CN108134549A - A kind of method for improving permanent magnet synchronous motor speed estimate stability - Google Patents

A kind of method for improving permanent magnet synchronous motor speed estimate stability Download PDF

Info

Publication number
CN108134549A
CN108134549A CN201711421023.XA CN201711421023A CN108134549A CN 108134549 A CN108134549 A CN 108134549A CN 201711421023 A CN201711421023 A CN 201711421023A CN 108134549 A CN108134549 A CN 108134549A
Authority
CN
China
Prior art keywords
matrix
permanent magnet
magnet synchronous
synchronous motor
covariance
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201711421023.XA
Other languages
Chinese (zh)
Inventor
尹忠刚
高峰涛
张延庆
刘静
钟彦儒
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian University of Technology
Original Assignee
Xian University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian University of Technology filed Critical Xian University of Technology
Priority to CN201711421023.XA priority Critical patent/CN108134549A/en
Publication of CN108134549A publication Critical patent/CN108134549A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/13—Observer control, e.g. using Luenberger observers or Kalman filters
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
    • H02P21/18—Estimation of position or speed
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P25/00—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details
    • H02P25/02—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details characterised by the kind of motor
    • H02P25/022—Synchronous motors
    • H02P25/024—Synchronous motors controlled by supply frequency
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P2203/00—Indexing scheme relating to controlling arrangements characterised by the means for detecting the position of the rotor
    • H02P2203/09—Motor speed determination based on the current and/or voltage without using a tachogenerator or a physical encoder

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Control Of Ac Motors In General (AREA)
  • Control Of Electric Motors In General (AREA)

Abstract

A kind of method for improving permanent magnet synchronous motor speed estimate stability disclosed by the invention, include the following steps, adaptive factor is solved first with prediction error vector and measurement noise covariance matrix, then adaptive factor is introduced into the status predication covariance matrix and measurement noise auto-covariance and Cross-covariance of traditional Unscented Kalman filter, structure is based on adaptive covariance Unscented Kalman filter, real-time online adjusts status predication covariance matrix, when there are unusual fluctuations in System State Model, influence of the system state amount for permanent magnet synchronous motor vector control system can be adaptively adjusted with reference to covariance observation in adaptive factor.The present invention by adaptive factor by introducing UKF algorithms, form the permanent magnet synchronous motor method for estimating rotating speed based on adaptive covariance UKF, the influence of initial value deviation and System State Model fluctuation is restrained effectively, so as to improve the stability of permanent magnet synchronous motor speed estimate.

