35 results about "Robust filter" patented technology

The invention discloses a system state estimation method based on an improved nonlinear robust filtering algorithm and relates to the technical field of space navigation. The method is based on a Huber nonlinear regression algorithm, so that high filtering accuracy and stable filtering output can be obtained under the condition that a measuring output equation is nonlinear, and measuring noises are in non-Gaussian distribution.

The invention provides a target tracking method based on maximum correntropy cubature particle filter (MCCPF). The target tracking method based on MCCPF is characterized by completing state estimationduring the target tracking process by using the MCCPF algorithm; during the target tracking process, reconstructing a state equation and a measurement equation of target tracking into a non-linear recursion model, and performing processing by means of the maximum correntropy criterion; and in a framework of standard Particle Filter (PF), using MCCKF (Maximum Correntropy Cubature Kalman Filter) togenerate an importance probability density function required by PF, and according to the algorithm process of PF, acquiring estimation of the tracking target state so as to realize real-time trackingof the target. The invention provides a target tracking method based on maximum correntropy cubature particle filter can obtain better performance than the current particle filter, improved particlefilter and robust filter in the target tracking process in which outliers appear in measurement of noise.

The invention relates to an integrated navigation system recursion optimizing initial-alignment method based on polarization information. The method is characterized in that positions and speed output by a ship's inertial navigation system (SINS) are corrected with the help of a global position system (GPS), and azimuth angles output by the SINS are corrected by using polarizing angle information provided by a polarization sensor. A recursion robust filter is adopted to restrain model errors in an inertial device and filter interference like noise of a part of devices. The initial-alignment method has the advantages that compared with an SINS / GPS integrated navigation system without the polarizing angle information, alignment time is shortened on equal conditions. The integrated navigation system recursion optimizing initial-alignment method has the advantages of being fast in alignment, strong in anti-interference capability and the like, and the integrated navigation system recursion optimizing initial-alignment method can be applied to an initial-alignment process of integrated navigational devices for airplanes, steamships and vehicles.

Embodiments of the present invention employ robust filtering at each of a number of scales, or resolutions, within a signal, such as a still image or video sequence. In certain embodiments of the present invention, robust filtering comprises or includes a non-linear neighborhood operation at each scale in order to produce both a denoised, sharpened and contrast-enhanced signal and a correction signal at each scale. Correction signals and sharpened and contrast-enhanced signals are combined to produce a denoised, sharpened and contrast-enhanced output signal. Scale-appropriate enhancements at each scale or resolution, including denoising, sharpening and contrast enhancement, ameliorate artifacts introduced by re-enhancement, at a second scale or resolution, features previously enhanced at another scale or resolution and inhibit over enhancement.

The invention relates to a nonlinear filtering method based on polytope differential inclusion, which belongs to the technical field of system filtering and control. The method comprises the steps of describing a nonlinear filtering error system with a PLDI (Programming Language Design And Implementation) model, converting a nonlinear filtering algorithm design problem to a linear uncertain system robust filtering algorithm design problem, designing a dynamic equation for estimated error correction solving by a hybrid robust H2 / H-to-infinity filtering method, then designing a discrete nonlinear filtering equation by combining an EKF (Extended Kalman Filter) one-step prediction equation, and applying the discrete nonlinear filtering equation to a nonlinear discrete system to obtain the state estimation of the nonlinear discrete system in real time. According to the method, the nonlinear filtering design is simplified, the filtering gain is not required to be updated in real time, and the jacobian matrix is not required to be computed in real time in the implementation process, so that the computation amount is reduced greatly, the real-time performance of the nonlinear filtering is improved effectively, and the method is applicable to the design of a nonlinear filter.

The invention discloses a robust intelligent cooperative calibration method for an INS / DVL integrated navigation system. The method comprises the following steps: 1, establishing a traditional INS / DVLsystem model including a state equation and a measurement equation; 2, correcting the marine unmanned motion platform by adopting an alternate correction algorithm based on acoustic ranging; and 3, filtering an output abnormal value by using a robust Kalman filter based on statistical similarity measurement. The alternate correction algorithm based on underwater acoustic ranging does not need tobe supported by a reference station or a mother ship, so that the working range of the marine unmanned motion platform is expanded, and the positioning precision of the marine unmanned motion platformis further improved; and the calculated amount of the robust filter is smaller than that of an existing robust filter, the method has more advantages in navigation application with high sampling frequency, the robustness of a non-Gaussian scene caused by an abnormal value is higher than that of the existing robust filter, and the positioning precision of the method is superior to that of a traditional INS / DVL integrated navigation system.

