22 results about "Subspace algorithms" patented technology

The invention relates to the technical field of underwater acoustic positioning, discloses an OFDM (Orthogonal Frequency Division Multiplexing) sound source distance and depth estimation method based on a single hydrophone, and aims to realize parallel processing of communication positioning tasks and efficient utilization of underwater time-frequency resources and promote development of light-weight and low-energy-consumption equipment adaptive to an underwater unmanned small platform. According to the OFDM sound source distance and depth estimation method based on the single hydrophone, the frequency domain input and output relation of a communication system under a multi-path channel is combined, an observation model used for target positioning is constructed through comb-shaped pilot frequency, a multi-snapshot measurement matrix is constructed through a space domain smoothing means, multi-path time delay estimation is achieved based on a rotation invariant subspace algorithm, and the estimation accuracy is improved. And then target positioning is realized by adopting a channel time delay feature matching method. The OFDM sound source distance and depth estimation method based on the single hydrophone provided by the invention is suitable for integrated information processing of the single hydrophone, has the advantages of miniaturization, light weight and portability, and is suitable for an underwater unmanned small platform.

The invention discloses a two-dimensional arrival angle estimation method under fluid antenna energization, and the method comprises the steps: enabling two orthogonally distributed fluid antennas to respectively move along an x axis and a z axis at the same speed, moving for a fixed distance each time, and receiving K narrowband pilot signals which are not related to each other at each position; based on received data after the fluid antenna is moved for multiple times, covariance vectors of virtual aperture expansion in the x-axis direction and the z-axis direction are respectively constructed by jointly utilizing auto-covariance and cross-covariance information; based on covariance vectors expanded by two virtual apertures, rough estimation is carried out firstly and then fine estimation is carried out on an included angle between an incident central transmission path of a narrowband pilot signal and an x-axis forward direction and an included angle between the incident central transmission path of the narrowband pilot signal and a z-axis forward direction by using two subspace algorithms based on non-eigenvalue decomposition; and finally completing high-precision pairing of the two-dimensional arrival angles. The method has the advantage that high-performance unambiguous estimation of the two-dimensional angle of arrival can be completed with relatively low calculation cost and system overhead.

The invention relates to a direction-of-arrival estimation method based on a linear-nonlinear branch fusion neural network, and the method comprises the steps: obtaining an original signal of an antenna array, calculating a reference empirical covariance matrix based on the original signal, and obtaining an input tensor based on the reference empirical covariance matrix; based on the input tensor, processing by using a dual-path neural network to obtain a residual matrix; the dual-path neural network comprises a linear branch and a nonlinear branch; obtaining an enhanced covariance matrix based on the residual matrix and the reference empirical covariance matrix; and based on the enhanced covariance matrix, obtaining a DOA estimation value by using a Root-MUSIC algorithm. Compared with the prior art, the direction-of-arrival estimation method provided by the invention has the advantages of a traditional subspace algorithm and a neural network participation type method.

This invention belongs to the field of information and marine technology, and discloses a method for estimating the DOA (Direction of Aperture) of a UUV (Ultra-UV) hull-side array using conjugate enhancement and virtual extension. This method performs conjugate enhancement processing on the received signal from a parallel coprime array, calculates the augmented statistical matrix including autocorrelation, conjugate autocorrelation, cross-correlation, and conjugate cross-correlation, and constructs an extended correlation matrix through weighted combination. A virtual uniform linear array is constructed using the differential co-array characteristics of the coprime array. A redundant integration matrix is ​​designed to remove redundancy from the covariance matrix, obtaining the equivalent signal of the virtual array. Then, the covariance matrix of the virtual array with a Toeplitz structure is reconstructed. Finally, a dual rotation-invariant subspace algorithm is used to achieve high-precision, automatic pairing estimation of the underwater target's azimuth and pitch angles. This invention extends the array aperture, improves the estimation robustness in low signal-to-noise ratio environments, and eliminates the need for spectral peak search, providing technical support for UUV underwater target positioning.

