Abstract:Landslide displacements are characterized by time lag and nonlinearity, and common correlation analysis methods between landslide factors and cycle term displacements cannot intuitively determine the causal relationship between factors. In this paper, the CEEMD-CIWOA-BP landslide displacement prediction model considering time lag is established by taking the displacement monitoring data of Bazhimen landslide and Baishui River landslide as an example. Firstly, the landslide displacement monitoring data are decomposed into multiple signal components and res components using the CEEMD method, and reconstructed into landslide trend term and period term displacements; then, the Cubic chaotic mapping and inertia weight factor are introduced to optimize the WOA algorithm, and the optimized WOA algorithm is used to assign connection weights and bias terms to the BP neural network model; considering the time-lag effect of rainfall and reservoir level on landslide displacement, a Granger model is developed to predict the landslide displacement using the CEEMD-CIWOA-BP model, and the time-lag effect of rainfall and reservoir level on the landslide displacement is considered. Considering the time lag effect of rainfall and reservoir water level on landslide displacement, Granger causality test is used to determine the causal relationship between rainfall and reservoir water level and cycle displacement and MIC method is used to determine the number of time lag periods, and the cycle displacement is predicted using the CIWOA-BP model; finally, the cumulative predicted values of landslide displacement are obtained by superimposing the results of the various components, and the predictive accuracy of the model is evaluated, and the results show that the performance of the CEEMD-BP-CIWOA model proposed in this paper is very good. The results show that the CEEMD-BP-CIWOA model proposed in this paper outperforms other models, which verifies the feasibility of the proposed model and provides certain reference value for landslide disaster early warning and prediction.