Description

A kind of method for improving permanent magnet synchronous motor speed estimate stability
Technical field
The invention belongs to permanent magnet synchronous motor control technology fields, and it is steady to be related to a kind of raising permanent magnet synchronous motor speed estimate Qualitatively method.
Background technology
Permanent-magnet Synchronous-motor Speed Servo System has become the research hotspot in Prospect of AC Adjustable Speed Drive field.It is controlled in modern electrical machine Technical field, due to the use of velocity sensor, to destroy permanent magnet motor structure simple, reliable, at low cost, easy to maintain excellent Gesture limits its scope of application, and reduces the robustness of system, and therefore, deadlock_free scheduling becomes modern AC biography One important research direction of dynamic control technology.At present, domestic and foreign scholars have carried out numerous studies to deadlock_free scheduling, And many practical effective methods are proposed, such as sliding mode observer, adaptive full order observer, model reference are adaptive Should, High Frequency Injection, Extended Kalman filter etc., wherein, Extended Kalman filter (Extended Kalman Filtering, EKF) it is widely used in speed-sensorless control speed estimate, it provides a kind of iteration form Non-linear estimation algorithm is avoided and is differentiated to measured value, and can be corrected by gain matrix and new breath error Estimated value makes state estimation error tend to be minimum.But EKF is there is also the deficiency of itself, since there is the mistakes of linearisation Journey and component more than second order is had ignored, therefore there are large errors on filtering accuracy；On the other hand, for it is certain can not The nonlinear function of differential is difficult often to seek its Jacobi matrix, and which limits the applications of EKF methods.
In order to make up the deficiency of EKF algorithms, people carry out approximation by the probability density distribution to complex nonlinear function, So as to which instead of the approximation to nonlinear function, nonlinear propagation is solved the problems, such as using the method that sampled point approaches.It is a kind of new Non-linear filtering method-unscented Kalman filter (Unscented Kalman Filtering, UKF) converted using U (Unscented Transform, UT) carrys out the probability density distribution of approximate non-linear function, therefore UKF algorithms are in computational accuracy On be higher than EKF, while UKF does not need to calculate the Jacobi matrixes of state-transition matrix, this causes it using more convenient. Although UKF can overcome the problems, such as that EKF is existing, in practical application, UKF still has computationally intensive, real-time not High and noise is required to have the statistical property of Gaussian Profile, more sensitive to the value of initial value, robustness is not strong, ability of tracking The problem of poor.
Invention content
The object of the present invention is to provide a kind of methods for improving permanent magnet synchronous motor speed estimate stability, solve existing UKF speed estimates are disturbed the problem of error is larger when influencing by filtering initial value and System State Model.
The technical solution adopted in the present invention is a kind of method for improving permanent magnet synchronous motor speed estimate stability, tool Body operates in accordance with the following steps：
Step 1., which utilizes, predicts that error vector and measurement noise covariance matrix solve adaptive factor αk；
The adaptive factor α that step 2. is obtained using step 1kStructure is based on adaptive covariance unscented Kalman filter Device, real-time online adjustment status predication covariance matrix, obtains state variableWith the optimal estimation value of variance P.
Other features of the present invention also reside in,
It is utilized in step 1 and predicts that error vector and measurement noise covariance matrix solve adaptive factor αkSpecific calculating Process is as follows：
UKF equations are expressed as：
In formula,Estimated value for state variable；Estimated value for observed quantity；Y is observed quantity；A is state matrix；B is Input matrix；H is observing matrix；U is input variable；K is gain matrix, and t represents the time；
By formula (1), (2) discretization structure UKF digitization systems, obtain：
Calculate state variable predicted value
In formula, A ', B ', Hk、KkState matrix, input matrix, observing matrix, gain matrix after respectively discrete；Subscript " k " represents the discrete sampling moment；Subscript "~" is predicted value；Subscript " ^ " be estimated value, it is discrete after state variable xk= (iα,k iβ,k ωr,k θk), it is discrete after observed quantity yk=(iα,k iβ,k)T, it is discrete after input variable uk=(uα,k uβ,k)T, iα、iβRespectively α, β axis stator current, ωrFor rotor velocity, θ is electromagnetic point；
Definition prediction error vector
Adaptive factor can be obtained by the following formula：
Wherein, PyFor measurement noise covariance matrix.
The concrete operation step of step 2 is as follows：
