CN101902187A - Control system for low-speed running of permanent magnet motor - Google Patents

Abstract

The invention discloses a control system for low-speed running of a permanent magnet motor, which consists of a speed controller, a current loop, an angle sensor, a disturbing estimator, a tooth socket torque estimator and a tooth socket torque calculator. The sizes of the tooth socket torque and the rest resisting moments are quickly calculated by the disturbing estimator and the tooth socket torque estimator; the influence of the tooth socket torque on the speed is further accurately calculated through the tooth socket torque calculator so as to make compensation; the stability of the control system is increased by controlling integral coefficients of the speed controller in real time, and equivalently, output current is changed through changing the integral coefficients to compensate for aperiodic resisting moments at the same time; and the response speed of the control system is increased by carrying out the advanced compensation of the delay of the controller so that the disturbance is quicker to attenuate. The control system has higher correctness and stability, and is particularly suitable for realizing the low-speed control of the permanent magnet motor under the condition of the larger change range of the resisting moments.

Description

A kind of control system that is used for low-speed running of permanent magnet motor

Technical field

The present invention relates to a kind of electric machine control system, more particularly, the present invention relates to a kind of control system that is used for low-speed running of permanent magnet motor.

Background technology

Cogging torque is a kind of intrinsic phenomenon of magneto, because the existence of teeth groove causes the motor output torque pulsation, has influenced the precision of control system, and this phenomenon is particularly evident under the low speed situation.Prior art except that improving on electric machine structure, mainly is being the indemnifying measure that increases in control algolithm cogging torque to the inhibition of cogging torque aspect the control system, thereby reduces the fluctuation of motor speed.The system of introducing as " electrical technology " 2009 the 3rd phase 21-24 pages or leaves " directly driving annular permanent magnet torque motor low speed cogging torque ripple compensation research " literary composition, by speed control, electric current loop, motor, velocity transducer and disturbance estimator (observer) are formed, wherein speed control is based on the PI controller of classical control theory, the disturbance estimator is based on the theoretical estimator of being constructed of state observer in the modern control theory (estimator), it is input as the current value I of electric current loop output and the actual measurement speed omega of velocity transducer output, suppose that external disturbance has only the cogging torque disturbance, then the disturbance estimator is output as cogging torque estimated value T _c' and velocity estimation value ω ', with cogging torque estimated value T _c' send into the electric current loop input can compensate to cogging torque, the feedback speed of control system can also can be with the actual measurement speed omega of velocity transducer output with velocity estimation value ω, uses estimating speed ω ' as the control system block diagram of feedback speed as shown in Figure 1.

The weak point of this control system is not have to consider the influence of other moment of resistance except that cogging torque.The estimator constructed based on modern control theory is that cost exchanges high response speed for to increase error, and the response speed of estimator design is fast more, though can make the quicker decay of disturbance, compensating error is also big more.Because cogging torque is a periodic disturbance, and other moment of resistance such as moment of friction or elastic resistance square can be regarded aperiodic disturbance as, aperiodic disturbance is to the influence and the periodic disturbance and inequality of control system time-delay, single estimator can't accurately reduce the phase place of external disturbance, under the bigger situation of external disturbance excursion, the error that is caused by phase place can be bigger.If the excursion of aperiodic disturbance is bigger in addition, then the overshoot of control system has reduced the stability of control system also in continuous variation, has limited the maximum bandwidth that control system can obtain, and has reduced the response speed of control system.Therefore need extract cogging torque and all the other moments of resistance respectively and be compensated.

Summary of the invention

Based on the problems referred to above, the objective of the invention is to propose a kind of accuracy and the higher magneto of the stability control system that slowly runs.

The present invention is made up of speed control, electric current loop, motor, angular transducer, disturbance estimator, cogging torque estimator and cogging torque calculator; The output current I that is input as electric current loop of described disturbance estimator and actual measurement speed omega, wherein survey speed omega by the output angle θ of angular transducer through differentiating acquisition, be output as external disturbance T _tEstimated value T _t' and estimating speed ω ', the state space equation of describing its feature is:

$\left[\begin{array}{c}\frac{d{\mathrm{\ω}}^{\′}}{\mathrm{dt}}\\ \frac{{d{T_t}^{\′}}^{}}{\mathrm{dt}}\end{array}\right]=\left[\begin{array}{cc}0& -\frac{1}{J}\\ 0& 0\end{array}\right]\begin{array}{}\end{array}\left[\begin{array}{c}{\mathrm{\ω}}^{\′}\\ {T_t}^{\′}\end{array}\right]+\left[\begin{array}{c}\frac{K_t}{J}\\ 0\end{array}\right]I+\left[\begin{array}{c}{g}_1\\ g_2\end{array}\right][\mathrm{\ω}-\left[\begin{array}{cc}1& 0\end{array}\right]\begin{array}{}\end{array}\left[\begin{array}{c}{\mathrm{\ω}}^{\′}\\ {T_t}^{\′}\end{array}\right]]---\left(1\right)$

