Papers by Keyword: Deformation Prediction

Abstract: In the process of metal parts fabricated by Laser Melting Deposition (LMD), a high temperature gradient will generate due to the instantaneous high laser energy input, which will cause residual stress in the formed part of metal parts, the residual stress will result in defects like warping deformation or even cracking. In this paper, a finite element method based on inherent strain method is proposed to predict the deformation of metal parts fabricated by LMD. Firstly, combing with the birth and death element technology, a local model is established to simulate the layer-by-layer deposition in the LMD forming process, and the values of inherent strain is obtained. Secondly, the obtained inherent strain values is applied to large-sized part layer by layer, and the final deformation of large-sized part is calculated. Based on the proposed method, the efficiency of deformation prediction of large-sized metal parts fabricated by LMD could be effectively improved.

Abstract: Thin-walled aluminum alloy parts are widely used in the aviation industry. In order to predict the deformation of milling aluminum alloy 7075-T7451 thin-walled parts, a deformation prediction method based on BP artificial neural network is presented. Firstly, the orthogonal experiment is designed to acquire the experimental data. Secondly, the BP neural network model of deformation prediction based on the experimental data is established. The comparison of the simulated values with the experimental results is performed to validate the proposed model. Lastly, the result shows that the proposed deformation prediction model is reasonable and can be used to predict the milling deformation.

Abstract: Combining with the localized analysis capability of the wavelet and the learning capability of the neural network, this article investigates the wavelet neural network nonlinear time series model and the wavelet neural network architecture integrated model which combines the affine transformation and the rotation transformation ,moreover, based on wavelet neural network algorithm that substitutes the Morlet wavelet function for the Sigmoid function, researchers analyze the application of Wavelet Neural Network in settlement deformation of high-speed rail. The example this paper discussed shows that the wavelet neural network has good results in the deformation prediction, especially in large complicated engineering.

Abstract: Hydropower reservoirs change the local geo-environment greatly and often induce reservoir bank landslides. Hongyanzi landslide is a large rejuvenated ancient landslide in Pubugou Hydropower Station reservoir area in western China Sichuan Province. To study the deformation of the Hongyanzi landslide, we arranged multi-parameter monitoring using GPS, fixed borehole tilt-meter, osmometer, and crack-meter and collected continuous data of displacement and ground water table. The monitoring data, together with the reservoir water level data, show that the landslide deforms obviously during reservoir water level is falling, while it deforms little in the other period. Comprehensive analysis of the deformation and reservoir water level implies that the seepage force caused by water level difference between the inner and outer slope may be the key factor. Through regression analysis, the authors established a forecast model for landslide deformation based on ground water table and reservoir water level; thus concluded that reducing the falling velocity of reservoir water level may slow the landslide deformation down.

Abstract: For there are many problems such as large amount interference factors, the amount of data collected is small, and so on occur in monitoring and forecasting, use the model that combine grey system and neural network to forecast. That is predicting with single model first, and then taking the prediction result of grey system as input sample value of BP neural network. Considering the measured values as objective sample value of network, neural network used to deformation forecast can be obtained through training. Prediction results show that this method can obtain a good forecasting result.

Abstract: Bipolar plate is the key component of proton exchange membrane (PEM) fuel cell and represents a significant part of the overall cost and the total weight in a fuel cell stack. Many research have been done on the manufacturing methods of bipolar plate, among which stamping is very popular. With the increasing of the channel number and complexity, its dimensional error caused by sprinkback will change a lot, even under the same forming process. And the risk of crack is also different. These all impact the quality of bipolar plate. In order to predict deformation of channels and the plate’s quality, the displacement along X-axis, the strain and stress state, and the displacement along Z-axis are measured. The results show that 1) the risk of crack increases with the increasing of channel number; 2) the springbacks increase with the increasing of channel number; 3) the most dangerous point locates on the right internal fillet of the plate’s last section.

Abstract: Extreme learning machine (ELM), as an emergent technique for training feed-forward neural networks, has shown good performance on various learning domains. This work evaluates the effectiveness of a new Gaussian kernel function-based extreme learning machine (KELM) algorithm for the deformation prediction of mine slope surface utilizing various kinds of meteorological influence factor data including the temperature, atmospheric pressure, cumulative rainfall, relative humidity and refractive index of the mining slope. The KELM model was applied to the deformation of Anjialing diggings, which is an open-pit mine of the China Coal PingShuo Group Co., Ltd. in China. The prediction performance on real data suggests that the proposed KELM model can effectively express the non-linear relationship between the mine open-pit slope surface deformation and various kinds of meteorological influence factors. The prediction results are compared with the most widely used algorithms – Support vector machine (SVM) and back-propagation neural networks (BP NN) in terms of the ease of use ( for example, the number of user-defined parameters), regression accuracy and computation cost. The comparison shows that the new algorithm is similar to SVM and BP NN but more accurate, and has notable lower computational cost and stronger generalization ability.

Abstract: Landslide prediction was one of the important tasks in the monitoring and early warning of the Three Gorges reservoir. On the basic of the BP neural network which was combined with the algorithm model, we took the typical Baishuihe landslide in Three Gorges reservoir for example. With the use of GPS monitoring data in surface displacement and its predisposing factors which was reservoir level, rainfall and so on, we built BP neural network prediction model. The monitoring of data recent years was used in the training and validation of the network .The results show that the BP neural network prediction model prediction can be used in the prediction of landslide deformation and the accuracy is high.

Abstract: Deformation analysis and prediction is a multidisciplinary important research topic. Considering that fuzzy theory has been widely used in prediction methods and models fields at present,the paper introduces fuzzy time series to deformation analysis and prediction. In this paper,the modeling steps of fuzzy time series are briefly introduced and experimental results show that fuzzy time series can be effectively applied to deformation analysis and prediction and improve the accuracy of prediction.

Abstract: This paper principally studies the prediction of slope deformation based on Support Vector Machine (SVM). To explore the prediction process, phase space is reconstructed. The geological body’s displacement data obtained from chaotic time series are used as SVM’s training samples. Slope displacement caused by multivariable coupling is predicted by means of single variable. Results show that this model is of high fitting accuracy and generalization, and provides reference for deformation prediction in slope engineering.