CN102354107A

CN102354107A - On-line identification and control method for parameter of alternating current position servo system model

Info

Publication number: CN102354107A
Application number: CN2011101427369A
Authority: CN
Inventors: 李向国; 刘向红; 梅志千
Original assignee: Changzhou Campus of Hohai University
Current assignee: Changzhou Gugao Intelligent Equipment Technology Research Institute Co ltd
Priority date: 2011-05-30
Filing date: 2011-05-30
Publication date: 2012-02-15
Anticipated expiration: 2031-05-30
Also published as: CN102354107B

Abstract

The invention discloses an on-line identification and control method for a parameter of an alternating current position servo system model. According to the invention, a model reference adaptive identification algorithm based on a Lyapunov stability theory is utilized; and on-line identification is carried out on a rotary inertia J and a viscous damping coefficient B of a controlled object of an alternating current servo system; after convergence is carried out on an identification parameter, an on-line design on a position controller is carried out according to J and B values and the design is automatically switched to position control. According to the invention, an efficiency on an alternating current servo system design can be substantially improved.

Description

On-line identification of a kind of AC position servo system model parameter and control method

Technical field

The present invention relates to permanent magnet synchronous motor (PMSM) AC servo system, on-line identification of particularly a kind of AC position servo system model parameter and control method belong to the synchronous motor technical field.

Background technology

After a new AC servo was set up, its many systematic parameter all was unknown, and the controller parameter of control system all to be model parameter according to system design.In addition in many control problems, towards object normally uncertain, the change of operating load can cause the variation of servo system models parameter.Therefore; A high performance AC servo not only requires system to make response fast and accurately to servo instruction; But also require when load characteristic changes; Can guarantee that still system has the better controlling performance, this just requires the controller parameter of servo-drive system to do suitable adjustment with load characteristic.

The method for designing of AC servo control system generally is divided into three kinds; First kind is that first off-line identification goes out the system model parameter; Again according to the model parameter CONTROLLER DESIGN parameter that picks out; This method efficient is lower; And off-line identification needs a large amount of experimental datas; To handle the parameter that just can obtain system to multi-group data; Amount of calculation is very big; And, the variation of load to carry out the offline parameter identification again when causing the system model parameter to change; Increased Control System Design time and workload, number of patent application is 200710024700.4 and is called that the Chinese patent of " observation of disturbing in a kind of AC position servo system and compensation method " introduced the influence to the controlled device control performance such as parameter variation.Second kind is to pick out the system model parameter earlier, again according to the online CONTROLLER DESIGN parameter of model parameter that picks out, and automatically switches to position control.The third is the position adaptive control algorithm; Different with second method is the modification of in the control system operational process, constantly carrying out parameter identification and controller; This kind method system design more complicated is difficult to realization, and higher to the hardware requirement of control system.

Summary of the invention

The technical matters that the present invention will solve is, to the defective on the prior art, on-line identification of a kind of AC position servo system model parameter and control method is provided, to improve the efficient of design control system.

For solving the problems of the technologies described above, the present invention provides on-line identification of a kind of AC position servo system model parameter and control method, it is characterized in that, may further comprise the steps:

1) Set the AC servo system controlled object model: the use of a first-order differential equation is expressed as

where ω is the system output speed, u is the input speed signal controlled object, J, and B is the AC servo system controlled object model parameters, are unknown parameters, namely the moment of inertia and viscous damping coefficient;

Set the reference model of controlled device identification:

J_{m} {\overset{\cdot}{ω}}_{m} = - B_{m} ω_{m} + ω_ref,

ω _ ref is the external reference signal of identification input, ω _mFor control system is hoped the control performance index that reaches, parameter J _mAnd B _mBeing respectively moment of inertia and viscous damping coefficient hopes the control performance index that reaches to be positive number in control system;

2) in plant model Model Reference Adaptive Control system; The model parameter moment of inertia J of identification and the identifier and the moment of inertia J and the viscous damping coefficient B in its last sampling period of viscous damping coefficient B are compared; Stop identification, the parameter value J and the B of output identification during less than given in advance performance index ε up to their difference;

3) according to step 2) the online calculating location controller of parameter value of controlled device identification of output, and automatically switch to position control.

