In terms of lacking environmental awareness and autonomous path planning of snow cleaning robot in unknown environment, the designing idea of snow cleaning was confirmed through analyzing simultaneous localization and mapping(SLAM) based on robot operating system(ROS) combined with laser radar, and path planning technology.SLAM based on improved Rao-Blackwellized particle filter was applied to realize simultaneous localization and mapping of robots in unknown environments.Raster map was constructed through laser radar scanning and matching,A* algorithm and DWA algorithm were used for global path planning and local dynamic obstacle avoidance in the established raster map, and the feasibility was verified with the independently developed snow clearing robot as an experimental platform. The experimental results show that SLAM based on laser radar effectively solve the pose estimation and map construction of snow cleaning robot problemin the actual environment. The snow cleaning robot can move according to the designated route and avoid obstacles real-time, which verifies the possibility of the designed snow cleaning robot.