106results about How to "Realize online estimation" patented technology

The invention discloses a quick prototype design method for gas path fault diagnosis for an aeroengine. The quick prototype design method comprises the following steps of: building an adaptive model of the engine; designing a gas path fault diagnosis logic in the full service life of the aeroengine, and implementing abnormal monitoring and gas path performance on-line estimation; and designing a quick prototype design platform for gas path fault diagnosis for the engine, and checking abnormal monitoring and gas path performance on-line estimation functions of the platform. The adaptive model of the engine is built by a compound interference method; an engine on-board adaptive model is used for estimating performance parameters; an engine adaptive base line model is used for monitoring abnormities and performs off-line period updating; and the quick prototype design method is used for checking the engine gas path fault diagnosis logic. The invention also discloses a corresponding quick prototype design platform. The method and the platform are relatively high in capacity of monitoring abnormities and estimating the performance in the overall service life of the engine and have great significance for shortening the development period and reducing the test risk and the test cost.

The invention discloses a multiple input multiple output (MIMO) radar array position error self-correcting method based on a genetic algorithm, belonging to an MIMO radar array position error self-correcting method. The method comprises the following steps of: carrying out one-dimensional spectral peak search according to an ideal array element position by utilizing a MUSIC method to obtain the initial value theta k of direction of arrival (DOA); generating an initial group, carrying out individual fitness evaluation according to f(G), and finding out f max; carrying out genetic evolution operations of selection, crossing, heteromorphosis and the like to generate a new generation of search group, and carrying out fitness evaluation to reduce a deviation factor sigma; revising the theta k according to the optimal individual in the population; and carrying out genetic iteration until a greatest iteration number is set. By using the method, online estimation and correction of array position error can be finished while the DOA estimation is carried out, thereby the robustness of estimation of system parameters is improved.

The invention discloses an electronic transformer error prediction method. The method comprises: collecting error data and environmental parameter data of an electronic transformer to generate sampledata and rejecting abnormal data among the collected data; on the basis of a Z-score standardization method, carrying out standardization processing on the sample data; carrying out clustering processing on historical data of environmental parameters and establishing an electronic transformer error prediction model by training and learning; and according to the environmental parameter values, predicting a ratio error and an angular difference of the electronic transformer based on a prediction model. The method has the following advantages: no physical model needs to be established; on the basis of the multi-dimensional data driving method, the online estimation of the electronic transformer error can be realized according to the error data and environmental parameter data of the electronic transformer, so that a problem that no deterministic function relation exists between the error and the environmental parameter of the electronic transformer is solved and the accuracy of electronictransformer error prediction is improved.

The invention discloses a system for estimating the position of a microphone in a distributed microphone array. The system comprises a sound source signal generating module, a loudspeaker, a reference microphone, the microphone to be positioned and a microphone position estimation module. Signals output by the output end of the sound source signal generating module are converted into sound through the loudspeaker to be played in a work place of the distributed microphone array; the sound is received by the reference microphone and the microphone to be positioned; the signals received by the reference microphone and the signals received by the microphone to be positioned are simultaneously transmitted to the microphone position estimation module corresponding to the microphone to be positioned. Three signals different in frequency band are used as sounding signals of three sound sources, the three sound sources are collected and separated in the microphone position estimation process in a one-off mode, and then the position of the microphone can be estimated on line. The system combines the ranging methods based on energy and time delay, rough estimation and accurate estimation of the distance are performed successively, and thus the position can be estimated with less calculation amount and higher accuracy.

The invention relates to state-of-charge estimation method and system for compensating non-smooth hysteresis in power batteries. The method includes: firstly, acquiring battery output voltage and current, establishing a neural network OCV (k), (open circuit voltage) (k), prediction model according to the relation of parameters obtained by a battery equivalent circuit model, solving the parameters, and estimating OCV (k) online; and secondly, serially connecting an SDH (simple dynamic hysteresis) model with an RBF2 (radial basic function 2) to form a dynamic hysteresis hybrid model. The SDH model using the OCV (k) obtained in the first step as input outputs y(k), OCV (k) and OCV (k-1); the RBF2 using the y(k), the OCV (k) and the OCV (k-1) as input is weighted to learn so as to indirectly adjust parameters of the SDH model; approximately actual complex hysteresis relation is obtained; online-estimated SOC (k), (state of charge) (k), is output finally. The system comprises a microprocessor and other parts installed on a battery circuit, such as current and voltage sensors, stores a program executing the method, and obtains an SOC (k) estimate. Non-smooth hysteresis in power batteries is compensated by the aid of a neural network, and precision in online estimation of the SOC (k) is increased.

