Prediction of Landslide Displacement Based on KPCA and PSO-SVR

Taking Baijiabao landslide in the Three Gorges reservoir area for example,the kernel principal component analysis(KPCA) was employed to extract main features from influential factors data in order to obtain the principal component,which was used as the feature of support vector machine(SVM),and then building the displacement prediction model,the SVM model parameter of which was optimized by the particle swarm optimization(PSO) algorithm.Finally,the cooperative optimization model based on KPCA and PSO-SVR was proposed.And it was applied to predict displacement of landslide based on influential factors.The average absolute error and relative error of the prediction results were 0.760 and 7.563% respectively.By comparison with other forecast models,it was found that the fitting and generalization of this model are the best.The results indicate that the predicted value of the model is consistent with the monitoring data.It plays a key role in landslide hazard prediction and warning.