Due to the decrease of population diversity in the later iteration, which leads to local optimization and precocious maturity, an algorithm that introduces simulated annealing operation into particle swarm algorithm is proposed, and the parameter setting of the algorithm is improved. Through the test of three standard test functions and compared with the particle swarm algorithm and simulated annealing algorithm, it is proved that the algorithm has the characteristics of fast convergence speed, high convergence accuracy and strong robustness. Finally, the algorithm is applied to the identification of DC motor transfer function parameters, and the simulation results show that the algorithm has high identification accuracy.