By adaptive factor αkIt is introduced into status predication covariance matrix and the measurement of traditional Unscented Kalman filter In noise covariance matrix, structure is based on adaptive covariance Unscented Kalman filter real-time online adjustment status predication association Variance matrix establishes permanent magnet synchronous motor under α β shaftings and is based on adaptive covariance Unscented Kalman filter digital system System, carries out speed estimate, and concrete operation step is as follows：
Step 2.1 sets original state variable and variance, then constructs and calculates Sigma points；
Step 2.2 asks state variable prediction mean value and variance；
Step 2.3 asks noise covariance matrix to update；
Step 2.4 calculates Kalman filtering gain matrix；
Step 2.5 is modified predicted value update.
The calculating process of step 2.1 is as follows：
Init state variable and its variance
Wherein,For original state mean value, E [] is asks for expectation computing, and x (0) is state variable initial value, P (0) Initial value for state error covariance matrix；
The constructive formula of Sigma points is as follows：
Wherein, χ0Represent first Sigma sampled point, χiRepresent i-th of Sigma sampled point；λ=α2(n+ κ)-n is ratio The example factor；α is a scale parameter；κ is adjustable parameter, when state variable is for multivariable and when being Gaussian Profile, κ=3-n；I-th row of matrix square root or the i-th row, are asked for using Cholesky decomposition methods；
It is as follows that Sigma points are chosen according to formula (9)：
Wherein, k represents sampling instant.
Corresponding weights are assigned to each point after Sigma points have been constructed, weight coefficient computational methods are as follows：
Wherein,Weight coefficient for mean value；Weight coefficient for variance；α is scale parameter, decides that Sigma points exist The distribution of surrounding；β is non-negative weight coefficient.
The calculating process of state variable prediction mean value and variance is as follows in step 2.2：
Wherein, k represents sampling instant,Represent the Sigma points after nonlinear transformation, χi(k-1) it is selection Sigma points, f [] are non-linear system status function,For the predicted value of state variable,For the weight coefficient of mean value,For the predicted value of state error covariance matrix,For the weight coefficient of variance, Q (k) is system noise covariance matrix,For the predicted value of observed quantity, h [] is non-linear observation function.
The renewal equation of noise covariance matrix is as follows in step 2.3：
Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, αkFor adaptive factor,For variance Weight coefficient,Represent the Sigma points after nonlinear transformation,For the predicted value of k moment observed quantities, R (k) is surveys Measure noise covariance matrix,The predicted value of state variable, Pxy(k) it is Cross-covariance, h [] is non-linear observation letter Number.
The accounting equation of Kalman filtering gain matrix is in step 2.4：
Kk=Pxy(k)Py -1(k) (18)
Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, Pxy(k) it is Cross-covariance.
The equation for updating to obtain state variable and variance in step 2.5 to predicted value amendment is as follows：
Wherein, k represents sampling instant,For the predicted value of state variable, y (k) be it is discrete after observed quantity,For The predicted value of observed quantity, KkFor kalman gain matrix, αkFor adaptive factor,For the pre- of state error covariance matrix Measured value, Py(k) it is measurement error covariance matrix, Kk TFor KkTransposition.
The invention has the advantages that a kind of method for improving permanent magnet synchronous motor speed estimate stability, based on adaptive Covariance unscented Kalman filter permanent magnet synchronous motor method for estimating rotating speed is answered, with traditional unscented Kalman filter algorithm phase Than adaptive factor is introduced into the update of status predication covariance matrix and measurement noise covariance matrix by the present invention, is passed through Adjust gain matrix is come to the on-line tuning of status predication covariance matrix and measurement noise covariance matrix, when system mode mould When unusual fluctuations occurs in type, system state amount can be adaptively adjusted for permanent magnetism with reference to covariance observation in adaptive factor The influence of synchronous motor vector control system.Adaptive correction auto-covariance matrix and Cross-covariance, effectively inhibit Initial value deviation and the influence of System State Model fluctuation, so as to improving the stability of permanent magnet synchronous motor speed estimate.
Description of the drawings
Fig. 1 be the present invention a kind of raising permanent magnet synchronous motor speed estimate stability method in permanent magnet synchronous motor swear Amount control system block diagram；
Fig. 2 is to ask for adaptive factor in a kind of method of raising permanent magnet synchronous motor speed estimate stability of the present invention Structure diagram；
Fig. 3 be the present invention a kind of raising permanent magnet synchronous motor speed estimate stability method in speed estimate flow Figure.
