| Being unlimited by the limitation of earth's curvature, High Frequency (HF) radar can effectively detect the ship targets and flying targets beyond range of visibility. The HF radar plays a vital role in military and civilian applications such as marine environmental monitory, smuggling monitory, etc.However, the HF radar normally involves in severely harsh electromagnetic environments, for instance, the sea clutter will obscure the target signal and the unexpected transient interference will level up the basement of Doppler spectrum which will cause extraordinary obstacles for ship targets monitory. Therefore, the rejection of transient interference has always been a popular issue, and this topic possesses a significant meaning in realizing the all weather operations of HF radar. There is a frequently happened situation that ships in different moving statuses exist in a range-azimuth cell of HF radar, especially in the prosperous international sea channels. Therefore, the key technology for realizing practical use of HF radar is to effectively detect multi-targets in strong sea clutter background. However, currently effective solutions to this problem have not been researched yet. The research in this paper was carried out under above-mentioned topic background, and major contributions include:1. Analysis of complex signal environments'characteristics of HF radarBased on Barrick equation of HF radar sea surface scattering cross section, and taking a comprehensive consideration of signal-affected environmental factors including sea-state (wind speed, wind direction and ocean surface currents), ionosphere phase contamination and multi-mode propagation, it suggested that sea clutter mode should be improved, and base on the improved mode, an in-depth discussion regarding the effects of environmental factors on target detection performance of HF radar from the angle of detection blind area. Subsequently, based on measured data of ground wave radar and sky-wave radar, this paper particularly analyzed the external interferences characteristics of time-domain, frequency-domain and range-doppler spectrum, providing thoughts and possible problems-solving approaches for the research of transient interference rejection algorithm.2. An algorithm of transient interference rejection of HF radar was proposed based on compressed sensing theory A new Modified Sparsity Adaptive Matching Pursuit (MSAMP) algorithm is proposed for signal high-precision reconstruction without prior information of the sparsity, and the analytical theory and simulation results show that the problem of over or under estimation in SAMP algorithm is almost resolved. In order to improve the inaccuracy of conventional transient interference rejection method in predicting the recovery of time-domain incomplete signals and to shorten the length of signal samples needed by the algorithm, a transient interference rejection method based on compressed sensing theory was proposed. Through suppressing the transient interference in measured data of sky wave radar and ground wave radar, the algorithm was tested in terms of effectiveness and superiority.3. A new multi-targets detection algorithm was proposed based on time-frequency analysis method and image processing techniquesRegarding maneuvering effect caused by non-radial uniform motion of HF radar target, the corresponding kinematic model and echo model were established, which emphasizingly analyzed the Doppler spread phenomenon of across first-order peak targets. Based on RS-method time-frequency distribution and Radon transform, a multi-targets detection algorithm (RRST method) was proposed, which converting multi-targets detection of time-frequency plane into identifying the peak point of Radon domain. According to the simulation experiment and real data process results, this algorithm can not only alleviate cross-term interference caused by multi-component signal, but also accurately detecte multi-target signals in low false alarm condition, and even conducts effective detection for across first-order peak targets.4. A CFAR multi-target detection algorithm was proposed based on curve estimation of instantaneous frequency and CFAR detection techniqueCombining with RS-method time-frequency distribution and Viterbi algorithm, a new instantaneous frequency (IF) estimation method (RSMVF) was proposed, which substantially decrease the influence of sea clutter on IF estimation accuracy. RSMVF constant false alarm rate(CFAR) multi-target detection algorithm was proposed based on time-frequency energy difference between target IF curve and noise IF curve. Through accumulating the time-frequency energy of IF curves and conducting threshold detection, this scheme overcame the RSMVF algorithm's defect of being unable to eliminate false-alarm, and acquired the ability to detect multi-target signals without the prior information on target number by using the method of successive detection and successive reset. The stimulation experiment and measured data processing results demonstrate that the proposed algorithm can precisely detect multi-targets and obtain accurate target motion parameters.5. An adaptive multi-target detection scheme was proposed based on IF curve estimation and linear goodness of fit testCombining the linear goodness of fit test, an adaptive multi-target detection algorithm was proposed based on linear characteristic difference between target signal IF and noise IF. Through adopting the previous five groups of data as reference, conducting low threshold detection on each group of data of added simulation targets and adding up the samples determination coefficients of related IF curves, the adaptive detection threshold can be obtained. Then, with the application of method of successive detection and successive reset, the algorithm can realize the adaptive detection for multi-targets. According to the results of stimulation experiment and measured data, in the condition of no prior information on target numbers, the algorithm can adaptively detect multi-targets in background of heavy clutter.6. By comparing the performance of the three proposed multi-target detection algorithms, an order of precedence was given regarding selecting the three algorithms in real target detection practiceDue to the current limited amount of literature reference, this paper conducted analysis on performances of the three proposed multi-target detection algorithms only. A experiment was conducted by randomly adding two stimulation targets in practical sea clutter background, and the results showed in practical multi-target detections of HF radar, the selecting priority of the three multi-target detection algorithms in this paper are:RRST method> RSMVF adaptive method> RSMVF CFAR method. |
PDF Full Text Request