27 results about "Frequency weighting" patented technology

The invention relates to the technical field of audio signal processing, and discloses a wind-noise-resistant self-adaptive volume adjustment method and system for a riding earphone, and the method comprises the following steps: S1, collecting environment sound and a source audio signal in parallel; s2, analyzing environment sound, and determining macroscopic adjustment intensity; s3, analyzing the source audio in parallel, and identifying a content type and a transient signal of the source audio; s4, based on the psychological acoustic model, generating a frequency-related target method tone quality compensation gain; s5, performing dynamic smoothing and frequency weighting on the target gain according to the content and the transient characteristics to generate an application gain; s6, fusing the macroscopic adjustment intensity and the application gain, and synthesizing a final gain parameter; and S7, applying the final gain to the source audio, and outputting the compensated audio. According to the method, the psychological acoustic model and audio content parallel analysis are fused, frequency-level accurate compensation is realized, the problems of traditional degradation and sudden hearing sense change are solved, and the audio definition and comfort are improved.

The invention relates to a probability-based building vibration comfort evaluation method under a multi-load working condition, and the method comprises the steps: carrying out the FFT of an obtained original acceleration time-history signal of a complete time length of a building under each single-load working condition, and obtaining an original acceleration frequency spectrum; performing frequency weighting on the original acceleration frequency spectrum to obtain a weighted acceleration frequency spectrum; carrying out IFFT on the weighted acceleration frequency spectrum to obtain a weighted acceleration time history signal with complete duration; segmenting the weighted acceleration time history signal into 1-second time segment signals; setting a probability distribution model of the total sampling frequency and the 1-second time segment signal of each single load working condition; randomly sampling from the 1-second time segment signals of each single load working condition, and synthesizing into 1-second time segment signals under a multi-load working condition; and calculating the root mean square of the 1-second time segment signal under the multi-load working condition, and then calculating the Z vibration level for vibration comfort evaluation of the building under the multi-load working condition.

The present application relates to the technical field of underwater acoustic signal processing, and particularly relates to a method for detecting underwater target echo in a strong reverberation background, comprising the following steps: performing band-pass filtering on received sonar echo signals to obtain filtered signals; performing sliding window segmentation on the filtered signals to form a plurality of short-time signal segments; extracting the envelope of the short-time signal segments and performing normalization processing on the envelope; performing differential processing on the normalized envelope to obtain a differential envelope signal; performing frequency spectrum transformation and frequency weighting processing on the differential envelope signal to obtain a weighted envelope spectrum; performing feature extraction on the weighted envelope spectrum, calculating the spectral entropy and spectral variance, and constructing a spectral variance-entropy ratio; and determining whether a target echo exists. The method can reduce the false alarm rate while maintaining a high detection probability, and avoids the performance degradation of existing matched filtering methods in a strong reverberation condition.

The application belongs to the technical field of power system frequency stability analysis, and provides a frequency response analysis method and system based on a unified equivalent model of heterogeneous devices, provides a correction mechanism for the time sequence coupling effect between devices, considers the difference in dominant dynamic characteristics of different devices at different stages of the frequency response process, divides the frequency response process into an inertia response stage, a primary frequency modulation stage and a frequency recovery stage, and respectively adopts a targeted frequency weighting function for parameter identification of each type of device at each stage, so that the parameters of the unified equivalent electromechanical model focus on the dominant dynamic characteristics of the stage in each stage, and under the premise of keeping the structure of the 2-order unified model unchanged, through independent identification and adaptive switching of each stage, the prediction accuracy of key indicators such as the frequency minimum point, the frequency change rate peak value and the power response is obviously improved compared with the fixed parameter method in the whole period.

The application relates to a probability-based building vibration comfort evaluation method under a multi-load working condition, which comprises the following steps: performing FFT on original acceleration time history signals of a building under each single load working condition to obtain original acceleration frequency spectra; performing frequency weighting on the original acceleration frequency spectra to obtain weighted acceleration frequency spectra; performing IFFT on the weighted acceleration frequency spectra to obtain weighted acceleration time history signals of the building with a complete time length; dividing the weighted acceleration time history signals into 1-second time length segmented signals; setting a total sampling number and a probability distribution model of the 1-second time length segmented signals under each single load working condition; randomly sampling the 1-second time length segmented signals under each single load working condition and combining the signals into 1-second time length segmented signals under a multi-load working condition; calculating the root mean square of the 1-second time length segmented signals under the multi-load working condition, and then calculating a Z vibration level, which is used for vibration comfort evaluation of the building under the multi-load working condition.

