42 results about "Noise power spectrum" patented technology

The invention relates to the technical field of noise filtering, in particular to a communication line noise filtering system and method, and the system comprises a temperature coupling module, a voltage response module, a parameter joint debugging module and a link checking module. According to the method, inductance and capacitance node temperature data are collected and input into a piecewise linear regression model, the temperature-to-resonant frequency deviation gradient is calculated, a noise suppression bandwidth threshold value is dynamically matched in combination with a reference table, the frequency stability under temperature fluctuation is improved, Kalman filtering is used for processing voltage sampling, and a fluctuation predicted value is generated. According to the method, interval indexes are extracted based on a parameter mapping table, instant inductance and capacitance parameters are generated through least square fitting, voltage transient fluctuation self-adaption is achieved, noise power spectrum density is calculated through fast Fourier transform, a 3dB attenuation point is compared with a threshold value difference value to trigger parameter rollback, closed-loop control is formed, it is ensured that suppression bandwidth is continuously matched with a noise frequency band, and noise suppression is achieved. The bandwidth offset is reduced to be within 1.2%, and the optimization response time is shortened to 200 ms level.

The invention discloses a spatial positioning system and method for a partial discharge source of a transformer, and the system can accurately calculate the three-dimensional coordinates of the partial discharge source through the spatial layout and phase difference measurement of multiple sensors in combination with a geometric calculation method. Aiming at complex electromagnetic interference in the operation environment of the transformer, a dynamic phase difference compensation algorithm is adopted, and the electromagnetic interference is adaptively suppressed by analyzing the phase difference of signals received by each sensor, so that accurate extraction of partial discharge signals is ensured; the compensation algorithm can be automatically adjusted according to the change of the interference intensity, and the robustness of the system is effectively improved; according to the invention, through the adaptive weighting function, the partial discharge signal and the environmental noise signal can be effectively distinguished, and accurate processing of the signal is ensured; particularly, through dynamic adjustment of the exponential attenuation function and the noise power spectrum density, the influence of interference signals can be further weakened, and the signal processing precision is improved.

The invention relates to a pipeline leakage detection method based on acoustic signal spectrum noise reduction, and belongs to the technical field of pipeline leakage monitoring. The method comprises the steps that pure background noise signals are collected through an acoustic sensor to serve as noise analysis samples, noise power spectrum estimation is conducted, and average noise power spectrum estimation is obtained; collecting an original acoustic signal containing potential leakage; performing spectrum analysis and phase reservation on the original acoustic signal after framing processing; a continuous noise reduction acoustic signal is output through spectral subtraction operation and signal reconstruction; performing feature extraction on the continuous noise reduction acoustic signal, performing anomaly detection on a current signal segment through a pipeline leakage detection model, and outputting a leakage signal segment; and leakage point positioning is carried out according to the time difference of the leakage signal segment reaching different sensors. Efficient and accurate detection of pipeline leakage is realized, and interference of background noise on a detection result is effectively overcome.

The invention discloses a voice interactive chart dynamic generation method and system based on a large language model, and relates to the technical field of chart generation, and the method comprises the steps: collecting a user voice signal, carrying out the preliminary framing, calculating the short-time energy, and analyzing a mute segment set signal; frequency domain signals are extracted from the set signals, gain is calculated through a filter, noise reduction is carried out on the frequency domain signals, the zero-crossing rate and noise reduction short-time energy are calculated, effective voice frames are screened, peak detection is carried out, the average voice speed is calculated, voice speed normalization is carried out, and signal values of sampling points in the frames are extracted. According to the method, the noise power spectrum density is extracted through FFT on the mute section, a frequency domain model of background noise is effectively established, a follow-up filter can accurately act on an actually existing frequency band interference area, weakening of the voice main signal spectrum is avoided, the Mel filter bank is accessed after the power spectrum is calculated through the FFT after frame windowing, and the noise power spectrum density is calculated through the FFT after frame windowing. Energy can be redistributed on the logarithmic Mel scale according to human ear perception characteristics.