This invention relates to a defect detection method for carbon fiber cables, comprising the following steps: S1. Constructing a simulation model of the carbon fiber composite cable and obtaining a sensitivity matrix; S2. Setting up a detection probe; S3. The detection probe acquires the detection voltage of the carbon fiber cable under test; S4. Calculating the initial grayscale value set of all pixels on the cross-section detected by the detection probe at the location of the carbon fiber cable under test based on the detection voltage, sensitivity matrix, and defect-free voltage; S5. Calculating the initial residual based on the results output from S1-S4, and updating the initial grayscale value set of all pixels on the cross-section of the carbon fiber cable under test using the Krylov subspace algorithm; S6. Updating the initial grayscale value set of all pixels on the cross-section of the carbon fiber cable under test based on the last iteration update and creating an image. When a defect is found in the image, the location detected by the detection probe is the location of the defect. This invention reconstructs the cross-sectional image using the Krylov subspace algorithm, making the cross-section of the carbon fiber cable clearer.

The present invention discloses an INCK-based Direction of Arrival (DOA) estimation method in an impulse noise environment. The method receives signals using a uniform linear array antenna to obtain received signal information #imgabs0#. Based on the received signal information, the local information entropy #imgabs1# for a single snapshot is calculated using a local entropy calculation formula. The infinite norm for a single snapshot is calculated using the infinite norm calculation formula. A semi-infinite norm Cauchy kernel and a full infinite norm Cauchy kernel are then constructed. The effects of the two Cauchy kernels are adjusted by the magnitude of the local information entropy #imgabs2#, and a weight #imgabs3# corresponding to each received signal information is established accordingly. #imgabs4# is weighted to obtain #imgabs5#, and a covariance matrix #imgabs7# is constructed for #imgabs6#. Finally, a rotationally invariant subspace algorithm is used to perform DOA estimation. The method uses the local information entropy #imgabs8# to adjust the effect of the infinite norm Cauchy kernel, suppress outliers in the received signal, and effectively estimate the DOA of the signal in the presence of alpha-stable distributed impulse noise.

The application discloses a real-time positioning method based on adaptive weight coefficient space. The specific steps are as follows: an array signal receiving model is established; a spatial smoothing method based on a virtual subarray is applied to the array signal receiving model; a weight coefficient space is established according to a signal covariance matrix; angle estimation and position solution are carried out based on a subspace algorithm; a position error function is established according to a real position result and an estimated position result; and an adaptive adjustment feedback mechanism of the weight coefficient space is established according to the position error function. The application has the beneficial effects that the method starts from the definition of the weight coefficient space, estimates the position error by setting the position error function, compares the position error with an error threshold, dynamically adjusts the weight coefficient space parameters, realizes the self-adaptation of the weight coefficient space parameters, and can improve the indoor positioning precision.

The invention belongs to the technical field of information and marine science and technology, and discloses a UUV broadside array DOA estimation method based on conjugate enhancement and virtual extension. According to the method, conjugate enhancement processing is carried out on parallel co-prime array receiving signals, augmented statistical matrixes including autocorrelation, conjugate autocorrelation, cross-correlation and conjugate cross-correlation are calculated, and weighted combination is carried out to construct an extended correlation matrix. The method comprises the following steps: constructing a virtual uniform linear array by using the differential common-array characteristic of a co-prime array, designing a redundancy integration matrix to perform redundancy elimination processing on a covariance matrix, obtaining an equivalent signal of a virtual array, and further reconstructing a virtual array covariance matrix with a Toeplitz structure, and finally, high-precision and automatic pairing estimation of the azimuth angle and the pitch angle of the underwater target is realized through a dual-rotation invariant subspace algorithm. According to the method, the array aperture is expanded, the estimation robustness in a low signal-to-noise ratio environment is improved, spectral peak search is not needed, and technical support is provided for UUV underwater target positioning.

The invention discloses a micro-grid linearization model order reduction method, system, equipment and medium, and belongs to the field of micro-grid stability analysis and control. Models of all elements in a micro-grid are obtained, a nonlinear differential algebraic equation set is established, a stable equilibrium point of the nonlinear differential algebraic equation set is solved, linearization processing is carried out, and a linearized differential algebraic equation set is obtained; solving a rightmost characteristic value by adopting a characteristic value algorithm, if the system is not stable, performing leftward translation transformation on the system to stabilize the system, generating a Lyapunov equation according to the translated system, solving a controllable and observable Gramb matrix by adopting a Krylov subspace algorithm, and performing order reduction on the translated system by adopting a balanced truncation method to obtain a stable system; according to the method, an existing balance truncation method is improved, the precision, the stability, the controllability and the observability of a system after order reduction are guaranteed, meanwhile, the Grubrum matrix can be efficiently solved, and the method is made to be suitable for an unstable system.

