79 results about "Sigma point" patented technology

A method of estimating the navigational state of a system entails acquiring observation data produced by noisy measurement sensors and providing a probabilistic inference system to combine the observation data with prediction values of the system state space model to estimate the navigational state of the system. The probabilistic inference system is implemented to include a realization of a Gaussian approximate random variable propagation technique performing deterministic sampling without analytic derivative calculations. This technique achieves for the navigational state of the system an estimation accuracy that is greater than that achievable with an extended Kalman filter-based probabilistic inference system.

A method of estimating the navigational state of a system entails acquiring observation data produced by noisy measurement sensors and providing a probabilistic inference system to combine the observation data with prediction values of the system state space model to estimate the navigational state of the system. The probabilistic inference system is implemented to include a realization of a Gaussian approximate random variable propagation technique performing deterministic sampling without analytic derivative calculations. This technique achieves for the navigational state of the system an estimation accuracy that is greater than that achievable with an extended Kalman filter-based probabilistic inference system.

A pulse oximeter system comprises a data processor configured to perform a method that combines a sigma point Kalman filter (SPKF) or sequential Monte Carlo (SMC) algorithm with Bayesian statistics and a mathematical model comprising a cardiovascular model and a plethysmography model to remove contaminating noise and artifacts from the pulse oximeter sensor output and measure blood oxygen saturation, heart rate, left-ventricular stroke volume, aortic pressure and systemic pressures.

The invention discloses an electrical power system dynamic state estimation method based on an unscented transformation Kalman filter, which comprises the following steps: selecting a sigma point sampling strategy according to statistics of input variable to obtain a sigma point set of the input variable and corresponding weight; secondly, performing non-linear transformation on each sigma point in the sampled sigma point set of the input variable to obtain a transformed sigma point set; and finally, performing weight processing on the transformed sigma point set to obtain the statistics of output variable. The method reduces the application difficulty, and simultaneously can obtain higher accuracy and robustness.

A strong maneuver-based target tracking method comprises the following steps: initializing parameters, interactively inputting models, judging the covariance matrix, filtering in parallel, updating the model probabilities, interactively outputting the models, filtering in a fixed-delay smoothing manner, and judging whether the state updating is completed; on the basis of the IMM algorithm, the IMM algorithm for recalculating the weight is used, that is, RIMM. The method not only uses the model probabilities, but also takes full advantage of the filtering covariance matrix, so that the tracking accuracy is higher. In addition, an SRCK method is used in the filter forecasting stage, and adopts the spherical integrating criterion and the radial integrating criterion. Compared with a wide UKF algorithm in nonlinear filtering, the target tracking method optimizes the sigma point sampling strategy and the weight distribution in the UKF. Meanwhile, QR factorization is introduced in SRCKF, so that the matrix squiring operation is avoided, and the filtering stability is improved. On the basis, the filtering in a fixed-delay smoothing manner is also introduced, so that the real-time property and the accuracy of target tracking are further improved.

A method of the present invention using a prediction process including a battery equivalent circuit model used to predict a voltage and a state of charge of a battery. The equivalent circuit battery model includes different equivalent circuit models consisting of at least an ideal DC power source, internal resistance, and an arbitrary number of representative parallel resistors and capacitors. These parameters are obtained a priori by fitting the equivalent circuit model to battery testing data. The present invention further uses a correction process includes determining a corrected predicted state of charge of the battery; and storing the corrected state of charge of the battery in a storage medium. In the present invention, an expectation of the predicted voltage of the battery and an expectation of the predicted state of charge of the battery are obtained by an unscented transform with sigma points selected by a Gaussian process optimization.

