209 results about "Forgetting factor" patented technology

The invention provides a combined estimation method for lithium ion battery state of charge, state of health and state of function. The combined estimation method comprises the steps that the state of he---alth of a battery is estimated online: open circuit voltage and internal resistance are identified online by adopting a recursive least square method with a forgetting factor, the state of charge is indirectly acquired according to a pre-established OCV-SOC corresponding relation, and then the size of battery capacity is estimated according to cumulative charge and discharge electric charge between two SOC points; the state of charge of the battery is estimated online: the state of charge of the battery is estimated by adopting the Kalman filter algorithm based on a two-order RC equivalent circuit model, and the battery capacity parameter in the Kalman filter algorithm is updated according to the estimation result of battery capacity; and the state of function of the battery is estimated online: the maximum chargeable and dischargeable current is calculated based on the voltage limit and the current limit of the battery according to internal resistance obtained by online identification, and then the maximum chargeable and dischargeable function can be obtained through further calculation.

An apparatus and method facilitate combined query based searching with serendipitous browsing in a multimedia collection. A user selects objects to label from a local map, which may include representations of objects retrieved from the collection as being responsive to a text or image base query. The text and image portions of the object can be independently labeled. Unlabeled objects are scored and ranked based on the applied labels of labeled objects, which may take into account cross-media pseudo-relevance and user selectable (or default) parameters, such as a forgetting factor, which tends to place greater weight on more recently labeled objects, and a modality parameter, which laces greater weight on the modality (text, image, or hybrid) currently selected by the user. The local map is modified, based on the ranking, optionally after reranking of objects to improve the diversity of the displayed objects.

The invention provides a personalized film recommendation system and method based on attribute description, comprising a candidate film data base, a viewing record data base, a user interest preference model calculation module and a film recommendation model. The personalized film recommendation system uses the attribute description to represent the film feature and is hierarchized according to the types and analyses the user interest hot spot of concentration ratios of different hierarchy and tracks the user interest change using the time forget factor and film model linear superposition method, responds the feedback of the user on the film at real time, adjusts the user interest preference model and increases the accuracy and adaptability of recommended film.

A recursive algorithm is provided for adaptive multi-parameter regression enhanced with forgetting factors unique to each regressed parameter. Applications of this algorithm can include lead acid batteries, nickel-metal hydride batteries, and lithium-ion batteries, among others. A control algorithm is presented, having an arbitrary number of model parameters, each having its own time-weighting factor. A method to determine optimal values for the time-weighting factors is included, to give greater effect to recently obtained data for the determination of a system's state. A methodology of weighted recursive least squares is employed, wherein the time weighting corresponds to the exponential-forgetting formalism. The derived mathematical result does not involve matrix inversion, and the method is iterative, i.e. each parameter is regressed individually at every time step.

A method for data transmission, contemplates calculating a cell-load ratio by:rcl,n=(1-α)·rcl,n-1+dnSn,where rcl,n-1 is a cell-load ratio calculated in a previous stage, rcl,n is an instant cell-load ratio, α is a forgetting factor, dn is an instantaneous cell load, and Sn is a frame capacity; and broadcasting the calculated cell-load ratio by carrying the cell-load ratio in a designated message.Methods of a cell selection and a handover for a subscriber which are initiated by a subscriber are in dependence upon both of the cell-load ratio and the signal strength provided by the base stations.

The invention discloses a method and system for estimating SOC (State-of-Charge) of a power battery based on dynamic parameters. The method comprises the following steps: carrying out a discharge-standing experiment on the battery, obtaining OCV (Open Circuit Voltage)-SOC characteristic curves of the battery at different temperatures, and fitting out an OCV-SOC relational expression; carrying out a constant current pulse discharge-standing experiment on the battery, recording voltage response during the experiment, and identifying the initial value of the parameter of a battery second-order RC equivalent circuit model by an offline method; carrying out dynamic parameter identification on the second-order RC equivalent circuit model by using a forgetting factor-containing recursive least squares method RRFLS; carrying out online estimation on the SOC of the battery by using an EKF (Extended Kalman Filter) algorithm. The estimation method overcomes the defects of inaccuracy and cumulative error of the initial value of SOC in an ampere-hour integral method, and adapts to the dynamic change of battery characteristics, the battery model is high in precision and convergence speed, and is stable and reliable, and the precision of SOC online estimation is improved. The method and system can be widely used in fields of electric vehicles and energy storage battery management systems.