On-line identification of aforesaid a kind of AC position servo system model parameter and control method is characterized in that: in said step 2) in, the tracing deviation between the reference model of controlled device identification and the realistic model parameter adopts scale operation, and control law is u=θr(t) ω _ ref-θy(t) ω, wherein θr(t) and θy(t) being is respectively the time change feedback gain of t identifier constantly and reference model parameter, and the closed-loop system function does<maths num=" 0002 "><[CDATA[<math><mrow><mover><mi>ω</mi><mo>·</mo></mover><mo>=</mo><mo>-</mo><mrow><mo>(</mo><mi>B</mi><mo>+</mo><msub><mi>θ</mi><mi>y</mi></msub><mo>)</mo></mrow><mi>ω</mi><mo>/</mo><mi>J</mi><mo>+</mo><msub><mi>θ</mi><mi>r</mi></msub><mi>ω</mi><mo>_</mo><mi>Ref</mi><mo>/</mo><mi>J</mi><mo>;</mo></mrow></math>]]></maths>Tracing deviation is e=ω-ωm, the dynamic expression formula of tracing deviation does<maths num=" 0003 "><[CDATA[<math><mrow><mover><mi>e</mi><mo>·</mo></mover><mo>=</mo><mover><mi>ω</mi><mo>·</mo></mover><mo>-</mo><msub><mover><mi>ω</mi><mo>·</mo></mover><mi>m</mi></msub><mo>=</mo><mo>-</mo><mfrac><msub><mi>B</mi><mi>m</mi></msub><msub><mi>J</mi><mi>m</mi></msub></mfrac><mi>e</mi><mo>+</mo><mrow><mo>(</mo><mfrac><msub><mi>B</mi><mi>m</mi></msub><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>-</mo><mfrac><msub><mi>θ</mi><mi>y</mi></msub><mi>J</mi></mfrac><mo>-</mo><mfrac><mi>B</mi><mi>J</mi></mfrac><mo>)</mo></mrow><mi>ω</mi><mo>+</mo><mrow><mo>(</mo><mfrac><msub><mi>θ</mi><mi>r</mi></msub><mi>J</mi></mfrac><mo>-</mo><mfrac><mn>1</mn><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mo>.</mo></mrow></math>]]></maths>

On-line identification of aforesaid a kind of AC position servo system model parameter and control method is characterized in that: in said step 2) in, utilize Liapunov's stability criterion to judge whether plant model Model Reference Adaptive Control system is stable, and concrete grammar is:

Liapunov Lyapunov function does

V (e, θ_{r}, θ_{y}) = \frac{1}{2} (e^{2} + \frac{J}{γ} {(\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}})}^{2} + \frac{J}{γ} {(\frac{θ_{r}}{J} - \frac{1}{J_{m}})}^{2}),

Wherein γ is an adaptive gain, as e=0, controller parameter θ _rAnd θ _yWhen equaling their optimal value, Lyapunov function V is zero, and the derivative of V does

\frac{dV}{dt} = e \frac{de}{dt} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) \frac{{dθ}_{y}}{dt} + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) \frac{{dθ}_{r}}{dt}

= - \frac{B_{m}}{J_{m}} e^{2} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) (\frac{{dθ}_{y}}{dt} - γωe) + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) (\frac{{dθ}_{r}}{dt} + γeω_ref)

