462 results about "Optimal estimation" patented technology

Neural networks for optimal estimation (including prediction) and / or control involve an execution step and a learning step, and are characterized by the learning step being performed by neural computations. The set of learning rules cause the circuit's connection strengths to learn to approximate the optimal estimation and / or control function that minimizes estimation error and / or a measure of control cost. The classical Kalman filter and the classical Kalman optimal controller are important examples of such an optimal estimation and / or control function. The circuit uses only a stream of noisy measurements to infer relevant properties of the external dynamical system, learn the optimal estimation and / or control function, and apply its learning of this optimal function to input data streams in an online manner. In this way, the circuit simultaneously learns and generates estimates and / or control output signals that are optimal, given the network's current state of learning.

Disclosed is a system and method for onboard optimal estimation of heading, pitch, and roll through real-time measurement of magnetic field, acceleration and angular motion in three dimensions. Magnetometer information is used to create an initial reference from which movement is measured. Thus, the process does not have to start when the body is in a known position. Further, the device does not have to continually rely on accelerometer data to get roll and pitch. To do this magnetic field data is used to complement gyro information. The magnetic data is used to estimate pitch, roll, and heading.

Disclosed is a method of estimating repeatable runout (RRO) for a disk drive. The method includes: determining a plurality of position error signal (PES) values for a plurality of servo wedges of a track of a disk; and estimating repeatable runout (RRO) in the PES values for the servo wedges of the track by averaging determined PES values over a pre-determined number of revolutions of the disk of the disk drive and scaling the averaged PES values by a pre-determined ratio of the variance of non-repeatable runout (NRRO) to the variance of an optimal estimation of the RRO for the disk drive.

A method, computer program product, and system to estimate the number of distinct value combinations for a set of attributes in a database system is disclosed. The method comprises utilizing frequency information within the set of attributes to provide a best estimate for the number of distinct value combinations. In a preferred embodiment, the utilizing step comprises estimating the number of distinct value combinations utilizing frequency information of the set of attributes based upon probability theory and further includes providing bounds on the distinct value information combinations utilizing the frequency information of the set of attributes. In so doing, an estimate for the number of distinct value combinations is provided.

Neural networks for optimal estimation (including prediction) and / or control involve an execution step and a learning step, and are characterized by the learning step being performed by neural computations. The set of learning rules cause the circuit's connection strengths to learn to approximate the optimal estimation and / or control function that minimizes estimation error and / or a measure of control cost. The classical Kalman filter and the classical Kalman optimal controller are important examples of such an optimal estimation and / or control function. The circuit uses only a stream of noisy measurements to infer relevant properties of the external dynamical system, learn the optimal estimation and / or control function, and apply its learning of this optimal function to input data streams in an online manner. In this way, the circuit simultaneously learns and generates estimates and / or control output signals that are optimal, given the network's current state of learning.

The invention provides a positioning service method; firstly, a positioning management platform is connected with a plurality of LEs (logic elements) and LIUs (line internet units) through a communication network, wherein the LEs have the functions of positioning measurement, wireless sensing and communication, the LIUs have the functions of the wireless sensing and the communication; the LEs report positioning measurement data to the positioning management platform in a timing manner, and the positioning management platform obtains an optimal estimation of LE positions through an information fusion technology; when approaching a detection range of a wireless sensing unit of the LIUs, the LEs can be automatically identified, wherein an identification number ID-1 can be automatically read by the LIUs and is reported to the positioning management platform, and the positioning management platform corrects position data corresponding to the LEs. In the method of the invention, the data processing of the LEs can be greatly simplified, the power consumption is reduced, various positioning value-added services can be favorably developed, and particularly, the positioning services with requirements of low power consumption and hidden nature to the LEs can be favorably developed.

A system able to locate and identify aircraft and vehicles based on the reception and processing, with novel means and methods, of signals emitted by the transponder of the secondary surveillance radar, shortly SSR. The system has a number of fixed stations distributed in the area of interest, e.g. in the airport area; any signal (the well known SSR reply / squitter) transmitted by the on-board transponder is received by four or more stations and the measurement of three or more differences of times of arrival (TOA) permits the reconstruction of the position of the transponder in spite of the fact that the transmission time is unknown. Suitable algorithms based on optimal estimation enhance both the accuracy of TOA measurements and the accuracy of the reconstructed position. The effects of possible overlapping of signal in time are avoided or mitigated by multiple source separation techniques based on least squares algebraic processing.

