55 results about "Linear kalman filter" patented technology

Automotive electronic control unit programmed to realtime estimate either or both of vehicle mass and road slope, wherein; a. road slope, is estimated; a1. when vehicle is considered stopped based on an accelerometer signal indicative of vehicle acceleration, wherein the vehicle is considered stopped in the presence of substantially zero values of a speed signal indicative of vehicle speed, and a2. when vehicle is in rectilinear and curvilinear motion by implementing a road slope observer based on a linear Kalman filter, which is designed to: a21. operate based on signals indicative of vehicle speed and acceleration, and a22. compensate for accelerometric disturbances due to; a221. vehicle static pitch resulting from vehicle load distribution, and a222. vehicle dynamic pitch due to acceleration to which vehicle is subjected during motion, and a223. accelerometric disturbance components due to vehicle lateral dynamics; b. vehicle mass is estimated: b1. when vehicle is in motion, and b2. based on a recursive least square algorithm with forgetting factor, and b3. based on an accelerometric signal indicative of vehicle acceleration, on a vehicle speed signal, and other signals representing a vehicle propulsive/resistive torque, and b4. at different low gears, to provide a mass estimation and an associated variance for each gear, and b5. based on mass estimations and corresponding variances for each gear, and b6. compensating for accelerometer disturbances due to: b61, vehicle dynamic pitch; and b62. accelerometrie disturbance components due to vehicle lateral dynamics; and b7. minimizing uncertainties on propulsive/resistive torque due to gear efficiency and roiling resistance.

The invention provides a hybrid derivative-free expansion Kalman filter-based USBL/SINS tight combined navigation positioning method particularly applicable to positioning of underwater equipment. Thehybrid derivative-free expansion Kalman filter-based USBL/SINS tight combined navigation positioning method involves a USBL acoustic positioning system and an SINS, and combined navigation is performed by a hybrid derivative-free expansion Kalman filter (HDEKF). The method comprises the steps of firstly, obtaining a slope distance measurement value between a transponder and a hydrophone by the USBL system through calculating one-way propagation time of an ultrasonic wave signal between the transponder and the hydrophone, and obtaining an observation equation by a coordinate conversion formula; secondly, building an error state equation according to an error transfer formula of the SINS; and finally, performing hybrid derivative-free expansion Kalman filtering, performing time updating bya standard linear Kalman filter, and performing measurement updating by a derivative-free expansion filter. By the method, the navigation accuracy and the stability of the USBL/INS combined navigationsystem can be effectively improved, and the real-time calculation quantity is reduced.

The invention discloses a method for estimating lateral force of front wheels of a distributed-type electrically-driven vehicle. The method is suitable for realtime monitoring of tire lateral force when the vehicle travels and used for safety evaluation. The method includes collecting state signals of the vehicle, and utilizing a vehicle dynamics equation to estimate longitudinal force and vertical force of tires in real time; transmitting estimated longitudinal force of each wheel, longitudinal acceleration signals, lateral acceleration signals, yawing velocity signals and steering wheel angle signals to a Kalman lateral force observer in a vehicle controller, and acquiring Kalman lateral force estimation values and rear axle lateral force estimation values of the front wheels; further processing the estimated lateral force by utilizing the vertical force of each wheel and front wheel steering angle difference to acquire final lateral force estimation values. The method has the advantages that only a linear Kalman filter is adopted, so that calculating instantaneity is guaranteed; acquiring of information of the tires and roads is not needed, so that the method is enabled to have robustness on different roads and tires, and estimated lateral force results can be used for tire characteristic analysis and road condition recognition.

The invention relates to a vehicle positioning method based on vision, GPS and high-precision map fusion. The method comprises steps that firstly, inertial navigation data and vehicle GPS coordinatesin the map are converted to spatial plane coordinates, a road arrow closest to the current vehicle position is obtained through matching, the vision positioning scope is determined, a Kalman filter isthen utilized to fuse GPS positioning and vision positioning results, vision positioning data is utilized to correct the GPS positioning result, the more accurate position information of a vehicle ina global coordinate system can be obtained, and positioning is completed. The method is advantaged in that traditional nonlinear constraints based on the distance between points are overcome, the distance from the point to arrow sides is utilized as a constraint, a linear Kalman filter is constructed, system accuracy is improved, problems of map data missing and road sign wear of some complex road segments of the map are solved, problems of low GPS positioning accuracy, accumulation errors of vision positioning and incapability in global positioning are overcome through fusion positioning, and thereby centimeter-level positioning accuracy is achieved.

