90 results about "QR decomposition" patented technology

The invention discloses a tire longitudinal force and lateral force prediction method based on vehicle state parameter estimation, which adopts a seven-order spherical radius criterion to replace the original three-order spherical radial volume criterion on the basis of the traditional volume Kalman filter (CKF), and improves the estimation precision by optimizing sampling point distribution and weight configuration. And the stability of numerical calculation is ensured in combination with QR decomposition. A vehicle-mounted sensor is used for collecting real-time data such as a front wheel turning angle and longitudinal / lateral acceleration under a complex working condition, a time sequence feature set is built by combining longitudinal speed, lateral speed, a side slip angle and yaw velocity estimated by an algorithm, and meanwhile tire force calculated by a Dugoff tire model serves as target output. The TCN-GRU-based tire force prediction method is designed, the calculation complexity can be reduced while key time sequence information is kept, and the real-time prediction performance of the tire force is improved.

The invention provides an image restoration method based on adaptive weighted tensor completion, and relates to the technical field of image processing and application, and the method comprises the steps: obtaining to-be-restored image data, and carrying out the tensor of the to-be-restored image data, and obtaining input tensor data; constructing a tensor completion model based on an adaptive weighted tensor nuclear norm; wherein a weight matrix in the tensor completion model can be adaptively updated along with input tensor data; and based on an alternating direction multiplier method or an approximate singular value decomposition method based on tensor QR decomposition, solving the tensor completion model, and outputting restored tensor data to realize image restoration. According to the scheme, the image restoration quality can be improved.

The invention relates to the technical field of telephone communication, and discloses a robust recursive least square adaptive filter based on inverse QR decomposition and a storage medium, and the method comprises the steps: calculating an inner product of an adaptive weight vector at a moment and an input signal vector, and obtaining an output signal at the moment of the adaptive filter; calculating a difference value between the expected signal and the output signal at the moment, obtaining an estimation error signal, and constructing a nonlinear weighting function; recursively updating and acquiring an inverse correlation matrix of the moment input signal vector based on the moment input signal vector and transpose thereof, a nonlinear weighting function, the inverse correlation matrix of the input signal vector and the covariance of the estimated error signal, and constructing a block matrix containing the inverse correlation matrix of the moment input signal vector; inverse QR matrix decomposition and unitary rotation operation are sequentially carried out on the block matrix, an array before inverse QR decomposition and an array after inverse QR decomposition at the moment are obtained, a gain vector at the moment is constructed, and a self-adaptive weight vector at the moment is updated and obtained.

The invention provides a method, medium and system for estimating the orientation of a horizontal array in an uncertain shallow sea environment, and belongs to the technical field of geophysics, and the method comprises the steps: firstly, receiving a far-field narrow-band signal through a horizontal array, obtaining a time domain sampling snapshot, then calculating an array sampling covariance matrix, and constructing a signal subspace through the first J feature vectors; meanwhile, according to a standard shallow sea sound velocity profile and a parameter uncertainty range, a horizontal wave number range of a propagation mode is obtained, sampling is carried out in the range to construct a mode matrix, and feature vectors corresponding to the first K maximum feature values are extracted to construct a mode subspace. Next, constructing an expansion matrix by using base vectors in the signal subspace and the modal subspace, performing QR decomposition to obtain an upper triangular matrix, and extracting diagonal elements of the upper triangular matrix to construct a sound source orientation estimation function; and finally, the angle corresponding to the peak value of the sound source orientation estimation function is the sound source orientation estimation value.

The invention discloses an interpolation curved surface generation method based on a five-point difference format, and relates to the technical field of three-dimensional modeling and curved surface generation, four NURBS curves are input, a parameter domain Omega is equally divided along a u direction and a v direction, and (n + 1) * (m + 1) grid points are obtained through discretization on the parameter domain Omega, and each grid point is a control point {Pi, ji = 0, 1, 2,..., n, j = 0, 1, 2,..., m} of an interpolation curved surface; wherein P0, j, Pn, j, Pi, 0, Pi, m are control points on the four NURBS curves, namely known control points; (Pi, ji = 1, 2,..., n-1, j = 1, 2,..., m-1) is an internal control point of the interpolation curved surface, namely an unknown control point; a five-point difference format is used for representing an internal control point and recording the internal control point as a vector P, solving of the internal control point is converted into solving AP = b, QR decomposition is carried out on a matrix A to make A = QR, solving of the internal control point is converted into solving RP = QTb, then solving is carried out to obtain the vector P, and an interpolation curved surface S (u, v) can be generated according to the control point. The invention provides a curved surface modeling method for interpolating a given boundary NURBS (Non-Uniform Rational B-Spline) curve based on a partial differential equation five-point difference format. The curved surface modeling method can be adaptive to the condition that the given boundary curve is complex to generate a regular curved surface interpolated to the boundary curve.

