Since X-ray pulsar signals cannot be directly detected on the ground, and the space flight detection is both time-consuming and costly, simulation of X-ray pulsar signals with true physical characteristics is of great importance to the validation of various X-ray pulsar signal processing algorithms and X-ray pulsar-based navigation strategies. In this paper, a new simulation method of X-ray pulsar signals is proposed, in which according to the pulsar signal model at the solar system Barycenter (SSB) and the trajectory information of the spacecraft, the real-time photon arrival rate function at the spacecraft is established, then based on this, a scale transforming method is employed to directly generate the photon event time stamps at the spacecraft which follow a non-homogeneous Poisson process. The proposed simulation method takes into account the pulsar spin down law and the influences of the largescale time-space effects introduced in the process of dynamic detection, and thus avoids the complicated iteration procedure involved in the state of the art simulation methods. Finally, a series of simulations are designed to evaluate the performance of the proposed simulation method. The main results can be concluded as follows: 1) The simulated photon event timestamps have a slowly changing period, which are consistent with the pulsar spin down law. 2) The observed pulsar profile accurately reflects how the radiation intensity of pulsars changes over time within a phase cycle, and it has a Pearson correlation coefficient of up to 0.99 with a standard profile. 3) The simulated average fluxes of the pulsars are very close to the true values, and thereby verifies the correctness of the proposed simulation method from an overall point of view. 4) The simulated photon series are very similar to the real data detected by the RXTE explorer, and when the simulation time is longer than 50 s, the relevancy between the simulated profile and the profile obtained from the real data is higher than 0.9. 5) The computational cost of the scale transforming method is much less than that for the commonly used Poisson sifting method and the inverse mapping method. The above results show the validity and high efficiency of the proposed method in terms of the period property, the profile and flux accuracy, the similarity to the RXTE real data and the computational cost.