Abstract: Accurate prediction of open-pit mine slope deformation can safeguard the safe production of open-pit mines. Traditional slope deformation prediction models often have significant deficiencies in dealing with parameter sensitivity, which affects the accuracy and reliability of the prediction results. To effectively address this issue, this study proposes a novel slope deformation prediction method (referred to as the SA-PSO-MLP model) based on the time series characteristics and nonlinear evolution laws of historical slope displacement data from Heishantou Open-pit Mine. This model combines the simulated annealing (SA) algorithm and the particle swarm optimization (PSO) algorithm to optimize the multi-layer perceptron (MLP) model. Specifically, the model improves the global search ability of PSO through the SA algorithm, optimizes the initial weights and threshold parameters of MLP, and enhances the prediction accuracy and stability of the model. To verify the effectiveness and feasibility of this method, this paper compares and analyzes six models: MLP, genetic algorithm-optimized backpropagation neural network model (GA-BP), support vector regression model (SVR), grey prediction model GM(1,1), particle swarm optimization-optimized multi-layer perceptron model (PSO-MLP), and SA-PSO-MLP. The experimental results show that the mean squared error (MSE) of the SA-PSO-MLP model is 0.509 9, which is lower than the error values of the other five models; the coefficient of determination is 0.922 9, which is 0.17 to 21.40 percentage points higher than that of the comparison models, demonstrating a significant predictive advantage. The research proves that the SA-PSO-MLP model effectively overcomes the shortcomings of traditional models in terms of parameter sensitivity through a dual optimization mechanism and can improve the prediction accuracy by relying solely on single historical slope displacement data, providing new idea and method for the prediction of open-pit mine slope deformation.