In order to solve the problems of high calculation complexity and low efficiency in hydraulic fracturing numerical simulation, the invention provides a method for parallel simulation of hydraulic fracturing crack propagation. A three-dimensional fracture grid is constructed by adopting a displacement discontinuity method (DDM) and a finite volume method (FVM), and stress and fluid pressure distribution in fractures are calculated. Crack expansion is driven by optimizing DDM stress coefficient matrix generation and combining FVM to calculate crack tip pressure. OpenMP multi-thread scheduling and Intel MKL and Eigen library optimization are adopted, LU or QR decomposition is dynamically selected according to the matrix scale, and therefore the calculation efficiency is improved. According to the method, by initializing a three-dimensional fracture grid, optimizing a stress matrix, constructing a fluid-solid coupling equation set, carrying out iterative solution, checking convergence and generating a three-dimensional fracture expansion model, efficient fracture expansion simulation is achieved, and an efficient and accurate numerical tool is provided for fracture design of unconventional oil and gas reservoirs.

The invention relates to the technical field of compressed sensing, and particularly discloses a sparse adaptive system identification method and system based on inverse QR decomposition, and the method comprises the steps: a receiving signal obtaining step: transmitting a transmitting signal to a detected target, and then obtaining a receiving signal, the receiving signal comprises a direct coupling interference signal and a target signal; an estimation signal obtaining step of obtaining an estimation signal corresponding to the transmitting signal based on the transmitting signal through an adaptive filtering model based on inverse QR decomposition; and an identification step: eliminating the direct coupling interference signal through the estimation signal to obtain the target signal.

The invention belongs to the field of intelligent manufacturing, and discloses a component state real-time sensing method based on parameterized loads. According to the method, parameterized representation is carried out on the load, an overall stiffness matrix is obtained in combination with finite element analysis, and a displacement data set of a component under the parameterized load is constructed through Monte Carlo statics simulation. On the basis of POD, in combination with parameterized loads, a parameterized intrinsic orthogonal decomposition method is provided, then a projection base vector used for expressing a component displacement deformation mode is obtained, sensor layout optimization is conducted on a component deformation sensitive area through QR decomposition, and finally a real-time and accurate displacement state sensing result is obtained. The calculation amount of real-time sensing of the displacement state of the component can be effectively reduced, and the operation efficiency is improved; and by optimizing the layout of the sensor, the displacement sensing precision of the component is improved.

The present invention discloses a method for minimal-delay fine-grained parallel Givens rotation of ultra-large-scale matrix QR decomposition, relating to the technical field of signal and information processing. The method comprises: an outer loop and an inner loop, wherein the outer loop is responsible for calculating a row vector of a Q matrix and an R matrix, and the inner loop is responsible for the Givens rotation of the two row vectors; the outer loop is decomposed into multiple loop blocks and a multi-level PE unit array is used to realize parallel operation of the outer loop slices, while the parallel calculation of the inner loop slices is realized by a step-by-step decreasing element update module inside the PE unit; the collaborative design of the inner and outer loop slices significantly reduces the idle time of hardware resources and optimizes the generation and transmission of key data, so that the system starts calculation with minimal delay; the method improves the utilization efficiency of hardware accelerators and is suitable for ultra-large-scale matrix decomposition tasks in the fields of signal processing, image processing, and scientific computing.

The application discloses a low-complexity deep learning MIMO detection method based on partial MAP, comprising the following steps: obtaining multiple groups of training and test samples, and performing hierarchical and data preprocessing on the training and test data samples through QR decomposition to obtain K training sublayers and calculate the input and output labels of each training sublayer; training a DNN for generating local LLR in each detection stage of the K training sublayers based on a partial MAP method; performing layer-by-layer inspection from the Kth layer to the first layer, for each detection stage, calculating the local LLR by using the trained DNN, updating the LLR result of the current layer in combination with the prior information transmitted by the previous layer, and transmitting the LLR result to the next layer as prior information until the final LLR result is obtained; inputting the final LLR result into an FEC decoder to perform error correction and obtain a final signal estimation result; and the partial MAP directly outputs log-likelihood ratio soft information, and the combination of the partial MAP and a forward error correction code can obtain better detection performance.

