21 results about "Adaptive filter coefficients" patented technology

The invention provides a hearing aid intelligent protection device and method integrated on a headgear, and the method comprises the steps: converting an environment sound wave signal into a frequency domain signal through a signal processor in an interlayer of the headgear, and separating a noise dominant frame and a voice dominant frame in a hearing aid; determining a voice masking index between the environmental noise and the voice in the environmental sound wave signal according to a distribution association relationship between the noise dominant frame and the voice dominant frame; according to the energy distribution characteristics of the noise spectrum in the frequency domain signal, determining the noise interference suffered by the hearing aid in the current wearing scene, and determining the transient noise dynamic characteristics of the environment sound wave signal under the current noise interference; and performing disturbance trend prediction on voice signal noise in the hearing aid according to the transient noise dynamic characteristics and the voice masking index, and further updating an adaptive filtering coefficient of the hearing aid according to a prediction result. By adopting the scheme of the invention, the prediction of the noise disturbance trend in a complex and changeable environment can be realized so as to adaptively adjust the noise reduction filter coefficient of the hearing aid.

The purpose of the present invention is to update and easily set an adaptive filter coefficient with an optimal value for a control filter that generates a noise-cancelled signal. An active vibration noise reduction device (100) is characterized by comprising: a speaker (20) that outputs interference sound (y) for cancelling noise (d); a microphone (30) that generates an error signal (e) from the noise (d) and the interference sound (y); a control filter W that generates a control signal (u) for controlling the interference sound (y) from the reference signal (r); and a secondary path filter (C ^) that represents an estimated value of a transfer function from the speaker (20) to the microphone (30). The control filter (W) adaptively updates using an update amount [Delta] W obtained by multiplying the error signal (e), a step parameter ([mu] W (t)) calculated on the basis of the error signal (e), and the reference signal (r) by a convolution operation result of the secondary path filter (C ^).

The invention relates to the technical field of power system automation and intelligent operation and maintenance, and discloses a multi-parameter real-time monitoring and anomaly diagnosis method and device for a high-voltage direct-current power transmission system, and the method comprises the steps: obtaining a monitoring interface image frame according to a preset period, and generating a semantic state vector containing a color flag bit and a working condition mode code through a semantic analysis module; the numerical value processing module determines a state self-adaptive filtering coefficient to execute self-adaptive state gating filtering on an original numerical value in combination with the character confidence coefficient and the semantic state vector; the anomaly analysis module executes energy conservation calculation and bipolar gradient analysis based on the smooth monitoring value, and generates an alarm signal. According to the method, through fusion of semantic perception and adaptive filtering, the contradiction between non-intrusive acquisition of steady-state jitter and transient delay is effectively solved, and the reliability of diagnosis of system logic conflicts and dynamic asynchronous anomalies is improved by utilizing physical mechanism constraints.

To provide an active noise reduction device that can shorten a delay concerning output of a correlation sound.SOLUTION: An active noise reduction device 10 includes: a first signal processor 11 that applies an adaptive filter to a reference signal having a correlation with noise in a space in a vehicle 50 to generate a cancellation signal for outputting a cancellation sound for reducing noise; a second signal processor 12 that updates a coefficient of the adaptive filter; and a sample rate converter 17. The first signal processor 11 operates in a cycle T1. The second signal processor 12 operates in a cycle T2 longer than the cycle T1. The sample rate converter 17 upsamples the coefficient of the adaptive filter updated by the second signal processor 12 for output to the first signal processor 11. The cycle T1 is longer than a difference between the maximum value and the minimum value of a processing time required until the second signal processor updates the coefficient of the adaptive filter after acquiring the reference signal.SELECTED DRAWING: Figure 1

