CN102269971A - Self-adapting servo controller based on model tracking - Google Patents

Abstract

The invention discloses a self-adapting servo controller based on model tracking, belonging to the technical field of motor control. The controller provided by the invention realizes the self-adapting control of a servo system based on real-time tracking of an ideal model; the system is provided with a reference model with ideal dynamic and static characters and an adjustable model which comprises a speed ring and a current ring of an actual alternating-current servo motor control system; and the two models have output qualities with same physical meanings. When the two models simultaneously work, the parameters of the adjustable model can be adjusted in real time by utilizing error amount of the output qualities according to a suitable self-adapting rate; the parameters of the adjustable model are dynamically adjusted so that the error amount of the output of the two models is close to zero; meanwhile, the process enables the adjustable model, which comprises the speed ring and the current ring of the actual alternating-current servo motor control system, to be close to an ideal reference model, so that the aim of enabling the actual servo motor control system to have the ideal dynamic and static characters can be achieved.

Description

Adaptive servo controller based on model following

Technical field

The present invention relates to a kind of adaptive servo controller, belong to the electric machines control technology field based on model following.

Background technology

Servo-drive system is widely used in industry manufacturing and the defence equipment.Development along with technology, each application is had higher requirement to the control performance of servo-drive system, application scenarios such as particularly high-end numerically-controlled machine driving, robot driving, national defence and space equipment control, the operating mode complexity, vary within wide limits such as inertia, load, adopt traditional PID regulator difficulty of adjusting, and be difficult to take into account the different loads situation, when bigger variations takes place systematic parameter, can cause the reduction of control system performance.For this reason, this patent proposes to adopt the regulator parameter automatic adjusting of the method for ideal model dynamic tracing being realized servo-drive system, satisfies above-mentioned high-end servo performance requirement.Be applied to speed observation in the electric machine control system in the past based on the adaptive approach of reference model, obtained effect necessarily.But in servo regulator control, also there is not actual application.Adopt this method can obviously improve the adaptability of servo-drive system to the disturbance of inertia, load and systematic parameter.

Summary of the invention

The object of the present invention is to provide a kind of adaptive servo controller based on model following, sort controller is based on the adaptive control that the real-time tracing of ideal model is realized servo-drive system, reference model and an adjustable model that comprises actual AC servo motor control system speed ring and electric current loop with desirable static and dynamic performance is set in the system, and two models have the output quantity of same physical meaning.Two models are worked simultaneously, utilize its output quantity between the margin of error, come the parameter of real-time regulated adjustable model according to suitable adaptive rate, dynamically adjust the output error amount that makes two models by the parameter of adjustable model and level off to zero, this process makes the adjustable model that comprises actual servo system speed ring and electric current loop level off to the desired reference model simultaneously, makes the actual servo system have the purpose of desirable static and dynamic performance thereby reach.

Servo system self-adaptive controller architecture disclosed in this invention as shown in Figure 1, r (t) is the instruction of servo-drive system speed ring, G (s) is the desired reference model, adopting the servo control system of PWM inverter is the actual servo control gear, it is the part of adjustable model, D (s) is for guaranteeing a linearity compensator of system stability, the instruction of q shaft current, y _mBe the speed output of desired reference model, y _pPromptly comprise the speed output of the actual servo electric system of PWM inverter control for adjustable model, υ (t) is y _mAnd y _pError through the value of linear compensation gained.The self-adaptation adjustment process of servo-drive system is presented as a series of from setting parameter K that adaptation mechanism is realized in the accompanying drawing 1 _A0, K _A1, K _rAnd K _ErAdaptive rate calculate process, and the q shaft current command value i that draws by the last computing of these parameters _QrefComputation process, its specifically adjust and computation process be described below:

1) from setting parameter K _A0, K _A1, K _rAnd K _ErAdaptive rate be respectively:

By y _mAnd υ (t) calculates K through following formula _A0,,

$K_{a0}={\∫}_0^t{l}_{1}v\left(\mathrm{\τ}\right)y_m\left(\mathrm{\τ}\right)\mathrm{d\τ}+l_{2}v\left(t\right)y_m\left(t\right)+K_{a00}$

By y _mAnd υ (t) calculates K through following formula _A1,

$K_{a1}={\∫}_0^t{l}_3\mathrm{\υ}\left(\mathrm{\τ}\right)d{y}_m\left(\mathrm{\τ}\right)/d\left(t\right)\mathrm{d\τ}+l_4\mathrm{\υ}\left(t\right)d{y}_m\left(t\right)/d\left(t\right)+K_{a10}$