The invention discloses a novel vehicle combination positioning algorithm, and the algorithm employs the technology of federated Kalman filtering data fusion to achieve the information fusion of three positioning subsystems: RSU (road-side unit) positioning, satellite positioning (GNSS) and DR (dead reckoning). In other words, a linear Kalman filter serves as a local filter of an RSU positioning subsystem, and is named as LF1, wherein the corresponding information distribution coefficient is beta1; the GNSS positioning subsystem also employs the linear Kalman filter as the local filter which is named as LF2, wherein the corresponding information distribution coefficient is beta2; and the DR positioning subsystem employs an extended Kalman filter as the local filter which is named as LF3, wherein the corresponding information distribution coefficient is beta3, and a main filter is used for information fusion. Residual error x2 detection and residual hardware detection are employed for building a two-stage fault detection method, and the method can detect a hard fault and a soft fault of the subsystems at the same time. Meanwhile, the algorithm dynamically adjust the information distribution coefficients of the combined system according to positioning precision factors of the positioning subsystems, enables the system to be able to quickly adapt to the environment change, and improves the vehicle positioning precision and reliability.

The invention discloses an aero-engine gas circuit parameter online prediction method. The method comprises the steps: A, obtaining an engine measurable parameter measurement value at the current moment k through a sensor; B, performing incremental estimation on the health parameters and the state quantity of the engine by using a linear Kalman filter; C, updating health parameters and external input of the engine model, and updating a state space model and a gas path parameter prediction state space model required by the Kalman filter based on the component-level partial derivative model; andD, transmitting the control input sequence to the gas circuit parameter prediction state space model, and conducting online prediction on engine gas circuit parameter changes. According to the method, the real engine health state can be effectively evaluated in the dynamic process, and the gas path parameter change can be quickly and accurately predicted on line according to the given input sequence.

A method and system is proposed for use by a moving station (such as a jetfighter) for radar tracking of a moving target (such as an air-to-air missile). The proposed method and system involves the use of a hybrid FSK/LFM (Frequency Shift Keying & Linear Frequency Modulation) scheme for acquiring a collection of raw radar data, a first Gaussian-noise filter array of one-stage linear Kalman filters for S/N-enhancement of the raw radar data, a trilateration module, and a second Gaussian-noise filter array of one-stage linear Kalman filters for S/N-enhancement of the trilateration-resulted radar data. These features allow the radar tracking of moving objects to be more fast and accurate.

The invention discloses an improvement code tracking method of a satellite navigation signal receiver and a loop. A method integrating the least square algorithm and the Kalman filtering algorithm is utilized to improve and optimize a code ring of a tracking ring, coherent integration results of all the related branches are processed through the least square algorithm to form a set of new observation points having properties of no bias, smallest mean square error sum and determined spatial distribution, the set of new observation points is further utilized for subsequent linear Kalman filtering to solve code phases of direct signals and multi-path signals, so the Kalman filtering algorithm is made to realize higher precision and high sensitivity, moreover, adding a divergence control module to a code ring Kalman filtering module is further proposed to discriminate and inhibit filter divergence. Through the method, not only can multi-path interference be effectively inhibited, weak satellite navigation signals with lower signal to noise ratios are tracked, but also higher code tracking precision is realized.

The invention discloses a substation inspection robot navigation method and system based on neural network assistance. According to the method, a GPS/INS integrated navigation system is used for providing navigation and positioning services for an inspection robot, the positioning precision is higher than that of a single set of navigation equipment, error estimation and correction are conducted on the navigation and positioning system of the inspection robot through unscented Kalman filtering, and the positioning precision is improved compared with traditional linear Kalman filtering. A radial basis function neural network is introduced and used for improving the problem that the positioning accuracy of the GPS/INS integrated navigation system becomes poor under the condition that GPS signals are lost. When the GPS is not available, the system predicts the navigation error through the neural network, and the prediction result is used for error correction of the system, so that the positioning precision and reliability of the system are improved. The requirement for hardware is low, a series of algorithms can be achieved through software programming, and the cost is low.

