Speed control method for permanent magnet synchronous motor considering current saturation and disturbance suppression

Abstract

A speed control method for a permanent magnet synchronous motor considering current saturation and disturbance suppression aims to effectively ensure that a current of the motor is always within a given range to avoid the problem of control performance reduction caused by the fact that the current gets into a saturation state, ensure the safety of a system, do not need to use unavailable state variables such as motor acceleration and the like, effectively estimate and compensate disturbances including parameters uncertainty and unknown load torque disturbance existing in a permanent magnet synchronous motor system, and rapidly and accurately control a speed of the motor finally. There is no need to configure a plurality of sensors in practical industrial application, so system building costs can be reduced on the one hand, and the stability of the system can be improved on the other hand.

Claims

1. A speed control method for a permanent magnet synchronous motor considering current saturation and disturbance suppression, comprising steps of: step 1: determining mathematical model of speed loop of permanent magnet synchronous motor: by taking rotor coordinates d−q axes of the motor as reference coordinates system, on the premise of fully considering system parameters uncertainty and unknown load torque disturbance, building the mathematical model of speed loop of permanent magnet synchronous motor: $\stackrel{.}{w}=\left(\frac{K_{\mathrm{to}}}{J_o}+\Delta a\right)i_q-\left(\frac{B_o}{J_o}+\Delta b\right)w-\frac{T_L}{J};$ where w represents speed of the motor, i.sub.q represents stator current of q axis, is permanent magnet synchronous motor speed controller to be designed, T.sub.L represents unknown load torque of the system, J.sub.o, K.sub.to, and B.sub.o respectively represent nominal values of moment of inertia, torque constant, and viscous friction coefficient, J, K.sub.t, and B respectively represent true values of the moment of inertia, the torque constant, and the viscous friction coefficient, and Δa=K.sub.t/J−K.sub.to/J.sub.o and Δb=B/J−B.sub.o/J.sub.o represent differences between true values of the system parameters and nominal values; intensively expressing disturbances caused by system parameters uncertainty and unknown load torque disturbance as lumped disturbances: $d\left(t\right)=-\Delta a{i}_q+\Delta bw+\frac{T_L}{J},$ which satisfies the following bounded conditions: |d(t)|<l.sub.1, |{dot over (d)}(t)|<l.sub.2, where l.sub.1 and l.sub.2 represent positive constants; further expressing the mathematical model of the speed loop of the permanent magnet synchronous motor as $\stackrel{.}{w}=\frac{K_{to}}{J_o}{i}_q-\frac{B_o}{J_o}w-d;$ step 2: determining control objectives of a speed-governing system of the permanent magnet synchronous motor: 2-1) rapid and accurate tracking of speed of the motor: $\underset{t.fwdarw.\infty }{\lim }w=w^{*};$ where w* represents given speed of the permanent magnet synchronous motor; 2-2) current saturation constraint: |i.sub.q(t)|<I.sub.max, ∀t≥0; where I.sub.max represents the maximum allowable current of the permanent magnet synchronous motor during normal operation; 2-3) disturbance estimation: estimating system disturbances including parameters uncertainty and unknown load torque disturbance on line by means of disturbance observer; step 3: designing a super-twisting disturbance observer to accurately estimate lumped disturbances of the permanent magnet synchronous motor: defining a speed estimation error signal: ε=ŵ−w; where ŵ represents an estimated speed of the permanent magnet synchronous motor; then giving an integral sliding surface as shown below: $s=\left(\frac{d}{dt}+\lambda \right){\int }_0^t\epsilon \mathrm{dt};$ where λ represents a positive constant greater than 0; a form of the super-twisting disturbance observer is as follows: $\stackrel{.}{\hat{w}}=\frac{K_{to}}{J_o}{i}_q-\frac{B_o}{J_o}\hat{w}-f;$ where ƒ(t) represents control law of the super-twisting disturbance observer, and is equivalent to estimated value of lumped disturbances d(t) when the disturbance observer tends to be stable, specific structure of ƒ(t) is as follows: $f=\left(\lambda -\frac{B_o}{J_o}\right)\epsilon +k_1{.Math.s.Math.}^{\frac{1}{2}}\mathrm{sign}\left(s\right)+k_2{\int }_0^t\mathrm{sign}\left(s\right)\mathrm{dt};$ where sign(•) represents signum function, and k.sub.1 and k.sub.2 represent positive constants satisfying following condition: $\{\begin{array}{c}32l_2<k_1^2<8(k_2-l_2)\\ k_2>5l_2\end{array};$ step 4: constructing a permanent magnet synchronous motor speed controller based on system model considering system parameters uncertainty and unknown load torque influence; defining motor speed tracking error: e=w*−w; in order to achieve control objective of current constraint, introducing a saturation function: $\Psi \left(\right)=\frac{}{\sqrt{n+{}^2}};$ where n represents constant greater than or equal to 0; it is known that |Ψ(•)|≤1; on a premise of avoiding introducing unavailable state variables, constructing a quasi-integral signal: E=Λ+k.sub.i∫.sub.0.sup.tedt, {dot over (Λ)}=−k.sub.i(Λ+k.sub.i∫.sub.0.sup.tedt); where Λ represents auxiliary signal, k.sub.i represents positive constant and is also control gain of subsequently given permanent magnet synchronous motor speed controller, a dynamic equation of the quasi-integral signal is: Ė=−k.sub.iE+k.sub.ie; in combination with the super-twisting disturbance observer, designing a permanent magnet synchronous motor speed controller of a following form: $i_q=\frac{J_o}{K_{to}}\left[f+\frac{B_o}{J_o}{w}^{*}+k_p\psi \left(e\right)+k_i\psi \left(E\right)+k_l\mathrm{sign}\left(e\right)\right];$ where k.sub.p, k.sub.i, k.sub.l represent positive adjustable control gains; step 5: determining control gains of the permanent magnet synchronous motor speed controller: according to the controller, it is known that values of control gains k.sub.p, k.sub.i, k.sub.l determine control input, is the magnitude of the current of the motor, in order to ensure an objective of the current saturation constraint in 2-2), the control gains shall be determined according to a following method: $\{\begin{array}{c}{k}_p+k_i+k_l\le \frac{K_{\mathrm{to}}}{J_o}{I}_{\max }-l_1-\mu -\frac{B_o}{J_o}{w}^{*}\\ k_l>\mu \end{array};$ where all that on a right side of a first inequality are all available constants, μ represents an upper limit of an estimation error of the super-twisting disturbance observer and satisfies |d|=|d(t)−ƒ(t)|≤μ, ∀t>0, {tilde over (d)} represents an estimation error of the super-twisting disturbance observer; step 6: achieving a control method: controlling an operation of the permanent magnet synchronous motor by means of the permanent magnet synchronous motor speed controller so as to control the speed of the motor, constrain the current of the motor, estimate and compensate system disturbances, to achieve control objectives of a speed-governing system of the permanent magnet synchronous motor.

