103 results about "Marginal spectrum" patented technology

The invention relates to a comprehensive single-end fault positioning method applicable to power distribution network lines. The comprehensive single-end fault positioning method comprises the following steps: firstly, extracting intrinsic frequency of transient state traveling waves by using a signal marginal spectrum, calculating by using a frequency domain method so as to obtain initial fault distance, estimating transition resistance according to the initial fault distance and power frequency, and confirming a reflection wave recognition time window and reflection wave polarity according to the estimation result of the initial fault distance and the transition resistance, thereby realizing reliable reflection wave recognition and finally completing precise single-end fault positioning on the power distribution network lines on the basis. By adopting the method, an impedance method, a frequency domain method and a traveling wave method are fused, the position of a fault point of the power distribution network lines is precisely confirmed, and the reliability and the precision in positioning faults of the power distribution network are improved.

The invention discloses a mill load parameter soft measurement method, which comprises the steps of firstly decomposing mill cylinder vibrations and vibration sound signals into sub-signals (intrinsic mode functions (IMFs)) with different time scales and different physical meanings by adopting an ensemble empirical mode decomposition (EEMD) technology; then selecting of three kinds of features of the multi-scale IMF by adopting mutual information (MI) based adaptive feature selection approach, wherein the three kinds of features are the frequency spectrum, the marginal spectrum, and the mean and the standard deviation of the instantaneous amplitude and the frequency of Hilbert transformation; and finally, constructing a soft measurement model based on selective integration kernel partial least squares (KPLS) method on the basis of selected spectrum features and a training sample. A small ball mill based simulation experiment result shows that the method disclosed by the invention can effectively detect load parameters.

The invention provides a heartbeat frequency detection algorithm of a non-contact vital sign detection system; the steps are: respectively doing bandpass filtering for I and Q two way signals outputted by a continuous wave doppler radar; using a center estimate algorithm to carry out useful dc component recovery; using a complete cluster experience modal decomposition algorithm to separate a heartbeat signal from a human body jitter signal, a breathing signal and environment interference noises; resolving a Hilbert marginal spectrum of each decomposition signal and doing peak value detection; finding out the Hilbert marginal spectrum corresponding to the heartbeat signal according to a marginal spectrum peak position and energy concentration degree close to the spectrum peak; obtaining heartbeat frequency information according to the spectrum peak position. The heartbeat frequency detection algorithm can effectively extract the heartbeat signal under unstable human body and large environment interference noises, thus obtaining accurate heartbeat frequency information, improving anti-interference property of the non-contact vital sign detection system, and satisfying heartbeat frequency detection accuracy demands of medical affairs personnel.

The invention discloses a time-varying non-stable-signal time-frequency analyzing method. The method comprises the steps of providing definitions of single-frequency signals and i(th)-level local extremum, conducting single-frequency signal decomposition on the time-varying non-stable signals according to the definitions so as to obtain a plurality of single-frequency components, then obtaining a plurality of instantaneous frequency and instantaneous amplitude of the components by Hilbert, further, drawing the amplitude on a time-frequency plane in an amplitude-frequency-time three-dimensional space to obtain a Hilbert amplitude spectrum, and conducting integration on the time through the amplitude spectrum to obtain a Hilbert marginal spectrum. According to the method, the complete adaptability is realized; meanwhile, the method is not restricted by uncertainty principle, and the time and the frequency have high resolution ratio; the decomposition is thorough, so that the respectively independent single-frequency components, namely the substantive characteristics, are obtained, and cross terms among all the components do not exist; and the calculating time is saved as the accelerating method is adopted in the decomposition process.

A vibration signal processing method based on HHT (Hilbert-Huang Transformation) and related analyses includes steps as follows: using EMD (empirical mode decomposition) to decompose vibration signals; carrying out related analyses to each mode component obtained through decomposition; carrying out Hilbert transformation to denoised signals and obtaining a Hilbert spectrum. Aiming at the defect that noisy signals cannot be distinguished in signals if the HHT method is applied directly, the invention provides the method based on HHT and the related analyses, and denoises noisy signals. Through analyses on the Hilbert spectrum of the extracted noisy mode components and a marginal spectrum, the frequency and amplitude information of noisy vibration signals can be effectively extracted. The method can be used for processing signals of metallurgical machinery, aerospace, hydropower engineering, aeromancy and so on, and effectively remove noise.