Calculate K by υ (t) and r (t) through following formula _r,

$K_r={\&Integral;}_0^t{l}_5\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)r\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_6\mathrm{\&upsi;}\left(t\right)r\left(t\right)+{\mathrm{<figure-callout\; id="0"\; label="K\; r"\; filenames="HSA00000127377400012.png"\; state="\{\{state\}\}">K</figure-callout>}}_r$

Calculate K by υ (t) and e (t) through following formula _Er,

$K_{\mathrm{er}}={\&Integral;}_0^t{l}_7\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)e\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_8\mathrm{\&upsi;}\left(t\right)e\left(t\right)+K_{\mathrm{er}0}$

L wherein ₁, l ₃, l ₅And l ₇Be respectively K _A0, K _A1, K _rAnd K _ErIntegration constant, l ₂, l ₄, l ₆And l ₈Be respectively K _A0, K _A1, K _rAnd K _ErProportionality constant.

2) by from setting parameter K _A0, K _A1, K _rAnd K _ErDerivation actual servo motor q shaft current command value i _QrefThe reckoning process, step is as follows:

By K _A0, K _A1, y _mCalculate U through following formula _P1,

U _p1＝K _a0·y _m(t)+K _a1·dy _m(t)/d(t)

By K _r, r (t) calculates U through following formula _P2,

U _p2＝K _r·r(t)

By K _Er, e (t) calculates U through following formula _P3,

U _p3＝K _er·e(t)

By U _P1, U _P2, U _P3Calculate i through following formula _Qref

i _qref＝U _p1+U _p2+U _p3

The present invention compares with existing traditional PI D servo control technique, have the following advantages and effect: it is better to the adaptability of disturbances such as system inertia variation, load variations and parameter of electric machine variation that model is followed the trail of adaptive control, can keep consistent rotating speed steady state controling precision, the overshoot of rotating speed response is less simultaneously, under perturbation action, can dynamically adjust the servo controller parameter, reach good dynamic characteristics by adaptive rate.

Embodiment

Below in conjunction with accompanying drawing principle of the present invention, embodiment are further described.

1) calculation procedure is as follows as shown in Figure 1 from the adaptive rate of setting parameter, is respectively:

By y _mAnd υ (t) calculates K through following formula _A0,,

$K_{a0}={\&Integral;}_0^t{l}_{1}v\left(\mathrm{\&tau;}\right)y_m\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_{2}v\left(t\right)y_m\left(t\right)+K_{a00}$

By y _mAnd υ (t) calculates K through following formula _A1,

$K_{a1}={\&Integral;}_0^t{l}_3\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)d{y}_m\left(\mathrm{\&tau;}\right)/d\left(t\right)\mathrm{d\&tau;}+l_4\mathrm{\&upsi;}\left(t\right)d{y}_m\left(t\right)/d\left(t\right)+K_{a10}$

Calculate K by υ (t) and r (t) through following formula _r,

$K_r={\&Integral;}_0^t{l}_5\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)r\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_6\mathrm{\&upsi;}\left(t\right)r\left(t\right)+{\mathrm{<figure-callout\; id="0"\; label="K\; r"\; filenames="HSA00000127377400012.png"\; state="\{\{state\}\}">K</figure-callout>}}_r$

Calculate K by υ (t) and e (t) through following formula _Er,

$K_{\mathrm{er}}={\&Integral;}_0^t{l}_7\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)e\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_8\mathrm{\&upsi;}\left(t\right)e\left(t\right)+K_{\mathrm{er}0}$

2) by from setting parameter K _A0, K _A1, K _rAnd K _ErDerivation actual servo motor q shaft current command value i _QrefThe reckoning process, step is as follows:

By K _A0, K _A1, y _mCalculate U through following formula _P1,

U _p1＝K _a0·y _m(t)+K _a1·dy _m(t)/d(t)

By K _r, r (t) calculates U through following formula _P2,

U _p2＝K _r·r(t)

By K _Er, e (t) calculates U through following formula _P3,

U _p3＝K _er·e(t)

By U _P1, U _P2, U _P3Calculate i through following formula _Qref

i _qref＝U _p1+U _p2+U _p3

3) simulation example of model method for tracing

Below the model method for tracing is carried out emulation.Under environment, build realistic model based on Matlab/Simulink.Given speed command is the square-wave signal of positive and negative 100rpm, and the cycle is 0.4s.Accompanying drawing 2 is the second order reference model of setting dynamic response processes to this square wave instruction, and accompanying drawing 3 is servo-drive system moment of inertia wide variation, the rotating speed response when system's moment of inertia is 1J, 2J, 4J, 6J, and wherein J is rotor self inertia.