The invention discloses a nonlinear-model-based SINS / DVL (strapdown inertial navigation system / doppler velocity log) integrated navigation method. The method comprises the following steps of: building a quaternion-based SINS nonlinear speed, posture and position error model according to the working principles of an SINS and a DVL, and confirming an error model of the DVL; building a state equation of the systems according to the error models of the two systems, measuring by taking the difference between the actually-measured speeds of the SINS and the DVL as the quantity, and building a measurement equation of the system; discretizing an actual continuous system to obtain a discrete nonlinear model which is convenient to compute; initializing the system, and computing sampling points and corresponding weight numbers by the discrete nonlinear model and unscented conversion; and sequentially carrying out time update and measurement update on unscented kalman filter on the basis of the discrete nonlinear model according to the constructed Sigma point. After the SINS / DVL integrated navigation system is used, information of each subsystem can be effectively used, and the best of each subsystem can be taken, so that the overall positioning accuracy can be greatly improved; the UKF (unscented kalman filter) estimation can be carried out by the nonlinear model of the SINS / DVL integrated navigation system, so that the positioning error of the system can be effectively reduced and the accurate positioning of the navigation system can be realized better.

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 an FPGA (Field Programmable Gata Array)-based unscented kalman filter system mainly solving the problem that the traditional unscented kalman filter hardware has great implementation difficulty and poor instantaneity and comprising a covariance matrix Cholesky decomposition model A, a Sigma point generation module B, a time updating module C, an observation and prediction module D, a part-mean value and covariance matrix computation module E, a population mean value computation module F, a population covariance matrix computation module G, an observation and predictioncovariance matrix inversion module H, a gain computation module I and a state quantity and state covariance matrix estimating module J, wherein the module A generates K group of column vector to the module B and the B, C, D and E modules are connected in series and respectively comprise K submodules adopting a parallel arithmetic modular construction; the F and G modules receive and process the Kgroup of results of the module E and the processed results pass through the modules H, I and J in sequence to obtain the present result. The invention has the advantages of quick filter speed and easy hardware implementation and can be used for target tracking and parameter estimation.

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 an image scene type discrimination method based on covariance features, wherein based on local Sigma point features of a covariance matrix of an object of interest, a semantic model is established for scenes, and the scenes are classified. The method comprises that on the basis of the interactive image segmentation, the object of interest, which is a typical object when a user selects such a scene, is extracted from the scene; the covariance matrix which represents the object of interest is formed by combining pixel positions, colors, Gabor features and LBP features; and the covariance matrix is converted into Sigma point features in an European-style space, and learning and discrimination of a scene classifier is completed via a support vector machine.

The invention discloses a network flow estimating method which concretely comprises the steps of initializing, obtaining a sigma point set, predicting the state, estimating the state and updating the process. In the method, the flow matrix estimating problem is modeled into a nonlinear system which more meets the OD (Origin-Destination) flow real characteristic; the sigma point set obtained by UT (Unscented Transformation) is subjected to nonlinear transformation; the coefficient of the system sate equation needed by nonlinear transformation is obtained concretely by Chebyshev polynomial fitting instead of approximately obtaining the system state equation by the traditional local linearization; accordingly, the state equation of the system does not need to meet the available linear function approximating condition; a symmetrical sampling policy is adopted for the UT; the particle point set approaches the probability density function distribution of a nonlinear function to obtain the higher-order approximation of state estimation, the result of OD flow estimation has higher precision, and the calculating complexity of the system is reduced.

The invention belongs to the technical field of fault diagnosis and forecasting in the chemical industry production process and relates to a dynamic process fault forecasting method based on fuzzy self-adaptive prediction. Fault parameters and state variables are sequentially determined based on a fuzzy self-adaptive unscented Kalman prediction method, and whether future multi-step prediction is needed or not is determined according to fault parameter standard deviations; then, a future multi-step prediction value is worked out; finally, the fault prediction probability is worked out through sigma points of output variables, and fault prediction is carried out according to the fault prediction probability. The whole technical process is simple, principles are reliable, calculating parameters are accurate, fault state estimation precision is high, the application range is wide, logicality is strong, and the method is environmentally friendly.

The invention discloses a self-adaptive kalman filter estimation algorithm for state of charger (SOC) for a power battery. According to the algorithm, the weighted sum of residual errors between output values measured each time and output values obtained by estimation and residual errors of the output values obtained by sigma points under various states can be obtained, then the weighted sum is regarded as an innovation to estimate the current noise covariance, the current noise covariance can be reduced along with the time, and the real-time feedback can be performed. Proved by theoretic and actual data validation, the algorithm can greatly improve the estimated accuracy of SOC.

