CN103684188A - Method and system for identifying rotational inertia of motion control system - Google Patents

Abstract

The invention discloses a method for identifying rotational inertia of a motion control system. The method comprises the following steps of building a reference model similar to an actual system; generating a speed command through a speed command generation unit; controlling a torque command to output through a controller according to the speed command, and enabling motor feedback speed to trace the speed command; taking the motor feedback speed as a speed command of the reference model, controlling a disturbance torque observed value and a torque command of the reference model to output, and enabling feedback speed of the reference model to trace the motor feedback speed; outputting the disturbance torque observed value and the torque command of the reference model to an identification unit; and identifying the rotational inertia of the system by the identification unit through MIT self-adaption rule according to the disturbance torque observed value and the torque command of the reference model. According to the method, the rotational inertia of the whole system can be identified with higher accuracy within shorter time without the specific speed command. The invention also discloses a system for identifying the rotational inertia of the motion control system.

Description

A kind of method for identification of rotational inertia of kinetic control system and system

Technical field

The present invention relates to motion control parameter self-tuning technical field, more particularly, relate to a kind of method for identification of rotational inertia and system of kinetic control system.

Background technology

The load rotating inertia of AC servo in running and the variation of perturbing torque easily cause not mating of controller parameter, thereby cause the deterioration of control performance, the stability margin of reduction system and bandwidth, restricted the raising of systematic function to a great extent.

Current existing method for identification of rotational inertia mainly contains off-line identification and the large class of on-line identification two.Off-line identification is by inputting specific speed preset signal (at least comprising an Acceleration and deceleration time section) at motor speed instruction end, gather torque and the speed responsive of motor reality simultaneously, then by methods such as cycle integrated or linear least squares, pick out the moment of inertia of motor according to torque instruction and acceleration feedback signal.

Off-line inertia identification need to pick out in advance moment of inertia in the out-of-work situation of kinetic control system, then the control parameter of servo-driver is set according to the inertia of identification.Off-line inertia identification method is applicable to the application scenario that system inertia is constant or variation is very little, yet in a lot of actual motion control system, as reel machine control system, whole moment of inertia can increase along with the increase of coil paper, if servo-driver parameter still remains unchanged, the response of system will reduce, and even causes system unstable.

Existing online inertia identification method is by analyzing the response of motor to outside disturbing signal, take observer or Kalman filtering, the methods such as Landau discrete-time parameter identification are estimated moment of inertia, because the attribute of moment of inertia must just can embody in acceleration and deceleration process, therefore Acceleration and deceleration time that need to be longer, can not accomplish real real-time identification.

Summary of the invention

In view of this, the invention provides a kind of method for identification of rotational inertia and system of kinetic control system, without under specific speed command, only need just can pick out to degree of precision the moment of inertia of whole system in the short period of time.

For solving the problems of the technologies described above, the technical solution used in the present invention is: a kind of method for identification of rotational inertia of kinetic control system, comprising:

Build the reference model approximate with real system;

Negotiation speed instruction generation unit formation speed instruction ω _r;

By controller according to described speed command ω _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _pfollow the tracks of described speed command ω _r;

By described motor feedback speed omega _pas the speed command of described reference model, the disturbing moment measured value of Reference model for control system

with reference model torque command T' _refexport, and make the feedback speed ω ' of described reference model _ptracking motor feedback speed ω _p;

By the disturbing moment measured value of described reference model

with reference model torque command T' _refexport identification unit to;

Described identification unit is according to the disturbing moment measured value of described reference model with reference model torque command T ' _ref, by MIT adaptive law, pick out the moment of inertia of system.

Preferably, described identification unit is according to the disturbing moment measured value of described reference model with reference model torque command T' _ref, the moment of inertia that picks out system by MIT adaptive law is specially:

By described reference model torque command T ' _refby high pass filter, carry out filtering and obtain HPFT' _ref;

By the disturbing moment measured value of described reference model

by high pass filter, carrying out filtering obtains

Will according to MIT adaptive law

after multiplying each other with inertia identification gain alpha again with HPFT' _refmultiply each other, then product is carried out to the moment of inertia that integration draws system

Preferably, described by controller according to described speed command ω _rcontrolling torque instruction T _refthe computing formula of output is:

$T_{\mathrm{ref}}=\frac{(K_{p}s+K_i)\&CenterDot;\mathrm{Js}}{\frac{J}{\hat{J}}\&CenterDot;{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r;$

Wherein: K _pfor the proportional gain of controller, K _ifor the storage gain of controller, s is Laplacian, the true moment of inertia that J is system,

for needing the moment of inertia of the system of identification.