The invention discloses a rolling bearing fault sound feature extraction method based on anti-noise Hough index guided optimal demodulation. The method comprises the following steps: collecting a sound signal during operation of rotating mechanical equipment through a sound sensor; determining a preset filtering bandwidth and a center frequency sequence; respectively carrying out band-pass filtering on the sound signal according to the filtering bandwidth and the center frequency sequence; calculating an anti-noise Hough index of each filtering signal to obtain a corresponding sequence; outputting the filtering signal corresponding to the maximum anti-noise Hough index as a target filtering signal; performing element-by-element adaptive sparse processing on the target filtering signal; demodulating the signal after the sparse processing by using an enhanced frequency weighted energy operator; and performing square envelope spectrum analysis on the demodulated signal to realize extraction of each fault characteristic frequency of the rolling bearing and recognition of a bearing fault mode. According to the method, the weak fault characteristic frequency of the rolling bearing can be adaptively and accurately extracted based on the non-contact acoustic measurement signal with a lower signal-to-noise ratio.

The application discloses a high-proportion new energy through extra-high voltage transmission system stability evaluation method and relates to the technical field of power system stability evaluation. In view of the problems that the new energy through extra-high voltage transmission system is prone to multi-frequency band resonance due to the lack of rotational inertia and damping of the new energy, the existing method is difficult to adapt to dynamic scenes, only single-frequency band analysis and lacks engineering quantitative indexes, and the scheme points are as follows: an extensible system node impedance matrix is established, multi-frequency band characteristics are extracted by deducing frequency domain response, a composite evaluation index system is formed, comprehensive risk is calculated by frequency weighting integral, and visual results are output. The application is mainly used for new energy station parameter setting, extra-high voltage channel operation optimization and power grid regulation and control, and guarantees the safety of the power grid.

The invention relates to the technical field of magnetic resonance detection, in particular to a double-domain magnetic resonance groundwater detection same-frequency harmonic noise processing method and system, and the method comprises the steps: decomposing an obtained noisy magnetic resonance signal to obtain frequency weight coefficients of the intensity of the noisy magnetic resonance signal at different components; generating a frequency intensity matrix by using the plurality of frequency weight coefficients; adjusting the frequency intensity matrix to obtain a frequency weight matrix, multiplying the frequency weight matrix by the input data set channel by channel, and retaining signal components near the Larmor frequency to obtain a preliminary de-noising signal; establishing a time domain mapping relation between the preliminary de-noising signal and the pure magnetic resonance signal, performing network training to obtain a trained time domain mapping relation, and adjusting a frequency intensity matrix according to a result of each round of training; according to the trained time domain mapping relation, the de-noising signal is constructed, the same-frequency harmonic noise in the magnetic resonance signal can be well eliminated, and an ideal de-noising effect is achieved.

The invention relates to the technical field of underwater acoustic signal processing, in particular to an underwater target echo detection method under a strong reverberation background, which comprises the following steps of: carrying out band-pass filtering processing on a received sonar echo signal to obtain a filtering signal; performing sliding window segmentation processing on the filtering signal to form a plurality of short-time signal segments; extracting an envelope of the short-time signal segment, and carrying out normalization processing on the envelope; performing differential processing on the normalized envelope to obtain a differential envelope signal; performing frequency spectrum transformation and frequency weighting processing on the differential envelope signal to obtain a weighted envelope spectrum; performing feature extraction on the weighted envelope spectrum, calculating a spectrum entropy and a spectrum variance, and constructing a spectrum variance entropy ratio; and judging whether target echoes exist or not. According to the method, the false alarm rate can be reduced, meanwhile, the high detection probability is kept, and the defect that the performance of an existing matched filtering method is seriously reduced under the strong reverberation condition is overcome.

