1069 results about "Discrete Fourier transform" patented technology

The embodiment of the invention provides an image semantic segmentation method and device fused with Fourier transform, and the method comprises the steps: constructing a segmentation model based on a U-shaped network, and carrying out the feature extraction through employing a four-layer convolution pooling network as an encoder; constructing a Fourier transform module, performing two-dimensional discrete Fourier transform on the input image to obtain a spectrogram, and performing amplitude spectrum and phase spectrum decomposition on the spectrogram to obtain an initial frequency domain feature; performing preset size scaling and pixel normalization processing on the input image to obtain a pre-processed image, and inputting the pre-processed image into a Fourier transform module to obtain four groups of adjusted frequency domain features; a bilinear interpolation algorithm is adopted to adjust the size of the prediction probability graph to the size of an original input image, pixel points with probability values larger than 0.5 in the prediction probability graph are marked as foreground areas, and a final semantic segmentation result is obtained; according to the method, the semantic segmentation task in a complex scene can be better processed, and the segmentation precision and the model robustness are improved.

The invention discloses an interference suppression method for a random phase modulation electromagnetic metasurface. The interference suppression method comprises the following steps: constructing a mathematical model for a modulated radar echo signal; wherein a random phase modulation electromagnetic metasurface is installed on the surface of a target, a radar transmitting signal enters the random phase modulation electromagnetic metasurface, and after random time domain phase modulation, a radar echo signal is generated; discrete sampling processing is carried out after digital down-conversion and dechirping are carried out on radar echo signals, and then additional phase information is estimated after self-mixing processing is carried out. Phase compensation is carried out on the radar echo signal after discrete sampling processing under the condition that noise is considered, and then time domain differential transformation and discrete Fourier transformation are carried out; estimating a phase mutation moment by using a subspace algorithm, and reconstructing an interference modulation sequence based on an estimation result; and demodulating random phase interference by using the reconstructed interference modulation sequence, and performing matched filtering processing to obtain an output signal.

The invention discloses a self-calibration method based on a planar two-dimensional time grating sensor, which comprises a movable ruler base body and a fixed ruler base body which are arranged up and down, excitation electrodes are uniformly arranged on the fixed ruler base body, and two induction electrode groups are arranged on the movable ruler base body along the diagonal direction. Each induction electrode group is composed of a plurality of induction electrode arrays arranged in the X direction and the Y direction. Equal-amplitude and same-frequency sine excitation voltages with the phase difference of pi / 2 are applied to different types of excitation electrodes respectively, and the induction electrodes of the induction electrode group are subjected to combined operation, so that displacement signals of the single induction electrode group in the X and Y directions can be decoupled and output; the method comprises the following steps of: performing combined operation on output signals among sensing electrode groups arranged along a moving direction to obtain an output signal containing measurement error information, performing discrete Fourier transform on data obtained after processing the output signal, solving amplitude and phase parameters to construct an error function, and calculating the measurement error of the sensing electrode groups. And self-calibration correction is carried out on measurement errors of the sensor along the X direction and the Y direction.

The invention discloses a lithium battery state estimation method based on relaxation time scale modeling, and the method comprises the steps: 1, carrying out a sudden pulse charging test on a lithium battery, collecting voltage and current data, and calculating the open-circuit voltage, direct-current internal resistance and a time-domain impedance sequence of the lithium battery; and 2, converting the time domain impedance sequence of the lithium battery into a frequency domain impedance sequence after discrete Fourier transform, establishing a fitting equation set of a distribution function of the relaxation time scale DRT of the lithium battery and the frequency domain impedance sequence, and solving by adopting a regularized least square method. And 3, taking the solved DRT model distribution function as the input of a neural network, taking the state parameters of the lithium battery as the output of the neural network, and establishing a prediction model of the state parameters of the lithium battery. According to the method, EIS test required by traditional relaxation time scale modeling is avoided, the control complexity and design cost of test acquisition equipment are effectively reduced, and the method can be effectively applied to occasions such as battery life prediction, battery classification and screening, battery echelon utilization and the like.

