Study On Frequency Monitoring Subsystem Of High-Frequency Ground Wave OTH Radar

High-Frequency ground wave OTH radar is a new radar system, which is widely used in the military to detect and track aircraft, missiles, ships and other moving objects, especially in low-flying targets and moving targets at sea horizon accidents beyond the goal. However, high-frequency ground wave OTH radar works in shortwave band of 3-30MHz that electromagnetic environment is very complicated and the frequency is very crowded, as well as has a large number of interference. In order to ensure the high-frequency ground wave OTH radar to work normally in harsh electromagnetic environment, we must monitor its working spectrum in real-time and continuous, and chose the suitable "silence" band. The purpose of this task is to study on the frequency monitoring system combine specific high-frequency ground wave OTH radar.This paper describes the structure and characteristics of high-frequency ground wave OTH radar frequency monitoring systems, analyses the composition of the high-frequency radar, work timing and signal processing and put forward the function that need to be realized. We have completed some work contained following three aspects about signal processing of frequency monitoring systems。Firstly, we analyses the time-frequency characteristics of the radar interference in depth and the impact on the monitoring system. Anti-lightning double threshold algorithm in the frequency domain is proposed, which can remove the frequency of lightning interference successfully and achieve good results.Then we go on to predict the frequency monitoring data from frequency monitoring system. Empirical Mode Decomposition (EMD) is particularly suitable for the analysis of non-stationary signals, and Extreme Learning Machine (ELM) has simple structure and can quickly predict. Based on the above theory, this paper proposes a new method that the use of EMD method for spectral data preprocessing and ELM network for spectrum prediction. The simulation results show that the predictive outcome of this method is more effective than using a single model of ELM and SVM using data. Combined with the actual continuous processing scenarios, we simulate spectrum prediction based on Online Sequential Extreme Learning Machine (OS-ELM) through experiment, which has a very important practical significance.Finally, we study on the frequency selecting of frequency monitoring system and achieve it. Frequency selection is based on sliding window method. We can extract the average power spectrum of the sliding window and the ups and downs of the value through the sliding window, and then follow certain guidelines for their scoring. Considering the two-factor, we can calculate the total score for each sliding window and finally select the best frequency in accordance with the scores. In addition, the pretreatment of frequency selection is carried out, then silence frequency can be chosen based on each sliding window’s total score and the steps of frequency selection combine the requirements by using the actual collection data of high frequency ground wave OTH radar’s frequency subsystem monitoring...