234 results about "Unscented kalman filtering" patented technology

A method for estimating joint angles of a multi-segment limb from inertial sensor data accurately estimates and tracks the orientations of multiple segments of the limb as a function of time using data from a single inertial measurement unit worn at the distal end of the limb. Estimated joint angles are computed from measured inertial data as a function of time in a single step using a nonlinear state space estimator. The estimator preferably includes a tracking filter such as an unscented Kalman filter or particle filter. The nonlinear state space estimator incorporates state space evolution equations based on a kinematic model of the multi-segment limb.

The invention discloses a lithium battery SOC online estimation method comprising the following steps that 1) the open-circuit voltage of a battery is measured, and the state-of-charge initial value of the battery is obtained according to an OCV-SOC curve; 2) the second-order RC equivalent model of the battery is established and the parameter initial value of the battery equivalent model is estimated; 3) the estimation program is started, and the matching coefficient initial value of a state equation is set according to the battery state-of-charge initial value and the parameter initial value of the battery equivalent model; 4) the current battery state-of-charge value is obtained by using an adaptive unscented Kalman filtering algorithm, and the current open-circuit voltage is obtained according to the OCV-SOC curve; 5) the least square method with the forgetting factor is started to identify the parameters of the current battery equivalent model, the matching coefficient of the state equation is updated by the identified parameters and the battery state-of-charge value of the next moment is solved; and 6) the steps 4) and 5) are repeated so that the battery state-of-charge value of each moment is obtained. Compared with the conventional unscented Kalman filtering algorithm, the method has higher accuracy and higher error convergence.

The invention discloses an indoor positioning method based on improved square root unscented Kalman filtering. The method comprises the following steps of firstly, establishing a UWB positioning modeland a pedestrian dead reckoning PDR model in a fusion indoor positioning method; then, according to a residual identification method criterion, determining whether an NLOS criterion exists in a positioning result of the UWB positioning model; if the NLOS exists, using an improved square root unscented Kalman filtering algorithm to fuse a coordinate of inertial navigation dead reckoning with a positioning coordinate of UWB, and correcting a UWB solving coordinate data; and if the NLOS does not exist, using the UWB to analyze a current coordinate and completing real-time positioning. In the invention, the pedestrian dead reckoning PDR is used to compensate a UWB signal loss or non-line of sight condition, which improves the positioning accuracy of the system and the advantages of high precision and high robustness are possessed.

The invention discloses a high-precision integrated navigation positioning method for an underwater glider. The method comprises the steps that output of an inertial measurement unit (IMU) is subjected to rough processing and fine processing; higher-precision data is obtained; data obtained by dead reckoning based on a Runge-Kutta method (RK4) is combined; adaptive Kalman filtering (AKF) and unscented Kalman filtering (UKF) are conducted; the filtered data is fed back to the IMU; an accumulative error of the IMU is corrected; and relatively accurate information such as position, speed and gesture is finally output. The autonomous navigation method for the underwater glider has the advantages of high precision, real-time performance, stability, and the like, so that a whole navigation system can stably operate at long endurance, high precision and low power consumption, the current position and gesture information of an underwater aircraft can be obtained rapidly and accurately, and navigation path and position parameters can be provided for the underwater aircraft.

The invention aims to provide a GPS (Global Positioning System)-pseudo-range-differential-based cooperative positioning method for vehicles. A standard GPS receiver with low cost is used for precisely estimating relative positions of any automobiles. The dominant motive by adopting pseudo-range differential is that the observed quantity of the GPG receiver close to an automobile is ordinarily affected by common errors (for example, same troposphere delay and ionized layer delay), and the errors can be counteracted through a differential process, so that an obtained relative vector is more precise than that of absolute positioning. As the automobile is not directly connected usually, a GPS measured value needs to be exchanged through V2V (Vehicle to Vehicle) wireless communication equipment. According to the method provided by the invention, by combining a constant turning angular velocity and velocity (CTRV) vehicle motion model and adopting an unscented Kalman filter to fuse the GPS measured value and a kinematical sensor measured value, stable estimation to the system state of a target vehicle is provided. The special sensor errors, such as excursion, can be estimated anytime and reduced through unscented Kalman filtering.

