Study on Automatic Identification Algorithm of Wild Fire Near Transmission Lines Based on CINRAD-net Monitoring

In order to overcome the defects of existing methods on monitoring and warning of the wild fire near transmission lines, such as satellite and image monitoring, an automatic identification algorithm of wild fire near transmission lines based on the CINRAD-net monitoring was proposed in this paper. According to a sample database of CINRAD echoes for historical wild fires in a certain area, the characteristics of the CINRAD echoes for wild fire were analyzed. Thereby, the features characterizing CINRAD echoes were constructed and the threshold criteria were given. Then, the four-neighborhood algorithm was used to automatically identify the wild fire, and the radar clutter was filtered to improve the identifying accuracy. Combined with multithread processing method, the automatic identification for wild fire was realized. Based on an improved tower location method, fast locating and warning of wild fire were achieved for transmission line towers. The application in a provincial power grid has showed that the proposed method has an identification accuracy of nearly 80% in sunny days. This paper provided a new way for monitoring and warning of the wild fire near transmission lines, which can be verification and a supplement for satellite monitoring method. © 2020 Chin. Soc. for Elec. Eng.