An in-line compensation method for the match error of multiple sensors integrated tracking system features that the maximal probability estimation method is used to determine the initial error, extracting the system error by use of the information measured by sensors, and the filtered tracking information, and performing in-line compensation.

The present invention relates to an online estimation method of an SOC (State of Charge), an electronic device and a storage medium. The method comprises the steps of: obtaining a quantitative function relation Uocv(SOCk) of an open-circuit voltage Uocv and an SOCk, obtained based on a pulse charge-discharge experiment, of a battery, and determining an initial parameter value based on a second-order RC equivalent circuit model; establishing a battery model function based on the second-order RC equivalent circuit model, and obtaining a state equation and an observation equation of the SOC and apolarization voltage according to the battery model function, the quantitative function relation Uocv(SOCk) and an SOC defined function; obtaining a state equation and an observation equation of internal parameters of the battery corresponding to the initial parameter value based on the battery model function; and performing SOC online estimation. The online estimation precision of the SOC of thebattery is high in precision and low in workload.

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.

The invention provides a determining method and device of the fatigue damage of a tower part of a wind generating set. The method includes the steps of obtaining the tower top absolute displacement, wherein the tower top absolute displacement refers to the displacement of the top of a tower relative to the bottom of the tower; obtaining the tower upper part relative displacement, wherein the tower upper part relative displacement refers to the displacement of a preset position on the tower relative to the axis; obtaining a tower load according to the obtained tower top absolute displacement and the obtained tower upper part relative displacement; determining the fatigue damage value of a preset part on the tower within a preset time duration according to the obtained tower load. By means of the method and device, the fatigue damage value of the preset part on the tower is determined through the tower top absolute displacement and the tower upper part relative displacement collected in real time without depending on historical running data of the wind generating set, and the estimation reliability and accuracy of the residual life of the whole tower of the wind generating set are improved.

The invention relates to a GPR (gaussian process regression) online soft measurement method with model updating. Characteristics and various parameters of the actual industrial process are changed with the passage of time, and a soft measurement model built by a traditional off-line method cannot adaptively adjust the model parameters. The invention provides a model updating method integrating local updating and overall updating so as to solve the type of problems. The method provided by the invention comprises the steps of firstly carrying out off-line modeling on a training sample by using a method of gaussian process regression (GRP) so as to acquire a predicted output value and a predicted error; analyzing the predicted error acquired by the off-line model, carrying out overall updating on the GPR model when the error mean is greater than a certain preset threshold, that is, parameters of a covariance matrix and a covariance function of the GPR model are updated simultaneously; otherwise, carrying out local updating on the GPR model, that is, only the covariance matrix of the GPR model is updated; and finally, carrying out compensation on the predicted output of the updated GPR model by using an error gaussian mixture model (EGMM) so as to acquire a final predicted result. The effectiveness of the method provided by the invention is verified through carrying out modeling and simulation on the actual industrial sewage treatment process.

The invention discloses a relative pose and inertia estimation method of a spatial complete non-cooperative target. The method comprises the steps that 1, two industrial computers which are arranged on a tracking spacecraft in a left-right spaced manner and are the same in parameters obtain image information of the non-cooperative target in real time, and 3D position and speed of each characteristic point in the camera coordinate system are calculated; 2, the relative angular velocity of the non-cooperative target is calculated from 3D positions and speeds of at least three characteristic points according to a rigid body motion model, and the relative pose of the non-cooperative target at random time is estimated; 3, the centroid position, the centroid speed and the relative positions of the characteristic points of the non-cooperative target are estimated according to the 3D positions and speeds as well as the relative pose and angular speed of the rigid body; and 4, a rotation inertia parameter of the non-cooperative target is estimated. Thus, the centroid position, the centroid speed, the relative pose, the relative angular speed and the inertial parameter of the complete non-cooperative target can be resolved needless of knowing the geometric shape and characteristic point positions of the complete non-cooperative target in advance.

