Knowledge of pore-water pressure(PWP)variation is fundamental for slope stability.A precise prediction of PWP is difficult due to complex physical mechanisms and in situ natural variability.To explore the applicability and advantages of recurrent neural networks(RNNs)on PWP prediction,three variants of RNNs,i.e.,standard RNN,long short-term memory(LSTM)and gated recurrent unit(GRU)are adopted and compared with a traditional static artificial neural network(ANN),i.e.,multi-layer perceptron(MLP).Measurements of rainfall and PWP of representative piezometers from a fully instrumented natural slope in Hong Kong are used to establish the prediction models.The coefficient of determination(R^2)and root mean square error(RMSE)are used for model evaluations.The influence of input time series length on the model performance is investigated.The results reveal that MLP can provide acceptable performance but is not robust.The uncertainty bounds of RMSE of the MLP model range from 0.24 kPa to 1.12 k Pa for the selected two piezometers.The standard RNN can perform better but the robustness is slightly affected when there are significant time lags between PWP changes and rainfall.The GRU and LSTM models can provide more precise and robust predictions than the standard RNN.The effects of the hidden layer structure and the dropout technique are investigated.The single-layer GRU is accurate enough for PWP prediction,whereas a double-layer GRU brings extra time cost with little accuracy improvement.The dropout technique is essential to overfitting prevention and improvement of accuracy. 相似文献
The phenomenon of moisture increase under an impervious cover in soils due to thermal gradients is defined as the pot cover effect, which may lead to an obvious soil moisture increase in the shallow soil. This paper explores a measure of laying an impervious layer at an appropriate depth in the process of foundation treatment to eliminate the potential moisture increase in the covered soil. Because the impervious cover above the soil and the impervious layer inside the soil constitute a “double pot cover” structure, the moisture migration in the covered soil with an impervious layer is generalized as the double pot cover effect. To investigate the mechanism of the double pot cover effect and further determine the optimal depth of the impervious layer, a numerical model is established to simulate this problem. Analysis results indicate that the moisture increase under the cover varies with the depth of the impervious layer. As the impervious layer is laid at a certain depth, the moisture increase reaches a minimum value. Moreover, the double pot cover effect under different boundary temperatures is further discussed. Results show that the moisture increase in the covered soil can be significantly reduced by laying the impervious layer slightly below the freezing front (0 °C).
We present high-precision,multi-band CCD photometry of two less-studied close binaries V1123 Tau and V1128 Tau.Complete covered light curves and a number of new times of light minima of the two eclipsing systems were obtained,based on which,revised orbital elements and new ephemerides were given.By adopting the Wilson-Devinney method,the light curves were analyzed.The photometric solutions confirm the W UMa-type nature of the binary systems.With the less-massive secondary slightly cooler than the primary,V1... 相似文献
The development of large-scale time-domain surveys provides an opportunity to study the physical properties as well as the evolutionary scenario of B-type subdw... 相似文献
In order to simulate the soil response during principal stress rotation, anisotropic unified hardening (UH) model is developed within the framework of elastoplastic theory. Without introducing any additional mechanism to display the role of stress rotation specifically, this model achieves the simulation by considering the material anisotropy. The effect of inherent anisotropy is reflected using the anisotropic transformed stress method, but a new formula for the stress mapping is adopted to keep the mean stress unchanged. Analysis indicates that from the view of the transformed stress tensor, the anisotropic soil is subjected to loading during pure rotation of principal stress axes, so that plastic strains can be calculated. To represent the induced anisotropy, a fabric evolution law is proposed based on laboratory and numerical test results. At the critical state, the fabric tensor reaches a stable value determined by the stress state, while the critical state line is unique in the plane of void ratio versus mean stress. The anisotropic UH model has concise formulation and explicit elastoplastic flexibility matrix and can provide reasonable predictions for the deformation of anisotropic soils when principal stresses rotate. 相似文献