机场停机坪环境具有场景范围较大、特征点稀疏的特征，随机特征点数量的增加会导致建图失真和里程计漂移现象，基于通过平滑和映射的紧耦合雷达惯性里程计算法（LIO-SAM）进行改进，首先对激光雷达（LiDAR）和惯性测量单元（IMU）进行外参和内参标定，将激光雷达点云格式转换为Velodyne格式，并采用自适应特征点曲率计算，以确保不同距离下特征点所计算曲率的准确性，然后引入Scan-Context回环检测，提高回环效率和建图精度，最终生成高精度的停机坪环境点云地图，通过实验平台采集园区道路和模拟停机坪两组数据集，使用本文算法与主流的A-LOAM、LeGO-LOAM和LIO-SAM算法进行对比，并利用GNSS数据作为轨迹真值。结果表明，改进后的多传感器融合算法在停机坪环境下相较于原始LIO-SAM算法的绝对位姿误差均值降低了13.8%。相较于其他主流算法，本文算法的绝对位姿误差均值最低，提供了更高的建图精度，有效降低了系统的累积误差，生成的点云地图更好地还原了停机坪的真实场景。

The airport apron environment features a large scene range and sparse feature points. Increased random feature points can the lead to mapping distortion and odometer drift. This paper presents improvements based on the tightly-coupled Lidar Inertial Odometry via Smoothing and Mapping （LIO-SAM） method. Firstly， the external and internal parameter calibration of Light Detection and Ranging （LiDAR） and Inertial Measurement Unit （IMU） is conducted， the LiDAR point cloud format is converted into the Velodyne format， and adaptive feature point curvature calculation is used to obtain accurate curvature at various distances. Then， Scan-Context loop detection is introduced to enhance loop efficiency and mapping accuracy， so as to generate a high-precision apron environment point cloud map. Two datasets， park roads and a simulated apron， are collected using the experimental platform for comparison. The proposed algorithm is evaluated against the A-LOAM， LeGO-LOAM， and LIO-SAM algorithms， using GNSS data as trajectory ground truth. Results show that the improved multi-sensor fusion algorithm reduces the mean absolute pose error by 13.8% compared to the original LIO-SAM algorithm in apron environments. The proposed algorithm also has the lowest mean absolute pose error among mainstream algorithms， providing higher mapping accuracy， reducing cumulative error， and better restoring the real scene of the apron.