The invention discloses a power-cell SOC online closed-loop estimation method based on an N-2RC model. In the invention, an electrochemical model and an equivalent circuit model are combined and a novel power cell model is provided. The N-2RC model uses a Nernst electrochemical model to replace an electromotive force portion of a second-order RC equivalent circuit model so that one to one correspondence of a cell electromotive force and SOC can be accurately reflected. Based on the model, a recursive least-square method based on a forgetting factor is used to identify a model parameter, and then an expansion Kalman filtering algorithm is used to realize on-line closed loop estimation of the cell SOC. The electrochemical model can well describe a cell characteristic on an electrochemical aspect, but the structure is complex and the model is not suitable for individually individual usage. And the equivalent circuit model belongs to an external characteristic model and can well express a volt-ampere characteristic relationship of the cell, but can not reflect an internal characteristic of the cell. By using the method in the invention, the above problems are overcome.

The invention provides a vehicle mass identification method and a device. The vehicle mass identification method comprises the following steps of: acquiring wheel rotating speed and driving torque of a vehicle during running; acquiring a longitudinal acceleration value of the running vehicle; constructing a longitudinal kinetic equation of the vehicle; and on the basis of the longitudinal kinetic equation, setting a roughly estimated vehicle mass by taking the wheel rotating speed, the driving torque and the longitudinal acceleration value of the running vehicle as measurement input parameters and taking a sum of rolling resistance and air resistance as a variable, and identifying the mass of the vehicle through a recursive least square method with a plurality of forgetting factors.

A recursive algorithm is provided for adaptive multi-parameter regression enhanced with forgetting factors unique to each regressed parameter. Applications of this algorithm can include lead acid batteries, nickel-metal hydride batteries, and lithium-ion batteries, among others. A control algorithm is presented, having an arbitrary number of model parameters, each having its own time-weighting factor. A method to determine optimal values for the time-weighting factors is included, to give greater effect to recently obtained data for the determination of a system's state. A methodology of weighted recursive least squares is employed, wherein the time weighting corresponds to the exponential-forgetting formalism. The derived mathematical result does not involve matrix inversion, and the method is iterative, i.e. each parameter is regressed individually at every time step.

Provided is a power battery SOC estimation method based on backward difference discrete model and a system thereof; the method comprises the following steps: step one, establishing a backward difference discrete model of a power battery, identifying parameters of the backward difference discrete model by a least square method including forgetting factors; step two, on the basis of the backward difference discrete model of the power battery obtained in step one, using self-adaptive extended Kalman filter in combination with a non-linear relationship between an open-circuit voltage and the SOC to complete an effective estimation of the power battery SOC. In the system, voltage and current sensors connected with the power battery are connected with an embedded microcontroller via an analog-digital conversion module. The microcontroller comprises a low-pass filter pre-processing module, a backward difference discrete battery model parameter online identification module, and an AEKF algorithm SOC estimation module. The obtained SOC result is sent to a CAN network of a display device. The power battery SOC estimation system based on backward difference discrete model is simple in structure; the parameter identification speed and precision are increased; the affection to the identification caused by history data is reduced; the calculation is convenient; and the SOC estimation precision is high.

The invention discloses a lithium battery SOC online estimation method comprising the following steps that 1) the open-circuit voltage of a battery is measured, and the state-of-charge initial value of the battery is obtained according to an OCV-SOC curve; 2) the second-order RC equivalent model of the battery is established and the parameter initial value of the battery equivalent model is estimated; 3) the estimation program is started, and the matching coefficient initial value of a state equation is set according to the battery state-of-charge initial value and the parameter initial value of the battery equivalent model; 4) the current battery state-of-charge value is obtained by using an adaptive unscented Kalman filtering algorithm, and the current open-circuit voltage is obtained according to the OCV-SOC curve; 5) the least square method with the forgetting factor is started to identify the parameters of the current battery equivalent model, the matching coefficient of the state equation is updated by the identified parameters and the battery state-of-charge value of the next moment is solved; and 6) the steps 4) and 5) are repeated so that the battery state-of-charge value of each moment is obtained. Compared with the conventional unscented Kalman filtering algorithm, the method has higher accuracy and higher error convergence.

