Cloud-Verhulst hybrid prediction model for dam deformation under uncertain conditions

Uncertainties existing in the process of dam deformation negatively influence deformation prediction. However, existing deformation pre- diction models seldom consider uncertainties. In this study, a cloud-Verhulst hybrid prediction model was established by combing a cloud model with the Verhulst model. The expectation, one of the cloud characteristic parameters, was obtained using the Verhulst model, and the other two cloud characteristic parameters, entropy and hyper-entropy, were calculated by introducing inertia weight. The hybrid prediction model was used to predict the dam deformation in a hydroelectric project. Comparison of the prediction results of the hybrid prediction model with those of a traditional statistical model and the monitoring values shows that the proposed model has higher prediction accuracy than the traditional sta- tistical model. It provides a new approach to predicting dam deformation under uncertain conditions.

Time effect and prediction of broken rock bulking coefficient on the base of particle discrete element method

Bulking characteristics of gangue are of great significance for the stability of goafs in mining overburden in the caving zones.In this paper,a particle discrete element method with clusters to represent gangue was adopted to explore the bulking coefficient time effect of the broken rock in the caving zone under three-dimensional triaxial compression condition.The phenomena of stress corrosion,deformation,and failure of rock blocks were simulated in the numerical model.Meanwhile,a new criterion of rock fragments damage was put forward.It was found that the broken rock has obvious viscoelastic properties.A new equation based on the Burgers creep model was proposed to predict the bulking coefficient of broken rock.A deformation characteristic parameter of the prediction equation was analyzed,which can be set as a fixed value in the mid-and long-term prediction of the bulking coefficient.There are quadratic function relationships between the deformation characteristic parameter value and Talbot gradation index,axial pressure and confining pressure.

A Novel Hybrid Intelligent Prediction Model for Valley Deformation: A Case Study in Xiluodu Reservoir Region, China

The narrowing deformation of reservoir valley during the initial operation period threatens the long-term safety of the dam,and an accurate prediction of valley deformation(VD)remains a challenging part of risk mitigation.In order to enhance the accuracy of VD prediction,a novel hybrid model combining Ensemble empirical mode decomposition based interval threshold denoising(EEMD-ITD),Differential evolutions—Shuffled frog leaping algorithm(DE-SFLA)and Least squares support vector machine(LSSVM)is proposed.The non-stationary VD series is firstly decomposed into several stationary subseries by EEMD;then,ITD is applied for redundant information denoising on special sub-series,and the denoised deformation is divided into the trend and periodic deformation components.Meanwhile,several relevant triggering factors affecting the VD are considered,from which the input features are extracted by Grey relational analysis(GRA).After that,DE-SFLA-LSSVM is separately performed to predict the trend and periodic deformation with the optimal inputs.Ultimately,the two individual forecast components are reconstructed to obtain the final predicted values.Two VD series monitored in Xiluodu reservoir region are utilized to verify the proposed model.The results demonstrate that:(1)Compared with Discrete wavelet transform(DWT),better denoising performance can be achieved by EEMD-ITD;(2)Using GRA to screen the optimal input features can effectively quantify the deformation response relationship to the triggering factors,and reduce the model complexity;(3)The proposed hybrid model in this study displays superior performance on some compared models(e.g.,LSSVM,Backward Propagation neural network(BPNN),and DE-SFLA-BPNN)in terms of forecast accuracy.

Comparative Study on Deformation Prediction Models of Wuqiangxi Concrete Gravity Dam Based on Monitoring Data

The deformation prediction models of Wuqiangxi concrete gravity dam are developed,including two statistical models and a deep learning model.In the statistical models,the reliable monitoring data are firstly determined with Lahitte criterion;then,the stepwise regression and partial least squares regression models for deformation prediction of concrete gravity dam are constructed in terms of the reliable monitoring data,and the factors of water pressure,temperature and time effect are considered in the models;finally,according to the monitoring data from 2006 to 2020 of five typical measuring points including J23(on dam section 24^(#)),J33(on dam section 4^(#)),J35(on dam section 8^(#)),J37(on dam section 12^(#)),and J39(on dam section 15^(#))located on the crest of Wuqiangxi concrete gravity dam,the settlement curves of the measuring points are obtained with the stepwise regression and partial least squares regression models.A deep learning model is developed based on long short-term memory(LSTM)recurrent neural network.In the LSTM model,two LSTMlayers are used,the rectified linear unit function is adopted as the activation function,the input sequence length is 20,and the random search is adopted.The monitoring data for the five typical measuring points from 2006 to 2017 are selected as the training set,and the monitoring data from 2018 to 2020 are taken as the test set.From the results of case study,we can find that(1)the good fitting results can be obtained with the two statistical models;(2)the partial least squares regression algorithm can solve the model with high correlation factors and reasonably explain the factors;(3)the prediction accuracy of the LSTM model increases with increasing the amount of training data.In the deformation prediction of concrete gravity dam,the LSTM model is suggested when there are sufficient training data,while the partial least squares regression method is suggested when the training data are insufficient.