The present invention discloses a robust adaptive beamforming method. First, a linear array is divided into two sub-arrays, and the initial arrival directions of the desired signal and interference and the nominal steering vector are obtained by using the rotational invariant subspace algorithm. A projection matrix 1 is constructed to obtain the optimal steering vector of the desired signal. A projection matrix 2 is constructed to remove the desired signal component and estimate the power of the interference. The noise power is estimated by using the eigenvalue matrix corresponding to the noise subspace. Based on the nominal steering vector of the interference and the estimated power, the interference-plus-noise covariance matrix is reconstructed by using the zero broadening method of virtual interference. Based on the optimal steering vector of the desired signal and the interference-plus-noise covariance matrix, robust beamforming is performed on the array received data. The present invention can obtain a more accurate steering vector of the desired signal, and the reconstructed interference-plus-noise covariance matrix is used to calculate the weight vector, which has a stronger suppression effect on the interference and improves the robustness of the beamformer.

The present application relates to the technical field of inertia evaluation of power system based on electromechanical response characteristics, and particularly relates to a method for evaluating inertia of power system based on electromechanical response characteristics. The method comprises the following steps: extracting tie-line power and bus frequency data under steady-state operation of the power system by using a sliding data window method, and converting the data into equivalent electromechanical power and speed signals of the region; extracting electromechanical response characteristics, and extracting the electromechanical response characteristics by using a stochastic subspace algorithm; evaluating regional inertia, and evaluating the effective inertia of the system region according to the coupling relationship between the electromechanical response characteristics and the system inertia derived from the foregoing theory. The present application is mainly applied to power system stability analysis and control, has good recognition accuracy and practicability, and can effectively guide safe and stable operation of the power system according to the actual operation characteristics of the power system to obtain the equivalent inertia of the regional power grid.

The application relates to a linear-nonlinear branch fusion neural network-based direction of arrival estimation method, which comprises the following steps: obtaining original signals of an antenna array, calculating a reference empirical covariance matrix based on the original signals, and obtaining an input tensor based on the reference empirical covariance matrix; based on the input tensor, a residual matrix is obtained by using a double-path neural network; the double-path neural network comprises a linear branch and a nonlinear branch; an enhanced covariance matrix is obtained based on the residual matrix and the reference empirical covariance matrix; and a DOA estimation value is obtained by using a Root-MUSIC algorithm based on the enhanced covariance matrix. Compared with the prior art, the application provides a direction of arrival estimation method which can have the advantages of both a traditional subspace algorithm and a neural network participation type method.

The present invention relates to the field of underwater acoustic positioning technology, and discloses an OFDM sound source distance and depth estimation method based on a single hydrophone, aiming to achieve parallel processing of communication and positioning tasks and efficient utilization of underwater time-frequency resources, and to promote the development of lightweight and low-power consumption devices suitable for underwater unmanned small platforms. The OFDM sound source distance and depth estimation method based on a single hydrophone proposed by the present invention combines the frequency-domain input-output relationship of a communication system under a multipath channel, uses comb-shaped pilots to construct an observation model for target positioning, constructs a multi-snapshot measurement matrix through spatial smoothing means, realizes multipath delay estimation based on the rotational invariance subspace algorithm, and then uses the channel delay feature matching method to achieve target positioning. The OFDM sound source distance and depth estimation method based on a single hydrophone provided by the present invention is applicable to the integrated information processing of a single hydrophone, has the advantages of miniaturization, lightweight, and portability, and is suitable for underwater unmanned small platforms.