The present invention discloses a low-cost method for estimating antenna attitude of satellite communication on the move measurement and control system. The method comprises the steps of: 1) sending a control command from a master controller to a servo driver, and according to the control command, controlling a motor to align the beam pointing of the antenna in the mobile-satellite communication system on the move by the servo driver; 2) acquiring an attitude angle differential equation of a carrier; 3) acquiring an pitch angle theta m and roll angle phi m of the carrier; 4) acquiring a pseudo course angle Psi v; and 5) conducting attitude resolution by using improved unscented Kalman filtering according to the attitude angle differential equation of the carrier, theta m, phi m, and Psi v obtained in the previous steps, so as to obtain a three dimensional attitude angle of the carrier, and then controlling the three motors to adjust the antenna beam pointing according to the three dimensional attitude angle of the carrier. The invention can accurately align the carrier to an object satellite and has low cost.

The invention discloses a power generator dynamic state estimation method based on unscented transformation strong tracking filtering. The power generator dynamic state estimation method comprises two steps, namely, prediction and filtering, wherein in the prediction step, sigma point sampling is carried out through the symmetrical sampling strategy according to a filtering mean value and a filtering covariance matrix at a previous moment, a predicted measurement calculation value is calculated, a residual equation is obtained, and a fading factor is introduced to correct a predicted covariance matrix; in the filtering step, a gain matrix is adjusted in online mode, and estimate values of the power angle and electrical angle speed of a power generator in the electromechanical transient process are obtained after correction. Compared with unscented Kalman filtering and strong tracking filtering, the power generator dynamic state estimation method provided by the invention is high in tracking speed, tracking precision and noise robustness.

The invention discloses a bayesian fitering-based general data assimilation method. The method comprises the steps of: inputting an initial value set into an analysis model in a prediction step so as to obtain a prediction set value; calculating prediction error covariance matrix by using set kalman filtering in an updating step, and updating each prediction set according to the observation value and kalman gain matrix; or, calculating importance weight of each set sample by adopting particle filtering through set prediction value, calculating the number of effective particles by utilizing normalization importance, resampling the set according to the weight to obtain updated analysis value and analysis set; or, calculating prediction error covariance matrix by adopting unscented kalman filtering, and updating each prediction set according to the observation value and kalman gain matrix; conducting next prediction and assimilation by taking the updated analysis set as the initial values of the analysis model, and repeating the prediction step and the updating step. The method can enable Earth remote-sensing observation information and land surface process model information to be effectively integrated, thus forming a land surface process information prediction system with small errors.

The invention discloses a method used for positioning indoor moving targets by improved unscented Kalman filtering. In order to solve problems of traditional unscented Kalman filtering method that positioning precision is low and real-time performance is poor, the traditional unscented Kalman filtering method is improved by an iteration algorithm. Wherein according to the iteration algorithm, a noise covariance matrix is regulated dynamically by using scale factors, i.e. in filtering processes, filtering values of last moments are continuously substituted into a time updating equation; prior estimation of current states is established; estimated values of state variables of current states and estimated values of error covariance are calculated; and observation noise and systematic noise are updated according to the estimated values so as to ensure convergence and stability of the algorithm.

The invention puts forward an indoor moving target positioning method based on trajectory smoothing, comprising the following steps: (1) a mobile UWB (Ultra Wideband) node acquires the distance between the node and each anchor node through polling; (2) the distance measured by each anchor node at a moment is corrected by an unscented Kalman filter; (3) the position coordinate of a moving target is acquired through a trilateration positioning approach based on the filtered UWB measured value at the moment; (4) the positioning result is corrected by the unscented Kalman filter; and (5) the filtered positioning result is smoothed by an adaptive smoothing algorithm so as to reach the position estimated value of the moving target at the current moment. According to the invention, the measured values and estimated values at the current moment and at a previous moment are fully utilized, an unscented Kalman filtering mode is adopted, so the influence of environment, device and other factors on the measurement result is reduced, and the positioning accuracy of the trilateration approach is improved; and the positioning result is filtered secondarily and smoothed adaptively, and the positioning accuracy is further improved.

