The accuracy of virtual sensor output data is crucial for ensuring the reliability of simulation results in the field of autonomous driving simulation. As a key sensor for vehicle environment perception, the accuracy of point cloud data collected by LiDAR is critical for achieving a comprehensive understanding of the three-dimensional environment. However, in the virtual environment, the simulated point cloud data generated using 3D rendering techniques often fail to accurately reflect the variations of the sensor under complex conditions. In this study, we propose a modeling method for virtual LiDAR sensor in autonomous driving simulation. Firstly, a geometric measurement model for LiDAR is constructed based on the Unity 3D engine. Subsequently, a simplified LiDAR physical model is derived by incorporating the decay characteristics of real sensors. Finally, a Monte Carlo-based noise simulation is performed on the stochastic model to achieve high-fidelity LiDAR data output. The proposed method is validated by combining it with finely detailed virtual scenarios. Experimental results demonstrate the effectiveness of the method in accurately simulating LiDAR data in virtual environments, thus facilitating the verification of autonomous driving simulation algorithms