Intelligent identification and localization of base pile defects based on PSO-GBDT
The low-strain reflected wave method is an important means to realize the diagnosis and health assessment of pile foundation defects.However,at present,the judgment of the detection results of this method is still carried out manually,and the judgment of the results carried out manually inevitably leads to misjudgment or inaccurate judgment and other problems as the defect waveform is not obvious.In order to solve this problem,the gradient boosting decision tree(GBDT)is used to establish a nonlinear relationship between the detection results of the low-strain reflected wave method and the location of the pile defects,and to realize the rapid identification and localization of the pile defects.And the particle swarm optimization(PSO)algorithm is introduced to optimize the key parameters of the model to improve the accuracy and gener-alization ability of the model.In addition,the kernel principal component analysis(KPCA)algorithm is used to downscale the multidomain features of the low-strain reflected wave,so as to reduce the difficulty of model training.Finally,the feasibility and accuracy of the model are verified by a large number of measured experi-mental data.The experimental results show that the model has the ability of rapid identification and localiza-tion of defects in foundation piles.
万方数据
维普