In figure, 1. three-phase inverters, 2. permanent magnet synchronous motors, 3. current signal detection circuits, 4.Clark transformation, 5.Park is converted, 6.AUKF speed estimate modules, 7.Park inverse transformations, and module occurs for 8.PWM.
Specific embodiment
The present invention is described in detail with reference to the accompanying drawings and detailed description.
A kind of method of raising permanent magnet synchronous motor speed estimate stability of the present invention, permanent magnet synchronous motor use vector Control system, as shown in Figure 1, system use 3 pi regulators, formed rotating speed, Current Feedback Control double closed loop AC regulation speed System.Input of the output of rotating speed outer shroud pi regulator as electric current pi regulator, the output control electric power electricity of current regulator Sub- converter.
Current signal detection circuit 3 detects three-phase current of the motor under three-phase static coordinate system by Hall sensor, 4 (3s/2s) are converted by Clark, the current value i under two phase coordinate system of convert to staticsα、isβ, then by giving in speed outer shroud Determine rotating speedWith the feedback speed estimated by AUKF speed estimates module 6The error to compare, by speed outer ring controller After adjusting, the q shaft currents under output rotor rotating coordinate systemCurrent value i under static two phase coordinate systemsα、isβAnd AUKF Speed estimate module 6, which estimates electromagnetic point θ and converts 5 (2s/2r) by Park, is converted to two-phase under rotor rotating coordinate system Feedback calculates exciting current idWith torque current iq.Given exciting currentExciting current i is calculated with feedbackdIt compares, by electricity After flowing PI adjustings, the d axis output voltages of two cordic phase rotators are obtainedTorque currentWith feeding back calculating torque electric current iqPhase After comparing, after overcurrent PI adjustings, the q axis output voltages of two cordic phase rotators are obtainedTwo-phase under rotating coordinate system VoltageWithTwo-phase voltage after Park inverse transformations 7 (2r/2s) inverse transformation under two phase coordinate system of convert to staticThe adjusting of module 8 occurs by PWM, generates PWM wave, after three-phase inverter 1, drives permanent magnet synchronous electric Machine 2 works.
A kind of method of raising permanent magnet synchronous motor speed estimate stability of the present invention, is specifically grasped in accordance with the following steps Make：
Step 1., which utilizes, predicts that error vector and measurement noise covariance matrix solve adaptive factor αk；
The adaptive factor structure that step 2. is obtained using step 1 is based on adaptive covariance unscented Kalman filter Device, real-time online adjustment status predication covariance matrix, obtains state variableWith the optimal estimation value of variance P.
The method of a kind of raising permanent magnet synchronous motor speed estimate stability of the present invention, using predicting error in step 1 Vector sum measurement noise covariance matrix solves adaptive factor αk, adaptive factor can adaptively equilibrium state equation it is pre- Notify breath and the weights of observation information, but can influence of regulating system model when being disturbed to filtering, as shown in Fig. 2, specifically Calculating process is as follows：
UKF equations are expressed as：
In formula,Estimated value for state variable；Estimated value for observed quantity；Y is observed quantity；A is state matrix；B is Input matrix；H is observing matrix；U is input variable；K is gain matrix, and t represents the time；
By formula (1), (2) discretization structure UKF digitization systems, obtain：
Calculate state variable predicted value
In formula, A ', B ', Hk、KkState matrix, input matrix, observing matrix, gain matrix after respectively discrete；Subscript " k " represents the discrete sampling moment；Subscript "~" is predicted value；Subscript " ^ " be estimated value, it is discrete after state variable xk= (iα,k iβ,k ωr,k θk), it is discrete after observed quantity yk=(iα,k iβ,k)T, it is discrete after input variable uk=(uα,k uβ,k)T, iα、iβRespectively α, β axis stator current, ωrFor rotor velocity, θ is electromagnetic point；
Definition prediction error vector
Adaptive factor can be obtained by the following formula：
Wherein, PyFor measurement noise covariance matrix.
By formula (7) it is found that adaptive factor αkThe improvement of UKF algorithms is：It chooses to exist when filtering initial value and miss When there is disturbance in difference or System State Model, αk＜ 1, meaning represent the mathematical model prediction information of system to filtering algorithm Influence should be small as possible；When there are unusual fluctuations in System State Model, αkIt is approximately 0.From the above analysis, αkIt can combine Each quantity of state of system is adaptively adjusted for above-mentioned permanent magnet synchronous motor vector control system shown in FIG. 1 in covariance observation Influence.
The concrete operation step of step 2 is as follows：