The invention provides a method for recognizing over-voltage fault types of a power distribution network. The method comprises the following steps: acquiring zero sequence voltage waveform sampling data x(t); taking an IMF (Intermediate Frequency) signal with a maximum correlation coefficient C as a principle mode component y(t); calculating a marginal spectrum h(omega) of the y(t); judging that a frequency dividing resonance fault occurs when an energy ratio of a frequency band of 10-40Hz in h(omega) exceeds 50%; judging that a high-frequency resonance fault occurs when an energy ratio of a frequency band of 100-150Hz in h(omega) exceeds 50%; otherwise calculating overall similarity S of the x(t) and standard single-phase earth fault zero sequence voltage waveform; judging that a fundamental resonance fault occurs when S is smaller than a preset similarity threshold K, otherwise judging that a single-phase earth fault occurs. The method disclosed by the invention is capable of automatically recognizing the over-voltage resonance type and the single-phase earth fault, simple in algorithm, less in calculation time, high in anti-jamming capability and high in recognition accuracy.

An embodiment of the invention discloses a method for predicting a failure of an aero-engine, which comprises the following steps: collecting parameter data in an engine recording system to obtain anengine vibration signal, and determining a training data set and test data based on the parameter data; constructing an LSTM neural network model by using a python platform, and determining a signal margin spectrum according to the training data set and the LSTM (Long Short Term Memory) neural network; constructing a BP neural network model by using the python platform, training a BP (Backward Propagation) neural network by using a marginal spectrum characteristic training set, and acquiring and storing parameters of the BP neural network after convergence of training; and predicting a failureof the aero-engine according to the trained BP neural network model. Through adoption of the method of the invention, the failure of the aero-engine can be predicted.

The invention discloses a shock echo signal analysis method based on variational mode decomposition, which comprises the following steps: 1) selecting measurement parameters and working parameters ofequipment according to requirements of field measurement environment, and collecting shock echo signals; 2) setting parameters in variational modal decomposition according to the shock echo signals; 3) decomposing the collected shock echo signals into several eigen modulus functions by a variational modal decomposition method to solve the variational problem; 4) using Hilbert transform to obtain the Hilbert time-frequency spectrum of the eigen modulus functions; 5) integrating the Hilbert time-frequency spectrum of different frequencies in the time domain to obtain the final marginal spectrum.The method identifies the defects of the concrete component by using the marginal spectrum of the shock echo signals, and can suppress the noise interference under strong noise, and achieves higher resolution compared with the traditional Fourier transform.

The invention discloses a method for diagnosing operating state of windings in short-circuiting of a transformer. The method includes the steps of firstly, acquiring a vibration signal of a box wall of the transformer when the short-circuiting of the transformer occurs; secondly, pre-continuing left and right ends of the vibration signal; thirdly, decomposing the continued vibration signal into a plurality of intrinsic-mode function components; fourthly, subjecting all the intrinsic-mode function components obtained by decomposing to Hilbert transformation so as to obtain Hilbert spectrum of the vibration signal; fifthly, acquiring Hilbert marginal spectrum and Hilbert energy according to the Hilbert spectrum; and sixthly, judging the state of the transformer windings according to changes in the Hilbert marginal spectrum and the Hilbert energy.

The invention provides a method and apparatus for diagnosing blades of a wind turbine. The method includes steps of acquiring, via a microphone, an operation sound of the wind turbine when the wind turbine is under operation; transforming the operation sound into a time-frequency spectrum; integrating the time-frequency spectrum over time to generate a marginal spectrum; determining whether any blade of the wind turbine is damaged according to the marginal spectrum and a reference curve.

The invention discloses a current transformer (CT) saturation detection method based on Hilbert-Huang transformation (HHT), and the method comprises the followings steps that instantaneous frequency spectrum, Hilbert spectrum and Hilbert marginal spectrum are obtained through EMD decomposition and Hilbert diversion by collecting differential current signals on both sides of a transformer; the detection of saturation faults and CT saturation outside a transformer area is completed respectively arranged to the three types of spectrum so as to realize the quick action and the reliable action of differential protection under CT saturation; the method serves as a preferential improvement to a traditional CT saturation detection method and combines instantaneous frequency criteria, Hilbert spectrum criteria and Hilbert marginal spectrum criteria in the forms of 'and' and 'or'; when a judgment result has errors, another method can correctly judge, so that the reliability of differential protection is improved; and in addition, fault current outside the transformer area, fault current in the area, conversion fault current and the like under the two circumstances of CT saturation and unsaturation are fully considered, and the CT saturation detection method based on HHT has the advantages of stronger functions, higher efficiency, higher reliability and the like.

