CN104124908A - Inertia ratio on-line identifying system and method - Google Patents

Abstract

The invention provides an inertia ratio on-line identifying system and method. The inertia ratio on-line identifying system comprises a sampling unit, a first computing unit and a second computing unit, wherein the sampling unit is used for collecting rotation speeds and corresponding current of a plurality of servo motors in fixed sampling time when the servo motors are started or reversed; the first computing unit is used for calculating electromagnetic torque according to the current, and subjecting the collected rotation speeds and the calculated electromagnetic torque to iterative computation to obtain an identification vector; the second computing unit is used for calculating an inertia ratio according to the identification vector. According to the inertia ratio on-line identifying system and method, iterative computation of the identification vector from the sampled rotation speeds is performed, and the inertia ratio is calculated according to the identification vector, so that inertia ratio parameters of the servo motors can be obtained without off-line operation, and user's operation is simplified.

Description

The system and method for ONLINE RECOGNITION ratio of inertias

Technical field

The present invention relates to servomotor control field, more particularly, relate to a kind of system and method for ONLINE RECOGNITION ratio of inertias.

Background technology

In servo system, ratio of inertias is an important governing factor, is the basis of setting up loop model.Ratio of inertias is the ratio between load inertia and motor inertia, and according to ratio of inertias, whether the Acceleration and deceleration time that can estimate servo system can meet apparatus and process requirement.

The most frequently used ratio of inertias discrimination method is based on off-line identification at present, by acceleration and deceleration instruction, obtains its acceleration change amount, and then calculates ratio of inertias.The identification of above-mentioned off-line ratio of inertias no doubt can obtain servo ratio of inertias more accurately, but it cannot tackle the situation that outside inertia changes.If control parameter, do not follow outside ratio of inertias variation and change, often can cause servomotor control performance to decline, affecting result of use.

Frictional force is outer force-disturbance important in SERVO CONTROL, and its break during to servomotor inverted running has decisive influence, in order to do corresponding compensation, often by the mode of friction model, calculates the coulomb friction of zero crossing.Current friction model, when calculating coulomb friction, owing to being subject to the impact of friction model accuracy, is difficult to obtain correct result, and model as complicated in use, in the time of can causing practical application again, amount of calculation increases greatly, is difficult to real time execution and realizes.

Summary of the invention

The technical problem to be solved in the present invention is, for above-mentioned off-line ratio of inertias, identification cannot be tackled the problem that outside inertia changes, and a kind of system and method for ONLINE RECOGNITION ratio of inertias is provided.

The technical scheme that the present invention solves the problems of the technologies described above is, a kind of system of ONLINE RECOGNITION ratio of inertias is provided, comprise sampling unit, the first computing unit and the second computing unit, wherein: described sampling unit, be used for when servomotor starts or be reverse, with the sampling time of fixing, gather rotating speed and the corresponding electric current of a plurality of servomotors; Described the first computing unit, for according to described Current calculation electromagnetic torque, and with the rotating speed of described collection and calculate the electromagnetic torque iterative computation identification vector obtaining; Described the second computing unit, for according to described identification vector calculation ratio of inertias.

In the system of ONLINE RECOGNITION ratio of inertias of the present invention, the identification vector that described the first computing unit calculates is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ And vectorial by identification described in following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ ψ (k-1)=[y (k-1) T wherein _e(k-1)-1] ^t, and y is the servomotor rotating speed of sampling acquisition, T _efor electromagnetic torque, B is coefficient of friction, T _sfor the sampling time, T _lfor load torque, J be motor and load inertia and, β is recursion variable gain.

In the system of ONLINE RECOGNITION ratio of inertias of the present invention, described recursion variable gain $\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)},$ Wherein λ is greater than 0 constant.

In the system of ONLINE RECOGNITION ratio of inertias of the present invention, described the first computing unit comprises data transaction subelement and floating point arithmetic subelement, wherein: it is the power series that 16 systems represent that described data transaction subelement is used for the rotating speed of the servomotor of sampling and corresponding current conversion, and power series described in each are converted to self-defining data structure, described self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number; Described floating point arithmetic subelement is used described self-defining data structure to replace floating number the corresponding calculating formula of substitution to complete computing.

In the system of ONLINE RECOGNITION ratio of inertias of the present invention, described the second computing unit is also for according to described identification vector calculation load torque; Described system also comprises compensation computing unit, for carry out forward friciton compensation and reverse friction compensation according to described load torque.