The invention provides an MCAUPF based target tracking method. An MCAUPF is used to complete state estimation in the target tracking process. In the target tracking process, target tracking equation and measurement equations are reconstructed into a nonlinear recursion model, in the framework of unscented particle filtering, a significance probability density function needed by filtering particlesgenerated by maximum correntropy unscented Kalman filtering is used, a Kullback-Leibler distance resampling method is used to resample the generated particles, the state of the tracking target is estimated according to a UPF algorithm flow, and the target is tracked in real time. The MCAUPF method is applied to target tracking in which outliers occur in noise measurement, the precision is highercompared with that of present PF, improved PF and robust filtering, and the computing complexity is lower than that of a present improved particle filter algorithm.

The invention relates to a state estimation method based on high-order unscented Kalman filtering. A high-order unscented Kalman filter is used for finishing the state estimation task in the target tracking process. According to the state estimation method based on the high-order unscented Kalman filtering, the state estimation task in the target tracking process is finished by the high-order unscented Kalman filter. In the target tracking process, the state equation and the measurement equation of target tracking are established; a sigma point required for the target tracking filter is obtained by high-order unscented transformation, and the weight of the sigma point is calculated; and the state estimation is obtained by iterating the sigma point and the weight of the sigma point to realize the real-time tracking of the target. The tracking precision of the state estimation method is higher than those of the existing target tracking methods based on other filters, a proper performance parameter k is selected to further improve the precision of the proposed high-order unscented Kalman filtering (UKF) target tracking method, and the high-precision real-time tracking to the target is realized. The state estimation method disclosed by the invention is applied to the technical field of the target tracking.

The invention discloses a body language-based emotion perception method adopting deep learning and UKF. The method comprises the following steps of monitoring a person entering a Kinect work area by adopting Kinect, and then calculating skeleton points of the person in real time; estimating the positions of the skeleton points by using unscented Kalman filtering, and calculating a measurement error generated by a tracking error and noises of a device; adopting a convolutional neural network method for static body actions, and adopting bidirectional long-short-term memory conditional random field analysis for dynamic body actions; and directly putting output items of characteristics obtained after action processing into a softmax classifier for performing identification, thereby identifyingeight emotions. The body language-based emotion perception has the following advantages that firstly, body languages can be more easily captured by a sensor; secondly, emotion perception-based body language noises are relatively small; thirdly, the body languages have less fraudulence; and fourthly, the capture of the body actions does not influence or disturb the actions of participants.

The invention discloses a battery SOC estimation method based on the fusion of multi-scale Kalman filtering and unscented Kalman filtering. The method comprises a first step of performing charging anddischarging tests on a to-be-tested battery at different temperatures and under different working conditions through a battery charging and discharging test system and a temperature box; a second step of constructing a battery actual capacity calculation model according to the sample parameters of the to-be-tested battery collected in the test process, and obtaining the actual capacity of the to-be-tested storage battery through the calculation of the calculation model; a third step of establishing a first-order RC equivalent circuit model of the storage battery to be tested according to thesample parameters of the battery collected in the test process, and obtaining a state and observation equation of the first-order equivalent circuit model according to the Kirchhoff law; and a fourthstep of performing state estimation by using a multi-scale adaptive unscented Kalman filter algorithm, inputting the test current and voltage, and obtaining an optimal estimated value of the SOC valueof the storage battery to be measured by taking the minimum error between the voltage of the actual measurement end and the estimated value as a target, namely, the SOC estimated value of the storagebattery.

The invention relates to an autonomous celestial navigation method for a deep space explorer on a swing-by trajectory. The autonomous celestial navigation method comprises the following steps: establishing the corresponding equation of state according to the distance to a planet; calculating the location information of the planet by using the sensitive starlight angle distance detected by a star sensor and calculated by pure astronomic and geometry analytic method; and estimating the parameters of the navigation system by using a fuzzy multiple-model adaptive unscented Kalman filtering (FMMUKF) method. The invention is suitable for the navigation location of a deep space probe on a swing-by trajectory and applicable to the determination of navigation parameters of the deep space probe on a multi-celestial-body intersection swing-by trajectory, belonging to the technical field of aerospace navigation.