The invention relates to an early fault determining method for a bearing. The method comprises the steps of cutting an acquired bearing vibration time domain signal into N groups of secondary vibration signals according to the same length; 2, calculating local spectral band energy Mg of each secondary vibration signal, namely, 1) performing frequency domain conversion for each secondary vibration signal; 2) selecting a local spectral band from the whole Hilbert marginal spectrum based on the formula shown in specification as the interval, wherein fp is the bearing fault characteristic frequency calculated based on the bearing structural dimension, and delta f is 2Hz; 3) calculating the local spectral band energy through the formula shown in the specification, wherein h(f) is frequency amplitude in the local spectral band; 3, creating a local spectral band energy sequence through N Mg, wherein the local spectral band energy value of each vibration signal is Mf when the bearing is free of a fault; Mg is not less than the product of K and Mf at M times in the sequence, wherein K is a constant; the characteristic power rate (CPR) is determined according the formula shown in the specification; 4, determining the early fault when the CPR is more than or equal to some constant A.

The invention discloses a wheel polygon identification method based on the wheel-rail vertical force and a device thereof. The identification method comprises the following steps: collecting the wheel-rail vertical force signal by the shear stress method; and analyzing the collected wheel-rail vertical force signal Carry out the EEMD decomposition of the ensemble empirical mode to obtain the aggregated intrinsic mode function IMF component; calculate the marginal spectrum of the intrinsic mode function IMF containing the wheel polygonal disease characteristic information; according to the obtained marginal spectrum of the intrinsic mode function IMF, obtain The damage characteristic frequency f is used to identify whether the wheel has polygons and the type of polygons according to the preset judgment criteria. Its advantages are: it can carry out real-time monitoring and damage evaluation on the vehicles passing through the monitoring section, and it has the characteristics of fast and accurate; The marginal spectrum of the wheel-rail vertical force damage characteristic frequency after polygonal damage of the wheel is obtained.

The invention discloses an identity recognition algorithm based on heart sound multi-dimension feature extraction and a system thereof. The method comprises the steps that heart sound signals are collected through sensors, and the collected heart sound signals are first processed through a filter and then processed through discrete wavelet transform to obtain relative pure signals; the Mel-frequency cepstral coefficients and the Hilbert marginal spectrum of the heart sound signals are extracted through a computer; template features are extracted and formed; normalization is carried out on the template features, and feature selection is carried out through a PCA algorithm to establish a more perfect low-dimension template feature space; training features are trained through a KNN algorithm; test features to be tested are tested through an established classifier. According to the identity recognition algorithm based on the heart sound multi-dimension feature extraction and the system thereof, pattern recognition is introduced into an identity matching algorithm, the learning capacity and the rapid operational capability of the computer are utilized, training matching is carried out by adopting a great amount of data, and recognition speed and accuracy are further improved.

The invention discloses a spectrum sensing method and device based on Hilbert-Huang transform, wherein the method comprises the steps of: obtaining a signal to be detected, and performing EMD decomposition of the signal to be detected, so that n corresponding IMF components are obtained, wherein the signal to be detected is a wireless spectrum signal; obtaining the Hilbert spectrum of each IMF component, and superposing the Hilbert spectrums respectively corresponding to the first IMF component to the ith IMF component, so that the total Hilbert spectrum is obtained, wherein i is less than n; accumulating the total Hilbert spectrum in terms of time to obtain a marginal spectrum; and, judging whether the marginal spectrum is greater than or equal to the pre-set judgement threshold value or not, if so, determining that a main user signal exits in the signal to be detected, and if not, determining that the main user signal does not exist in the signal to be detected. By means of the technical scheme provided by the invention, the local feature of the wireless spectrum signal can be reflected accurately; and thus, the spectrum sensing accuracy rate can be greatly improved.

The invention relates to a high speed railway rail corrugation acoustic diagnosis method based on an IMF (Intrinsic Mode Function) energy ratio, and belongs to the technical field of high speed railway vibration noise. The method comprises the following steps that: (1) testing rail roughness; (2) carrying out ensemble empirical mode decomposition; and (3) obtaining an IMF energy ratio: according to the energy ratio of an IMF signal corresponding to fault characteristic frequency, carrying out rail corrugation fault identification, screening to obtain an IMF component corresponding to the railcorrugation, and obtaining a Hilbert marginal spectrum and instantaneous frequency through HHT (Hilbert-Huang Transform). By use of the high speed railway rail corrugation acoustic diagnosis method based on the IMF energy ratio, a direct method is adopted on an operation line to actually measure the rail roughness characteristics of a corrugation-free zone, the IMF energy ratio obtained after an acoustic signal is subjected to EEMD (Ensemble Empirical Mode Decomposition) is subjected to component screening, and IMF energy ratio distortion characteristics are used for carrying out rail corrugation identification. Compared with theoretical acoustic frequency corresponding to the rail roughness actually measured by the direct method, the invention puts forward an efficient high speed railwayrail corrugation acoustic diagnosis strategy.