The application discloses a Beidou precise point positioning method applied to a flood environment, and comprises the following steps: S1, collecting original observation values of Beidou satellites in the flood environment; S2, eliminating the influence of pseudo-range multipath error by a multi-path correction algorithm based on a time-space dual-domain strategy; S3, eliminating the influence of cycle slip by an adaptive threshold cycle slip detection and repair algorithm based on a signal-to-noise ratio constraint; S4, weighting the observation values by a random model based on a signal-to-noise ratio weighting, so as to improve positioning precision; S5, eliminating the influence of different Beidou satellite orbit types by a satellite type weighting strategy algorithm; S6, eliminating the influence of signal quality differences of different Beidou satellite frequencies by a frequency weighting strategy; and S7, solving a position by a Beidou precise point positioning algorithm, and saving and outputting a result. The method can effectively overcome problems such as large pseudo-range multipath error, high cycle slip mis-detection rate and decreased positioning precision caused by the flood, reduces the pseudo-range multipath error, and improves the pseudo-range observation value precision.

The invention discloses a noise evaluation method and system for a high-frequency transformer, and relates to the technical field of transformer noise testing, and the method comprises the steps: obtaining the noise data of a plurality of training transformers, carrying out the data preprocessing of the noise data of each training transformer, and obtaining a corresponding training noise set; inputting the training noise set into a preset initial noise evaluation model for training to obtain a corresponding target noise evaluation model, and when noise data of a to-be-tested transformer is received, performing variational mode decomposition processing on the noise data of the to-be-tested transformer based on a preset mode number to obtain a plurality of noise mode components, and adopting the target noise evaluation model and the modal number to perform adaptive evaluation on each noise modal component to obtain a corresponding noise evaluation result. The technical problem that the noise evaluation accuracy of the high-frequency transformer is reduced due to the fact that the traditional noise evaluation method mainly reflects the comprehensive influence of the noise on the human body by performing frequency weighting on the sound signal of the high-frequency transformer is solved.

A method for equalizing an audio frequency signal broadcast in a broadcasting environment. Such a method includes: estimating a frequency profile of a noise signal representing a background noise present in the broadcasting environment, based on, on the one hand, a signal captured by at least one microphone implemented in the broadcasting environment and, on the other hand, the audio frequency signal; determining a desired frequency profile for the broadcast audio frequency signal; determining an acoustic frequency mask representing, for each frequency component, a difference between the frequency profile of the noise signal and the desired frequency profile; and equalizing the audio frequency signal via a weighting of a spectrum of the audio frequency signal by applying a frequency weighting mask that is a function of the acoustic frequency mask, delivering the equalized audio frequency signal.

The invention relates to the field of partial discharge quantitative calibration and electromagnetic simulation analysis, in particular to a partial discharge quantitative calibration method based on modal decomposition and frequency weighting. The method comprises the following steps: acquiring a structure parameter, a defect excitation parameter, an ultrahigh frequency sensor position parameter and a sensitive direction parameter, establishing an electromagnetic model, and performing finite difference time domain simulation to obtain a frequency domain response spectrum; obtaining each modal frequency domain component set through modal expansion and waveguide eigenmode orthogonal projection; performing inverse transformation and wavelet transformation on each modal frequency domain component set to obtain time-frequency energy distribution, and extracting total energy, energy weighted arrival time, dispersion degree and frequency energy spectrum to generate a modal characteristic quantity set; and normalizing the modal characteristic quantity set to obtain a modal quality factor weight and generate a frequency correlation weight function, synthesizing an equivalent calibration transfer function, inverting a discharge current in combination with an actually measured voltage frequency spectrum, and integrating to obtain an apparent discharge quantity. According to the invention, the quantification reliability of apparent discharge capacity can be effectively improved.

The invention relates to a borehole radar data high-resolution processing method based on mixed time-frequency analysis, and the method comprises the steps: constructing a discrete frequency based on the scale resolution of borehole radar data; performing Hilbert transform on the drilling radar data to generate an analysis signal; and filtering the signals, performing signal reconstruction, further obtaining weighted energy of different frequency bands, and finally outputting data after high-resolution processing. Dynamic weighting is realized through low-frequency and high-frequency weighting factors, and low-frequency / high-frequency components are flexibly enhanced or suppressed; through envelope normalization, pulse noise is suppressed, and the signal-to-noise ratio is improved; original phase information is reserved during frequency domain filtering through conjugate symmetric filling, signal distortion is avoided, and time positioning precision can be kept in time-frequency analysis; through different frequency band energy weighting processing and high and low frequency joint analysis, simultaneous broadening of high and low frequencies is realized, and high-resolution amplitude preservation processing is realized.