By adaptive factor αkIt is introduced into status predication covariance matrix and the measurement of traditional Unscented Kalman filter In noise covariance matrix, structure is based on adaptive covariance Unscented Kalman filter real-time online adjustment status predication association Variance matrix, when there are unusual fluctuations in System State Model, adaptive factor αkIt is adaptive covariance observation can be combined Influence of the ground regulating system quantity of state for above-mentioned permanent magnet synchronous motor vector control system shown in FIG. 1, so as to improve forever The stability of magnetic-synchro motor speed estimation establishes permanent magnet synchronous motor under α β shaftings and is based on adaptive covariance without track card Thalmann filter digitization system carries out speed estimate, as shown in figure 3, being as follows：
Step 2.1 sets original state variable and variance, then constructs and calculates Sigma points；
Step 2.2 asks state variable prediction mean value and variance；
Step 2.3 updates noise covariance matrix；
Step 2.4 calculates Kalman filtering gain matrix；
Step 2.5 is modified predicted value update.
The calculating process of step 2.1 is as follows：
Init state variable and its variance
Wherein,For original state mean value, E [] is asks for expectation computing, and x (0) is state variable initial value, P (0) Initial value for state error covariance matrix；
In UKF algorithms, there are different sampling policies, are sampled herein using traditional symmetry, selected Sigma Point number is 2n+1, and the constructive formula of Sigma points is as follows：
Wherein, χ0Represent first Sigma sampled point, χiRepresent i-th of Sigma sampled point；λ=α2(n+ κ)-n, is one A scale factor；α is a scale parameter, decides that Sigma points existThe distribution of surrounding；κ is adjustable parameter, adjusts this ginseng Number can improve approximation accuracy, when state variable is for multivariable and when being Gaussian Profile, κ=3-n；Matrix is put down I-th row of root or the i-th row, generally use Cholesky decomposition methods are asked for；
It is as follows that Sigma points are chosen according to formula (9)：
It needs to assign corresponding weights to each point after Sigma points have been constructed, weight coefficient computational methods are as follows：
Wherein,Weight coefficient for mean value；Weight coefficient for variance；α is a scale parameter, decides that Sigma points existThe distribution of surrounding, is typically set at a smaller positive number 1e-4≤α≤1；β is a non-negative weight coefficient, Ke Yihe And in covariance higher order term moment, the influence of higher order term can be thus included, thus adjust β can improve association side The approximation quality of difference, for Gaussian Profile, it is optimal that β, which takes 2,.
The calculating process of state variable prediction mean value and variance is as follows in step 2.2：
Wherein, k represents sampling instant,Represent the Sigma points after nonlinear transformation, χi(k-1) it is selection Sigma points, f [] are non-linear system status function,For the predicted value of state variable,For the weight coefficient of mean value,For the predicted value of state error covariance matrix,For the weight coefficient of variance, Q (k) is system noise covariance matrix,For the predicted value of observed quantity, h [] is non-linear observation function.
The renewal equation of noise covariance matrix is as follows in step 2.3：
Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, αkFor adaptive factor,For variance Weight coefficient,Represent the Sigma points after nonlinear transformation,For the predicted value of k moment observed quantities, R (k) is surveys Measure noise covariance matrix,The predicted value of state variable, Pxy(k) it is Cross-covariance, h [] is non-linear observation letter Number.
The accounting equation of Kalman filtering gain matrix is in step 2.4：
Kk=Pxy(k)Py -1(k) (18)
Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, Pxy(k) it is Cross-covariance.
It is as follows to correct newer equation for predicted value in step 2.5：
Wherein, k represents sampling instant,For the predicted value of state variable, y (k) be it is discrete after observed quantity,For The predicted value of observed quantity, KkFor kalman gain matrix, αkFor adaptive factor,For the pre- of state error covariance matrix Measured value, Py(k) it is measurement error covariance matrix, Kk TFor KkTransposition.
The present invention is based on adaptive covariance unscented Kalman filter permanent magnet synchronous motor method for estimating rotating speed, with tradition Unscented Kalman filter algorithm is compared, and adaptive factor is introduced the present invention status predication covariance matrix and measurement noise is assisted In the update of variance matrix, adjusted by the on-line tuning to status predication covariance matrix and measurement noise covariance matrix Gain matrix, when System State Model occurs abnormal, adaptive factor can be adaptively adjusted with reference to covariance observation Influence of the system state amount for permanent magnet synchronous motor vector control system.Adaptive correction auto-covariance matrix and mutual association side Poor matrix restrained effectively the influence that initial value deviation and System State Model are disturbed, so as to improve permanent-magnet synchronous The stability of motor speed estimation.