The invention discloses a broadband oscillation detection and tracing method, a storage medium and a processor, relates to the technical field of power systems, and solves the technical problem of low accuracy of detection and tracing of broadband oscillation signals in the prior art. The method comprises the following steps: constructing a mathematical model of a broadband oscillation signal; converting the broadband oscillation signal to a frequency domain based on discrete Fourier transform to obtain a plurality of frequency decomposition values; detecting the broadband oscillation signal based on Hilbert transform; respectively carrying out image matrix transformation on the plurality of coded data to obtain an oscillation characteristic matrix; and training the deep learning model based on the oscillation characteristic matrix and the corresponding oscillation position label to obtain an oscillation positioning model. According to the invention, through the strong sequence feature capturing capability of the convolutional neural network, each node feature vector and the system adjacency matrix are input to the oscillation positioning model, high-precision fitting of the broadband oscillation signal is realized, and the accuracy of detection and traceability of the broadband oscillation signal is improved.

The invention relates to the technical field of geotechnical mechanics and engineering, and particularly discloses a soft soil foundation settlement monitoring method and system based on multi-field multi-source information, and the method comprises the steps: collecting multi-source data of a target soft soil area; performing constitutive parameter inversion, data standardization and discrete Fourier transform frequency domain conversion on the data to obtain a standardized frequency domain multi-field multi-source data set; based on the data set, a soft soil constitutive parameter library and a soil mechanics physical equation, constructing an FD-PINN frequency domain physical information neural network, embedding the physical equation as a prior constraint, and training the model by adopting an alternating optimization strategy; inputting the real-time frequency domain data flow into the model, and outputting the current settlement amount, the settlement rate and the multi-physical field frequency domain distribution; and in combination with a pre-established large scale model test result, through IDFT inverse transformation, a time-space domain settlement field is reconstructed, multi-stage early warning is triggered, a targeted reinforcement scheme is recommended, the soft soil foundation settlement monitoring precision and the engineering practicability are remarkably improved, and the method is suitable for construction, operation and maintenance of infrastructures such as high-speed rails and highways.

The present disclosure is generally related to wireless communications technologies, network topologies, and communication device implementations, and in particular, to multipletransmission reception point (multi-TRP) uplink (UL) transmission schemes. The multi-TRP UL transmission schemes include timing advance (TA) associations for multi-TRP operation, channel state information (CSI) enhancements, and codebook structural enhancements. To enhance the performance of Rel-16 codebook, a new codebook structure is constructed by preserving the Rel-16 codebook structure in the spatial dimension (SD) and frequency dimension (FD) while extending the Rel-16 codebook in the time dimension (TD) using a mutually orthogonal Discrete Fourier Transform (DFT) basis.

The invention discloses a time series data anti-noise anomaly detection method and system based on dynamic decomposition, and the method comprises the steps: obtaining a to-be-detected time series, inputting an anomaly detection model, and carrying out the dynamic decomposition of the time series; based on a channel-time mixed attention mechanism, obtaining channel feature representation of a trend component and time feature representation of a seasonal component; and obtaining a reconstruction sequence according to the channel feature representation and the time feature representation, calculating a difference value of each corresponding time step in the reconstruction sequence and the time sequence, and comparing and marking the difference value with an abnormal threshold value to obtain a marked abnormal detection sequence so as to complete sequence abnormal detection. According to the method, the time sequence needing to be detected is subjected to dynamic trend seasonal decomposition through learnable one-dimensional convolution and discrete Fourier transform of the anomaly detection model, the potential time pattern of the time sequence can be effectively extracted, the adaptability of the model to complex trends and periodic changes is improved, and the accuracy of abnormal data recognition is improved.

