34 results about "Linear least squares" patented technology

The invention discloses a physical field solving method based on a Bayesian physical information extreme learning machine, and the method comprises the steps: constructing a single-layer full-connection neural network, carrying out the random initialization, and fixing the weight of an input layer; based on a partial differential equation of a physical system and boundary conditions thereof, defining a training loss item containing physical information; a physical system solving problem is converted into a linear least square problem, and a linear equation set is constructed; supposing that an output layer weight parameter obeys Gaussian prior distribution with the mean value being zero, and controlling a covariance matrix by an adjustable hyper-parameter; constructing a Gaussian likelihood function based on the observation data, and calculating posterior distribution of the output weight in combination with the prior distribution; carrying out iterative optimization on the hyper-parameter by adopting an evidence maximization method to obtain a mean value and a covariance of posterior distribution; based on posterior distribution, adopting a Monte Carlo integral method to generate prediction output of the physical system; and performing uncertainty quantization based on the variance of prediction output, and outputting a prediction value and a confidence interval thereof.

The invention relates to an electric vehicle wireless charging system LPV-Hammerstein model identification method based on an alternating least square iteration strategy, and belongs to the technical field of power electronic system modeling and parameter identification. The method comprises the following steps: firstly, constructing an LPV-Hammerstein model structure based on a WPT system circuit topology and a working mechanism, and then collecting discrete data of a system input control signal U, an output DC voltage Vo and a scheduling variable Req; then, alternately fixing other parameters by adopting an alternate least square iterative algorithm, and converting a to-be-estimated parameter problem into a linear least square sub-problem to be solved; and when a convergence condition is satisfied, a final LPV-Hammerstein model parameter estimation result is obtained. According to the method, the common nonlinear and parameter change coupling characteristics in the wireless charging system of the electric vehicle can be effectively processed, a systematized and efficient way is provided for obtaining a high-precision system dynamic model, and the limitation that the system cannot be accurately described by a traditional linear model or a simple nonlinear model is overcome.

The invention relates to an AUV (Autonomous Underwater Vehicle) cooperative positioning method based on multi-ranging circle intersection. The invention relates to the technical field of underwater AUV (Autonomous Underwater Vehicle) cooperative positioning. The method comprises the following steps: establishing a kinematic model and a measurement model of an AUV; obtaining a distance measurement circle equation at the current moment and current control input, and jointly resolving an equation set constructed by a historical distance measurement circle; removing wrong candidate points; and determining error covariance matrixes of all candidate solutions and cross covariance matrixes among the candidate solutions to obtain a joint covariance matrix of all position estimations, and performing optimal weighted fusion on the candidate solutions by using linear least squares to obtain optimal position estimation and corresponding estimation error covariance. According to the method, a plurality of potential position estimation values are creatively generated through combined solution of historical distance measurement equations, fuzzy solutions are systematically removed according to the shape of an uncertain region, an accurate inner point set is obtained, and the method has high state estimation precision and low time complexity while estimation consistency is guaranteed.

The invention relates to a robust elliptical target positioning method and device, communication equipment and a storage medium, and the method comprises the steps: taking a to-be-detected target under target pulse noise as a positioning object, and carrying out the positioning of a to-be-detected target according to a minimum deviation criterion, a pseudo-linear technology, multistage weighted least square estimation, and a bistatic distance measurement equation of a distributed multiple-input-multiple-output radar positioning system; respectively constructing an iterative reweighted pseudo-linear least square estimator and a two-stage iterative reweighted least square estimator; and carrying out positioning processing on a to-be-detected target according to the iterative reweighted pseudo-linear least square estimator and the two-stage iterative reweighted least square estimator to obtain a target position estimation result. According to the invention, through cascading the iterative reweighted pseudo-linear least square estimator and the two-stage iterative reweighted least square estimator, approximately unbiased target position estimation can be derived, and the positioning precision of a to-be-detected target is effectively improved.