Wherein J is a load rotating inertia, K _tBe motor torque coefficient, g ₁And g ₂Be feedback oscillator; The output angle θ that is input as angular transducer of described cogging torque estimator is output as cogging torque estimated value T _c', computing formula is:

T _c′＝T _cmsin[N _p(θ-θ ₀)] (2)

N wherein _pBe the spatial distribution of cogging torque, θ ₀Be cogging torque zero point, T _CmPeak value for cogging torque; With disturbance estimated value T _t' with cogging torque estimated value T _c' subtract each other, and divided by feedback speed ω _-, obtain a procedure parameter b, wherein feedback speed ω _-Can also can use estimating speed ω ' with the actual measurement speed omega; Described speed control is a self-adaptive PID controller, its proportionality coefficient K _p, integral coefficient K _iWith differential coefficient K _dHave following relation:

K _d＝K _pτ _p (3)

$K_i=K_p\frac{\frac{b}{\mathrm{<figure-callout\; id="1"\; label="J"\; filenames="HSA00000152212100013.png"\; state="\{\{state\}\}">J</figure-callout>}}}{1+{\mathrm{\&tau;}}_i\frac{b}{J}}---\left(4\right)$

τ wherein _pAnd τ _iBe respectively the time-delay that ratio and integral operation are caused; Described cogging torque calculator be input as described procedure parameter b, feedback speed ω _-With angle θ, be output as velocity perturbation ω _c', computing formula is:

${{\mathrm{\&omega;}}_c}^{\&prime;}=\frac{T_{\mathrm{cm}}}{\sqrt{J^2{N}_p^2{\mathrm{\&omega;}}_{-}^2+b^2}}\sin [N_p(\mathrm{\&theta;}-{\mathrm{\&theta;}}_0)-{\tan }^{-1}\frac{{\mathrm{JN}}_p{\mathrm{\&omega;}}_{-}}{b}]---\left(5\right)$

With velocity perturbation ω _c' send into speed ring error end speed is compensated.The present invention uses estimating speed ω ' as the block diagram of feedback speed as shown in Figure 2; The block diagram of described self-adaptive PID controller as shown in Figure 3, τ wherein _p, τ _iAnd τ _dThe time-delay that is respectively ratio, integration and differentiates and caused; The block diagram of described disturbance estimator as shown in Figure 3.

Principle of the present invention is: increase an estimator cogging torque and all the other moments of resistance are made a distinction; Further accurately calculate the influence of cogging torque by the cogging torque calculator, reduce the error brought by estimator, more accurate cogging torque is compensated speed; By real-time control to the speed control integral coefficient, be equivalent to real-time control to zero point that controller produces, and make it offset the time-delay that causes by aperiodicity moments of resistance such as moment of friction, elastic resistance squares all the time, the retentive control system has fixing overshoot, increased the stability of control system, also be equivalent to simultaneously change output current by changing integral coefficient, thereby the aperiodicity moment of resistance is compensated, and this compensation way makes control system have better adaptability under the bigger situation of moment of resistance excursion; By the lead compensation (differentiating) of delaying time of controller being increased the response speed of control system, disturbance is decayed faster.

Description of drawings

Fig. 1 tradition band estimator also utilizes the control system block diagram of estimating speed as feedback.

Fig. 2 the present invention utilizes the control system block diagram of estimating speed as feedback.

Fig. 3 PID controller of the present invention block diagram.

Fig. 4 disturbance estimator of the present invention block diagram.

Embodiment

1, builds control system according to block diagram shown in Figure 2; Wherein the block diagram of self-adaptive PID controller as shown in Figure 3; The block diagram of disturbance estimator as shown in Figure 4; The feedback speed ω of the outer shroud of control system _-Both can use the also available actual measurement speed omega of estimating speed ω ', generally speaking, resolution as transducer is enough high, and the output of then available estimator is as feedback, as the not high but response of sensor resolution enough soon then the speed of available sensors differential gained as feedback.