Methods, systems and apparatus for simulating quantum systems. In one aspect, a method includes the actions of obtaining a first Hamiltonian describing the quantum system, wherein the Hamiltonian is written in a plane wave basis comprising N plane wave basis vectors; applying a discrete Fourier transform to the first Hamiltonian to generate a second Hamiltonian written in a plane wave dual basis, wherein the second Hamiltonian comprises a number of terms that scales at most quadratically with N; and simulating the quantum system using the second Hamiltonian.

The invention relates to a reconfigurable battery energy storage system grid-connected and off-grid seamless switching control method and device. The method comprises the following steps of: sensing a real-time power grid state, pre-judging a system operation mode, and acquiring a millisecond-level panoramic information flow of the power grid state through a high-speed sampling voltage sensor, a high-speed sampling frequency sensor, a high-speed sampling phase sensor and a high-speed sampling communication interface. And pre-judging the stability and disturbance trend of the power grid through a deep learning algorithm, executing a power distribution and voltage support algorithm in grid-connected and off-grid modes, and calculating an optimal switching opportunity point and a system reconstruction target topology scheme. Meanwhile, the harmonic distortion rate and the voltage sag depth are calculated in real time through discrete Fourier transform, so that a mode pre-judgment mechanism is automatically triggered. After switching, key parameters such as voltage, frequency and power are automatically corrected through real-time sampling and observation, and zero-sensing switching is achieved. The method can effectively improve the response speed and stability of the battery energy storage system when the power grid fails, and guarantees the safe operation of the power grid.

The invention belongs to the field of edge computing acceleration, and relates to a fast Fourier transform and neural network reasoning collaborative computing acceleration method, which comprises the following steps of: deploying a computing method which is based on butterfly computing merging and a tensor mapping strategy and is mixed with DFT (Discrete Fourier Transform) and FFT (Fast Fourier Transform) in an operator deployment level; on the interface integration level, a bus interface of a register access path in a tensor accelerator control path is subjected to lightweight reconstruction, so that the bus interface is adaptive to an edge computing platform; in an operation level, a user-defined instruction is introduced into a tensor calculation unit, and an accelerator is enabled to independently complete a whole-process task of FFT signal processing and NN intelligent identification. According to the method, an operation instruction set is expanded on hardware, and delay overhead caused by returning a large amount of intermediate data to a processor is avoided, so that efficient execution of FFT and neural network tasks on unified hardware is guaranteed, and multiple requirements of high real-time performance, high precision and low power consumption are met.

The invention relates to the field of partial discharge source positioning, in particular to a large-scale substation multi-partial discharge source positioning method and system, and the method comprises the steps: forming a sensor array through a movable ultrasonic sensor and an ultrahigh frequency sensor, and collecting ultrasonic signals and ultrahigh frequency signals generated by partial discharge sources respectively; denoising the collected discharge signals, and calculating the distance between the partial discharge source and the sensor array; denoised ultrasonic array signals are divided into a plurality of narrowband signals by discrete Fourier transform, weighted covariance matrixes of all narrowband signals after focusing transformation are subjected to mean value calculation to obtain weighted covariance matrixes of focusing frequency points, and the weighted covariance matrixes of the focusing frequency points are subjected to unitary transformation to obtain a weighted covariance matrix of a plurality of narrowband signals. Estimating the number of partial discharge sources according to a signal source number estimation equation; and carrying out spectrum peak search in a specified space parameter range according to a preset step length to obtain a space parameter corresponding to each partial discharge source, and obtaining a partial discharge source coordinate in a three-dimensional space.

The invention relates to a cloth flaw detection method based on high-frequency feature constraint and edge entropy minimization, and the method comprises the steps: 1, collecting a flawless and abnormal cloth training picture data set, carrying out the initialization of hyper-parameters, and carrying out the initialization and fine adjustment of model parameters; 2, randomly sampling an image sample x, obtaining vector quantized image features through a feature encoder, generating flaw noise through an anomaly generator, adding the flaw noise into original features of an image, respectively inputting features with anomalies into an image reconstruction decoder and an image restoration decoder, and carrying out image reconstruction; and reconstructing an abnormal image and an abnormal image respectively, and the like. The detection method has the advantages that the high-frequency feature constraint module and the edge entropy minimization module are introduced, so that the cloth defect anomaly detection accuracy is remarkably improved; and the high-frequency feature constraint module converts the real image and the reconstructed image into frequency domain representation by using discrete Fourier transform.