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 discloses a method for estimating the longitudinal adhesion coefficient of a road. The method comprises the following steps: preliminarily estimating a longitudinal road adhesion coefficient in real time on the road surface of a flat expressway by using a recursive least squares (RLS) method with a forgetting factor specific to a front-wheel steering four-wheel automobile on the basis of an overall longitudinal dynamics model and a simplified magic formula tire model; further filtering signal noise by using an extended Kalman filter (EKF) algorithm by taking the estimated longitudinal road adhesion coefficient and a tire model parameter as expansion states, and realizing self-adaptive adjustment of a tire model coefficient; finally, acquiring accurate and robust longitudinal road adhesion coefficient estimation in real time. The method can adapt to working conditions of high slip rate and low slip rate of flat expressways at the same time.

The invention discloses a power battery SOC (state of charge) estimation system and method based on ARMAX (autoregressive-moving-average exogenous) model. The method comprises the steps of representing polarization reaction of a battery base on RC (resistor-capacitor) structure, discretizing a Thevenin battery model, and introducing the ARMAX model; identifying parameters of the Thevenin battery model; calculating the parameters of the Thevenin battery model based on a least squares estimation method; establishing a following linear relationship between open-circuit voltage and SOC of the battery, and acquiring an estimated value of the SOC. The ARMAX model is introduced based on the RC equivalent circuit model; a recursive least squares method with forgetting factors is used to identify parameters of the model; the estimated value of SOC is found according to an OCU (Uoc)-SOC relation table, and accordingly, the problems of difficulty in establishing a battery mathematic model and precisely predicting SOC are solved; influences of model errors upon parameters are fully considered, and SOC estimation accuracy is improved.

The invention discloses a heavy vehicle weight real-time identification method based on CAN information and a function principle. According to the method, when a vehicle runs, vehicle wheel speed, engine rotating speed, engine practical torque percentage, a gear position, accelerating pedal opening degree, a clutch pedal signal and a braking pedal signal are acquired through a vehicle CAN bus; engine torque, accelerated speed and a rotating weight conversion coefficient are obtained by computing; a vehicle longitudinal kinetic equation based on the function principle is established, on the basis of the kinetic equation, the work done by equivalent outer force generated by driving force and air resistance is regarded as system input quantity, the work done by the equivalent accelerated speed generated by the sum of rolling resistance, slope resistance and acceleration resistance is regarded as an observable data vector, and heavy vehicle weight is estimated in real time through a recursive least-squares method with which a forgetting factor is selected according to the feature of the vehicle weight.

A system and method for self-tuning a PID controller utilized with an exciter and generator, which includes a power source, an exciter electrically connected to the power source, a generator that is electrically energized by the exciter, and a processor that provides a PID controller that calculates system gain an estimated exciter time constant and an estimated generator time constant with a recursive least square with forgetting factor algorithm, wherein the estimated exciter time constant and the estimated generator time constant are utilized to calculate PID gains by the processor, wherein the processor includes a random input generator that is summed with the output of the PID controller as input to determine the PID gains using the estimated exciter time constant and the estimated generator time constant and the processor compares a digital value of rms generator voltage against a reference voltage as input into the PID controller.

The invention relates to noise identification based on CNN and a voice enhancement model--SFTRLS-CNN combining the CNN and smooth and fast recursive least squares (SFTRLS). Firstly, 648-dimensional characteristics of MFCC with a noise in a noise frequency band and the like are extracted and enter into a first kind of trained convolutional neural network so as to identify the environment type of the noise. Then, an extracted audio frequency characteristic, a signal to noise ratio and a noise type value are used to form 658-dimensional characteristics. A second kind of convolutional neural network is adopted to be adaptively matched to an SFTRLS algorithm so as to carry out the optimum forgetting factor of voice enhancement. Finally, through the smooth and fast recursive least squares, the noise reduction processing of various environments is realized. By using the algorithm, the enhancement model is suitable for the different noise environments and an adaptive capability is increased. Compared to a traditional SFTRLS, the algorithm has a better voice quality evaluation index value.

The invention discloses a lithium battery model parameter recognition method through a least square method with a forgetting factor, and the recognition method comprises the steps: 1, building a second-order RC equivalent circuit model; 2, carrying out the bilinear transformation, and obtaining a difference equation of the system input and output of the equivalent circuit model; 3, building a recursive least square method with the forgetting factor (shown in the description); 4, building a coefficient equation; 5, collecting the parameters of a battery; 6, recognizing the parameters of the equivalent circuit model. According to the invention, the OCV-SOC function relation and the least square method with the forgetting factor are employed for the dynamic parameter recognition of the second-order RC equivalent circuit model, and the forgetting factor is introduced to a normally used recursive least square method. The method solves a problem of old data overlapping in an operation recursion process, and avoids the phenomenon of data saturation.