The invention discloses a method for performing transfer alignment on a large azimuth misalignment angle of a ship in a polar region environment based on unscented Kalman filtering. The method comprises the following steps: I, completing starting and preheating preparation of a slave inertial navigation system, wherein the slave inertial navigation system completes primary binding by utilizing navigation parameters transmitted by a master inertial navigation system; II, performing inertial navigation calculation on the slave inertial navigation system, and synchronously acquiring the speed and attitude information of the master and slave inertial navigation systems output in a grid system to obtain speed and attitude errors so as to constitute and measure Z; III, according to navigation mechanics arrangement in the grid system, combining a grid navigation error equation, adopting a 'speed + attitude' matching manner, and establishing a system state equation and a measurement equation in the grid system; and IV, performing unscented Kalman filtering calculation, estimating the attitude misalignment angle of the slave inertial navigation system and the speed state estimated value, and completing transfer alignment. According to the method disclosed by the invention, the problem that the large azimuth misalignment angle cannot be subjected to transfer alignment in the polar region environment is solved, and the transfer alignment accuracy and applicability of the polar region are improved.

The invention relates to a GPS/INS integrated navigation method based on unscented Kalman filtering. The method is mainly characterized in that the method comprises the following steps that INS data and corresponding GPS data are combined together, and an unscented Kalman filtering method is adopted to build a GPS/INS integrated navigation model to obtain the online optimal value of a GPS; the INS data are used as training data and are combined with the online optimal value of the GPS as a training target to build an SVM regression model, and the SVM regression model is used for predicting the off-line navigation data of the GPS. According to the method, the precision of the GPS is usually combined with the reliability of an inertial navigation system to achieve the information fusion of the filtering, the GPS and an INS, and when the GPS and the INS are used together, a precise filtering result can be obtained; an SVM is adopted to train INS data obtained when GPS signals are lost, the parameters of a kernel function and a penalty function in an SVM algorithm are optimized through a simulated annealing algorithm, and the precision of the off-line GPS is kept close to that of the online GPS.

The invention discloses a FastSLAM method based on particle proposal distribution improvement and adaptive particle resampling. The method comprises the steps that 1, a square root transformation unscented Kalman filter is utilized to estimate optimal particle proposal distribution, and the pose state of a robot is sampled; 2, a square root volume Kalman filter is utilized to update feature map information corresponding to each particle; 3, an adaptive particle resampling method based on relative entropy is utilized to determine the quantity of particles needed at the current moment; 4, the pose stat of the robot and guidepost feature map information are determined according to a particle set obtained after resampling. A traditional FastSLAM algorithm is improved from the two aspects of quality and quantity of the sampling particles at the same time, thus, the numerical stability and accuracy of the algorithm in the estimation process are enhanced, and the quality of the sampling particles is improved; in the particle resampling process, the least quantity of the needed particles is dynamically determined according to estimation uncertainty, and therefore the calculation efficiency of the algorithm is improved.

The invention discloses a method for estimating dynamic states of power generators on the basis of unscented particle filtering theories. The method includes utilizing fourth-order dynamic equations of the power generators as state equations for estimating the dynamic states of the power generators, simulating PMU (phasor measurement units) by the aid of power system analysis software to acquire measurement data of power angles, angular speeds and the like of the power generators, and creating measurement equations for the power generators; acquiring static estimation values at state estimation initial moments, utilizing the static estimation values as initial values for the power generators at dynamic state start moments, generating initial particles adjacent to the initial values, carrying out tracking filtering on state variables of the power angles, the angular speeds and the like of the power generators by the aid of unscented particle filtering algorithms to ultimately obtain estimation values of the state variables of the power generators. The method has the advantages that the quantity demands on the particles can be lowered, and the filtering accuracy and the computational efficiency of the method are superior to the filtering accuracy and the computational efficiency of the traditional particle filtering processes; the dispersibility of the particles is improved by the aid of the method, and accordingly the robustness of the method is superior to the robustness of the traditional particle filtering processes and unscented Kalman filtering processes.

