Application Of Improved Support Vector Machine In Dam Deformation Prediction

Dam dam deformation throughout the entire construction process and operations being late.The dam deformation may cause the occurrence of the disaster if the maximum deformation of the dam body is exceeded.It can reduce the probability of the occurrence of disaster if it can make accurate prediction of dam deformation.However,there are some shortcomings in the current domestic forecasting methods,which lead to the low prediction accuracy,so it is significant to predict the deformation of the dam rapidly and accurately.Thesis research work is mainly done as follows:First,the paper describes the background and significance of status quo dam deformation prediction home and abroad,for at the current domestic dam deformation prediction method deficiencies exist,it proposed based on gene expression programming method of improved support vector machine dam deformation prediction and analysis.Secondly,it uses wavelet analysis of the raw data preprocessing.To by Matlab 2012 b platform development tools,design codes and procedures related to the original observation data denoising;and then using gene expression programming algorithm powerful global search capability to find out the optimum parameters of support vector machine combination,establishment of the gene expression programming optimization support vector machine model for dam deformation prediction.Finally,improved support vector machine deformation prediction model of a dam on the dam in Yunnan ship lift monitoring data to predict application,the results show that the improved support vec tor machine model for dam deformation prediction than the traditional support vector machine dam deformation prediction accuracy of the model predictions derived from increased nearly three-fold,thus forecasting model gene expression programming instructions improved support vector machine has good predictive effect in deformation forecast of dam.