Description

DESCRIPTION OF DRAWINGS

(1) FIG. 1 is a flow chart of a speed control method for a permanent magnet synchronous motor of the present invention;

(2) FIG. 2 is a block diagram of speed and current double closed loop vector control of a permanent magnet synchronous motor of the present invention;

(3) FIG. 3 is a diagram of simulation results of a speed control method for a permanent magnet synchronous motor proposed by the present invention; and

(4) FIG. 4 is a diagram of simulation results of a PI control method in the prior art.

DETAILED DESCRIPTION

(5) In order to more visually and clearly reflect the technical solution and the advantages of the present invention, the present invention will be further described below in detail in combination with the drawings and embodiments.

Embodiment 1

(6) This embodiment discloses a speed control method for a permanent magnet synchronous motor considering current saturation and disturbance suppression, as shown in FIG. 1, comprising:

(7) (I) Determining a mathematical model of a speed loop of a permanent magnet synchronous motor:

(8) a mathematical model of a permanent magnet synchronous motor taking rotor coordinates (d-q axes) as reference coordinates system is shown as follows:

(9) $\begin{array}{cc}\left(\begin{array}{c}{\stackrel{.}{i}}_d\\ {\dot{i}}_q\\ \stackrel{.}{w}\end{array}\right)=\left(\begin{array}{ccc}-\frac{R}{L}& n_{p}w& 0\\ -n_{p}w& -\frac{R}{L}& \frac{n_p{\psi }_f}{L}\\ 0& \frac{K_t}{J}& -\frac{B}{J}\end{array}\right)\left(\begin{array}{c}{i}_d\\ i_q\\ w\end{array}\right)+\left(\begin{array}{c}\frac{u_d}{L}\\ \frac{u_q}{L}\\ -\frac{T_L}{J}\end{array}\right)& \left(1\right)\end{array}$

(10) where u.sub.q, u.sub.d represent stator voltages of q axis and d axis, i.sub.q, i.sub.d represent stator currents of q axis and d axis, w represents a speed of the motor, n.sub.p represents the number of the pole pairs of the permanent magnet synchronous motor, L and R represent stator inductance and stator resistance respectively, ψ.sub.ƒ represents a permanent magnet flux linkage, K.sub.t represents a torque constant, and T.sub.L, B, J represent a load torque, a viscous friction coefficient and a moment of inertia respectively.

(11) The vector control framework is one of the most widely used control frameworks in the field of permanent magnet synchronous motor control at present. FIG. 2 shows a control block diagram of a speed-governing system of a permanent magnet synchronous motor based on vector control, which adopts a cascade structure of a speed loop and two current loops. In order to decouple speed control from current control, the reference current of d axis is set to 0. In these two current loops, two classical PI controllers are used to stabilize current errors of d-q axes. The present invention mainly provides a design scheme of a permanent magnet synchronous motor speed controller.