The invention discloses an asynchronous motor rotor fault detection method and device based on a random subspace algorithm, and belongs to the field of signal processing and motor fault detection. According to the method, the random subspace algorithm is improved, fundamental wave information in the stator current is eliminated, and meanwhile the influence of noise is avoided. The characteristic fault frequency is calculated for solving the system matrix, and the amplitude of the characteristic fault frequency is solved by using a least square method. According to the method, an improved random subspace algorithm is utilized, the influence of fundamental waves and noise is suppressed, stator current characteristic fault information is estimated more accurately, the limitation that the frequency resolution of a traditional method based on fast Fourier transform analysis is insufficient is broken through, and the purpose of accurately identifying the rotor broken bar fault and the rotor composite fault of the asynchronous motor is achieved. According to the method, the problem of inaccurate feature extraction caused by interference and noise superposition of a traditional algorithm is solved, and the frequency identification precision and the fault judgment reliability are improved.

The application discloses a modal identification method based on an unsupervised optimization covariance stochastic subspace method, and comprises the following steps: arranging vibration sensors on a to-be-measured structure with noise interference or weak excitation modes, and acquiring structural dynamic response data; calculating a covariance matrix and constructing a toply matrix; calculating an extended observable matrix and a controllable matrix based on the weighted toply matrix; defining a toply matrix row block number and a model order parameter range, and iteratively calculating a parameter optimization index; calculating a cumulative contribution rate based on a singular entropy increment of the toply matrix and determining a critical model order; calculating a cumulative parameter optimization index in an interval from a minimum model order to the critical model order under different toply matrix row block numbers, and determining an optimal parameter combination through an average minimum value; substituting the optimal parameter combination into a covariance stochastic subspace algorithm to identify a system matrix and calculate structural modal parameters; and automatically identifying each order physical mode from candidate modes by using a DBSCAN clustering method. The application further discloses a system and an electronic device.

The invention discloses a self-adaptive multi-user real-time respiration and heartbeat monitoring method and system based on millimeter wave radar. The method comprises the following steps: performing one-dimensional Fourier transform on radar echoes and removing static clutters; performing rough target detection of an adaptive threshold by using global statistical characteristics; dynamically updating a threshold value by using a noise intensity ratio of front and back frames, and performing target verification and accurate positioning through 2D-CFAR; using a micro-motion feature enhanced subspace MUSIC algorithm to measure an angle, and combining with specific frequency filtering to determine target space information; performing angle compensation on multi-channel data to obtain phase information, preliminarily separating signals by utilizing dual-channel heterogeneous wavelet packet transformation, and introducing a respiration frequency multiplication harmonic removal technology to eliminate interference, and reconstructing respiration and heartbeat signals with a high signal-to-noise ratio; and finally, carrying out feature tracking in combination with Kalman filtering and short-time Fourier transform. According to the invention, the problems of target positioning and interference of flexible switching under different conditions such as different distances between a single person and two persons, the same distance between two persons and multiple persons are effectively solved, and high-precision self-adaptive vital sign monitoring is realized.

The invention provides a user tag prediction method and device, relates to the field of artificial intelligence, and can be used in the financial field or other fields. The method comprises the steps of obtaining a behavior transaction information table authorized by a user, performing key information extraction on the behavior transaction information table to obtain multiple pieces of user feature information, and determining a feature missing rate corresponding to each piece of user feature information; comparing a preset missing rate threshold with the feature missing rate, performing data cleaning on the user feature information, and performing data recovery on the user feature information after data cleaning by using a sparse subspace algorithm; and processing the user feature information after data recovery by using a pre-established prediction model to obtain a user tag prediction result. According to the method, the characteristics and behaviors of the user can be described more accurately, the personalized requirements of clients are better met, the data accuracy is improved, the label division efficiency is improved, accurate division of user labels is realized, the user experience is improved, the resource waste is reduced, and the popularization efficiency is improved.

The invention discloses a polarization vector robust beam online optimization method and system of an antenna array. An antenna array model is obtained, deformation of the antenna array is caused by applying an external load, and a model of the antenna array after deformation is constructed; based on an improved feature subspace algorithm, constructing a steering vector estimation problem of a model after antenna array deformation, and further solving to obtain a steering vector after deformation; and solving an optimal beam weight under a polarization constraint condition according to the deformed steering vector through a second-order cone programming algorithm so as to realize actual polarization beam optimization of the antenna array. By integrating real-time sensing, high-precision estimation, dynamic polarization control and efficient optimization solution, a closed-loop online optimization system is formed, the beam robustness problem of a large-scale antenna array in a dynamic environment is effectively solved, and the overall performance, environmental adaptability and engineering practical value of the system are remarkably improved.