This invention discloses a method and device for reconstructing the deformation of mesh antenna nodes, belonging to the field of spaceborne mesh antennas. Based on the geometric position and topological connection relationship of the mesh antenna, a mapping model between node displacement and cable element strain is established. RBF radial basis interpolation is introduced, and the displacement information of a small number of control points is used to characterize the deformation field of the entire reflector surface, establishing an explicit relationship between cable element strain and control point deformation. Singular value decomposition is performed on this mapping matrix to extract the dominant strain characteristics; then, combined with column-pivot QR decomposition, the most linearly independent cable elements with the most significant response to the dominant strain are automatically identified as the placement positions for strain sensors. This invention achieves high-precision reconstruction of mesh antenna node displacements with a limited number of sensors, effectively reducing the number of sensors required while ensuring reconstruction accuracy, thus solving the problems of large sensor count and high system complexity in on-orbit deformation monitoring of spaceborne mesh antennas.

The invention discloses a super-large scale matrix QR decomposition minimum delay fine granularity parallel Givens rotation method, and relates to the technical field of signal and information processing, the method comprises an outer layer cycle and an inner layer cycle, the outer layer cycle is responsible for calculation of a row vector of a Q matrix and a row vector of an R matrix, and the inner layer cycle is responsible for Givens rotation of the two row vectors; an outer layer cycle is decomposed into a plurality of cycle blocks, parallel operation of an outer layer cycle piece is achieved through a multi-stage PE unit array, and meanwhile parallel calculation of an inner layer cycle piece is achieved in a PE unit through a step-by-step decreasing element updating module; the idle time of hardware resources is remarkably shortened through the collaborative design of the inner circulation piece and the outer circulation piece, generation and transmission of key data are optimized, and the system starts calculation with the minimum delay; the method improves the utilization efficiency of the hardware accelerator, and is suitable for super-large-scale matrix decomposition tasks in the fields of signal processing, image processing, scientific calculation and the like.

The application discloses a radar target single-station RCS fast scanning angle method, which comprises the following steps: firstly, a triangle grid is used to divide a created radar target geometric model; secondly, an ACA algorithm is used to perform low-rank compression processing on an excitation matrix; thirdly, a double-layer mixed compression method combining an ACA algorithm and QR decomposition is used; fourthly, QR decomposition is performed on the compressed matrix, and ACA algorithm is used for re-compression; fifthly, the number of excitations required for iterative solving in the solving process is further reduced; and finally, a multilevel fast multipole method is used to combine a near-zone block diagonal preconditioning technology to construct a matrix equation, so that the number of iterative solving times of the matrix equation is greatly reduced, the solving of the matrix equation is accelerated, and the solving efficiency of the radar target single-station RCS under the excitation of a plane wave at multiple angles is improved.

The application relates to a QR decomposition-based matrix-vector multiplication implementation method based on a vector processor, and the QR decomposition algorithm is a vsip_cqrd_f function in a VSIP library, the VSIP library is a vector, signal and image processing library launched by a GE intelligent platform, and is an open standard application programming interface specially provided for developers who need to use intensive signal and image processing. The technical scheme of the application mainly comprises the following steps: step one, judging whether input matrix C needs transposition conjugate processing and whether the matrix C is on the left or right in matrix multiplication, so as to enter different branch processing; step two, pre-processing the A matrix; step three, adopting different processing schemes according to the quantity size to realize matrix-vector multiplication operation. The application is designed and realized for the QR decomposition-based matrix-vector multiplication in the light of the advantages of the vector processor, such as large storage space, on-chip cache and efficient and fast data transmission, and can play a great advantage in complex matrix operation.

The application discloses a spherical decoding detection method and device, electronic equipment and a readable storage medium, and belongs to the technical field of wireless communication. The application obtains a Q matrix and an R matrix by performing orthogonal triangular QR decomposition on a channel response matrix, wherein each column in the Q matrix is arranged according to the size of channel energy; determines a target equalization signal according to the Q matrix and the R matrix; performs an ML path search process according to the target equalization signal to select N branches with minimum metrics as survivor paths; performs ML complement path search on the survivor paths to obtain ML complement paths; takes the survivor paths with minimum metrics as ML paths, and obtains the likelihood ratio information of each bit of each symbol at each layer according to the ML paths and the ML complement paths. The application can reduce the operation complexity of the spherical decoding algorithm while maintaining the decoding detection performance.