The invention discloses a mixed interactive multi-model filtering method. The method comprises the following steps: firstly, building a system model and an observation model of an application object,building a total model set according to a change condition of noise statistic characteristics of external information, and describing current noise characteristics by adopting the adjacent three models; storing a robust filter real-time noise variance matrix in a historical sequence by adopting a sliding window, and adaptively selecting a proper model according to the smoothed noise statistic characteristics; setting an initial probability of each model and a mixed initial state and a covariance matrix of each filter, and according to the system model and the observation model, performing state estimation and covariance matrix updating processes, and performing model updating by adopting a Bayesian hypothesis testing method; and performing an output interactive process according to a weight, and outputting a final filtering result. By taking an inertia / Doppler combined navigation system as an example, the interference of noise sudden change type errors and outliers existing in Dopplerspeed measurement information can be effectively relieved.

The invention discloses a method for estimating an attitude angle of a rescuing wrecker based on robust filtering. The method comprises the following steps: firstly, constructing a kinematic model of the rescuing wrecker according to the working characteristics of the rescuing wrecker, and then utilizing a robust filtering algorithm to acquire accurate estimated values of a pitch angle and a lateral inclination angle of the rescuing wrecker, on the basis of the model. According to the method provided by the invention, the pitch angle and the lateral inclination angle of the rescuing wrecker under the working condition of higher gradient can be estimated under the condition that the statistical properties of the interference signal cannot be accurately known. The method has the advantages of high precision, low cost, excellent instantaneity, wide application scope, etc., and can be used for monitoring the posture of the rescuing wrecker and early warning for the dangerous posture.

The invention belongs to the field of satellite analysis, and especially relates to a method used for increasing satellite attitude determination precision. The method comprises following steps: a uncertainty system model of a satellite attitude determination system is constructed based on the relative installation error of a star sensor; the uncertainty system model is subjected to standardization so as to obtain a Robust filtering algorithm iteration structure form; the weighting matrix in the Robust filtering algorithm iteration structure form is determined based on estimated error least squares optimization; a novel Robust filtering algorithm is constructed based on the Robust filtering algorithm iteration structure form and the weighting matrix; the novel Robust filtering algorithm isadopted for data processing of satellite attitude measuring data. The method is capable of inhibiting influences of star sensor relative installation error on satellite attitude determination, and increasing satellite attitude determination precision.

The invention discloses a high-speed train set robustness filtering fault detection method considering event driving and output saturation and belongs to the signal processing field. The method comprises the following steps of establishing a high-speed train set state space model considering the output saturation and designing a trigger strategy; designing the robust filter of event driving and designing a filter gain in an off-line mode: setting the initial value of the filter, calculating a middle variable, calculating one step prediction state estimation error covariance upper bound, calculating the filter gain, calculating a state estimation error covariance upper bound, and calculating a state estimation value in an online mode: setting a state estimation initial value, calculating one step state prediction value, calculating information, and calculating a state estimation value; and designing a fault detection strategy, calculating a residual error, calculating a fault detectionstatistics amount, calculating a residual error secondary moment upper bound, calculating a fault detection threshold and setting a fault detection logic. By using the fault detection method, a systemfault can be detected in an online mode, and the safety and reliability of a high-speed train set are increased.

The invention discloses a robust filtering method of near space aerocraft transfer alignment model uncertainty. The method comprises the following four steps: step 1, based on working principle and characteristics of a near space aerocraft transfer alignment system, establishing a math platform misalignment angle error equation, a velocity error equation, a position error equation and an observation equation of the system; step 2, establishing a state equation and the observation equation of model uncertainty according to the error equation of the system; step 3, providing a state variable initial value (x0) and a forecast error variance matrix initial value (sigma 0|0), providing a sparse grid quadrature point set (zeta j, epsilon j; j=1, 2,... Np), and providing robust filtering parameters gamma and epsilon; and step 4, estimating system state by robust filtering, performing error correction for a secondary inertial navigation system, and finishing the transfer alignment process. The method provided by the invention is applicable for transfer alignment in the case that primary and secondary inertial navigation systems have model uncertainty when the near space aerocraft is in dynamic state.