The invention discloses a single-sensor double closed-loop urea-selective catalytic reduction (SCR) feedback control method based on an observer. According to a calibration test of a relation map between cross-sensitivity factors of an NOx sensor and the exhaust gas temperature and on the basis of a urea-SCR system mathematical model, a non-linear Kalman filtering observation theory is adopted and the diesel engine urea-SCR system C<NOx> and C<NH3> parameter observer based on data/mechanism is designed. Two estimators of C<NOx> and C<NH3> and the difference value of the respective target values are used as inputs of a controller, and a urea injection double closed-loop feedback controller is designed. The observer/controller and the NOx sensor arranged at the exhaust end of a urea-SCR system are used as main parts to establish the urea injection feedback control system. By means of the single-sensor double closed-loop urea-SCR feedback control method based on the observer, the hardware cost is reduced; the compromise control effect can be achieved through a double closed-loop system under the condition of single control input, limits of authority, double-output constraint and mutual contradiction; the observer is simple and reasonable in design; too many calibration processes during development of an engine electronic control unit can be reduced through the data/mechanism observer, and the strategic flexibility and universality are controlled.

The invention discloses an airborne distributed POS transfer alignment method based on relative strapdown calculation, and aims to improve airborne distributed POS spatial motion information measurement precision in a dynamic environment. The method comprises the steps: a pseudo single subsample rotation vector is constructed by utilizing angular velocities output by gyroscopes at the current andprevious N sampling moments of a main subsystem and a rotation vector error compensation coefficient derived based on conical motion, and a relative attitude quaternion is solved in real time based onthe rotation vector to realize relative attitude updating; meanwhile, a relative speed and position updating algorithm is established according to the space position vector relation of the main subsystem, and relative speed and position updating is achieved through the output specific force and the relative posture of an accelerometer of the main subsystem; and finally, a transfer alignment statemodel and a measurement model are established based on the relative strapdown calculation error differential equation and the three-dimensional relative position and the three-dimensional relative attitude measured by a fiber bragg grating, and transfer alignment is carried out by adopting a linear Kalman filtering estimation method to obtain high-precision motion information.

The invention discloses a human body posture capturing system, and belongs to the field of posture detection. The human body posture capturing system is used for solving a problem that capturing of human body postures using a conventional optical capturing system is difficult to realize because the conventional optical capturing system is easily influenced by the environment. The human body posture capturing system comprises 12 data acquisition nodes, 2 data receiving nodes, and a host computer used for receiving posture data; the 12 data acquisition nodes are placed at different positions ofa human body for measuring of the movement postures of different human body parts, and are used for sending posture data to the data receiving nodes; the data receiving nodes are connected with the host computer through serial ports, and sending received data from the data acquisition nodes to the host computer for process. The human body posture capturing system can be used for measuring triaxialaccelerated speed, triaxial angular velocity, and triaxial magnetometers; a speedometer and a magnetometer are adopted for posture direction estimation, accurate three dimensional all posture solvingis realized via combination of a gyroscope and linear kalman filtering; influences of the external factors are avoided; real-time capturing of human posture is realized; and accurate recurrence of human body motions is realized.

The invention relates to the technical field of unmanned surface vehicle environment sensing, in particular to an unmanned surface vehicle environment sensing method based on a single-line laser radar. For the near-distance obstacle avoidance requirement of a small-sized high-speed unmanned surface vehicle during autonomous sailing, invalid echoes of the water surface are filtered according to thepoint cloud reflection intensity in the detection step; DBSCAN clustering parameters are adjusted according to an obstacle distance; inter-frame data matching is performed by adopting a nearest neighbor data correlation method in the tracking step; motion state estimation is performed by using interactive multiple models under a linear Kalman filter framework; and a noise covariance matrix methodis adaptively adjusted according to the maneuverability of targets so as to enhance the adaptability of a filter. According to the method, the detection and tracking within a 40-meter range can be well completed in a low-sea condition scene; the laser radar almost does not cause leak detection in an environment free of wind waves and where the unmanned surface vehicle stably runs; the clusteringerror rate is relatively low in the point cloud clustering and target correlation problems; and the inter-frame targets can be well matched.

The invention discloses an attitude measurement method of fusion of noninertial system vision and dual-gyroscope multirate CKF. The method includes steps of building a measuring system and positioninga coordinate system; performing vision gesture measurement, inertia angle speed differential calculation and coordination system normalization; building a non-linear Kalman filtering system model; carrying out the multirate CKF fusion of residual compensation. The method combines the single-eye vision and the dual-MEMS gyroscope, and combines fixed problems of inconsistent sampling frequencies ofmultiple sensor combination systems, the fusion of multiple sensor data is carried out by using the multirate CKF based on the residue compensation; the high-frequency output of updated measurement system is guaranteed while the high-precision and stability of the measurement system are ensured.

