CN101044673A - Motion controller and system identifying method - Google Patents
Motion controller and system identifying method Download PDFInfo
- Publication number
- CN101044673A CN101044673A CNA2005800357840A CN200580035784A CN101044673A CN 101044673 A CN101044673 A CN 101044673A CN A2005800357840 A CNA2005800357840 A CN A2005800357840A CN 200580035784 A CN200580035784 A CN 200580035784A CN 101044673 A CN101044673 A CN 101044673A
- Authority
- CN
- China
- Prior art keywords
- hat
- aforementioned
- value
- calculate
- carried out
- 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
Links
Images
Landscapes
- Control Of Electric Motors In General (AREA)
Abstract
A motion controller accurately identifies system constants, such as inertia, viscous friction coefficient, and constant disturbance, without requiring operation, even in cases where viscous friction and/or constant disturbance are present. An identifier includes a time-differentiator for calculating an acceleration detection value afb by time-differentiating a speed detection value Vfb, a first filter for calculating Fafb by filtering the acceleration detection value afb, a second filter for calculating Fvfb by filtering the speed detection value Vfb, a third filter for calculating Ftref by filtering a torque command value Tref, and a JDC estimator for calculating an inertia identification value J_hat, a viscous friction coefficient identification value D_hat, and a constant disturbance identification value C_hat of a control object by performing time-differentiation and four arithmetic operations based on Ftref, Fvfb and Fafb.
Description
Technical field
But the present invention relates to precision inertia, viscous friction coefficient and the fixing motion control device and the system identifying method that disturbs of calculation control object well.
Background technology
Calculate the device of the inertia of existing motor, adopt following method: the torque instruction value to torque instruction value and model carries out integration within a certain period of time, making their ratio multiply by the inertia nominal value calculates, at this moment, with specific action, remove viscous friction or static friction, gravity etc. fixedly interference to improve precision (for example, with reference to patent documentation 1).
In Fig. 9, the 3rd, motor, the 4th, be incorporated into the machinery of motor.On motor 3, attach detector 5 is arranged.The 71st, instruction generating unit, the responsiveness instruction vref of its output motor.The 72nd, speed controlling portion, it carries out proportional plus integral control so that instruction is consistent with motor speed, and output torque instruction value Tref.The 75th, the current control division of output current value motor is moved according to torque instruction Tref.The 73rd, estimation portion, it has the model of motor, carries out proportional plus integral control so that instruction is consistent with the speed of model, and the torque instruction Tref ' of output model.The 74th, the Identification Division, it carries out integration to the torque instruction Tref of reality with by the torque instruction Tref ' that estimation portion 73 is input to model in time [a, b] interval, and makes the nominal value J ' that their ratio multiply by inertia try to achieve the inertia estimated value.In the method, the inertia estimated value was not consistent with actual inertia J in theory when interference did not exist fully.
But, at viscous friction or static friction, gravity etc. when fixedly interference exists, must make the integrated value of the torque instruction value Tref of time [a, b] in interval not comprise the restrictive condition of the influence of viscous friction, static friction etc.In addition, in order to remove fixedly interference such as gravity, also need special way.
In patent documentation 1,, use action as shown in figure 10 to be used as restrictive condition as the method for these conditions of realization.
(A)～(F) of Figure 10 is that to get transverse axis be the time, and the longitudinal axis is the curve chart of speed, under which kind of situation, all is the motion of satisfying the restrictive condition of action shown below no matter.
(being used to remove the restrictive condition of the action of viscous friction, static friction etc.) integrated value at interval [a, b] interior speed Vfb is zero.This is the such motion of for example reciprocating action.
When (being used to remove the restrictive condition of the fixing action of disturbing) obtained and moves according to certain speed command at interval [a1, b1] in the inertia J 1 that obtains, and the inertia J 2 that in interval [a2, b2], obtains when moving according to the reverse instruction of carrying out positive and negative counter-rotating, obtain the mean value of J1 and J2.
So, come the identification inertia J on the action of the motion limits condition that more than satisfying, illustrates of existing control constant device for identifying.
Patent documentation 1: No. 96/37039 communique of the WO of Japan (5-8 page or leaf, Fig. 4 and Fig. 5)
Summary of the invention
Existing inertia estimation unit is owing to have above two restrictive conditions enumerating, and for example, existence can't be removed viscous friction or the fixedly influence of interference such as static friction, gravity in one-time positioning action etc., can't calculate the problem that inertia is such.
In addition, when ignoring restrictive condition identification inertia, then exist the inertia error of calculation to become big problem.
The present invention proposes in view of above problem, purpose provides a kind of controller for motor and system identifying method, it is not influenced by action can, even exist viscous friction or static friction, gravity etc. fixedly during the influencing of interference, also can calculate inertia identification value J_hat, the viscous friction coefficient identifier D_hat of controlling object and fixing disturbance identification value C_hat well by simple computational accuracy; The influence of the interference that can be produced by the time diffusion computing of carrying out Filtering Processing minimizing use in calculating thereafter; Can only come identification inertia except that the simple computation of calculation and 5 time diffusion computings by 3 times; Can be only calculate and take advantage of for 1 time also identification constant of friction rarely of the simple computation of calculation and amount of calculation by subtracting for 1 time; Can only take advantage of subtracting of calculation and 1 time 3 variable to calculate and the also fixing interference of identification rarely of amount of calculation by 2 times; Because compare the number of times of denominator with the number of times of the molecule of filter big more than 3 times, therefore can in the identification computing of inertia, carry out 3 times time diffusion.