Claims (8)

  1. A kind of 1. method for improving permanent magnet synchronous motor speed estimate stability, which is characterized in that specifically grasp in accordance with the following steps Make：
    Step 1., which utilizes, predicts that error vector and measurement noise covariance matrix solve adaptive factor αk；
    The adaptive factor α that step 2. is obtained using step 1kStructure is real based on adaptive covariance Unscented Kalman filter When on-line tuning status predication covariance matrix, obtain state variableWith the optimal estimation value of variance P.
  2. A kind of 2. method for improving permanent magnet synchronous motor speed estimate stability as described in claim 1, which is characterized in that institute It states to utilize in step 1 and predicts that error vector and measurement noise covariance matrix solve adaptive factor αkSpecific calculating process such as Under：
    UKF equations are expressed as：
    In formula,Estimated value for state variable；Estimated value for observed quantity；Y is observed quantity；A is state matrix；B is input Matrix；H is observing matrix；U is input variable；K is gain matrix, and t represents the time；
    By formula (1), (2) discretization structure UKF digitization systems, obtain：
    Calculate state variable predicted value
    In formula, A ', B ', Hk、KkState matrix, input matrix, observing matrix, gain matrix after respectively discrete；Subscript " k " Represent the discrete sampling moment；Subscript "~" is predicted value；Subscript " ^ " be estimated value, it is discrete after state variable xk=(iα,k iβ,k ωr,k θk), it is discrete after observed quantity yk=(iα,k iβ,k)T, it is discrete after input variable uk=(uα,k uβ,k)T, iα、iβ Respectively α, β axis stator current, ωrFor rotor velocity, θ is electromagnetic point；
    Definition prediction error vector
    Adaptive factor can be obtained by the following formula：
    Wherein, PyFor measurement noise covariance matrix.
  3. A kind of 3. method for improving permanent magnet synchronous motor speed estimate stability as described in claim 1, which is characterized in that institute The concrete operation step for stating step 2 is as follows：
    By adaptive factor αkIt is introduced into the status predication covariance matrix of traditional Unscented Kalman filter and measurement noise association In variance matrix, structure adjusts status predication covariance square based on adaptive covariance Unscented Kalman filter real-time online Battle array establishes permanent magnet synchronous motor under α β shaftings and is based on adaptive covariance Unscented Kalman filter digitization system, into Row speed estimate, concrete operation step are as follows：
    Step 2.1 sets original state variable and variance, then constructs and calculates Sigma points；
    Step 2.2 asks state variable prediction mean value and variance；
    Step 2.3 asks noise covariance matrix to update；
    Step 2.4 calculates Kalman filtering gain matrix；
    Step 2.5 is modified predicted value update.
  4. A kind of 4. method for improving permanent magnet synchronous motor speed estimate stability as claimed in claim 3, which is characterized in that institute The calculating process for stating step 2.1 is as follows：
    Init state variable and its variance
    Wherein,For original state mean value, E [] is state variable initial value to ask for expectation computing, x (0), and P (0) is shape The initial value of state error co-variance matrix；
    The constructive formula of Sigma points is as follows：
    Wherein, χ0Represent first Sigma sampled point, χiRepresent i-th of Sigma sampled point；λ=α2(n+ κ)-n is ratio because Son；α is a scale parameter；κ is adjustable parameter, when state variable is for multivariable and when being Gaussian Profile, κ=3-n；I-th row of matrix square root or the i-th row, are asked for using Cholesky decomposition methods；
    It is as follows that Sigma points are chosen according to formula (9)：
    Wherein, k represents sampling instant；
    Corresponding weights are assigned to each point after Sigma points have been constructed, weight coefficient computational methods are as follows：
    Wherein,Weight coefficient for mean value；Weight coefficient for variance；α is scale parameter, decides that Sigma points existAround Distribution；β is non-negative weight coefficient.
  5. A kind of 5. method for improving permanent magnet synchronous motor speed estimate stability as claimed in claim 3, which is characterized in that institute It is as follows to state the calculating process of state variable prediction mean value and variance in step 2.2：
    Wherein, k represents sampling instant,Represent the Sigma points after nonlinear transformation, χi(k-1) it is selection Sigma points, f [] are non-linear system status function,For the predicted value of state variable,For the weight coefficient of mean value,For the predicted value of state error covariance matrix,For the weight coefficient of variance, Q (k) is system noise covariance matrix,For the predicted value of observed quantity, h [] is non-linear observation function.
  6. A kind of 6. method for improving permanent magnet synchronous motor speed estimate stability as claimed in claim 3, which is characterized in that institute The renewal equation for stating noise covariance matrix in step 2.3 is as follows：
    Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, αkFor adaptive factor,Power system for variance Number,Represent the Sigma points after nonlinear transformation,For the predicted value of k moment observed quantities, R (k) makes an uproar to measure Sound covariance matrix,The predicted value of state variable, Pxy(k) it is Cross-covariance, h [] is non-linear observation function.
  7. A kind of 7. method for improving permanent magnet synchronous motor speed estimate stability as claimed in claim 3, which is characterized in that institute The accounting equation for stating Kalman filtering gain matrix in step 2.4 is：
    Kk=Pxy(k)Py -1(k) (18)
    Wherein, k represents sampling instant, Py(k) it is measurement error covariance matrix, Pxy(k) it is Cross-covariance.
  8. A kind of 8. method for improving permanent magnet synchronous motor speed estimate stability as claimed in claim 3, which is characterized in that institute State predicted value amendment is updated to obtain in step 2.5 state variable and variance equation it is as follows：
    Wherein, k represents sampling instant,For the predicted value of state variable, y (k) be it is discrete after observed quantity,For observation The predicted value of amount, KkFor kalman gain matrix, αkFor adaptive factor,Prediction for state error covariance matrix Value, Py(k) it is measurement error covariance matrix, Kk TFor KkTransposition.
CN201711421023.XA 2017-12-25 2017-12-25 A kind of method for improving permanent magnet synchronous motor speed estimate stability Pending CN108134549A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711421023.XA CN108134549A (en) 2017-12-25 2017-12-25 A kind of method for improving permanent magnet synchronous motor speed estimate stability

