6 results about "Time frequency filtering" patented technology

The invention relates to a real-time water-turbine generator set frequency filtering method and system. The method comprises the steps that a frequency measurement value of a water-turbine generator set is calculated, and a state equation and a measurement equation are established for parameter prediction; updating a state estimation value in real time by adopting an improved Kalman filtering algorithm; calculating a measurement trust coefficient R value according to a concept of fusion variance of a measurement equation, and further calculating a frequency prediction value at the moment; calculating a standard deviation according to the state estimation value and the average value at the previous moment, performing nonlinear transformation on the standard deviation according to a state equation to obtain a prediction trust coefficient Q value, and calculating a state prediction value at the moment; when the Kalman gain tends to be smooth, calibrating a state estimation value; according to the method, on the premise that hardware is not modified, fusion filtering and noise covariance estimation are carried out on the collected data through the improved efficient Kalman filtering algorithm, the measurement precision is improved, and the purpose of measuring the real frequency value is achieved.

The invention discloses an adaptive method for removing ringing artifacts in single-element endoscopic ultrasonic based on time-frequency analysis filtering. Steps: obtain original data of endoscopic single-element ultrasonic imaging, perform time-frequency spectrum analysis on adjacent frame signals; determine mutation of the time-frequency spectrum of the signal to be processed relative to reference signal by using an optical flow method; determine a threshold coefficient and spectral distribution factors to set parameters of a time-frequency filter; perform time-frequency filtering on the signal to be processed, and then perform time-domain recovery; calculate index of filtered and original data, and optimize and adjust threshold parameters based on whether index meets standard; perform time-frequency filtering, time-domain reconstruction, and ultrasonic endoscopic image reconstruction. The method nearly eliminates ringing artifacts, restores effective echo signals, and maintains structural integrity during close-range imaging.

Provided is a self-adaptive single-array-element endoscopic ultrasonic ringing artifact removal method based on time-frequency analysis filtering. The method is as follows: acquiring endoscopic single-array-element ultrasonic imaging original data, and performing time-frequency spectrum analysis on adjacent frame signals; according to an optical flow method, obtaining a time-frequency spectrum mutation condition of a time-frequency spectrum of a signal to be processed relative to a reference signal; determining a threshold coefficient and a frequency spectrum distribution condition factor, and determining parameters of a time-frequency filter by means of the two parameters; performing time-frequency filtering on the signal to be processed by means of a time-frequency filter obtained by means of preliminary calculation and performing time domain recovery; performing index calculation on data after the artifact is filtered out by using time-frequency filtering and the original data, and optimizing and adjusting threshold parameters according to whether an index reaches a standard or not; and performing time-frequency filtering, time domain reconstruction, and ultrasonic endoscopic image reconstruction on the data by using a time-frequency filter parameter with the index reaching the standard. According to the method, the ringing artifact can be almost completely removed, an effective echo signal is recovered, and the structural integrity during short-distance imaging is maintained.

The application discloses a kind of motion imagination eeg signal decoding methods based on full-scale filtering double granularity interaction, comprising: obtaining motion imagination eeg signal and pretreatment;Frequency decoupling is carried out to filtered motion imagination signal;Based on the time-frequency filter designed, complete global time-domain features and all important local time-domain features are captured;Time convolution and space convolution are applied to map to space with more abundant potential features;Based on double granularity feature interaction module, space coarse-grained information and space fine-grained information are guided to mutually excavate meaningful supplementary information;Full connection layer is applied to output the probability that motion imagination eeg signal belongs to certain motion imagination category, and finally realize the decoding of motion imagination eeg signal.The application creatively realizes the capture of complete global time-domain features and all important local time-domain features in a parameter-friendly manner, and innovatively designs a double granularity feature interaction module to comprehensively and systematically excavate meaningful supplementary information.

The invention discloses a motor imagery electroencephalogram signal decoding method based on full-scale filtering dual-granularity interaction. The motor imagery electroencephalogram signal decoding method comprises the steps that motor imagery electroencephalogram signals are obtained and preprocessed; performing frequency decoupling on the filtered motor imagery signal; capturing a complete global time domain feature and all important local time domain features based on a designed time frequency filter; mapping to a space with richer potential features by applying time convolution and space convolution; the space-time coarse-grained information and the space-time fine-grained information are guided to mutually mine meaningful supplementary information based on a dual-granularity feature interaction module; and outputting the probability that the motor imagery electroencephalogram signal belongs to a certain motor imagery category by applying a full connection layer, and finally realizing the decoding of the motor imagery electroencephalogram signal. Complete global time domain features and all important local time domain features are captured creatively in a parameter-friendly mode, and a dual-granularity feature interaction module is creatively designed to comprehensively and systematically mine meaningful supplementary information.

The present invention belongs to the field of electronic countermeasure technology, and discloses a method, system for individual identification of communication radiation sources based on time-frequency filter coefficient fusion. The method includes: using a constant modulus algorithm in the time domain to perform equalization filtering on the transmitted signal carrying the radiation source fingerprint, extracting the equalizer coefficients after equalization convergence; converting the equalized, filtered signal to the frequency domain, applying a Mel filter bank to perform homomorphic filtering on the signal in the frequency domain to obtain the filtered Mel filter coefficients; performing frame-wise convolution on the equalizer coefficients, the Mel filter coefficients to fuse the time-domain, frequency-domain features of the communication radiation source, extract the communication radiation source fingerprint; constructing an individual identification model for communication radiation sources; identifying, classifying the extracted communication radiation source fingerprint features based on the individual identification model for communication radiation sources.