This application discloses a deconvolution method for gas chromatography-mass spectrometry (GC-MS) data, relating to the field of GC-MS data analysis technology. This method utilizes a non-negative matrix factorization function to initially decompose the time-mass-charge ratio two-dimensional intensity matrix, obtaining an initial elution curve matrix and an initial mass spectrometry matrix. An exponentially modified Gaussian peak shape modeling function is then used to fit the initial peak shape parameter set. For each component: based on the initial peak shape parameter set, an exponentially modified Gaussian peak shape modeling function is used to generate the original elution curve matrix. Non-negative least squares optimization and non-linear least squares optimization methods are used to iteratively optimize the mass spectrometry matrix, peak shape parameter set, and elution curve matrix. Finally, a comprehensive scoring function is used to select the optimal component set, and the optimal peak shape parameter set, mass spectrometry matrix, and elution curve matrix for the optimal component set are sorted and output. This application enables precise separation of co-eluted components, improving the accuracy and reliability of complex mixture analysis.

The application discloses a kind of based on track keeping's unmanned aerial vehicle thermal infrared image splicing method and system, it is related to image splicing technical field, including steps: constructing image set;Obtain the matching feature point of any pair of to-be-spliced unmanned aerial vehicle's thermal infrared image respectively, and the re-projection error of matching feature point is generated alignment item by the similar transformation matrix of to-be-spliced thermal infrared image;The actual distance between the center of reference image and the thermal infrared image of any unmanned aerial vehicle is calculated, and the track trajectory protection item is obtained between the center point of similar transformation matrix and reference image by pixel distance;Alignment item and track trajectory protection item are weighted and summed to construct global optimization objective function, and global optimization objective function is converted into linear least square model with similar transformation constraint to solve, and generate spliced thermal infrared panoramic image.The application minimizes the cumulative error of splicing process, while ensuring the integrity of global structure.

The present invention discloses a method for online identification of human driver style, comprising the following steps: Step 1: constructing an online identification model of human driving behavior using the modeling approach of inverse optimal control; Step 2: simplifying the online identification model of human driving behavior based on the Pontryagin minimum principle; Step 3: transforming the model into a linear least squares problem based on the simplified model result; Step 4: solving the model based on the model transformation using the Lagrange multiplier method to obtain driving behavior parameters that describe the human driver's style preferences, thereby identifying the human driver's driving style. The method of the present invention can accurately identify the driving style of human drivers of environmental vehicles, thereby effectively improving the responsiveness and adaptability of the autonomous driving system, and enhancing driving safety and comfort.

The invention relates to the technical field of electric power operation and maintenance, in particular to a harmonic source modeling method and device considering background harmonic fluctuation and a medium, and the method comprises the steps: collecting voltage and current data of load outer ports of various harmonic sources; establishing a load equivalence time domain circuit model, and solving load equivalence impedance parameters by using linear least squares under a non-negative condition; and constructing a harmonic source load model, and dynamically adjusting an impedance parameter in real time according to the background harmonic fluctuation by using a recursive least square method to obtain a harmonic source model adapted to the background harmonic fluctuation. Compared with the prior art, the constructed harmonic source load model can accurately represent harmonic characteristics changing along with background harmonic fluctuation, and the adaptability and accuracy of the model in a dynamic environment are improved.

The invention discloses a target harmonic signal complex frequency estimation method based on frequency domain least square, and belongs to the technical field of signal processing, and the method comprises the steps: obtaining a geophysical observation time sequence containing a target harmonic signal; windowing Fourier transform is adopted to obtain an initial power spectrum, and a spectrum peak of the target harmonic signal is identified; selecting power spectrum values at a plurality of frequency points near the target spectrum peak, and obtaining complex frequency parameters of the target harmonic signal by using linear least square fitting; zero-filling expansion is adopted to increase effective frequency points of a power spectrum, and the estimation precision of complex frequency parameters is improved through multiple least squares. According to the method, Fourier transform and linear least square fitting are combined, high-precision estimation of the complex frequency of the target harmonic signal can be achieved, the method is simple, efficient and easy to implement, and the method is successfully applied to scenes such as earth normal mode frequency extraction and ocean partial tide parameter inversion and can be expanded to any time sequence signal analysis such as communication and biomedicine.