A Characteristic Analysis of the Dynamic Evolution of Preseismic-Coseismic-Postseismic Interferometric Deformation Fields Associated with the M 7.9 Earthquake of Mani,Tibet in 1997

By using the D-InSAR technique, we have acquired the temporal-spatial evolution images of preseismic.cosesimci-postseismic interferometric deformation fields associated with the M 7.9 earthquake of Mani, Tibet on 8 November 1997. The analysis of these images reveals the relationships between the temporal-spatial evolution features of the interferometric deformation fields and locking, rupturing, and elastic restoring of the source rupture plane, which represent the processes of strain accumulation, strain release, and postseismic restoration. The result shows that 10 months prior to the Mani event, a left-lateral shear trend appeared in the seismic area, which was in accordance with the earthquake fault in nature. The quantity of local deformation on the north wall was slightly larger than that on the south wall, and the deformation distribution area of the north wall was relatively large. With the event impending, the deformation of the south wall varied increasingly, and the deformation center shifted eastward. Two and half monthd before the event, the west side of the fault was still locked while the east side began to slide, implying that the whole fault would rupture at any moment. These features can be regarded as short-term precursors to this earthquake. Within the period from 16 April 1996 to two and half months before the earthquake, the most remarkable deformation zones appeared in the north and south walls, which were parallel to and about 40 km apart from the fault, with accumulated local displacements of 344 mm and 251 mm on the north and south walls, respectively. The south wall was the active one with larger displacements. Five months after the earthquake, the distribution feature of interferometric fringes was just opposite to that prior to the event, expressing evident right-lateral shear. The recovered displacements are -179 mm on the north wall and -79 mm on the south wall, close to the east side of the fault. However, in the area of the south wall far from the fault there still existed a trend of sinistral motion. The deformation of the north wall was small but recovered fast in a larger area, while the active south wall began to recover from the east section of the fault toward the WSW.

Influence of adjacent shield tunneling construction on existing tunnel settlement: field monitoring and intelligent prediction

Urban subway tunnel construction inevitably disturbs the surrounding rock and causes the deformation of existing subway structures. Dynamic predictions of the tunnel horizontal displacement, tunnel ballast settlement, and tunnel differential settlement are important for ensuring the safety of buildings and tunnels. First, based on the Hangzhou Metro project, we analyzed the influence of construction on the deformation of existing subway structures and the difficulties and key points in monitoring. Then, a deformation prediction model, based on a back propagation(BP) neural network, was established with massive monitoring data. In particular, we analyzed the influence of four structures of the BP neural network on prediction performance, i.e., single input–single hidden layer–single output, multiple inputs–single hidden layer–single output, single input–double hidden layers–single output, and multiple inputs–double hidden layers–single output, and verified them using measured data.

Deformation resistance prediction of tandem cold rolling based on grey wolf optimization and support vector regression

In the traditional rolling force model of tandem cold rolling mills,the calculation of the deformation resistance of the strip head does not consider the actual size and mechanical properties of the incoming material,which results in a mismatch between the deformation resistance setting and the actual state of the incoming material and thus affects the accuracy of the rolling force during the low-speed rolling process of the strip head.The inverse calculation of deformation resistance was derived to obtain the actual deformation resistance of the strip head in the tandem cold rolling process,and the actual process parameters of the strip in the hot and cold rolling processes were integrated to create the cross-process dataset as the basis to establish the support vector regression(SVR)model.The grey wolf optimization(GWO)algorithm was used to optimize the hyperparameters in the SVR model,and a deformation resistance prediction model based on GWO–SVR was established.Compared with the traditional model,the GWO–SVR model shows different degrees of improvement in each stand,with significant improvement in stands S3–S5.The prediction results of the GWO–SVR model were applied to calculate the head rolling setting of a 1420 mm tandem rolling mill.The head rolling force had a similar degree of improvement in accuracy to the deformation resistance,and the phenomenon of low head rolling force setting from stands S3 to S5 was obviously improved.Meanwhile,the thickness quality and shape quality of the strip head were improved accordingly,and the application results were consistent with expectations.

An improved artificial bee colony-random forest(IABC-RF)model for predicting the tunnel deformation due to an adjacent foundation pit excavation

An improved artificial bee colony-random forest(IABC-RF)model is proposed for predicting the tunnel deformation due to the excavation of an adjacent foundation pit.A new search strategy of the artificial bee colony(ABC)algorithm is herein developed and incorporated,with the results showing that a much higher computational efficiency can be achieved with the new model,while high computational accuracy can also be maintained.The improved ABC algorithm is thereafter utilised and combined with the random forest(RF)model,where four important hyper-parameters are optimized,for a tunnel deformation prediction.Results are thoroughly compared with those of other prediction methods based on machine learning(ML),as well as the monitored data on the site.Via the comparisons,the validity and effectiveness of the proposed model are fully demonstrated,and a more promising perspective can be seen of the method for its potential wide applications in geotechnical engineering.

Analysis of meso-inhomogeneous deformation on a metal material surface under low-cycle fatigue

A polycrystalline Voronoi aggregation with a free surface is applied as the representative volume element（RVE）of the nickel-based GH4169 superalloy.Considering the plastic deformation mechanism at the grain level and the Bauschinger effect,a crystal plasticity model reflecting the nonlinear kinematic hardening of crystal slipping system is applied.The microscopic inhomogeneous deformation during cyclic loading is calculated through numerical simulation of crystal plasticity.The deformation inhomogeneity on the free surface of the RVE under cyclic loading is described respectively by using the following parameters：standard deviation of the longitudinal strain in macro tensile direction,statistical average of first principal strains,and standard deviation of longitudinal displacement.The relationship between the fatigue cycle number and the evolution of inhomogeneous deformation of the material’s free surface is investigated.This research finds that：（1）The inhomogeneous deformation of the material free surface is significantly higher than that of the RVE inside;（2）the increases of the characterization parameters of inhomogeneous deformation on the free surface with cycles reflect the local maximum deformation of the RVE growing during cyclic loading;（3）these parameters can be used as criteria to assess and predict the low-cycle fatigue life rationally.