The invention discloses a driving behavior analysis method based on a three-axis accelerometer. The three-axis accelerometer is arranged on an automobile. The method comprises the following steps that a plurality of sets of three-axis accelerations are collected; low-pass filtering is carried out on the multiple sets of three-axis accelerations in a moving average manner; the average value of the multiple sets of three-axis accelerations obtained after filtering is calculated; the average value and expectation values (0,0,g) are subjected to subtraction to work out the static error, and the three-axis accelerations obtained after low-pass filtering and the static error are subjected to subtraction to eliminate the static error; the average value of the multiple sets of three-axis accelerations without the static error is calculated; sigma point moving average window filtering is carried out on the characteristic quantity; and the maximum value among the absolute values of the characteristic quantities exceeding the preset value and including Hx, Hy and Hz is judged so that the corresponding driving behavior can be judged, the sum acceleration is not sensitive to the intermediate frequency or the high frequency, data are locked when the sum acceleration exceeds the preset value, and therefore multiple times of misjudgment caused by high-frequency shaking is eliminated, and the accuracy of the driving behavior is improved.

The invention discloses a driver eye movement characteristics handover detecting and tracing method based on light sensing, comprising the following steps: detecting illuminance by a light sensor, positioning a high illuminance condition by a Harr human eye detection method, and positioning a low illuminance condition by a bi-dimensional orthotropic Log-Gabor human eye detection method started by a computer; maintaining all-day position without influence of illuminance; and adding two dimensional parameters of system noise and observation noise on the basis of an uniform model after positioning, performing logarithmic transform on the two dimensional parameters, establishing a human eye non-linear model and a sampling strong tracking non-linear filter, adjusting system process noise and observation noise covariance matrix by fading factors, improving the non-linear strong tracking capability and model state robustness of a UKF filter, adapting to driver normal emergent operation reaction, and improving real time property of an algorithm by adopting a sampling strategy that (n+2) numbered sampling points perform strong tracking on Sigma point in the UKF filter.

The invention provides a novel adaptive Kalman attitude calculation method for unmanned aerial vehicles, comprising: selecting an NED (north east down) right-handed coordinate system to create a discrete time-space model for a unmanned aerial vehicle dynamic attitude calculation system; determining a Sigma point set according to the dimensionality of a state variable; calculating weight value of the Sigma point set; calculating one-step prediction of the Sigma point set; subjecting a prediction value of the Sigma point set to weighted summation to obtain one-step prediction and covariance matrix of a system state quantity; generating a new Sigma point set via UT conversion according to the one-step prediction value; substituting the new Sigma point set to an observation equation in the discrete time-space model to obtain an observed quantity of prediction; subjecting the observed quantity of prediction to weighted summation to obtain a prediction average and covariance of the system; updating a Kalman gain matrix and the state and covariance of the system.

The invention discloses a target tracking system and tracking method applied to a driverless car. The tracking system comprises a vehicle-borne radar system and a robot operating system (ROS), whereinthe vehicle-borne radar system is composed of a laser radar and a millimeter wave radar. The laser radar is mounted at the top of the car and has relatively high measurement accuracy, and but is susceptible to environmental factors such as illumination, the millimeter wave radar is installed at a license plate at the front of the car, has relatively low measurement accuracy, but is just slightlyaffected by external environment, and therefore, the vehicle-borne radar system integrating the laser radar and the millimeter wave radar combined can improve measurement accuracy; and the node set ofthe robot operating system (ROS) includes a radar sensor node and an unscented Kalman filter estimator node. According to the unscented Kalman filter algorithm, a series of sigma points are adopted to generate states through nonlinear transformation, and the estimated sigma points are used to cover state estimation points and covariances, and therefore, the fusion of the data received by the radar system can be realized, and target tracking can be achieved.