A handover method provides intrasatellite transmitter handover detection and decision using an unbiased weighted Signal Noise Ratio (SNR) estimator based on forgetting factors for SNR monitoring, and a threshold delay is introduced to ensure the reliability of handover decision to prevent undesirable handovers during beam switching.

The invention belongs to the video target tracking technology field and especially relates to an anti-occlusion target tracking method based on particle filtering and weighting Surf. The method comprises the following steps of firstly, initializing a target template; then, establishing a particle state transfer and observation model and using the particle filtering to predict a target candidate area; secondly, calculating an occlusion factor and determining whether a target generates occlusion; and then, if the target generates the occlusion, using extended Kalman filter to predict a target position again; if the target does not generate the occlusion, for the target candidate area determined by the particle filtering, extracting Surf characteristic points and matching with the target template, and accurately positioning the target position and an area; finally, according to the number of registering characteristic point pairs, deciding whether to use a forgetting factor mode to dynamically update the template. In the method, technologies of the particle filtering, occlusion determination, the extended Kalman filter, weighting Surf registering and the like are combined, tracking accuracy and robustness when the target generates the occlusion are increased and a good application prospect is possessed.

A communication control apparatus includes: a weighted average processor configured to update an average value of measured values at predetermined timings by use of a weighted average process using a forgetting factor; a forgetting factor calculator configured to increase the forgetting factor gradually; and a communication controller configured to control communications based on the average value.

The invention provides a multi-sensor passive synergic direction finding and positioning method.By the adoption of the method, target positioning capacity under the condition that a multi-passive-sensor observed value hops, time is asynchronous and precision difference is large can be improved remarkably, and high-precision synergic direction finding and positioning can be achieved.According to the technical scheme, multi-sensor measurement data is input into passive synergic direction finding and positioning software, outliers distinguishing is conducted on the observed value by means of a predicted residual, and extrapolation is conducted with the extension forgetting factor recursive least-square filtering algorithm to the same moment; then a weighting matrix is calculated with a measurement covariance matrix, the observation linear equation system of multiple sensors is established, the initial estimate value of a target position is solved, increment is calculated with the weighting gauss-newton descent method, iterative solution is conducted on the target position, the significance testing statistics of the increment is judged, and iterative computation is stopped and the iterative estimation solution obtained at the moment is output to serve as final target positioning information when it is judged that estimation convergence occurs.

The invention discloses a mobile medical care remote monitoring system and a data transmission method thereof. The mobile medical care remote monitoring system comprises a mobile terminal and a mobile medical care data center system which are in communication connection through a service platform. The data transmission method comprises the following steps of: evaluating physiological status data of the human body, further sending a control instruction of further providing status data to the mobile terminal when the health status of a patient is worsened, and sending a control instruction of reducing status data to the mobile terminal when the health status of the patient gets right; carrying out health evaluation by adopting multiple types of status data; processing multiple types of physiological status data of the human body by adopting a data fusion technology, and mapping the processed physiological status data into new parameters representing the health status of the patient; and smoothening the data by adopting a forgetting factor. According to the mobile medical care remote monitoring system and the data transmission method thereof, the data transmission quantity of the mobile terminal of the patient is self-adaptively adjusted, the processing capacity of the mobile medical care remote monitoring system is improved, the unnecessary mobile communication flow cost is saved and the popularization and generalization of a mobile medicine heath care system are facilitated.

The invention discloses a method for forecasting travel time between bus stops on the basis of self-adaptive attenuation Kalman filtering, including steps of acquiring data of the history travel time of a bus between bus stops by the aid of a vehicle-mounted GPS (global position system) device, and then forecasting the future travel time of the bus between bus stops by a self-adaptive attenuation filtering forecasting model. Affection of history data to filtering is suppressed by introducing forgetting factors so as to ensure the vantage ground of new information during the forecasting process, and accordingly the defects of filtering divergence and computing divergence of a conventional forecasting model are overcome. Besides, an exponential declined noise model is used as a filtering start value, the forecasting precision and efficiency of the model is improved effectively. In addition, the method overcomes the defect of poor robustness of the conventional Kalman filtering, and has the advantages of high forecast precision, convenience in computing and high practicability.