The application discloses a two-dimensional angle of arrival estimation method under fluid antenna energization, which comprises the following steps: moving two orthogonally distributed fluid antennas along the x-axis and the z-axis at the same speed, respectively, moving a fixed distance each time, and receiving K narrowband pilot signals irrelevant to each other at each position; based on the received data after the fluid antenna moves multiple times, jointly using the autocovariance and cross-covariance information, respectively constructing the virtual aperture extended covariance vectors in the x-axis and z-axis directions; based on the two virtual aperture extended covariance vectors, using a two-step non-eigenvalue decomposition based subspace algorithm to coarsely estimate and then finely estimate the angle between the central transmission path of the narrowband pilot signal incidence and the x-axis positive direction and the angle between the central transmission path of the narrowband pilot signal incidence and the z-axis positive direction; and finally completing high-precision pairing of the two-dimensional angle of arrival; the method has the advantages that high-performance non-ambiguous estimation of the two-dimensional angle of arrival can be realized at a lower calculation cost and system overhead.

The present invention provides a single base station positioning method and system based on joint estimation of arrival angle and arrival time, and the positioning method comprises: step S1, using 3D-MUSIC algorithm to make a rough estimate of the received signal, and obtaining a rough estimated azimuth, elevation angle and delay; step S2, further refining the rough estimation result obtained in step S1, using the finely estimated three-dimensional angle and delay calculation formula to estimate the azimuth offset, elevation angle offset and delay offset, and further obtaining an updated azimuth, an updated elevation angle and an updated delay; repeating this step, and iterating repeatedly until complete convergence; step S3, outputting the azimuth, elevation angle and delay obtained after the iteration is completed. The accuracy of the technical solution of the present invention is significantly better than the accuracy of other types of subspace algorithms at present, and compared with the latest subspace algorithm for rough and fine estimation, the accuracy is improved while the complexity is greatly reduced.

The invention relates to a defect detection method of a carbon fiber inhaul cable, which comprises the following steps: S1, constructing a simulation model of a carbon fiber composite inhaul cable to obtain a sensitivity matrix; s2, arranging a detection probe; s3, the detection probe collects the detection voltage of the detected carbon fiber inhaul cable; s4, according to the detection voltage, the sensitivity matrix and the defect-free voltage, calculating an initial gray value set of all pixels of a section detected by the detection probe at the position of the detected carbon fiber inhaul cable; s5, calculating an initial residual error according to results output from S1 to S4, and updating an initial gray value set of all pixels of the section of the tested carbon fiber inhaul cable by using the initial residual error through a Krylov subspace algorithm; and S6, obtaining an initial gray value set of all pixels of the cross section of the detected carbon fiber inhaul cable according to the last iteration update, imaging, and determining the position detected by the detection probe as the position of the defect when the imaging result has the defect. According to the method, the cross section image is reconstructed through the Krylov subspace algorithm, so that the cross section of the carbon fiber inhaul cable is clearer.

The invention discloses a DOA estimation method based on INCK in an impulse noise environment. The method comprises the following steps: receiving a signal through a uniform linear array antenna to obtain received signal information # imgabs0 #; the method comprises the following steps: calculating a local information entropy # imgabs1 # under single snapshot by a local entropy calculation formula according to received signal information, calculating an infinite norm under single snapshot by combining an infinite norm calculation formula, and constructing a semi-infinite norm Cauchy kernel and a full infinite norm Cauchy kernel; adjusting the action size of the two Cauchy kernels according to the size of the local information entropy # imgabs2 #, and establishing a weight # imgabs3 # corresponding to each piece of received signal information according to the action size of the two Cauchy kernels; and performing weighting processing on the # imgabs4 # to obtain a # imgabs5 #, constructing a covariance matrix # imgabs7 # for the # imgabs6 #, and performing DOA estimation by adopting a rotation invariant subspace algorithm. According to the method, the effect of the infinite norm Cauchy kernel is adjusted through the local information entropy # imgabs8 #, the abnormal value in the received signal is suppressed, and effective estimation of the signal DOA under the alpha stable distribution impulse noise is realized.