The method comprises the following steps of: decomposing a measurement matrix of each strapdown triaxial magnetometer into a unit orthogonal matrix and a symmetric positive definite matrix by utilizing the characteristic that a magnetic field modulus is independent of the unit orthogonal matrix and adopting a polar decomposition technology, and obtaining an equivalent zero offset and the symmetric positive definite matrix of each strapdown triaxial magnetometer by adopting ellipsoid equation iteration linear least square fitting; a virtual coordinate system is introduced, a vector modulus value correction value of measurement data of each strapdown triaxial magnetometer is calculated, a correction value of a magnetic field horizontal modulus gradient is further obtained, and calibration of the strapdown triaxial cross magnetic array is completed. According to the method, the ellipsoid equation is solved with high precision by adopting the iterative linear least square method, the strapdown three-axis cross magnetic array does not need to be accurately rotated, high-precision calibration can be carried out under the condition of providing the geomagnetic field modulus, and the operation is simple and convenient; the iterative linear least square algorithm is simple, the calculation amount is small, and engineering implementation is easy.

The invention discloses a deception jamming method and device for an azimuth phase coding SAR (synthetic aperture radar), and belongs to the technical field of electronic support system synthetic aperture radar countermeasure. The method comprises the following steps: intercepting an SAR signal, carrying out two-dimensional structured phase stripping by constructing an inverse distance modulation operator and an inverse azimuth modulation operator, and estimating to obtain an azimuth phase sequence; carrying out complex domain differential operation on the estimated azimuth phase sequence, constructing a phase consistency measurement index and minimum consistency order mapping, and identifying a coding structure of the phase consistency measurement index and the minimum consistency order mapping; constructing a linear least square problem based on an identification result, and reconstructing an azimuth phase sequence; and performing phase consistent modulation including convolution modulation on the intercepted signal to generate a deception jamming signal. According to the method, complete matching of the interference signal and the SAR receiving filter is realized, and effective deception interference can be implemented on a new system SAR adopting random, primary, secondary and grouped phase coding.

The application discloses a kind of pure angle target trajectory estimation methods, this method will typical target motion model be converted into the autoregressive moving average model of position component to realize state decoupling, by adjusting the relationship between pseudo-linear least square cost function and process noise energy to obtain regularized pseudo-linear least square model.To reduce the estimation deviation generated by the correlation between pseudo-linear measurement matrix and noise, the regularized instrumental variable least square position estimation is obtained by using instrumental variable estimation method.The method of the application obtains the analytical expression of batch estimation of azimuth position component in the framework of regularized least square estimation.The method has linear time calculation complexity, and its estimation performance is irrelevant to initial state.

The invention discloses a federal causal portrait and teaching intervention optimization method for a multi-school online education scene. The method comprises the following steps: constructing a local sample set and a local variable matrix; correlation calculation is carried out on education variables in the local variable matrix, and a local candidate neighborhood matrix is constructed; constructing a local candidate father node set; constructing a father variable set and a target variable, and performing linear least square fitting on the relationship between the father variable set and the target variable; obtaining a first residual dependency degree between the residual and the parent variable matrix; constructing a local causal adjacency matrix according to the local direction score; obtaining a global causal direction score by adopting federal causal structure weighted aggregation; constructing a global causal skeleton matrix; mining a causal path set pointing to the learning result node from a teaching related node set, and obtaining path causal intensity; calculating an overall causal influence score of learning result nodes of the teaching action set; and sorting according to the overall causal influence scores to obtain an intervention strategy.

This invention relates to the field of optical inspection technology, and in particular provides a method and apparatus for sub-aperture stitching inspection of large-aperture optical elements. The method includes the following steps: S1, setting up an inspection system, including an electrically controlled rotary table, a display screen, and a camera; S2, establishing a mirror coordinate system with the center of the mirror surface under test as the origin; S3, rotating the mirror under test and transforming the coordinates of sampling points within each sub-aperture to the world coordinate system; S4, ray tracing to calculate the local slope data of the sampling points; S5, uniformly transforming the slope data of each sub-aperture to the mirror coordinate system, and constructing a full-aperture slope field based on a slope stitching fusion method; S6, constructing an overdetermined linear least squares equation system based on Zernike polynomials to obtain the full-aperture reconstructed surface shape of the mirror surface under test. This invention acquires sub-aperture data through rotational scanning and uses a slope stitching fusion method to achieve full-aperture surface shape reconstruction and high-precision complete inspection of the surface of large-aperture optical elements.

