For space-borne gravitational wave detection missions based on the heterodyne interferometry principle,tilt-to-length(TTL)coupling noise is an important optical noise source,significantly influencing the accuracy of t...For space-borne gravitational wave detection missions based on the heterodyne interferometry principle,tilt-to-length(TTL)coupling noise is an important optical noise source,significantly influencing the accuracy of the measurement system.We present a method for analyzing TTL coupling noise under the joint influence of multiple factors.An equivalent simulated optical bench for the test mass interferometer was designed,and Gaussian beam tracing was adopted to simulate beam propagation.By simulating the interference signal,it can analyze the impact of various factors on the TTL coupling noise,including positional,beam parameters,detector parameters,and signal definition factors.On this basis,a random parameter space composed of multiple influential factors was constructed within a range satisfying the analysis requirement,and the corresponding simulation results from random sampling were evaluated via variance-based global sensitivity analysis.The calculated results of the main and total effect indexes show that the test mass rotation angle and the piston effect(lateral)significantly influence the TTL coupling noise in the test mass interferometer.The analysis provides a qualitative reference for designing and optimizing space-borne laser interferometry systems.展开更多
In this work,the generation of high signal-to-noise ratio(SNR)single-frequency microwave signal without noise sidebands is demonstrated based on the interaction of integrated all-fiber lasers.The microwave signals are...In this work,the generation of high signal-to-noise ratio(SNR)single-frequency microwave signal without noise sidebands is demonstrated based on the interaction of integrated all-fiber lasers.The microwave signals are generated by the interference between a narrow linewidth Brillouin pump light from a single-frequency laser and the Stokes light generated by it.Firstly,the linewidths of the Stokes lights are compressed to~43 Hz based on the stimulated Brillouin scattering(SBS)effect,which ensures that the frequency noise is as low as possible.And then,the relative intensity noise(RIN)of the first order Stokes light is reduced by 21 dB/Hz based on the noise dynamics principle in cascaded SBS effect.By simultaneously reducing the frequency noise and the intensity noise of the coherent signals,the noise sidebands of microwave signals are completely suppressed.As result,the SNR of the microwave signal is improved from 48 dB to 84 dB at the first-order Brillouin frequency shift of 9.415 GHz.Meanwhile,a microwave signal with a SNR of 70 dB is generated at the second-order Brillouin frequency shift of 18.827 GHz.This kind of microwave signals with narrow linewidth and high SNR can provide higher detection resolution and higher transmission efficiency for applications on radar,satellite communication and so on.展开更多
The spin-exchange relaxation-free atomic gyroscope,with its exceptionally high theoretical precision,demonstrates immense potential to become the next-generation strategic-grade gyroscope.However,due to technological ...The spin-exchange relaxation-free atomic gyroscope,with its exceptionally high theoretical precision,demonstrates immense potential to become the next-generation strategic-grade gyroscope.However,due to technological noise,there is still a significant gap between its actual precision and theoretical precision.This study identifies the key factor limiting the precision of the SERF gyroscope as coupling noise.By optimizing the detection loop structure,a distinction between the dual-axis signals'response to optical and magnetic fields was achieved-where the optical errors responded similarly,while the response to magnetic noise was opposite.Based on the differences in the optical-magnetic response of the dual-axis signals,empirical mode decomposition was used to decompose the dual-axis gyroscope signals into multiple intrinsic mode functions,and Allan deviation analysis was applied to analyze the noise characteristics of the intrinsic mode functions over various periods.This study successfully reveals that optical errors caused by thermal-optical coupling and long-period magnetic noise induced by thermal-magnetic coupling are the dominant factors limiting the long-term stability of the SERF gyroscope.Based on these analyses,the study concludes that to achieve strategic-grade precision for the SERF gyroscope,it is essential to effectively address the noise issues caused by multi-physical field couplings.展开更多
In this paper,we study the orthogonal time frequency space signal transmission over multi-path channel in the presence of phase noise(PHN)at both sides of millimeter wave(mmWave)communication links.The statistics char...In this paper,we study the orthogonal time frequency space signal transmission over multi-path channel in the presence of phase noise(PHN)at both sides of millimeter wave(mmWave)communication links.The statistics characteristics of the PHN-induced common phase error and inter-Doppler interference are investigated.Then,a column-shaped pilot structure is designed,and training pilots are used to realize linear-complexity PHN tracking and compensation.Numerical results demonstrate that the proposed scheme enables the signal to noise ratio loss to be restrained within 1 dB in contrast to the no PHN case.展开更多