The invention discloses a joint sorting QR decomposition method and related device for multi-antenna detection, and belongs to the technical field of multi-input multi-output signal detection, and the method comprises the steps: initializing an orthogonal matrix and an upper triangular matrix; acquiring a channel matrix H; extracting candidate columns from residual columns in the channel matrix H based on the sum of residual energy and self energy joint weighting, updating an orthogonal matrix and an upper triangular matrix according to the candidate columns until all columns in the channel matrix H are traversed, and obtaining a final orthogonal matrix and a final upper triangular matrix; and taking the final orthogonal matrix and the upper triangular matrix as a joint sorting QR decomposition result for multi-antenna detection, the method and the related device can shear out the nodes with the super radius as early as possible, reduce access nodes, and improve the speed of multi-antenna detection while ensuring the detection performance.

The invention discloses a QR decomposition matrix inversion optimization method and system. The method comprises the following steps: receiving signal data, and carrying out FFT (Fast Fourier Transform) preprocessing and data downsampling; double-buffer optimization parallel processing is adopted, and FFT signal data are stored based on a double-buffer FIFO data storage structure; calculating a covariance matrix of the FFT signal data; adopting Givens rotation to gradually eliminate off-diagonal elements of the covariance matrix, carrying out Givens rotation calculation, and adopting an optimized CORDIC calculation method to calculate a Givens rotation factor; a calculation process of Givens rotation is accelerated by adopting a Systolic array architecture; and outputting a final QR decomposition result according to a Givens rotation calculation result. According to the method, the data flow is optimized, the calculation throughput rate is improved, hardware resource occupation is reduced, a low-power-consumption and high-performance hardware implementation scheme is provided for efficient matrix calculation, and the method has wide engineering application value.

The invention relates to a TSN (Time Sensitive Network) timestamp recovery method based on an FPGA (Field Programmable Gate Array), which belongs to the technical field of TSNs and comprises the following steps of: S1, collecting synchronous messages of different TSN nodes in parallel by multiple ports of the FPGA, and analyzing timestamp data by taking a message serial number as an anchor point; s2, constructing an incomplete timestamp matrix and a mask matrix by adopting a time sliding window mechanism, and automatically marking missing data; s3, a random index set is generated through LFSR, and matrix sparse sampling is achieved through a bitmap judgment unit; and S4, based on an FPGA accelerated CUR matrix decomposition algorithm, solving a pseudo-inverse matrix through parallel MGS-QR decomposition, and reconstructing a complete timestamp matrix through block parallel matrix chain multiplication. The FPGA parallel architecture, ping-pong cache and fixed-point number processing are adopted, microsecond-level low-delay recovery is achieved, hardware resource occupation is greatly reduced, robustness in a severe network environment is improved, and the method is suitable for a high-precision clock synchronization scene of an industrial TSN node.

This invention discloses an efficient QR matrix decomposition method based on FPGA for implementing systolic matrices, belonging to the field of software-defined radio signal transmission technology for airborne networks. It is based on the design of a multi-channel digital front-end implemented on an FPGA chip. The method includes: designing a general handshake coordination mechanism, designing a timing flow for systolic matrix multiplication, performing fast QR matrix decomposition single-loop operations by calculating the Householder mirror transformation matrix, designing a top-level loop pipeline for QR matrix decomposition, and finally verifying the hardware scheme of the fast QR decomposition algorithm. This invention converts resource consumption into timing consumption by designing the overall pipeline logic and loop operations, enabling fast processing and accurate decomposition of digital information in a multi-channel digital front-end on an FPGA, effectively saving on-chip resources. In addition, this invention has excellent scalability, supporting 64 to 128-bit high-precision fixed-point number formats, and the data matrix order can reach up to 128.

This application provides a method and system for ad push based on user browsing behavior, belonging to the field of data processing technology. First, this application simultaneously acquires the user's historical search terms and current page interaction behavior, including click weights and dwell time, and constructs an intent vector representing long-term preferences and a behavior vector reflecting current interests accordingly. Next, the two vectors are combined into an observation matrix, and a QR decomposition technique is used to extract an orthogonal basis vector representing the user's independent, core intent as a reference standard. Subsequently, the grey relational coefficient and correlation degree between this reference vector and the attribute vectors of each candidate ad are calculated to quantify the degree of matching between the two. Finally, the ads with the highest correlation are selected and accurately pushed to the user. This application achieves accurate mining of the user's core intent and matching and pushing highly relevant ads in complex page environments.