The present invention also provides the method for the servo ratio of inertias of a kind of ONLINE RECOGNITION, comprises the following steps:

(a), when servomotor starts or be reverse, with the sampling time of fixing, gather a plurality of rotating speeds and the corresponding electric current of this servomotor;

(b) according to described Current calculation electromagnetic torque, and with the rotating speed of described collection and the electromagnetic torque iterative computation identification vector of calculating acquisition;

(c) according to described identification vector calculation ratio of inertias.

In the method for ONLINE RECOGNITION ratio of inertias of the present invention, in described step (b), described identification vector is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ And this identification is vectorial by following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ ψ (k-1)=[y (k-1) T wherein _e(k-1)-1] ^t, y is servomotor rotating speed, T _efor electromagnetic torque, B is coefficient of friction, T _sfor the sampling time, T _lfor load torque, J be motor and load inertia and, β is recursion variable gain.

In the method for ONLINE RECOGNITION ratio of inertias of the present invention, described recursion variable gain $\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)},$ Wherein λ is greater than 0 constant.

In the method for ONLINE RECOGNITION ratio of inertias of the present invention, described step (b) comprises the following steps:

(b1) by the rotating speed of the servomotor of sampling and corresponding current conversion, be the power series that 16 systems represent;

(b2) power series described in each are converted to self-defining data structure, described self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number;

(b3) use described self-defining data structure to replace floating number the corresponding calculating formula of substitution to complete computing.

In the method for ONLINE RECOGNITION ratio of inertias of the present invention, described step (c) comprising: according to described identification vector calculation load torque; Described step (c) comprises afterwards: according to described load torque, carry out forward friciton compensation and reverse friction compensation.

The system and method for ONLINE RECOGNITION ratio of inertias of the present invention, also obtains ratio of inertias by identification vector calculation by rotating speed, the electric current iterative computation identification vector of sampling, and can obtain the ratio of inertias parameter of servomotor without off-line operation, has simplified user's operation.And the present invention carries out friciton compensation by the ratio of inertias parameter obtaining, can reduce the pause of servomotor when reverse.

Accompanying drawing explanation

Fig. 1 is that servo-driver is controlled schematic diagram.

Fig. 2 is the schematic diagram of the system embodiment of ONLINE RECOGNITION ratio of inertias of the present invention.

Fig. 3 is the schematic diagram of friciton compensation.

Fig. 4 is the schematic diagram of another friciton compensation.

Fig. 5 is the schematic flow sheet of the embodiment of the method for ONLINE RECOGNITION ratio of inertias of the present invention.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

As shown in Figure 1, be the schematic diagram of the control of servo-driver.In this servo-driver, in order to simplify calculating, frictional force can be considered as to the linear function relevant with rotating speed, thereby the equation of motion of servo driver drives servomotor operation can use formula (1) description (ω ' be rotor acceleration):

$\mathrm{\ω}\text{'}=-\frac{B}{J}\mathrm{\ω}+\frac{1}{J}(T_e-T_l)---\left(1\right)$

Wherein, J is motor inertia and load inertia sum, T _efor electromagnetic torque, T _lfor load torque, B is coefficient of friction.A part using formula (1) as state equation (2), and utilize the described discretization method of formula (3) to calculate, finally can obtain the expressed discretization equation of calculating formula (4).

$\left\{\begin{array}{c}x\text{'}=\mathrm{Ax}+\mathrm{Bu}\\ y=\mathrm{Cx}\end{array}\right.---\left(2\right)$

$\left\{\begin{array}{c}\mathrm{\Φ}\left(t\right)=e^{\mathrm{At}}=L^{-1}{[\mathrm{sI}-A]}^{-1}\\ G\left(T\right)=\mathrm{\Φ}\left(T\right)=\mathrm{\Φ}\left(t\right){|}_{t=T}\\ H\left(T\right)={\∫}_0^T{e}^{\mathrm{At}}\mathrm{Bdt}\\ x\left[(k+1)T\right]=G\left(T\right)x\left(\mathrm{kT}\right)+H\left(T\right)u\left(\mathrm{kT}\right)\end{array}\right.---\left(3\right)$

$\left\{\begin{array}{c}\mathrm{\ω}\left(k\right)=e^{-\frac{\mathrm{BTs}}{J}}\mathrm{\ω}(k-1)+\frac{1-e^{-\frac{\mathrm{BTs}}{J}}}{B}{T}_e(k-1)+\frac{1-e^{-\frac{\mathrm{BTs}}{J}}}{B}{T}_l(k-1)(-1)\\ y\left(k\right)=\mathrm{\ω}\left(k\right)\end{array}\right.---\left(4\right)$