The invention discloses a relay station for digitally locking an optical fiber link and monitoring phase noise in real time, which belongs to the technical field of relay stations for optical fiber optical frequency transmission and comprises a main control circuit, a link locking circuit and a laser locking circuit. The link locking circuit is bidirectionally connected with the main control circuit, the link locking circuit sends a link locking error signal to the main control circuit, and the main control circuit sends a link locking instruction to the link locking circuit; the laser locking circuit is bidirectionally connected with the main control circuit, the laser locking circuit sends a laser locking error signal to the main control circuit, and the main control circuit sends a laser locking instruction to the laser locking circuit; the link locking circuit comprises an optical frequency digital phase-locked loop; and the optical frequency digital phase-locked loop adopts a DDS (Direct Digital Synthesizer) to drive an acousto-optic modulator to adjust the frequency and the phase of an optical frequency signal. An optical frequency digital phase-locked loop is adopted to measure a phase noise power spectrum in real time, link locking can be achieved, and meanwhile whether the power spectrum of a locked signal reaches a theoretical noise suppression limit or not can be judged.

The invention discloses an engineering scale bump aerodynamic noise prediction method based on data fusion, and belongs to the technical field of spacecraft aerodynamic design. The method comprises the following steps: firstly, acquiring wind tunnel test data, simulation data, corresponding test characteristic data and simulation characteristic data of each scaling model of a bulge; secondly, constructing a recurrent neural network, and forming a model for predicting test power spectral density data based on simulation power spectral density data by taking a mapping relation of active subspace vectors to simulation conditions as mapping from a neural network input layer to a first layer of an intermediate layer; and acquiring simulation data of the engineering scale bump, inputting the simulation data into the model, and predicting flight test noise power spectral density data of the engineering scale bump. Through the application of the method, the physical constraint and interpretability of the neural network are improved, the model confidence is improved, the aerodynamic noise of the projection in the engineering scale can be predicted more accurately, and the aircraft performance can be improved.

The present application relates to the technical field of noise processing, and discloses a noise whitening method, device, equipment and storage medium, the method comprises the following steps: constructing an error function based on an equalization signal and a training sequence; calculating based on the error function to obtain gradient information; updating the tap coefficient of the noise whitening filter based on the gradient information to obtain a target tap coefficient; whitening the colored noise based on the target tap coefficient. The present application updates the tap coefficient of the noise whitening filter by formulating the error function and calculating the gradient, solves the problem that when the channel equalization technology is used to overcome the bandwidth limitation, the equalization enhances the colored noise, leads to uneven noise power spectrum, and thus reduces the IM-DD transmission rate, realizes the whitening of the colored noise, makes the power spectrum flat, and thus significantly improves the performance of the IM-DD system and increases the communication capacity.

The present invention discloses a speech enhancement method based on nonlinear residual noise estimation, including: obtaining a complex spectrum of a noisy speech stream; calculating the noise power spectrum of the complex spectrum based on nonlinear residual noise estimation; calculating the posterior signal-to-noise ratio and the prior signal-to-noise ratio, and updating the gain when speech exists according to the posterior signal-to-noise ratio and the prior signal-to-noise ratio; estimating the gain when speech does not exist and the probability of speech existence; calculating the overall gain according to the gain when speech exists, the gain when speech does not exist and the probability of speech existence, and then calculating the spectral amplitude of the denoised speech signal; synthesizing the time-domain denoised speech signal of the corresponding frame based on the spectral amplitude of the denoised speech signal and the phase of the complex spectrum of the noisy speech of the same frame, and then obtaining the time-domain denoised speech signal of the noisy speech stream. The method of the present invention not only improves the estimated noise level, but also does not cause excessive noise reduction, and finally achieves the improvement of the noise reduction effect.

The present invention relates to the field of noise filtering technology, specifically a communication line noise filtering system and method, the system comprising: a temperature coupling module, a voltage response module, a parameter joint adjustment module, and a link verification module. In the present invention, by collecting the temperature data of the inductor and capacitor nodes and inputting them into a piecewise linear regression model, the temperature deviation gradient of the resonant frequency is calculated, the noise suppression bandwidth threshold is dynamically matched in combination with the reference table, the frequency stability under temperature fluctuations is improved, the Kalman filter processes the voltage sampling to generate a fluctuation prediction value, the interval index is extracted based on the parameter mapping table, the least squares method is used to fit and generate the real-time inductor and capacitor parameters, and the voltage transient fluctuation is adaptively realized. The noise power spectrum density is calculated by fast Fourier transform, and the parameter rollback is triggered by comparing the 3dB attenuation point with the threshold difference, thereby forming a closed-loop control to ensure that the suppression bandwidth is continuously matched with the noise frequency band, reduce the bandwidth offset to within 1.2%, and optimize the response time to shorten it to 200ms.