If the controller parameters based on adaptive law <maths num="0007"> <! [CDATA [<math> <mrow> <mfrac> <msub> <mi> dθ </ mi> <mi> r </ mi> </ msub> <mi> dt </ mi> </ mfrac> <mo> = </ mo> <mo> - </ mo> <mi> γeω </ mi> <mo > _ </ mo> <mi> ref </ mi> <mo>, </ mo> </ mrow> </ math>]]> </maths> <maths num="0008"> <! [CDATA [<math> <mrow> <mfrac> <msub> <mi> dθ </ mi> <mi> y </ mi> </ msub> <mi> dt </ mi> </ mfrac> <mo> = </ mo> <mi> γωe </ mi> </ mrow> </ math>]]> </maths> update, then get <img file = " BDA0000064938950000041.GIF " he =" 120 " img-content =" drawing " img-format =" tif " inline =" yes " orientation =" portrait " wi =" 290 "/> If the error e is not equal to zero, the function V is decremented small, we can draw the error will tend to zero, determine the controlled object model reference adaptive control system is stable.The difference equation of self-adaptation rule does<maths num=" 0009 "><[CDATA[<math><mrow><mfrac><mrow><msub><mi>θ</mi><mi>r</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><msub><mi>θ</mi><mi>r</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths><maths num=" 0010 "><[CDATA[<math><mrow><mfrac><mrow><msub><mi>θ</mi><mi>y</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><msub><mi>θ</mi><mi>y</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths>Then the difference equation of moment of inertia and viscous damping coefficient does<maths num=" 0011 "><[CDATA[<math><mrow><mfrac><mrow><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>J</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths><maths num=" 0012 "><[CDATA[<math><mrow><mfrac><mrow><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>B</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths>TsBe the sampling period, θr(k), θy(k), e (k), ω (k), ω _ ref (k), J (k), B (k) and θr(k-1), θy(k-1), J (k-1), B (k-1) are respectively the value in k and k-1 sampling period.

On-line identification of aforesaid a kind of AC position servo system model parameter and control method is characterized in that: in said step 3), positioner adopts PD control, and transport function is expressed as G _c(s)=K _p+ K _dS, then the closed loop transfer function, of position ring does

Φ (s) = \frac{K_{p} + K_{d} s}{{Js}^{2} + (B + K_{d}) s + K_{p}},

Canonical form according to second-order system Obtain ω _n ²=K _p/ J, 2 ζ ω _n=(B+K _d)/J, so

K _d=2 ζ ω _n-B, wherein, K _P+ K _dS is positioner, wherein K _PBe scale-up factor, K _dBe differential coefficient, s is Laplce's variable; Make ω _nThe expectation free-running frequency of expression position ring, ζ representes the expectation damping ratio of position ring.

The invention has the beneficial effects as follows; Utilization is based on the model reference adaptive identification algorithm of Liapunov (Lyapunov) stability theory; Model parameter moment of inertia J and viscous damping coefficient B to the AC servo controlled device carry out on-line identification; After the identified parameters convergence; According to the online design attitude controller of the value of J and B; And automatically switch to position control, improve the efficient of system design greatly.

Description of drawings

Fig. 1 is of the present invention based on model parameter on-line identification and position control schematic diagram;

Fig. 2 is plant model on-line parameter identification and position control process flow diagram;

Fig. 3 is the block diagram of plant model Model Reference Adaptive Control system;

Fig. 4 is the model reference adaptive identification block diagram of first-order system;

Fig. 5 is the position closed loop theory diagram after simplifying;

Fig. 6 is a position control output tracking curve.

Embodiment

Below in conjunction with accompanying drawing and embodiment the present invention is further specified.

Among Fig. 1

Expression AC servo controlled device mathematical model, J and B are the model parameter moment of inertia and the viscous damping coefficients of AC servo controlled device, K _p+ K _dS is a positioner, and θ _ ref is the position control input reference signal of system, and ω _ ref is the external reference signal of identification input.Controlled object model with first-order differential equation is expressed as

where ω is the system output speed, u is the velocity loop input.

Among Fig. 2<maths num=" 0014 "><[CDATA[<math><mrow><mfrac><mrow><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>J</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow></mrow></math>]]></maths>With<maths num=" 0015 "><[CDATA[<math><mrow><mfrac><mrow><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>B</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow></mrow></math>]]></maths>Difference equation for plant model parameter moment of inertia and viscous damping coefficient Adaptive Identification; ε is for judging the convergent performance index; The identifier and the value in its last sampling period of moment of inertia J and viscous damping coefficient B are compared; Stop identification, the parameter value J and the B of output identification during less than given in advance performance index ε up to their difference; According to the online design attitude controller of the value of J and B, and automatically switch to position control.

Among Fig. 3

Be the reference model of controlled device identification, wherein J _mAnd B _mBe constant, ω _mHope the control performance index that reaches for control system; E representes the tracing deviation between each dynamic transient real process and the reference model; According to this difference; Constantly revise controller parameter; The control performance index that just can make real system is accomplished the plant model on-line parameter identification as far as possible near reference model.