Order $b\left(k\right)=\left[\begin{array}{ccc}{e}^{-\frac{\mathrm{BTs}}{J}}& \frac{1-e^{-\frac{\mathrm{BTs}}{J}}}{B}& \frac{1-e^{-\frac{\mathrm{BTs}}{J}}}{B}{T}_l(k-1)\end{array}\right]$ For identification vector, be the amount of calculation in simplifying procedures, utilize Taylor series expansion can obtain identification vector:

$b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right]---\left(5\right)$

Make ψ (k-1)=[ω (k-1) T _e(k-1)-1] ^tfor known inputoutput data, ω (k-1) is the rotating speed (a upper sampling instant) of sampling, T _e(k-1) be the electromagnetic torque (a upper sampling instant) obtaining according to the Current calculation of sampling.Now can utilize the recurrence formula of calculating formula (6) to carry out identification:

$b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1)---\left(6\right)$

Like this, after calculating the value that obtains above-mentioned identification vector, can obtain a ternary linear function group in conjunction with calculating formula (5), and can obtain coefficient of friction B, motor and load inertia and J and load torque T by solving this ternary linear function group _l, further to obtain ratio of inertias.

As shown in Figure 2, be the schematic diagram of the system embodiment of ONLINE RECOGNITION ratio of inertias of the present invention.The system of the present embodiment comprises sampling unit 21, the first computing unit 22 and the second computing unit 23, above-mentioned sampling unit 21, the first computing unit 22 and the second computing unit 23 can be integrated into the digital signal processor of servo-driver, certainly also can adopt the one or more chips that are connected to above-mentioned digital signal processor to form.

Sampling unit 21 is for starting at servomotor or the (initial time at servomotor zero crossing as sampling oppositely time, so that the frictional force of calculating is Coulomb friction), with the sampling time Ts fixing, gather the rotating speed of a plurality of servomotors and the electric current of correspondence.The input of above-mentioned sampling unit 21 can be connected to the encoder on servomotor, and according to the output signal of encoder, calculates the rotating speed of servomotor.And the output that above-mentioned sampling unit 21 can directly pass through the AD converting unit of sampling servo-driver obtains current data.Especially, this sampling unit 21 can the sample rotating speed of 200 groups of servomotors and corresponding current data, with guarantee ratio of inertias identification accuracy (the data group of sampling is more, identify more accurate, but relative recognition speed can be slower).And can be 1 millisecond of left and right the interval time of the sampling of sampling unit 21 (being sampling time Ts).

The first computing unit 22 calculates electromagnetic torque (the corresponding electromagnetic torque of each electric current) according to the electric current of sampling and in conjunction with the moment coefficient of servomotor, and with the rotating speed that gathers and calculate the electromagnetic torque iterative computation identification vector obtaining.Particularly, this identification vector is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ T wherein _sfor the sampling time, T _lfor load torque, J be motor and load inertia and.

Particularly, above-mentioned identification vector can obtain by following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ And ψ (k-1)=[y (k-1) T _e(k-1)-1] ^t, the servomotor rotating speed that y (being y (k)) obtains for sampling, T _e(be T _e(k-1)) be electromagnetic torque, β is recursion variable gain.

Above-mentioned β value is if a constant basis, recursive process cannot correctly reflect the convergence process of recursion (being iterative computation) so, if input data (being rotating speed and electromagnetic torque) have sudden change, to there is saltus step in recursion result, for avoiding this situation, need to make β value along with the carrying out of recursion decays, its change procedure can represent by calculating formula (7) formula:

$\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)}---\left(7\right)$

Wherein λ is greater than 0 constant, and in actual applications, above-mentioned λ can directly be set to 1.

The second computing unit 23 is for according to identification vector calculation ratio of inertias, the value combination of calculating according to the first computing unit 22 the identification vector obtaining $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right]$ Obtain a ternary linear function group and obtain motor and load inertia and J, load torque T by solving this ternary linear function group _land coefficient of friction B, and then according to motor and load inertia and J and motor inertia calculation acquisition ratio of inertias.

The system of above-mentioned ONLINE RECOGNITION ratio of inertias, can obtain ratio of inertias parameter without independent off-line identification and control for follow-up servomotor, has simplified the operation of servo-driver.

The variable relating in computational process due to the identification vector at the first computing unit 22 is more, and computer capacity is indefinite, consider that the conventional fixed DSP of SERVO CONTROL realizes, and above-mentioned computational process cannot directly be calculated by calibrating, therefore can in fixed DSP, by floating-point arithmetic, complete above-mentioned computational process.In order to reduce floating number amount of calculation, sampled data can be regarded as is the power series that 16 systems represent, is defined as follows data structure:

And complete addition subtraction multiplication and division computing with above-mentioned data structure, finally calculate identification vector.The first computing unit 22 comprises data transaction subelement and floating point arithmetic subelement, wherein data transaction subelement is for being the power series that 16 systems represent by the rotating speed of the servomotor of sampling and corresponding current conversion, and each power series is converted to self-defining data structure, this self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number; Floating point arithmetic subelement is used self-defining data structure to replace floating number the corresponding calculating formula of substitution to complete computing.