The present invention relates to an ultra-directional speaker system comprising: a first envelope calculation unit for calculating the envelope of the currently input audio input signal; a square root calculation unit for generating a square root signal of a first envelope signal by calculating a square root of the first envelope signal calculated by the first envelope calculation unit; a pre-distortion adaptive filter unit for generating a compensation signal by performing distortion compensation in which an adaptive filter coefficient update term according to an adaptive filter coefficient determined in a previous stage is applied to the currently input audio input signal; a second envelope calculation unit for generating a second envelope signal by calculating the envelope of the compensation signal; an error calculation unit for generating an error signal by comparing and calculating the second envelope signal with the square root signal of the first envelope signal; an adaptive filter coefficient update unit for calculating an adaptive filter coefficient update term and an adaptive filter coefficient from the error signal; a dynamic signal modulation unit for generating a modulation signal by dynamically modulating the compensation signal into an ultrasonic band; an ultrasonic transducer model for modeling an inverse filter corresponding to frequency characteristics of an ultrasonic transducer, and generating a filtering signal by applying the inverse filter to the modulation signal; an ultrasonic amplifier for amplifying the filtering signal; and the ultrasonic transducer for converting the amplified signal into an ultrasonic signal, wherein the dynamic signal modulation unit includes: an AM signal modulation unit for generating an AM modulation signal of transmission data; a first sideband signal modulation unit, which receives the AM modulation signal and performs a Hilbert operation thereon so as to obtain a first sideband modulation signal from which a carrier wave has been removed; and a second sideband signal unit, which receives the first sideband modulation signal and adds the carrier wave thereto so as to obtain a second sideband modulation signal.

The embodiment of the invention provides an image decoding method for adaptive loop filtering. The image decoding method may include a step of acquiring information on adaptive loop filtering from a bitstream. The image decoding method may include a step of obtaining a first filtered residual sample point corresponding to a current sample point by performing filtering using a residual sample point of the current sample point and a first filter. The image decoding method may include a step of obtaining a second filtered residual sample corresponding to a current sample by performing filtering using a first filtered residual block and a second filter. The image decoding method may include a step of acquiring an adaptive loop filtering sample point by using a first filtering residual sample point, a second filtering residual sample point, and at least one adaptive filter coefficient acquired from information on adaptive loop filtering.

The application provides a kind of multimodal synergistic industrial voice control method.Application to intelligent manufacturing technology field, the method includes collecting industrial field original voice stream, generates noise spectrum entropy value by frequency domain feature analysis, according to the noise interference level determined by dynamic matching array beam forming weight and adaptive filter coefficient, and gets pure voice instruction by noise reduction;Extract its voiceprint features and semantic text, generate instruction semantic structure tree combined with industrial knowledge graph, calculate identity matching degree and semantic analysis confidence respectively, and get instruction comprehensive effectiveness coefficient by weighting;Satisfy the benchmark time to map it to sub-instruction, evaluate process conflict and equipment risk through digital twin pre-play, determine execution strategy;Real-time acquisition of running data in execution, whether to trigger abnormal rollback is determined according to trajectory deviation rate.This improves the efficiency of industrial voice control.

A method of operating a repeater comprises selecting some of adaptive filter coefficients among a plurality of adaptive filter coefficients of an adaptive filter used for interference cancellation, based on the size of a coefficient, generating a predicted interference signal using the selected adaptive filter coefficients and generating an interference-canceled communication signal from a received signal using the generated predicted interference signal.

To update an adaptive filter coefficient with the optimal value to easily set the filter coefficient to a control filter that generates a signal for cancelling a noise.SOLUTION: An active vibration noise reduction device 100 comprises: a speaker 20 that outputs a cancellation sound y for cancelling a noise d; a microphone 30 that generates an error signal e from the noise d and the cancellation sound y; a control filter W that generates, from a reference signal r, a control signal u for controlling the cancellation sound y; and a secondary path filter C^ indicating an estimated value of a transmission function from the speaker 20 to the microphone 30. The control filter W is adaptively updated by an update amount ΔW determined by multiplying the error signal e, a step size parameter μW(t) calculated on the basis of the error signal e, and a result of convolution operation of the reference signal r and the secondary path filter C^.SELECTED DRAWING: Figure 1

The embodiment of the invention provides a time domain channel estimation method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining a pre-stored filtering coefficient group of a data symbol based on a time domain correlation function of a line-of-sight path and the symbol number of a demodulation reference signal; according to the type of the current data symbol and a preset coherence time judgment mechanism, selecting a filter coefficient used by the current data symbol from a pre-stored filter coefficient group; and performing interpolation filtering according to the filtering coefficient used by the current data symbol and the frequency offset compensation strategy to obtain a time domain channel estimation result of the current data symbol. According to the method, a sight distance path is utilized to obtain a time domain correlation function closer to an NTN scene, so that an accurate filtering coefficient is generated; on the basis of a judgment mechanism of coherence time, adaptive filtering coefficients are dynamically selected and used, and the filtering effect of edge data symbols is improved; meanwhile, in combination with a frequency offset compensation mechanism, the Doppler frequency shift influence is effectively suppressed, and the channel estimation precision and stability are improved.

