| 51 | 0 | 27 |
| 下载次数 | 被引频次 | 阅读次数 |
针对雾霾天气下能见度降低，轨道交通市域铁路列车运行安全受威胁，且激光雷达易受雾天环境干扰、目标识别与测距精度大幅下降的问题，开展相关技术优化研究。介绍了激光雷达在城市轨道交通领域的应用价值，以及雾气、噪声对激光脉冲信号检测造成的不良影响。在此基础上，提出一种基于改进YOLOv5模型的激光雷达雾天目标检测系统(MobileNet+MPDIoU-V3YOLO-based Multi-Pulse Lidar system，MM-YMPL)。将多次一维脉冲检测数据整合为二维图像，依托改进后的YOLOv5算法完成真实目标回波识别与位置解算。研究对该系统开展仿真与实物实验测试，结果表明，MM-YMPL系统可有效抑制雾气与噪声对激光信号的干扰，不仅提升了目标探测概率、延长有效探测距离，还具备优异的抗目标抖动性能。在后向散射率40%的环境中，系统50 m探测范围内平均绝对误差仅为0.016 m，测距精度显著优于现有方法。综上，该检测系统能够有效改善雾天激光雷达的工作性能，可保障市域铁路在雾霾天气下的列车运营安全，具备较强的实际应用价值。
Abstract：In response to the reduced visibility under haze weather conditions， which threatens the operational safety of urban rail transit metro trains， and the issues of LiDAR being susceptible to fog interference with significant declines in target recognition and ranging accuracy， this paper conducts relevant technical optimization research. The study first introduces the application value of LiDAR in urban rail transit systems and the adverse effects of fog and noise on laser pulse signal detection. Based on this， an improved YOLOv5 model-based LiDAR foggy weather target detection system ((MobileNet+MPDIoU-V3YOLO-based Multi-Pulse Lidar system，MM-YMPL) is proposed. This system integrates multiple one-dimensional pulse detection data into a two-dimensional image and utilizes the enhanced YOLOv5 algorithm to achieve real target echo identification and position calculation. Simulation and physical experiments were conducted on the system， demonstrating that MM-YMPL effectively suppresses fog and noise interference with laser signals， not only improving target detection probability and extending effective detection range but also exhibiting excellent resistance to target jitter. In an environment with a backscatter rate of 40%， the system's average absolute error within a 50-meter detection range was only 0.016 m， significantly outperforming existing methods in ranging accuracy. Overall， this detection system can effectively enhance LiDAR performance under foggy conditions， ensuring operational safety of metro trains during haze weather and possessing strong practical application value.
[1] 康学亮，王晓川. 基于车载激光雷达的点云道路标线提取方法[J] . 无线电工程，2023，53(5) :1228-1234. KANG Xueliang， WANG Xiaochuan. Road Marking Extraction from Point Cloud Based on Vehicular Lidar [J] . Radio Engineering，2023，53(5) :1228-1234.
[2] Zang S， Ding M， Smith D， et al. The impact of adverse weather conditions on autonomous vehicles: How rain， snow， fog， and hail affect the performance of a self-driving car[J]. IEEE vehicular technology magazine， 2019， 14(2): 103-111.
[3] Zhang G， Sun J， Zhou Y， et al. Imaging process and signal-to-noise ratio improvement of enhanced self-heterodyne synthetic aperture imaging ladar[J]. Chinese Optics Letters， 2017， 15(10): 102801.
[4] Shakil S， Billings J C， Keilholz S D， et al. Parametric dependencies of sliding window correlation[J]. IEEE Transactions on Biomedical Engineering， 2017， 65(2): 254-263.
[5] Huang N E， Long S R， Shen Z. The mechanism for frequency downshift in nonlinear wave evolution[J]. Advances in applied mechanics， 1996， 32: 59-117C.
[6] 李京，梅浩，何成林，等.基于脉冲宽度调制技术的激光引信抗干扰方法[J].红外与激光工程，2020，49(04):52-57.
[7] Mau J， Trumpf J， Day G， et al. An image feature-based approach to improving SPAD flash LiDAR imaging through fog[C]Emerging Imaging and Sensing Technologies for Security and Defence VII. SPIE， 2022， 12274: 23-32.
[8] Donoho D L. De-noising by soft-thresholding[J]. IEEE transactions on information theory， 1995， 41(3): 613-627.
[9] Du W， Xu H， She S， et al. Tracking Assisted LiDAR Target Detection Method During Rainy and Foggy Weather[C]//2023 35th Chinese Control and Decision Conference (CCDC). Yichang : IEEE， 2023: 207-212.
[10] Szegedy C， Vanhoucke V， Ioffe S， et al. Rethinking the inception architecture for computer vision[C]Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 2818-2826.
[11] Cai B， Xu X， Jia K， et al. Dehazenet: An end-to-end system for single image haze removal[J]. IEEE transactions on image processing， 2016， 25(11): 5187-5198.
[12] Heinzler R， Piewak F， Schindler P， et al. Cnn-based LiDAR point cloud de-noising in adverse weather[J]. IEEE Robotics and Automation Letters， 2020， 5(2): 2514-2521.
[13] Yan X， Yang J， Huang H， et al. AdverseNet: A Unified LiDAR Point Cloud Denoising Network for Autonomous Driving in Adverse Weather[J]. IEEE Sensors Journal， 2025，25（5）：8950-8961.
[14] Xu X，Huang Q.MD-DOA:A model-based deep learnin-g DOA estimation architecture[J].IEEE Sensors Journal，2024，24(12):20240-20253.
[15] Xiao L， Xie Y， Gao S， et al. Generalized radar range equation applied to the whole field region[J]. Sensors， 2022， 22(12): 4608.
[16] Xu X， Wang J， Wu J， et al. Full-waveform LiDAR echo decomposition method based on deep learning and sparrow search algorithm[J]. Infrared Physics & Technology， 2023， 130: 104613.
[17] Casasanta G， Garra R. Towards a generalized Beer-Lambert law[J]. Fractal and Fractional， 2018， 2(1): 8.
[18] 李明芳.DMP-YOLO:面向自动驾驶的多尺度目标检测算法[J].无线电工程，2025，55(11):2142-2152.
[19] Chambi S， Lemire D， Kaser O， et al. Better bitmap performance with roaring bitmaps[J]. Software: Practice and Experience， 2016， 46(5): 709-719.
[20] Howard A， Sandler M， Chu G， et al. Searching for mobilenetv3[C]//Proceedings of the IEEE/CVF international conference on computer vision. 2019: 1314-1324.
[21] Ma S， Xu Y M. A loss for efficient and accurate bounding box regression. arXiv 2023[J]. arXiv preprint arXiv:2307.07662， 2023.
[22] Wang S， Li M， Yang T， et al. Cone-shaped space target inertia characteristics identification by deep learning with compressed dataset[J]. IEEE Transactions on Antennas and Propagation， 2022， 70(7): 5217-5226.
基本信息:
中图分类号:U298;TP183;TN958.98
引用信息:
[1]龚福祥,吴龙,杨旭.基于YOLO的激光雷达雾霾天气下目标探测系统[J].无线电工程().
基金信息:
中国国家自然科学基金（62301493，62371163）; 中国基础科学研究计划（JCKY2024603C034）; 激光空间信息国家重点实验室基金
2026-07-01
2026-07-01
2026-07-01