In the system of above-mentioned ONLINE RECOGNITION ratio of inertias, also can comprise a compensation computing unit, this compensation computing unit is for calculating according to the second computing unit 23 the load torque T obtaining _l, and by this load torque T _las the frictional force of servomotor zero crossing, and then compensate.For simplified operation, compensation computing unit can be set forward friciton compensation value a and reverse friction offset-a according to command speed direction, as shown in Figure 3.

The prerequisite of above-mentioned compensation way is, the load torque that while supposing the positive and negative operation of servomotor, identification obtains equals frictional force, but this kind of situation while being practical application.Also having a kind of situation is when servomotor at right angle setting, in the load torque calculating, has comprised gravity according to the ratio of inertias of identification, has following expression way while moving up and down due to servomotor:

$\left\{\begin{array}{c}\mathrm{Te}-T_g-T_f={\mathrm{Ja}}^{+}\\ \mathrm{Te}+T_g-T_f={\mathrm{Ja}}^{-}\end{array}\right.---\left(8\right)$

T wherein _gfor suffered gravity, T _ffor friction, a ⁺for load climb acceleration, a ^-for load decline acceleration, the load torque therefore picking out can be expressed as:

$\left\{\begin{array}{c}{\mathrm{Tl}}^{+}=T_f+T_g\\ {\mathrm{Tl}}^{-}=T_f-T_g\end{array}\right.---\left(9\right)$

Tl wherein ⁺load torque while rising for load, Tl ^-load torque while declining for load.Now, forward friciton compensation a and reverse friction compensation-a need be combined with gravity compensation, its compensation way as shown in Figure 4.

As shown in Figure 5, be the schematic flow sheet of the embodiment of the method for the servo ratio of inertias of ONLINE RECOGNITION of the present invention.The method can directly be moved in servo-driver, and comprises the following steps:

Step S51: when servomotor starts or be reverse, gather a plurality of rotating speeds and the corresponding electric current of this servomotor with the sampling time Ts fixing.

In this step, it is the initial time as sampling at servomotor zero crossing, so that the frictional force of calculating is Coulomb friction.And the output signal that the rotating speed of servomotor can be arranged on the encoder on servomotor by sampling realizes, current signal directly the output of the AD sampling unit by sampling servo-driver obtain.Especially, in this step, can the sample rotating speed of 200 groups of servomotors and corresponding current data, to guarantee the accuracy of ratio of inertias identification, (the data group of sampling is more, identifies more accurate, but recognition speed can be slower relatively), and can be 1 millisecond of left and right the interval time of sampling.

Step S52: according to described Current calculation electromagnetic torque, and with the rotating speed of described collection and the electromagnetic torque iterative computation identification vector of calculating acquisition.

Step S53: according to identification vector calculation ratio of inertias, load torque and coefficient of friction.

In this step, above-mentioned identification vector is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ And this identification is vectorial by following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ ψ (k-1)=[y (k-1) T wherein _e(k-1)-1] ^t, y is servomotor rotating speed, T _efor electromagnetic torque, B is coefficient of friction, T _sfor the sampling time, T _lfor load torque, J be motor and load inertia and, β is recursion variable gain.And, for avoiding the sampled data of input to have sudden change to make recursion result produce saltus step, can make above-mentioned recursion variable gain $\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)},$ Wherein λ is greater than 0 constant.

And, for guaranteeing the operation efficiency of fixed DSP in servo-driver, can be the power series that 16 systems represent by the rotating speed of the servomotor of sampling and corresponding current conversion, and each power series is converted to self-defining data structure, and use the corresponding calculating formula of above-mentioned self-defining data structure substitution to complete computing.Above-mentioned self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number.

In addition after step S53, also can comprise: according to the load torque picking out, carry out forward friciton compensation and reverse friction compensation.

The above; be only the present invention's embodiment preferably, but protection scope of the present invention is not limited to this, is anyly familiar with in technical scope that those skilled in the art disclose in the present invention; the variation that can expect easily or replacement, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claim.