The invention relates to an autonomous integrated navigation system which belongs to the technical field of navigation systems. The SINS (Strapdown Inertial Navigation System) / SAR (Synthetic Aperture Radar) / CNS (Celestial Navigation System) integrated navigation system takes SINS as a main navigation system and SAR and CNS as aided navigation systems and is established by the following steps: firstly, designing SINS / SAR and SINS / CNS navigation sub-filters, calculating to obtain two groups of local optimal estimation values and local optimal error covariance matrixes of the integrated navigation system state, then transmitting the two groups of local optimal estimation values into a main filter by a federal filter technology for fusion to obtain an overall optimal estimation value and an overall optimal error covariance matrix, and finally, performing real-time correction on the error according to the overall optimal estimation value so as to obtain an optimal estimation fusion algorithm of the SINS / SAR / CNS integrated navigation system. The autonomous integrated navigation system, disclosed by the invention, is less in calculation amount and high in reliability, is applicable to aircrafts in near space, aircrafts flying back and forth in the aerospace, aircrafts for carrying ballistic missiles, orbit spacecrafts and the like, and has wide application prospect.

The invention discloses an aircraft-mounted multi-navigation-source comprehensive navigation simulation system. In the system, each navigation simulation sub system really simulates functions, performances and interfaces which are same as those of a model according to a self-navigation principle. A scheduling system responds a display control command, a load command and load data information sent by a mission system, and drives the navigation simulation sub systems to perform simulation calculation with combination of environment simulation data. A combined navigation system, with an INS simulation sub system as a main navigation source, performs integrated navigation based on and fault detection federated Kalman filter with GNSS, TACAN and IFDL as auxiliary navigation sources. Information fusion is achieved through a generalized federated Kalman filter in a two-stage structure constructed through a federated filter algorithm, wherein a main filter performs total information fusion and performs information distribution and information reset to each sub filter. Optimal estimation is carried out to a result of each sub filter through a dynamic information distribution coefficient. Meanwhile, the INS simulation sub system is subjected to closed-loop negative feedback correction through an estimated error by the system.

The invention relates to a scene matching / visual odometry-based inertial integrated navigation method. The method comprises the following steps: calculating the homography matrix of an unmanned plane aerial photography real time image sequence according to a visual odometry principle, and carrying out recursive calculation by accumulating a relative displacement between two continuous frames of real time graph to obtain the present position of the unmanned plane; introducing an FREAK characteristic-based scene matching algorithm because of the accumulative error generation caused by the increase of the visual odometry navigation with the time in order to carry out aided correction, and carrying out high precision positioning in an adaption zone to effectively compensate the accumulative error generated by the long-time work of the visual odometry navigation, wherein the scene matching has the advantages of high positioning precision, strong automaticity, anti-electromagnetic interference and the like; and establishing the error model of the inertial navigation system and a visual data measuring model, carrying out Kalman filtering to obtain an optimal estimation result, and correcting the inertial navigation system. The method effectively improves the navigation precision, and is helpful for improving the autonomous flight capability of the unmanned plane.

The invention discloses a dual-wavelength binocular vision seam tracking method. The method comprises the steps including image acquisition, data processing, seam tracking and the like. The invention further discloses a tracking system for implementing the method. The system adopts a near-infrared and structured light dual-wavelength binocular vision sensing system, images of a molten pool and different wavelengths of seam area images with of seams are measured simultaneously and transmitted to a miniature industrial control computer, and the seam positions are measured accurately with a multi-information fusion algorithm and a seam image three-dimensional reconstruction algorithm; the miniature industrial control computer adopts the Kalman filter algorithm to perform optimal estimation on the seam tracking deviation state according to a seam position detection result, a servo driver drives a servo motor to move so as to control a 3-axis motion workbench to generate corresponding motion, a welding torch or a laser head is controlled for deviation correction, and the seams are tracked accurately. The system can eliminate hard light, splashing and electromagnetic interference on a welding site and improve the seam tracking accuracy and reliability.

