Aiming at solving the problems of poor noise sensitivity and accuracy in the filtering process of mobile robot LiDAR point cloud data, a weighted Gaussian filtering method was proposed. In this method, Gaussian filter and angle cosine were used as the weighted value to filter the radar point cloud data, which reduced the fluctuation of object contour scanned by LiDAR. Finally, the method was applied to the actual robot chassis of robot operation system. The experimental results show that the method can better solve the problems in the process of LiDAR point cloud filtering compared with the traditional Gaussian filtering method, and the scanned object profile is more smooth, and has good stability and accuracy, which is beneficial to the effect of LiDAR mapping.