In addition, the purpose of this invention is to provide a kind of controller for motor and system identifying method thereof, even it exists viscous friction or static friction, gravity etc. fixedly during the influencing of interference, the inertia identification value J_hat and the viscous friction coefficient identifier D_hat that also can calculate controlling object by simple computational accuracy well; The influence of the interference that can be produced by the time diffusion computing of carrying out Filtering Processing minimizing use in calculating thereafter; Can be only by 6 time diffusion computings, take advantage of calculation for 6 times, subtract for 3 times calculate and remove for 2 times the simple computation of calculating come identification inertia and constant of friction the two.
The system identifying method of the present invention program's 1 described motion control device, its motion control device has command generator, according to the instruction of aforementioned command generator output with control and export the controller of the current value that is used for drive motor from the detected value that the detector that is additional to motor obtains, and the inertia identification value J_hat of identification controlling object, viscous friction coefficient identifier D_hat and the fixing identifier of disturbance identification value C_hat, its processing sequence is: speed detected value Vfb is carried out time diffusion to calculate acceleration detection value afb, calculate acceleration detection value afb is carried out signal Fafb after the Filtering Processing, calculate speed detected value Vfb is carried out signal Fvfb after the Filtering Processing, calculate torque instruction value Tref is carried out signal Ftref after the Filtering Processing, with Ftref, Fvfb and Fafb are that time diffusion and arithmetic are carried out in the basis, calculate the inertia identification value J_hat of controlling object, viscous friction coefficient identifier D_hat and fixing disturbance identification value C_hat.
The system identifying method of the present invention program's 2 described motion control devices is the system identifying method as scheme 1 described motion control device, the processing sequence of calculating inertia identification value J_hat is: Ftref, Fvfb, Fafb are carried out time diffusion respectively to calculate Dtref, Dvfb, Dafb, by making Dtref calculate W1 divided by Dvref, by making Dafb calculate W2 divided by aforementioned Dvref, W1 is carried out time diffusion to calculate DW1, W2 is carried out time diffusion to calculate DW2, and then the value that obtains divided by aforementioned DW2 of DW1 is inertia identification value J_hat.
The system identifying method of the present invention program's 3 described motion control devices is the system identifying method as scheme 2 described motion control devices, the processing sequence of calculating aforementioned viscous friction coefficient identifier D_hat is: after calculating inertia identification value J_hat, calculate W3 by making W2 multiply by inertia identification value J_hat, then to deduct the value that W3 obtains be viscous friction coefficient identifier D_hat to W1.
The system identifying method of the present invention program's 4 described motion control devices is the system identifying method as scheme 3 described motion control devices, calculating fixedly, the processing sequence of disturbance identification value C_hat is: after calculating inertia identification value J_hat and viscous friction coefficient identifier D_hat, calculate TF by making Fvfb multiply by aforementioned viscous friction identifier D_hat, calculate TA by making Fafb multiply by inertia identification value J_hat, then Ftref deducts value that TF and TA obtain and is fixing disturbance identification value C_hat.
The system identifying method of the present invention program's 5 described motion control devices is the system identifying method as scheme 1 described motion control device, and it is to be used for the filter of Filtering Processing for compare the big at least filter more than 3 times of number of times of denominator with the number of times of molecule.
The system identifying method of the present invention program's 6 described motion control devices is the system identifying method as scheme 1 to 5 described motion control device, it is to monitor inner quantity of state in advance, only reach set point when above in the aforesaid state amount, calculate inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly processing of disturbance identification value C_hat of controlling object, quantity of state is when set point is not above, and the identifier of J_hat, D_hat, C_hat is kept the value of last time respectively.
The system identifying method of the present invention program's 7 described motion control devices is the system identifying method as scheme 6 described motion control devices, its quantity of state for monitoring in advance is speed Vfb, acceleration afb, torque instruction Tref or these signals is carried out Fvfb, Fafb, Ftref after the Filtering Processing, perhaps any one among Dvfb, the Dw2, perhaps their combination.
The present invention program's 8 described motion control devices have command generator, according to the instruction of command generator output with carry out the positioner of Position Control from the detected value that the detector that is additional to motor obtains, carry out the speed control of speed control, output is used for the current controller of the electric current of drive motor, and the inertia identification value J_hat of identification controlling object, viscous friction coefficient identifier D_hat and the fixing identifier of disturbance identification value C_hat, its identifier constitutes and comprises: speed detected value Vfb is carried out time diffusion to calculate the time diffusion device of acceleration detection value afb, acceleration detection value afb is carried out Filtering Processing to calculate the 1st filter of Fafb, speed detected value Vfb is carried out Filtering Processing to calculate the 2nd filter of Fvfb, torque instruction value Tref is carried out Filtering Processing to calculate the 3rd filter of Ftref, with Ftref, Fvfb and Fafb are that time diffusion and arithmetic are carried out in the basis, calculate the inertia identification value J_hat of aforementioned controlling object, viscous friction coefficient identifier D_hat and the fixing JDC estimator of disturbance identification value C_hat.
The present invention program's 9 described motion control devices are that its JDC estimator comprises as scheme 8 described motion control devices: calculate the inertia estimator of inertia, calculate the viscous friction estimator of viscous friction, calculate the fixing fixedly interference estimator of disturbing.
The system identifying method of the present invention program's 10 described motion control devices, its motion control device has command generator, according to the instruction of aforementioned command generator output be used for the controller of the current value of drive motor from the detected value output that the detector that is additional to motor obtains, and the inertia identification value J_hat of identification controlling object, viscous friction coefficient identifier D_hat and the fixing identifier of disturbance identification value C_hat, its processing sequence is: speed detected value Vfb is carried out time diffusion to calculate acceleration detection value afb, calculate aforementioned acceleration detection value afb is carried out signal Fafb after the Filtering Processing, calculate aforementioned speed detected value Vfb is carried out signal Fvfb after the Filtering Processing, calculate aforementioned torque command value Tref is carried out signal Ftref after the Filtering Processing, calculate aforementioned Ftref, Fvfb, Fafb carries out the Dtref behind the time diffusion respectively, Dvfb, Dafb and carry out time diffusion once more after D2tref, D2vfb, D2afb, by following equation
J_hat＝(D2tref·Dvfb-Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (a)
D_hat＝(D2afb·Dtref-Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (b)
Calculate inertia identification value J_hat and viscous friction coefficient identifier D_hat.