The invention provides a satellite attitude orbit determining system and a satellite attitude orbit determining method. The method comprises the following steps of: acquiring a pseudo-range, a pseudo-range rate and a carrier phase data from satellite navigation signals from an attitude orbit acquisition module by an attitude orbit determining module in an on-satellite subsystem, generating inter-satellite relative position information, inter-satellite relative velocity information and three-dimensional position information by using the pseudo-range, the pseudo-range rate and the carrier phase data, performing a federal filtering algorithm on the three-dimensional position information and a satellite starlight elevation to generate position information and velocity information of a relative satellite inertia system, performing joint attitude determining by using the satellite navigation signals, the satellite attitude angle information and the satellite starlight elevation to generate a triaxial angle, and performing optimal estimation on the triaxial angle to acquire a triaxial attitude angle and a triaxial attitude angular velocity of a satellite; and performing orbital transfer and attitude adjusting control by an attitude orbit control computer according to the information output by the attitude orbit determining module. By adopting the system and the method, the attitude determining precision and the orbit determining instantaneity can be improved.

The present invention relates to a method and an apparatus for adjusting an average interval of channel estimation dynamically based on Doppler-shift. The method comprises the steps of estimating Doppler-shift by using level cross rate (LCR) according to differently moving speeds of mobile terminals, determining the optimal average interval of channel estimation based on the relationship between the existing Doppler-shift and the optimal average interval of channel estimation, dynamically adjusting the average interval of channel estimation according to the determined the optimal average interval of channel estimation to make the coherent receiver obtain the optimal estimation performance at different moving speeds. The apparatus of the present invention comprises a channel estimation module, a valid / strongest finger selection module, a RAKE demodulation and combining module, a RAKE output decision module, an LCR detection and Doppler-shift estimation module, a Gaussian noise power estimation module, a strongest path signal power estimation module, and an optimal average interval calculation module.

The invention discloses a three-dimensional position reconstructing method based on an ISAR (inverse synthetic aperture radar) image sequence for a scattering point, which comprises the following four links: a target ISAR image gives the distribution information of a target strong scattering point in a radial direction and a transverse direction based on a distance-Doppler high resolution basic principle; data correlation gives a corresponding relationship between two-dimensional projection points in the ISAR image sequence and is realized by utilizing a flight path initialization method through extracting the one-dimensional radial distance information of all the scattering points in an image sequence; an observing matrix is obtained through the following steps: obtaining a target ISAR image sequence through a period of time of sampling, and after the data correlation, and combining the two-dimensional position coordinates of all the corresponding projection points in a sequence to form the observing matrix, so as to form three-dimensional reconstructed known information; and a position matrix is solved through the following step: solving the optimal estimation of a three-dimensional position matrix of a target scattering point from a subspace through carrying out singular value decomposition on the observing matrix and by utilizing a rank theory and using the orthogonality of a projection space as a constraint condition, so as to obtain a target three-dimensional reconstructed image.

The invention relates to a high dynamic satellite navigation signal rapid capture and tracking system and method. The method comprises the following implementation steps that S1, related parameters are set according to the capture requirements, and intermediate frequency signals are read to be mixed with local carrier; S2, windowing is performed on PMF and FFT modules based on the conventional PMF+FFT to implement capturing; S3, an adaptive threshold adjustment method is adopted on signal peak judgment after FFT operation so as to obtain the capture result; S4, the process enters the tracking stage, frequency mixing and pseudo code correlation are performed on the intermediate frequency data and the result after correlation is integrally cleared so as to obtain a coherent integral value; S5, dimension enlarged unscented Kalman filter is adopted for carrier tracking, and the state vector optimal estimation value is obtained according to the dynamic model and the measurement model of the system; and S6, the conventional code ring under carrier aiding is adopted in code tracking. According to the method, the problems of rapid capture and tracking of the high dynamic satellite navigation signals can be solved, and the dynamic performance can be 100g.