In the Adaptive Identification system, supposing the system parameter J and B are unknown.The dynamic perfromance of desired system is made as the single order reference model

J_{m} {\overset{\cdot}{ω}}_{m} = - B_{m} ω_{m} + ω_ref,

ω _ ref is the external reference signal of input.Parameter J _mRequirement is strict positive, and reference model is stable like this.Be without loss of generality B _mAlso elect strict positive number as.Reference model can be expressed as ω with its transport function G _m=G ω _ ref, wherein

G = \frac{1}{J_{m} s + B_{m}} .

Tracing deviation between controlled device reference model and the realistic model parameter will carry out suitable computing, proportional control of the algorithm of comparative maturity and proportional plus integral control, and control law of the present invention adopts scale operation, therefore selects control law u=θr(t) ω _ ref-θy(t) ω, wherein θrAnd θyBecome feedback gain when being.Closed-loop system does<maths num="0018"><![CDATA[<math> <mrow> <mover> <mi>ω</mi> <mo>·</mo> </mover> <mo>=</mo> <mo>-</mo> <mrow> <mo>(</mo> <mi>B</mi> <mo>+</mo> <msub> <mi>θ</mi> <mi>y</mi> </msub> <mo>)</mo> </mrow> <mi>ω</mi> <mo>/</mo> <mi>J</mi> <mo>+</mo> <msub> <mi>θ</mi> <mi>r</mi> </msub> <mi>ω</mi> <mo>_</mo> <mi>ref</mi> <mo>/</mo> <mi>J</mi> <mo>,</mo> </mrow></math>]]></maths>Select u=θr(t) ω _ ref-θy(t) ω is a control law, and clearly: its system that makes can realize the accurate model coupling.In fact, if object parameters is known, allow formula<maths num=" 0019 "><[CDATA[<math><mrow><mover><mi>ω</mi><mo>·</mo></mover><mo>=</mo><mo>-</mo><mrow><mo>(</mo><mi>B</mi><mo>+</mo><msub><mi>θ</mi><mi>y</mi></msub><mo>)</mo></mrow><mi>ω</mi><mo>/</mo><mi>J</mi><mo>+</mo><msub><mi>θ</mi><mi>r</mi></msub><mi>ω</mi><mo>_</mo><mi>Ref</mi><mo>/</mo><mi>J</mi></mrow></math>]]></maths>And formula<maths num=" 0020 "><[CDATA[<math><mrow><msub><mover><mi>ω</mi><mo>·</mo></mover><mi>m</mi></msub><mo>=</mo><mfrac><mrow><mo>-</mo><msub><mi>B</mi><mi>m</mi></msub><msub><mi>ω</mi><mi>m</mi></msub><mo>+</mo><mi>ω</mi><mo>_</mo><mi>Ref</mi></mrow><msub><mi>J</mi><mi>m</mi></msub></mfrac></mrow></math>]]></maths>Equate:<maths num=" 0021 "><[CDATA[<math><mrow><msup><msub><mi>θ</mi><mi>r</mi></msub><mo>*</mo></msup><mo>=</mo><mfrac><mi>J</mi><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>,</mo></mrow></math>]]></maths><maths num=" 0022 "><[CDATA[<math><mrow><msup><msub><mi>θ</mi><mi>y</mi></msub><mo>*</mo></msup><mo>=</mo><mfrac><mrow><msub><mi>B</mi><mi>m</mi></msub><mi>J</mi></mrow><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>-</mo><mi>B</mi><mo>,</mo></mrow></math>]]></maths>If promptly select θr*, θy*Such controller parameter, then the dynamic perfromance of closed-loop system and reference model is identical, thus odd tracking error.