The invention provides a gyroscope fault diagnosis method based on K-S (Kolmogorov-Smirnov) distribution check and HHT (Hilbert-Huang Transform), relates to a fault diagnosis method of a gyroscope, and mainly solves the problems of an existing gyroscope fault diagnosis method that a virtual frequency component is generated and the fault diagnosis precision is low. The gyroscope fault diagnosis method comprises the following steps: step 1: decomposing an original gyroscopic angle speed output signal Xp by adopting an EMD (Empirical Mode Decomposition) method to obtain IMF (Intrinsic Mode Function) components of different frequency bands; step 2: carrying out a correlation test on the IMF components of the different frequency bands in the step 1 by using a K-S distribution check method and judging that whether the IMF components of the different frequency bands are effective components of the original gyroscopic angle speed output signal or not; and step 3: carrying out the HHF on the IMF components checked by the K-S method in the step 2 to further obtain time-frequency spectrums and marginal spectrums of the IMF components, and combining with energy and frequency change of signals on the time-frequency spectrums and signal frequency distribution on the marginal spectrums to judge whether the system has faults in an operation process or not. The gyroscope fault diagnosis method is applied to the field of signal processing.

The invention discloses a vibration signal analysis method for a main bearing of a wind power transmission system based on improved HHT and fuzzy entropy. The method comprises the steps of collectinga vibration signal, performing EMD decomposition on the vibration signal, screening a real IMF component by a relative entropy theory, and performing Hilbert spectrum analysis to obtain marginal spectrum through synthesis. According to the method, the fault characteristics of the main bearing of the wind power transmission system can be stably and effectively extracted, and the interference of thenoise to a characteristic frequency can be effectively filtered. A false eigenmode function component generated by empirical mode decomposition is eliminated with proposed improved Hilbert-Huang transform, the marginal spectrum synthesized by the real component is made, and the true frequency distribution of the signal is reflected. Combined with the fuzzy entropy theory, an eigenmode function which best reflects fault bearing characteristics is screened out.

The invention relates a method for analyzing a time-intensity curve of a myocardium ultrasonic angiography image, in particular to a method for measuring myocardium ultrasonic angiography image physiological parameters based on EMD. The method includes dividing a myocardial area into six areas according to a circumference, extracting perfusion signals and extracting a time-intensity curve of each of divided myocardial areas respectively; performing EMD on an extracted time-average gray value intensity curve to obtain an expanded signal intrinsic mode function, performing Hilbert transform on the obtained first intrinsic mode function to obtain instantaneous spectrum parameters, and adding amplitude values of each frequency point to obtain a marginal spectrum; and performing threshold judgment on a marginal spectrum through energy thresholds of the marginal spectrum. According to the method, an analytical approach based on the myocardium ultrasonic angiography image is provided for diagnosing myocardial infarction or myocardial microcirculation, and further physiological parameter measurement accuracy based on the myocardium ultrasonic angiography image is improved, and doctor subjective dependence is reduced.

The present invention provides a method and device for extracting a vibration signal characteristic frequency band. The method comprises a step of extracting the marginal spectrum of a vibration signal, a step of dividing the marginal spectrum into a plurality of window marginal spectrums by using a sliding window and carrying out clustering analysis on the window marginal spectrum sets under the same frequency band in different fault states so as to calculate and generate the clustering effect evaluation index of each frequency band window marginal spectrum set, and a step of extracting a fault sensitive characteristic frequency band according to the clustering effect evaluation index. The vibration signal characteristic frequency band extracted by the method and device of the invention has the advantages of high fault recognition rate and good noise immunity ability quality.