Preferably, described in, make motor feedback speed omega _pfollow the tracks of described speed command ω _rcomputing formula be:

${\mathrm{\&omega;}}_p=\frac{K_{p}s+K_i}{\frac{J}{\hat{J}}\&CenterDot;{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r.$

Preferably, by described motor feedback speed omega _pas the speed command of described reference model, the disturbing moment measured value of Reference model for control system

with reference model torque command T' _refthe computing formula of output is:

${\hat{T}}_{\mathrm{dis}}=\frac{(K_{p}s+K_i)}{{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+K_{p}s+K_i}(1-\frac{J}{\hat{J}})s\&CenterDot;{\mathrm{\&omega;}}_p;$

$T_{\mathrm{ref}}^{\&prime;}=\frac{T_{\mathrm{ref}}}{J}+{\hat{T}}_{\mathrm{dis}}.$

Preferably, by described reference model torque command T' _refby high pass filter, carry out filtering and obtain HPFT' _refcomputing formula be:

${\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}=\frac{K_{p}s}{K_{p}s+K_i}\&CenterDot;(\frac{T_{\mathrm{ref}}}{\hat{J}}+{\hat{T}}_{\mathrm{dis}}).$

Preferably, by described reference model torque command

by high pass filter, carrying out filtering obtains

computing formula be:

$\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}=\frac{{\mathrm{<figure-callout\; id="2"\; label="K\; p\; s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">K</figure-callout>}}_p{s}^{}{}^2\&CenterDot;{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="p\; s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">p</figure-callout>}}}{s^{}}(1-\frac{J}{\hat{J}}).$

Preferably, according to MIT adaptive law, draw the moment of inertia of system

computing formula be:

$\hat{J}=\&Integral;\mathrm{\&alpha;}\frac{K_{p}s+K_i}{{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+K_{p}s+K_i}\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}.$

Preferably, according to MIT adaptive law, simplify the moment of inertia of the system that draws

computing formula be:

$\hat{J}=\&Integral;\mathrm{\&alpha;}*\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}.$

An identification of rotational inertia system for kinetic control system, comprising: velocity command, controller, reference model, identification unit and motor; Wherein:

Described velocity command is connected with described controller, for formation speed instruction ω _r, and by described speed command ω _rexport described controller to;

Described controller is according to the described speed command ω receiving _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _pfollow the tracks of described speed command ω _r;

Described reference model is connected with described controller, motor and identification unit respectively, by the feedback speed ω of motor _pas speed command, the disturbing moment measured value of output reference model

with reference model torque command T' _ref, and make the feedback speed ω ' of described reference model _ptracking motor feedback speed ω _p;

Described identification unit is connected with controller with described reference model respectively, by the disturbing moment measured value of the reference model of described reference model output

with reference model torque command T ' _ref, by MIT adaptive law, pick out the moment of inertia of system.

From above-mentioned technical scheme, can find out, the method for identification of rotational inertia of a kind of kinetic control system disclosed by the invention, in the process of identification, the instruction of Negotiation speed instruction generation unit formation speed, without the specific speed command of input, just can in the situation that not affecting normal production operation, realize online identification of rotational inertia; By building the reference model approximate with real system, the output or the state that compare reference model and actual controlled object, by MIT adaptive law conditioning controller parameter, it can making the dynamic characteristic of controlled object, and the dynamic characteristic of reference model is consistent or deviation is as far as possible little; By MIT adaptive law, regulate inertia identification to gain, can regulate according to demand the speed of inertia identification.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.

Fig. 1 is the flow chart of the method for identification of rotational inertia of a kind of kinetic control system disclosed by the invention;

Fig. 2 is the flow chart of the method for identification of rotational inertia of another kind of kinetic control system disclosed by the invention;

Fig. 3 is the block diagram of the identification of rotational inertia system of a kind of kinetic control system disclosed by the invention;

Fig. 4 is the controlling party block diagram of realizing identification of rotational inertia disclosed by the invention;

Fig. 5 is the simulation result figure of identification of rotational inertia disclosed by the invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only a part of embodiment of the present invention, rather than whole embodiment.Embodiment based in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.