The invention discloses a pipeline geographic position information measurement method based on thick-tail robust filtering and relates to the technology of pipeline geographic position information measurement. The method is characterized in that an inertia / odometer wheel combined positioning system is composed of an MSINS and an odometer wheel, the inertia / odometer wheel combined positioning system is driven by a system power device to operate in a pipeline to obtain sensor data related to the trend of a pipeline, and strapdown inertia calculation and dead reckoning are conducted respectively,a difference value between the strapdown inertial calculation position and the dead reckoning position is used as the measurement information, the position measurement outlier information caused by odometer wheel slipping, sliding faults and pipeline motion constraint condition failure is filtered out through a thick-tail robust filter, moreover, strapdown inertia calculation errors are estimated, pipeline geographic position information output is corrected, and thereby a system can provide the continuous and high-precision pipeline geographic position information.

Embodiments of the present invention employ robust filtering at each of a number of scales, or resolutions, within a signal, such as a still image or video sequence. In certain embodiments of the present invention, robust filtering comprises or includes a non-linear neighborhood operation at each scale in order to produce both a denoised, sharpened and contrast-enhanced signal and a correction signal at each scale. Correction signals and sharpened and contrast-enhanced signals are combined to produce a denoised, sharpened and contrast-enhanced output signal. Scale-appropriate enhancements at each scale or resolution, including denoising, sharpening and contrast enhancement, ameliorate artifacts introduced by re-enhancement, at a second scale or resolution, features previously enhanced at another scale or resolution and inhibit over enhancement.

The invention discloses a rescue vehicle overall gravity center plane position estimation method. Plane positions of the gravity center of the rescue vehicle itself and the gravity center of a rescued vehicle are measured respectively; the plane position of the overall gravity center is acquired through derivation of a resultant moment theorem; and on the basis, a robust filtering algorithm is used for calculating the accurate estimation value of the overall gravity center plane position. The method of the invention can be used for safety monitoring and dangerous condition early warning for the rescue vehicle, and has the advantages of high precision, low cost, good real-time performance and the like.

The invention belongs to the technical field of integrated navigation under non-ideal conditions, and particularly relates to an integrated navigation robust filtering method based on statistical similarity measurement. According to the method, the situation that normal DVL beam measurement information and large-error beam measurement information appear at the same time in an SINS / DVL tight combination system is considered, and aiming at the problem that normal measurement information is prone to being lost due to the fact that a robust filter of an existing combined navigation system processes the measurement information coarsely at present, a new robust filtering method is provided based on multi-dimensional measurement equation decomposition and statistical similarity measurement. According to the method, a statistical similarity measurement (SSM) theory is introduced while a multi-dimensional measurement equation of an SINS / DVL tightly integrated navigation system is decomposed, so that the measurement noise variance of each beam is assisted to complete respective self-adaptive updating when a large measurement error occurs, and finally, the independence of measurement information processing of each beam is ensured. The method can be used in the field of underwater vehicle integrated navigation under non-ideal conditions.

Embodiments of the present invention employ robust filtering at each of a number of scales, or resolutions, within a signal, such as a still image or video sequence. In certain embodiments of the present invention, robust filtering comprises or includes a non-linear neighborhood operation at each scale in order to produce both a denoised, sharpened and contrast-enhanced signal and a correction signal at each scale. Correction signals and sharpened and contrast-enhanced signals are combined to produce a denoised, sharpened and contrast-enhanced output signal. Scale-appropriate enhancements at each scale or resolution, including denoising, sharpening and contrast enhancement, ameliorate artifacts introduced by re-enhancement, at a second scale or resolution, features previously enhanced at another scale or resolution and inhibit over enhancement.

The invention discloses a target estimation method based on multi-source information fusion, and the method comprises the steps: respectively observing the same target at the same time through a plurality of aircrafts, and obtaining the sight angle information between each aircraft and the target, the position information of each aircraft relative to the target, and the speed information of each aircraft relative to the target in real time; and sharing the optimal estimation state of the target obtained by each aircraft, and calculating the state information of the target at the next moment through a linear distributed consistent robust filtering algorithm.

The embodiment of the invention discloses an adaptive model online reconstruction robust filtering method, device and system. The method comprises the steps that a model set, a Markov probability transfer matrix and the model probability of each model in the model set are initialized; obtaining an input state quantity and a state covariance matrix of each model; filtering and fusing the data information acquired by the integrated navigation system by using a robust filter, and outputting a state vector corresponding to each model; updating the model probability of each model; and mixing the covariance matrix of each model, the state vector corresponding to each model and the updated model probability to obtain the state vector and the state covariance matrix of the carrier. According to the method, the robust filter is utilized to perform filtering fusion on the integrated navigation system, so that adverse effects caused by measurement of outliers can be effectively resisted, and the total size of the interactive multi-model preset model set is reduced.

