Collision is the main issue in Aprons safe transportation, including in the railway system.Sensor systems have been developed to detect obstacles to prevent a collision, such as using cameras.One disadvantage of the camera systems is that performance detection decreases in a not clean environment, like the target position behind the fogs.This paper discusses the development of frequency modulated continuous wave (FMCW) radar signal processing for high-speed railway collision avoidance.
The development of radar signal processing combines a two-dimensional constant false alarm rate (2D-CFAR) and robust principal component analysis (RPCA) to detect moving targets under clutter.Cell average (CA) and Greatest of CA 3 Piece Sectional with Audio System (GOCA) CFAR are evaluated under a cluttered wall environment along the railway track.From the experiment, the development of FMCW radar can detect stationary or moving obstacles around 675 meters in front of the locomotive.Combining 2D-CFAR and RPCA algorithm outperforms average background subtraction in extracting moving targets from strong clutter signals along the railway track.