As far as the randomness and uncertainty of switch machine fault was concerned, the equipment fault prognostics and health management (PHM) model based on the adaptive particle swarm optimization is proposed to optimize the hidden semimarkov. Firstly, the degradation process of mechanical parts of S700K switch machine is divided according to the whole life cycle, then the general HSMM model of equipment degradation state is established. Secondly, APSO algorithm is selected to optimize the PHM model of switch machine intelligently. Thirdly, the parameters of the optimized model are estimated using the forwardbackward algorithm. Finally, the effectiveness of the model for state assessment and remaining life prognostics is verified by an example analysis.