The embodiment of the invention discloses a kind of method for identification of rotational inertia and system of kinetic control system, without under specific speed command, only need just can pick out to degree of precision the moment of inertia of whole system in the short period of time.

As shown in Figure 1, a kind of method for identification of rotational inertia of kinetic control system, comprising:

The reference model that S101, structure and real system are approximate;

S102, Negotiation speed instruction generation unit formation speed instruction ω _r;

S103, by controller according to speed command ω _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _ptracking velocity instruction ω _r;

S104, by motor feedback speed omega _pas the speed command with reference to model, the disturbing moment measured value of Reference model for control system

with reference model torque command T' _refexport, and make the feedback speed ω ' of reference model _ptracking motor feedback speed ω _p;

S105, with reference to the disturbing moment measured value of model

with reference model torque command T ' _refexport identification unit to;

S106, identification unit are according to the disturbing moment measured value of reference model with reference model torque command T ' _ref, by MIT adaptive law, pick out the moment of inertia of system.

The operation principle of the method for identification of rotational inertia of above-mentioned kinetic control system and process are: the present invention is online identification of rotational inertia, first need to build the reference model approximate with real system, then Negotiation speed instruction generation unit formation speed instruction ω _r, system controller is according to speed command ω _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _pfollow the tracks of described speed command ω _r, then by motor feedback speed omega _pas the speed command with reference to model, the disturbing moment measured value of Reference model for control system with reference model torque command T' _refexport, and make the feedback speed ω ' of reference model _ptracking motor feedback speed ω _p; Then with reference to the disturbing moment measured value of model

with reference model torque command T' _refexport identification unit to; Last identification unit is according to the disturbing moment measured value of reference model

with reference model torque command T' _ref, by MIT adaptive law, pick out the moment of inertia of system.

Above-described embodiment is in the process of identification, and the instruction of Negotiation speed instruction generation unit formation speed without the specific speed command of input, just can realize online identification of rotational inertia in the situation that not affecting normal production operation; By building the reference model approximate with real system, the output or the state that compare reference model and actual controlled object, by MIT adaptive law conditioning controller parameter, it can making the dynamic characteristic of controlled object, and the dynamic characteristic of reference model is consistent or deviation is as far as possible little; By MIT adaptive law, regulate inertia identification to gain, can regulate according to demand the speed of inertia identification.The present invention be take ideal model as with reference to control unit, the output or the state that compare reference model and actual controlled object, and go conditioning controller parameter by MIT adaptive law, it making every effort to make the dynamic characteristic of controlled object, and the dynamic characteristic of reference model is consistent or deviation is as far as possible little, although it is characterized in that having load torque in identification process acts in system, because the disturbing moment with reference to model has been passed through high pass filter, and the time constant of dividing with speed controlling portion of the coefficient setting of high pass filter is consistent, thereby effectively eliminate the impact of fixing interference or load torque, improved the accuracy of inertia identification.

As shown in Figure 2, the method for identification of rotational inertia for another kind of kinetic control system disclosed by the invention, comprising:

The reference model that S201, structure and real system are approximate;

S202, Negotiation speed instruction generation unit formation speed instruction ω _r;

S203, by controller according to speed command ω _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _ptracking velocity instruction ω _r;

S204, by motor feedback speed omega _pas the speed command with reference to model, the disturbing moment measured value of Reference model for control system with reference model torque command T' _refexport, and make the feedback speed ω ' of reference model _ptracking motor feedback speed ω _p;

S205, with reference to the disturbing moment measured value of model

with reference model torque command T' _refexport identification unit to;

S206, with reference to model torque command T' _refby high pass filter, carry out filtering and obtain HPFT ' _ref;

S207, with reference to the disturbing moment measured value of model

by high pass filter, carrying out filtering obtains

S208, according to MIT self adaptation regulation rate will

after multiplying each other with inertia identification gain alpha again with HPFT' _refmultiply each other, then product is carried out to the moment of inertia that integration draws system

Below in conjunction with Fig. 3 and Fig. 4, the operation principle of the method for identification of rotational inertia of the disclosed kinetic control system of the present embodiment and process are described in detail.

With reference to Fig. 3, inertia identification embodiment of the present invention comprises velocity command, controller, motor, reference model and identification unit.

Wherein, velocity command formation speed instruction ω _r, and by speed command ω _rexport controller to, in inertia identification process, this speed command ω _rneed to there is acceleration and deceleration process.