The invention discloses a broadband lightning very high frequency radiation source positioning method based on an ESPRIT algorithm, and the method comprises the steps: 1, constructing an orthogonal uniform linear very high frequency antenna array, and collecting broadband lightning very high frequency signals; 2, discrete Fourier transform is carried out on the broadband thunder and lightning very-high-frequency signals, the broadband thunder and lightning very-high-frequency signals are divided into J frequency points, and the frequency domain value of each frequency point signal is worked out; step 3, dividing the orthogonal uniform linear very high frequency antenna array into two sub-arrays in an x-axis direction and a y-axis direction, positioning each frequency point value of the signals in the two directions by using an ESPRIT algorithm, and respectively obtaining a calculation incident angle of each frequency point in the two directions; step 4, carrying out phase ambiguity resolution on the calculated incident angles of each frequency point in the two directions so as to obtain a real incident angle; and 5, calculating the azimuth angle and the elevation angle of the radiation source according to the real incident angle. According to the invention, rapid and refined positioning imaging of the lightning radiation source can be realized.

The invention discloses a photovoltaic grid-connected broadband oscillation suppression system based on network construction type energy storage, and belongs to the technical field of photovoltaic grid connection. The invention aims to solve the problem of broadband oscillation in photovoltaic grid connection. Comprising the following steps: acquiring voltage and current of a public grid-connected point and energy storage voltage and current by adopting a data acquisition module, and preprocessing; a data processing module is adopted to carry out autoregressive empirical wavelet transform on voltage and current of a common grid-connected point to obtain an oscillation signal, and then interpolation discrete Fourier transform of a Hanning window is adopted to obtain modal information; filtering a fundamental component of the energy storage voltage based on the modal information by adopting an adaptive impedance remodeling module, and then filtering to obtain a filtered energy storage voltage harmonic component; self-adaptive impedance remodeling is carried out to obtain an energy storage voltage harmonic component after amplitude and phase adjustment; and inputting the harmonic component of the energy storage voltage after amplitude and phase adjustment as a voltage feedback signal to a PWM control link of the inverter, thereby realizing broadband oscillation suppression at the public grid-connected point. The method is used for photovoltaic grid-connected oscillation suppression.

The invention provides a cable defect positioning method based on a broadband impedance-clustering algorithm and a related device, and belongs to the technical field of cable defect localization, and the method comprises the steps: calculating the broadband impedance spectrum of a defective cable according to the distribution parameters of the defective cable, carrying out the discrete Fourier transform of the broadband impedance spectrum, extracting the characteristic information of the broadband impedance spectrum, and carrying out the clustering of the broadband impedance. According to the method, a cable defect initial positioning spectrogram is obtained, a clustering result of initial positioning spectrum data points is obtained from data in the cable defect initial positioning spectrogram through a DBSCAN algorithm, a final diagnosis spectrogram is drawn, defect peak data points and non-defect section data points in the final diagnosis spectrogram are distinguished, and therefore extraction and positioning of defect points are completed. According to the method, centralized weak defects such as single-point and multi-point high-resistance faults of the cable can be effectively positioned, the identification degree of the weak defects of the cable can be improved, and misjudgment caused by human subjective factors and objective factors on a test site is reduced.

