Support Vector Machine Model for Dam Security Monitoring
XU Guolong · Water Resources Informatization · 2011
Monitoring model and indicators can be effectively applied for the judgment of dam performance to realize early warning for dam security.Based on sudden change suppression of measured data by wavelet analysis,small-sample model establishment of support vector regression and high generalization ability are applied to set up variable weight support vector regression model for dam monitoring,with consideration on that measured value of different ranges will play different role in early warning for dam monitoring.Engineering cases and theoretical analysis show that the model has certain theoretical and practical values.