The system identifying method of the present invention program's 11 described motion control devices is as the system identifying method of scheme 10 described motion control devices, it is characterized by: the filter that is used for aforementioned Filtering Processing is for comparing the big at least filter more than 3 times of number of times of denominator with the number of times of molecule.
The system identifying method of the present invention program's 12 described motion control devices is as the system identifying method of scheme 10 described motion control devices, it is characterized by: the device of filter for a plurality of moving average filters are connected in series that is used for aforementioned Filtering Processing.
The system identifying method of the present invention program's 13 described motion control devices is characterized as: the signal of monitoring moving control device inside in advance, only reach set point when above at aforementioned signal, calculate inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly processing of disturbance identification value C_hat of aforementioned controlling object, the aforesaid state amount is when set point is not above, and the identifier of J_hat, D_hat, C_hat is kept the value of last time respectively.
Being characterized as of the system identifying method of the present invention program's 14 described motion control devices: the aforementioned signal of monitoring in advance is speed Vfb, acceleration afb, torque instruction Tref or these signals is carried out Fvfb, Fafb, Ftref after the Filtering Processing, perhaps these signals are carried out Dtref, Dvfb, Dafb, the D2tref behind the time diffusion, any one among D2vfb, the D2afb, perhaps their combination.
The present invention program's 15 described motion control devices, have the command generator of the instruction of generating, according to aforementioned instruction with generate current controller, and inertia identification value J_hat, the viscous friction coefficient identifier D_hat of identification controlling object and the fixing identifier of disturbance identification value C_hat of the electric current that is used for drive motor from the detected value that the detector that is additional to motor obtains, it is characterized by, have: speed detected value Vfb is carried out time diffusion to calculate the acceleration arithmetic unit of acceleration detection value afb; Calculate the 1st filter that aforementioned acceleration detection value afb is carried out the signal Fafb after the Filtering Processing; Calculate the 2nd filter that aforementioned speed detected value Vfb is carried out the signal Fvfb after the Filtering Processing; Calculate the 3rd filter that aforementioned torque command value Tref is carried out the signal Ftref after the Filtering Processing; And calculate D2tref, the D2vfb after aforementioned Ftref, Fvfb, Fafb carried out Dtref, Dvfb, the Dafb behind the time diffusion respectively and carry out time diffusion once more, the differentiator of D2afb, by following equation,
J_hat＝(D2tref·Dvfb_Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (a)
D_hat＝(D2afb·Dtref_Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (b)
Calculate inertia identification value J_hat and viscous friction coefficient identifier D_hat.
According to scheme 1 described the present invention, because to torque instruction, viscous friction and fixing interference use are not the relations that is similar to but are out of shape and (use differential, remove to calculate etc.) equation, utilization is to speed, acceleration and torque instruction value carry out the signal after the Filtering Processing, carry out identification and calculate by only they being carried out time diffusion and arithmetic, not influenced by action, even have viscous friction or static friction, during fixing interference such as gravity, inertia identification value J_hat that also can precision identification controlling object well, viscous friction coefficient identifier D_hat and fixing disturbance identification value C_hat.In addition, by carrying out Filtering Processing can reduce the interference that the time diffusion computing used produced in calculating thereafter influence.
According to scheme 2 described the present invention, owing to can be only calculate and 5 time diffusion computings come identification inertia by removing for 3 times, so can carry out identification with less amount of calculation by simple calculating.
According to scheme 3 described the present invention,, therefore can carry out identification with less amount of calculation by simple calculating owing to can only subtract calculation and take advantage of for 1 time calculation to come the identification constant of friction by 1 time.
According to scheme 4 described the present invention,, therefore can carry out identification with less amount of calculation by simple calculating owing to can only take advantage of subtracting of calculation and 1 time 3 variable to calculate the fixing interference of identification by 2 times.
According to scheme 5 described the present invention, because therefore the number of times of comparing denominator with the number of times of the molecule of filter can carry out 3 times time diffusion more than 3 times in the identification computing of inertia.
According to scheme 6 and scheme 7 described the present invention, because not carrying out identification when set point is following, do not calculate the quantity of state in the monitoring, therefore can prevent to reduce because of small value being removed the precision that causes, no matter can both guarantee identification precision in any case.
According to scheme 8 and scheme 9 described the present invention, a kind of motion control device can be provided, it influences for can not being subjected to action, even exist viscous friction or static friction, gravity etc. fixedly during the influencing of interference, also can precision calculate inertia identification value J_hat, the viscous friction coefficient identifier D_hat of controlling object and fixing disturbance identification value C_hat well.
According to scheme 10 described the present invention, because will be to torque instruction, viscous friction and fixing disturb that to use be not that the resulting equation of time diffusion is carried out on the both sides of equation of the relation that is similar to but distortion, and further carry out the resulting equation of time diffusion and be used as two simultaneous equationss and obtain and separate on the both sides of this equation, and utilize speed, acceleration and torque instruction value carry out the signal after the Filtering Processing, carry out identification and calculate by only they being carried out time diffusion and arithmetic, not influenced by action, even have viscous friction or static friction, gravity etc. are fixedly during the influencing of interference, inertia identification value J_hat that also can precision identification controlling object well, viscous friction coefficient identifier D_hat and fixing disturbance identification value C_ hat.In addition, by carrying out Filtering Processing can reduce the interference that the time diffusion computing used produced in calculating thereafter influence.
According to scheme 11 described the present invention,, therefore can interference-free influence can in the identification computing of inertia, carry out 3 times time diffusion because compare the number of times of denominator with the number of times of the molecule of filter big more than 3 times.