The invention provides a radar tide level data non-stationary noise adaptive filtering method, which belongs to the technical field of radar tide levels, and comprises the following steps: carrying out segmented time-frequency domain conjoint analysis on radar tide level original echo signals to extract time-varying noise feature vectors; after abnormal noise nodes are recognized and eliminated by adopting a spectral clustering algorithm, cubic spline interpolation repair is performed, a preprocessed signal is input into a frequency domain self-adaptive denoising model, the model extracts multi-band noise features by utilizing a frequency domain convolution acceleration mechanism and parallel branches, a self-adaptive filter coefficient is calculated according to a noise estimation result, and a noise estimation result is obtained. The updating step length of the filter is dynamically adjusted according to the noise energy change rate, residual signal energy is monitored and iteratively optimized until the precision requirement is met, and the technical problem that a fixed parameter filter cannot accurately track noise instantaneous change due to non-stationary noise characteristic time variation is solved.

The application relates to the technical field of voice processing, in particular to a voice enhancement method and system of a loudspeaker. The method comprises the following steps: acquiring a sound signal, performing frame processing, extracting multi-dimensional voice features, presetting a classification rule, outputting a human voice confidence, generating human voice existence information; acquiring a loudspeaker reference signal, performing acoustic echo cancellation processing to obtain an echo cancellation output signal, and adjusting an adaptive update strategy according to the human voice confidence when human voice exists, namely, freezing adaptive filter coefficient update and / or reducing the update strength of the adaptive filter; identifying a main noise frequency band of fluctuant noise, performing noise suppression processing to obtain a noise suppression output signal; acquiring state information, adjusting update parameters and / or suppression strength, and outputting a voice-enhanced loudspeaker signal. The method solves the problem that an existing loudspeaker system is difficult to effectively adapt to background noise and echo influence in a complex acoustic environment, thereby causing the intelligibility of voice to decrease.

An active noise control device includes an adaptive filter that generates a control signal based on a vibration frequency generated by a vibration source and sequentially updating the adaptive filter, a reference signal generation device that generates a reference signal, an amplitude characteristic calculation device that acquires an acoustic characteristic including amplitude and phase information, the acoustic characteristic being an acoustic characteristic of a secondary path between a speaker and a microphone and calculating an amplitude characteristic of the secondary path, a smoothed signal generation device having a low-pass filter, a correction coefficient calculation device that calculates a correction coefficient according to a frequency, an adaptive filter update device that updates the adaptive filter by subtracting an update term including the correction coefficient from a first adaptive filter coefficient that is an immediately preceding adaptive filter coefficient, and a control signal generation device that generates the control signal.

The application provides a radar tide level data non-stationary noise adaptive filtering method, and belongs to the technical field of radar tide level.The original echo signal of the radar tide level is subjected to segmented time-frequency domain joint analysis to extract a time-varying noise characteristic vector, a spectrum clustering algorithm is used to identify and remove abnormal noise nodes, and then cubic spline interpolation repair is carried out, the pretreated signal is input into a frequency domain adaptive denoising model, the model uses a frequency domain convolution acceleration mechanism and a parallel branch to extract multi-frequency noise characteristics, adaptive filter coefficients are calculated according to noise estimation results, the filter update step is dynamically adjusted according to the noise energy change rate, and the residual signal energy is monitored and iteratively optimized until the accuracy requirement is met, so that the technical problem that a fixed parameter filter cannot accurately track the instantaneous change of noise due to the time-varying of non-stationary noise characteristics is solved.

This invention discloses a line spectrum enhancement method and system with adaptive sparse constraint strength. The method includes: initializing relevant parameters; processing the input time-domain signal to obtain a frequency-domain input signal; weighting the frequency-domain input signal based on the frequency-domain coefficient vector to obtain a line spectrum enhanced output signal and calculating the corresponding error signal; constructing a cost function based on the error signal and introducing a sparse constraint term weighted by the sparse constraint strength; minimizing the KL divergence between the probability distribution of the frequency-domain coefficient vector and the sparse prior distribution, and adaptively iteratively updating the sparse constraint strength; updating the frequency-domain coefficient vector based on the error signal and the iteratively updated sparse constraint strength; and repeating the above process until the time-domain signal iteration is complete. This invention achieves dynamic optimization of sparse constraint parameters by introducing an adaptively adjustable sparse constraint strength during the adaptive filter coefficient vector update process, thereby improving the robustness and performance of line spectrum enhancement.