The invention relates to the technical field of image recognition, and discloses a laboratory animal identity recognition method and system based on image recognition, and the method comprises the steps: carrying out the preprocessing of an animal motion image sequence collected by a high-speed camera in a laboratory, and obtaining the image data of a target animal; performing two-dimensional discrete Fourier transform and inverse Fourier transform to obtain animal image high-frequency feature data and animal image low-frequency feature data; respectively inputting into a double-branch diffusion network for feature processing to obtain fused animal feature data; constructing a positive sample pair, constructing a negative sample pair through difficult sample mining, calculating a feature distance based on cosine similarity and performing momentum updating to obtain an animal identity feature template; and based on the double-branch diffusion network and the animal identity feature template, feature extraction and similarity calculation are performed on the real-time animal image sequence to obtain animal identity data, so that the distinguishing capability of identity features is improved, and the time sequence consistency of an identity recognition result is enhanced.

The invention discloses a self-adaptive threshold SAMP reconstruction method for a power quality disturbance signal. According to the method, a compression observation value is obtained by constructing a random Gaussian observation matrix, and sparse representation is carried out on an original signal by using discrete Fourier transform. In the iterative reconstruction process, the spectrum amplitude difference is introduced for the first time to serve as an adaptive termination basis, and automatic adaptation of different noise levels and different disturbance characteristics is achieved in combination with a dynamic threshold update function. According to the method, the problems of traditional SAMP sparseness overestimation and redundant iteration are effectively avoided, and the calculation load is remarkably reduced. Compared with an OMP method, an original SAMP method and the like, the method has the advantages that the number of iterations can be reduced by 30%-60%, the reconstruction signal-to-noise ratio is increased by 2-5 dB, the root-mean-square error is reduced by 10%-25%, higher robustness and real-time performance are achieved in power quality disturbance signal reconstruction, and the method is quite suitable for scenes such as compressed sampling, edge calculation and high-speed signal reconstruction in a power quality monitoring system.

The invention relates to the technical field of food processing parameter optimization, and particularly discloses a Baijiu brewing process parameter optimization method based on artificial intelligence, which comprises the following steps of: acquiring temperature and steam flow data in a distillation process in real time, performing discrete Fourier transform processing on the temperature and steam flow data, and calculating a temperature fluctuation index and a steam flow fluctuation index; the two indexes are used as input characteristics, a wine quality prediction model is established by using a deep neural network technology, the model can predict whether the wine quality is qualified or not according to real-time data, and for the wine with an unqualified prediction result, the system can automatically analyze reasons, so that the wine quality prediction accuracy is improved. According to the method, through real-time monitoring and intelligent adjustment, the product quality problem caused by parameter fluctuation in a traditional brewing process is effectively solved, the production efficiency is improved, and the stability of the white spirit quality is ensured.

The invention discloses a noise suppression optimization method for a high-sensitivity camera module. The method comprises the following steps: S1, acquiring an original image and preprocessing the original image; s2, executing discrete Fourier transform, and converting the image into a frequency domain image; s3, inputting the frequency domain image into a Fourier neural operator model, and extracting frequency domain features; s4, inverse Fourier transform is executed, and a preliminary noise reduction image is generated to serve as a reference image; s5, adopting an improved jellyfish group optimization algorithm to perform global search and optimization on model structure parameters and hyper-parameters; s6, updating the Fourier neural operator model according to an optimization result, and performing closed-loop iterative optimization; and S7, performing image quality evaluation on the noise-reduced image and the reference image until performance indexes are met. According to the method, the Fourier neural operator and the improved jellyfish group optimization algorithm are fused, so that efficient noise suppression and image quality improvement of the high-sensitivity camera module in a low-light environment are realized.

A radar system may include transmitter modules configured to transmit multiple transmit signals in accordance with a Doppler Domain Multiplexing (DDM) scheme, receiver modules configured to receive reflections from the plurality of transmit signals reflected off an object and to generate corresponding digital signals, and a signal processor configured to generate a range-Doppler antenna cube representing the digital signals, and, for each of range bin of the cube, to generate a decoded range-Doppler bit map (RDBM) from the range-Doppler antenna cube by generating decoded RDBM rows by extracting a range bin matrix, performing a Discrete Fourier Transform on range bin matrix to generate an output matrix, determining peak locations in the output matrix, comparing associated peak locations across transmit channels of the output matrix to identify location-matched peaks associated with the Doppler bin, and generating a decoded RDBM row based on the location-matched peaks.

