Abstract: To change the traditional method of applying the least square regression (LSM) in solving the statistical early warning model for dams, the stochastic search optimizing ability of particle swarm optimization (PSO) is employed to ascertain the regression coefficients of model. In order to solve the slow convergence rate of PSO used for a high-dimensional space optimization problem, a new self-adapting strategy that can adjust the learning factors, and combine with the crossover and mutation operators of genetic algorithm (GA) is proposed. The results show that the present method has better ability of searching diverse solutions and can adjust the flight length of particles by self-adapting, and can enhance the convergence rate of PSO; compared with the traditional least square regression and PSO, the data mining ability of this model is strong. The early warning evaluation results even more correspond with the practical operating condition, thus efficiently enhancing the forecasting precision of statistical models.