41 results about "Kalman estimation" patented technology

The invention provides an indoor WLAN / MEMS fusion cross-stair three-dimensional positioning method. The indoor WLAN / MEMS fusion cross-stair three-dimensional positioning method is mainly characterized in that a WLAN / MEMS fusion positioning robust extended Kalman filter is designed, available information of two systems is fully utilized, and resolution of a high-precision two-dimensional position is realized; a height is calculated by using output of a barometer, a geographic position and stair height information, and a position fingerprint database is selected for positioning a wireless local area network (WLAN), and indoor space cross-stair three-dimensional positioning is realized. The system and method are high in positioning precision by introducing robust extended Kalman estimation, and can be used for effectively overcoming an accumulative error existing in positioning of a micro electro mechanical system (MEMS) sensor and received signal strength (position information and actual height information of the previous time) in WLAN positioning to realize indoor space cross-stair three-dimensional positioning.

The invention belongs to the field of image processing, and discloses a space-time correlated channel massive MIMO transmission method. According to the space-time correlated channel massive MIMO transmission method, a time-shifted pilot frequency system structure is adopted, a user selection scheme based on the position of a user is utilized, user sub-groups having large interference in surrounding cells are omitted, so that data interference of the adjacent cells is reduced in a channel estimation stage, meanwhile, interference among the cells is reduced in a downlink data transmission stage, and system capacity is promoted. On the basis of the user selection scheme, a Kalman estimation method is adopted, a space-time correlation between channels is used, residual interference among the cells is eliminated further, and channel estimation accuracy is promoted. By means of combination of a user selection process and Kalman channel estimation, more accurate channel estimation results are obtained under the space-time correlated channel of a multi-cell massive MIMO system, interference among the cells in a pilot frequency estimation stage is restrained, and meanwhile the throughput rate of downlink data of the system is increased.

The invention discloses a wheel hub motor driving vehicle speed estimation method based on multi-model fusion. The method comprises the following steps that a vehicle-mounted sensor signal is collected, filtering processing is carried out on an original collected signal, a vehicle speed Kalman estimation filter is established, self-adaptive adjustment is carried out on noise variances by combiningvehicle driving state information, and an exponential damping memory factor is utilized to correct a state vector estimation error to obtain a vehicle speed estimation value based on self-adaptive exponential weighted damping memory Kalman filtering; a vehicle body acceleration integral vehicle speed estimator is designed; and the two kinds of vehicle speed estimation models, namely the vehicle speed estimation value based on self-adaptive exponential weighted damping memory Kalman filtering and the vehicle body acceleration integral vehicle speed estimator are weighted and fused based on theprinciple of minimum total mean square error in combination with vehicle driving conditions. The wheel hub motor driving vehicle speed estimation method based on the multi-model fusion aims to achieve the effects of being good in real-time performance, high in precision and strong in applicability.

The invention discloses a permanent magnet synchronous motor direct torque control method based on model predictive control, and belongs to the technical field of control. The invention aims to provide the manent magnet synchronous motor direct torque control method based on model predictive control, wherein by means of an electric vehicle wheel hub motor drive control method based on extended Kalman estimation and model predictive control, the defects in the prior art can be well overcome. According to the permanent magnet synchronous motor direct torque control method based on model predictive control, firstly, an extended Kalman principle is used for completing motor drive system estimator designing, secondly, torque and flux linkage values obtained by the estimator are used as feedbackvalues which are transferred to a motor control system controller along with reference set values, and finally, by solving an optimum control problem corresponding to a target function, fast accuratetracking control of motor torque and flux linkage is completed by the system. By means of the permanent magnet synchronous motor direct torque control method based on model predictive control, the problems of complex coordinate change and current magnitude coupling in a motor model are effectively solved, the system complexity is reduced, and dynamic response speed of the system and the intuitionism and the reliability of a system control effect are improved.

The invention discloses a method for adjusting pressure of an anesthesia machine and a breathing machine through flow and pressure common control, which is characterized by comprising the following steps of: calculating a pressure value under the current flow by using a formula P=F*R in a pressure control mode, wherein P is pressure; F is the current flow; and R is air resistance; estimating the pressure at a next moment under the current pressure value by using a Kalman estimation algorithm into which sub-items with control parameters influencing an overall estimation result are added; substituting the estimated pressure value into an error system to calculate a control error quantity, performing proportion integration differentiation (PID) control on the control error quantity, and converting into control pressure required at the next moment; converting the control pressure into standard control flow by using the formula P=F*R; and inputting the converted control flow into a flow PID ratio controller for flow feedback control.

