56 results about "Gauss newton method" patented technology

A method of parameter adaptation is used to modify the parameters of a model to improve model performance. The model separately estimates the contribution of each model parameter to the prediction error. It achieves this by transforming to the time-scale plane the vectors of output sensitivities with respect to model parameters and then identifying the regions within the time-scale plane at which the dynamic effect of individual model parameters is dominant on the output. The method then attributes the prediction error in these regions to the deviation of a single parameter from its true value as the basis of estimating individual parametric errors. The proposed Signature Isolation Method (PARSIM) then uses these estimates to adapt individual model parameters independently of the others, implementing, in effect, concurrent adaptation of individual parameters by the Newton-Raphson method in the time-scale plane. The Signature Isolation Method has been found to have an adaptation precision comparable to that of the Gauss-Newton method for noise-free cases. However, it surpasses the Gauss-Newton method in precision when the output is contaminated with noise or when the measurements are generated by inadequate excitation.

The invention discloses a VSLAM method based on a multi-characteristic visual odometer and a graph optimization model, and belongs to the robot SLAM field; the method comprises the following steps: firstly using a FAST (features from accelerated segment test) and an improved LSD algorithm to extract point and line features in a color image; further using different descriptors to describe characteristics; then carrying out characteristic matching; finally using an improved NICP (normal iterative closest point) algorithm and a PnL (perspective n line) algorithm to estimate a robot initial posture. The method can extract image line features so as to enlarge algorithm application scenes, can obtain a well robot initial posture, uses a Bayes network to express a multi-characteristic visual odometer, obtains a factor graph on the basis of the Bayes network, uses a maximum posterior probability to estimate a robot global posture in the factor graph, and uses a Gauss-Newton method to solve themaximum posterior probability to obtain the updated posture graph; finally, the posture graph and three dimensional points of corresponding frames are fused to obtain a reconstructed three dimensional map.

The invention provides a monocular infrared video three-dimensional reconstruction method based on a visual odometer. Firstly, a thermal infrared imager is calibrated; then, a direct method and sparse method visual odometer model is constructed; then, frames are managed, a key frame is generated from a continuous infrared image sequence to be processed, and the key frame is marginalized; point management is conducted on the pixel on the key frame; finally, sliding window optimization is conducted on the key frame, a photometric error of the direct method and sparse method visual odometer model is minimized, the reverse depths of thermal infrared imager pose, internal references and spatial points are solved in an iterative algorithm by a gauss-newton method, and finally, three-dimensional point cloud of a scene is obtained. According to the monocular infrared video three-dimensional reconstruction method based on the visual odometer, the direct method is applied to photometric error minimization, traditional feature point detection and matching are skipped, through direction operation on a grey value of the pixel, all variables which the photometric error to be optimized depend on is calculated directly, and therefore three-dimensional reconstruction of a night vision scene is achieved, the real-time capability of the three-dimensional reconstruction is ensured, and the spaciousness of the scene is enhanced.

The invention discloses a quaternion-based attitude resolving method for an extended Kalman filter algorithm. The method comprises the following steps: acquiring multisensor data of a vector in a fixed coordinate system; filtering data acquired by an accelerometer and a magnetometer, and subjecting the data acquired by the two sensors to normalization processing; constructing a state equation of avector system according to a quaternion differential equation and an attitude matrix, and determining a process noise variance matrix of the system; constructing a systematic observation model by using a rapid Gauss-Newton method, and determining a system measured noise variance matrix; constructing a Kalman filtering recurrence equation according to the constructed system state equation and theobservation model; resolving three attitude angles of the vector by using an optimal quaternion obtained through recurrence. According to the method, the computing capacity can be greatly simplified,and the existing problem that parameters are unclear in computing is solved.

A method of parameter adaptation is used to modify the parameters of a model to improve model performance. The model separately estimates the contribution of each model parameter to the prediction error. It achieves this by transforming to the time-scale plane the vectors of output sensitivities with respect to model parameters and then identifying the regions within the time-scale plane at which the dynamic effect of individual model parameters is dominant on the output. The method then attributes the prediction error in these regions to the deviation of a single parameter from its true value as the basis of estimating individual parametric errors. The proposed Signature Isolation Method (PARSIM) then uses these estimates to adapt individual model parameters independently of the others, implementing, in effect, concurrent adaptation of individual parameters by the Newton-Raphson method in the time-scale plane. The Signature Isolation Method has been found to have an adaptation precision comparable to that of the Gauss-Newton method for noise-free cases. However, it surpasses the Gauss-Newton method in precision when the output is contaminated with noise or when the measurements are generated by inadequate excitation.