Claims (10)

1. the system of an ONLINE RECOGNITION ratio of inertias, it is characterized in that: comprise sampling unit, the first computing unit and the second computing unit, wherein: described sampling unit, be used for when servomotor starts or be reverse, with the sampling time of fixing, gather rotating speed and the corresponding electric current of a plurality of servomotors; Described the first computing unit, for according to described Current calculation electromagnetic torque, and with the rotating speed of described collection and calculate the electromagnetic torque iterative computation identification vector obtaining; Described the second computing unit, for according to described identification vector calculation ratio of inertias.

2. the system of ONLINE RECOGNITION ratio of inertias according to claim 1, is characterized in that: the identification vector that described the first computing unit calculates is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ And vectorial by identification described in following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ ψ (k-1)=[y (k-1) T wherein _e(k-1)-1] ^t, and y is the servomotor rotating speed of sampling acquisition, T _efor electromagnetic torque, B is coefficient of friction, T _sfor the sampling time, T _lfor load torque, J be motor and load inertia and, β is recursion variable gain.

3. the system of ONLINE RECOGNITION ratio of inertias according to claim 2, is characterized in that: described recursion variable gain $\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)},$ Wherein λ is greater than 0 constant.

4. according to the system of the ONLINE RECOGNITION ratio of inertias described in claim 2 or 3, it is characterized in that: described the first computing unit comprises data transaction subelement and floating point arithmetic subelement, wherein: it is the power series that 16 systems represent that described data transaction subelement is used for the rotating speed of the servomotor of sampling and corresponding current conversion, and power series described in each are converted to self-defining data structure, described self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number; Described floating point arithmetic subelement is used described self-defining data structure to replace floating number the corresponding calculating formula of substitution to complete computing.

5. the system of ONLINE RECOGNITION ratio of inertias according to claim 1, is characterized in that: described the second computing unit is also for according to described identification vector calculation load torque; Described system also comprises compensation computing unit, for carry out forward friciton compensation and reverse friction compensation according to described load torque.

6. a method for the servo ratio of inertias of ONLINE RECOGNITION, is characterized in that: comprise the following steps:

(a), when servomotor starts or be reverse, with the sampling time of fixing, gather a plurality of rotating speeds and the corresponding electric current of this servomotor;

(b) according to described Current calculation electromagnetic torque, and with the rotating speed of described collection and the electromagnetic torque iterative computation identification vector of calculating acquisition;

(c) according to described identification vector calculation ratio of inertias.

7. the method for ONLINE RECOGNITION ratio of inertias according to claim 6, is characterized in that: in described step (b), described identification vector is: $b\left(k\right)\≈\left[\begin{array}{ccc}1-\frac{\mathrm{BTs}}{J}& \frac{\mathrm{Ts}}{J}& \frac{\mathrm{Ts}}{J}{T}_l(k-1)\end{array}\right],$ And this identification is vectorial by following calculating formula iterative computation: $b\left(k\right)=\frac{\mathrm{\β\ψ}(k-1)[y\left(k\right)-{\mathrm{\ψ}}^T(k-1)b(k-1)]}{1+{\mathrm{\ψ}}^T(k-1)\mathrm{\β\ψ}(k-1)}+b(k-1),$ ψ (k-1)=[y (k-1) T wherein _e(k-1)-1] ^t, and y is servomotor rotating speed, T _efor electromagnetic torque, B is coefficient of friction, T _sfor the sampling time, T _lfor load torque, J be motor and load inertia and, β is recursion variable gain.

8. the method for ONLINE RECOGNITION ratio of inertias according to claim 7, is characterized in that: described recursion variable gain $\mathrm{\β}\left(k\right)=\mathrm{\β}(k-1)-\frac{\mathrm{\β}(k-1)\mathrm{\ψ}(k-1){\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)}{\mathrm{\λ}+{\mathrm{\ψ}}^T(k-1)\mathrm{\β}(k-1)\mathrm{\ψ}(k-1)},$ Wherein λ is greater than 0 constant.

9. according to the method for the ONLINE RECOGNITION ratio of inertias described in claim 7 or 8, it is characterized in that: described step (b) comprises the following steps:

(b1) by the rotating speed of the servomotor of sampling and corresponding current conversion, be the power series that 16 systems represent;

(b2) power series described in each are converted to self-defining data structure, described self-defining data structure comprises that floating number splits rear character array, floating number significance bit length, floating number power series, positive and negative four defined variables of floating number;

(b3) use described self-defining data structure to replace floating number the corresponding calculating formula of substitution to complete computing.

10. the method for ONLINE RECOGNITION ratio of inertias according to claim 6, is characterized in that: described step (c) comprising: according to described identification vector calculation load torque; Described step (c) comprises afterwards: according to described load torque, carry out forward friciton compensation and reverse friction compensation.