The invention provides a vibration type flow meter characteristic signal extraction method based on improved assemble average empirical mode decomposition. The method includes the steps of conducting end continuation processing on a collected vibration type flow meter flow vibration signal through a waveform-matched self-adaption end continuation method, conducting envelope line fitting on the collected vibration signal through a cubic B-spline method, conducting MEEMD decomposition to obtain a plurality of IMF components, conducting relevance analysis on the IMF components and the original signal, selecting the useful IMF components, conducting HHT conversion on the IMF components, and obtaining the Hilbert time-frequency spectrum and the marginal spectrum of the flow signal, wherein the Hilbert time-frequency spectrum and the marginal spectrum are the signal characteristics of the vibration type flow meter flow vibration signal. The method is suitable for accurately and rapidly metering the pipe network fluid flow in the industrial field.

The invention provides a rolling bearing fault diagnosis method based on optimized variational mode decomposition, and the method comprises the steps: selecting 4096 sampling points of an original vibration signal as input signals of variational mode decomposition; optimizing the modal number and the secondary penalty factor of variational modal decomposition by adopting an improved bat algorithmand taking the minimum average envelope entropy as an optimization target; decomposing the original vibration signal by using the optimized parameters, and solving an energy entropy and an energy spectrum entropy of a decomposed component; taking the kurtosis, the correlation coefficient and the marginal spectrum entropy as screening criteria to screen the components, and solving main frequency distribution characteristics of the reserved components; and taking the energy entropy, the energy spectrum entropy and the main frequency distribution characteristics as characteristic vectors and inputting the characteristic vectors into a support vector machine so as to realize fault diagnosis. According to the method, the variational mode decomposition parameters are optimized through the improved bat algorithm, and the feature vectors are obtained according to the optimized parameters, so that manual parameter determination is avoided, the optimal solution can be found more quickly, and therecognition rate of the fault state is improved.

The invention discloses a speech emotion recognition method based on variational modal decomposition and an extreme learning machine, and belongs to the field of artificial intelligence and speech recognition. The speech emotion recognition method first preprocesses an emotional speech signal through a variational modal decomposition method, the emotional speech signal is decomposed into a plurality of intrinsic mode function (IMF) components and a residual component, and the components can reflect the change of a original sequence more accurately and retain the emotional characteristics of the speech signals; then, the IMF components are subjected to hilbert conversion to obtain hillbert marginal spectrum characteristics of the IMF components; and in addition, the IMF components are reaggregated to obtain the speech signal removing the residual component, and then an MEL cepstrum function is extracted for the signal. The extracted new characteristics are added into a traditional speech emotional characteristic set, and an extreme learning machine model is constructed for classification and recognition. The speech emotion recognition method has the advantage of obtaining new speechcharacteristics through the variational modal decomposition. Compared with the traditional speech emotional characteristics, the characteristics have a higher recognition rate for speech emotion recognition.

The invention discloses a transformer winding state diagnosis method. The transformer winding state diagnosis method comprises a first step of collecting vibration signals of a transformer box wall when a transformer is short-circuited; a second step of performing prolongation pretreatment on left end points and right end points of the vibration signals; a third step of decomposing the vibration signals undergone prolongation into a plurality of product function components; a fourth step of enabling all the product function components obtained by decomposition to undergo Hilbert transform, and obtaining a Hilbert spectrum of the vibration signals; a fifth step of obtaining a Hilbert marginal spectrum and Hilbert energy according to the Hilbert spectrum; and a sixth step of distinguishing a state of a transformer winding according to the Hilbert marginal spectrum and variations of the Hilbert energy.

The invention discloses a grating-scale-measuring-error dynamic compensation method based on deep learning. The rating-scale-measuring-error dynamic compensation method includes the steps that error data is collected and obtained, and meanwhile effect intensity values of a plurality of interference factors corresponding to the error data are measured and obtained through a plurality of sensors; the error data is decomposed into a plurality of IMF components based on the empirical mode decomposition algorithm, and Hilbert marginal spectrums of all the IMF components are solved and obtained; the effect intensity values of the multiple interference factors corresponding to the error data and the Hilbert marginal spectrums of the multiple IMF components serve as input data, and identification and calculation are carried out through a trained CNN neural network to obtain correspondingly-output label functions; trend terms corresponding to the error data are obtained and accumulated to serve as the error compensation value of a grating scale; the grating scale is subjected to measurement compensation through the obtained error compensation value. The grating-scale-measuring-error dynamic compensation method is easy to operate, low in cost and good in compensation effect, effective compensation of a grating-scale system can be achieved, and the grating-scale-measuring-error dynamic compensation method can be widely applied to grating-scale measuring industries.

