37 results about "Modulation spectrum" patented technology

Techniques for speech recognition models using modulation spectrum are disclosed herein. A modulation spectrum is generated from time series data output of an encoder layer of a speech recognition model and used as input into a decoder layer of the speech recognition model to improve accuracy of the model such as for recognizing subword units. The modulation spectrum is determined by applying a convolution filter to the output of the encoder layer of the speech recognition model. The time series data and / or the modulation spectrum can be normalized. A rectified linear unit activation function can be applied to the output of the convolution filter. The output of the encoder layer may be residually connected to the output of the rectified linear unit activation function prior to being input into the decoder layer.

The invention discloses an automobile maintenance engine fault dynamic diagnosis system and method. The system comprises a data acquisition module, a data preprocessing module, a dynamic monitoring and recording module, a sound signal extraction algorithm module, a wavelet transform algorithm module, a neural network module and a fault diagnosis and prediction module. The method comprises the following steps: step 1, data acquisition; step 2, data preprocessing; step 3, real-time monitoring and recording; step 4, sound signal extraction and analysis; step 5, fault feature extraction; step 6, constructing and training a neural network; step 7, fault diagnosis and prediction; the dynamic monitoring function is designed, the vibration frequency, surface abnormity and working condition parameters of the automobile engine can be comprehensively monitored, real-time alarming is carried out on abnormal conditions, potential risks existing in the engine can be found in time, powerful support is provided for fault prevention and processing, and the working efficiency is improved. And a fast cyclic modulation spectrum peak graph algorithm is designed through a multi-stage decomposition method.

The present invention belongs to the technical field of power grid dispatching, and provides a fault detection and system for a power grid multimedia dispatching system, comprising the following steps: acquiring audio data information of a device and constructing an audio database; extracting modulation spectrum features of the audio database; performing clustering analysis and calculation of the extracted modulation spectrum features of the audio database based on a clustering algorithm to obtain a standard deviation variable; and judging the operating status of the device according to the audio to be tested of the device and the obtained standard deviation variable.

The invention provides a stimulated Brillouin scattering suppression device and method of an optical fiber amplification system, and the device comprises a seed source, a signal generator, a radio frequency amplifier, a signal modulator and a spectrum filter. And performing power amplification on the LFM radio-frequency signal output by the signal generator by using a radio-frequency amplifier so as to drive the signal modulator to modulate the seed laser and generate a modulation spectrum containing a carrier and a plurality of sidebands, and finally screening out a single target sideband through a spectrum filter to obtain a sweep-frequency light source. Effective suppression of stimulated Brillouin scattering is realized in an optical fiber amplification system. According to the invention, high modulation voltage can be prevented from being applied, and the voltage-withstanding range of an existing commercial signal modulator is adapted; meanwhile, the modulation period range is large and controllable, the short modulation period can meet the short optical fiber scene requirement of the optical fiber amplification system, and the SBS suppression effect of the optical fiber amplification system is effectively improved.

This invention relates to the field of passive underwater target identification technology, and in particular to a propeller noise cyclic coherent demodulation method based on an improved real-number genetic algorithm. Based on cyclic stationarity analysis theory, an optimal coherent weighted quadratic envelope spectrum is constructed by replacing the fixed integration frequency band with a weighting function, and an adaptive detection threshold is constructed based on the false alarm probability. The observed spectrum is modeled as the sum of harmonics and noise, and a sparse non-negative optimization problem is solved to extract the harmonics. Using the total harmonic intensity of the envelope spectrum as the cost function, an improved real-number genetic algorithm is used to optimize and search for the optimal weighting coefficient vector. The demodulated spectrum is obtained by coherently weighting and integrating the normalized cyclic modulation spectrum. This invention achieves autonomous allocation of weighting coefficients and adaptive selection of demodulation frequency bands, effectively improving the signal-to-noise ratio gain of cyclic demodulation, and is applicable to passive target identification in data backtracking and review systems.