The invention provides a power grid frequency detection method, system and device and a medium, and the method comprises the steps: carrying out the narrowband signal decomposition of an obtained input voltage signal of a power grid through a null-space tracking algorithm, and obtaining a fundamental main mode component of the input voltage signal; performing low-pass processing on the fundamental wave main mode component to obtain a fundamental wave signal of the fundamental wave main mode component; performing discrete Fourier transform on the fundamental wave signal to obtain phase information of the fundamental wave signal; according to the phase information of the fundamental wave signal, calculating phase deviation information of the fundamental wave signal in pre-selected adjacent cyclic waves to obtain frequency estimation information of the power grid; according to the invention, narrow-band decomposition is carried out on an input voltage signal through a null-space tracking algorithm, so that a fundamental main mode can be accurately extracted; and phase deviation information can be determined based on discrete Fourier transform, so that when the interference frequency of the power grid is close to the fundamental frequency, high separation precision can be kept, and the accuracy of power grid frequency estimation can be improved.

A method of beamforming for downlink (DL) transmission in a wireless communication system is provided. The method includes generating, by a base station of the wireless communication system, configuration information required to signal channel state information (CSI) to the BS, wherein the configuration information includes parameters of a code book, wherein the code book defines an overall set of spatial beams in which data streams can be transmitted from the BS, wherein each spatial beam is represented by a discrete Fourier transform (DFT) vector in the code book, and a parameter (NDFT) indicative of a maximum number of DFT vectors with one signal polarization in a precoding matrix, transmitting, by the BS, the configuration information from the base station to a user device (UE), based on CSI reference signals (CSI-RSs) received from the BS, choosing, by the UE, a number (L) of Multiple Input-Multiple Output (MIMO) layers.

The invention relates to a power load prediction system based on time sequence self-supervised representation learning, and the system comprises a data preprocessing module, an expansion time convolution network, a trend representation decoupler, a seasonal representation decoupler, a feature converter, and a predictor. The expansion time convolution network performs dynamic adaptive grouping on different variables in the same data processing unit, establishes an intra-group variable relationship by sharing a convolution kernel weight and stacking a plurality of expansion time convolution layers, and establishes an inter-group variable relationship through subsequent multi-group representation splicing operation and single-layer expansion time convolution; the trend representation decoupler is used for separating trend representation Z (T) from feature representation extracted from the expansion time convolution network based on a plurality of parallel one-dimensional causal convolution blocks with different scales; the seasonal representation decoupler adopts discrete Fourier transform to separate seasonal representation Z (S) from the feature representation extracted from the expansion time convolutional network; the feature converter maps a combined feature vector of trend representation and seasonal representation back to an original space from a potential space by stacking a plurality of deconvolution layers; and the predictor maps the feature representation of the original space into a power load prediction result by using a linear projection layer, and the adaptability, robustness and generalization ability of the model for modeling complex power load time series data can be improved.

Embodiments of the present disclosure relate to methods and devices for wake-up signals. A method implemented at a network device comprises forming an OOK subsequence with multiple OOK symbols that comprise at least one OOK ON-symbol and at least one OOK OFF-symbol. The multiple OOK symbols are associated with one OFDM symbol. The method further comprises mapping the OOK subsequence into multiple pre-Discrete Fourier Transform (pre-DFT) subsequences respectively to form a pre-DFT sequence, and performing transform precoding on the pre-DFT sequence. The method further comprises mapping output of the transform precoding into frequency resources allocated to a Wake-up Signal (WUS) and generating the WUS, and transmitting to a terminal device the WUS.

