Research On Image Recognition Based On TMS320DM642 For Switching State Of Substation Switch Gear

The provisions of article 2.3.6.5 in "State Grid Corporation of electric power safety work procedures(substation part)" proposes double confirmation request to the remote control operation of switch gear, and in the operation site of switch it would not be able to achieve safety regulations requirements without the staff to confirm the mechanical instructions. This issue restricts the development of unattended and one-key sequential control key technologies in intelligent substations. Therefore, this thesis designs a system to recognize the opening and closing state of the substation's switch gear to solve this problem.The algorithm used in the switch image recognition technology is studied, and the SIFT scale invariant feature transformation algorithm is used to locate the switch image accurately. The key information of the foreground (switch) region is extracted by the gray scale and binarization of the image. The Hough Transform algorithm is used to detect the straight line to judge the switch's opening and closing state,and the recognition result is obtained.In the aspect of hardware design, the characteristics and internal resources of TMS320DM642, which are used in the recognition system, are introduced. The overall structure of the switch image recognition system is designed as the overall program: the CCD camera collects the switch image and then perform A/D conversion of the switch image through the SAA7115HL decoder chip, and output YUV format digital signal to the TMS320DM642. The TMS320DM642 caches the input digital image firstly, and after the image is processed, the final recognition result is output. Then the detailed circuit schematic diagram of each module is given,which includes the video image acquisition module, external memory module(SDRAM, FLASH), power module, clock module and serial communication module.Then the software design's principle of the whole recognition system is described,including the loading and starting program of the DSP chip, the initialization of each module of the system, the configuration of the key parameters and the analysis of key instructions of the switch image processing module.Finally, the advantages of MATLAB in visualization of image processing and recognition are combined with the DSP development environment CCS, and the recognition system is validated. The result shows that the algorithm is effective and the recognition system can quickly recognize the switch opening and closing state and achieve the design function expected.