The invention discloses a close combination navigation method of a multi-satellite system and a strapdown inertial navigation system. The method comprises the steps that first, strapdown calculating is conducted on IMU output data to acquire the current gesture, speed and position information of an SINS, ephemeris calculating is conducted on a satellite ephemeris message to obtain the positions and the speeds of all satellites, all satellite systems achieve time-space uniformity, and satellite selecting is conducted; secondary, the selected satellite positions and the speed information are used for being calculated with the SINS position and the speed information which are obtained through strapdown calculating to obtain an SINS pseudorange and pseudo-range rate; then, the original measuring data output by a GNSS, namely a code measuring pseudorange and doppler frequency shift, are used for being compared with the SINS pseudorange and the pseudo-range rate, a difference value is used as the observed quantity of a filter, through optimal estimation of the filter, the compensation dosage for correcting the SINS is given, and closed loop correction is conducted to obtain the optimal solution of the SINS gesture, speed and position. The close combination navigation method of the multi-satellite system and the strapdown inertial navigation system solves the problems that the number of satellites is too small in the complex environment with high building blocking and trees covering of combination navigation of inertial navigation and one single satellite, so that the combination navigation precision is influenced.

The invention discloses an AFC data-based public transport passenger flow OD real-time estimation method. The method utilizes card swiping data of a public transport AFC system, performs sorting to obtain bus station passenger flow, on this basis excavates a mapping relation between public transport passenger flow OD and the bus station passenger flow, and builds a public transport passenger flow OD real-time estimation model based on Kalman filtering, thereby realizing real-time estimation of the public transport passenger flow OD. The AFC data-based public transport passenger flow OD real-time estimation method provided by the invention performs deep excavation on passenger flow data of the public transport AFC system, and obtains an optimal estimation value of a public transport passenger flow distribution probability through the Kalman filtering model, thereby realizing real-time estimation of the public transport passenger flow OD, which is of great importance to accurately describing the public transport passenger flow demand distribution law and realizing real-time optimization of bus route operation dispatching.

The invention discloses a fast sparse Bayesian learning based direction-of-arrival estimation method and mainly solves the problems of heavy computation and large location estimation error in the prior art. The method includes the implementation steps: (1) adopting antenna receivers to form a uniform linear array; (2) sampling space signals and computing an array covariance matrix R; (3) vectorizing R to obtain a sparse model vector y; (4) dividing space domain grids, and constructing an over-complete base phi(theta) according to the structure of the sparse model vector y; (5) establishing a sparse equation according to the sparse representation relation of the sparse model vector and the over-complete base; (6) defining a hyper-parameter vector alpha, and adopting a fast sparse Bayesian learning algorithm to solve the sparse equation; (7) drawing a magnitude spectrogram according to an optimal estimation value of alpha to obtain a direction-of-arrival value. By the method, estimation accuracy of target reconnaissance and passive location under the conditions of low signal to noise ratio and small snapshot number is improved, computational complexity is lowed, and the method can be applied to target reconnaissance and passive location.

The invention discloses a device and a method for fusing sensor information. The sensor-information fusion method comprises the steps of adopting a self-adaptive weighted-data fusion algorithm to allocate corresponding weight numbers for induction information from a plurality of sensors, adopting a Kalman filtering algorithm to perform optimal estimation to the induction information of the corresponding weight numbers so as to acquire the locally fused induction information, adopting a D-S theory estimation algorithm to perform interval estimation to the uncertain induction information in the locally fused induction information, adopting a multi-Bayesian estimation algorithm to perform individual association probability distribution for the induction information after interval estimation so as to synthesize a joint posteriori probability distribution function, and utilizing the joint distribution function to output a final fusion value of the induction information. The invention has the advantages of improving the reliability, confidence and efficiency of sensor information fusion.