An active noise control device includes: a first signal processor that generates a cancellation signal for outputting a cancellation sound for reducing noise in a space inside an automobile, by applying an adaptive filter to a reference signal correlating with the noise; a second signal processor that updates a coefficient of the adaptive filter; and a sample rate converter. The first signal processor operates in cycle T1, the second signal processor operates in cycle T2 that is longer than cycle T1, and the sample rate converter upsamples the coefficient of the adaptive filter updated by the second signal processor and outputs the coefficient upsampled to the first signal processor. Cycle T1 is longer than a difference between maximum and minimum values of a processing time required from when the second signal processor obtains the reference signal to when the second signal processor updates the coefficient of the adaptive filter.

An image decoding method may include obtaining information regarding adaptive loop filtering from a bitstream, obtaining a first filtered residual sample corresponding to a current sample by performing filtering based on a residual sample for the current sample and a first filter, obtaining a second filtered residual sample corresponding to the current sample by performing filtering based on the first filtered residual block and a second filter, and obtaining an adaptive loop filtered sample based on the first filtered residual sample, the second filtered residual sample, and at least one adaptive filter coefficient obtained from the information regarding the adaptive loop filtering.

The invention discloses an in-vehicle echo cancellation method based on a variable proportion decorrelation least mean square algorithm. The method comprises the following steps: sampling a vehicle signal; updating coefficients of the decorrelation filter; calculating partial decorrelation signals; self-adaptive filter coefficient updating; echo cancellation; calculating signal power; updating a proportional control factor; and repeating the steps until the dialogue is finished. The method has the advantages of being high in convergence speed and good in steady-state error.

The application discloses a full duplex communication multistage self-interference suppression and channel equalization integrated method, which comprises the following steps: S1, giving the expression of a received signal and a coupled self-interference signal of a receiving end through a RRC filter with a sampling rate of K times a baseband sampling rate; S2, discretely sampling the received signal and a feedback signal at the receiving end; S3, performing minimum mean square error estimation of an expected symbol; S4, estimating the symbol in a multistage iteration mode, and obtaining an optimal coefficient equation of an adaptive filter according to the minimum mean square error estimation; and S5, updating the adaptive filter coefficient through the iteration mode by using an RLS algorithm, and obtaining an estimated value of the expected signal by using the updated coefficient. The application has lower system complexity and lower inter-stage delay, reduces overall quantization noise based on channel estimation and equalization.

The invention relates to the technical field of ocean detection and navigation, in particular to a deep sea underwater vehicle positioning system and method. Comprising a multi-sensor fusion positioning module for performing weighted fusion on acoustic, inertial and magnetic field sensor data by using a data fusion algorithm; the self-adaptive signal processing module integrates a noise monitoring unit and a self-adaptive signal processing algorithm in the acoustic sensor, detects the intensity and frequency characteristics of marine environment noise in real time and adjusts the coefficient of a self-adaptive filter and the gain of a signal in real time; the map matching correction module is used for establishing a submarine topography map and a geomagnetic field distribution map, matching map data with measurement data, detecting data deviation and performing positioning correction; and the underwater relay node auxiliary positioning module is used for arranging relay nodes in the deep sea, receiving signals sent by the underwater vehicle and forwarding the signals to the water surface base station, and the water surface base station performs positioning and calculation to obtain position coordinates of the underwater vehicle. According to the invention, the positioning precision, the operating distance and the anti-interference capability of the underwater vehicle can be improved.

The invention discloses a robust distributed echo cancellation method based on information entropy, which is executed in a network formed by a plurality of audio processing nodes, and comprises the following steps of: at each node, filtering a local reference signal by using a current local adaptive filter coefficient to generate a local echo estimation signal; subtracting the local echo estimation signal from a near-end signal acquired by a local microphone to obtain a local echo cancellation error signal; taking minimization of the information entropy of the error signal as a criterion, and updating an intermediate coefficient vector of a local adaptive filter through a normalized Euclidean direction search algorithm; and finally, carrying out global coefficient diffusion fusion. According to the method, the minimum error entropy criterion and the normalized Euclidean direction search are combined, the diffusion network topology is fused, and the robustness, the convergence speed and the calculation complexity can be effectively balanced.