We select parameter θ now _rAnd θ _yThe self-adaptation rule.The note tracking error is e=ω-ω _m, the dynamic expression formula of tracking error does

\overset{\cdot}{e} = \overset{\cdot}{ω} - {\overset{\cdot}{ω}}_{m} = - \frac{B_{m}}{J_{m}} e + (\frac{B_{m}}{J_{m}} - \frac{θ_{y}}{J} - \frac{B}{J}) ω + (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) ω_ref,

If parameter θ _rAnd θ _yEqual their desired value θ _r ^*And θ _y ^*, then error e just goes to zero.Introduce the Lyapunov function

V (e, θ_{r}, θ_{y}) = \frac{1}{2} (e^{2} + \frac{J}{γ} {(\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}})}^{2} + \frac{J}{γ} {(\frac{θ_{r}}{J} - \frac{1}{J_{m}})}^{2}),

When e=0 and controller parameter equaled their optimal value, function V was zero.The derivative of V does

\frac{dV}{dt} = e \frac{de}{dt} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) \frac{{dθ}_{y}}{dt} + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) \frac{{dθ}_{r}}{dt}

= - \frac{B_{m}}{J_{m}} e^{2} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) (\frac{{dθ}_{y}}{dt} - γωe) + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) (\frac{{dθ}_{r}}{dt} + γeω_ref)

If the controller parameters based on adaptive law <maths num="0027"> <! [CDATA [<math> <mrow> <mfrac> <msub> <mi> dθ </ mi> <mi> r </ mi> </ msub> <mi> dt </ mi> </ mfrac> <mo> = </ mo> <mo> - </ mo> <mi> γeω </ mi> <mo > _ </ mo> <mi> ref </ mi> <mo>, </ mo> </ mrow> </ math>]]> </maths> <maths num="0028"> <! [CDATA [<math> <mrow> <mfrac> <msub> <mi> dθ </ mi> <mi> y </ mi> </ msub> <mi> dt </ mi> </ mfrac> <mo> = </ mo> <mi> γωe </ mi> </ mrow> </ math>]]> </maths> update, then get <img file = " BDA0000064938950000077.GIF " he =" 120 " img-content =" drawing " img-format =" tif " inline =" yes " orientation =" portrait " wi =" 290 "/> So as long as the error e is not zero, the function V on reduced, whereby the error can be derived will tend to zero, can be obtained according to the Lyapunov stability criterion of this system is stable.

Figure 4 is a first-order model reference adaptive system identification diagram, where <img file = "BDA0000064938950000078.GIF" he = "117" img-content = "drawing" img-format = "tif" inline = "yes" orientation = "portrait" wi = "351" /> is the reference model, <img file = "BDA0000064938950000079.GIF" he = "107" img-content = "drawing" img-format = " tif " inline =" yes " orientation =" portrait " wi =" 273 "/> is the controlled object model.Adjusting time tsBe the composite target of a system dynamic characteristic of reaction, in the present invention just according to the adjusting time t of systemsConfirm J in the reference modelmAnd BmValue.According to the computing formula of the time of adjusting, the adjusting time of choosing institute's design system in the present invention is 0.025 second, just can draw JmAnd BmValue be 0.00625 and 1, computation process repeats no more.Basis again<maths num=" 0029 "><[CDATA[<math><mrow><msup><msub><mi>θ</mi><mi>r</mi></msub><mo>*</mo></msup><mo>=</mo><mfrac><mi>J</mi><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>,</mo></mrow></math>]]></maths><maths num=" 0030 "><[CDATA[<math><mrow><msup><msub><mi>θ</mi><mi>y</mi></msub><mo>*</mo></msup><mo>=</mo><mfrac><mrow><msub><mi>B</mi><mi>m</mi></msub><mi>J</mi></mrow><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>-</mo><mi>B</mi><mo>,</mo></mrow></math>]]></maths>Can get J=θr*Jm,<maths num=" 0031 "><[CDATA[<math><mrow><mi>B</mi><mo>=</mo><mfrac><mrow><msub><mi>B</mi><mi>m</mi></msub><mi>J</mi></mrow><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>-</mo><msup><msub><mi>θ</mi><mi>y</mi></msub><mo>*</mo></msup><mo>.</mo></mrow></math>]]></maths>

Fig. 5 is the position closed loop theory diagram after simplifying, and positioner adopts PD control among the present invention, and its transport function is expressed as G _c(s)=K _p+ K _dS, then the closed loop transfer function, of position ring does

Φ (s) = \frac{K_{p} + K_{d} s}{{Js}^{2} + (B + K_{d}) s + K_{p}},

Make ω _nThe expectation free-running frequency of expression position ring, ζ representes the expectation damping ratio of position ring, according to the canonical form of second-order system

Can obtain ω _n ²=K _p/ J, 2 ζ ω _n=(B+K _d)/J, so K _d=2 ζ ω _n-B.