A Log-Likelihood Ratio (LLR) generating apparatus and method in a communication system using a spatial multiplexing scheme. The reception method includes acquiring at least one estimated transmit vector by demodulating a receive vector using at least one Multiple Input Multiple Output (MIMO) detection process; selecting one of the at least one estimated transmit vector as an optimum estimated transmit vector; calculating LLRs with respect to the optimum estimated transmit vector; calculating a weight to be applied to each of elements constructing the optimum estimated transmit vector; and applying the weight to each of the calculated LLRs. Accordingly, the present invention can generate the LLRs with high reliability similar to the LLRs of a Maximum Likelihood receiver by applying the weight to the LLRs generated from the estimated transmit vector.

An image processing device includes: an estimated plane calculation portion calculating, for inputted image data, estimated planes defined by luminance values in blocks each containing a predetermined number of pixels including a reference pixel; an optimal estimated plane selection portion selecting an optimal estimated plane having a least summation of errors between luminance values of respective pixels in a block and an estimated plane among the estimated planes; a weighting factor calculation portion calculating a weighting factor for the reference pixel base on luminance values of the reference pixel and in the optimal estimated plane; and a weighted smoothing portion computing a sum of products of relative luminance differences between the luminance values of the respective pixels in the block and in the optimal estimated plane and the weighting factor for the reference pixel and adding a luminance value in the optimal estimated plane at a reference pixel position to a result of computation.

The invention discloses an information fusion method for an airborne inertia / Doppler radar integrated navigation system. The method comprises the following six major steps of: I, establishing an error model of an inertia device and a Doppler radar; II, determining a state variable based on the application of inertia / Doppler integrated navigation on an airborne fixed-wing aircraft; III, establishing an inertia / Doppler radar integrated navigation state equation; IV, establishing an inertia / Doppler radar integrated navigation measurement equation; V, discretizing a Kalman filtering continuous system; and VI, performing Kalman filtering data fusion resolving, obtaining the sub-optimal estimation related to navigation parameters through state estimation, and correcting the inertial navigation and Doppler radar in a closed-loop feedback mode at the same time. The method disclosed by the invention makes full use of the Doppler information and inertial navigation information, puts forward a data fusion scheme for airborne inertia / Doppler radar integrated navigation, effectively solves the problem of a navigation error caused by an imperfect error model, and improves the system precision.

The invention discloses an underwater terrain matching aided inertial navigation positioning method and a positioning system. The method mainly comprises the following steps: (1) compensating an error for a closed loop through optimal estimation of Kalman filtering and compensating an error for an open loop through speed backtracing; (2) building an accurate three-dimensional ray tracing model taking an attitude angle of an underwater vehicle into consideration, and providing a method for extracting a sampling point in a beam footprint; (3) providing a method for establishing an underwater digital terrain model according to a water depth constraint and a Delauney triangulation network of a fractal dimension theory; (4) providing an underwater terrain adaption area evaluation method according to a gray decision theory; (5) providing the underwater terrain matching aided navigation method according to the self-adaptive gauss and the filtering. The error compensation methods for the closed loop and the open loop are utilized, the accumulative error of an inertial navigation system is greatly weakened, and the navigation positioning accuracy of the underwater vehicle is greatly improved by constructing the high-accuracy high-resolution underwater digital terrain model and the self-adaptive gauss and filtering algorithms.

The invention relates to a balance control method for a power battery pack of an electric vehicle. The method comprises the steps of building a second-order thevenin model and an one-state hysteresis characteristics model of the single battery; performing charging / discharging test for single power battery; recording the ampere per hour of charging / discharging; calculating SOC theoretical value; performing current excitation on real time according to sampling time intervals; treating the actual end voltage of the single battery and the SOC theoretical value as the system input; performing the adaptive unscented kalman filter based SOC estimation method for the second-order thevenin model; performing the unscented kalman filter based SOC estimation method for the one-state hysteresis characteristics model; performing weighted summation and conversion for the two estimation results according to a model probability matrix to obtain interactive multi-model kalman filter; solving the optimal SOC estimation value; performing balance control for the battery pack through a Buck-boost circuit based on adjacent battery energy transfer and of a balance topological structure according to the SOC optimal estimation value. The method is simple and reliable, and high in practicability and performability.

