| 融合无线通信与视觉感知的牵引走板运行过程智能监控研究 |
投稿时间：2026-05-21 修订日期：2026-06-21  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 32 |
| 全文下载次数: 0 |
|
|
| 中文摘要:特高压输电线路张力放线施工面临复杂的非结构化环境与强电磁干扰，牵引走板在非视距工况下的姿态失稳极易诱发翻转掉线事故。针对现有监测手段在观测盲区、通信链路稳定性及姿态解算精度等方面的不足，本文提出一种融合异构无线通信与视觉-惯性感知的智能监控系统。首先，构建基于编码正交频分复用与自组网技术的通信架构，利用子载波并行传输与保护间隔机制，有效克服了电晕放电噪声与多径衰落对视频回传的干扰；其次，提出一种引入线特征几何约束的视觉-惯性里程计（PL-VIO）算法，并结合改进YOLOv8网络中的双向特征金字塔与协同注意力机制，解决了弱纹理、高动态场景下的位姿漂移与微小缺陷漏检难题。工程实测结果表明：该系统在强干扰下视频传输稳定，端到端预警延迟低于185ms；在剧烈震荡及天空背景下，轨迹跟踪误差较主流算法降低约59.1%~67.9%，微小缺陷检测mAP达到86.2%。研究成果实现了放线过程的全维态势感知，为特高压工程的智能化施工与主动安全防御提供了关键技术支撑。 |
| 中文关键词:特高压输电 张力放线 牵引走板 视觉-惯性里程计 COFDM 智能监控 |
| |
| Intelligent Monitoring of Running Boards in Tension Stringing: A Visual-Communication Fusion Approach |
|
|
| Abstract:During the tension stringing of UHV transmission lines, the stability of traction plates is critical but often compromised by unstructured environments and strong electromagnetic interference (EMI). To address the challenges of monitoring blind spots, unstable communication links, and low pose estimation accuracy in non-line-of-sight (NLOS) conditions, this paper proposes an intelligent monitoring system integrating heterogeneous wireless communication and visual-inertial perception. First, a communication architecture based on Coded Orthogonal Frequency Division Multiplexing (COFDM) and Mesh networking is constructed. By leveraging sub-carrier parallel transmission and guard intervals, it effectively overcomes video transmission interruptions caused by corona discharge noise and multipath fading. Second, a Point-Line Visual-Inertial Odometry (PL-VIO) algorithm incorporating line feature geometric constraints is proposed. Combined with the Bi-directional Feature Pyramid Network (BiFPN) and Coordinate Attention (CA) mechanism in an improved YOLOv8 network, this approach solves the problems of pose drift and missed detection of small defects in weak-texture and high-dynamic scenarios. Field test results demonstrate that the system maintains stable video transmission under strong interference with an end-to-end warning latency of less than 185ms. In scenarios with severe vibration and sky backgrounds, the trajectory tracking error is reduced by approximately 59.1%~67.9% compared to mainstream algorithms, and the detection mAP for small defects reaches 86.2%. This study realizes full-dimensional situational awareness during stringing operations, providing key technical support for intelligent construction and active safety defense in UHV projects. |
| keywords:UHV transmission tension stringing traction plate visual-inertial odometry COFDM intelligent monitoring |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