According to scheme 12 described the present invention, because filter uses moving average, can eliminate phase lag poor of each variable, therefore precision is carried out identification well more.
According to scheme 13 and scheme 14 described the present invention, because not carrying out identification when set point is following, do not calculate the signal in the monitoring, therefore can prevent to reduce because of small value being removed the precision that causes, no matter can both guarantee identification precision in any case.
Description of drawings
Fig. 1 is the block diagram of the structure of the expression motion control device of using method of the present invention.
Fig. 2 is the block diagram of the structure of expression the 2nd motion control device of using method of the present invention.
Fig. 3 is the block diagram of the identifier structure of the expression motion control device of using the present invention's the 1st method.
Fig. 4 is the block diagram of the structure of expression inertia identification portion.
Fig. 5 is the block diagram of the structure of expression viscous friction Identification Division.
Fig. 6 is the fixedly block diagram of the structure of disturbance identification portion of expression.
Fig. 7 is the block diagram of the identifier structure of the expression motion control device of using the present invention's the 2nd method.
Fig. 8 is the block diagram of the structure of expression inertia viscous friction Identification Division.
Fig. 9 is the block diagram of the structure of the expression motion control device of using existing method.
Figure 10 is the figure of expression by the action of existing method qualification.
Symbol description
1 command generator
2 controllers
3 motor
4 machineries
5 detectors
6 identifiers
7 positioners
8 speed controls
9 current controllers
10 time diffusion devices
11 the 1st filters
12 the 2nd filters
13 the 3rd filters
14 JDC estimators
15 inertia identification devices
16 viscous friction identifiers
17 fixing disturbance identification devices
18 inertia viscous friction identifiers
41,42,43 time diffusion devices
44,45 dividers
46,47 time diffusion devices
48 dividers
49,4A, 4B time diffusion device
4C, 4D, 4E, 4F, 4G, 4H multiplier
4I, 4J, 4K subtracter
4L, the 4M divider
51 multipliers
52 subtracters
61,62 multipliers
63 subtracters
71 command generators
72 speed controls
73 estimators
74 identifiers
75 current controllers
100 motion control devices
Embodiment
Below with reference to the accompanying drawings the specific embodiment of the inventive method is described.
Fig. 1 is the block diagram of the structure of the expression motion control device of implementing the inventive method.Among the figure, the 1st, send the command generator of position command Xref.2 expression controllers, it controls computing and output current I according to position command, position probing value Xfb and speed detected value Vfb.The computing of controller 2 inside can be arbitrary form, in this specific embodiment, it is served as reasons according to the positioner 7 of position command Xref and position probing value Xfb output speed instruction Vref, constitutes according to the speed control 8 of speed command Vref and speed detected value Vfb output torque instruction value Tref and for the current controller of controlling according to the torque instruction value output current 9.Herein, the speed detected value also can be the value of the position detected value being carried out time diffusion, and when digital control, the time diffusion computing also can be used after the difference approximate differential divided by control cycle.The 3rd, motor, 4 mechanical bond is in motor 3.5 expressions detect the position of motor 3 and the detector of speed.The 6th, identifier, it calculates inertia identification value J_hat, the viscous friction coefficient identifier D_hat of controlling object and fixing disturbance identification value C_hat according to torque instruction Tref and speed detected value Vfb.
Fig. 3 is the block diagram of the identifier inter-process of expression 6.Among the figure, 10 express time differentiators carry out time diffusion to calculate acceleration detection value afb to speed detected value Vfb.Herein, when digital control, the time diffusion computing also can be used behind the signal differential of current and last time the approximate differential divided by the control cycle gained.
11,12,13 represent identical filter respectively.The filter of Shi Yonging can be the big filter more than 3 times of number of times of comparing denominator with the number of times of molecule herein, and for example the transmission characteristic of filter is so long as use the represented characteristic of transfer function Gfil by formula (1) to get final product.
Gfil＝r
3/(s+r)
3 (1)
S represents Laplacian herein.
In addition, the filter of use can be the filter of arbitrary form, even the series connection of the low-pass first order filter more than 3 use is also no problem fully.
14 expression JDC estimators, its input is carried out Ftref, Fvfb, Fafb after the Filtering Processing to torque instruction Tref, speed detected value Vfb and acceleration detection value afb, and calculates inertia identification value J_hat, the viscous friction coefficient identifier D_hat of controlling object and fixing disturbance identification value C_hat.Also have, 14 inside is made of 15 inertia identification device, 16 viscous friction identifier and three devices of fixedly disturbance identification device of 17.
Below, describe the inter-process of the fixedly disturbance identification device of 15 inertia identification device, 16 viscous friction identifier and 17 in detail.
(I) processing of inertia identification device inside
Fig. 4 is the block diagram of the processing of expression inertia identification device 15.Among the figure, 41,42,43,46,47 express time differentiators, 44,45,48 expression dividers.Inertia identification value J_hat carries out computing by following order.
At first, Ftref, Fvfb, Fafb are carried out time diffusion respectively to calculate Dtref, Dvfb, Dafb, next, calculate W1 divided by aforementioned Dvfb by making Dtref, next, by making Dafb calculate W2, next, W1 is carried out time diffusion to calculate DW1 divided by aforementioned Dvfb, next, W2 is carried out time diffusion to calculate DW2, next, is inertia identification value J_hat with the value that obtains divided by aforementioned DW2 with DW1.
When representing this processing with mathematical expression then suc as formula (2) to shown in the formula (9).
Dtref＝dFtref/dt (2)
Dvfb＝dFvfb/dt (3)
Dafb＝dFafb/dt (4)
W1＝Dtref/Dvfb (5)
W2＝Dafb/Dvfb (6)
DW1＝dW1/dt (7)
DW2＝dW2/dt (8)
J_hat＝DW1/DW2 (9)
Herein, when digital control, the time diffusion computing also can be used behind the signal differential of current and last time the approximate differential divided by control cycle.