Controller is according to the speed command ω receiving _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _ptrace command speed omega _r, motor feedback speed omega wherein _paccording to encoder position feedback signal, calculate.

When controller adoption rate integral controller (PI controller), controller is according to speed command ω _rcontrolling torque instruction T _refthe computing formula of output is as follows:

$T_{\mathrm{ref}}=\frac{(K_{p}s+K_i)\&CenterDot;\mathrm{Js}}{\frac{J}{\hat{J}}\&CenterDot;s^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r;$

Wherein: K _pfor the proportional gain of controller, K _ifor the storage gain of controller, s is Laplacian, the true moment of inertia that J is system,

for needing the moment of inertia of the system of identification.

Motor feedback speed omega _ptracking velocity instruction ω _rcomputing formula as follows:

${\mathrm{\&omega;}}_p=\frac{K_{p}s+K_i}{\frac{J}{\hat{J}}\&CenterDot;s^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r.$

Above-mentioned reference model is the state observer model of actual controlled device, be used for monitor state variable, control object using electric current loop and servomotor as broad sense in the present embodiment, because electric current loop bandwidth is much larger than speed ring bandwidth, therefore can not consider electric current loop response process and delay, it is 1 model that electric current loop is approximately to gain, ignores friction simultaneously.

Specifically, as shown in Figure 4 by motor feedback speed omega _pas the speed command with reference to model, the disturbing moment measured value of Reference model for control system

with reference model torque command T' _refthe computing formula of output is as follows:

${\hat{T}}_{\mathrm{dis}}=\frac{(K_{p}s+K_i)}{s^2+K_{p}s+K_i}(1-\frac{J}{\hat{J}})s\&CenterDot;{\mathrm{\&omega;}}_p;$

$T_{\mathrm{ref}}^{\&prime;}=\frac{T_{\mathrm{ref}}}{J}+{\hat{T}}_{\mathrm{dis}};$

From above formula, in acceleration and deceleration process, the error of inertia can show with the form of disturbance.Meanwhile, in above-mentioned reference model, in order to simplify, do not express external loading moment, yet the relative identification time of conversion rate of common load is relatively slow, therefore in order to eliminate the disturbing moment measured value of load torque to reference model

impact, with reference to the disturbing moment measured value of model

with reference model torque command T' _refby the high pass filter in identification unit, wherein high pass filter function is:

$\mathrm{high}\_\mathrm{pass}\_\mathrm{filter}=\frac{K_{p}s}{K_{p}s+K_i}$

Disturbing moment measured value with reference to model

by high pass filter, can obtain:

$\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}=\frac{K_p{s}^2\&CenterDot;{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="p\; s"\; filenames="HDA0000451499920000041.png"\; state="\{\{state\}\}">p</figure-callout>}}}{s^{}}(1-\frac{J}{\hat{J}});$

Again with reference to the reference model torque command T' of model _refby high pass filter, can obtain:

${\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}=\mathrm{high}\_\mathrm{pass}\_\mathrm{filter}\&CenterDot;T_{\mathrm{ref}}^{\&prime;}=\frac{K_{p}s}{K_{p}s+K_i}\&CenterDot;T_{\mathrm{ref}}^{\&prime;}=\frac{K_{p}s}{K_{p}s+K_i}\&CenterDot;(\frac{T_{\mathrm{ref}}}{\hat{J}}+{\hat{T}}_{\mathrm{dis}})$

From above formula, in acceleration and deceleration region, can pick out the moment of inertia of system, and go conditioning controller parameter by MIT adaptive law.If desired the moment of inertia of the system of identification equate with the true moment of inertia J of system,

?

the reason that the reason producing or error produce is inaccurate causing because inertia arranges.

In the present invention, need the parameter of identification to be

the moment of inertia that needs the system of identification, the error of real system and reference model is simultaneously

known according to MIT adaptive law, will

multiply each other with the gain-α of inertia identification, the size of α need adjust according to the variation of gain, and to guarantee the stability of whole system, then the two product is multiplied by by this

obtaining inertia sensitivity is:

$\stackrel{\stackrel{\&CenterDot;}{^}}{J}=\frac{\&PartialD;\hat{J}}{\&PartialD;t}=-\mathrm{\&alpha;}\&CenterDot;\mathrm{\&epsiv;}\&CenterDot;\frac{\&PartialD;\mathrm{\&epsiv;}}{\&PartialD;\hat{J}}$

Wherein, α is the gain of inertia identification, determines the speed of identification.

