Hybrid Particle swarm optimization (HPSO) algorithm is a stochastic global optimization technique with many advantages, such as quick convergence, simple regulation and easy implementation. In order to determine the time-varying parameters of creep constitutive model of rock, in this article, a new method is presented using HPSO algorithm and fish language, which was contained in FLAC. At first, the stochastic values of parameters are initialized and the difference between the value computed and the datum measured during creep was regarded as fitness function to evaluate quality of the parameters. Then the parameters are updated continually using HPSO until the optimal parameters are found. Thus time-varying parameters of creep constitutive model of rock are identified adaptively during computation. Simulations was done for shale creep experiment, the results show that hybrid particle swarm optimization algorithm is effective in identifying the time-varying parameters of creep constitutive model of rock and viscoelastic characteristics of shale can be described better by using inconstant creep constitutive model.