The invention discloses an unscented Kalman filtering and neural network-based photovoltaic power generation prediction method. The method comprises the steps of 1: taking illumination and temperature as input signals u(k) and taking active power and reactive power as output yk; 2: establishing a BP neural network and taking a weight coefficient and a threshold of the neural network as state variables xk; 3: initializing the state variables; 4: calculating a Sigma point; 5: improving state updates and error variance updates of time updates in Kalman filtering; 6: calculating covariances of the state variables and measurement variables; 7: measuring updated state updates and error covariance updates in the Kalman filtering; and 8: judging whether an updated state variance matrix is converged or not. The method has high calculation speed and high prediction precision, and can be adaptive to dynamic changes under different weather conditions; and a prediction model built in the method has wider adaptability to the weather conditions.

A method of the present invention using a prediction process including a battery equivalent circuit model used to predict a voltage and a state of charge of a battery. The equivalent circuit battery model includes different equivalent circuit models consisting of at least an ideal DC power source, internal resistance, and an arbitrary number of representative parallel resistors and capacitors. These parameters are obtained a priori by fitting the equivalent circuit model to battery testing data. The present invention further uses a correction process includes determining a corrected predicted state of charge of the battery; and storing the corrected state of charge of the battery in a storage medium. In the present invention, an expectation of the predicted voltage of the battery and an expectation of the predicted state of charge of the battery are obtained by an unscented transform with sigma points selected by a Gaussian process optimization.

The invention discloses a method for acquiring available power transmission capability of a power system with a wind power plant. Nonlinear transformation of unscented transformation is applied to calculation of probability available power transmission capability of the power system with the wind power plant, the problem that the air speed relevancy of an adjacent wind power plant is difficult to process in the calculation process is solved, and the calculation efficiency is improved under the premise that the accuracy is ensured. A series of Sigma points are selected according to a certain rule, infinite points are utilized to express the whole sample space, and an interior point method and an optimal power flow method are utilized to conduct definitive solution calculation on each sample, probability indexes such as the expectation and the standard difference of the probability available power transmission capability of the power system with the wind power plant are obtained, and the abundant probability indexes can provide reference to the operating personnel.

The invention provides a smoothing constraint unscented Kalman filtering method. The smoothing constraint unscented Kalman filtering method comprises the following steps: 1, acquiring an original prior probability density function of a target state at a current target observation moment according to unscented transformation; 2, calculating a mean value and a variance of an original prior probability through numerical expectation; 3, introducing noise constraint information, and calculating the center of the approximate feasible region to obtain corrected prior probability density; 4, seeking aGaussian distribution mean value and a variance meeting the constraint condition through posteriori iterative optimization, and generating a new Gaussian sigma point meeting the constraint condition;and 5, carrying out weighted calculation on the Gaussian sigma points to complete a filtering process. The smoothing constraint unscented Kalman filtering method has the advantages in the aspects ofaccuracy and robustness, and meanwhile, the real-time performance of the smoothing constraint unscented Kalman filtering method is superior to that of a particle filtering algorithm. Correspondingly,the invention further provides a target tracking method.

The invention provides a particle filtering method based on spherical simplex unscented Kalman filter. The particle filtering method comprises the steps of: initializing a particle and a weight value thereof; generating a particle through importance sampling; updating and normalizing the particle weight value; sampling again; outputting a result; and entering a next time step. The invention mainly improves an importance sampling step and obtains importance probability density by adopting a SSUKF (Spherical Simplex Unscented Kalman Filter) algorithm based on SSUT (Spherical Simplex Unscented Transformation). Compared with PF (Particle Filter), EKPF (Extended Kalman Particle Filter) and standard unscented particle filter, SSUPF (Spherical Simplex Unscented Particle Filter) can acquire the precision equivalent to UPF (Unscented Particle Filter). On the other hand, because the SSUT adopts sampling points, i.e. sigma points, which are distributed in a spherical way, the quantity of the sampling points is far less than the UT (Unscented Transformation), and the advantage on the aspect of computing efficiency is gradually obvious in a high dimensional system.

