复杂环境下无人机智能巡检轨迹规划方法研究  

Research on intelligent inspection trajectory planning method of Unmanned Aerial Vehicle under complex environment

作　　者：田小壮 石辉 刘家辛 曹飞 常雪婷 TIAN Xiaozhuang;SHI Hui;LIU Jiaxin;CAO Fei;CHANG Xueting(State Grid Xinjiang Maintenance Company,Urumqi 830000,China;Beijing Zhongfei Aiwei Aviation Technology Co.,Ltd.,Beijing 100176,China)

机构地区：[1]国网新疆检修公司,新疆乌鲁木齐830000 [2]北京中飞艾维航空科技有限公司,北京100176

出　　处：《电子设计工程》2021年第20期77-81,共5页Electronic Design Engineering

基　　金：国家自然科学基金(61171106)。

摘　　要：针对无人机在变电站智能巡检中的应用问题,文中推导了载波相位观测单差分、双差分模型,分析了基于实时动态差分法(Real-Time Kinematic,RTK)的GPS定位原理。从机载系统与地面系统两个部分,构建了变电站无人机智能巡检系统架构;利用八叉树模型进行激光点云数据处理分析,构建了变电站三维激光点云仿真模型;分析快速拓展随机树(Rapidly-exploring Random Tree,RRT)算法存在不足的基础上提出去除冗余节点的改进方法,并将其应用于变电站复杂环境下无人机智能巡检轨迹规划。仿真算例与实验测试结果表明,文中所提基于RTK技术的无人机定位方法具有更高的定位精度,基于改进RRT算法能够显著减少无人机的转弯次数与巡检轨迹长度,从而有效降低无人机的能耗。Aiming at the application of unmanned aerial vehicle in intelligent inspection of substations,the paper derives the carrier phase observation single-difference and double-difference models,and analyzes the GPS positioning principle based on the Real-Time Kinematic(RTK)method.From the two parts of the airborne system and the ground system,the structure of the substation UAV intelligent inspection system is constructed.A 3D laser point cloud simulation model of the substation is constructed through data processing and analysis,using the octree model.The shortcomings of the Rapidly-exploring Random Tree(RRT)algorithm are analyzed.An improved method for removing redundant nodes is proposed and applied to the intelligent inspection trajectory planning of UAVs in the complex environment of substations.Simulation and test results showed that the UAV positioning method based on RTK technology proposed in the article has higher positioning accuracy,and the improved RRT algorithm can significantly reduce the number of UAV turns and inspection trajectory length,thereby effectively reducing energy consumption of drones.

关 键 词：无人机 变电站 智能巡检 轨迹规划 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]