Track Matching Based on ELM for HFSWR

High frequency surface wave radar (HFSWR) plays an important role in marine surveillance and target tracking. However, in the complex environment of interference, the ship target signals will be submerged, which lead to track breakage and miscorrelation. How to effectively detect and track specific targets continuously is the key problem to be solved in signal processing for HFSWR. In this paper, we propose a track matching method based on extreme learning machine (ELM). In this method, track features are fully explored: 1. Extract the key features of ship attributes based on track information; 2. Establish the matching relationship between track data and ship model based on ELM. The track matching accuracy is verified by both the simulation and field experiments, validating the effectiveness of the new method.