| The salient feature of cognitive radar is not only the ability to perceive the working environment,but also the ability to match the working characteristics with the environment,so as to achieve high efficiency,energy saving,low pollution and high performance.However,the working characteristics of the current marine radar do not match the radar working environment,which results in low working efficiency,high power consumption,high pollution to the electromagnetic environment and unsatisfactory working effect.In view of the existing problems of marine radar.In this article,the concept of cognitive radar is applied to marine radar.According to the application requirements and market competition characteristics of marine radar,the core technology of the cognitive function of marine radar is deeply studied.The structural model,signal model and cognitive control mode of cognitive marine radar are presented,and the working mode of traditional marine radar with scanning and tracking is preserved,and the features of simple structure and high cost performance are also preserved.(1)This article puts forward the overall design idea of separating the transceiver control of marine radar from the display control of radar,forming two independent parts,and constructs the structure model of cognitive marine radar.The structure model retains the track-while-scan working mode of conventional marine radar and the basic structure relationship of antenna and transceiver.At the same time,the model adds radar environment sensor,sensing memory,radar environment recognizer,working memory and control actuator to realize the perception-action cycle and memory,and establishes the information feedback link between the transmission signal generation,radar reception and target detection.On this basis,a transmitting signal model and a signal generation method are proposed.This method is convenient for realizing adaptive control of the waveform parameters of the transmitted signal and matching the working environment of the radar according to the spatial distribution of the target and the sensing information of the geographical environment of the water area.The signal generation method can be used to realize dozens or even thousands of different parameters of the transmitted signal waveform,(2)This article studies the waveform control method based on target space distribution and the waveform control method based on target detection performance.The formula for calculating the minimum detectable Signal-to-Clutter Ratio(SCR)of the target that meets the detection performance requirement(the probability of discovery(POD)under certain false alarm probability)is derived.A perceptual estimation algorithm of signal-to-clutter ratio and a waveform cognitive control algorithm based on signal-to-clutter ratio are proposed.According to the basic working characteristics of cognitive marine radar's perception-control cycle and the characteristics of marine radar's working environment changing greatly with time,this article puts forward two working stages of cognitive radar's working environment perception and working performance cognitive control,and designs the working flow of these two stages.On this basis,the experimental platform is designed and built.Experiments show that,based on the structure model,signal model and control flow,under the experimental environment,the radar emission energy can be reduced by 15.9dB,which is only 2.57%of the total energy under normal conditions,achieving the matching of radar and working environment.The method is convenient to realize efficient information perception and waveform control strategy.(3)This article presents a simple and fast coastline extraction algorithm and an improved GM-PHD multi-target tracking algorithm for piecewise scoring track management that does not depend on the strength of new targets,so as to adapt to the target perception requirements under high false alarm and high density environment.In the method of shoreline perception,a simple and fast algorithm of coastline extraction based on variable size template smoothing and Harr wavelet transform is adopted.The window size of the variable-size smoothing template is automatically adjusted with the distance to adapt to the relationship between the azimuth resolution of the radar and the distance.For the smoothed radar video array,threshold processing and Harr wavelet transform are used to obtain the coastline of radar image.In the method of target perception under high false alarm and high Density distribution environment,the GM-PHD multi-objective tracking algorithm is applied to the perception of spatial distribution characteristics of the target by taking advantage of the characteristics that the Gaussian Mixture Probability Hypothesis Density(GM-PHD)multi-objective tracking algorithm can achieve both target state estimation and target quantity estimation at the same time.In this article,a piecewise score track management GM-PHD multi-target tracking algorithm which does not depend on the strength of new targets is constructed,which improves the existing GM-PHD multi-target tracking algorithm and makes it adapt to the requirements of cognitive marine radar for target perception.In this algorithm,piecewise track management and improved pruning merge are adopted to enable the filter to adapt to the multi-target scene with fast speed,detect the new target and protect the new track more effectively,and enhance the filtering ability of false track under high false alarm.The simulation results show that,in the multi-target environment,the algorithm can effectively detect and protect the new high-speed moving target with unknown intensity,and can achieve good target state estimation effect and higher target number estimation accuracy and estimation speed.This shortens the stable estimation time of the number and status of targets to 5 data update cycles,which shortens nearly 10 to 20 data update cycles compared with the track management algorithm.(4)This article presents a comprehensive Constant False Alarm Rate(CFAR)target detection algorithm based on perception.In order to effectively suppress the effect of clutter peak interference on the target detection performance,this article firstly defines the peak-to-ratio coefficient that can reflect the correlation difference between target echo and clutter spike interference,and constructs an adaptive two-parameter logarithmic processor controlled by the peak-to-ratio coefficient to suppress the trailing effect of clutter distribution.Then,a matrix combination algorithm of radar synthesis video based on radar signal waveform selection control parameters is proposed,which can realize the generation of synthetic video signal and the average processing of non-coherent accumulation at the same time,and realize the normal approximation of clutter.Then,an adaptive CFAR processor based on normal clutter is constructed by using a window structure based on clutter uniformity and target distance extension perception,and a comprehensive Constant False Alarm detection algorithm is realized.On the basis of these processing methods,the integrated constant false alarm detection processing algorithm is implemented.The integrated CFAR algorithm can achieve adaptive adjustment of the width of the reference window and the protection window,so that the radar detection performance can match the clutter environment and target size,and obtain robust detection performance under both uniform and non-uniform clutter.The experimental results show that comp-CFAR detection algorithm can achieve a signal-to-clutter ratio improvement of about 0.3dB compared with NCI-CFAR processing algorithm at 80%detection probability. |
PDF Full Text Request