Abstract: The deformation prediction of high slopes in open pit mines is an important means to ensure the safety of mining production. This paper takes the slopes of a mine in Tibet as an object and uses high-precision synthetic aperture radar (SAR) interferometry to monitor the displacements of the southern slopes of the mining area in all weather conditions, and analyzes the basic law of slope deformation; The displacement monitoring data was denoised using wavelet denoising theory, and to prevent the prediction model from falling into local minima, genetic algorithm was introduced into the training steps of the BP neural network to optimize the initial weights and thresholds of the BP neural network. A GA-BP neural network slope deformation time series prediction model was established and compared with the BP neural network slope deformation time series prediction model. The research results show that the GA-BP model has improved the predictive accuracy of the BP model by over 10%, with an average prediction error of only 2.43%. The model convergence speed has increased by more than 10 times, and the regression coefficients of the GA-BP model are superior to those of the BP model. Therefore, the GA-BP model has improved the accuracy, convergence speed, and generalization ability of slope deformation prediction, making it more reliable for predicting slope deformations in mining areas and providing security for mining slope safety production.