针对电力线信道噪声干扰大、噪声建模复杂、重建无噪信号样本困难等问题,提出了一种面向一维时间序列的Noise2Noise神经网络优化模型,并验证分析Noise2Noise算法抑制电力线噪声的可行性。首先,分析了Noise2Noise算法的原理,对该算法原...针对电力线信道噪声干扰大、噪声建模复杂、重建无噪信号样本困难等问题,提出了一种面向一维时间序列的Noise2Noise神经网络优化模型,并验证分析Noise2Noise算法抑制电力线噪声的可行性。首先,分析了Noise2Noise算法的原理,对该算法原理进行了理论推导。其次,通过选取合适的网络结构,改进网络输入输出,使神经网络适合处理实验数据,并采用正弦波数据进行网络测试与模型验证。然后,通过搭建正交频分复用(orthogonal frequency division multi-plexing,OFDM)调制模型,产生OFDM符号,并添加特定噪声,形成带噪样本。最后,基于改进的网络结构实现了对电力线接收信号的OFDM前导序列进行带噪样本测试,增强了Noise2Noise方法的可行性与有效性,具有较好的实用性。展开更多
Recently, the digital image blind forensics technology has received an increasing attention in academic community. This paper aims at developing a new identification approach based on the statistical noise and exchang...Recently, the digital image blind forensics technology has received an increasing attention in academic community. This paper aims at developing a new identification approach based on the statistical noise and exchangeable image file format （EXIF） information of image for images authen- tication. In particular, the authors can identify whether the current image has been modified or not by utilizing the relevance between noise and EXIF parameters and comparing the real values with the estimated values of the EXIF parameters. Experimental results validate the proposed method. That is, the detecting system can identify the doctored image effectively.展开更多
In radar target tracking application, the observation noise is usually non-Gaussian, which is also referred as glint noise. The performances of conventional trackers degra de severely in the presence of glint noise. A...In radar target tracking application, the observation noise is usually non-Gaussian, which is also referred as glint noise. The performances of conventional trackers degra de severely in the presence of glint noise. An improved particle filter, Markov chain Monte Carlo particle filter (MCMC-PF), is applied to cope with radar target tracking when the measurements are perturbed by glint noise. Tracking performance of the filter is demonstrated in the present of glint noise by computer simulation.展开更多
To predict the remaining useful life(RUL) for a class of nonlinear multi-degradation systems, a method is presented. In the real industrial processes, systems are usually composed by several parts or components, and t...To predict the remaining useful life(RUL) for a class of nonlinear multi-degradation systems, a method is presented. In the real industrial processes, systems are usually composed by several parts or components, and these parts or components are working in the same environment, thus the degradations of these parts or components will be influenced by common factors. To describe such a phenomenon in degradations, a multi-degradation model with public noise is proposed. To identify the degradation states and the unknown parameters, an iterative estimation method is proposed by using the Kalman filter and the expectation maximization(EM) algorithm. Next, with known thresholds,the RUL of each degradation can be predicted by using the first hitting time(FHT). In addition, the RUL of the whole system can be obtained by a Copula function. Finally, a practical case is used to demonstrate the method proposed.展开更多
Stochastic noises of fiber optic gyroscope （FOG） mainly contain white noise and fractal noise whose long-term dependent component causes FOG a rather slow drift. In order to eliminate this component, a two-step filt...Stochastic noises of fiber optic gyroscope （FOG） mainly contain white noise and fractal noise whose long-term dependent component causes FOG a rather slow drift. In order to eliminate this component, a two-step filtering methodology is proposed. Firstly, fractional differencing （FD） method is introduced to trans-form fractal noise into fractional white noise based on the estima-tion of Hurst exponent for long-term dependent fractal process, which together with the existing white noise make up of a gener-alized white noise. Further, an improved denoising algorithm of wavelet maxima is developed to suppress the generalized white noise. Experimental results show that the basic noise terms of FOG greatly decrease, and especially the slow drift is restrained effectively. The proposed methodology provides a promising ap-proach for filtering long-term dependent fractal noise.展开更多
A minimum geometric power distortionless response beamforming approach against impulsive noise （including all α- stable noise） of unknown statistics is proposed. Due to that definite logarithmic moments require no ...A minimum geometric power distortionless response beamforming approach against impulsive noise （including all α- stable noise） of unknown statistics is proposed. Due to that definite logarithmic moments require no priori knowledge of impulsive noise, this new beamformer substitutes the logarithmic moments for the second-order moments and iteratively minimizes the ＂ge- ometric power＂ of the beamformer.s output snapshots, subjected to a linear constraint. Therefore, the proposed beamformer can provide significantly higher output geometric signal-to-noise-andinterference ratio. Moreover, the optimum weight vector is obtained by using a new iteration process. The simulation results prove that the new method is effective.展开更多