(II) processing of viscous friction identifier inside
Fig. 5 is the block diagram of the processing of expression viscous friction identifier 16.Among the figure, 51 expression multipliers, 52 expression subtracters.Viscous friction identifier D_hat carries out computing by following order.
At first, calculating W3 by making W2 multiply by aforementioned inertia identification value J_hat, next, is viscous friction coefficient identifier D_hat to deduct the value that aforementioned W3 obtains with W1.When representing this processing with mathematical expression then suc as formula (10) to shown in the formula (11).
W3＝W2·J_hat (10)
D_hat＝W1-W3 (11)
(III) the fixedly processing of disturbance identification device inside
Fig. 6 is the fixedly block diagram of the processing of disturbance identification device 17 of expression.Among the figure, 61,62 expression multipliers, 1 variable of 63 expressions deducts the subtracter of 2 variablees.Fixedly disturbance identification device D_hat carries out computing by following order.
At first, calculate TF by making Fvfb multiply by aforementioned viscous friction identifier D_hat, next, calculate TA by making Fafb multiply by aforementioned inertia identification value J_hat, next, be the fixing identifier C_hat that disturbs to deduct the value that aforementioned TF and aforementioned TA obtain with Ftref.
When representing this processing with mathematical expression, then suc as formula (12) to shown in the formula (14).
TF＝Fvfb·D_hat (12)
TA＝Fafb·J_hat (13)
C_hat＝Ftref-TF-TA (14)
So, by in identifier 6, carrying out above processing, inertia identification value J_hat, viscous friction coefficient identifier D_hat that can the identification controlling object and fixing disturbance identification value C_hat.
In addition, needn't carry out computing to above processing all the time, only need one or its combination among monitor speed Vfb, acceleration afb in advance or the torque instruction value Tref, only reach pre-set threshold when above at these variablees, by implementing above-mentioned processing, just can improve identification precision.
In addition, the variable of monitoring also can be that Vfb, afb, Tref are carried out Fvfb, the Fafb after the Filtering Processing, or its combination among the Ftref.
In addition, the variable in the monitoring hour does not only carry out computing than threshold value, also keeps as the value of this identifier by the final value that will carry out after the computing, can only upgrade each identifier when computing.
In addition, the variable in the monitoring hour is not inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly whole computings of disturbance identification value C_hat of not carrying out controlling object than threshold value, can be a part of computing of only not carrying out wherein yet.
The application's invention also can use current-order to replace torque instruction, and as long as this moment is with torque constant (thrust constant), the unit of torque and electric current is carried out conversion get final product.
In addition, this sentences rotating machinery is that example is illustrated, and for the machinery that carries out translational motion, by same calculating, can calculate mass M rather than inertia J.
In addition, in this specific embodiment 1, though narrated the formation of carrying out Position Control, but as shown in Figure 2, even also can be when carrying out speed control by identical method, the inertia identification value J_hat of identification controlling object, viscous friction coefficient identifier D_hat and fixing disturbance identification value C_hat.At this moment, 1 command generator sends speed command Vref, and the computing of speed control and current controller is only carried out in 2 controller inside.
The deriving method of below representing the mode of embodiment 1.
Viscous friction (coefficient of friction D) and fixing when disturbing C to act on inertia J mechanical, torque instruction Tref that give this moment and the relation of speed Vfb are as the formula (15).
Tref＝J·afb+D·Vfb+C (15)
Both sides carry out obtaining formula (16) after the Filtering Processing
Ftref＝J·Fafb+D·Fvfb+C (16)
Time diffusion is carried out on both sides can the cancellation constant C after handling, and obtains formula (17).
Dtref＝J·Dafb+D·Dvfb (17)
Both sides obtain formula (18) after divided by Dvfb.
Dtref/Dvfb＝J·Dafb/Dvfb+D (18)
With W1 (=Dtref/Dvfb) and W2 (=Dafb/Dvfb) obtain formula (19) after rewriting.
W1＝J·W2+D (19)
Time diffusion is carried out on both sides can cancellation constant D after handling, and obtains formula (20).
DW1＝J·DW2 (20)
Therefore through type (21) can calculate J.
J＝DW1/DW2 (21)
Through type (19) and formula (21) can calculate D as (22).
D＝W1-J·W2 (22)
Through type (16) can calculate C as (23).
C＝Ftref-J·Fafb-D·Fvfb (23)
As mentioned above, do not use fully approximate, but the equation that can derive each value of identification.
Below the 2nd method of the present invention is described.
Fig. 1 is the block diagram of the structure of the expression motion control device of implementing the inventive method, owing in embodiment 1 this figure is illustrated, therefore omits its explanation.
Fig. 7 is the block diagram of expression identifier 6 inter-process, since different with Fig. 3, therefore make the following instructions.Among the figure, 10 express time differentiators, it carries out time diffusion to calculate acceleration detection value afb to speed detected value Vfb.Herein, when digital control, the time diffusion computing also can be used behind the signal differential of current and last time the approximate differential divided by control cycle.11,12,13 represent the 1st filter, the 2nd filter, the 3rd filter respectively.The filter of Shi Yonging can be the big filter more than 3 times of number of times of comparing denominator with the number of times of molecule herein, and for example the transmission characteristic of filter is so long as use the represented characteristic of transfer function Gfil by formula (24) to get final product.
Gfil＝r
3/(s+r)
3 (24)
S represents Laplacian herein.
In addition, the filter of use can be the filter of arbitrary form, even the series connection of the low-pass first order filter more than 3 use is also no problem fully.In addition, the series connection of the moving average filter more than 3 use is also no problem fully.At this moment, owing to the phase lag that does not cause, therefore can further improve computational accuracy because of Filtering Processing.