The invention relates to an acoustoelectric bimodal fusion measuring method of two-phase flow process parameters. The method includes: arranging a conductivity sensor and a capacitance sensor on a measuring pipeline, and arranging an ultrasonic sensor on the measuring pipeline; extracting flow rate characteristic quantity from data measured by the conductivity sensor as related flow rate; acquiring Doppler flow rate from data measured by the ultrasonic sensor; respectively performing state estimation on water content acquired by utilizing the capacitance sensor for measuring and water content acquired by the conductivity sensor; fusing the capacitance water content and the conductivity water content; respectively performing state estimation on the Doppler flow rate and the related flow rate; respectively performing Kalman estimation on the flow rates acquired by measuring based on two principles, and calculating flow rate state estimation at all moments; fusing the Doppler flow rate and the related flow rate.

The invention relates to a self-localization method of an intelligent wheelchair in corner areas. The wheelchair comprises a camera, sonar sensors, a first Kalman filter, and a second Kalman filter. Environmental information is collected by the sensors. The information is processed through two times of Kalman estimation, such that corresponding weight values of the sensors are obtained. Then a corner position is monitored, and an accurate distance between the wheelchair and the corner is obtained through calculation, such that self-localization is achieved.

The invention relates to a self-adaption Kalman estimation and fusion method of the moisture content of two-phase flow. The method includes the steps of calculating measurement values of sensors capable of directly reflecting the moisture content to serve as the judgment standard of the sudden change of the moisture content; setting a threshold value to be compared with judgment parameters of the sudden change of the moisture content to judge whether the moisture content is suddenly changed or not so as to switch the working states of Kalman estimation, wherein the working states of the Kalman estimation include the stable state and the sudden change state; conducting Kalman estimation updating on each sensor; correcting the estimation value according to measurement information; conducting normalization calculation on self-adaption weights by means of different algorithms in the stable state and the sudden change state; calculating the fused moisture content. By means of the method, the estimation precision of the moisture content can be improved.

The invention discloses a vehicle state estimation method, which aims to find a more effective implementation scheme of vehicle state estimation, and comprises the following steps of: establishing a seven-degree-of-freedom vehicle model according to kinematic characteristics of a four-wheel independent drive electric vehicle; simplifying the seven-degree-of-freedom vehicle model to obtain a discrete model suitable for an estimator, and further improving to obtain an extended state equation only containing white noise; and estimating the state by adopting a Kalman filtering algorithm. Accordingto the method, the seven-degree-of-freedom kinematic model of the vehicle is adjusted, and the Kalman estimation method is designed based on the improved model, so that model errors and measurement errors of non-Gaussian noise can be effectively processed, the defects of an existing four-wheel independent drive electric vehicle in control are overcome, and the state of the vehicle can be estimated more accurately.

The invention relates to a centralized two-stage Kalman estimation method with related measurement noise. Aimied at the filtering problem of a multi-sensor measurement system with related measurementnoise influencing a measurement value, a measurement equation with measurement noise irrelevance is re-established by introducing a decorrelation technology, and an optimal estimation value of a system state is obtained through a two-stage Kalman filter. Compared with a two-stage Kalman method directly used, the method provided by the invention has the advantages that although the data fusion results are the same, the calculation complexity is greatly reduced.

The invention relates to a monitoring method of a value R and a value C of an anesthesia apparatus and a respirator. The method comprises the following steps of in the respiration stage of each respiration cycle of a patient, collecting the values of air supply volume V, pressure P and flow speed F of the time in a real-time way at certain time intervals, and transmitting the collected data to a CPU (central processing unit) of a controller; according to the collected values of volume V (t), pressure P (t) and flow speed F (t) at certain time t, and the collected values of volume V (t+1), pressure P (t+1) and flow speed F (t+1) at the next time (t+1), calculating approximate values of the value R and the value C through the CPU of the controller, transmitting the calculated approximate values of the value R and the value C into a Kalman estimation controller to estimate, and classifying and storing the estimating results into a storage unit of a controller according to the values; and after the respiration stage of the cycle is finished, solving the normal distribution of a plurality of stored value R and value C samples, and respectively calculating the average values of the value R and the value C with widest normal distribution as the results of the value R and the value C.