The application discloses a synchronous measurement and analysis method and system for torsional vibration of a steam turbine shaft system and blade vibration, and belongs to the technical field of monitoring of steam turbine generator units. The method comprises noise suppression processing on multi-mode synchronous signals and correlation analysis, so as to obtain coupling modal characteristics representing the comprehensive state of the steam turbine shaft system-blade, and to evaluate and analyze the coupling modal characteristics based on a pre-constructed coupling analysis model. In the application, an adaptive filter is used to filter the blade vibration signal, and cross modulation spectrum of the torsional vibration demodulation timing information and the vibration modal characteristic component is determined based on a mutual information function, so as to accurately determine the cross phase relationship at the sampling point, obtain the coupling modal characteristics representing the comprehensive state of the steam turbine shaft system-blade, and more comprehensively and accurately reflect the overall operation state of the steam turbine.

The application provides a modulation spectrum circuit and a power amplifier, wherein the modulation spectrum circuit comprises an operational amplifier, a field effect transistor, a first resistor, a first control circuit, a second control circuit, a second resistor, a third control circuit, a third resistor and a comparison circuit; a first end of the first resistor and a drain of the field effect transistor are connected and serve as an output end of the modulation spectrum circuit, and are used for outputting a modulation signal with a preset modulation spectrum to a final stage amplifier in the power amplifier; an input end of the second control circuit is used for connecting a feedback signal formed by an output current of the final stage amplifier; and the comparison circuit is used for controlling the conduction or the turn-off of the first control circuit, the second control circuit and the third control circuit. The modulation spectrum circuit of the application can not only meet the modulation spectrum under the condition of large power output, but also optimize the modulation spectrum under the condition of small power output.

The modulation spectrum adaptive feature purification method based on feature energy entropy reconstruction belongs to the field of underwater acoustic target recognition. The method is used for solving the problem that the existing modulation spectrum feature extraction method cannot solve the influence of harmonic interference on the final estimation result. The method comprises the following processes: in step one, the cavitation noise in the ship radiated noise is demodulated through a plurality of preset demodulation frequency bands, a plurality of demoded signals are obtained, and the optimal demodulation frequency band is selected from the demodulation frequency bands corresponding to the demoded signals by adopting the minimum entropy criterion; in step two, the demoded signal obtained under the optimal demodulation frequency band is decomposed through the VMD algorithm, and the signals of the intrinsic mode functions of all orders are obtained; in step three, the feature energy entropy of the signals of the intrinsic mode functions of all orders is calculated, the feature energy entropy is reconstructed by using the FEER algorithm, and the reconstructed signal is obtained; in step four, the power spectrum analysis is carried out on the reconstructed signal, and the purified modulation spectrum is obtained. The method is mainly used for purifying the modulation spectrum.

A method of utilizing beamforming to selectively locate one or more ambient electromagnetic power harvesting (AEPH) chips, implemented by a reader apparatus comprising multiple antennas, the method comprises transmitting, to the AEPH chips, an interrogation signal from multiple transmit antennas by a reader application executing on a processor of the reader apparatus in a first frequency band to activate the AEPH chips; receiving, from at least one AEPH chip by multiple receive antennas of the reader apparatus, one or more signals in response to the interrogation signal; processing the one or more signals to determine a direction of the at least one AEPH chip; and adjusting, using a beamformer, a beam direction of the multiple transmit antennas of the reader apparatus in the direction of the at least one AEPH chip to deliver the maximum power to the at least one AEPH chip.

The invention discloses a communication signal carrier frequency estimation method based on a cyclic modulation spectrum, and the method comprises the steps: carrying out the discretization of a communication signal received by a receiving end, and obtaining a discrete real signal; windowing the discrete real signal, and calculating a modulo 2 norm of a window function; performing short-time Fourier transform on the discrete real signal after windowing, and performing module value square calculation on the signal after the short-time Fourier transform; performing discrete Fourier transform by using the signal after module value square calculation to obtain a cyclic modulation spectrum of a discrete real signal; calculating a cyclic frequency corresponding to the maximum amplitude value of the cyclic modulation spectrum on a positive frequency axis; substituting the cyclic frequency corresponding to the maximum amplitude value into the cyclic modulation spectrum to obtain a function about discrete frequency; and calculating the discrete frequency corresponding to the maximum module value of the function related to the discrete frequency, and taking the discrete frequency as the carrier frequency of the communication signal. The method has better generalization, and is high in application value in various actual scenes.