14 expression JDC estimators, its input is carried out Ftref, Fvfb, Fafb after the Filtering Processing to torque instruction Tref, speed detected value Vfb and acceleration detection value afb, and calculates inertia identification value J_hat, the viscous friction coefficient identifier D_hat of controlling object and fixing disturbance identification value C_hat.Also have 14 JDC estimator inside to constitute by 18 inertia viscous friction identifier and two devices of fixedly disturbance identification device of 16.
Below, will describe the inter-process of the fixedly disturbance identification device of 18 inertia viscous friction identifier and 16 in detail.
(IV) processing of inertia viscous friction identifier inside
Fig. 8 is the block diagram of the processing of expression inertia viscous friction identifier 15.Among the figure 41,42,43,49,4A, 4B express time differentiator, 4C, 4D, 4E, 4F, 4G, 4H represents multiplier.4I, 4J, 4K represents subtracter, 4L, 4M represents divider.Inertia identification value J_hat and viscous friction coefficient identifier D_hat carry out computing by following order.
At first, calculate Ftref, Fvfb, Fafb are carried out Dtref, Dvfb, Dafb behind the time diffusion respectively, next, carry out time diffusion once more to calculate D2tref, D2vfb, D2afb.Herein, when digital control, the time diffusion computing also can be used behind the signal differential of current and last time the approximate differential divided by control cycle.
In addition, use these variablees to carry out the computing of formula (25) and (26), to calculate inertia identification value J_hat and viscous friction coefficient identifier D_hat.
J_hat＝(D2tref·Dvfb-Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (25)
D_hat＝(D2afb·Dtref-Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (26)
Fig. 6 is the fixedly block diagram of the processing of disturbance identification device 17 of expression.Owing in embodiment 1, the fixing processing of disturbance identification device inside is illustrated, so omits its explanation.
In addition, needn't carry out computing to above processing all the time, only need one or its combination among monitor speed Vfb, acceleration afb in advance or the torque instruction value Tref, only reach pre-set threshold when above at these variablees, by implementing above-mentioned processing, just can improve identification precision.
In addition, the variable of monitoring also can be speed Vfb, acceleration afb, torque instruction Tref or these signals are carried out Fvfb, Fafb, Ftref after the Filtering Processing, perhaps these signals are carried out Dtref, Dvfb, Dafb, the D2tref behind the time diffusion, any one among D2vfb, the D2afb, perhaps its combination.
In addition, the variable in monitoring hour does not only carry out computing than threshold value, and final value that can also be by will carry out computing the time keeps as the value of this identifier, only upgrades each identifier when computing.
In addition, the variable in the monitoring hour is not inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly whole computings of disturbance identification value C_hat of not carrying out controlling object than threshold value, can be a part of computing of only not carrying out wherein yet.
The present invention also can use current-order to replace torque instruction, as long as utilize torque constant (thrust constant) this moment, the unit of torque and electric current is carried out conversion get final product.
In addition, this sentences rotating machinery is that example is illustrated, and for the machinery that carries out translational motion, by same calculating, can calculate mass M rather than inertia J.
In addition, in this specific embodiment 2, though narrated the formation of carrying out Position Control, but as shown in Figure 2, even also can be when carrying out speed control by identical method, the inertia identification value J_hat of identification controlling object, viscous friction coefficient identifier D_hat and fixing disturbance identification value C_hat.At this moment, 1 command generator sends speed command Vref, and the computing of speed control and current controller is only carried out in 2 controller inside.
The deriving method of below representing the mode of embodiment 2.
Viscous friction (coefficient of friction D) and fixing when disturbing C to act on inertia J mechanical, torque instruction Tref that give this moment and the relation of speed Vfb are as the formula (27).
Tref＝J·afb+D·Vfb+C (27)
Both sides carry out obtaining formula (28) after the Filtering Processing
Ftref＝J·Fafb+D·Fvfb+C (28)
Time diffusion is carried out on both sides can the cancellation constant C after handling, and obtains formula (29).
Dtref＝J·Dafb+D·Dvfb (29)
Both sides carry out obtaining formula (30) after time diffusion is handled once more.
D2tref＝J·D2afb+D·D2vfb (30)
Therefore, formula (29) and formula (30) being used as simultaneous equations calculates the estimated value of J and D and then obtains formula (31) and formula (32).
J＝(D2tref·Dvfb-Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (31)
D＝(D2afb·Dtref-Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (32)
Through type (27) can calculate C as (33).
C＝Ftref-J·Fafb-D·Fvfb (33)
As mentioned above, do not use fully approximate, but the equation that can derive each value of identification.
The variable of Filtering Processing has been carried out in utilization, after this, only handle and arithmetic by time diffusion, inertia identification value J_hat, viscous friction coefficient identifier D_hat that just can precision identification controlling object well and fixing disturbance identification value C_hat can be applied to the purposes of the System Discrimination of industrial robot, lathe etc.
Claims (15)
1. the system identifying method of a motion control device, the aforementioned movement control device has command generator, according to the instruction of aforementioned command generator output be used for controller, and inertia identification value J_hat, the viscous friction coefficient identifier D_hat of identification controlling object and the fixing identifier of disturbance identification value C_hat of the current value of drive motor from the detected value output that the detector that is additional on the motor obtains, it is characterized by
Handle in the following order:
Speed detected value Vfb is carried out time diffusion calculating acceleration detection value afb,
Calculate aforementioned acceleration detection value afb carried out signal Fafb after the Filtering Processing,
Calculate aforementioned speed detected value Vfb carried out signal Fvfb after the Filtering Processing,
Calculate aforementioned torque command value Tref carried out signal Ftref after the Filtering Processing,
With aforementioned Ftref, Fvfb and Fafb is that time diffusion and arithmetic are carried out in the basis,
Calculate inertia identification value J_hat, the viscous friction coefficient identifier D_hat of aforementioned controlling object and fixing disturbance identification value C_hat.