The invention discloses an on-line judgment method for capacity of a backup storage battery. The prior storage battery capacity measuring technology is lagged. The on-line judgment method for the capacity of the backup storage battery comprises the following steps: obtaining a corresponding relationship between the internal resistance and the capacity of the storage battery through actual measurement, setting the internal resistance of the battery obtained according to the capacity of the storage battery from the corresponding relationship as R1, setting the internal resistance of the storagebattery according to the actual measurement as R2, and calculating a reference value R0 of the internal resistance according to the following formula: R0=omega R1 + (1-omega) R2, wherein in the formula, the omega is a weighting coefficient, the measured current internal resistance of the storage battery correspondingly searched from a graph obtained through the actual measurement is Ri, then the percentage of delta R is equal to (Ri-R0)/R0, and the current capacity range of the storage battery can be determined by a corresponding relationship table of the actual measurement according to the numerical value of the percentage of delta R. The method is used for on-line measurement for the capacity of the storage battery.

The invention discloses a lithium ion battery state of health linear evaluation method. The method includes the following steps that: step S1, m health factor indicators are extracted in n times of the constant-current constant-voltage charging process of a single battery or a plurality of batteries, so that a m*n-dimension original data matrix A is formed; step S2, each row of vectors of the matrix A is normalized, so that various types of normalized health factor vectors are obtained; step S3, an ampere-hour integral method is adopted to solve available battery capacity during the n times ofthe constant-current constant-voltage charging process; step S4, standardization processing is performed on the available battery capacity Qchg, i during the n times of the constant-current constant-voltage charging process, so that a vector q is obtained; and step S5, the Pearson's correlation coefficients pq,xi of the various types of normalized health factor vectors xi and the vector q are calculated. With the method adopted, the problems of low robustness, complicated modeling process, low computational efficiency and low accuracy of methods in the prior art can be solved.

The invention discloses a soft sensing method based on a probabilistic principal component regression model for online detection of butane content of a debutanizer. Principal component regression model is a common modeling approach for soft sensing. However, as noise information of process data is not taken into consideration well, the principal component regression model fails in many real processes. The method introduces a probabilistic modeling approach which converts the conventional principal component regression model into a probabilistic form, which is a soft-sensing model based on probabilistic principal component regression. Compared to the conventional principal component regression model, the probabilistic principal component regression model allows modeling on the process data and the noise information at the same time and provides more complete data information, so that the soft sensing results are more reliable.

The invention discloses a single-battery-based on-line estimation method for a state of charge (SOC) of a series battery pack. The method comprises: a central controller sends a time reference and a soft synchronization sampling time interface to all local controllers; the central controller caries out current sampling and records a time stamp; all local controllers carry out voltage and temperature sampling on single batteries and record time stamps; all local controllers carry out soft time stamp synchronization on voltages, currents, and temperatures of all single batteries; all local controllers carry out recursion and on-line identification of parameters of all single batteries; all local controllers carry out adaptive estimation on SOCs of all single batteries; and the central controller calculates an SOC of the series battery pack. With the method, on-line estimation of the SOCs of all single batteries in the series battery pack as well as on-line estimation of the SOC of the series battery pack based on the on-line estimation of the SOCs of all single batteries can be realized. The method has characteristics of high SOC estimation precision and good adaption to different application environment conditions, the operation condition, and the lift cycle of the series battery pack.

The invention discloses a double-MEMS accelerometer-based online tilt angle measuring method. According to the method, the three-axis zero offset error registers of a first MEMS accelerometer and a second MEMS accelerometer are established; the three-axis output of the two MEMS accelerometers at one moment of a moving object is obtained; whether the three-axis output of the two MEMS accelerometersat the moment is both in a stationary state is judged; and if the three-axis output of the two MEMS accelerometers at the moment is both in the stationary state, the object is considered to be in a stationary state, otherwise, the object is considered to be in a moving state. A double-MEMS accelerometer zero offset online estimation compensation algorithm is adopted to realize the online estimation and compensation of the zero offset of the MEMS accelerometers, and the problem of the accuracy decrease of tilt angle measurement caused by the zero offset variation of the MEMS accelerometers canbe solved, the accuracy of the tilt angle measurement is improved.

The present invention discloses a GPR modeling method based on kernel slow feature analysis and time delay estimation, which is applied to chemical processes with time delay and non-linearity. The method is characterized in that: by using fuzzy curve analysis, the delay information in industrial data is fully tapped to obtain the optimal delay in the data, and the reconstruction of the modeling data is carried out; the kernel slow feature analysis method is further used to carry out nonlinear feature extraction on the reconstructed data; and finally, based on the extracted features, a Gaussian process regression model is established to realize accurate prediction of the key variables, so as to improve product quality and reduce production costs.

