铁路接触网异物检测系统的设计

发布时间：2018-04-27 08:36

本文选题：铁路接触网 + DSP　；参考：《辽宁科技大学》2016年硕士论文

【摘要】：近年来,高速电气化铁路不断发展,其安全问题也备受关注。接触网是电气化铁路牵引供电系统的重要组成部分,一旦发生故障,将会直接影响牵引供电系统的正常运行,甚至还会中断电气化铁路的行车功能。由于接触网的露天布置,受环境影响大,运行中受到各种恶劣天气和外侵异物的影响,调查发现,接触网经常会悬挂有塑料袋、风筝、气球等异物,主要悬挂在承力索、吊弦、接触线、附加导线上,接触网悬挂异物对运输的影响不容忽视,轻则造成动车组降弓或停车,重则造成供电设备故障或弓网故障,使动车组大量晚点,造成严重的社会影响。所以,对接触网异物检测系统的研究也是非常必要的。随着计算机技术的发展,数字图像处理技术也被应用在铁路检测系统中。由于铁路系统的复杂性,其对图像采集和识别的技术的要求也不断提高。目前大多数接触网检测系统都是通过摄像头来检测,其具有不定性,且摄像头所拍摄的画面受外界影响很大,本文提出一种可以应用在接触网异物检测系统中的图像采集和识别的硬件系统。接触网异物检测系统的关键就是图像的采集和处理,要求具有较高的实时性和可靠性。本文采用了DSP+FPGA的硬件结构,DSP和FPGA的结合能够充分发挥DSP芯片的处理速度和运算能力以及FPGA可重复配置的灵活性以及实时性等优点。大大的提高了图像处理的速度和准确性。在硬件上,本文选择了TI的8核Keystone架构的TMS320C6678芯片作为核心的DSP处理器,Altera Cyclone IV系列的EP4CE15作为FPGA协处理器,设计了一套图像采集和识别系统。该系统的设计主要包括图像采集部分的设计、图像处理部分的设计以及一些外围电路的设计。在算法上,本文完成了图像预处理的滤波算法以及图像的锐化方法,对比了几种常用的滤波算法,设计了基于非局部滤波的时空联合滤波方法、针对拉普拉斯算法设计了拉普拉斯锐化滤波器,并进行了仿真。最后,针对Sobel算子边缘检测,提出了一种改进的方法,并对其效果进行验证。
[Abstract]:In recent years, high-speed electrified railway has been developing, and its safety has been concerned. Catenary is an important part of the traction power supply system of electrified railway. Once the fault occurs, it will directly affect the normal operation of the traction power supply system and even interrupt the running function of the electrified railway. Because of the open-air arrangement of the catenary, which is greatly affected by the environment, and affected by all kinds of bad weather and foreign bodies in operation, the investigation found that the catenary often has foreign objects such as plastic bags, kites, balloons, etc., which are mainly suspended on the load cables and strings. The influence of foreign body suspended by catenary on the contact line and additional wire can not be ignored. The light causes the lower bow or stop of the EMU, and the heavy causes the fault of the power supply equipment or the pantograph, which makes the EMU delayed a lot and causes serious social impact. Therefore, it is necessary to study the foreign body detection system of catenary. With the development of computer technology, digital image processing technology is also used in railway detection system. Due to the complexity of railway system, the requirements of image acquisition and recognition technology are also increasing. At present, most of the catenary detection systems are detected by the camera, which is uncertain, and the images taken by the camera are greatly influenced by the outside world. In this paper, a hardware system for image acquisition and recognition is presented, which can be used in the detection system of foreign bodies in catenary. The key of the foreign body detection system of catenary is the acquisition and processing of image, which requires high real-time and reliability. In this paper, the hardware structure of DSP FPGA and the combination of FPGA and DSP can give full play to the processing speed and computing ability of DSP chip, the flexibility of FPGA reconfigurable configuration and real-time performance. The speed and accuracy of image processing are greatly improved. In terms of hardware, this paper chooses the TMS320C6678 chip of TI's 8-core Keystone architecture as the core DSP processor and the EP4CE15 of Cyclone IV series as FPGA coprocessor, and designs a set of image acquisition and recognition system. The design of the system mainly includes the design of the image acquisition part, the image processing part and the design of some peripheral circuits. In the algorithm, the filtering algorithm of image preprocessing and the sharpening method of image are completed in this paper. Several common filtering algorithms are compared, and a spatio-temporal joint filtering method based on non-local filtering is designed. Laplace sharpening filter is designed for Laplace algorithm and simulated. Finally, an improved method for edge detection of Sobel operator is proposed and its effect is verified.
【学位授予单位】：辽宁科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U226.8;TP391.41

【参考文献】