2. the system identifying method of motion control device as claimed in claim 1 is characterized by, and the processing of calculating aforementioned inertia identification value J_hat comprises:
Aforementioned Ftref, Fvfb, Fafb are carried out time diffusion respectively calculating Dtref, Dvfb, Dafb,
By making aforementioned Dtref calculate W1 divided by aforementioned Dvfb,
By making aforementioned Dafb calculate W2 divided by aforementioned Dvfb,
Aforementioned W1 is carried out time diffusion calculating DW1,
Aforementioned W2 is carried out time diffusion calculating DW2,
The value that obtains divided by aforementioned DW2 with aforementioned DW1 is inertia identification value J_hat.
3. the system identifying method of motion control device as claimed in claim 1 is characterized by, and the processing of calculating aforementioned viscous friction coefficient identifier D_hat comprises:
After calculating aforementioned inertia identification value J_hat, calculate W3 by making aforementioned W2 multiply by aforementioned inertia identification value J_hat,
Deducting the value that aforementioned W3 obtains with aforementioned W1 is viscous friction coefficient identifier D_hat.
4. the system identifying method of motion control device as claimed in claim 1 is characterized by, and the processing of calculating aforementioned fixation disturbance identification value C_hat comprises:
After calculating aforementioned inertia identification value J_hat and aforementioned viscous friction coefficient identifier D_hat, calculate TF by making aforementioned Fvfb multiply by aforementioned viscous friction identifier D_hat,
Calculate TA by making aforementioned Fafb multiply by aforementioned inertia identification value J_hat,
Deduct value that aforementioned TF and aforementioned TA obtain with aforementioned Ftref and be fixing disturbance identification value C_hat.
5. the system identifying method of motion control device as claimed in claim 1 is characterized by: the filter that is used for aforementioned Filtering Processing is for comparing the big at least filter more than 3 times of number of times of denominator with the number of times of molecule.
6. as the system identifying method of the described motion control device of claim 1 to 5, it is characterized by: monitor inner quantity of state in advance, only reach set point when above, inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly processing of disturbance identification value C_hat of calculating aforementioned controlling object in the aforesaid state amount; The aforesaid state amount is when set point is not above, and the identifier of J_hat, D_hat, C_hat is kept the value of last time respectively.
7. the system identifying method of motion control device as claimed in claim 6, it is characterized by: the aforementioned quantity of state of monitoring in advance is speed Vfb, acceleration afb, torque instruction Tref or these signals is carried out Fvfb, Fafb, Ftref after the Filtering Processing, among perhaps aforementioned Dvfb, the Dw2 any one, perhaps their combination.
8. motion control device, it has command generator, controls and export current controller, and inertia identification value J_hat, the viscous friction coefficient identifier D_hat of identification controlling object and the fixing identifier of disturbance identification value C_hat of the current value that is used for drive motor according to the instruction of aforementioned command generator output with from the detected value that the detector that is additional to motor obtains, and it is characterized by:
Aforementioned identifier constitutes and comprises:
Speed detected value Vfb is carried out time diffusion calculating the time diffusion device of acceleration detection value afb,
Aforementioned acceleration detection value afb is carried out Filtering Processing calculating the 1st filter of Fafb,
Aforementioned speed detected value Vfb is carried out Filtering Processing calculating the 2nd filter of Fvfb,
Aforementioned torque command value Tref is carried out Filtering Processing calculating the 3rd filter of Ftref,
With aforementioned Ftref, Fvfb and Fafb is that time diffusion and arithmetic are carried out with the inertia identification value J_hat, the viscous friction coefficient identifier D_hat that calculate aforementioned controlling object and the fixing JDC estimator of disturbance identification value C_hat in the basis.
9. motion control device as claimed in claim 8 is characterized by:
Aforementioned JDC estimator comprises:
Calculate the inertia estimator of inertia,
Calculate the viscous friction estimator of viscous friction,
Calculate the fixing fixedly interference estimator of disturbing.
10. the system identifying method of a motion control device, the aforementioned movement control device has the command generator of the instruction of generating, according to aforementioned instruction with generate current controller, and inertia identification value J_hat, the viscous friction coefficient identifier D_hat of identification controlling object and the fixing identifier of disturbance identification value C_hat of the electric current that is used for drive motor from the detected value that the detector that is additional to motor obtains, it is characterized by
Handle in the following order:
Speed detected value Vfb is carried out time diffusion calculating acceleration detection value afb,
Calculate aforementioned acceleration detection value afb carried out signal Fafb after the Filtering Processing,
Calculate aforementioned speed detected value Vfb carried out signal Fvfb after the Filtering Processing,
Calculate aforementioned torque command value Tref carried out signal Ftref after the Filtering Processing,
Calculate D2tref, D2vfb, D2afb after aforementioned Ftref, Fvfb, Fafb carried out Dtref, Dvfb, the Dafb behind the time diffusion respectively and carry out time diffusion once more,
By following equation,
J_hat＝(D2tref·Dvfb-Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (a)
D_hat＝(D2afb·Dtref-Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (b)
Calculate inertia identification value J_hat and viscous friction coefficient identifier D_hat.
11. the system identifying method of motion control device as claimed in claim 10 is characterized by: the filter that is used for aforementioned Filtering Processing is for comparing the big at least filter more than 3 times of number of times of denominator with the number of times of molecule.
12. the system identifying method of motion control device as claimed in claim 10 is characterized by: the device of filter that is used for aforementioned Filtering Processing for a plurality of moving average filters are connected in series.