Finally to inertia sensitivity

carry out integration and obtain the moment of inertia of system

for:

$\hat{J}=\&Integral;-\mathrm{\&alpha;\&epsiv;}\frac{\&PartialD;\mathrm{\&epsiv;}}{\&PartialD;\hat{J}}=\&Integral;\mathrm{\&alpha;\&epsiv;}\frac{K_{p}s+K_i}{s^2+K_{p}s+K_i}{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}=\&Integral;\mathrm{\&alpha;}\frac{K_{p}s+K_i}{s^2+K_{p}s+K_i}\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}$

In order to simplify calculating, above-mentioned expression formula is reduced to:

$\hat{J}=\&Integral;\mathrm{\&alpha;}*\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}$

Adopt MIT adaptive law, can be easily by regulating inertia gain coefficient α to regulate the speed of inertia identification, use the inventive method compared with prior art, can obtain quickly inertia result, according to the conversion of inertia, constantly revise the PI parameter of actual control system simultaneously, make torque output more steady, the speed responsive of real system more approaches ideal curve.

Fig. 5 is for being used identification of rotational inertia simulation result figure of the present invention.In figure, the whole load inertia of reality is set to 5 times of motor inertia, and mechanical system is carried out to modeling by single order inertia link, will fix to disturb being set as 0.005Nm, the moment of inertia of initial system when identification is started

be set to 0.5 times of motor inertia (being the true moment of inertia of system).As can be seen from the figure, in the first two acceleration and deceleration cycle after identification starts, the moment of inertia of system

converge to soon near the true inertia of whole system, then according to the moment of inertia of system

revise in real time PI parameter, can realize speed command tracking response very reposefully.

In sum, the present invention be take ideal model as with reference to control unit, the output or the state that compare reference model and actual controlled object, and go conditioning controller parameter by MIT adaptive law, it making every effort to make the dynamic characteristic of controlled object, and the dynamic characteristic of reference model is consistent or deviation is as far as possible little, although it is characterized in that having load torque in identification process acts in system, because the disturbing moment with reference to model has been passed through high pass filter, and the time constant of dividing with speed controlling portion of the coefficient setting of high pass filter is consistent, thereby effectively eliminate the impact of fixing interference or load torque, improved the accuracy of inertia identification.

Adopt in addition MIT adaptive law, can be easily by regulating inertia gain coefficient α to regulate the speed of inertia identification, use the inventive method compared with prior art, can obtain quickly inertia result, according to the conversion of inertia, constantly revise the PI parameter of actual control system simultaneously, make torque output more steady, the speed responsive of real system more approaches ideal curve.

Use in addition inertia identification method of the present invention, without the specific speed command of input, can in the situation that not affecting normal production operation, realize online inertia identification, and the identification time is short, only need duration acceleration and deceleration process seldom, from simulation result shown in Fig. 5, can find out, within two acceleration and deceleration stages, can pick out result, and identification precision is high, for load inertia, change situation greatly, there is good adaptability.

In this specification, each embodiment adopts the mode of going forward one by one to describe, and each embodiment stresses is the difference with other embodiment, between each embodiment identical similar part mutually referring to.

Above-mentioned explanation to the disclosed embodiments, makes professional and technical personnel in the field can realize or use the present invention.To the multiple modification of these embodiment, will be apparent for those skilled in the art, General Principle as defined herein can, in the situation that not departing from the spirit or scope of the present invention, realize in other embodiments.Therefore, the present invention will can not be restricted to these embodiment shown in this article, but will meet the widest scope consistent with principle disclosed herein and features of novelty.

Claims (10)

1. a method for identification of rotational inertia for kinetic control system, is characterized in that, comprising:

Build the reference model approximate with real system;

Negotiation speed instruction generation unit formation speed instruction ω _r;

By controller according to described speed command ω _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _pfollow the tracks of described speed command ω _r;

By described motor feedback speed omega _pas the speed command of described reference model, the disturbing moment measured value of Reference model for control system

with reference model torque command T ' _refexport, and make the feedback speed ω ' of described reference model _ptracking motor feedback speed ω _p;

By the disturbing moment measured value of described reference model

with reference model torque command T ' _refexport identification unit to;

Described identification unit is according to the disturbing moment measured value of described reference model

with reference model torque command T ' _ref, by MIT adaptive law, pick out the moment of inertia of system.