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711421023.XA CN108134549A (en) 2017-12-25 2017-12-25 A kind of method for improving permanent magnet synchronous motor speed estimate stability

Publications (1)

Publication Number Publication Date
CN108134549A true CN108134549A (en) 2018-06-08

Family

ID=62392423

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711421023.XA Pending CN108134549A (en) 2017-12-25 2017-12-25 A kind of method for improving permanent magnet synchronous motor speed estimate stability

Country Status (1)

Country Link
CN (1) CN108134549A (en)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108900128A (en) * 2018-09-06 2018-11-27 吉林大学 Direct torque control method for permanent magnetic synchronous electric machine based on Model Predictive Control
CN109391202A (en) * 2018-11-08 2019-02-26 吉林大学 Permanent magnet synchronous motor model prediction-Direct Torque Control
CN109981015A (en) * 2019-03-23 2019-07-05 广东希塔变频技术有限公司 Motor drive control method, device, circuit and transducer air conditioning
CN110417378A (en) * 2019-06-11 2019-11-05 浙江工业大学 A kind of gravity value estimation method based on cold atom interference-type gravimeter
CN110739893A (en) * 2019-10-23 2020-01-31 西安理工大学 improved self-adaptive trackless Kalman filtering rotational inertia identification method
CN110768600A (en) * 2019-11-08 2020-02-07 江苏科技大学 PMSM speed sensorless rotor detection method
CN111290008A (en) * 2020-03-23 2020-06-16 兰州交通大学 Dynamic self-adaptive extended Kalman filtering fault-tolerant algorithm
CN112083349A (en) * 2020-08-01 2020-12-15 南通长江电器实业有限公司 Method for diagnosing turn-to-turn short circuit fault of stator winding of permanent magnet synchronous motor
CN112671373A (en) * 2020-12-21 2021-04-16 北京科技大学 Kalman filtering self-adaptive filtering algorithm based on error control
CN112688596A (en) * 2020-12-04 2021-04-20 西安理工大学 Position sensorless control method based on UPF permanent magnet linear synchronous motor
CN112968643A (en) * 2021-02-01 2021-06-15 南京邮电大学 Based on self-adaptation extension H∞Filtering brushless direct current motor parameter identification method
CN114257150A (en) * 2022-01-26 2022-03-29 合肥倍豪海洋装备技术有限公司 Permanent magnet synchronous motor dead zone compensation method based on improved Kalman filter
CN114499310A (en) * 2022-02-14 2022-05-13 南京理工大学 Rocker arm servo control method based on Kalman filter
CN117526795A (en) * 2023-11-20 2024-02-06 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor based on two-dimensional EKF
CN117544030A (en) * 2023-11-14 2024-02-09 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor for reducing estimation complexity
CN117725687A (en) * 2024-02-06 2024-03-19 华东交通大学 Electric automobile road surface adhesion coefficient estimation method based on permanent magnet synchronous motor

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103338003A (en) * 2013-06-28 2013-10-02 西安交通大学 Online simultaneous identifying method for load torque and inertia of motor
CN103684183A (en) * 2013-11-14 2014-03-26 西安理工大学 Revolving speed estimation method of asynchronous motor
CN106253782A (en) * 2016-07-27 2016-12-21 西安理工大学 EKF Rotational Speed of Asynchronous Motor method of estimation based on method of least square
CN106487297A (en) * 2016-11-24 2017-03-08 北京邮电大学 A kind of PMSM parameter identification method based on covariance matching Unscented kalman filtering algorithm
CN106849801A (en) * 2016-12-20 2017-06-13 江苏大学 A kind of induction-type bearingless motor method for estimating rotating speed

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103338003A (en) * 2013-06-28 2013-10-02 西安交通大学 Online simultaneous identifying method for load torque and inertia of motor
CN103684183A (en) * 2013-11-14 2014-03-26 西安理工大学 Revolving speed estimation method of asynchronous motor
CN106253782A (en) * 2016-07-27 2016-12-21 西安理工大学 EKF Rotational Speed of Asynchronous Motor method of estimation based on method of least square
CN106487297A (en) * 2016-11-24 2017-03-08 北京邮电大学 A kind of PMSM parameter identification method based on covariance matching Unscented kalman filtering algorithm
CN106849801A (en) * 2016-12-20 2017-06-13 江苏大学 A kind of induction-type bearingless motor method for estimating rotating speed

