总体来说，本文提出的算法具有一定的应用价值，实验结果也表明了这一点，算

In general, the algorithm proposed in this paper has certain application value, and the experimental results also show this. The map constructed by the algorithm completely reflects the actual structure and details of the environment. In terms of operating efficiency, the algorithm has added a 3D point cloud Compared with pure 2D lasers, the requirements for computing power are higher, but due to the use of descriptors to describe 3D point clouds, the algorithm can also meet the actual operating requirements in real time.

In general, the algorithm proposed in this paper has a certain application value, the experimental results also show this, the algorithm to build a map completely reflects the actual structure and details of the environment, in terms of operational efficiency, the algorithm because of the addition of 3D point cloud, compared to the pure 2D laser, the requirements for calculation is high, but due to the use of description sub-description 3D point cloud, the algorithm in real-time can also meet the actual operating requirements.

In general, the algorithm proposed in this paper has a certain application value, and the experimental results also show this point. The map constructed by the algorithm completely reflects the actual structure and details of the environment. In terms of operation efficiency, the algorithm requires higher computational power than pure 2D laser because of the addition of 3D point cloud. However, due to the use of descriptors to describe 3D point clouds, the algorithm is in real-time It can also meet the actual operation requirements.<br>