This study establishes a whole-process simulation model of a coherent wind lidar considering the transmission process of the lidar's optical detection path and atmospheric stratification theory. Using the existing system parameters, we simulate the detection process under two models: a constant wind speed model and the typical wind shear model of NASA. We also consider two systems: a high-pulse energy low-pulse repetition frequency (HPE-LPRF) system and a low-pulse energy high-pulse repetition frequency (LPE-HPRF) system. The detection performances of both systems are compared by computing the mean square error of their inversed wind speeds. The simulated signal-to-noise ratios of both systems well agree with the theoretical values. The accuracy of the measured wind speed is improved by incoherent accumulation average process. At an incoherent accumulative time of 0.1 s, the mean square error of the wind speeds detected under the constant wind speed model and NASA shear model is lower in the LPE-HPRF system (0.75 and 1.03 m/s, respectively) than in the HPE-LPRF system (0.93 and 1.25 m/s, respectively).