The invention discloses a mean-line-based blade front and back edge fitting and section line smooth reconstruction method. The method is used for solving the technical problem that an existing method is poor in practicability. According to the technical scheme, on the basis of a fitting problem of front and back edge measuring points, a restrained least square fitting method is built for fitting of front and back edges meeting the design requirement. Different from unrestraint least square method front and back edge fitting, the method adopts a mean line as constraint conditions for front and back edge fitting, the restrained least square method is built, a Gauss-Newton method is used for iterative solving, finally, a fitting front and back edge curve approximating the designed front and back edges is obtained, front edge radius errors are lowered to 0.007% from original 0.670%, and the back edge radius errors are lowered to 0.062% from original 1.018%. The fitting front and back edge and the blade back and blade basin achieve smooth connection, and the smooth section line curve is built. The method has the advantages of being high in calculation precision, high in convergence rate and effective in constraint.

The application relates to a mobile side vision fusion positioning method and system, and electronic equipment. The method comprises the following steps: step a, acquiring an initial location of a mobile terminal as the current location of a positioning target based on the calibrated starting location and sensor information; step b, acquiring a video frame by using the mobile terminal; and step c,detecting a static object in the video frame, acquiring geographic coordinate information of the static object through a BIM space database, and substituting the coordinate information of the staticobject into a multi-target object positioning model, iteratively solving the positioning model through a Gaussian newton method, acquiring the current location of the mobile terminal, and combining the current location of the mobile terminal and the coordinate information of the static object to obtain a positioning result of the positioning target. Through the positioning method disclosed by theapplication, the more convenient, precise and cheaper positioning method can be realized.

An airframe butt joint attitude-adjusting method meeting stringer reference alignment and straightness requirements includes the steps: measuring end face reference points and linear guide rail reference points of a fixed airframe section and an attitude-adjusting airframe section and coordinates of the position of an attitude-adjusting system three-dimensional positioner connector connected with the attitude-adjusting airframe section by a laser tracker; building airframe end face reference point matching and straightness matching target functions, and accordingly building a mathematic model for airframe butt joint attitude-adjusting calculation; performing iterative calculation for the airframe butt attitude-adjusting model by a Gauss-Newton method to obtain coordinates of optimal butt joint positions of the attitude-adjusting airframe section and the fixed airframe section and obtain a transformational matrix from an initial position to a target position; calculating the coordinates of the target position of an attitude-adjusting system three-dimensional positioner connector connected with the attitude-adjusting airframe section according to the obtained transformational matrix, and butting the attitude-adjusting airframe section and the fixed airframe section by controlling movement of the attitude-adjusting system three-dimensional positioner from the initial position to the target position.

The invention provides a passive sensor error registering method and apparatus based on maximum likelihood estimation. When an observed point is positioned, coordinates of the observed point and three observation points under the same coordinate system, azimuth measurement values between the observation points and the observed point and random measurement deviations are obtained. By use of the maximum likelihood estimation and a Gauss-Newton method, system deviations and random deviations in an observation process are processed, and positioning results satisfying precision requirements are obtained by use of an iteration method. Compared to a positioning method in the prior art, the method brought forward by the invention takes system deviations between the observation points into consideration and improves the positioning precision of a passive sensor.

The invention discloses a method for calculating direct-current transmission line electric field intensity and radio interference. The method solves the problems that in the prior art and a traditional method, a transmission line lead section has waved relief, so that lead mirror image charge cannot be directly set, and ill-posed problems of an original equation can be solved when a direct-current transmission line adopts linear equations such as a Gauss-Newton method. Compared with a traditional partial discharge method, experimental efficiency is improved, and the method has practicability and feasibility through experimental verification.