The invention discloses an electrochemical impedance spectroscopy detection method and system and related equipment. The electrochemical impedance spectroscopy detection system comprises: a battery pack comprising a plurality of cells; the battery cell management module is connected with the plurality of battery cells and is used for acquiring response voltage of the battery pack and performing discrete Fourier transform on the excitation current and the response voltage to obtain an electrochemical impedance spectrum of the battery pack under the target excitation frequency; the battery management module is connected with the excitation source module and used for sending the target excitation frequency to the excitation source module; and the excitation source module is used for generating excitation current according to the target excitation frequency sent by the battery management module.

The invention discloses a weight function phase shift method for deformation field measurement, belongs to the technical field of optical measurement, is suitable for phase extraction and deformation field measurement of periodic structure images such as gratings and stripes, and has the characteristics of high speed and high precision. Weight function windows (a rectangular window, a triangular window, a cosine window and a Gaussian window) are applied in image processing, discrete Fourier transform is carried out after pixel gray values in the windows are multiplied by corresponding weights, and image phases at the center pixel positions of the windows are obtained. After the weight function window traverses all image pixels, the phase distribution of the whole image is obtained. And solving an image deformation field by using the phase difference of the images before and after deformation. Compared with the existing phase-based deformation field method, the method does not need complex image processing operation, and can solve the high-precision phase in one step and obtain full-field deformation only by using the weight function and the global pixel gray scale.

The invention discloses an online moment compensation and control method for an elastic joint, belongs to the technical field of robot joint precision control, and aims at performing online moment error compensation for the elastic joint based on adaptive filtering and a neural network and realizing accurate moment control based on compensated moment measurement information. Specifically, the method comprises the following steps: respectively collecting angle information of double encoders, carrying out noise reduction processing, converting a time domain signal of an angle into a frequency domain signal through discrete Fourier transform, and carrying out adaptive filtering based on a recursive least square algorithm (RLS); taking the obtained angle error information as input, performing adaptive prediction and compensation on a nonlinear error of an encoder angle based on an RBF neural network, solving the rigidity of an elastic element, and verifying the accuracy of torque identification through a torque sensor; finally, the solved torque value serves as reference input, an elastic joint torque control algorithm is designed, and high-precision torque control is achieved.

According to the Mama network solenoid valve fault diagnosis method based on the frequency domain characteristics, the problem of solenoid valve system fault diagnosis can be solved. According to the method, a pneumatic solenoid valve fault data set is constructed by collecting operation signals such as voltage and current, and multi-scale frequency domain features of the signals are extracted by adopting Wavelet Packet Transform (WPT) and Discrete Fourier Transform (DFT). WPT-DFT preprocessing features are embedded into an improved channel-space joint attention mechanism module, and the improved channel-space joint attention mechanism module is combined with a Mamba network with extremely high sequence modeling capability to construct an end-to-end fault diagnosis model. Compared with a traditional convolutional neural network, the method can more effectively capture deep dynamic features in a time sequence and highlight a key frequency region, and experimental results show that the method has higher diagnosis precision and generalization ability and is suitable for intelligent detection of the electromagnetic valve under complex working conditions.

The invention relates to the technical field of geophysical exploration, in particular to a multi-scale three-dimensional gravitational-seismic joint frequency domain inversion method based on hybrid constraints. The method comprises the following steps: determining observation gravity anomaly on the ground by adopting discrete Fourier transform, and discretizing the observation gravity anomaly by adopting a method of weighted average of values at a plurality of Gaussian points; calculating spherical harmonics and spherical harmonic coefficients of two-dimensional density distribution of each layer in the inversion range, and calculating gravity anomalies on the observation surface based on the spherical harmonics and the spherical harmonic coefficients; converting the inversion range data into a three-dimensional gravitational field forward kernel matrix; constructing a three-dimensional gravitational field hybrid constraint inversion objective function; inputting pre-collected earthquake three-dimensional velocity model data, and converting the earthquake three-dimensional velocity model data into a reference density model; and calculating a gradient operator in the inversion objective function, simplifying the inversion objective function based on the gradient operator obtained through calculation, and then solving the simplified inversion objective function by using an alternating direction multiplier method.