Abstract: The deformation serves as a crucial indicator reflecting the structural changes of dams. Due to the non-linear characteristics of deformation data and the intricate mechanisms underlying them, enhancing the predictive accuracy of deformation is of paramount significance for the safety and structural control of dams. In this context, a combined approach for dam deformation prediction is proposed based on the integrated modeling concept. This method integrates an improved wavelet threshold denoising and a Pelican Optimization Algorithm (POA) optimized Bidirectional Long Short-Term Memory (BiLSTM). Initially, the deformation measurement data sequence is processed using the improved wavelet threshold denoising method. Subsequently, the POA is employed to search for the optimal hyperparameter combination to optimize the BiLSTM model. Finally, dam deformation prediction is conducted based on the BiLSTM model with the optimal hyperparameters. Engineering case studies demonstrate that the improved wavelet threshold method exhibits superior denoising effects, and the POA-BiLSTM accurately predicts dam deformation. On the ultimate test set, the average MAE, MAPE, RMSE, and R2 are 0.244, 0.041, 0.301, and 0.906, respectively. Compared to other methods, it exhibits higher predictive accuracy and robustness, offering valuable insights for dam deformation monitoring.