial compressive strength (Rc), the integrity coefficient of rock mass, groundwater seepage, and joint condition. The aim of this study is to compute correlation coefficients among various indexes and verify the effectiveness of lattice reduction. Alpha stable distribution is used to replace the commonly used Gauss distribution in probabilistic neural networks. A prediction model for the stability of roadway surrounding rock is then established based on a concept lattice and improved probabilistic neural network. 100 groups of training sample data are plugged into this model one by one to examine its rationality. The established model is employed for engineering application prediction with ten indiscriminate sample groups from the Jianlinshan mining area of the Daye iron mine, revealing accuracy of up to 90%. This demonstrates that our prediction model based on a concept lattice and improved probabilistic neural network has high reliability and applicability.

Underground space engineering structures are generally subject to extensive damages and significant deformation. Given that composite rocks are prone to shear failure, which cannot be accurately monitored, the piezoelectric active sensing method and wavelet packet analysis method were employed to conduct a shear failure monitoring test on composite rocks in this study. For the experiment, specimens were prepared for the simulation of the composite rocks using cement. Two pairs of piezoelectric smart aggregates (SAs) were embedded in the composite specimens. When the specimens were tested using the direct shear apparatus, an active sensing-based monitoring test was conducted using the embedded SAs. Moreover, a wavelet packet analysis was conducted to compute the energy of the monitoring signal; thus allowing for the determination of the shear damage index of the composite specimens and the quantitative characterization of the shear failure process. The results indicated that upon the shear failure of the composite specimens, the amplitudes and peak values of the monitoring signals decreased significantly, and the shear failure and damage indices of the composite specimens increased abruptly and approached a value of 1. The feasibility and reliability of the piezoelectric active sensing method, with respect to the monitoring of the shear failure of composite rocks, was therefore experimentally demonstrated in this study.

The morphology of the joint surface is multi-scale, and it can be divided into first-order asperity (waviness) and second-order asperity (unevenness). At present, the joint roughness characterization formula considers only the morphology contribution of waviness and unevenness components and does not fully consider their mechanical contribution. At same time, the relationship between the mechanical contribution and the morphology contribution is still unclear. Thus, the characterization formula considering the mechanical contribution of waviness and unevenness needs to be further studied. In this study, the standard joint roughness coefficient (JRC) profiles were first decomposed into waviness and unevenness. Then, three types of joint specimens with different asperity orders (flat, the standard JRC profile, and the profile containing only waviness) were prepared by the 3D engraving technique. Finally, direct shear tests were carried out on 39 sets of red sandstone joint specimens under three normal stresses. The mechanical contributions of waviness and unevenness were studied, the relationship between the mechanical contribution and the morphology contribution of waviness and unevenness was analyzed, and the characterization formula considering the mechanical contribution of waviness and unevenness was established. The results showed that the following: (1) the method combining the ensemble empirical mode decomposition (EEMD) and the critical decomposition level could be used to separate the waviness and unevenness from the joint surface; (2) the mechanical contribution of the waviness and unevenness decreased with the increase in normal stress; (3) the relationship between the mechanical contribution ratio and the statistical parameter ratio of the waviness and unevenness can be describe by power function; and (4) the roughness characterization formula considering the mechanical contribution and morphology contribution was established. This study will enhance the accurate evaluation of the roughness coefficient and shear strength of the joint specimen.

The shear resistance of filling joints is an important factor affecting the stability of rock joints. Pressure-shear tests of cement-filled joints were carried out. Combined with the acoustic emission (AE) technique, the effects of normal stress, roughness and filling degree on the shear strength, damage morphology and damage evolution of cement-filled joints were investigated. The results show that with the increase of roughness, the failure mode is more complicated. When the roughness is low, only the bonding surface of the interface between the filler and the joint surface is damaged, and the filling degree has a weak effect on the failure mode. When the roughness is high, with the increase of normal stress and filling degree, the failure of the filled joint is from the joint failure of the bonding surface and the filling material to the serious failure of the bonding surface, the filling material and the joint. The peak shear strength of filled joints is positively correlated with roughness and negatively correlated with filling degree. With the increase of filling degree, the influence of roughness will be weakened by filling material, and the normal stress will amplify the effect of roughness. The evolution characteristics of AE show that the damage degree of filled joints is positively correlated with normal stress and roughness, and negatively correlated with filling degree. The joint surface is damaged locally at first, then failure near the main raised body of the joint, and finally spreads to the whole joint surface.

Different from the goaf formed by normal mining, the goaf formed by the continuous caving of the lower illegal goafs is extremely irregular and unstable, which leads to the difference in the goaf caving characteristics. Taking the complex irregular goaf of the Shirengou Iron Mine in China as the research object, the 3D laser scanning technology is used to detect the goaf shape and analyze the cause of the formation. Then, the caving characteristics of the goaf are analyzed by the methods of the numerical simulation, the monitoring of the crash of the caving rock, and the monitoring of the deformation of the arch foot. The simulation results show that the transverse expansion of the irregular goaf boundary is mainly along the perpendicular to the direction of the ore body strike when the goaf caves upward in the initial stage. Additionally, the monitoring results show that the caving characteristics can be divided into two stages: the paroxysmal caving stage and the periodicity caving stage. The paroxysmal caving characteristics are intense and irregular, while the periodicity caving characteristics have obvious periodicity, which is mainly reflected in the periodic alternation between the mass caving and the temporary stability. The interval period of the goaf mass caving is about 15–30 days and is changed by the progress of filling treatment. The conclusions in this work reveal the particularity of the caving characteristics in the complex irregular goaf and contribute to the improvement of the caving characteristics and the prediction of mass caving.