An improved ensemble empirical mode decomposition(EEMD) algorithm is described in this work, in which the sifting and ensemble number are self-adaptive. In particular, the new algorithm can effectively avoid the mode ...An improved ensemble empirical mode decomposition(EEMD) algorithm is described in this work, in which the sifting and ensemble number are self-adaptive. In particular, the new algorithm can effectively avoid the mode mixing problem. The algorithm has been validated with a simulation signal and locomotive bearing vibration signal. The results show that the proposed self-adaptive EEMD algorithm has a better filtering performance compared with the conventional EEMD. The filter results further show that the feature of the signal can be distinguished clearly with the proposed algorithm, which implies that the fault characteristics of the locomotive bearing can be detected successfully.展开更多
This paper investigates the main scale analysis of the aerodynamic noise in the foremost bogie area by the large-eddy simulation(LES)and the Ffowcs Williams-Hawkings(FW-H)analogy.The mechanism of the aerodynamic noise...This paper investigates the main scale analysis of the aerodynamic noise in the foremost bogie area by the large-eddy simulation(LES)and the Ffowcs Williams-Hawkings(FW-H)analogy.The mechanism of the aerodynamic noise in this area has been excavated.The aerodynamic excitation results show that the bogie divides the bogie compartment into two cavities,each of which contains a large circulating flow and presents multi-peak characteristics in the frequency domain.The far-field noise results suggest that in the speed range of 200−350 km/h,the aerodynamic noise mechanism in the bogie area is the same.Cavity noise is the main noise mechanism in the foremost bogie area,and the bogie divides the bogie cabin into two cavities,thereby changing the aerodynamic noise in this area.展开更多
Performance of Turbo-Codes in communication channels with impulsive noise is analyzed. First, mathematical model of impulsive noise is presented because it has non-Gaussian nature and is found in many wireless channel...Performance of Turbo-Codes in communication channels with impulsive noise is analyzed. First, mathematical model of impulsive noise is presented because it has non-Gaussian nature and is found in many wireless channels due to impulsive phenomena of radio-frequency interference. Then, with linear Log-MAP decoding algorithm for its low complexity, Turbo-Codes are adopted and analyzed in such communication channels. To confirm the performance of the proposed method, simulations on both static and fully interleaved flat Rayleigh fading channels with impulsive noise have been carried out. It is shown that Turbo-Codes have a better performance than the conventional methods （e.g. convolutionally coded system）.展开更多
For the robustness problem of open-loop P-type iterative learning control under the influence of measurement noise which is inevitable in actual systems, an adaptive adjustment algorithm of iterative learning nonlinea...For the robustness problem of open-loop P-type iterative learning control under the influence of measurement noise which is inevitable in actual systems, an adaptive adjustment algorithm of iterative learning nonlinear gain matrix based on error amplitude is proposed and two nonlinear gain functions are given. Then with the help of Bellman-Gronwall lemma, the robustness proof is derived. At last, an example is simulated and analyzed. The results show that when there exists measurement noise, the proposed learning law adjusts the learning gain matrix on line based on error amplitude, thus can make a compromise between learning convergence rate and convergence accuracy to some extent： the fast convergence rate is achieved with high gain in initial learning stage, the strong robustness and high convergence accuracy are achieved at the same time with small gain in the end learning stage, thus better learning results are obtained.展开更多