The invention provides a lithium ion battery service life prediction method based on DAUPF. The method comprises the following steps of in a sampling portion, adding a double-adaptive factor based ona UKF algorithm, and using a state value and a covariance obtained after one-step prediction of a Sigma point set as a guide, carrying out another UT conversion to get a new Sigma point set, bringinginto an observation equation and acquiring a new observed quantity so as to obtain a first cycle sample mean value and the covariance; in an improved UKF algorithm portion, updating an adaptive factorafter primary circulation is completed, and then carrying out next improved UKF algorithm cycle; after sampling is completed, entering into a PF process, and after one output prediction value is obtained, updating another adaptive factor to complete a DAUPF process; and finally predicting test data. In the invention, a UPF algorithm sampling portion is improved, through adding the double-adaptivefactor, the algorithm has higher robustness, through two-step UT conversion, the adaptive factor can be better integrated into the algorithm, and a algorithm prediction effect is accurate.

The invention provides a strong tracking UKF filter method based on sampling point changing. The strong tracking UKF filter method based on the sampling point changing comprises the steps that (1) initial parameter setting is carried out on a system; (2) Sigma points are sampled according to an orthogonal transformation sampling point method, a corresponding prediction equation is determined, and time and measurement are updated; (3) fading factors are calculated; (4) new one-step prediction covariance is calculated by using the fading factors, the Sigma points are recalculated, and auto-covariance and cross covariance after the fading factors are introduced are obtained through nonlinear measurement function propagation; (5) filter updating is carried out to the end. According to the strong tracking UKF filter method based on the sampling point changing, the problem of non-local sampling of the system is solved effectively, the precision of the system is improved, and the system is made to have certain strong tracking capability. The strong tracking UKF filter method based on the sampling point changing can be used for solving the problems of poor robustness and filtering divergence when a model of the system is uncertain, solves the problem of the non-local sampling in the high-dimensional system, and expands the application range of strong tracking filter. In an MEMS / GPS combined navigation system, the positioning and attitude determination performance of the MEMS / GPS combined navigation system can be improved through the method.

Disclosed herein are an apparatus for integrating multiple rate systems and a method of operating an apparatus for integrating multiple rate systems. In the method of operating the apparatus, navigation information is calculated through an inertial measurement unit, and mean values and variances of initial state variables of the navigation information are set. Sigma points are calculated using the mean values and the variances. The mean values are time-propagated until measurement information is input through a Global Positioning System (GPS). When the measurement information is input, the sigma points are time-propagated at intervals of a frequency of the measurement information. Variances are calculated using the time-propagated sigma points. The navigation information is updated using the time-propagated mean values, the calculated variances, and the measurement information.

The invention relates to a model parameter distinguishing method. The method comprises the steps that an intermediate variable and parameters to be distinguished in the change process of a study object are selected as system state variables, and the relevant parameters are initialized; a system state change model and a resilience measuring model are set up; a system state predicated value of the next time point is calculated according to the state change model and the original values of all the system state variables; the resilience measured and predicted value and system resilience measured and predicted value of each sigma point are calculated; unscented Kalman filter gain is calculated; a system state optimizing estimated value is calculated; the state change model is updated. A Bouc-Wen model is pulled away from a specific civil structure model, data is adopted as drive to solve the accurate distinguishing problem of the model parameters, universality is achieved, the method can be expanded and applied to the distinguishing problem meeting or approximately meeting Bouc-Wen models among various kinds of relevant variables.

The invention discloses a stochastic optimization scheduling method for a power grid in an emergency situation. The method comprises the steps of firstly, building an uncertainty model with wind poweroutput and an interruptible load response and setting a motivation factor and a penalty factor for an interruptible load to constrain an interrupt behavior of the interruptible load; secondly, with the goal of minimum total cost of stochastic optimization scheduling in the emergency situation, building a stochastic optimization scheduling model, in which the interruptible load participates, for the power grid in the emergency situation, and then generating a certain number of Sigma point sets in a multi-dimensional space by using an unscented transformation method; and finally solving a certainty optimization scheduling result of the power grid in the emergency situation on each Sigma sampling point and carrying out analysis and evaluation according to the statistical characteristics. Theinvention provides an effective, practical and scientific stochastic optimization scheduling method for the power grid in the emergency situation. The calculation accuracy and calculation cost of themodel are taken into account, guarantee of electric power and energy balance between supply and demand is facilitated and safe and stable operation of the system is maintained.