The application provides a method, device, equipment and storage medium for extracting a spectrum of spot noise in a along-track area, which comprises the following steps: when the along-track direction is perpendicular to the sea wave propagation direction, determining a boundary spectrum of spot noise of the along-track area and a non-along-track area by using a spectrum of spot noise in the non-along-track area; determining a sea wave modulation spectrum of the along-track area under a corresponding relative observation angle by using the boundary spectrum of spot noise; determining a spectrum of spot noise in the along-track area under the relative observation angle by using the sea wave modulation spectrum of the along-track area; and determining a spectrum of spot noise in a target observation angle of a target along-track area under the same sea surface condition by using the spectra of spot noise in the along-track area under different relative observation angles, wherein the sea surface condition comprises a sea surface wind speed and an effective wave height. The method can improve the accuracy of extracting the spectrum of spot noise in the along-track area.

The invention relates to the technical field of power system equipment protection, in particular to a phase modifier protection method based on fusion of vibration characteristics and non-electrical parameters. Comprising the following steps: synchronously acquiring vibration signals, thermal parameters and operation condition parameters; extracting a vibration feature vector through a wavelet packet energy entropy modulation spectrum, and establishing a thermal parameter health reference model to calculate a residual feature to form a thermal feature vector; generating a dynamic adjacency matrix according to the current working condition by adopting a dynamic weight generation network, and outputting node update features through graph convolution; when the node update feature exceeds the limit, calculating the abnormal confidence coefficient according to the over-limit state of the cross-domain strong association node; diagnosing the fault through cross-domain attention fusion; and establishing an equipment risk, power grid risk and economic cost multi-objective function, and performing Pareto optimization on the candidate protection action set to select an optimal protection action. According to the invention, working condition self-adaptive multi-physics field depth coupling analysis and intelligent protection decision are realized.

The present invention discloses an underwater target recognition method and device. The method comprises: obtaining an underwater target radiation noise data set, comprising N pieces of underwater target radiation noise data information, where N is an integer; extracting characteristic parameters of any piece of underwater target radiation noise data information to obtain characteristic information of the underwater target radiation noise, including radiation noise line spectrum characteristic information, modulation spectrum characteristic information, continuous spectrum characteristic information and auditory spectrum characteristic information; the underwater target radiation noise characteristic information constitutes an underwater target radiation noise characteristic information set; dividing the underwater target radiation noise characteristic information set to obtain a training set and a test set; using the training set to train a preset underwater target recognition model to obtain a training underwater target recognition model; using the training underwater target recognition model to process the test set to obtain an underwater target recognition result.

The present invention relates to a radar target robust identification method based on multi-domain feature fusion under short dwell conditions, comprising: step 1: obtaining a test data set; step 2: generating a test feature matrix according to the test data set; step 3: inputting the test feature matrix into a trained SVM classifier to obtain a target classification result; wherein the test data set includes multiple radar echo signals, and each X radar echo signals constitute a test sample, X≥3; using the mean feature and variance feature of the time domain entropy matrix, the time domain variance matrix, the Doppler domain entropy matrix and the Doppler domain variance matrix of each test sample in the test data set, the test feature matrix is ​​constructed. The method of the present invention overcomes the problem that the radar has a relatively short dwell time when observing a target, and the resolution of the micro-Doppler modulation spectrum of the target is reduced, resulting in a reduced target classification accuracy.

The present invention provides a method and system for extracting parameters of wavelength modulated laser gas absorption spectrum. The system includes a signal generation module, a laser, an interferometer, a fiber splitting module, an optical fiber, a signal acquisition module, a modulation spectrum wavelength recording module and an online harmonic analysis module based on FPGA. According to the time-wavelength relationship and the gas absorption function relationship recorded by the wavelength recording module, a training structure is a neural network model whose input is a normalized high-order harmonic peak point and output is a gas absorption spectrum parameter, and the model parameters are generated. The model structure and parameter values ​​are then transplanted to the FPGA to achieve online solution of harmonic signals and absorption parameters. The feature of the present invention is that it utilizes the determinism of the wavelength modulation model and the online calculation advantage of the FPGA, which not only reduces the complexity of the modulation absorption spectrum signal acquisition, but also broadens the measurement range of the wavelength modulation method for gas parameters such as temperature and concentration, and has broad application prospects.