The present invention provides a composite three-axis Hall sensor based on an acoustic wave excitation magnetic antenna, comprising a Hall probe and an acoustic wave excitation magnetoelectric antenna; the acoustic wave excitation antenna is applied to a three-axis Hall sensor to form a composite three-axis Hall sensor, the frequency band of a low-frequency signal to be measured is transferred to a higher frequency band through the acoustic wave excitation antenna, and the magnetic field signal to be measured in the output signal of the Hall probe of the three-axis Hall sensor is restored through a demodulation circuit; the noise power spectral density of the Hall probe in the higher frequency band is low, and the sensor noise is reduced by frequency transfer. The present invention transfers the frequency of the signal to be measured to a higher frequency by modulation and demodulation, and the noise power spectral density at the high frequency is low, so the noise sensitivity of the probe can be improved.

The embodiment of the invention provides a noise estimation method and device, electronic equipment and a storage medium. The noise estimation method comprises the following steps: carrying out time delay alignment on a first original voice signal and a second original voice signal to obtain a first target voice signal and a second target voice signal; determining a main path signal and an auxiliary path signal from the first target voice signal and the second target voice signal; and calculating an energy standard ratio and a dual-channel weighted feature of the main path signal and the auxiliary path signal, and performing noise estimation on the main path signal based on the energy standard ratio and the dual-channel weighted feature to obtain a noise power spectrum of the main path signal. According to the embodiment of the invention, when noise estimation is carried out on the main path signal, the energy standard ratio of the dual-channel voice signals and the dual-channel weighted features are added, so that the correlation information between the dual-channel voice signals is combined during noise estimation, the accuracy of noise estimation can be improved, and the method is more suitable for dual-channel noise estimation.

The application discloses a laser phase noise measurement method based on delay self-heterodyne and digital demodulation, and relates to the technical field of laser measurement, and comprises the following contents: after being adjusted by a variable optical attenuator, to-be-measured laser is divided into a frequency shift arm and a delay arm, the frequency shift arm is frequency-shifted by an acousto-optic modulator, the delay arm is delayed by a delay optical fiber, and original electric beat frequency signals are obtained by a balanced photoelectric detector after the two arms are coupled; homologous and synchronous quadrature demodulation is carried out on the original electric beat frequency signals, envelope credibility judgment and abnormal sample repair are combined, and a continuous differential phase sequence is obtained; actual optical time delay is determined according to the beat frequency spectrum characteristics and the continuous differential phase sequence, and a differential phase power spectral density is formed; then, transmission relationship conversion is carried out according to the actual optical time delay, bounded spectrum compensation is implemented on a zero point neighborhood and a low credibility frequency point, and an intrinsic phase noise power spectral density is obtained. The application can improve the stability, continuity and result credibility of phase noise measurement under the condition of short delay.

The embodiment of the invention discloses a noise measurement method for an acceleration sensor, and the method comprises the steps: obtaining the reference self-noise power spectrum density of a reference acceleration sensor, collecting the current acceleration data of a to-be-measured acceleration sensor, and the current acceleration data of the reference acceleration sensor, and finally, determining the current self-noise power spectral density of the acceleration sensor to be tested according to the current acceleration data of the acceleration sensor to be tested, the current acceleration data of the reference acceleration sensor and the reference self-noise power spectral density. The current self-noise power spectral density of the acceleration sensor to be measured is determined based on the reference self-noise power spectral density according to the current acceleration data of the acceleration sensor to be measured and the current acceleration data of the reference acceleration sensor, so that the noise of the acceleration sensor to be measured can be measured flexibly, efficiently and accurately. The noise of the acceleration sensor to be measured can be measured flexibly, efficiently and accurately.