From simulated effect, it is better to the adaptability of system inertia that model is followed the trail of adaptive control, can keep consistent rotating speed steady state controling precision, and the overshoot of rotating speed response is less simultaneously, and dynamic property is better.

4) experimental verification of model method for tracing

To the checking that experimentizes of model method for tracing, in the experiment with typical second-order system be example as the desired reference model, be taken as:

$G\left(s\right)=\frac{1600}{{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HSA00000127377400033.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+68s+1600}$

Linearity compensator is corresponding to be taken as:

D(s)＝0.1s+1

The given speed command of experimental system is the square-wave signal of positive and negative 300rpm, and the cycle is 0.4s.The dynamic response process that accompanying drawing 4 instructs to this square wave for the second order desired reference model of choosing in the experiment, model was followed the trail of the dynamic response process of adaptive controller to this square wave instruction when accompanying drawing 5 was unloaded, model is followed the trail of the dynamic response process of adaptive controller to this square wave instruction when accompanying drawing 6 is twice rotor inertia, when adding perturbation load, and accompanying drawing 7 is the adaptive rate adjustment process from setting parameter that model is followed the trail of adaptive controller.

Emulation and experimental result show, when load inertia changes, the servo-drive system that adopts model to follow the trail of adaptive controller can keep the dynamic and steady state speed tracking error of basically identical, when the change load rotating inertia is twice rotor inertia, the rotating speed steady-state error maintains in 1.0%, rotating speed response is more steady, and system is preferable to the adaptability of the parameter of electric machine and load variations.

Description of drawings

Accompanying drawing 1 is the speed control synoptic diagram based on the model method for tracing

Accompanying drawing 2 is followed the trail of the reference model synoptic diagram of adaptive control for model

Model was followed the trail of adaptive controller control effect synoptic diagram when accompanying drawing 3 was moment of inertia wide variation (1J, 2J, 4J, 6J)

The reference model synoptic diagram of accompanying drawing 4 for choosing in the experiment

Model was followed the trail of adaptive controller control effect synoptic diagram when accompanying drawing 5 was unloaded

Model is followed the trail of the control effect synoptic diagram of adaptive controller when accompanying drawing 6 is twice rotor inertia, when adding perturbation load

Accompanying drawing 7 is followed the trail of the parameter adjustment process synoptic diagram of adaptive controller for model.

Claims (1)

1. adaptive servo controller based on model following is characterized in that:

1) from setting parameter K _A0, K _A1, K _rAnd K _ErAdaptive rate be respectively:

By y _mAnd υ (t) calculates K through following formula _A0,

$K_{a0}={\&Integral;}_0^t{l}_{1}v\left(\mathrm{\&tau;}\right)y_m\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_{2}v\left(t\right)y_m\left(t\right)+K_{a00}$

By y _mAnd υ (t) calculates K through following formula _A1,

$K_{a1}={\&Integral;}_0^t{l}_3\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)d{y}_m\left(\mathrm{\&tau;}\right)/d\left(t\right)\mathrm{d\&tau;}+l_4\mathrm{\&upsi;}\left(t\right)d{y}_m\left(t\right)/d\left(t\right)+K_{a10}$

Calculate K by υ (t) and r (t) through following formula _r,

$K_r={\&Integral;}_0^t{l}_5\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)r\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_6\mathrm{\&upsi;}\left(t\right)r\left(t\right)+K_{r0}$

Calculate K by υ (t) and e (t) through following formula _Er,

$K_{\mathrm{er}}={\&Integral;}_0^t{l}_7\mathrm{\&upsi;}\left(\mathrm{\&tau;}\right)e\left(\mathrm{\&tau;}\right)\mathrm{d\&tau;}+l_8\mathrm{\&upsi;}\left(t\right)e\left(t\right)+K_{\mathrm{er}0}$

L wherein ₁, l ₃, l ₅And l ₇Be respectively K _A0, K _A1, K _rAnd K _ErIntegration constant, l ₂, l ₄, l ₆And l ₈Be respectively K _A0, K _A1, K _rAnd K _ErProportionality constant.

2) by from setting parameter K _A0, K _A1, K _rAnd K _ErDerivation actual servo motor q shaft current command value i _QrefThe reckoning process, step is as follows:

By K _A0, K _A1, y _mCalculate U through following formula _P1,

U _p1＝K _a0·y _m(t)+K _a1·dy _m(t)/d(t)

By K _r, r (t) calculates U through following formula _P2,

U _p2＝K _r·r(t)

By K _Er, e (t) calculates U through following formula _P3,

U _p3＝K _er·e(t)

By U _P1, U _P2, U _P3Calculate i through following formula _Qref

i _qref＝U _p1+U _p2+U _p3

Images