The invention belongs to the technical field of magnetic measurement, and particularly relates to a calibration method of a magnetic dipole array. The method comprises the following steps: S1, obtaining an independent calibration model of each magnetic source unit; and S2, according to the independent calibration model, measuring and modeling an additional magnetic field generated by coupling between the magnetic source units, and establishing a compensation model. According to the method, the problem is decomposed into N low-dimensional nonlinear fitting and one linear least square problem, the calculation complexity and time consumption are reduced in order of magnitude, and high efficiency in engineering application is achieved.

The present invention relates to the field of image data processing or generation technology, and in particular to a method and system for chromatic aberration correction in single-molecule localization super-resolution imaging. The method comprises the following steps: eliminating misaligned points from multiple sets of lateral motion images of two different wavelength channels; performing registration between the retained points to obtain a chromatic aberration correction mapping function, thereby completing lateral chromatic aberration correction in single-molecule localization super-resolution imaging; performing Gaussian fitting and linear least squares fitting on multiple sets of axial height data for each wavelength channel to obtain the average deviation of the two channels at different axial heights, which is then compensated into the single-molecule localization super-resolution imaging chromatic aberration correction system to complete axial chromatic aberration correction in single-molecule localization super-resolution imaging. The method and system provided by the present invention achieve correction of lateral and axial chromatic aberration in multi-color single-molecule localization super-resolution imaging, improve the system's registration accuracy and correction speed, and reduce the cost of correction.

The invention discloses a laser spectrum adaptive algorithm for suppressing interference noise and improving time resolution, and belongs to the technical field of laser spectrum and sensors. The algorithm aims at solving the problems that in an existing laser spectrum technology, the interference noise suppression effect is limited, operation is complex, and the time resolution is low. According to the technical scheme, the method comprises two stages of spectral signal noise reduction and concentration inversion after spectral signal noise reduction, wherein in the spectral signal noise reduction stage, effective signal frequency is extracted through fast Fourier transform, and noise suppression is achieved in combination with linear least square fitting, regularization and residual autocorrelation analysis; in the concentration inversion stage, noise reduction signals are input into a Lorentz function model, parameter search is optimized by combining a particle swarm optimization algorithm with a prediction guide item of Kalman filtering, and the medium concentration is rapidly inverted. Through multi-algorithm cooperation, interference noise is effectively suppressed, the system time resolution is improved, operation is easy and convenient, adaptability is high, and the method can be applied to the fields of atmospheric monitoring, industrial control and the like.

A method is provided for using Light-Field Dynamic Homography (DH) to generate a large virtual focal plane from a non-rigid camera array with narrow overlaps between their fields of view (FOV). The method incorporates the 3D geometry of the cameras and employs non-linear least square optimization to dynamically estimate the inter-view homography transformations. Remarkably, only two feature correspondences are required between adjacent views to stitch the images and generate a wide virtual focal plane array, eliminating the need for significant FOV overlaps between the multiple cameras.

The invention discloses a surface shape anti-vibration measurement method for a large-aperture high-resolution phase-shifting interferometer. According to the method, vibration offset iterative estimation and high-resolution wavefront reconstruction are decoupled, and anti-vibration phase reconstruction is completed only by using a single group of high-resolution interference images collected in a vibration environment. The method comprises the following steps: acquiring a high-resolution interference image in a vibration environment and carrying out noise suppression and contrast compensation preprocessing; down-sampling the preprocessed image, and iteratively estimating the vibration offset of each frame of interference image in a low-resolution domain; and taking the vibration offset as a known parameter, carrying out linear least square fitting on the full-resolution pre-processed image, and reconstructing the measured wavefront phase distribution. According to the method, the calculation time is shortened by 8-10 times while the phase reconstruction precision is ensured, low-resolution data are not required to be collected, hardware is not required to be changed, and the method is suitable for the phase-shifting interferometer with the caliber being greater than or equal to phi 100mm and the resolution being greater than or equal to 1000 * 1000 pixels.

The invention discloses a porous medium seepage field rapid prediction method based on principal component analysis and a random neural network. Firstly, principal component analysis dimensionality reduction is performed on a permeability field and a pressure field, and an encoder and a decoder are constructed; then, a random neural network is constructed, the hidden layer weight and bias of the random neural network are fixed after being initialized, and the output layer weight is trained only by solving the linear least square problem so as to learn the potential space mapping relation; preferably, the prediction stability and precision can be improved by integrating a plurality of networks; furthermore, a prediction interval with statistical guarantee can be provided for the integrated model by using a conformal prediction method, and the prediction uncertainty is quantified. According to the method, the training time is shortened to the second level, the calculation cost is remarkably reduced while high prediction precision is kept, high robustness is achieved, and an efficient tool is provided for real-time simulation and rapid decision making of oil and gas exploitation, underground water simulation and other scenes.

