In view of the shortcomings of traditional random forest parameter optimization methods, the particle swarm optimization algorithm was improved by introducing equalizing inertia weight and adaptive mutation, and a dam deformation prediction model based on improved particle swarm optimization algorithm and random forest algorithm(improved PSO-RF algorithm) was proposed. The example analysis shows that in terms of computational efficiency, compared with traditional grid search method, the improved PSO-RF algorithm significantly improves the optimization speed of the model. In the aspects of prediction accuracy and stability, the dam deformation prediction model based on the improved PSO-RF algorithm is obviously better than long short-term memory, support vector machine and BP neural network.