The invention relates to an information fusion-based series battery pack SOC (state of charge) estimation method and belongs to the technical field of battery management. The method comprises the following steps that: S1, a series battery pack to be tested is selected, the technical parameters of the series battery pack to be tested are acquired, and the CMM+CDM model of the series battery pack isestablished, and model parameters required by the SOC estimation of the series battery pack are determined; S2, a trickle current charge/discharge test and an HPPC (hybrid pulse power characterization) test are carried out, so that a database of charge/discharge OCVs (open-circuit voltage) and circuit model parameters with respect to SOCs is established, and a DST (dynamic stress test) and a UDDSare simulated, and a working condition test experiment database is established; S3, an OCV-SOC relation curve is obtained, and the electrical characteristic parameters of the CDM+CDM model are obtained; and S4, the CDM+CDM model is combined with a decentralized filter, and a CMM and the SOCs of single cells are obtained on the basis of an optimal algorithm and the quantitative relation of the charge/discharge characteristic parameters with respect to the SOCs, and the SOCs of each cell are fused through weight distribution, so that the SOC of the series battery pack is realized.

The invention discloses an online soft measuring method of glutamate product concentration based on a multi-input wiener model. The method comprises the following steps: firstly, selecting key variables affecting the product concentration change and capable of being measured on line as input variables and establishing a soft measuring model based on the multi-input wiener model by utilizing the product concentration values obtained by offline analysis as output variables; and reasonably selecting the parameters of the multi-input wiener model, establishing an effective nonlinear relationship between the input variables and the output variables, determining the model parameters through a gradient iteration identification method, and evaluating the product concentration which is difficult to measure through variables liable to measure on line in the glutamate production process. The method can be used for effectively improving the prediction precision of the product concentration in the glutamate production process, guiding the production and improving the product quality.

The invention discloses a method for compensating inertia/GPS integrated navigation external lever arm effect errors under rotor wing disturbance, and belongs to the technical field of integrated navigation system measurement. The method comprises the following steps: establishing an external lever arm effect error model, extending the external lever arm effect errors as a system state quantity, establishing a filter state equation and a measurement equation, and achieving online estimation of the external lever arm effect errors based on a Kalman filter. The method provided by the invention can achieve estimation and compensation of the length of an external lever arm to further improve the accuracy of an inertia/GPS integrated navigation system, and is suitable for engineering application.

The invention discloses a central temperature estimation method of a crossed temperature measurement device of a blast furnace, belongs to the field of the automation control of blast furnace smelting, and particularly relates to a central temperature estimation method of the crossed temperature measurement device of the blast furnace on the basis of an M-SVR (Multi-output Support Vector Regression). The central temperature estimation method constructs a central temperature estimation model of crossed temperature measurement, wherein the central temperature estimation model is provided with an output self-feedback structure, considers input data of a previous moment and carries out on-line estimation on the temperature of five temperature measurement points of a crossed temperature measurement center. When the central temperature point of the crossed temperature measurement can not be normally measured, operation personnel can estimate the distribution of gas flow in the furnace according to the temperature estimation value of the model, timely regulates a burden distribution matrix and guarantee the stable direct motion of the blast furnace. The central temperature estimation method comprises the following steps: (1) selecting a model input variable; (2) determining a modeling strategy; and (3) determining a model training sample.

The invention provides a deep space probe angle measurement and time delay integrated navigation method based on on-line estimation, and the method comprises the steps: estimating the current positionand speed information of a probe and a reflection celestial body according to the orbit dynamics, and building a system state model through employing the current position and speed information as a system state quantity; obtaining starlight angular distance measurement between the detector and the reflection celestial body as well as between the detector and the background fixed star by using anangle measurement sensor, and establishing a starlight angular distance measurement model; respectively observing the direct sunlight and the reflected sunlight reflected by the reflection celestial body by using two atom frequency discriminators so as to obtain time delay measurement, and establishing a time delay measurement model; obtaining state estimation and error covariance estimation through implicit unscented Kalman filtering, correcting the actual position and speed information of the reflection celestial body, then obtaining the actual position and speed information of the detector,and achieving detector navigation. The influence of the ephemeris error of the reflection celestial body on the navigation precision is inhibited, and the autonomous navigation precision of the detector is improved.

The invention relates to an on-line estimation method of elbow joint torque under functional electrical stimulation, the musculoskeletal model is pre-modeled by off-line acquisition of joint torque, angle and angular velocity data, and then real-time acquisition of joint angle and angular velocity signals, real-time adaptive updating of the parameters of the pre-modeling musculoskeletal model system, so as to achieve online estimation of torque. The invention realizes on-line estimation of torque through real-time acquisition of joint angle and angular velocity signals.