The invention discloses a detection method of a steel pipe concrete cavity defect extracted on the basis of HHT characteristics. The method comprises the following steps: (1) applying mechanical impact on a detection point on the surface of steel pipe concrete to excite stress waves, using a sensor to receive the stress wave detection signals; (2) adopting an HHT algorithm to extract cavity characteristic signals from the stress wave detection signals to obtain a characteristic Hilbert marginal spectrum; (3) changing the detection point, repeating step (1) and (2); subjecting the characteristic Hilbert marginal spectrums corresponding to different detection points to an image wave train display; and judging the size of cavity of steel pipe concrete according to the characteristic Hilbert marginal spectrum. By accumulatively displaying the images of characteristic Hilbert marginal spectrums of multiple detection points, the situation of cavity defects in steel pipe concrete can be directly display and calculated, and the method has the advantages of strong reliability, and accurate detection result.

The invention discloses an energy identification method for slope earthquake damage. The method comprise the following steps that measured earthquake wave time interval is pretreated; empirical mode decomposition is carried out to generate a plurality of intrinsic mode functions; Hilbert transform is carried out on each intrinsic mode function to obtain a time-frequency spectrum of each intrinsicmode function; a marginal spectrum of a corresponding earthquake wave time interval is obtained according to the time-frequency spectrum of each intrinsic mode function; quantitative identification iscarried out on earthquake wave energy in a slope according to the obtained marginal spectrums; the earthquake wave energy distribution in a slope body is determined, and the position where earthquakedamage in the slope body occurs is concluded according to the earthquake wave energy distribution; according to the position where the earthquake damage in the slope body occurs, slop surface displacement and a fracture observation result are combined to conclude a damage mode of a consequent rock slope containing a weak intercalated layer. The method fundamentally reveals a failure mechanism ofthe slope under the action of earthquake, and has broad application prospects in the field of civil engineering and geological disaster prevention.

The invention discloses a power line two-way power-frequency communication uplink signal detecting method. The method is characterized by including the steps of firstly, making background power-frequency signals of modulating signals cancelled out and enhancing the modulating signals through a weighted summation method in the two-way power-frequency communication uplink signal detecting process, conducting derivation on the enhanced modulating signals, and conducting empirical mode decomposition on the modulating signals through ensemble empirical mode decomposition; secondly, finding the order, where the frequency of 150 HZ to 500 HZ is located, in an n-order intrinsic mode function through correlation operation, finding the threshold value of a corresponding order through the intrinsic mode function of the order where the frequency of 150 HZ to 500 HZ is located, conducting threshold value denoising, and calculating the Hilbert marginal spectrum of the denoised signals through Hilbert conversion, and therefore the time domain where the modulating signals are located is detected, most noise and harmonic interference can be effectively removed, and the time domain where the modulating signals are located can be accurately detected. The method has the advantages of being scientific, reasonable, easy to implement, high in detection accuracy and the like.

The invention relates to an underwater signal time-frequency endpoint parameter estimation method based on constant false alarm detection. On the basis of carrying out framing and band division processing on a signal, time-frequency transformation is carried out on the signal, on the basis, the constant false alarm detection is carried out on the signal, frequency marginal spectrum and time marginal spectrum of the detection result are calculated and judgment is carried out, so frequency and time endpoint estimation results of the signal can be obtained. According to the method, time and frequency endpoints of the signal can be estimated at the same time. Compared with a traditional method, the method has the advantages that higher detection probability is achieved. Information difficult to discover in a sonogram can be extracted, so corresponding target information is obtained. Prior information is not required in a detection process. Large scale data training is avoided. The method is applicable to an endpoint parameter estimation system under an known condition.

The invention provides an underwater maneuvering small target recognition method based on the HHT and an artificial neural network. The method includes the steps of extracting a HHT feature of a signal x(t) to be recognized: a Hilbert marginal spectrum, and extracting a HHT feature quantity from the Hilbert marginal spectrum; subjecting the signal x(t) to be recognized to Fourier transform, and extracting a peak frequency and 3dB bandwidth thereof; mixing the HHT feature quantity, the peak frequency and the 3dB bandwidth to construct a HHT mixed feature vector; and finally, inputting the HHT mixed feature vector into a trained artificial neural network classifier for recognition, and outputting the recognized type. The method of the invention makes full use of the adaptability of the HHT to signals, the advantages of processing non-stationary signals and the characteristics of the time-frequency domain of underwater small target radiated noise signals, and the extracted high-dimensional feature quantity can fully describe the features of an underwater target; and the recognition rate of the underwater maneuvering small target is improved.