To improve the feature extraction of ship-radiated noise in a complex ocean environment,a novel feature extraction method for ship-radiated noise based on complete ensemble empirical mode decomposition with adaptive s...To improve the feature extraction of ship-radiated noise in a complex ocean environment,a novel feature extraction method for ship-radiated noise based on complete ensemble empirical mode decomposition with adaptive selective noise(CEEMDASN) and refined composite multiscale fluctuation-based dispersion entropy(RCMFDE) is proposed.CEEMDASN is proposed in this paper which takes into account the high frequency intermittent components when decomposing the signal.In addition,RCMFDE is also proposed in this paper which refines the preprocessing process of the original signal based on composite multi-scale theory.Firstly,the original signal is decomposed into several intrinsic mode functions(IMFs)by CEEMDASN.Energy distribution ratio(EDR) and average energy distribution ratio(AEDR) of all IMF components are calculated.Then,the IMF with the minimum difference between EDR and AEDR(MEDR)is selected as characteristic IMF.The RCMFDE of characteristic IMF is estimated as the feature vectors of ship-radiated noise.Finally,these feature vectors are sent to self-organizing map(SOM) for classifying and identifying.The proposed method is applied to the feature extraction of ship-radiated noise.The result shows its effectiveness and universality.展开更多
文摘For space-borne gravitational wave detection missions based on the heterodyne interferometry principle,tilt-to-length(TTL)coupling noise is an important optical noise source,significantly influencing the accuracy of the measurement system.We present a method for analyzing TTL coupling noise under the joint influence of multiple factors.An equivalent simulated optical bench for the test mass interferometer was designed,and Gaussian beam tracing was adopted to simulate beam propagation.By simulating the interference signal,it can analyze the impact of various factors on the TTL coupling noise,including positional,beam parameters,detector parameters,and signal definition factors.On this basis,a random parameter space composed of multiple influential factors was constructed within a range satisfying the analysis requirement,and the corresponding simulation results from random sampling were evaluated via variance-based global sensitivity analysis.The calculated results of the main and total effect indexes show that the test mass rotation angle and the piston effect(lateral)significantly influence the TTL coupling noise in the test mass interferometer.The analysis provides a qualitative reference for designing and optimizing space-borne laser interferometry systems.
文摘In this work,the generation of high signal-to-noise ratio(SNR)single-frequency microwave signal without noise sidebands is demonstrated based on the interaction of integrated all-fiber lasers.The microwave signals are generated by the interference between a narrow linewidth Brillouin pump light from a single-frequency laser and the Stokes light generated by it.Firstly,the linewidths of the Stokes lights are compressed to~43 Hz based on the stimulated Brillouin scattering(SBS)effect,which ensures that the frequency noise is as low as possible.And then,the relative intensity noise(RIN)of the first order Stokes light is reduced by 21 dB/Hz based on the noise dynamics principle in cascaded SBS effect.By simultaneously reducing the frequency noise and the intensity noise of the coherent signals,the noise sidebands of microwave signals are completely suppressed.As result,the SNR of the microwave signal is improved from 48 dB to 84 dB at the first-order Brillouin frequency shift of 9.415 GHz.Meanwhile,a microwave signal with a SNR of 70 dB is generated at the second-order Brillouin frequency shift of 18.827 GHz.This kind of microwave signals with narrow linewidth and high SNR can provide higher detection resolution and higher transmission efficiency for applications on radar,satellite communication and so on.
基金supported by Hefei National Laboratory,Innovation Program for Quantum Science and Technology(2021ZD0300400/2021ZD0300402)the Beijing Natural Science Foundation(3252013)the China Postdoctoral Science Foundation(2024T171116).
文摘The spin-exchange relaxation-free atomic gyroscope,with its exceptionally high theoretical precision,demonstrates immense potential to become the next-generation strategic-grade gyroscope.However,due to technological noise,there is still a significant gap between its actual precision and theoretical precision.This study identifies the key factor limiting the precision of the SERF gyroscope as coupling noise.By optimizing the detection loop structure,a distinction between the dual-axis signals'response to optical and magnetic fields was achieved-where the optical errors responded similarly,while the response to magnetic noise was opposite.Based on the differences in the optical-magnetic response of the dual-axis signals,empirical mode decomposition was used to decompose the dual-axis gyroscope signals into multiple intrinsic mode functions,and Allan deviation analysis was applied to analyze the noise characteristics of the intrinsic mode functions over various periods.This study successfully reveals that optical errors caused by thermal-optical coupling and long-period magnetic noise induced by thermal-magnetic coupling are the dominant factors limiting the long-term stability of the SERF gyroscope.Based on these analyses,the study concludes that to achieve strategic-grade precision for the SERF gyroscope,it is essential to effectively address the noise issues caused by multi-physical field couplings.
基金supported by the National Natural Science Foundation of China(62071097)the Sichuan Science and Technology Program(2023NSFSC0458).