The present invention is according to the identifier of moment of inertia J and two model parameters of viscous damping coefficient B; Online design attitude controller; The identifier and the length of delay in its last sampling period that are about to model parameter compare; Stop identification during less than given in advance performance index up to their difference; The parameter value of output identification according to the online design attitude controller of identifier, and automatically switches to position control; The position reference of following the tracks of is a unit step signal, and final value is 1.

The AC permanent magnet synchronous motor model is MHMD042P1U in the experimental provision, the MATLAB2009a that Control Software adopts Mathworks company to produce.Selecting reference signal during test is ω _ ref=4sin100t.

Experimental result shows; The model reference adaptive parameter identification method can be correct the model parameter of identification control system; According to the online design attitude controller of the plant model parameter of identification, and automatically switch to position control, can improve the efficient of control system effectively.Below disclose the present invention with preferred embodiment, so it is not in order to restriction the present invention, and all employings are equal to replacement or the technical scheme that obtained of equivalent transformation mode, all drop within protection scope of the present invention.

Claims

1. AC position servo system model parameter on-line identification and control method is characterized in that, may further comprise the steps:

1) sets AC servo controlled device mathematical model: utilize differential equation of first order to be expressed as

J \overset{\cdot}{ω} = - Bω + u,

Wherein ω is system's output speed, and u is a controlled device input speed signal, and J and B are the model parameters of AC servo controlled device, are unknown parameters, is respectively moment of inertia and viscous damping coefficient;

Set the reference model of controlled device identification:

J_{m} {\overset{\cdot}{ω}}_{m} = - B_{m} ω_{m} + ω_ref,

3) according to step 2) export the online calculating location controller of parameter value of controlled device identification, and automatically switch to position control.

2. a kind of AC position servo system model parameter according to claim 1 on-line identification and control method; It is characterized in that: in said step 2) in; Tracing deviation between the reference model of controlled device identification and the realistic model parameter adopts scale operation, and control law is u=θr(t) ω _ ref-θy(t) ω, wherein θr(t) and θy(t) be respectively the time change feedback gain of t identifier constantly and reference model parameter, the closed-loop system function does<maths num=" 0003 "><[CDATA[<math><mrow><mover><mi>ω</mi><mo>·</mo></mover><mo>=</mo><mo>-</mo><mrow><mo>(</mo><mi>B</mi><mo>+</mo><msub><mi>θ</mi><mi>y</mi></msub><mo>)</mo></mrow><mi>ω</mi><mo>/</mo><mi>J</mi><mo>+</mo><msub><mi>θ</mi><mi>r</mi></msub><mi>ω</mi><mo>_</mo><mi>Ref</mi><mo>/</mo><mi>J</mi><mo>;</mo></mrow></math>]]></maths>Tracing deviation is e=ω-ωm, the dynamic expression formula of tracing deviation does<maths num=" 0004 "><[CDATA[<math><mrow><mover><mi>e</mi><mo>·</mo></mover><mo>=</mo><mover><mi>ω</mi><mo>·</mo></mover><mo>-</mo><msub><mover><mi>ω</mi><mo>·</mo></mover><mi>m</mi></msub><mo>=</mo><mo>-</mo><mfrac><msub><mi>B</mi><mi>m</mi></msub><msub><mi>J</mi><mi>m</mi></msub></mfrac><mi>e</mi><mo>+</mo><mrow><mo>(</mo><mfrac><msub><mi>B</mi><mi>m</mi></msub><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>-</mo><mfrac><msub><mi>θ</mi><mi>y</mi></msub><mi>J</mi></mfrac><mo>-</mo><mfrac><mi>B</mi><mi>J</mi></mfrac><mo>)</mo></mrow><mi>ω</mi><mo>+</mo><mrow><mo>(</mo><mfrac><msub><mi>θ</mi><mi>r</mi></msub><mi>J</mi></mfrac><mo>-</mo><mfrac><mn>1</mn><msub><mi>J</mi><mi>m</mi></msub></mfrac><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mo>.</mo></mrow></math>]]></maths>