The invention relates to a method for fitting a power function type stage-discharge relation, and belongs to the technical field of hydrology and water resources. The method includes utilizing a larger group of relation values and a smaller group of relation values in n groups of relation values to determine initial values theta (0)=[beta 0(0), beta (0)]' of parameters to be estimated in a power function type stage-discharge relation formula according to a stage-discharge relation value (Ze1, Q1) to be fit in a certain hydrometric station, or solving a nonlinear regression coefficient as initial values of the parameters beta 0 and beta to be estimated by a traditional method of converting nonlinear regression to linear regression; and determining regression coefficients beta 0 and beta by a Gauss-Newton method. An i ranges from 1 to n. The stage-discharge relation is one of important contents for reorganizing hydrologic data, and is one of important references for water-conservancy and water-power engineering planning programming, hydrological forecasting and reservoir regulation.

The invention provides an electrical impedance imaging method based on a movable deformation component, a device and a system. The electrical impedance imaging method first establishes a finite element electrode model for a target field and acquires a measured voltage value of each electrode corresponding to an object to-be-imaged; a plurality of movable deformation components are initially placed for each object to-be-imaged as an initial value of a Gauss-Newton iteration method, an iterative process is performed, a target conductivity distribution is obtained, and a target voltage value is calculated; and the reconstructed image is determined by iteratively solving the minimization of the preset loss function based on the Gauss-Newton method. The scheme is based on the movable deformation component for image reconstruction, and utilizes the flexibility control of the movable deformation component in structural topology optimization, and the resolution and imaging precision of the electrical impedance reconstruction image can be improved.

The invention discloses a bridge cable monitoring method based on an optimized tensioning string model and a bridge cable monitoring system thereof. The method comprises the steps that a steel cable having the same specification with the actually measured bridge cable is selected to construct the residual function of the Gauss-Newton method, then iterative solving is performed on the residual function of the Gauss-Newton method and the optimal eigen-frequency order number and the optimal steel cable flexural rigidity of the residual function of the Gauss-Newton method are obtained, and finally the tensioning string model is optimized according to the result of iterative solving; the theoretical eigen-frequency is solved; the vibration spectrum of the bridge cable is obtained, and then the frequency meeting the optimal eigen-frequency order number is selected from the obtained vibration spectrum to act as the actually measured eigen-frequency through screening; and an acceleration sensor node determines transmitting of acquisition data and sampling frequency for adjusting the acceleration sensor to an upper computer monitoring center according to the size of the ratio of the difference value of the actually measured eigen-frequency and the theoretical eigen-frequency to the theoretical eigen-frequency. The bridge cable monitoring method is low in error and considers analysis of both accuracy and power consumption so that the bridge cable monitoring method can be widely applied to the field of bridge monitoring.

The invention provides a maximum power estimation method for a photovoltaic power generation system. The maximum power estimation method comprises the steps of establishing a double-diode 7 parametermodel of a photovoltaic cell, and realizing conversion calculation from environmental parameters to standard conditions by adopting open-circuit voltage and short-circuit current parameters; and usingan L-M algorithm controlled by parameters, so that the algorithm shows the characteristics of a gradient descent method and a Gauss-Newton method in sections, and the advantages of the two methods are fully combined. Photovoltaic double-diode model parameter calculation based on the L-M algorithm is less affected by iterative initial value selection, and convergence speed and precision are increased. The maximum power value is obtained based on an open-circuit voltage approximation method. The process of one-time iteration is omitted, and the calculation speed is increased.

The invention discloses a three-dimensional reconstruction method based on optimized IMU tight coupling dense direct RGBD, and the method comprises the steps: (1) collecting optimized RGB data and optimized depth data from a local model reconstructed at the last moment, and obtaining original RGB data, original depth data and IMU data; (2) constructing a nonlinear cost function according to the inertial navigation error, the dense luminosity error and the ICP error, forming a nonlinear optimization problem by taking minimization of the nonlinear cost function as a target, and solving the nonlinear optimization problem by using a Gauss-Newton method so as to estimate an RGBD sensor optimization pose; (3) carrying out real-time three-dimensional reconstruction by utilizing the original RGB data and the original depth data at the current moment under the optimized pose of the RGBD sensor, and deforming the three-dimensional reconstruction model by using the nodes under the condition of finding the loopback constraint so as to ensure the global consistency. According to the method, the accuracy and efficiency of three-dimensional reconstruction are improved.