This invention discloses a method and system for diagnosing internal leakage faults in butterfly valves based on improved VMD and neural networks. The method includes: acquiring the original signal, which is the acoustic emission signal of an electro-hydraulic driven butterfly valve under different pressures and internal leakage conditions; adaptively optimizing and decomposing the original signal using an improved spectral information-guided variational mode decomposition method to obtain a finite number of IMF components; calculating the frequency-weighted energy index of each IMF component; adaptively selecting the optimal IMF component group based on the frequency-weighted energy index to construct a multi-channel tensor feature; training and optimizing a pre-constructed multi-channel attention convolutional neural network based on the multi-channel tensor feature to obtain a fault diagnosis model for internal leakage in ship butterfly valves; and acquiring online internal leakage acoustic emission signals for real-time fault diagnosis. This invention effectively overcomes strong background noise interference, avoids mode aliasing, and significantly improves the accuracy and robustness of internal leakage fault diagnosis in ship butterfly valves.

The invention discloses a BO-BiCGStab-based TI-DAC amplitude-frequency response pre-equalization method and system, belongs to the technical field of digital signal processing, and solves the problems that the TI-DAC is influenced by factors such as zero-order roll-off keeping characteristic, inconsistent sub-channel amplitude-frequency response and non-ideal analog link, so that the amplitude-frequency response of an output signal passband is uneven, and the signal quality is reduced. Firstly, TI-DAC passband amplitude-frequency response is modeled, influence factors of the TI-DAC passband amplitude-frequency response are analyzed, and precoder frequency response is converted into a linear equation set form in a cosine matrix form. And then a stable biconjugate gradient algorithm is adopted to iteratively solve the linear equation set, finally, a frequency weighting mechanism is introduced in a modeling stage, and Bayesian optimization is utilized to perform adaptive configuration on the frequency weight to obtain an optimal coefficient of the pre-equalizer. The method is suitable for scenes such as high-speed broadband communication.

The invention provides a voice robustness enhancement method and system for airport broadcast in severe weather, and the method comprises the steps: obtaining a flight state and meteorological data, selecting voice broadcast data according to the flight state if the flight state and the meteorological data meet preset conditions at the same time, and carrying out the voice broadcast according to the selected voice broadcast data; mapping the meteorological data to obtain a predicted noise power spectrum; calculating a voice compensation gain function according to the noise power spectrum and the frequency weighting function corresponding to the noise power spectrum; enhancing the voice broadcast data through a voice compensation gain function to obtain enhanced voice output; the speech spectrum distribution is dynamically adjusted, the speech broadcast structure is optimized, and the intelligibility and the transmission reliability of broadcast information in severe weather are remarkably improved.

The application relates to a kind of electromechanical integrated equipment vibration active inhibition method and system.The method constructs condensation graph reflecting structure vibration energy transmission path by graph condensation operation and calculates graph Laplacian matrix, to select sensor position in combination with modal shape matrix systematically, ensure that modal observation matrix row is full rank under limited measuring point;Then graph Laplacian matrix is used as smoothness constraint term, the essentially underdetermined modal reconstruction is converted into convex optimization problem with physical priori and is solved, and reconstruction matrix that can uniquely and stably reconstruct modal coordinates is obtained;Finally, linear mapping is carried out on real-time vibration signal based on the reconstruction matrix and differential to generate expected modal control force, which is distributed to actuator after frequency weighting optimization, so that observation overflow is effectively inhibited, and the robustness of closed-loop vibration suppression under limited measuring point is improved.