3. a kind of AC position servo system model parameter according to claim 2 on-line identification and control method; It is characterized in that: in said step 2) in; Utilize Liapunov's stability criterion to judge whether plant model Model Reference Adaptive Control system is stable, and concrete grammar is:

Liapunov Lyapunov function does

V (e, θ_{r}, θ_{y}) = \frac{1}{2} (e^{2} + \frac{J}{γ} {(\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}})}^{2} + \frac{J}{γ} {(\frac{θ_{r}}{J} - \frac{1}{J_{m}})}^{2}),

\frac{dV}{dt} = e \frac{de}{dt} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) \frac{{dθ}_{y}}{dt} + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) \frac{{dθ}_{r}}{dt}

= - \frac{B_{m}}{J_{m}} e^{2} + \frac{1}{γ} (\frac{θ_{y}}{J} + \frac{B}{J} - \frac{B_{m}}{J_{m}}) (\frac{{dθ}_{y}}{dt} - γωe) + \frac{1}{γ} (\frac{θ_{r}}{J} - \frac{1}{J_{m}}) (\frac{{dθ}_{r}}{dt} + γeω_ref)

If controller parameter is according to the self-adaptation rule<maths num=" 0008 "><[CDATA[<math><mrow><mfrac><msub><mi>Dθ</mi><mi>r</mi></msub><mi>Dt</mi></mfrac><mo>=</mo><mo>-</mo><mi>γ Eω</mi><mo>_</mo><mi>Ref</mi><mo>,</mo></mrow></math>]]></maths><maths num=" 0009 "><[CDATA[<math><mrow><mfrac><msub><mi>Dθ</mi><mi>y</mi></msub><mi>Dt</mi></mfrac><mo>=</mo><mi>γ ω e</mi></mrow></math>]]></maths>Upgrade, then<img file="FDA0000064938940000028.GIF" he="119" id="ifm0010" img-content="drawing" img-format="GIF" inline="yes" orientation="portrait" wi="290"/>If error e is not equal to zero, function V just reduces, and can draw error will go to zero, and it is stable judging plant model Model Reference Adaptive Control system, and the difference equation of self-adaptation rule does<maths num=" 0010 "><[CDATA[<math><mrow><mfrac><mrow><msub><mi>θ</mi><mi>r</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><msub><mi>θ</mi><mi>r</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths><maths num=" 0011 "><[CDATA[<math><mrow><mfrac><mrow><msub><mi>θ</mi><mi>y</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><msub><mi>θ</mi><mi>y</mi></msub><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths>Then the difference equation of moment of inertia and viscous damping coefficient does<maths num=" 0012 "><[CDATA[<math><mrow><mfrac><mrow><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>J</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>J</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths><maths num=" 0013 "><[CDATA[<math><mrow><mfrac><mrow><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>B</mi><mrow><mo>(</mo><mi>k</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow><msub><mi>T</mi><mi>s</mi></msub></mfrac><mo>=</mo><mo>-</mo><msub><mi>B</mi><mi>m</mi></msub><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mo>_</mo><mi>Ref</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>-</mo><mi>γ e</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mi>ω</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>,</mo></mrow></math>]]></maths>T whereinsBe the sampling period, θr(k), θy(k), e (k), ω (k), ω _ ref (k), J (k), B (k) and θr(k-1), θy(k-1), J (k-1), B (k-1) are respectively the value in k and k-1 sampling period.

4. according to claim 1 or 2 or 3 on-line identification of described a kind of AC position servo system model parameter and control methods, it is characterized in that: in said step 3), positioner adopts PD control, and transport function is expressed as G _c(s)=K _p+ K _dS, then the closed loop transfer function, of position ring does

Φ (s) = \frac{K_{p} + K_{d} s}{{Js}^{2} + (B + K_{d}) s + K_{p}},

K _d=2 ζ ω _n-B, wherein, K _p+ K _dS is a positioner, K _pBe scale-up factor, K _dBe differential coefficient, s is Laplce's variable; Make ω _nThe expectation free-running frequency of expression position ring, ζ representes the expectation damping ratio of position ring.