Papers by Keyword: Deformation Prediction

Abstract: Arch dam has gradually evolved as one of dam type as main large-scale hydraulic project, dam deformation prediction is an important part of dam safety monitoring, and it is difficult to forecast because of the complicated nonlinear characteristics of the monitoring data. Support Vector Machine (SVM) could solve the small sample, nonlinear high dimension problem due to the excellent generalization ability, and hence it has been widely used in the forecast of arch dam deformation. However, the forecast results considerably depend on the choice of SVM model parameters. In this paper, Particle Swarm Optimization (PSO), which has the characteristic of fast global optimization, was applied to optimize the parameters in SVM, and then the dam deformation prediction model based on PSO-SVM could be established. The model is applied to a certain arch dam foundation prediction. The accuracy of this employed approach was examined by comparing it with multiple regression method. In a word, the experimental results indicate that the proposed method based on PSO-SVM can be used in arch dam deformation prediction.

Abstract: In order to improve the accuracy and reliability of prediction of deformation monitoring data, a hybrid modeling and forecasting approach based on autoregressive model( AR) and the back-propagation( BP) neural network is proposed to forecast the deformation. The results of experiments show that this method can forecast the deformation precisely, and it is more suitable for those occasions where the deformation monitoring data should meet the high demand.

Abstract: Aiming at the problem that the Least Squares Support Vector Machines(LSSVM) was sensitive to noises or outliers, fuzzy idea was used to the Least Squares Support Vector Machines.The Fuzzy Least Squares Support Vector Machines(FLSSVM) was proposed and was applied to the Landslide Deformation Prediction. Experimental results show that this method can improve the accuracy of prediction and be effectively applied to landslide deformation prediction.

Abstract: The grey Verhulst nonlinear differential dynamic prediction model is applied to the prediction of the development of rock slope deformation in this paper. And experimental results show that grey Verhulst model is feasible to predict the slope rock mass deformation.

Abstract: Accurately estimating the deformation of dangerous rock is an important work for surveyors. Aiming at the limitation of the traditional GM (1,1) model, we propose that the error term in GM(1,1) model have an important influence on this model’s precision and adaptability. From this point of view, a novel new model termed SRGM (1, 1) is proposed. In this proposed model, the work modifies the algorithm of GM (1, 1) by integrate within semi-parametric regression model to eliminate the error term resulted from the traditional calculation of background value and initial value. According to the experimental results, our proposed SRGM (1, 1) model obviously can improve the precision of prediction and therefore can be adopted to deformation data analysis.

Abstract: Pit excavation was easy to cause the deformation of the supporting structure and surrounding soil, and brought serious harm to the surrounding buildings and urban underground pipelines. How to carry on a comprehensive analysis of inter-linked pit monitoring points, and improve the overall prediction accuracy was the urgent problem needed to be solved in scientific predictions of pit deformation. In order to establish the multi-variable gray theory GM(1,N) first-order linear dynamic model, using pit mutual influential settlement deformation monitoring data, and the correlation degree analysis, it filtered out the parent sequence WY09 point as the object to be analyzed, and the remaining points were as the systematic analysis of influencing factors, and WY09 point settlement predictions was calculated. According to the comparison analysis of the prediction results and engineering measured results, GM(1,N) model overall prediction accuracy was higher than GM(2,1) model, and prediction results were almost consistent with the measured results, so good effects was produced.

Abstract: Initial structure deformation prediction based on the regression analysis is an important means to evaluate the stability of tunnel, but the commonly used regression functions, such as exponential, logarithmic and hyperbolic function, have disadvantages of low accuracy. Aiming at the drawbacks, 3 kinds of multi-parameter growth functions (Weibull growth function, Richard function and Gompertz growth function) and 3 kinds of commonly used functions (exponential, logarithmic and hyperbolic function) are used to conduct the regression analysis for a loess tunnel’s initial structure in north-west China. Comparison results show that, (1) regressions based on multi-parameter growth functions are far more precise than the commonly used functions, reflecting its superiority in initial structure deformation prediction of the loess tunnel, (2) the more coefficients the function have, the higher the regression correlation is, and (3) the similarity of iterative and error analysis algorithm in different multi-parameter growth functions lays foundations for computer programming realization of regression analysis and model selection automation.

Abstract: This article from deformation characteristics of deep foundation pit which support by the pile anchor, then using neural network and Matlab software to establish the time series model to prediction and analysis the deep horizontal displacement of soil. The prediction results show that the overall shape of the curve is similar to "bow” ,and with the depth of excavation the maximum displacement occurred by the beginning of the location of 0.5 m from the surface to move to about half of the excavation depth That is the H/2 up and down position. The results have some reference to the practical engineering in a certain extent, this explain that the prediction is a kind of important means to realize information construction.

Abstract: Combining the advantages of basic genetic algorithm and neural network, analyze and set up GA & NN genetic neural network, explore and study the algorithm. The efficiency and effectiveness of this hybrid training has been significantly improved comparing with the single genetic evolution or BP training method, its versatility is better. The model is applied to predict the deformation of shield tunnel excavation. According to the effects of measured influence factors under construction, it can make the appropriate forecast to the surface settlement which is better than the conventional regression model. It shows that neural networks in the ground during tunneling shield analysis and prediction of settlement is practical and adaptable.

Abstract: "Zero waste" is the eternal pursuit of goals of global manufacturing. But the 2D error analysis has not yet comprehensive and accurate analysis and prediction of product quality information. Comprehensive 3D error analysis and judgments can be obtained through the high-precision 3D data acquisition on part and standard error of alignment for 3D data. According to the error information, the deformation of the part can be measured, traced and predicted, thus to reduce waste generation. 3D error analysis and deformation prediction system is proposed in this article, it has higher reliability and a certain role for the correct evaluation of processing conditions and comprehensive understanding of the process of evaluating the whole production process.