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘铮: "《UKF算法及其改进算法的研究》", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108900128A (en) * 2018-09-06 2018-11-27 吉林大学 Direct torque control method for permanent magnetic synchronous electric machine based on Model Predictive Control
CN108900128B (en) * 2018-09-06 2021-09-28 吉林大学 Direct torque control method of permanent magnet synchronous motor based on model predictive control
CN109391202A (en) * 2018-11-08 2019-02-26 吉林大学 Permanent magnet synchronous motor model prediction-Direct Torque Control
CN109391202B (en) * 2018-11-08 2021-09-28 吉林大学 Model prediction-direct torque control method for permanent magnet synchronous motor
CN109981015A (en) * 2019-03-23 2019-07-05 广东希塔变频技术有限公司 Motor drive control method, device, circuit and transducer air conditioning
CN110417378A (en) * 2019-06-11 2019-11-05 浙江工业大学 A kind of gravity value estimation method based on cold atom interference-type gravimeter
CN110417378B (en) * 2019-06-11 2023-07-11 浙江工业大学 Gravity value estimation method based on cold atom interference type gravity meter
CN110739893A (en) * 2019-10-23 2020-01-31 西安理工大学 improved self-adaptive trackless Kalman filtering rotational inertia identification method
CN110768600A (en) * 2019-11-08 2020-02-07 江苏科技大学 PMSM speed sensorless rotor detection method
CN111290008A (en) * 2020-03-23 2020-06-16 兰州交通大学 Dynamic self-adaptive extended Kalman filtering fault-tolerant algorithm
CN112083349A (en) * 2020-08-01 2020-12-15 南通长江电器实业有限公司 Method for diagnosing turn-to-turn short circuit fault of stator winding of permanent magnet synchronous motor
CN112688596A (en) * 2020-12-04 2021-04-20 西安理工大学 Position sensorless control method based on UPF permanent magnet linear synchronous motor
CN112671373A (en) * 2020-12-21 2021-04-16 北京科技大学 Kalman filtering self-adaptive filtering algorithm based on error control
CN112671373B (en) * 2020-12-21 2024-04-26 北京科技大学 Kalman filtering self-adaptive filtering algorithm based on error control
CN112968643A (en) * 2021-02-01 2021-06-15 南京邮电大学 Based on self-adaptation extension H∞Filtering brushless direct current motor parameter identification method
CN112968643B (en) * 2021-02-01 2022-08-26 南京邮电大学 Based on self-adaptation extension H ∞ Filtering brushless direct current motor parameter identification method
CN114257150A (en) * 2022-01-26 2022-03-29 合肥倍豪海洋装备技术有限公司 Permanent magnet synchronous motor dead zone compensation method based on improved Kalman filter
CN114499310A (en) * 2022-02-14 2022-05-13 南京理工大学 Rocker arm servo control method based on Kalman filter
CN117544030A (en) * 2023-11-14 2024-02-09 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor for reducing estimation complexity
CN117544030B (en) * 2023-11-14 2024-05-14 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor for reducing estimation complexity
CN117526795A (en) * 2023-11-20 2024-02-06 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor based on two-dimensional EKF
CN117526795B (en) * 2023-11-20 2024-04-30 江南大学 Sensorless control method and sensorless control system for permanent magnet synchronous motor based on two-dimensional EKF
CN117725687A (en) * 2024-02-06 2024-03-19 华东交通大学 Electric automobile road surface adhesion coefficient estimation method based on permanent magnet synchronous motor

Similar Documents

Publication Publication Date Title
CN108134549A (en) A kind of method for improving permanent magnet synchronous motor speed estimate stability
CN110739893B (en) Improved self-adaptive trackless Kalman filtering rotational inertia identification method
Naidu et al. Dynamic voltage restorer with quasi-Newton filter-based control algorithm and optimized values of PI regulator gains
CN111884555B (en) Filtering estimation method for rotating speed and position of permanent magnet synchronous motor rotor
Xia et al. Speed adaptive sliding mode control with an extended state observer for permanent magnet synchronous motor
CN103944481A (en) AC asynchronous motor vector control system model parameter online modifying method
Mansouri et al. Genetic algorithm optimized robust nonlinear observer for a wind turbine system based on permanent magnet synchronous generator
Aryza et al. Adaptive speed estimation of induction motor based on neural network inverse control
Li et al. A novel adaptive self-turned PID controller based on recurrent-wavelet-neural-network for PMSM speed servo drive system
Yin et al. A speed estimation method for induction motors based on strong tracking extended Kalman filter
Rigatos et al. A nonlinear optimal control approach for shipboard AC/DC microgrids
Fan et al. Rotor resistance online identification of vector controlled induction motor based on neural network
Chaouch et al. Optimized torque control via backstepping using genetic algorithm of induction motor
CN116805849A (en) Continuous set model prediction control method of permanent magnet synchronous motor
CN111092579A (en) Asynchronous motor self-adaptive vector control system with stator temperature on-line monitoring function
Tian et al. Speed-sensorless control of induction motors based on adaptive EKF
Xiao et al. Parameter identification of direct‐drive permanent magnet synchronous generator based on EDMPSO‐EKF
CN111092578B (en) Vector control method for accurately orienting rotor magnetic field of asynchronous motor
CN113644852B (en) Robust three-vector model prediction flux linkage control method for permanent magnet synchronous motor
CN115392110A (en) Data model hybrid drive wind power plant modeling method based on PPO algorithm
CN113965129A (en) Compensation method for current measurement offset error of permanent magnet synchronous motor control system
Rigatos et al. Derivative‐free nonlinear Kalman filtering for PMSG sensorless control
CN111181461A (en) DFIG stator flux observer and method based on resistance online identification
CN111614299A (en) Direct torque control method based on ant colony optimization PID permanent magnet synchronous motor
Aguilar Mejía et al. Adaptive Speed Controller for a Permanent Magnet Synchronous Motor

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20180608

RJ01 Rejection of invention patent application after publication