The invention discloses a Kalman estimation method of joint torque of an SCARA robot. The Kalman estimation method comprises the following steps: Step 1, acquiring a continuous nonlinearity torque expression of all joints; Step 2, utilizing the first-order Taylor expansion formula of a multivariate function to obtain a recursive linear discrete joint torque expression; Step 3, taking the linear discretized joint torque expression as a process equation, adding a corresponding measurement equation, and combining and arranging the process equation and the measurement equation into a discrete time state space model of the joint torque; and Step 4, determining a formula used in a stepping cycle by combining the discrete time state space model of the joint torque and according to the basic Kalman filtering method, and estimating the joint torque of the SCARA robot in real time by utilizing a read joint motor current signal and a motor encoder signal. According to the Kalman estimation method, the estimation precision of the joint torque is improved without using a sensor to provide reliable information for a robot control algorithm.

The invention belongs to the field of signal processing, and provides a self-adaptive clock synchronization method with high robustness for defects of the existing Kalman estimation clock synchronization method. The method provided by the invention comprises the steps that the topology model of a wireless sensor network is established; network nodes are divided into multiple levels; a subsequent clock synchronization method is carried out only between adjacent levels; progressive synchronization ultimately develops into global clock synchronization; a network node clock model and a transport model between nodes are established; an H infinity estimation filter is used to carry out relative clock parameter estimation on the nodes of adjacent levels; clock synchronization of adjacent levels is realized; and global clock synchronization is realized. According to the invention, H infinity estimation has the advantage of high robustness; and an H infinity clock synchronization method does not need the priori information of clock model noise, a transmission delay autocorrelation matrix and the like.

The invention discloses a blood oxygen content estimation method based on binary sensor Kalman fusion. The method comprises the following steps of: establishing a blood oxygen content dynamic physiological model and a binary measurement model, analyzing effective information in binary sensor measurement according to a blood oxygen content model, and obtaining a practical measurement model; designing a robust local Kalman estimator, giving a recursive computation process of an estimation error covariance upper bound, and through the minimization of the estimation error covariance upper bound, solving an optimal local estimation gain; and designing a distributed fusion Kalman filter used for blood oxygen content estimation, adopting a covariance insertion fusion criterion, and solving an optimal problem to obtain an optimal fusion weight matrix. The invention provides the blood oxygen content distributed fusion Kalman estimation method based on the binary sensor, and realizes the real-time non-invasive estimation of the blood oxygen content.

The invention provides an indoor WLAN / MEMS fusion cross-stair three-dimensional positioning method. The indoor WLAN / MEMS fusion cross-stair three-dimensional positioning method is mainly characterized in that a WLAN / MEMS fusion positioning robust extended Kalman filter is designed, available information of two systems is fully utilized, and resolution of a high-precision two-dimensional position is realized; a height is calculated by using output of a barometer, a geographic position and stair height information, and a position fingerprint database is selected for positioning a wireless local area network (WLAN), and indoor space cross-stair three-dimensional positioning is realized. The system and method are high in positioning precision by introducing robust extended Kalman estimation, and can be used for effectively overcoming an accumulative error existing in positioning of a micro electro mechanical system (MEMS) sensor and received signal strength (position information and actual height information of the previous time) in WLAN positioning to realize indoor space cross-stair three-dimensional positioning.

The invention discloses a vehicle state estimation method under the condition of abnormal measurement data of a vehicle-mounted sensor, which specifically comprises the following steps of: acquiring longitudinal acceleration, transverse acceleration, yaw velocity and front wheel rotation angle signals of a vehicle , combining with a nonlinear vehicle model, estimating axial force information of the vehicle by using strong tracking unscented Kalman filtering, estimating a tire road adhesion coefficient by utilizing interactive multi-model unscented Kalman based on the axial force information of the vehicle; wherein the vehicle axial force information includes longitudinal and lateral forces of a front axle of the vehicle and longitudinal and lateral forces of a rear axle of the vehicle. Through interaction, mixing, prediction and fusion, the invention provides an estimation method which can combine the advantages of a plurality of models to realize accurate estimation of the tire road adhesion coefficient under a complex driving condition, then updates a posteriori state and a covariance matrix P eta thereof, and adopts a prior and posteriori combined estimation method the technical blank that the axial force of the vehicle cannot be accurately estimated under the current mass mismatching condition can be filled.