The invention discloses a vehicle active safety control method and device. By establishing a state equation and an observation equation of robust filtering, a state vector and an observation vector are obtained, and the state vector and the observation vector are recursively calculated by filtering to obtain the first discrete moment The third vehicle mass and the third road longitudinal gradient, based on the third vehicle mass and the third road longitudinal gradient at the discrete time, the vehicle active safety control operation is performed. Through the above scheme, based on the robust filtering algorithm, the joint estimation of vehicle mass and road longitudinal gradient enables real-time and accurate acquisition of high-precision vehicle mass and road longitudinal gradient in the case of interference signals. By obtaining high-precision vehicle mass Quality and longitudinal slope of the road, and perform vehicle active safety control operations, thereby improving the control effect of vehicle active safety systems. In addition, while ensuring the high precision of vehicle quality and road longitudinal slope, it meets the application requirements under complex working conditions.

The invention relates to a spacecraft autonomous navigation robust filtering algorithm based on sequence image state expression, and belongs to the technical field of autonomous navigation. The method comprises the following steps: 1, constructing a spatial non-cooperative target autonomous relative navigation system state equation; 2, constructing a monocular camera sequence image measurement equation; 3, constructing a state expression model based on a sequence image, separating a measurement value from a measurement error value by using a trigonometric function formula, and obtaining a measurement noise expectation matrix epsilon k according to measurement error distribution; 4, obtaining a one-step state prediction value at the moment k and a one-step prediction covariance matrix Pkk-1 at the moment k; 5, calculating to obtain an optimal gain Kk of state estimation; step 6, calculating to obtain a state estimation value xkk and a state estimation covariance matrix Pkk at the moment k; according to the method, the anti-interference performance on the inaccurate initial value of the system state can be improved, so that a state estimation result with higher precision is obtained.

The invention relates to the technical field of positioning, and discloses a positioning method based on single-step robust filtering of course angle information of an electronic compass, and the method comprises the following steps: setting an initial position of a pedestrian; starting a pedestrian dead reckoning system according to the initial position of the pedestrian; acquiring sensor data through the mobile terminal, and storing the compass sensor data into a compass observation data storage file; carrying out the gait recognition according to accelerometer sensor data; carrying out single-step electronic compass data sequence robust filtering processing; and determining whether continued positioning is required or not. According to the method, single-step filtering processing is carried out on the course angle data sequence of the electronic compass, the influence of observation gross errors caused by magnetic field abnormality is weakened, noise in course angle observation datais reduced, and the robustness of a course angle filtering result is improved; the more reliable and stable course angle filtering value is used for track calculation, so the calculation error is greatly reduced, the jump phenomenon of the PDR track is weakened, and the precision and continuity of the PDR positioning result are improved.

The invention relates to a design method of a distributed k-order filter for communication topology random switching, and the method comprises the following steps: in a wireless sensor network environment, aiming at a linear time-varying system, modeling communication network random switching of the filter by utilizing a Markov theory, and designing a distributed robust k-order filter with topological structure random switching; regulating and controlling the order of the filter network by a parameter E to obtain a filter network distributed full-order filter and a distributed reduced-order filter with communication topology random switching; enabling the distributed filter of the linear time-invariant system to calculate the filtering gain of the whole network offline at one time, and calculating the gain of the distributed robust filter of the time-varying system online in real time by a group of LMIs. The method has the advantages that 1, a distributed robust k-order filter with topological structure random switching is designed; 2, the mean square index is stable, and the error peak value of the filter network does not exceed a given value;.

The invention relates to the technical field of filter design, and discloses a method for designing a filter of a multi-birth uncertain singular network system, which comprises the following steps of: exporting a network model of the filter of the multi-birth uncertain singular network system; wherein a relaxation matrix variable proves that the network model is permissible in performance; proving that the network model is asymptotically stable, regular and causal through a Lyapunov function method; wherein a sufficiency criterion that the error model is allowed and has an H infinity performance index is given; applying filter equivalence method, designing a gain matrix of a corresponding filter under event driving to acquire an event-driven robust filter and the method greatly reduces communication resources, has certain anti-interference capability, greatly reduces broadband resources and operation cost, and further greatly saves network communication cost.