文摘In this paper,we study the orthogonal time frequency space signal transmission over multi-path channel in the presence of phase noise(PHN)at both sides of millimeter wave(mmWave)communication links.The statistics characteristics of the PHN-induced common phase error and inter-Doppler interference are investigated.Then,a column-shaped pilot structure is designed,and training pilots are used to realize linear-complexity PHN tracking and compensation.Numerical results demonstrate that the proposed scheme enables the signal to noise ratio loss to be restrained within 1 dB in contrast to the no PHN case.
文摘针对电力线信道噪声干扰大、噪声建模复杂、重建无噪信号样本困难等问题,提出了一种面向一维时间序列的Noise2Noise神经网络优化模型,并验证分析Noise2Noise算法抑制电力线噪声的可行性。首先,分析了Noise2Noise算法的原理,对该算法原理进行了理论推导。其次,通过选取合适的网络结构,改进网络输入输出,使神经网络适合处理实验数据,并采用正弦波数据进行网络测试与模型验证。然后,通过搭建正交频分复用(orthogonal frequency division multi-plexing,OFDM)调制模型,产生OFDM符号,并添加特定噪声,形成带噪样本。最后,基于改进的网络结构实现了对电力线接收信号的OFDM前导序列进行带噪样本测试,增强了Noise2Noise方法的可行性与有效性,具有较好的实用性。
基金supported by the National Natural Science Foundation of China under Grant Nos.61370195and 11101048Beijing Natural Science Foundation under Grant No.4132060the National Cryptography Development Foundation of China under Grant No.MMJJ201201002
文摘Recently, the digital image blind forensics technology has received an increasing attention in academic community. This paper aims at developing a new identification approach based on the statistical noise and exchangeable image file format （EXIF） information of image for images authen- tication. In particular, the authors can identify whether the current image has been modified or not by utilizing the relevance between noise and EXIF parameters and comparing the real values with the estimated values of the EXIF parameters. Experimental results validate the proposed method. That is, the detecting system can identify the doctored image effectively.
文摘In radar target tracking application, the observation noise is usually non-Gaussian, which is also referred as glint noise. The performances of conventional trackers degra de severely in the presence of glint noise. An improved particle filter, Markov chain Monte Carlo particle filter (MCMC-PF), is applied to cope with radar target tracking when the measurements are perturbed by glint noise. Tracking performance of the filter is demonstrated in the present of glint noise by computer simulation.
基金supported by the National Natural Science Foundation of China(6129032461473164+1 种基金61490701)the Research Fund for the Taishan Scholar Project of Shandong Province of China(LZB2015-162)
文摘To predict the remaining useful life(RUL) for a class of nonlinear multi-degradation systems, a method is presented. In the real industrial processes, systems are usually composed by several parts or components, and these parts or components are working in the same environment, thus the degradations of these parts or components will be influenced by common factors. To describe such a phenomenon in degradations, a multi-degradation model with public noise is proposed. To identify the degradation states and the unknown parameters, an iterative estimation method is proposed by using the Kalman filter and the expectation maximization(EM) algorithm. Next, with known thresholds,the RUL of each degradation can be predicted by using the first hitting time(FHT). In addition, the RUL of the whole system can be obtained by a Copula function. Finally, a practical case is used to demonstrate the method proposed.
基金supported by Aviation Science Foundation(20070851011).
文摘Stochastic noises of fiber optic gyroscope （FOG） mainly contain white noise and fractal noise whose long-term dependent component causes FOG a rather slow drift. In order to eliminate this component, a two-step filtering methodology is proposed. Firstly, fractional differencing （FD） method is introduced to trans-form fractal noise into fractional white noise based on the estima-tion of Hurst exponent for long-term dependent fractal process, which together with the existing white noise make up of a gener-alized white noise. Further, an improved denoising algorithm of wavelet maxima is developed to suppress the generalized white noise. Experimental results show that the basic noise terms of FOG greatly decrease, and especially the slow drift is restrained effectively. The proposed methodology provides a promising ap-proach for filtering long-term dependent fractal noise.
基金supported by the National High Technology Research and Development Program of China (863 Program) (2007AA701403)
文摘A minimum geometric power distortionless response beamforming approach against impulsive noise （including all α- stable noise） of unknown statistics is proposed. Due to that definite logarithmic moments require no priori knowledge of impulsive noise, this new beamformer substitutes the logarithmic moments for the second-order moments and iteratively minimizes the ＂ge- ometric power＂ of the beamformer.s output snapshots, subjected to a linear constraint. Therefore, the proposed beamformer can provide significantly higher output geometric signal-to-noise-andinterference ratio. Moreover, the optimum weight vector is obtained by using a new iteration process. The simulation results prove that the new method is effective.
