VR技术在变电站仿真培训系统中的应用研究  

Application Research of VR Technology in Substation Simulation Training System

作　　者：人力资源管理 企业信息化建设 WAGN Lei;YANG Guolian;CHEN Wei;MOU Xudong;WANG Ruiqiang(Training Center,State Grid Gansu Electric Power Co.,Ltd.,Lanzhou 730070,China;Equipment Management Department,State Grid Gansu Electric Power Co.,Ltd.,Lanzhou 730900,China;Tianshui Power Supply Company,State Grid Gansu Electric Power Co.,Ltd.,Tianshui 741099,China)

机构地区：[1]国网甘肃省电力有限公司培训中心,甘肃兰州730070 [2]国网甘肃省电力有限公司设备管理部,甘肃兰州730900 [3]国网甘肃省电力有限公司天水供电公司,甘肃天水741099

出　　处：《自动化仪表》2021年第9期25-29,共5页Process Automation Instrumentation

摘　　要：为研究虚拟现实(VR)技术在变电站仿真培训系统中的优势,并分析应用前景,创新性地设计一种结合多种技术的330 kV变电站沉浸式仿真培训系统。该系统包括三维VR引擎技术、多通道视频输出技术、虚拟人驱动技术、Kinect体感交互和漫游技术。在分析系统总体架构的基础上,重点介绍多通道视频输出技术和Kinect体感的交互虚拟技术。330 kV变电站沉浸式仿真培训系统能够显示三维立体逼真效果图,同时能模拟人物的转向和漫游。330 kV变电站沉浸式仿真培训系统引入VR技术后,互动性和真实性更强。Kinect体态识别技术的最佳捕捉范围为2.0~3.5 mm。200次抽样数据中,识别准确率最高和最低的动作分别为左手朝胸部挥动作和手握拳动作,相应的样本数为32和18,识别准确率分别为97%和80%。试验收集到的真实数据能够为今后虚拟式仿真培训系统的完善提供支持。In order to study the advantages of virtual reality(VR)technology in the substation simulation training system and analyze the application prospect,an immersive simulation training system for 330 kV substation combining several technologies is innovatively designed.The system includes 3D VR engine technology,multi-channel video output technology,virtual human driving technology,Kinect motion-sensing interaction and roaming technology.Based on analyzing the overall construction of the system,the multi-channel video output technology and Kinect motion-sensing interactive virtual technology are emphatically introduced.The immersive simulation training system for 330 kV substation can display three-dimensional realistic effect pictures and simulate the turning and roaming of characters.After VR technology was introduced into the immersive simulation training system for 330 kV substation,the system’s interactivity and authenticity become better.The best capture range of Kinect body recognition technology is 2.0~3.5 mm.Among the 200 sample data,the movements which have the least and the most recogrition accuracies are left hand waving to the chest and hand clenching respectively,and the corresponding sample numbers are 32 and 18 respectively,and the recognition accuracies are 97%and 80%respectively.The real data collected from the experiment can provide support for perfecting the immersive simulation training system in the future.

关 键 词：变电站 虚拟现实技术 仿真培训 KINECT 多通道技术 系统 识别准确率 漫游 

分 类 号：TH-85[机械工程]