The invention provides a phasor measurement device laboratory calibrator and a phasor measurement method thereof which belongs to the technical field of phasor measurement and phasor measurement device test. The system includes a GPS synchronous time service module, a signal collection module and a controller module; the measurement method includes: model selection is performed on a hardware, influence of hardware error to phasor measurement is disclosed by utilizing an error transmission theory; measurement is performed by utilizing an improved energy center of gravity method and an improvedDFT algorithm during stable state test, a non-linear fitting model is constructed during dynamic test, solving is performed by utilizing a Gauss-Newton method, polynomial fitting is adopted for solving a frequency change rate of static and dynamic test signals. The calibrator can work reliably, and can limit influence on phasor measurement precision by a collection system within a small range andprovide a high precision reference value for error analysis. The measurement method makes the phasor measurement accuracy more than four times higher than an error standard, and has a good inhibitioneffect on harmonic wave and inter-harmonics interference.

The invention relates to a real-time high-precision visual mileage calculation method. According to the method, the mileage information of a system is acquired in real time by means of a closed-loop corner matching strategy under cross-check and the objective function weight setting based on a corner life cycle. The method comprises the following steps: (1) obtaining corner positions in stereo image pairs by partitioning HARRIS corner detection; (2) obtaining a normalized descriptor of corners according to pixel information around the corners; (3) carrying out corner cross-check matching on stereo image pairs, and then performing closed-loop detection on the corners between the stereo image pairs and the front and back image pairs so as to obtain a matching corner set; (4) constructing anobjective function based on the re-projection error of the matching corner set, and setting a weight according to the corner life cycle; (5) solving the mileage information by means of a RANSAC algorithm and a Gauss Newton method. The method provided by the invention has the advantages that the high-accuracy matching corner set is obtained, the interior point selection time of the RANSAC algorithmis shortened, and the real-time mileage information calculation can be realized.

The embodiment of the invention provides a signal determination method and device, electronic equipment and a storage medium. The method comprises the steps of acquiring a sample signal and a currentreceiving signal; determining a joint probability density function of the sample signal and the current received signal with respect to each parameter to be estimated; determining an expression corresponding to the parameter to be estimated when the joint probability density function obtains a maximum value through a maximum likelihood estimation algorithm; calculating doppler frequency offset andcode delay through a Gauss Newton method, and obtaining a target value of the Doppler frequency offset, a target value of the code delay, a target value of a carrier phase and a target value of a signal amplitude according to an expression corresponding to a parameter to be estimated when a calculation result of the Doppler frequency offset and the code delay meets preset conditions, wherein thetarget value of the Doppler frequency offset, the target value of the code delay, the target value of the carrier phase and the target value of the signal amplitude represent the current received signal. According to the invention, stable tracking of a signal with a small duty ratio is realized.

The invention discloses a high-precision fine synchronization device and method. The method comprises the steps that (S1) course capture of a synchronization sequence is carried out; (S2) multipath parallel correlation units are adopted for carrying out correlation operation in a chip before a peak value of the course capture and in a chip after the peak value of the course capture to obtain a sample sequence Y of a correlation function; (S3) the correlation function is approximated by means of the Gaussian function, the Gaussian function is initialized with the help of the sample sequence Y, and Gaussian function parameters are iterated through a Gaussian-Newton method; (S4) after iteration is finished, the parameter b of the Gaussian function is extracted, and the estimated value of the peak value position is obtained directly. According to the high-precision fine synchronization device and method, the shape features of the correlation function are utilized fully, the correlation function is approximated by means of the Gaussian function, the parameters are iterated through the Gauss-Newton method, synchronization results are extracted from the parameters, and the precision of precision synchronization is improved; the convergence sped is high, and parameter extraction is direct; achievement complexity is low, calculated amount is not increased greatly compared with a parallel correlation method under an equivalent sampling rate, and the device and the method are less influenced by noise, and are convenient to apply in reality.

The invention provides a ground transient electromagnetic method inversion method and device based on emission current full waveforms, and the method comprises the following steps: constructing a one-dimensional inversion target function based on a Tikhonov regularization mode, carrying out the optimization of the inversion target function through employing a Gaussian Newton method, and outputting a final inversion model. In a process of optimizing an inversion objective function, an emission current basic waveform used for one-dimensional inversion of a ground transient electromagnetic method is expanded from a slope step waveform to an emission current full waveform containing starting time, stabilization time, turn-off time and a duty ratio, and a cycle number parameter is included in the emission current basic waveform. The strategy is not only helpful for improving the interpretation precision of data processing, but also helpful for perfecting the design of instrument parameters.