Spectrum Prediction Method Based on EMD and ELM in HFSWR

High frequency surface wave radar (HFSWR) works in shortwave bands with complex electronmagnetic environment, so it is difficult to find a optimal frequency for the next working cycle. Frequency monitoring system (FMS) is introduced into the new system radar to provide the optimal frequency real-time. This paper proposes a spectrum prediction method based on Empirical mode decomposition (EMD) and Extreme learning machine (ELM) to find the optimal frequency in FMS. The simulation result shows that the method predicts spectrum more effective than traditional method and it provides a good method to select the optimal frequency in HFSWR.