【作者】 夏志宏；

【导师】 罗毅；

【作者基本信息】 华中科技大学 ， 电力系统及其自动化， 2005， 硕士

【摘要】 视觉是人类观察世界、认知世界的重要功能手段。随着电力系统自动化和管理现代化水平的提高,遥视系统(远程图像监控)和红外热成像技术在电力系统得到广泛的应用,为电力系统提供了丰富的现场运行图像信息。这些信息不仅能让电力系统操作运行人员直观,准确了解现场设备的运行状态和环境状况,而且基于图像信息的分析,能解决电力系统存在的许多顽固问题。由于电力系统过程的持续性、实时性、复杂性和强耦合性,远程图像监控和红外热成像取得的丰富的现场图像信息仅靠人为的监控和分析是远远不够的。本文就上述问题提出了解决方案。论文首先阐述了网络化遥视系统的结构,功能和关键技术问题,接下来介绍了遥视系统和红外热成像的集成,详细论述了遥视系统和红外热成像的硬、软件结构集成、红外热成像在电力遥视系统的预处理、伪彩色编码和传输编码等,探讨了集成系统的几个关键技术问题。论文在阐述电力遥视系统和红外热成像系统集成的基础上,引入计算机视觉技术,从解决刀闸开关状态自动识别问题入手,着重阐述了基于集成遥视系统的刀闸开关分合状态自动识别研究。论文详细阐述了刀闸开关图像的采集和编码、图像的灰度化、滤波、二值化和刀闸图像的深度分析等处理。重点阐述了利用红外热像解决不可见断口刀闸的状态识别和可见断口刀闸匹配困难等问题。并在自动识别仿真的基础上,分析了自动识别系统的性能。论文建立起系统抗干扰的数学模型,对影响识别准确性,可靠性和实时性的因素进行了仿真分析。接下来论述了基于集成遥视系统的继电保护压板投退位置的自动识别研究,对识别算法和识别效果进行了研究,并初步建立了自动识别系统和继电保护管理系统的互联模型,对集成遥视系统的高级应用作了一个拓展性研究。最后一章对全文进行了总结,并对集成遥视系统的进一步研究方向作了展望。更多还原

【Abstract】 Vision is the important means of observing and cognizing the world for human being. Following the improvement of power system automation and management modernization,power remote vision(Remote video monitoring) and the technology of infrared heat imaging are applied to power system widely. It supplies abundant local information for power system. from which not only power system operators can realize the state of local equipments and environment straightly and accurately,but also by analyzing these images,we can solve many difficult question of power system. Because of the property of duration, real time, complexity, strong coupling of power system process, the images of remote vision and infrared heat imaging are too abundant to monitor and analyze for human being. This article brings up the directions for the questions. Firstly, the framework, functions and key technology of network remote vision are expatiated. Following the integration of remote vision and technology of infrared heat imaging is introduced, the software and hardware integration of remote vision and infrared heat imaging, the pretreatment, fake color coding and transmission coding of infrared heat imaging in remote vision is particularly discussed. It discusses several key technology questions. On base of discussion of remote vision and technology of infrared heat imaging integration, it imports machine vision, proceeds with auto-recognition of switch, particularly discusses the state recognition of switch base of remote vision. It discusses the image collection, coding, binarization and depth analysis of switch. It elaborates matching arithmetic of switch state. Discusses the matching difficulty when the state of switch is difficult to confirm. On base of auto-recognition emluator, it analyzes the performance of auto-recognition system. It strikes up the mathematics model of anti-jamming, emulates the factors which affects veracity, reliability and real time. Secondly it dissertates image recognition of relay pressing on/off based on integrated remote vision, researches arithmetic and effects of recognition. It does a deploitation research for advanced apply of integrated remote vision. The researched problems of this article are summed up and more research directions are introduced about integrated remote vision lastly.更多还原