The invention relates to the technical field of medical health, in particular to a heart rate calculation method based on heart sound. The method comprises the following steps of: acquiring real-timeheart sound signal and processing the waveform of the heart sound signal to form a calculable characteristic waveform; detecting the time interval T between adjacent two peaks and calculating an instantaneous heart rate value; calculating arithmetic mean values of the instantaneous heart rate values in turn according to each successive five beats; utilizing a system estimation method to process the arithmetic average values to obtain corresponding estimation values; carrying out anti-misjudgment processing on the calculated values and the estimated values. According to the calculation method,under the premise of accurate detecting the heart rates, the calculation amount is small, the complexity is low and the real-time property is good, and the calculation method is suitable for calculation on the basis of an embedded chip, so that the development cost of a enterprise product can be reduced; Kalman estimation equation is used in the data processing, so that the final result cannot beinfluenced by the signal-to-noise ratio and the human and environmental interference, and meanwhile, stable and reliable heart rate results can be obtained.

The invention provides an electronic throttle valve opening degree estimation method and an electronic throttle valve opening degree estimation system based on Kalman-like filtering. The electronic throttle valve opening degree estimation method comprises the steps of: acquiring electronic throttle valve opening degree data acquired by means of an electronic throttle valve position sensor; and processing the opening degree data, measured by means of the position sensor, of an electronic throttle valve by adopting a Kalman-like estimation algorithm to obtain a filtered opening degree of the electronic throttle valve, and feeding the filtered opening degree value of the electronic throttle valve back to a vehicle-mounted controller for further controlling the opening degree of the electronicthrottle valve according to the filtered opening degree value. By adopting the Kalman-like estimation algorithm for filtering an output signal of the electronic throttle valve position sensor, interference of non-Gaussian white noise on a measurement result is reduced, and an estimation strategy is high in estimation precision and high in robustness.

The invention discloses a self-adaptive fuzzy Kalman estimation SOC algorithm. The method comprises the following steps: S1, establishing an equivalent circuit model of a battery, establishing a state-space equation and an observation equation by applying an extended Kalman algorithm, and estimating a short-time polarization end voltage variable Vst, a medium-time polarization end voltage variableVmt, a long-time polarization end voltage variable Vlt and a battery state-of-charge SOC variable; S2, under the condition that different SOCs are matched with the temperature T, setting equivalent internal resistance, polarization capacitance and polarization resistance of an equivalent circuit model in the charging and discharging process of the battery through a battery characteristic experiment; S3, realizing Kalman prediction and updating, and estimating the SOC value in each sampling period in real time; S4, calculating a corrected ampere-hour integral factor of the platform period by applying the EKF and the ampere-hour integral in the OCV-SOC non-platform period; and verifying the corrected ampere-hour integral of the platform period by applying the EKF again when the platform period is ended, introducing fuzzy control to perform error correction on a platform period correction factor, and finally applying the correction factor to the ampere-hour integral of a new round of non-platform period correction algorithm. The method has the advantages that the estimation precision and the algorithm debugging time of the algorithm are improved, and the precision of the extended Kalman filter can meet corresponding requirements by defining parameters in the automatic adjustment method.

The present application discloses UWB (Ultra-Wide Band) fine positioning method, device and equipment and a computer-readable storage medium. The method comprises the steps of balancing an optimal observation solution and an estimated solution through Kalman filtering by fully utilizing motion information to obtain a Kalman estimated solution; and compensating error caused by modeling and uncertain noise, of the Kalman filtering, through a BP (Back Propagation) neural network at the same time to obtain an optimal final estimated solution. The method has the advantages that the positioning result is more accurate and the technical problem that large error exists in a conventional UWB precise positioning method is solved.

The invention discloses a four-wheel independent drive electric vehicle side slip angle and tire lateral force estimation method which comprises the following steps: calculating tire longitudinal force according to a wheel kinetic equation; according to a longitudinal dynamic balance equation of the vehicle, estimating the mass of the whole vehicle based on a least square method with a forgetting factor; the robust volume Kalman estimation module is used for establishing a four-wheel drive electric vehicle dynamic model comprising three degrees of freedom of vehicle longitudinal, lateral and yawing and a semi-experience magic tire model for reflecting instantaneous mechanical characteristics of tires; and estimating the side slip angle and the tire lateral force based on the established robust volume Kalman filtering module. The method effectively improves the perturbation of filtering to model parameters and the anti-interference capability of unmodeled noise under the complex working condition, improves the accuracy, robustness and anti-interference performance of a joint estimation algorithm under different working conditions, and solves the problem of joint estimation of the side slip angle and the tire lateral force of the four-wheel-drive electric vehicle under the composite working condition.