This invention relates to a target detection device and its detection method, comprising: a transmitting unit for transmitting a detecting pulse to detect target which then reflects the detecting pulse to generate a reflected pulse; a plurality of measuring units, located at different positions respectively which receive said reflected pulse and generates measured values of distance and measured values of velocity according to the reflected pulse received; a plurality of two-stage linear Kalman filters, corresponding to said plural measuring units respectively, each of said plural two-stage linear Kalman filters proceeds an operation according to the measured values produced by corresponding measuring unit so as to generate respectively the estimation values of distance, velocity and acceleration; an arithmetic unit connecting to said plural two-stage linear Kalman filters, which proceeds a triangulation operation according to said estimation values so as to generate distance component values, velocity component values and acceleration component values with respect to the target.

The invention relates to a large azimuth misalignment angle aligning method based on additive quaternion. The method comprises the following steps: (1) according to a state equation and an observation equation of coarse alignment, by utilizing GPS observation information and linear Kalman filtering, carrying out coarse alignment; (2) in a sliding window, by utilizing the provided threshold conditions, directly calculating the convergence level of a current course angle through additive quaternion, judging whether the course angle completely satisfies the threshold conditions or not, if the result is positive, storing the system covariance information, carrying out fine alignment; and if the result is negative, carrying out coarse alignment (step (1)); and (3) keeping on the system state variable and position of coarse alignment, gesture error equation and observation equation, using a new speed error equation, taking the stored covariance matrix as the primary condition, and carrying out fine alignment, wherein the fine alignment can enter the convergence state rapidly until the alignment reaches an expected level. The provided method can realize stable model switch and satisfies the total state application conditions, and the calculation is simple.

The invention discloses a target positioning method and system, an unmanned aerial vehicle and a storage medium. The method comprises the steps of acquiring direction data of multiple targets measuredby the same unmanned aerial vehicle at a current sampling moment; and processing the direction data of each target at the current sampling moment by using a three-dimensional Pseudo-linear Kalman filter algorithm, and thus acquiring the state information of each target at the current sampling moment. According to the mode above, the multiple targets can be positioned by the same unmanned aerial vehicle at the same time, and the multiple targets can be more accurately positioned.

The invention discloses an energy storage converter system, a converter, and a control method and controller of a bidirectional converter. The control method of the bidirectional converter comprises the steps of subtracting a predetermined feed network current reference value from inductive current of an AC side of an energy storage converter to obtain a deviation signal, modulating the deviation signal through a quasi proportional resonant controller to obtain a voltage benchmark signal; carrying out linear kalman filter processing on feed network current of the AC side of the energy storage converter to obtain state estimation signals of harmonic current in the feed network current; carrying out corresponding gain processing on the state estimation signals of the harmonic current in the feed network current and then subtracting the state estimation signals from the voltage benchmark signal as decrements to obtain a voltage reference signal; and modulating the voltage reference signal through a PWM generator to obtain an SPWM control signal of controlling on and off of a power switch tube of the bidirectional converter. A high-frequency sub-harmonic within 11 times in the current of the AC side of the energy storage converter can be effectively filtered and the quality of output power is improved.

The invention discloses a model-based aero-engine performance recovery control method, and belongs to the technical field of aero-engine control. According to the method, an airborne model is established based on a component method, and real-time optimization is carried out from two aspects of aerodynamic thermal parameter calculation and equilibrium equation iterative solution, so that the calculation amount is greatly reduced; on the basis, in order to further utilize the advantages of simple structure and low calculation complexity of a linear Kalman filter, an airborne model online linearization method is designed so as to update the state matrix in real time and improve the estimation precision of the Kalman filter. The invention further discloses a model-based aero-engine performance recovery control device. Compared with the prior art, the technical scheme of the invention can realize online real-time performance recovery control on the aero-engine.

A motor drive control device includes a drive circuit configured to drive a motor with a drive control signal for driving the motor, and a control circuit configured to perform a vector control arithmetic operation based on a detection result of drive currents of coils of the motor, to generate the drive control signal and supply the drive control signal to the drive circuit. When generating the drive control signal, the control circuit estimates a rotation angle of a rotor of the motor and a rotation speed of the rotor with a q-axis current value of a two-phase rotating coordinate system calculated with a detection result of the drive current, and a q-axis voltage command value of the two-phase rotating coordinate system, by using a linear Kalman filter including a prediction step and an update step, using a stationary Kalman filter with the prediction step expressed linearly and time-invariantly.

The invention discloses an underwater self-adaptive maneuvering target rapid tracking method, which comprises the following steps: analyzing the motion situation of a target, tracking the target in real time by using a nonlinear Kalman filter, solving the motion state parameters of the target, and enabling an observation platform to adaptively track the target according to the estimated motion parameters.