The invention provides a method and device for positioning particles of a mobile robot. The method for positioning particles of the mobile robot comprises the following steps of: (a) estimating road signs by an unscented Kalman filtering (UKF) algorithm; (b) obtaining approximate distribution and position information of a plurality of sampling particles, and carrying out auxiliary positioning by adopting the unscented Kalman filtering algorithm; and (c) optimizing the sampled particles based on a genetic algorithm in a node. The method for positioning particles of the mobile robot has the advantages of: (1) estimating road signs by the UKF algorithm instead of the EKF (Extended Kalman Filter) algorithm in calculation of road sign estimation, and avoiding huge deduction of a Jocabian matrix; (2) disclosing the filtering and positioning method of particles distributed according to the UKF auxiliary advice; and (3) disclosing the concept of weight value based on topological node importance, improving diversity of particles, and preventing degradation of particles.

The invention discloses an HEO satellite-formation-flying automatic navigation method based on a star sensor and an inter-satellite link. The HEO satellite-formation-flying automatic navigation method includes the steps that two satellite-formation-flying structures and rail parameters are designed with HEO satellite earth observation as the task requirement, and according to a rail dynamic model of a primary satellite relative to a secondary satellite under a geocentric inertial coordinate system, an automatic navigation system state model is established; the theoretical light condition and the imaging condition which are required to be met when the secondary satellite is observed through the primary-satellite star sensor are proposed; the theoretical azimuth angle and the pitch angle of the secondary satellite relative to the primary satellite are calculated, the direction of the real optical axis of the primary-satellite star sensor and the theoretical direction are adjusted to be coincident, the secondary satellite is really observed, and an observation equation with the relative unit direction vector and the distance as observation quantities is established; the position and the speed of the satellite are estimated with Unscented Kalman filtering. By means of the HEO satellite-formation-flying automatic navigation method, errors of the relative positions of satellites can be effectively corrected, the relative navigation accuracy is improved, and the HEO satellite-formation-flying automatic navigation method is quite suitable for satellite-formation-flying automatic navigation.

The invention discloses an underwater robot state and parameter joint estimation method based on self-adaption unscented Kalman filtering (UKF). The method comprises building an expanding reference model of an underwater robot, and enabling the reference model to have a dynamical model of the underwater robot and a fault model of a thruster; adopting a main filter of the self-adaption UKF to deliver and update expanding states composed of poses and speed of the underwater robot state and faults of the thruster according to pose information detected by a position sensor, and timely estimating speed information of the underwater robot and fault message of the thruster; and simultaneously adopting an accessory filter of the self-adaption UKF to timely estimate noise information of a system according to innovation information of the main filter. The underwater robot state and parameter joint estimation method has good instantaneity, can conduct joint estimation on states and parameters of the system, and can achieve high estimation accuracy when priori information of process noise and measurement noise is unknown.

The invention discloses an unscented Kalman Filtering-based multisource error calibration method for a bionic polarization sensor. The method does not depend on high-precision instrument equipment; and by building a bionic polarization system model, multisource error calibration and compensation are realized by virtue of an unscented Kalman Filtering method. The method comprises the following steps of (1) performing multisource error analysis on the polarization light sensor, wherein multisource errors mainly include a sensor mounting error, a measurement noise and a scale factor; (2) selecting the multisource errors and a polarization azimuth angle as system states for building a system state model; (3) selecting a light intensity measurement value comprising the multisource errors as a system output for building a system measurement model; (4) establishing an experimental environment, and collecting experimental data; (5) designing an unscented Kalman Filter, and estimating the mounting error, the scale factor and the polarization azimuth angle; and (6) according to estimated values, compensating an actual measurement value of the polarization sensor. The method has the characteristics of low calibration cost, high efficiency and the like, and is suitable for quick calibration and compensation of the bionic polarization sensor.

The invention discloses an adaptive UKF (Unscented Kalman Filtering)-based lithium battery SOC (State Of Charge) estimation method which comprises the following steps: performing a quick calibration experiment on a lithium battery, thus obtaining an SOC and OCV (Open Circuit Voltage) relation curve; building a model for the lithium battery to be detected, identifying parameters of a battery model through battery charging and discharging experiments, and building a discrete state space model of a battery system; estimating a battery SOC through an improved UKF algorithm. The method disclosed by the invention solves the problems of reduction of filtering precision and even generation of filtering divergence possibly due to the uncertainty and an unknown noise statistical property of the battery model in the UKF algorithm.