The invention discloses a high-speed train set robustness filtering fault detection method considering event driving and output saturation and belongs to the signal processing field. The method comprises the following steps of establishing a high-speed train set state space model considering the output saturation and designing a trigger strategy; designing the robust filter of event driving and designing a filter gain in an off-line mode: setting the initial value of the filter, calculating a middle variable, calculating one step prediction state estimation error covariance upper bound, calculating the filter gain, calculating a state estimation error covariance upper bound, and calculating a state estimation value in an online mode: setting a state estimation initial value, calculating one step state prediction value, calculating information, and calculating a state estimation value; and designing a fault detection strategy, calculating a residual error, calculating a fault detectionstatistics amount, calculating a residual error secondary moment upper bound, calculating a fault detection threshold and setting a fault detection logic. By using the fault detection method, a systemfault can be detected in an online mode, and the safety and reliability of a high-speed train set are increased.

The present invention relates to a filter element (1) for a filter unit (30) for filtering a fluid, wherein the filter element (1), so as to be permeable to the fluid, consists of at least one structural body (6, 7) and a support body (5) and has a large number of apertures / pores (17). The structural boy (6, 7) and the support body (5) are each produced by means of a 3D printing process, wherein the structural body (6, 7) has a repeating three-dimensionally crosslinked lattice structure (15) and the support body (5) has a cell structure (10). The apertures / pores (17) of the lattice structure (15) are formed by inter-crosslinked and rigidly connected rod elements (16). The problem addressed by the filter element (1) is that of providing a new optimal and robust filter structure having a good filtering effect and a good deep loading capacity using a 3D printing process. The present invention is used to filter fluids in technical modular units / machine units or motors.

The invention discloses a robust filtering method of near space aerocraft transfer alignment model uncertainty. The method comprises the following four steps: step 1, based on working principle and characteristics of a near space aerocraft transfer alignment system, establishing a math platform misalignment angle error equation, a velocity error equation, a position error equation and an observation equation of the system; step 2, establishing a state equation and the observation equation of model uncertainty according to the error equation of the system; step 3, providing a state variable initial value (x0) and a forecast error variance matrix initial value (sigma 0|0), providing a sparse grid quadrature point set (zeta j, epsilon j; j=1, 2,... Np), and providing robust filtering parameters gamma and epsilon; and step 4, estimating system state by robust filtering, performing error correction for a secondary inertial navigation system, and finishing the transfer alignment process. The method provided by the invention is applicable for transfer alignment in the case that primary and secondary inertial navigation systems have model uncertainty when the near space aerocraft is in dynamic state.

The invention relates to a no-gyro satellite gesture determination method based on tensor product multi-cell robust heavy hydrogen (H2) filtering, and belongs to the technical field of aircrafts. As for the nonlinear characteristics of satellite gesture dynamic equations and kinematical equations, the no-gyro satellite gesture determination method based on the tensor product multi-cell robust H2 filtering provides the multi-cell robust H2 filtering based on tensor product conversion to transform a nonlinear filtering problem into a linear filtering problem. The no-gyro satellite gesture determination method based on the tensor product multi-cell robust H2 filtering includes the steps of first establishing a state equation and a star sensor measurement equation of a gesture determination system and transforming a nonlinear system into a linear variable parameter error system by utilizing Jacobian linearization, then establishing linear parameter varying (LPV) system multi-cell model description according to tensor product model transformation and acquiring state estimation correction amount of the gesture determination system through combining with the robust H2 filtering, and last correcting gesture single-step pre-measurement acquired through a extended Kalman filter (EKF) method by utilizing the estimation correction amount to obtain a gesture estimation value. Accordingly, real-time updated filtering gain in the EKF method is avoid, and filtering calculated amount is reduced greatly.

The invention discloses a robust filter-based fault detection method for a positive linear network control system. The method comprises the following steps of: establishing a discrete model of the positive linear network control system according to the current situation of the network control system with time delay and uncertainty, and designing a corresponding robust filter; establishing an augmented system according to the discrete model and the filter so as to generate a residual error; scaling upper and lower bounds of uncertain parameters to derive the augmented system as a positive system, wherein the augmented system is stable and meets the sufficient condition of 1 / l (robustness / fault sensitivity) performance; and solving filter parameters by an iterative algorithm and combining an optimization target, namely realizing fault detection of the positive linear network control system while considering the robustness and the fault sensitivity. By the iterative algorithm, the application range is greatly widened, the fault detection efficiency is improved, and the method can be applied to various sensor-based / sensorless fault systems with uncertain parameters.