This invention discloses a cooperative localization method for aircraft in ad hoc networks, relating to the field of aircraft cooperative localization. It utilizes the relative distance and angle measurements between aircraft within a two-hop link and the positions of observable anchor nodes within that two-hop link to construct a multi-hop link joint observation. A weight matrix is ​​constructed by combining the influence of ranging and angle measurement errors on the localization error. Cooperative localization of the aircraft in the ad hoc network is then achieved based on a weighted linear least squares algorithm. Compared to conventional cooperative localization methods, this method does not require aggregating all link measurement results to a central computing node for position calculation. It only needs to collect measurement results within a two-hop link to autonomously calculate the position of its own nodes, resulting in low communication overhead and low localization latency between aircraft.

The invention provides a joint timing synchronization and optical wireless positioning method based on continuous linear least square, and relates to the technical field of optical wireless positioning. An indoor optical wireless positioning system comprising a plurality of LED optical signal transmitting ends, a photodiode receiving end and user equipment located at the receiving end is constructed, an optical wireless communication signal waveform is used as measurement data of optical wireless positioning, and clock phase offset and clock frequency offset are used as time parameters; a time synchronization model is constructed by combining user equipment position parameters and user equipment direction parameters, then a time synchronization and positioning objective function is constructed based on measurement data of optical wireless positioning and the time synchronization model, and finally, the objective function is solved based on continuous linear least square and iterative update alternate optimization. The timing synchronization and optical wireless positioning joint estimation is realized, the optimal user equipment direction parameter, time parameter and user equipment position parameter are obtained, and the positioning precision is improved.

The present application relates to a soil moisture inversion method based on neighborhood change detection and passive microwave constraints, wherein the method includes: obtaining a synthetic aperture radar (SAR) image covering the target study area and preprocessing the SAR image; obtaining neighborhood homogeneous pixels of the central pixel using pseudo-scattering type parameters and pseudo-scattering entropy parameters; establishing a neighborhood change detection model based on scattering characteristic similarity weighting based on the co-polarization backscattering observations of the central pixel and the neighborhood homogeneous pixels; using the upper and lower bounds corresponding to the long-time series passive microwave soil moisture product as the constraint range, performing a linear least squares fitting solution on the soil moisture within the constraint range to obtain a soil moisture inversion result. The embodiment of the present application reduces the uncertainty and data acquisition cost brought by the temporal variability of vegetation to soil moisture inversion, and solves the problem in the related art that the soil moisture temporal change detection method is affected by the temporal variability of vegetation and has low performance in actual application.

According to the method, a polar decomposition technology is adopted, a measurement matrix of each strapdown triaxial magnetometer is decomposed into a unit orthogonal matrix and a symmetric positive definite matrix, and an ellipsoid equation is adopted for iterative linear least square fitting to obtain equivalent zero offset and the symmetric positive definite matrix of each strapdown triaxial magnetometer; and introducing a virtual coordinate system, calculating a measured data correction value of each strapdown three-axis magnetometer by using orthogonal Prusk decomposition, and further calculating a correction value of a magnetic gradient tensor measured by the strapdown three-axis cross-shaped magnetic array, thereby completing calibration of the strapdown three-axis cross-shaped magnetic array. According to the method, the iterative linear least squares method is adopted, the ellipsoid equation can be solved with high precision, the calibration precision is remarkably improved, the direction of the rotating shaft does not need to be kept stable in the rotating process, and the iterative linear least squares algorithm and the orthogonal Prushing analysis algorithm are small in calculated amount and easy to achieve in engineering. Under the condition that the geomagnetic field model and the magnetic array attitude are provided, calibration can be completed quickly, simply and conveniently, and the method is suitable for practical engineering application.