This invention belongs to the field of noise floor reduction and proposes a frequency-band noise floor reduction method based on a spectrum analyzer. This invention can solve the problems of uneven noise power distribution and excessive noise power estimation. It includes the following steps: acquiring the noise floor signal of the spectrum analyzer, wherein the noise floor signal is a low-pass or band-limited white noise signal with a specified bandwidth; dividing the specified bandwidth into a first number of sub-segments and dividing the signal power before denoising into a first number of sub-segments; estimating the noise power under the specified bandwidth and dividing it into a first number of sub-segments; denoising each sub-segment and calculating the noise power spectral density of each sub-segment after denoising based on the bandwidth after segmentation, the signal power before denoising after segmentation, and the estimated noise power after segmentation; and ensuring that the noise power spectral density after denoising is not negative by using the compression factor method or the square root method.

The present invention discloses a relay station for digitally locking and real-time monitoring phase noise of an optical fiber link, belonging to the technical field of relay stations for optical fiber optical frequency transfer, which includes a main control circuit, a link locking circuit and a laser locking circuit; the link locking circuit is bidirectionally connected to the main control circuit, the link locking circuit sends a link locking error signal to the main control circuit, and the main control circuit sends a link locking instruction to the link locking circuit; the laser locking circuit is bidirectionally connected to the main control circuit, the laser locking circuit sends a laser locking error signal to the main control circuit, and the main control circuit sends a laser locking instruction to the laser locking circuit; the link locking circuit includes an optical frequency digital phase-locked loop, and the optical frequency digital phase-locked loop uses a DDS to drive an acousto-optic modulator to adjust the frequency and phase of an optical frequency signal. The optical frequency digital phase-locked loop is used to measure the phase noise power spectrum in real time, which can not only achieve link locking, but also determine whether the power spectrum of the signal after locking reaches the theoretical noise suppression limit.

A multi-stage noise suppression system for reducing noise components in a noisy input signal has a first stage neural network that estimates a noise power spectrum for the noisy input signal. A first set of gain values corresponding to the noise power spectrum is generated by the first stage neural network. A second stage neural network estimates clean signal power spectrum values, which are derived from an application of a second set of gain values generated as a function of the clean signal power spectrum values and a first stage reduced noise signal power spectrum values.

The application provides a real-time sensing method for third-party construction damage events, and relates to the field of long-term monitoring of urban lifelines, which comprises the following steps: determining a cutoff frequency according to construction equipment information and an estimated noise power spectral density, and preprocessing an original vibration signal based on the cutoff frequency; calculating the product of the RMS values of P adjacent measuring points; combining D-S evidence theory to fuse the evidence of adjacent measuring points, eliminating abnormal interference signals by calculating a conflict coefficient K, and obtaining a processed signal; obtaining a dynamic threshold value, and determining the endpoints of the third-party construction damage event in the processed signal; and using a K nearest neighbor algorithm to perform feature matching and classification recognition on a to-be-identified sample, and identifying the event type of the third-party construction damage event. The application effectively improves the recognition accuracy and real-time performance, effectively eliminates abnormal data interference, enhances the robustness of feature extraction and classification, and significantly improves the intelligent level of the monitoring system.

The application discloses a ship radiated noise power spectrum measurement method based on frequency difference beam forming, comprising the following steps: obtaining a to-be-measured radiated noise signal received by a base array; constructing a difference frequency signal through the conjugate product of the to-be-measured radiated noise signal and a reference frequency signal; performing beam forming on the constructed difference frequency signal based on the difference frequency, and calculating the amplitude of the beam output signal of the difference frequency signal; correcting the amplitude of the beam output signal of the difference frequency signal by using the amplitude of the reference frequency signal to obtain the amplitude of the to-be-measured radiated noise signal, and calculating the power spectrum of the to-be-measured radiated noise signal. The application converts the signal outside the design frequency band of the base array to the inside of the design frequency band to perform beam forming, improves the array gain, better suppresses the environmental noise, improves the measurement performance of the base array on the radiated noise outside the design frequency band, greatly expands the frequency band width of the ship radiated noise measurement, and can realize the measurement of the radiated noise with a larger bandwidth by using a smaller aperture array.

