引用本文: | 贡文新, 余泽琰, 杨柳旺, 楚文静, 万相奎. 基于毫米波雷达的无人机障碍物分类方法[J]. 雷达科学与技术, 2025, 23(3): 317-327.[点击复制] |
GONG Wenxin, YU Zeyan, YANG Liuwang, CHU Wenjin, WAN Xiangkui. Millimeter-Wave Radar-Based Obstacle Classification Method for Unmanned Aerial Vehicles[J]. Radar Science and Technology, 2025, 23(3): 317-327.[点击复制] |
|
摘要: |
无人机巡线作为检测电力线的重要手段,在飞行过程中准确识别障碍物是保障巡线任务可靠完成的关键。但目前对无人机巡线过程中常见障碍物如电力线、电力塔、树木的识别受恶劣天气环境干扰严重,致使误判和漏判。为此,基于毫米波雷达传感器具有不受天气、光线因素影响,在复杂环境中工作稳定等特点,本文提出基于毫米波雷达的无人机障碍物分类方法。该方法首先通过毫米波雷达采集3类障碍物的原始数据并提取其距离-速度多普勒及距离-方位角多普勒信息,接着分别通过特征值分解及共生灰度矩阵实现特征提取,最后通过蛇鹭优化算法实现对3类障碍物的目标分类。实验结果表明,本文方法对电力线、电力塔和树木的整体识别准确率达89.4%,与传统方法相比具有较高的识别准确率及鲁棒性。 |
关键词: 毫米波雷达 障碍物分类 特征提取 蛇鹭优化算法 |
DOI:DOI:10.3969/j.issn.1672-2337.2025.03.009 |
分类号:TN958;TM755 |
基金项目:湖北省自然科学基金(No.2022CFA007); 武汉市知识创新专项项目(No.2022020801010258) |
|
Millimeter-Wave Radar-Based Obstacle Classification Method for Unmanned Aerial Vehicles |
GONG Wenxin, YU Zeyan, YANG Liuwang, CHU Wenjin, WAN Xiangkui
|
Hubei Key Laboratory for High Efficiency Utilization of Solar Energy and Operation Control of Energy Storage System, Hubei University of Technology, Wuhan 430068, China
|
Abstract: |
Using drones for power line inspection is a crucial method to ensure the reliability of power line maintenance. Accurately identifying obstacles during flight is essential for the successful completion of inspection tasks. However, the recognition of common obstacles during drone inspections, such as power lines, power towers, and trees, is often compromised by adverse weather conditions, leading to false detections or missed detections. To address this, a millimeter-wave radar-based obstacle classification method for drones is proposed, leveraging the radar’s stability in complex environments and immunity to weather and lighting interference. Firstly, by using millimeter-wave radar, the raw data of the three types of obstacles are collected, and the range-velocity Doppler and range-angle Doppler information are extracted. Then, the feature extraction is achieved through singular value decomposition (SVD) and gray-level co-occurrence matrix (GLCM), followed by classification of the three obstacle types using the secretary bird optimization algorithm (SBOA). Experimental results show that this method achieves an overall recognition accuracy of 89.4% for power lines, power towers, and trees, offering higher accuracy and robustness compared to traditional methods. |
Key words: millimeter-wave radar obstacle classification feature extraction SBOA |