This invention relates to a magnetic sensing deep learning-driven intelligent metamaterial on-demand design system and method. By setting up a terminal device, a unit cell structure device, a prediction device, and a simulation device, the unit cell structure device can generate a magnetic field modulation spectrum with the same dimension as the target band structure matrix using user-defined target parameters. Furthermore, it uses a magnetic conditional generative adversarial network model algorithm to transform the magnetic field modulation spectrum into multiple candidate unit cell structures. The prediction device can screen out all candidate unit cell structures whose bandgap attribute characteristics deviate from the target bandgap by less than a first threshold. The simulation device can perform finite element simulation on each candidate unit cell structure screened by the prediction device, selecting all candidate unit cell structures that simultaneously meet the bandgap performance threshold and engineering constraints. Ultimately, this achieves rapid and accurate prediction of the band structure of intelligent metamaterials under given excitation field conditions and on-demand design of unit cell structures.

The invention discloses a processing method and device of a measurement and control experiment result and a quantum computer, and by utilizing the scheme of the invention, the analysis efficiency of the experiment result of an energy spectrum experiment can be effectively improved, so that the execution efficiency of the quantum computer is improved. Besides, according to the scheme, the modulation spectrum with high accuracy can be obtained in the quantum chip rough test stage by directly utilizing the result of the energy spectrum experiment, the modulation spectrum can be used as a reference in the subsequent quantum chip fine test stage, and the modulation spectrum obtaining process in the scheme is simple and high in efficiency.

The invention discloses a delay acquisition method and device for a frequency control signal in a quantum chip, and the method comprises the steps: obtaining a first resonant frequency and a second resonant frequency according to the modulation spectrums of a first quantum bit and a second quantum bit; and setting the working frequency of the first quantum bit as a first resonant frequency and the working frequency of the second quantum bit as a second resonant frequency. And then performing a state exchange experiment on the first quantum bit and the second quantum bit, and obtaining resonance duration based on a result of the state exchange experiment. And finally, obtaining the time delay of the frequency control signal among the first quantum bit, the second quantum bit and the adjustable coupler. According to the scheme, the time delay of the frequency control signal in the quantum chip can be directly obtained, and compared with the prior art, the time consumed by the whole scheme is greatly shortened, so that the efficiency of executing a quantum computing task by a quantum computer is improved to a certain extent.

The invention provides a DEMON spectrum-based leaf frequency extraction method under a low signal-to-noise ratio. The method comprises the following steps: firstly, receiving an underwater target radiation noise signal by using a single hydrophone; then the modulation spectrum signal is demodulated through a square low-pass method, and a DEMON spectrum is obtained; secondly, background equalization is carried out on the DEMON spectrum through an alpha bidirectional filtering method, a line spectrum set K is preliminarily obtained, then line spectrum purification and leaf frequency extraction are carried out through different methods according to whether the axis frequency information is obtained or not, and finally the leaf frequency information can be accurately extracted under the low signal-to-noise ratio. Compared with a traditional leaf frequency extraction method, the leaf frequency can be accurately extracted at a lower signal-to-noise ratio, and a complex receiving array or an extraction algorithm is not needed; according to the method, different leaf frequency extraction methods can be selected according to whether the shaft frequency information is acquired or not.

The invention discloses a warning board processing method and device, a storage medium and electronic equipment, the method is applied to the field of metal detection, and the method comprises the following steps: sampling a target area through an eddy current detector to obtain original waveform data; performing interpolation processing on the original waveform data to obtain first waveform data; processing the first waveform data by adopting a complex modulation frequency spectrum refining algorithm to obtain first frequency spectrum data; metal information in the target area is determined according to the first frequency spectrum data, and installation of the warning board is controlled according to the metal information. According to the invention, the problem that the installation efficiency of a warning board is low due to the fact that the number of coils of a traditional eddy current type metal detector is limited and the traditional eddy current type metal detector is easily influenced by noise and interference in a complex environment in the prior art is solved.