基金Project(61573381)supported by the National Natural Science Foundation of ChinaProject(2012AA051601)supported by the National High-tech Research and Development Program of China
文摘An improved ensemble empirical mode decomposition(EEMD) algorithm is described in this work, in which the sifting and ensemble number are self-adaptive. In particular, the new algorithm can effectively avoid the mode mixing problem. The algorithm has been validated with a simulation signal and locomotive bearing vibration signal. The results show that the proposed self-adaptive EEMD algorithm has a better filtering performance compared with the conventional EEMD. The filter results further show that the feature of the signal can be distinguished clearly with the proposed algorithm, which implies that the fault characteristics of the locomotive bearing can be detected successfully.
基金Project(2017YFB1201103)supported by the National Key Research and Development Plan of ChinaProject(2019zzts540)supported by the Graduate Student Independent Innovation Project of Central South University,China。
文摘This paper investigates the main scale analysis of the aerodynamic noise in the foremost bogie area by the large-eddy simulation(LES)and the Ffowcs Williams-Hawkings(FW-H)analogy.The mechanism of the aerodynamic noise in this area has been excavated.The aerodynamic excitation results show that the bogie divides the bogie compartment into two cavities,each of which contains a large circulating flow and presents multi-peak characteristics in the frequency domain.The far-field noise results suggest that in the speed range of 200−350 km/h,the aerodynamic noise mechanism in the bogie area is the same.Cavity noise is the main noise mechanism in the foremost bogie area,and the bogie divides the bogie cabin into two cavities,thereby changing the aerodynamic noise in this area.
文摘Performance of Turbo-Codes in communication channels with impulsive noise is analyzed. First, mathematical model of impulsive noise is presented because it has non-Gaussian nature and is found in many wireless channels due to impulsive phenomena of radio-frequency interference. Then, with linear Log-MAP decoding algorithm for its low complexity, Turbo-Codes are adopted and analyzed in such communication channels. To confirm the performance of the proposed method, simulations on both static and fully interleaved flat Rayleigh fading channels with impulsive noise have been carried out. It is shown that Turbo-Codes have a better performance than the conventional methods （e.g. convolutionally coded system）.
基金supported by the Specialized Research Fund for the Doctoral Program of Higher Education(20106102110032)
文摘For the robustness problem of open-loop P-type iterative learning control under the influence of measurement noise which is inevitable in actual systems, an adaptive adjustment algorithm of iterative learning nonlinear gain matrix based on error amplitude is proposed and two nonlinear gain functions are given. Then with the help of Bellman-Gronwall lemma, the robustness proof is derived. At last, an example is simulated and analyzed. The results show that when there exists measurement noise, the proposed learning law adjusts the learning gain matrix on line based on error amplitude, thus can make a compromise between learning convergence rate and convergence accuracy to some extent： the fast convergence rate is achieved with high gain in initial learning stage, the strong robustness and high convergence accuracy are achieved at the same time with small gain in the end learning stage, thus better learning results are obtained.
基金supported by the National Natural Science Foundation of China under Grant 51709228。
文摘To improve the feature extraction of ship-radiated noise in a complex ocean environment,a novel feature extraction method for ship-radiated noise based on complete ensemble empirical mode decomposition with adaptive selective noise(CEEMDASN) and refined composite multiscale fluctuation-based dispersion entropy(RCMFDE) is proposed.CEEMDASN is proposed in this paper which takes into account the high frequency intermittent components when decomposing the signal.In addition,RCMFDE is also proposed in this paper which refines the preprocessing process of the original signal based on composite multi-scale theory.Firstly,the original signal is decomposed into several intrinsic mode functions(IMFs)by CEEMDASN.Energy distribution ratio(EDR) and average energy distribution ratio(AEDR) of all IMF components are calculated.Then,the IMF with the minimum difference between EDR and AEDR(MEDR)is selected as characteristic IMF.The RCMFDE of characteristic IMF is estimated as the feature vectors of ship-radiated noise.Finally,these feature vectors are sent to self-organizing map(SOM) for classifying and identifying.The proposed method is applied to the feature extraction of ship-radiated noise.The result shows its effectiveness and universality.
