In this article, a method for identification of equivalent stiffness was proposed based on a genetic algorithm by studying the stiffness characteristics of autoeliminating clearance auxiliary bearing devices (ACABDs) in static condition. Subsequently, the equivalent stiffness damping model and the Hertz contact theory were combined to establish the theoretical equations of the ACABD. Moreover, the linear equivalent model of the ACABD was established to eliminate the influence of contact of the revolute pair on mode shape and mode frequency of the rotor. In addition, simulations and experiments were carried out to verify the effectiveness of the genetic algorithm-based stiffness identification method. The results indicated that the relative errors between the equivalent stiffness in the X and Y directions identified by the linear equivalent model and the theoretical values were 6.22 and 7.19% respectively, demonstrating the feasibility of this identification method.