This invention provides a modulation spectrum circuit and a power amplifier. The modulation spectrum circuit includes an operational amplifier, a field-effect transistor (FET), a first resistor, a first control circuit, a second control circuit, a second resistor, a third control circuit, a third resistor, and a comparator circuit. The first terminal of the first resistor is connected to the drain of the FET and together serve as the output terminal of the modulation spectrum circuit, used to output a modulation signal with a preset modulation spectrum to the final stage amplifier in the power amplifier. The input terminal of the second control circuit is used to receive a feedback signal formed by the output current of the final stage amplifier in the power amplifier. The comparator circuit is used to control the on / off state of the first, second, and third control circuits. The modulation spectrum circuit of this invention can not only meet the modulation spectrum requirements under high power output conditions but also optimize the modulation spectrum under low power output conditions.

The application discloses a warning sign processing method and device, a storage medium and an electronic device. The method is applied to the metal detection field, and comprises the following steps: sampling a target area by using an eddy current detector to obtain original waveform data; performing interpolation processing on the original waveform data to obtain first waveform data; performing processing on the first waveform data by using a complex modulation frequency spectrum refinement algorithm to obtain first frequency spectrum data; determining metal information in the target area according to the first frequency spectrum data, and controlling installation of a warning sign according to the metal information. Through the application, the problem that the installation efficiency of the warning sign is low due to the fact that the traditional eddy current metal detector is easily affected by noise and interference in a complex environment because of the limited number of coils, and the detection accuracy is reduced is solved.

The invention discloses a local solvability analysis method based on a deep modulation network, and belongs to the technical field of mechanical equipment fault diagnosis. The method comprises the following steps: firstly, constructing a new frequency domain modulation spectrum as a signal processing layer of a deep modulation network so as to demodulate a modulation component from a complex vibration signal; then, fault features of different components are obtained by using a multi-branch structure of different depth modulation network slice scales, and weighting coefficients in the depth modulation network are optimized based on a frequency domain signal-to-noise ratio weighted fusion method so as to enhance fault feature components; and finally, learning fault features from the deep modulation network layer by using a convolutional neural network to identify fault classification of the mechanical equipment. According to the method, local fault feature components can be successfully analyzed, and a brand-new interpretable method is provided for fault diagnosis.

The invention relates to a multi-parameter coupling variable-speed navigation body radiation noise simulation method. According to the method, a continuous spectrum component, a line spectrum component and a modulation spectrum component are respectively generated by initializing a target category and a motion state parameter, and motion parameters such as a target navigational speed and an accelerated speed are innovatively and dynamically coupled with noise characteristics: a continuous spectrum changing along with the navigational speed is generated by using an FIR filter; doppler frequency shift and random fluctuation are introduced into the line spectrum to simulate time-varying characteristics; and dynamically calculating a modulation spectrum according to the propeller rotating speed. And finally, superposing the components to generate a high-fidelity variable-speed target radiation noise signal. The method solves the problem that the traditional model has fixed parameters and cannot reflect the dynamic evolution of maneuvering target noise, and can provide vivid and reproducible data support for sonar system test and target identification research.

The invention provides a compact range-based micro-Doppler characteristic test system and method, and aims to solve the problems of difficulty in micro-Doppler characteristic extraction, remarkable dynamic attitude influence and the like of a rotorcraft in the prior art. The system comprises a microwave anechoic chamber, a compact range reflecting surface, a feed source rotary table, a feed source antenna, a multi-degree-of-freedom low-scattering rotary table, a motion synchronous controller, a broadband vector signal generator, a receiver and an upper computer, and the method comprises the following steps: installing a tested piece on the multi-degree-of-freedom low-scattering rotary table, importing a flight attitude file and simulating the attitude of an aircraft; a broadband carrier signal is transmitted and reaches a measured object through a compact range reflecting surface, an echo signal is collected and conjugate matched filtering processing is carried out, and finally micro-Doppler modulation spectrum characteristic parameters are extracted; according to the invention, environmental electromagnetic interference and machine body strong scattering can be effectively suppressed, the signal-to-noise ratio of signals is improved by more than 20dB, and high-precision micro-Doppler feature extraction is realized.

