基于CPO-HKELM的库岸滑坡位移预测OA

In view of the difficulty of deformation prediction due to the complex operating environment,random load and strong nonlinearity of the bank slope of the library,Kernel Extreme Learning Machine(KELM)is proposed to model the monitoring data and hybrid kernel function is introduced to enhance the model mapping ability.In order to solve the problem that the mapping ability of the Hybrid Kernel Extreme Learning Machine(HKELM)is affected by hyperparameters,a combined prediction model named CPO-HKELM is constructed by applying the Crested Porcupine Optimizer(CPO)to optimize the nuclear parameters and penalty factors of HKELM.Taking the landslide of a certain reservoir bank as the research object,the H04 monitoring data of the landslide was modeled.In order to verify the feasibility and superiority of the proposed model,CPO-KELM,CPO-ELM and CPO-BP models were introduced for comparative analysis.The results show that the prediction accuracy of the proposed CPO-HKELM is obviously higher than the other two models,and the error is smaller,which has a good application prospect in landslide displacement prediction.