The invention discloses a method for diagnosing a fault sound signal of a rotating machine, and the method comprises the following specific steps: collecting a sound signal of a waterwheel room, and sequentially carrying out the framing and windowing, and FFT processing, thereby obtaining the amplitude and phase of the signal; the method comprises the following steps: calculating a noise power spectrum estimated value, a signal-to-noise ratio, an over-reduction factor and a denoised signal power spectrum of a waterwheel room sound signal after FFT of a current frame, recovering the signal through IFFT, carrying out pre-emphasis processing on the recovered signal to obtain a denoised clean signal, decomposing the denoised clean signal into IMF components through a CEEMDAN method, calculating a residual component, making a data set, and obtaining the waterwheel room sound signal. And inputting the data set into the trained CNN-TCN neural network model for fault identification. The method is high in fault diagnosis accuracy.

The invention discloses a fish culture net cage noise tracking method based on optimal smoothness and minimum statistics, and relates to the field of aquaculture. According to the method, the smoothing parameter is automatically adjusted according to the frequency energy distribution characteristic of the input signal, the power spectrum is smoothed, the smoothing result is estimated by combining the minimum value deviation compensation, accurate tracking of the background noise signal is realized, the noise power spectrum component can be effectively extracted in a complex noise environment, and extra detection does not need to be carried out on a target signal. According to the method, a time-frequency adaptive smoothing parameter mechanism is introduced, the background noise power spectrum is tracked and estimated by fusing the minimum statistic, and the noise condition of the sampling signal can be effectively determined under the complex noise condition without training data; the problems that an existing noise separation method is poor in performance and needs a large amount of training data when separating non-stationary noise are solved.

The invention relates to the technical field of signal processing, in particular to a signal-to-noise ratio estimation algorithm based on an auto-encoder, which comprises the following steps of: firstly, selecting four groups of parameter signals with different code element rates and roll-off factors, each group containing 16 mainstream modulation type sub-signals, segmenting the sub-signals, adding Gaussian white noise to construct a data set, and performing data processing on the data set; dividing into a training set and a verification set according to the ratio of 7: 3, and calculating a noise adding signal and a power spectrum of corresponding noise by adopting a Welch method; secondly, constructing a stacked full-connection auto-encoder, taking a training set noise adding signal power spectrum as input, taking a noise power spectrum as a target, and introducing L2 regularization suppression over-fitting according to set batchsize and epoch training; and finally, inputting a test set noise adding signal into the trained model to obtain the noise average power, and calculating the signal average power and the signal-to-noise ratio by combining the noise adding signal power spectrum. The method can cover multiple signal parameters and modulation scenes, the low signal-to-noise ratio interval precision is better, the process is clear, engineering landing is easy, and the method is suitable for a multi-standard communication system.

The ultrahigh frequency photoelectric detector frequency response calibration device has the advantages of being high in calibration precision, easy and convenient to operate and wide in coverage frequency range, high-resolution scanning is achieved through constant-power control and high-purity beat frequency signals in the double-laser beat frequency method, and the ultrahigh frequency photoelectric detector frequency response calibration device can be expanded to the hundred GHz level without a high-speed modulation device. The system structure is simplified, extra calibration requirements are avoided, the ASE frequency sweep method eliminates dependence of light source frequency response calibration by using a broadband noise source and a filter amplification mechanism, the dynamic range and the signal-to-noise ratio are remarkably improved through noise power spectral density measurement, power fluctuation and system noise can be effectively suppressed by the two methods, and the system performance is improved. And the accuracy and reliability of calibration in an ultrahigh frequency scene are ensured.

The present invention provides a method, apparatus, medium, and device for speech keyword recognition, comprising: obtaining a frequency domain amplitude spectrum of a current frame signal; performing noise estimation to obtain a noise power spectrum of the current frame signal; estimating an initial a posteriori signal-to-noise ratio (SNR) of the current frame signal based on the noise power spectrum of the current frame signal and the noise power spectrum of the previous frame signal; estimating an initial a priori SNR of the current frame signal based on the initial a posteriori SNR of the current frame signal and the posterior SNR of the previous frame signal; calculating a probability of the presence of a target speech in the current frame signal based on the initial a posteriori SNR and the initial a priori SNR; and determining whether speech is present in the current frame signal based on the target speech probability. If the determination result is that speech is present in the current frame signal, inputting the current frame signal into a preset decoding graph for recognition to obtain a recognition result. This method is used to improve the problem of high speech keyword misrecognition rate in noisy environments.