The present application relates to the field of digital image watermarking technology in information security, and provides a double color data image watermarking, image watermarking embedding, image watermarking extraction method and system. The double color data image watermarking method comprises an embedding process and an extraction process. In the embedding process, the color watermarking image W is encrypted by using HDNA, Mandelbrot graph and three-dimensional IPCM chaotic system to obtain an encrypted watermarking; the color carrier image is decomposed by QR decomposition, and then the encrypted watermarking is embedded into the color carrier image. In the extraction process, the encrypted watermarking information is extracted by QR decomposition, and then the original color watermarking information is extracted from the encrypted watermarking information by HDNA and Mandelbrot graph. The present application provides a safe, transparent and efficient double color image watermarking method, which has the advantages of high security, strong robustness, large copyright information capacity and fast operation speed, and has stronger application value.

The invention relates to the technical field of carbon emission calculation, and discloses an electric power carbon emission factor calculation method, system, equipment, medium and product based on QR matrix factorization. According to the method, a node power balance equation and a carbon emission conservation equation are determined by obtaining carbon emission information and tidal current power information of all generators in a power grid; the method comprises the following steps: constructing a carbon emission factor linear equation, solving the carbon emission factor linear equation based on QR matrix decomposition to obtain a carbon emission factor value of each node, so that efficient solving is realized by utilizing sparse matrix characteristics and QR decomposition, rapid response of multi-node carbon emission factor solving is realized, and the method does not depend on initial solution setting; the slow convergence or failure risk of an iteration method is avoided, the numerical stability is excellent, and the accuracy and reliability of carbon emission factor calculation are improved.

The invention discloses a temporary overvoltage calculation method and device for a new energy direct current delivery system. The method comprises the following steps: screening influence parameters of an overvoltage peak value of a new energy direct current delivery system in electromagnetic transient simulation based on a Sobol global sensitivity analysis method, and determining key influence factors; according to the value range of the key influence factor and a preset constraint condition, determining multiple groups of sample data of the key influence factor and a sample data set formed by a temporary overvoltage peak value of each group of data; fitting the sample data set by adopting a ridge regression method of QR decomposition to obtain a fitting relationship between the key influence factors and the transient overvoltage peak value; and calculating a simulation transient overvoltage peak value of the new energy alternating current delivery system according to the simulation key influence factor value of the new energy direct current delivery system by adopting the fitting relationship.

The present invention discloses an application method of a variable projection method based on QR decomposition in surface deformation. The fitting of surface deformation in a study area in time and space is taken as the research focus. On the basis of applying SBAS-InSAR time series analysis technology to explore the surface deformation in the study area, a complex exponential model with a separable structure is used to fit the settlement situation of the surface in spatial position. The model parameters are solved by a variable projection algorithm based on QR decomposition. The settlement profile curve is approximated with high precision in spatial position. At the same time, the surface is predicted in time in combination with the Logistic model, thereby realizing comprehensive monitoring of the surface deformation in time and space, and providing reliable technical support for surface deformation monitoring.

The present invention relates to a method for determining insulation resistance (R1, R2) and discharge capacitance (C1, C2) for ground (PE) of an ungrounded DC power supply system, wherein a linear difference equation is implemented as a function of a measurement voltage (UG(k)) and a grid voltage (UN(k)) with a measurement voltage (UM(k)), and a grid parameter (θi) formed from a shunt (RM), insulation resistance (R1, R2) to be determined, and discharge capacitance (C1, C2) to be determined. Using N linear differential equations, a measurement value equation system is implemented with a sample sequence ((UG(k), UN(k), UM(k))) and a grid parameter (θi) for measurement times from k = 1, 2 to N with a sampling period T. Another method step comprises a step (S6) of calculating the estimated grid parameter (θi) as an approximate solution to the measurement value equation system, wherein the sum of squared errors between the grid parameter (I) and the estimated grid parameter (θi) is minimized, a step (S7) of minimizing the sum of squared errors through QR decomposition of the measurement-value matrix (Ψ), wherein the QR decomposition is recursively calculated, and a step (S8) of calculating each insulation resistance (R1, R2) and each discharge capacitance (C1, C2) from the estimated grid parameter (I) and repeating the method step with the estimated grid parameter (I), each calculated to include samples available at the current measurement time.