The invention relates to a ship noise signal simulation method based on a deep neural network. The method comprises the following steps: simulating a continuous spectrum, a modulation spectrum and a line spectrum of ship noise; simulating the continuous spectrum by adopting an FIR filter design based on a Welch method; an improved model based on fusion of bidirectional LSTM and a self-attention mechanism is provided, so that the time-frequency characteristic of a modulation envelope is effectively captured, and high-precision simulation of a modulation spectrum is realized; a physical information neural network (PINN) is innovatively introduced, physical constraints are embedded into a loss function, and high-fidelity simulation of line spectrum features is achieved.

The invention discloses an absorption spectrum parameter solving method and device for extracting a same-phase point of a modulation spectrum, and the method comprises the steps: S1, enabling a laser generation module to drive a laser to emit modulation laser through employing a high-frequency sinusoidal signal and a low-frequency sinusoidal signal which are superposed as a modulation signal generated by laser, enabling the laser to pass through a to-be-detected gas area to carry absorption spectrum information; s2, a laser signal detection module and a high-speed data acquisition module are used for detecting and acquiring laser absorption spectrum signals; and S3, performing in-phase extraction and uploading on the acquired spectral signals by an in-phase extraction and uploading module, sequentially substituting wavelength change information and light intensity information of each grouped phase point into the direct absorption linear model for fitting iteration, and solving parameters of the gas to be detected. By adopting the technical scheme provided by the invention, the anti-noise capability of the modulation method is reserved, and meanwhile, the direct absorption spectrum parameters are quickly solved.

The invention relates to an elastic optical network routing modulation spectrum allocation method for deep Q network optimization, and belongs to the technical field of optical fiber communication. Aiming at the routing, modulation and spectrum resource allocation problems of the time-varying service in the elastic optical network, the invention designs an optimization method combining a deep Q network and a sliding window flow model. The method comprises the following steps of: taking information of the time-varying service and resource state information of the elastic optical network as a state set of the deep Q network; a greedy strategy and Softmax strategy combined action is adopted, or a sliding window flow model is adopted to postpone a delay action of distributing service routing resources, and service routing, modulation formats and spectrum resources are searched; designing Q values of an action reward function and a strategy; a deep Q network model is trained by using an improved priority experience playback mechanism, and neural network parameters of the deep Q network are updated to a target Q network in time, so that the deep Q network learns a more stable strategy, and the designed method can reduce the service blocking rate and improve the spectrum utilization rate.

The invention discloses an intelligent prediction method and system for large deformation of surrounding rock in a deep tunnel soft rock area, and the method comprises the steps: obtaining a displacement vector of each monitoring point in real time through a sensor disposed at each monitoring point in the tunnel surrounding rock area, taking the second derivative of the displacement vector as the dynamic reaction intensity of the monitoring point under construction disturbance, and calculating the dynamic response intensity of the monitoring point; performing Fourier transform on the dynamic reaction intensity to generate a modulation spectrum of surrounding rock disturbance; calculating the strain energy density of the current monitoring point according to the displacement vector of the current monitoring point, calculating the energy divergence of the current monitoring point in combination with the change speed of the displacement vector, and performing spectral analysis on the energy divergence to obtain the spectral energy density of the current monitoring point under each frequency; and carrying out spatial maximum value search on the change rate of the spectrum energy density of all the monitoring points along with time and the change rate of the modulation spectrum along with time, and taking the corresponding monitoring point as a position where unstable deformation is most likely to occur.

A modulation spectrum direction finding method based on automatic demodulation band optimization belongs to the field of underwater acoustic detection and array signal processing technology. To address the problem that existing modulation spectrum direction finding methods often use empirically selected fixed frequency bands for demodulation, which are difficult to adapt to changes in target modulation characteristics, leading to reduced signal-to-noise ratio (SNR) and insufficient direction finding accuracy, this invention first estimates the SNR of the cyclic modulation spectrum, uses a constant false alarm rate (CFAR) detection method for detection, performs spectral line tracking on the detected cyclic modulation spectrum, and optimizes the demodulation band of the modulation spectrum. Using the optimized demodulation band, the original signal is bandpass filtered, and the filtered signal is detected to obtain the corresponding envelope. Cross-spectral analysis is performed to obtain the phase information of all modulation frequencies within the band. The received signal delay difference between sensors is calculated from the phase information, the target's azimuth is calculated, and the azimuths measured from multiple modulation frequencies are weighted and fused to obtain the final target direction finding result.