The present invention is a method of generating a best estimate of an image attenuation map derived from a truncated image attenuation map and PET emissions data for the object being imaged by a morphological imaging modality. The method involves a plurality of steps beginning with the recordation and processing of PET emissions data. Next, the morphological imaging modality records image data which is processed to determine an attenuation map. The attenuation map, for image modalities such as CT and MR scanning systems integrated with PET, is truncated, resulting in a truncated attenuation map image. Pixels for which attenuation data needs to be determined are identified and attenuation coefficients for these pixels are estimated and combined with the truncated attenuation map to generate a full initial attenuation map for the image, which is iteratively processed together with the PET emission data until the improvement change in the emission image reaches a defined threshold improvement level.

The invention discloses a rapid establishing and dynamic updating system and a method for a WLAN position fingerprint database. According to the method, the extended Kalman fusion filtering algorithm is mainly designed, available information of an MEMS sensor and an indoor signal transmission model are fully used, high-precision two-dimensional position solution is realized, the optimal estimation outputted by an extended Kalman filter is corrected by employing known indoor map information, and final position information is obtained; and the WLAN position fingerprint database can be rapidly established and updated with the combination of the obtained position information and RSSI information collected at the current moment. According to the system and the method, the problem of reduced positioning precision due to big change of the WLAN position fingerprint database with the change of time is solved, and the time cost and the manpower cost of the conventional point-to-point acquisition method for establishing and updating the WLAN position fingerprint database are effectively reduced.

The invention discloses a dead reckoning-based low-cost Big Dipper and MEMS tight-coupling positioning system and method. Based on the acceleration, the magnetometer and the angular velocity information of an MEMS, the PDR algorithm is implemented. The acquired PDR position and velocity information is combined with the Ephemeris information of a Big Dipper receiver, and then the pseudo range and the pseudo-range rate value of a PDR terminal can be estimated. After that, the above pseudo range and the above pseudo-range rate value are compared with the pseudo range and the pseudo-range rate value that are outputted by the Big Dipper receiver, and the difference values therebetween are adopted as the observed values of a navigation filter. The optimal estimation on the error amount of a combination system is conducted by the filter, and then error-corrected data are obtained. According to the technical scheme of the invention, in an environment wherein a target is moved in such a manner that the signal of the target is decreased significantly or the target has no satellite signal, the target can still be continuously positioned. The invention also provides a federal kalman filter algorithm. The time accumulative error of a traditional inertial navigation system is inhibited, and the stability of the overall system is improved. At the same time, based on the adaptive adjustment algorithm, the soft handoff of the system can be realized in different environments, so that the seamless transition of the positioning effect is realized.

The invention discloses a pedestrian autonomous navigation calculation algorithm based on an MEMS-IMU. Two IMU systems are simultaneously and fixedly connected with the two feet of the user of a pedestrian navigation system; the two systems respectively perform a strapdown inertial navigation calculation algorithm and a zero-speed correcting algorithm on the basis of Kalman filtering, and then the positioning information of the two feet is fused; when the calculation distance between two feet exceeds the largest step size gamma between the two feet, inequality constraint is performed on navigation results of two IMUs by adopting a state constraint Kalman filtering algorithm, so that vague human physiological property problem is converted into a rigid mathematical problem, the optimal estimation of navigation results is achieved, and pedestrian navigational positioning function with higher accuracy is achieved.

The invention discloses a power system interharmonic estimation method based on an auto-regression (AR) model and Kalman filtering, relates to an interharmonic estimation method and aims to solve the problems the conventional fast Fourier transform method has poor real-time property and the conventional Burg algorithm has low spectral analysis performance and low anti-jamming capability. The method comprises the following steps of: A, acquiring signal data of a power system; B, establishing an autoregressive (AR) model according to the acquired signal data of the power system, and analyzing the signal of the power system by utilizing the AR model; C, establishing a state equation and an observational equation; D, determining the order of the AR model according to a final prediction error rule; E, estimating parameters of the AR model on line by utilizing an adaptive Kalman filtering method; and F, performing power spectrum estimation by using the AR model, and substituting the optimal estimated value into an AR model power spectrum density formula for power spectrum estimation. The method is used for estimating the power spectrum of power system interharmonics.