The application provides a sorting QR decomposition method for a MIMO system, comprising: determining an initial matrix to be sorted according to a channel matrix of a MIMO system to be detected; determining column vectors to be sorted of the matrix and a column number thereof, denoted as a target number; if the target number is greater than a preset value, performing a first sorting strategy: in a first round of iteration, exchanging column orders of a column vector to be sorted with a smallest norm value and a first column vector to be sorted, and in each subsequent round of iteration, exchanging column orders of a column vector to be sorted with a second smallest norm value and a column vector to be sorted in a current round, to obtain a first target matrix; and a last column vector of the first target matrix corresponds to a highest layer traversal constellation point of an FSD algorithm. The above scheme can obtain an optimized target matrix. The application also provides an FSD-based signal detection method, which performs FSD signal detection by using the above optimized target matrix, so that the number of highest layer traversal constellation points is reduced, and the detection accuracy and reliability are still ensured.

This invention relates to the field of digital signal processing technology, and discloses a DSP multi-core parallel QR decomposition method and apparatus based on tile algorithm. The method includes: dividing the matrix to be decomposed into blocks in the horizontal direction according to the number of cores of the DSP processor and setting tile blocks; performing QR decomposition on the main diagonal tile blocks; and updating all tile blocks to the right of the main diagonal tile blocks in the horizontal direction; pairwise coupling of the main diagonal tile blocks with each tile block in the vertical direction of the same column; performing QR decomposition on the coupling matrix formed by each pair of couplings; and updating all coupling matrices to the right of the pair of coupling matrices in the horizontal direction; and cyclically executing the decomposition and horizontal update of the main diagonal tile blocks, the coupling decomposition of the tile blocks in the same column, and the update of the right coupling matrix until all main diagonal tile blocks in the matrix to be decomposed have been processed, resulting in an upper triangular matrix. This invention can fully utilize the parallel processing advantages of multi-core processors to achieve inter-core parallel data processing.

The invention provides a vibration structure minimum norm control method based on a dynamic compensator, and the method comprises the steps: building a vibration structure model through a finite element analysis method, and designing the dynamic compensator; processing the unchanged feature structure matrix by adopting QR decomposition, and calculating a block matrix; performing QR decomposition on the correlation matrix of the closed-loop system, and calculating an actual representation matrix of an expected characteristic value and a matrix formed by an expected characteristic structure and compensator gain; a parameterized vector is selected, and parameterized dynamic compensator gain is calculated; if the cost function exceeds a set threshold value, the optimization function is optimized through a gradient optimization method according to any given initialization parameter vector, and a dynamic compensation gain matrix and a control gain matrix of the dynamic compensator are calculated through the optimal parameterization vector. According to the method, the calculation effectiveness is improved, the calculation complexity is reduced, meanwhile, the degree of freedom of the parameterized vector is not changed, and the sufficient optimization degree of freedom is ensured.

The invention discloses a temporary overvoltage calculation method and device for a new energy alternating current delivery system. The method comprises the following steps: screening influence parameters of an overvoltage peak value of a new energy alternating current delivery system in electromagnetic transient simulation based on a Morris global sensitivity analysis method, and determining key influence factors; according to the value range of the key influence factor and a preset constraint condition, determining a sample data set composed of multiple groups of sample data of the key influence factor and a temporary overvoltage peak value of each group of data; fitting the sample data set by adopting a QR decomposition least square method to obtain a fitting relationship between the key influence factors and the transient overvoltage peak value; and substituting the simulation key influence factor value of the new energy alternating current delivery system into the fitting relation to calculate the simulation transient overvoltage peak value of the new energy alternating current delivery system.

Embodiments of the present disclosure provide a visual inertia fused localization method and a device. The method comprises: acquiring target observation data corresponding to a target state vector; constructing a first residual model according to the target observation data; projecting the residual model according to a first-order partial derivative matrix, to obtain a second residual model, the second residual model comprising a product term of a second-order partial derivative square matrix and a system state matrix, and a covariance term corresponding to a first observation noise term; and eliminating a three-dimensional coordinate error vector of the second residual model, to obtain a target residual model, and performing EKF updating on the target state vector according to the target residual model. The embodiments do not involve left null space calculation and projection, and QR decomposition dimension reduction calculation, thereby reducing the amount of calculation, and improving positioning efficiency.