The invention relates to an acoustoelectric bimodal fusion measuring method of two-phase flow process parameters. The method includes: arranging a conductivity sensor and a capacitance sensor on a measuring pipeline, and arranging an ultrasonic sensor on the measuring pipeline; extracting flow rate characteristic quantity from data measured by the conductivity sensor as related flow rate; acquiring Doppler flow rate from data measured by the ultrasonic sensor; respectively performing state estimation on water content acquired by utilizing the capacitance sensor for measuring and water content acquired by the conductivity sensor; fusing the capacitance water content and the conductivity water content; respectively performing state estimation on the Doppler flow rate and the related flow rate; respectively performing Kalman estimation on the flow rates acquired by measuring based on two principles, and calculating flow rate state estimation at all moments; fusing the Doppler flow rate and the related flow rate.

A method for calculating heart rate on the basis of heart sounds, the method comprising: obtaining a real-time heart sound signal and processing a waveform of the heart sound signal so as to form a calculable characteristic waveform; detecting a time interval T between two adjacent peaks and calculating an instantaneous heart rate value; sequentially calculating an arithmetic mean value of instantaneous heart rate values according to every five consecutive beats; using a systematic estimation method on the arithmetic mean value to obtain a corresponding estimated value; and performing misjudgment prevention processing on the calculated value and the estimated value. On the premise of detecting heart rate accuracy, the described method has a small amount of calculation, low complexity and good real-time performance, and is suitable for calculation on the basis of embedded chips, which may reduce the costs of product development for enterprises; a Kalman estimation equation is used in data processing so that the final result may be free from the influence of signal-to-noise ratio and human and environmental interference, and stable and reliable heart rate results may be obtained.

The invention discloses a measurement method for the affection of noise variances on SOC (State of Charge) filtering effect, which includes the following steps: under the condition that P<0> and SOC<0> are not changed, a noise variance is adjusted and measured in Matlab / Simulink; under different measured noise variance R values, the difference of UKF (Unscented Kalman Filter) filtering effects of a lithium battery pack is great, and as the R value is increased, the filtering error is remarkably reduced; according to the measurement, during the Kalman estimation of lithium battery SOC, appropriate SOC<0> and P<0> are chosen. The measurement method for the affection of noise variances on SOC filtering effect can overcome the defects in the prior art, such as short service life, poor safety and low reliability, thus achieving the advantages of long service life, high safety and high reliability.

The invention provides a ship shaft power measurement method based on Monte Carlo Kalman filtering. The method comprises the specific process: abstracting a physical model of a shaft system according to the working principle of the shaft system, building a state equation and a measurement equation, and employing Monte Carlo Kalman filtering. The method comprises the following steps: firstly, simulating a random point set of state torque by adopting Monte Carlo according to a state estimation value and covariance at a previous moment, carrying out nonlinear transformation on the random point set according to a state equation, and calculating a state prediction value at the moment; secondly, simulating a random point set for the measured value of the rotating speed by adopting Monte Carlo, and fusing and calculating a predicted value of the shaft power according to a measurement equation; and finally, performing Kalman estimation on a torque state estimation value at the moment, and calculating the ship shaft power. Aiming at the defects of contact type and non-contact ship shaft power measurement, the invention aims to adopt a Monte Carlo Kalman filtering algorithm to carry out fusion filtering and noise covariance estimation on acquired data on the premise of not increasing hardware, so that the measurement precision is improved, and the purpose of monitoring the shaft power on line is achieved.

The present invention discloses a method for estimating vehicle state, aiming at finding a more effective solution for estimating vehicle state, which includes the following steps: building a vehicle model with seven degrees of freedom according to the kinematics characteristics of four-wheel independently driven electric vehicles; Simplify the seven-degree-of-freedom vehicle model to obtain a discrete model suitable for the estimator, and further improve the extended state equation containing only white noise; use the Kalman filter algorithm to estimate the state. The present invention adjusts the vehicle's seven-degree-of-freedom kinematics model, and designs a Kalman estimation method based on the improved model, which can effectively deal with the model error and measurement error of non-Gaussian noise, and overcomes the existing four-wheel independent drive electric vehicle. Insufficient, the vehicle state can be estimated more accurately.