13. system identifying method as the described motion control device of claim 10 to 12, it is characterized by: the signal of monitoring moving control device inside in advance, only reach set point when above, inertia identification value J_hat, viscous friction coefficient identifier D_hat and the fixedly processing of disturbance identification value C_hat of calculating aforementioned controlling object at aforementioned signal; The aforesaid state amount is when set point is not above, and the identifier of J_hat, D_hat, C_hat is kept the value of last time respectively.
14. the system identifying method of motion control device as claimed in claim 13, it is characterized by: the aforementioned signal of monitoring in advance is speed Vfb, acceleration afb, torque instruction Tref, perhaps these signals are carried out Fvfb, Fafb, Ftref after the Filtering Processing, perhaps these signals are carried out Dtref, Dvfb, Dafb, the D2tref behind the time diffusion, any one among D2vfb, the D2afb, perhaps their combination.
15. motion control device, it has the command generator of the instruction of generating, according to aforementioned instruction with generate current controller, and inertia identification value J_hat, the viscous friction coefficient identifier D_hat of identification controlling object and the fixing identifier of disturbance identification value C_hat of the electric current that is used for drive motor from the detected value that the detector that is additional to motor obtains, it is characterized by, have:
Speed detected value Vfb is carried out time diffusion to calculate the acceleration arithmetic unit of acceleration detection value afb;
Calculate the 1st filter that aforementioned acceleration detection value afb is carried out the signal Fafb after the Filtering Processing;
Calculate the 2nd filter that aforementioned speed detected value Vfb is carried out the signal Fvfb after the Filtering Processing;
Calculate the 3rd filter that aforementioned torque command value Tref is carried out the signal Ftref after the Filtering Processing;
And calculate D2tref, the D2vfb after aforementioned Ftref, Fvfb, Fafb carried out Dtref, Dvfb, the Dafb behind the time diffusion respectively and carry out time diffusion once more, the differentiator of D2afb,
By following equation,
J_hat＝(D2tref·Dvfb-Dtref·D2vfb)/
(Dvfb·D2afb-D2vfb·Dafb) (a)
D_hat＝(D2afb·Dtref-Dafb·D2tref)/
(Dvfb·D2afb-D2vfb·Dafb) (b)
Calculate inertia identification value J_hat and viscous friction coefficient identifier D_hat.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2004328574 | 2004-11-12 | ||
| JP328574/2004 | 2004-11-12 | ||
| JP019283/2005 | 2005-01-27 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101044673A true CN101044673A (en) | 2007-09-26 |
Family
ID=38809002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2005800357840A Pending CN101044673A (en) | 2004-11-12 | 2005-09-26 | Motion controller and system identifying method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101044673A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104423321A (en) * | 2013-08-26 | 2015-03-18 | 大隈株式会社 | Position control apparatus for identifying low-frequency disturbance |
| CN107872187A (en) * | 2016-09-26 | 2018-04-03 | 日本电产三协株式会社 | Control device of electric motor |
| CN111865171A (en) * | 2020-06-17 | 2020-10-30 | 苏州伟创电气科技股份有限公司 | Method and device for identifying device parameters |
-
2005
- 2005-09-26 CN CNA2005800357840A patent/CN101044673A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104423321A (en) * | 2013-08-26 | 2015-03-18 | 大隈株式会社 | Position control apparatus for identifying low-frequency disturbance |
| CN107872187A (en) * | 2016-09-26 | 2018-04-03 | 日本电产三协株式会社 | Control device of electric motor |
| CN107872187B (en) * | 2016-09-26 | 2020-08-18 | 日本电产三协株式会社 | Motor control device |
| CN111865171A (en) * | 2020-06-17 | 2020-10-30 | 苏州伟创电气科技股份有限公司 | Method and device for identifying device parameters |
| CN111865171B (en) * | 2020-06-17 | 2021-09-24 | 苏州伟创电气科技股份有限公司 | Method and device for identifying device parameters |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1181416C (en) | Pridiction controlling device | |
| CN1109390C (en) | Device and method for stabilizing electric power system | |
| CN1121293C (en) | Electric discharge machine and method of electric discharge machining | |
| CN101034865A (en) | Motor controller and motor control method | |
| CN1244196C (en) | Apparatus for controlling synchronous motor | |
| CN1969237A (en) | Method and device for adjusting and controlling manipulators | |
| CN1212701C (en) | Speed control device for AG electric motor | |
| CN1078420C (en) | Apparatus for converting frame format of television signal | |
| CN1264480C (en) | 3D back projection method and X-ray computer laminated imaging device | |
| CN1131640C (en) | Method for regenerating image point clock signal and apparatus thereby | |
| CN1762655A (en) | Method and apparatus for correcting thermal displacement of machine tool | |
| CN1067001C (en) | Robert controller | |
| CN1976212A (en) | Vector controller for a permanent magnet synchronous motor, inverter module, and permanent magnet synchronous motor constant display system | |
| CN1083279A (en) | Brshless DC motor and method for rotative driving thereof | |
| CN101074841A (en) | System and method for controlling air-separating tower dynamic matrix | |
| CN1661642A (en) | Target value processing unit, temperature controller, control process implementing system, process controlling method, target value processing program, and recording medium | |
| CN1399762A (en) | Robotic manipulators based on pre-defined time-optimum trajectory shape | |
| CN1725131A (en) | Three-parameter fastest self-anti-interference controller device and self-anti-interference control method | |
| CN1333487A (en) | Method and device for implementing optimized self anti-interference feedback control | |
| CN1716758A (en) | Control device and module for permanent magnet synchronous motor | |
| CN101062561A (en) | Method and device for adjusting the operating parameters of a robot, program and storage medium for this method | |
| CN1575765A (en) | X-ray ct imaging method and x-ray ct system | |
| CN1439338A (en) | Inverse projection method and computerized X-ray tomographic device | |
| CN1441434A (en) | Disc storage equipment and control method for disc storag equipment | |
| CN101044673A (en) | Motion controller and system identifying method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20070926 |