2, the limit op of selected disturbance estimator ₁And op ₂, can make op ₁=op ₂And be chosen in control system three dB bandwidth the 2-6 of corresponding limit doubly locate; Utilize the acker function calculation of Matlab software to go out the needed feedback oscillator g of disturbance estimator ₁And g ₂, the using method of acker function is:

[g ₁?g ₂]＝acker(A′，B′，op) (6)

Wherein

A′＝[0 0；-1/J?0]

C′＝[1；0]

op＝[op ₁?op ₂]

3, cogging torque estimator and cogging torque calculator are broken from the loop, regulate the PID controller to determine the optimal proportion COEFFICIENT K _pValue, this value also can be determined by the root locus diagram of open-loop transfer function.Allow control system control motor at the uniform velocity turn round, more than 10 times of the desirable real work speed of its speed, according to (5) formula, when speed is enough high, the b in the denominator ²Item can be ignored the output angle θ of acquisition angles transducer and feedback speed ω _-, be abscissa ω with θ _-Be the ordinate figure that draws, write down the peak value ω of this figure _m, average

Period T _θWith as feedback speed ω _-Equal average

And slope is any one angle value θ of timing _C0, cogging torque spatial distribution N then _p, cogging torque θ at zero point ₀With cogging torque peak value T _CmValue can determine by following several formulas:

$N_p=\frac{2\mathrm{\&pi;}}{T_{\mathrm{\&theta;}}}---\left(7\right)$

${\mathrm{\&theta;}}_0=\frac{N_p{\mathrm{\&theta;}}_{c0}-\frac{\mathrm{\&pi;}}{2}}{N_p}---\left(8\right)$

T _cm＝JN _pω _m (9)

Can be under friction speed survey several groups and get the limit and determine cogging torque peak value T _CmValue.

4., the cogging torque estimator is connected into loop with the cogging torque calculator, form complete control system, wherein the expression formula of cogging torque estimator is (2) formula, and the expression formula of cogging torque calculator is (5) formula.

Claims (1)

1. control system that is used for low-speed running of permanent magnet motor, it is made up of speed control, electric current loop, motor, angular transducer, disturbance estimator, cogging torque estimator and cogging torque calculator, it is characterized in that:

The output current I that is input as electric current loop of described disturbance estimator and actual measurement speed omega, wherein survey speed omega by the output angle θ of angular transducer through differentiating acquisition, be output as external disturbance T _tEstimated value T ' _tAnd estimating speed ω ', the state space equation of describing its feature is:

$\left[\begin{array}{c}\frac{d{\mathrm{\&omega;}}^{\&prime;}}{\mathrm{dt}}\\ \frac{{d{T_t}^{\&prime;}}^{}}{\mathrm{dt}}\end{array}\right]=\left[\begin{array}{cc}0& -\frac{1}{J}\\ 0& 0\end{array}\right]\begin{array}{}\end{array}\left[\begin{array}{c}{\mathrm{\&omega;}}^{\&prime;}\\ {T_t}^{\&prime;}\end{array}\right]+\left[\begin{array}{c}\frac{K_t}{J}\\ 0\end{array}\right]I+\left[\begin{array}{c}{g}_1\\ g_2\end{array}\right][\mathrm{\&omega;}-\left[\begin{array}{cc}1& 0\end{array}\right]\begin{array}{}\end{array}\left[\begin{array}{c}{\mathrm{\&omega;}}^{\&prime;}\\ {T_t}^{\&prime;}\end{array}\right]]---\left(1\right)$

Wherein J is a load rotating inertia, K _tBe motor torque coefficient, g ₁And g ₂Be feedback oscillator;

The output angle θ that is input as angular transducer of described cogging torque estimator is output as cogging torque estimated value T _c', computing formula is:

T _c′＝T _cmsin[N _p(θ-θ ₀)]

N wherein _pBe the spatial distribution of cogging torque, θ ₀Be cogging torque zero point, T _CmBe the peak value of cogging torque, with disturbance estimated value T _t' with cogging torque estimated value T _c' subtract each other, and divided by feedback speed ω _-, obtain a procedure parameter b, wherein feedback speed ω _-Can also can use estimating speed ω ' with the actual measurement speed omega;

Described speed control is a self-adaptive PID controller, its proportionality coefficient K _p, integral coefficient K _iWith differential coefficient K _dHave following relation:

K _d＝K _pτ _p

$K_i=K_p\frac{\frac{b}{J}}{1+{\mathrm{\&tau;}}_i\frac{b}{J}}$

τ wherein _pAnd τ _iBe respectively the time-delay that ratio and integral operation are caused;

Described cogging torque calculator be input as described procedure parameter b, feedback speed ω _-With angle θ, be output as velocity perturbation ω _c', computing formula is:

${{\mathrm{\&omega;}}_c}^{\&prime;}=\frac{T_{\mathrm{cm}}}{\sqrt{J^2{N}_p^2{\mathrm{\&omega;}}_{-}^2+b^2}}\sin [N_p(\mathrm{\&theta;}-{\mathrm{\&theta;}}_0)-{\tan }^{-1}\frac{{\mathrm{JN}}_p{\mathrm{\&omega;}}_{-}}{b}]$

With velocity perturbation ω _c' send into speed ring error end speed is compensated.

Images