Research On Chaotic Small Signal Detection Method Under The Background Of Sea Clutter

As a typical radar interference echo,sea clutter signal refers to the backscattered echo of sea surface radar,which is easily affected by various external natural factors,such as wind,tide,surge,etc.Its physical mechanism is complex and changeable,and its non-Gaussian,non-linear and non-stationary characteristics are significant,which can easily cause interference to radar target detection.With the deepening of research on the mechanism of sea waves and the characteristics of sea clutter,researchers have discovered that sea clutter has chaotic characteristics.The study of chaotic small signal detection methods under the background of sea clutter has important theoretical research and application value for the establishment of marine safety observation and monitoring and sea target detection systems.In order to improve the detection accuracy of the chaotic small signal detection model under the background of sea clutter,the variational modal decomposition algorithm(VMD)is used to compensate for the modal aliasing,false components and end effect defects of the empirical mode decomposition.This paper proposes two sea clutter denoising algorithms based on variational modal decomposition.In order to make up for the weak detection ability of the traditional weak signal detection method under the chaotic background,considering that the support vector machine algorithm(SVM)is better at predicting,the immune algorithm(IA)with better optimization ability is introduced,a chaotic small signal detection method based on the IA-SVM model is proposed.From the perspective of rational use of noise,combined with stochastic resonance theory,and using variant differential evolution algorithm with good optimization convergence,a small signal detection method of stochastic resonance chaos based on variant differential evolution algorithm is proposed.The specific research is as follows:In order to analyze the chaotic characteristics of sea clutter,regarding the question of how to choose the phase space reconstruction parameters of a chaotic system,considering that there are two main research directions for determining the phase space parameters(embedding dimension m and time delay?),the first is to study these two parameters separately.It is cumbersome to use different methods to determine these two parameters;the second is to study the embedding window?_w.The study shows that the main factors affecting the quality of the reconstructed phase space are the joint embedding dimension and the embedding window width.The results of comprehensive analysis of various phase space parameter determination methods are basically consistent.In this paper,the more mature C-C method is selected to determine the phase space embedding window,and the embedding dimension is determined to be 5 and the time delay is 1.In order to improve the detection accuracy of the chaotic small signal detection model,this paper proposes a VMD-based sea clutter hybrid denoising algorithm and a VMD-based sea clutter distributed denoising algorithm.Use VMD to decompose the sea clutter signal into variational modal components(VMF)with different center bandwidths,analyze the autocorrelation characteristics of the decomposed signal.The first algorithm only performs wavelet hard threshold denoising processing on noise signals,the second algorithm combines instantaneous half-period(IHP)and wavelet hard threshold algorithms to denoise all component signals.Then,after reconstructing each component signal,denoising is completed,and finally combined with least squares support vector machine(LSSVM)sea clutter prediction model,the prediction root mean square error before and after denoising is compared to judge its denoising effect.Experimental results show that the two denoising algorithms have significant denoising effects,and the root mean square error of prediction can be reduced by two orders of magnitude.Aiming at the defect that the traditional weak signal detection method has weak detection ability under the chaotic background,this paper proposes a chaotic small signal detection method based on the IA-SVM model.After constructing the phase space of the chaotic sequence by finding the width of the embedding window,the optimization ability of the IA algorithm is used to optimize the three parameters of the penalty coefficient,the kernel function and the insensitive loss parameter that affect the prediction accuracy in the SVM,thereby establishing the prediction of the chaotic time series model.Detection of chaotic small signals in chaotic noise background from prediction errors.The simulation experiment uses the chaotic data of the Lorenz system and the sea clutter of the measured radar as background noise,analyzes the signal-to-noise ratio and root mean square error of the IA-SVM model prediction signal to judge the prediction performance effect of the model,and compares it with other prediction models.The experimental verification results show that the root mean square error of the predicted signal is 0.0001463(signal-to-noise ratio is-104.2473d B).By comparing and analyzing the root-mean-square error predicted by other models,the IA-SVM in the case of higher signal-to-noise ratio The prediction error of the model is smaller,the prediction performance is better,and it is closer to the actual value.Aiming at the defect that traditional stochastic resonance small signal detection cannot synchronize multiple parameters,this paper proposes a method of stochastic resonance chaotic small signal detection based on variant differential evolution algorithm.Using the variant differential evolution algorithm to optimize the stochastic resonance system parameters a,b,k of the Duffing oscillator,and use the system output signal-to-noise ratio as the objective function of the optimization problem.In order to verify the feasibility of the algorithm,the simulation experiments of low-frequency and high-frequency small signal input are carried out respectively.In the low-frequency and small-signal detection experiment,the output signal-to-noise comparison is improved by 1.98d B on average by the chaotic variable-step firefly optimization algorithm.In the high frequency small signal detection experiment,combined with the heterodyne stochastic resonance theory,it can accurately recover the small signal at the low frequency band corresponding to the high frequency small signal,and further deduce the existence of the high frequency small signal.The simulation experiment on the measured sea clutter data shows that the method can effectively detect the chaotic small signals submerged in the sea clutter background.