The present application relates to the technical field of audio signal processing, and discloses an anti-wind-noise adaptive volume adjustment method and system for a riding earphone, the method comprising the following steps: S1, collecting ambient sound and source audio signals in parallel; S2, analyzing the ambient sound to determine a macro adjustment intensity; S3, analyzing the source audio in parallel to identify its content type and transient signals; S4, generating frequency-dependent target method sound quality compensation gain based on a psychoacoustic model; S5, dynamically smoothing and frequency-weighting the target gain according to the content and transient characteristics to generate an application gain; S6, fusing the macro adjustment intensity and the application gain to synthesize a final gain parameter; and S7, applying the final gain to the source audio to output the compensated audio. The present application fuses psychoacoustic model and parallel analysis of audio content, realizes frequency-level accurate compensation, solves the problems of traditional degradation and sudden changes in listening experience, and improves audio clarity and comfort.

The invention provides an intelligent detection method and device for a frequency abnormal channel, electronic equipment and a storage medium, and the method comprises the steps: carrying out the frequency spectrum transformation of seismic data by taking a channel as a unit, and extracting a plurality of channel frequency spectrum features based on the frequency spectrum data; comprise a frequency amplitude component attribute based on local maximization, a high-frequency amplitude proportion, a frequency percentage based on a frequency spectrum energy proportion, a frequency spectrum amplitude based on frequency weighting, a channel frequency spectrum third-order moment characteristic and the like. Due to the problems of multiple types of frequency abnormal signals, large scale range difference, different intensities, uncertain appearing positions and frequency ranges, no rigid standard and the like, deep learning and detection of the frequency abnormal signals are facilitated through optimal design of frequency spectrum features; the deep network model can be utilized to perform feature extraction and learning on sparse frequency abnormal samples in a large amount of data, and frequency abnormal channels can be quickly and accurately identified in a node channel set. The seismic data acquisition construction can be smoothly carried out, the quality of the seismic data can be guaranteed and the seismic data can be submitted on time, and the quality guarantee can be provided for the post-processing of the seismic data.

The invention relates to a method for carrying out building vibration evaluation by adopting a Z vibration level, and the method comprises the steps: carrying out the fast Fourier transform of an obtained original acceleration time-history signal of a complete duration of a building, and obtaining an original acceleration frequency spectrum; performing frequency weighting on the original acceleration frequency spectrum to obtain a weighted acceleration frequency spectrum; inverse fast Fourier transform is carried out on the weighted acceleration frequency spectrum to obtain a weighted acceleration time history signal with complete duration; the 1-second rectangular window function slides along the time axis of the weighted acceleration time history signal, a series of 1-second acceleration signals are extracted, and the overlapping rate between adjacent 1-second rectangular windows can be specified; calculating the root mean square of the acceleration signal in each 1-second time length, and then calculating the Z vibration level; after the Z vibration level is obtained, comfort evaluation of the building can be carried out.

The application discloses an anti-voice conversion watermark embedding method and program product based on a diffusion model, which comprises inputting real-time carrier voice into a trained anti-voice conversion watermark embedding network based on a diffusion model to generate watermark-containing voice; the training method of the anti-voice conversion watermark embedding network comprises: preprocessing received historical carrier voice to generate logarithmic mel spectrum and diffusion prior, and sending the historical carrier voice into the anti-voice conversion watermark embedding network together with the logarithmic mel spectrum and the diffusion prior to generate watermark-containing voice; the model parameters of the anti-voice conversion watermark embedding network are constantly updated with the minimum total loss function as the target, the total loss function comprises an adversarial loss function between the conversion results of the carrier voice and the watermark-containing voice, a diffusion loss function between the predicted noise and the actual added noise, and a frequency weighting loss function of the watermark-containing voice and the carrier voice under multi-scale short-time Fourier transform. The application can guarantee the imperceptibility of the watermark, improve the anti-voice conversion performance and robustness of the watermark-containing voice.

The invention relates to the technical field of music data processing, in particular to a violin playing auxiliary system based on artificial intelligence, and the system comprises an audio collection module which employs a microphone array to capture a violin playing sound field, and generates an original audio stream. A coding adaptation module generates a transmission code stream while extracting side information metadata including a noise masking threshold and a transient signal flag. The feature extraction module outputs a high-dimensional feature vector. The reconstruction module generates reconstructed audio. And the evaluation feedback module performs quantitative evaluation on the reconstructed audio, generates scores of intonation deviation degree, string rubbing depth and overtone purity, feeds the scores back to the coding adaptation module, and dynamically adjusts frequency weighting parameters of the psychological acoustic model. The music interface module outputs reconstructed audio and structured evaluation data. The system drives the coding adaptation module to adaptively optimize a compression strategy through score feedback of the evaluation feedback module, and the reconstruction module optimizes a generation process by using historical evaluation data to form a closed-loop data stream.