The present invention proposes a federated learning algorithm WD&DFD‑FL based on Wiener deconvolution and Diag‑Fisher pruning. First, after completing model training, the client uses the Diag‑Fisher pruning mechanism to prune the gradient through the Fisher information matrix. Then, the natural gradient optimization is applied to correct the pruned gradient and upload it to the server with homomorphic encryption. After receiving the encrypted pruned gradient, the server uses a federated mechanism based on Wiener deconvolution to filter the noise, estimates the noise power spectral density, and designs deconvolution to remove the noise introduced in the encryption process and enhance the useful signal. Finally, the server aggregates the gradients uploaded by each client, forms a global model and feeds it back to the client, completing the training iteration of federated learning; the present invention reduces communication overhead by pruning gradients with small weights and improves model accuracy by using natural gradients and Wiener deconvolution.

The present invention relates to the field of signal detection technology, and more particularly to a noise floor estimation and threshold calculation method for a low signal-to-noise ratio (SNR) spectrally dense signal. The method determines the noise power spectrum density by statistically analyzing a histogram of the minimum to mean interval and finding the histogram maximum value. This method avoids the problem of misdetecting the power spectrum of a broadband signal as the noise floor when the signal width accounts for a large proportion of the entire spectrum width, addresses the shortcoming that most noise floor estimation algorithms cannot detect dense signals on the spectrum, and improves the accuracy of noise floor estimation.

The present invention relates to a method for characterizing traps in the gate oxide layer of MOSFETs, and the steps are as follows: 1) Measure the transfer characteristics to extract the threshold voltage, and test the noise power spectral density S of the device in the linear region id ; 2) Draw a double y-axis graph according to the measurement data to judge the noise fluctuation mechanism; 3) Calculate the trap energy position distribution in the gate oxide layer at a certain gate voltage V g ; 4) Repeat step 3 for all gate voltages to draw the energy position distribution diagram of the oxide layer traps. This method breaks through the limitation of only using uniform distribution or exponential distribution to characterize the trap distribution with position, and at the same time takes into account the non-uniformity of trap position and energy distribution, and by drawing a three-dimensional image, the distribution of oxide layer traps with energy position can be clearly observed.

The present invention discloses a speech enhancement method suitable for transient noise suppression, comprising the following steps: (1) performing a short-time Fourier transform on speech data in a clean speech data set to obtain a power spectrum; (2) using the power spectrum to train a VAE model weight; (3) performing a short-time Fourier transform on the noisy signal to be enhanced to obtain a time-frequency spectrum and a power spectrum; (4) performing a NMF on the power spectrum to calculate the time-domain index and frequency-domain index of each mode obtained by decomposition, estimating the components belonging to transient noise therein and filtering them out; (5) inputting the power spectrum after filtering out transient noise into the trained VAE model, combining the NMF with the estimated speech power spectrum and noise power spectrum; (6) subjecting the estimated speech power spectrum and noise power spectrum as well as the time-frequency spectrum of the noisy signal to Wiener filtering to obtain the time-frequency spectrum of the enhanced speech. The present invention can effectively suppress speech noise in a scenario where transient noise exists, while relatively completely preserving the speech components.

A system and method of harmonizing computed tomography (CT) images to reduce variability introduced by differing reconstruction kernels. The method includes acquiring a CT image of a plurality of regions of a patient. The method includes calculating, by a processing device, a noise power spectrum (NPS) of a uniform region among the plurality of regions. The method includes selecting, from a library of reconstruction kernels, a first reconstruction kernel based on the NPS of the uniform region of the plurality of regions. The method includes generating a harmonized CT image based on the first reconstruction kernel and the CT image.

The invention discloses a GaN device life prediction method and device based on a noise power spectrum, a medium and a product, and relates to the technical field of semiconductors, and the method comprises the steps: obtaining a grid current noise power spectrum of a to-be-predicted GaN device; the horizontal axis of the grid current noise power spectrum is noise frequency, and the longitudinal axis is noise spectrum density; based on the grid current noise power spectrum of the to-be-predicted GaN device and the failure time prediction model, obtaining the failure time of the to-be-predicted GaN device; the failure time prediction model is a function about noise spectral density and failure time; and determining the failure time of the to-be-predicted GaN device as the service life of the to-be-predicted GaN device. According to the invention, the service life prediction of the GaN device is realized.