2. method according to claim 1, is characterized in that, described identification unit is according to the disturbing moment measured value of described reference model

with reference model torque command T' _ref, the moment of inertia that picks out system by MIT adaptive law is specially:

By described reference model torque command T' _refby high pass filter, carry out filtering and obtain HPFT ' _ref;

By the disturbing moment measured value of described reference model

by high pass filter, carrying out filtering obtains

Will according to MIT adaptive law

after multiplying each other with inertia identification gain alpha again with HPFT' _refmultiply each other, then product is carried out to the moment of inertia that integration draws system

3. method according to claim 2, is characterized in that, described by controller according to described speed command ω _rcontrolling torque instruction T _refthe computing formula of output is:

$T_{\mathrm{ref}}=\frac{(K_{p}s+K_i)\&CenterDot;\mathrm{Js}}{\frac{J}{\hat{J}}\&CenterDot;s^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r;$

Wherein: K _pfor the proportional gain of controller, K _ifor the storage gain of controller, s is Laplacian, the true moment of inertia that J is system,

for needing the moment of inertia of the system of identification.

4. method according to claim 3, is characterized in that, described in make motor feedback speed omega _pfollow the tracks of described speed command ω _rcomputing formula be:

${\mathrm{\&omega;}}_p=\frac{K_{p}s+K_i}{\frac{J}{\hat{J}}\&CenterDot;s^2+K_{p}s+K_i}\&CenterDot;{\mathrm{\&omega;}}_r.$

5. method according to claim 4, is characterized in that, by described motor feedback speed omega _pas the speed command of described reference model, the disturbing moment measured value of Reference model for control system

with reference model torque command T' _refthe computing formula of output is:

${\hat{T}}_{\mathrm{dis}}=\frac{(K_{p}s+K_i)}{s^2+K_{p}s+K_i}(1-\frac{J}{\hat{J}})s\&CenterDot;{\mathrm{\&omega;}}_p;$

$T_{\mathrm{ref}}^{\&prime;}=\frac{T_{\mathrm{ref}}}{J}+{\hat{T}}_{\mathrm{dis}}.$

6. method according to claim 5, is characterized in that, by described reference model torque command T' _refby high pass filter, carry out filtering and obtain HPFT' _refcomputing formula be:

${\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}=\frac{K_{p}s}{K_{p}s+K_i}\&CenterDot;(\frac{T_{\mathrm{ref}}}{\hat{J}}+{\hat{T}}_{\mathrm{dis}}).$

7. method according to claim 6, is characterized in that, by described reference model torque command

by high pass filter, carrying out filtering obtains computing formula be:

$\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}=\frac{K_p{s}^2\&CenterDot;{\mathrm{\&omega;}}_p}{s^2+K_{p}s+K_i}(1-\frac{J}{\hat{J}}).$

8. method according to claim 7, is characterized in that, draws the moment of inertia of system according to MIT adaptive law

computing formula be:

$\hat{J}=\&Integral;\mathrm{\&alpha;}\frac{K_{p}s+K_i}{s^2+K_{p}s+K_i}\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}.$

9. method according to claim 8, is characterized in that, simplifies the moment of inertia of the system that draws according to MIT adaptive law

computing formula be:

$\hat{J}=\&Integral;\mathrm{\&alpha;}*\mathrm{HPF}{\hat{T}}_{\mathrm{dis}}*{\mathrm{HPFT}}_{\mathrm{ref}}^{\&prime;}.$

10. an identification of rotational inertia system for kinetic control system, is characterized in that, comprising: velocity command, controller, reference model, identification unit and motor; Wherein:

Described velocity command is connected with described controller, for formation speed instruction ω _r, and by described speed command ω _rexport described controller to;

Described controller is according to the described speed command ω receiving _rcontrolling torque instruction T _refoutput, makes motor feedback speed omega _pfollow the tracks of described speed command ω _r;

Described reference model is connected with described controller, motor and identification unit respectively, by the feedback speed ω of motor _pas speed command, the disturbing moment measured value of output reference model

with reference model torque command T' _ref, and make the feedback speed ω ' of described reference model _ptracking motor feedback speed ω _p;

Described identification unit is connected with controller with described reference model respectively, by the disturbing moment measured value of the reference model of described reference model output with reference model torque command T ' _ref, by MIT adaptive law, pick out the moment of inertia of system.

Images