System Design Of Contact Wire Wear Detection Based On Fpga And DSP

With the development of CRH (China Railways High-speed), security issues of the train running at a rapid speed have aroused widespread concern. The detection of wear degree of contact wire becomes more prominent and important. The main way that the train gets power is the friction between pantograph and overhead contact line. Due to the long period of rubbing between the catenary and pantograph, overhead contact line system (OCS) will generate a certain amount of wear. In addition, the works environment of OCS is very poor. When the degree of wear exceed a certain range there will be insufficient power supply even power off, which may result in huge losses. In order to prevent such risks, the regular detection to wear of contact wire must be carried out and the catenary which can’t supply power normally must be replaced to ensure the stable development of railway transportation industry.After overviwe the reserch development about friction wear of OCS in domestic and foreign, a non-contact design to detect wear of catenary is proposed and verified. By constructing an optical system with an active lightsource, the target catenary images are captured as clearly as possible using high-resolution line array camera with Gigabit Ethernet interface, which facilitates subsequent catenary wear measurement and calculation. Nios II soft-core processor implanted into FPGA as the control chip controls and preprocesses the data of the entire image acquisition system. Based on minimum circuit system of FPGA, the hardware circuit of Gigabit Ethernet interface is designed. The principle of GigE camera standard communication protocal-GigE Vision is described in detail and then the software of image acquisition system is developed which is mainly composed by the data transmitting module, the data receiving module and the data transmission module. As a result the stable acquisition of image data is realized.After the initial pretreatment to image data such as binarization, filtering processing, image segmentation etc in FPGA, the image data processed is transmitted to processing system based on DSP via dual-port RAM for final abrasion value detection. The detection work is achieved by digital image processing technology and the result is displayed. The experiments show that the system is stable and reliable and provide some reference value to the study of catenary friction wear.