Magnetorheological (MR) dampers, as an intelligent vibration damping device, can quickly change the damping size of the material in milliseconds. The traditional semiactive control strategy cannot give full play to the ability of the MR dampers to consume energy and reduce vibration under different currents, and it is difficult to control the MR dampers accurately. In this paper, a semiactive control strategy based on reinforcement learning (RL) is proposed, which is based on “exploring” to learn the optimal value of the MR dampers at each step of the operation, the applied current value. During damping control, the learned optimal action value for each step is input into the MR dampers so that they provide the optimal damping force to the structure. Applying this strategy to a two-layer frame structure was found to provide more accurate control of the MR dampers, significantly improving the damping effect of the MR dampers.

Drill core lithology is an important indicator reflecting the geological conditions of the drilling area. Traditional lithology identification usually relies on manual visual inspection, which is time-consuming and professionally demanding. In recent years, the rapid development of convolutional neural networks has provided an innovative way for the automatic prediction of drill core images. In this work, a core dataset containing a total of 10 common lithology categories in underground engineering was constructed. ResNeSt-50 we adopted uses a strategy of combining channel-wise attention and multi-path network to achieve cross-channel feature correlations, which significantly improves the model accuracy without high model complexity. Transfer learning was used to initialize the model parameters, to extract the feature of core images more efficiently. The model achieved superior performance on testing images compared with other discussed CNN models, the average value of its Precision, Recall, F1−score for each category of lithology is 99.62%, 99.62%, and 99.59%, respectively, and the prediction accuracy is 99.60%. The test results show that the proposed method is optimal and effective for automatic lithology classification of borehole cores.

The traditional Three-dimensional (3D) model for stability analysis of a large underground cavern group only includes the main powerhouse, the main transformer chamber, the tailrace surge tank, the main electrical wire hall, and part of the diversion tunnel and tailwater tunnel. Other caverns excavation are not taken into consideration, which is inconsistent with the actual project. In this study, a 3D Finite Element Method (FEM) model, including all the cavities, and a 3D simplified model containing only part of the cavities were established based on the actual underground cavern group project. We study the structural characteristics of the cavern group during excavation via 3D nonlinear FEM. At the same time, three numerical calculation schemes are designed for comparison. The results show that the simulated distributions of the stress field, deformation field, and plastic zone of the two numerical models are similar, but the magnitudes of the 3D FEM model, which considers all caverns are generally higher than the simplified model. The difference between the two simulated horizontal displacements of the main powerhouse reaches 53.72%. Therefore, from the perspective of engineering safety and actual construction, it is advisable to adopt the 3D FEM model, including all the cavities, when conducting a simulation analysis of the underground cavern group.

Affected by various complex factors, dam deformation monitoring data usually reflect volatility and non-linear characteristics, and traditional prediction models are difficult to accurately capture the complex laws of dam deformation. A multi-scale deformation prediction model based on Variational Modal Decomposition (VMD) signal decomposition technology is proposed in this study. The method first decomposes the original deformation sequence into a series of sub-sequences with different frequencies, then the decomposed sub-sequences are modeled and predicted by Long Short-Term Memory neural network (LSTM) and Random Forest (RF) according to different frequencies. Finally, the prediction results of all sub-sequences are reconstructed to obtain the final deformation prediction results. In this process, it is proposed to use the instantaneous frequency mean method to determine the decomposition modulus of VMD. The innovation of this paper is to decompose the monitoring data with high volatility, and use LSTM and RF prediction, respectively, according to the frequency of the monitoring data, so as to realize the more accurate capture of volatility data during the prediction process. The case analysis results show that the proposed model can effectively solve the negative impact of the original data volatility on the prediction results, and is superior to the traditional prediction models in terms of stability and generalization ability, which has an important reference value for accurately predicting dam deformation and has far-reaching engineering significance.

No abstract