The application relates to a method for evaluating building vibration by using a Z vibration level, which comprises the following steps: performing fast Fourier transform on an original acceleration time history signal of a complete time length of a building to obtain an original acceleration frequency spectrum; performing frequency weighting on the original acceleration frequency spectrum to obtain a weighted acceleration frequency spectrum; performing inverse fast Fourier transform on the weighted acceleration frequency spectrum to obtain a weighted acceleration time history signal of the complete time length; sliding a 1-second rectangular window function along a time axis of the weighted acceleration time history signal to extract a series of 1-second acceleration signals, and a specified overlap rate can be set between adjacent 1-second rectangular windows; calculating the root mean square of each 1-second acceleration signal and then calculating a Z vibration level; and performing comfort evaluation of the building after the Z vibration level is obtained.

The invention discloses a micro-seismic event positioning confidence ellipsoid construction method based on event detection frequency weighting, and the method comprises the steps: operating a micro-seismic monitoring system for a long time, generating a record containing an event number and a marking amount for reflecting the availability of a station for each micro-seismic event, and constructing a historical micro-seismic event directory; performing event screening to obtain a micro-seismic event set; counting the number of micro-seismic events which are successfully recorded by the stations in the station network and participate in positioning; constructing a detection frequency index reflecting the actual monitoring capability of the station; constructing a weight matrix of all stations in the station network by taking the detection frequency index as the weight of the stations; solving an event positioning result; constructing a residual vector and a sensitivity matrix; estimating a parameter covariance matrix by adopting a weighted method; 3 * 3 sub-matrixes corresponding to the position parameters are extracted from the covariance matrix, and characteristic decomposition is carried out; and solving the half-axis length of the confidence ellipsoid under a given confidence level. According to the method, the confidence ellipsoid with higher precision can be constructed, and the precision and reliability of micro-seismic monitoring can be improved.

The application discloses an aerospace hot-end metal component simulation test platform corrosion thinning online monitoring method and system, wherein the method comprises the following steps: using laser vertical incidence of the measured hot-end metal component to excite ultrasonic waves; receiving the ultrasonic echo signal through the electromagnetic ultrasonic probe arranged on the surface of the measured hot-end metal component; extracting a plurality of resonance frequencies and corresponding resonance orders based on the ultrasonic echo signal; selecting adjacent resonance frequencies and non-adjacent resonance frequencies to calculate a plurality of groups of first-order resonance frequencies with different accuracies, performing weighted average on the plurality of groups of first-order resonance frequencies, and obtaining a first-order resonance frequency weighted average value and a weighted uncertainty; and further calculating the thickness of the measured hot-end metal component based on the relationship between the resonance frequency and the thickness. The application realizes online monitoring of the corrosion thinning of the high-temperature alloy material, and can overcome the flutter condition. A method for measuring uncertainty with different accuracies is proposed, and all effective data are used as much as possible to reduce the measurement error.

This application discloses a method and apparatus for particulate matter detection based on dual coupling of signal strength and signal fluctuation frequency. The method includes: acquiring a particulate matter detection signal and a current ambient temperature; determining a first weighting coefficient corresponding to the signal fluctuation frequency of the particulate matter detection signal based on the current ambient temperature; the first weighting coefficient being positively correlated with the current ambient temperature; determining a second weighting coefficient corresponding to the signal strength of the particulate matter detection signal based on a preset weighting relationship; the preset weighting relationship indicating that the second weighting coefficient is complementary to the first weighting coefficient; and weighting and fusing the signal strength and the signal fluctuation frequency based on the second weighting coefficient corresponding to the signal strength and the first weighting coefficient corresponding to the signal fluctuation frequency to obtain the target particulate matter concentration. This application improves the accuracy and adaptability of particulate matter concentration detection in complex environments.