The invention discloses a grasping force active compliance control method of an apple picking robot tail end actuator, and belongs to the technical field of picking robot control. The grasping force control method is characterized in that a force sensor and an encoder arranged on the apple picking robot tail end actuator obtain the position transformation quantity and the acting force on a grasping object through collection; the displacement and the force obtained through collection are used as the input of a recursive least squares method recognizer of a variable forgetting factor; the rigidity coefficient of an impedance controller is subjected to on-line identification; the rigidity parameters of the impedance controller for adapting to different environment requirements are automatically regulated in real time according to the output result of a second-order impedance controller. The control method has the advantages that the grasping damage rate of a picking robot to apples can be effectively reduced; in addition, the anti-jamming capability of the robot during field work is improved.

The invention relates to a power battery system identification and state estimation method for a wide sampling interval. Based on a power battery equivalent circuit model, a recursive least square algorithm is used, online parameter identification and state estimation on the power battery in multiple scales are realized, and the state of charge of the power battery is thus acquired. Based on an optimal forgetting factor curve in a time scale, accurate state estimation in a wide sampling interval condition is realized. The method has high precision, the stability is strong, and the needed computing power and the storage space are extremely small.

The invention discloses an online calibration method of a star sensor assisted gyroscope for ships, and relates to the technical field of inertial sensor error calibration in inertial navigation. Theonline calibration method comprises following steps: (1) initializing a star sensor / inertia combined navigation system; (2) collecting the output data of an inertia device and a star sensor; (3) calculating the inertia device to obtain attitude quaternion and navigation information; (4) obtaining the real value of the carrier attitude quaternion; (5) utilizing a fuzzy logic control method to determine the forgetting factor of a simplified Sage-Husa adaptive filtering method, and then subjecting the carrier attitude quaternion information to filter calculation; (6) compensating the gyroscope output angular velocity information based on the gyroscope output error, and carrying out navigation calculation; and (7) storing and outputting navigation information. The applicability of online calibration on the star sensor assisted gyroscope in a complicated environment is improved. The problem that the estimate on noise error is not sufficient in measurement is solved. The attitude precision of a star sensor is effectively improved.

An apparatus and method facilitate combined query based searching with serendipitous browsing in a multimedia collection. A user selects objects to label from a local map, which may include representations of objects retrieved from the collection as being responsive to a text or image base query. The text and image portions of the object can be independently labeled. Unlabeled objects are scored and ranked based on the applied labels of labeled objects, which may take into account cross-media pseudo-relevance and user selectable (or default) parameters, such as a forgetting factor, which tends to place greater weight on more recently labeled objects, and a modality parameter, which laces greater weight on the modality (text, image, or hybrid) currently selected by the user. The local map is modified, based on the ranking, optionally after reranking of objects to improve the diversity of the displayed objects.

The invention discloses a high-voltage parallel hybrid active power filter device and an iterative learning control method with a forgetting factor. The device is composed of a passive power filter, an active power filter and a coupling transformer. A steplength-variable fuzzy self-adapting harmonic wave detection method is adopted, so as to coordinate the requirements of steady-state performance and convergence rate of the system, and the detected electric network harmonic current is used as the system control target. An iterative learning control algorithm with the forgetting factor is adopted, and an improved Ziegler-Nichols method is utilized to realize the optimization of control parameters and improve the system control performance. The high-voltage parallel hybrid active power filter device has certain advantages in response time and control accuracy, can satisfy the requirements of large-capacity hybrid active filtering system, not only can well control the harmonic wave in the electric network, but also can compensate large-capacity reactive power, and is reliable in systems operation.

The invention discloses a design method for a digital filter of an inertial measurement unit (IMU) of a mechanically-dithered laser gyroscope. The invention is characterized in that the filter comprises a band-stop part and a low-pass part, and the method comprises the following steps of: firstly, designing the band-stop part of a linear phase infinite impulse response (IIR) digital filter according to the mechanically-dithered noise characteristic of the mechanically-dithered laser gyroscope; secondly, performing least square parameter estimation with forgetting factors in a time domain to make the low-pass part of the linear phase IIR digital filter approach a linear phase finite impulse response (FIR) low-pass filter, and thus obtaining an approximately linear phase; and finally, establishing a target function of the low-pass part of the linear phase IIR digital filter in a frequency domain, optimizing the target function by using a Daridon-Fletcher-Powell (DFP) method, and correcting a time domain design result to make the low-pass part of the linear phase IIR digital filter have expected frequency characteristic. The method has the advantages of approximately linear phase, low operation amount and small group delay, requirements of the filter for IMU hardware are reduced, the measurement real-time performance of the IMU is improved, time-delay compensation accuracy is ensured, and a foundation is laid for high-accuracy and real-time measurement of the IMU.