Described herein is NLS using a bounded linear initial search space and a fixed grid with pre-calculated variables. An initial first elevation angle corresponding to a direct reflection from an object and an initial second elevation angle corresponding to a multi-path reflection from the object are received along with a grid comprising pre-calculated variables for each of a plurality of grid points. The initial elevation angle pair is then refined by, for a plurality of iterations, determining a closest grid point to a current elevation angle pair and performing a non-linear least squares iteration on the closest grid point using the pre-calculated variables for the closest grid point to determine a next elevation angle pair. When the iterations have converged, a height of the object is calculated based on a final elevation angle pair.

The invention discloses an unmanned aerial vehicle thermal infrared image stitching method and system based on flight path keeping, and relates to the technical field of image stitching, and the method comprises the steps: constructing an image set; obtaining respective matched feature points of any one thermal infrared image of the to-be-spliced unmanned aerial vehicle, and generating an alignment item through a similar transformation matrix of the to-be-spliced thermal infrared image and a re-projection error of the matched feature points; the actual distance between the thermal infrared image of any unmanned aerial vehicle and the center of the reference image is calculated, and a route track protection item is obtained through the pixel distance between the center point of the similar transformation matrix and the reference image; and constructing a global optimization objective function through weighted summation of the alignment item and the route track protection item, and converting the global optimization objective function into a linear least square model constrained by similar transformation for solving to generate a spliced thermal infrared panoramic image. According to the method, accumulated errors in the splicing process are reduced to the maximum extent, and meanwhile the integrity of a global structure is guaranteed.

The invention relates to a CO2 concentration inversion method based on intelligent optimization cubic smooth spline fitting, electronic equipment and a storage medium. The method comprises the following steps: S1, preprocessing radar echo signals; s2, performing division and logarithm processing on the preprocessed radar echo signal, and calculating a differential absorption optical thickness; s3, performing inversion by using linear least square fitting to obtain a reference area and a reference concentration value; s4, using an intelligent optimization algorithm to find an optimal smoothing parameter of a cubic smoothing spline fitting algorithm; and S5, substituting the optimal smoothing parameter into a cubic smoothing spline fitting algorithm to calculate a concentration value. According to the invention, CO2 concentration inversion of laser radar echo signals in long-distance transmission and the like can be realized, and the accuracy and reliability of an inversion result are effectively improved.

The application discloses a target harmonic signal complex frequency estimation method based on a frequency domain least square, and belongs to the technical field of signal processing, and comprises the following steps: acquiring a geophysical observation time sequence containing a target harmonic signal; acquiring an initial power spectrum by adopting a windowed Fourier transform, and identifying a spectral peak of the target harmonic signal; selecting power spectrum values at multiple frequency points near the target spectral peak, and obtaining a complex frequency parameter of the target harmonic signal by utilizing a linear least square fitting; and adopting zero padding extension to increase effective frequency points of the power spectrum, and improving estimation accuracy of the complex frequency parameter by a multiple least square. The application combines the Fourier transform and the linear least square fitting, can realize high-precision estimation of the complex frequency of the target harmonic signal, is simple, efficient and easy to realize, has been successfully applied to scenes such as extraction of earth normal mode frequencies and inversion of ocean tide constituent parameters, and can be extended to communication, biomedical and other arbitrary time series signal analysis.

This invention relates to the field of indoor positioning technology, specifically to a UWB and IMU tightly coupled positioning method and system based on sliding window and motion constraints, comprising the following steps: S1, deploying UWB base stations in the space to be measured, installing UWB and IMU units on a ground robot, acquiring distance observation values ​​through bilateral bidirectional ranging, and establishing a set of distance equations in conjunction with the base station coordinates, and solving for the initial position using linear least squares; S2, initializing the initial position input error state Kalman filter, and predicting the position, velocity, and attitude based on IMU data; S3, constructing an innovation vector based on the UWB observation vector and the predicted state, updating the observation noise covariance through a sliding window, and fusing UWB and inertial information; S4, constructing a Z-axis pseudo-zero velocity observation model and configuring the observation noise covariance, updating the error state through the input filter, and correcting the nominal state. This invention achieves the collaborative fusion of UWB and IMU information, thereby improving the accuracy and stability of ground robot positioning.

The invention relates to a method and a device for detecting the service life of a transformer respirator. The method comprises the following steps of: 1, constructing a spectrum water vapor associated data set; 2, a linear least square fitting algorithm is adopted to evolve the spectrum water vapor model; 3, constructing a respirator service life diagnosis model; the device has the advantages of reasonable structure, convenience, rapidness and high accuracy.