(12) The following mathematical model of the speed loop of the permanent magnet synchronous motor is built on the premise that system parameters uncertainty and the unknown load torque disturbance are fully considered:

(13) $\begin{array}{cc}\stackrel{.}{w}=\left(\frac{K_{\mathrm{to}}}{J_o}+\Delta a\right)i_q-\left(\frac{B_o}{J_o}+\Delta b\right)w-\frac{T_L}{J}& \left(2\right)\end{array}$
where w represents a speed of the motor, i.sub.q represents a stator current of q axis, i.e. a permanent magnet synchronous motor speed controller to be designed, J.sub.o, K.sub.to, and B.sub.o respectively represent nominal values of a moment of inertia, a torque constant, and a viscous friction coefficient, and Δa=K.sub.t/J−K.sub.to/J.sub.o and Δb=B/J−B.sub.o/J.sub.o represent differences between the true values of the system parameters and the nominal values.
Disturbances caused by system parameters uncertainty and unknown load torque disturbance are intensively expressed as a lumped disturbances term:

(14) $\begin{array}{cc}d\left(t\right)=-{\Delta \mathrm{ai}}_q+\Delta bw+\frac{T_L}{J}& \left(3\right)\end{array}$
which meets the following bounded conditions:
|d(t)|<l.sub.1,|d(t)|<l.sub.2  (4)
where l.sub.1 and l.sub.2 represent positive constants.
The system model can be further expressed as

(15) $\begin{array}{cc}\stackrel{.}{w}=\frac{K_{\mathrm{to}}}{J_o}{i}_q-\frac{B_o}{J_o}w-d& \left(5\right)\end{array}$
(II) Determining control objectives of a speed-governing system of the permanent magnet synchronous motor:
in the process of operation, the control objectives of the speed-governing system of the permanent magnet synchronous motor includes the following three parts: 1) adjust the speed of the motor to reach a given speed w* quickly and accurately; 2) effectively ensure that the current i.sub.q(t) of the motor is within a given safety range in the whole control process; 3) estimate lumped disturbances d(t) of the system on line by means of a disturbance observer.
To sum up, the control objectives of the present invention can be described as the following mathematical form:

(16) $\begin{array}{cc}\underset{t.fwdarw.\infty }{\lim }w=w^{*}& \left(6\right)\end{array}$$\begin{array}{cc}\left|i_q\left(t\right)\right|<I_{\max },\forall t\ge 0& \left(7\right)\end{array}$$\begin{array}{cc}\underset{t.fwdarw.\infty }{\lim }\tilde{d}\left(t\right)=d\left(t\right)-f\left(t\right)=0& \left(8\right)\end{array}$
where l.sub.max represents the maximum allowable current of the permanent magnet synchronous motor during normal operation, ƒ(t) represents an estimated value of the lumped disturbances d(t), and {tilde over (d)}(t) represents an estimation error of the disturbance observer.
(III) Designing a super-twisting disturbance observer to accurately estimate lumped disturbances of permanent magnet synchronous motor:
based on the mathematical model of the speed loop of the permanent magnet synchronous motor in (5), designing a super-twisting disturbance observer of the following form:

(17) $\begin{array}{cc}\stackrel{.}{\stackrel{ˆ}{w}}=\frac{K_{\mathrm{to}}}{J_o}{i}_q-\frac{B_o}{J_o}\hat{w}-f& \left(9\right)\end{array}$
where ŵ represents a motor speed estimation signal, and ƒ(t) represents a super-twisting controller to be designed;
defining a speed estimation error signal:
ε=ŵ−w  (10)
then giving an integral sliding surface as shown below:

(18) $\begin{array}{cc}s=\left(\frac{d}{dt}+\lambda \right){\int }_0^t\epsilon \mathrm{dt}& \left(11\right)\end{array}$
where λ represents a positive constant greater than 0;
taking the derivative a sliding surface s with respect to time t, and combining with equations (5), (9) and (10), so it is easy to obtain that:

(19) $\begin{array}{cc}\stackrel{.}{s}=\stackrel{.}{\epsilon }+\lambda \epsilon =\left(\lambda -\frac{B_o}{J_o}\right)\epsilon -f+d& \left(12\right)\end{array}$
designing a super-twisting controller ƒ(t) as follows:

(20) 0$\begin{array}{cc}f=\left(\lambda -\frac{B_o}{J_o}\right)\epsilon +k_1{.Math.s.Math.}^{\frac{1}{2}}\mathrm{sign}\left(s\right)+k_2{\int }_0^t\mathrm{sign}\left(s\right)\mathrm{dt}& \left(13\right)\end{array}$
where k.sub.1 and k.sub.2 represent positive constant satisfying the following condition:

(21) $\begin{array}{cc}\{\begin{array}{c}32l_2<k_1^2<8\left(k_2-l_2\right)\\ k_2>5l_2\end{array}& \left(14\right)\end{array}$
substituting equation (13) into equation (12), obtaining:

(22) $\begin{array}{cc}\stackrel{.}{s}=-k_1{.Math.s.Math.}^{\frac{1}{2}}\mathrm{sign}\left(s\right)-k_2{\int }_0^t\mathrm{sign}\left(s\right)\mathrm{dt}+d& \left(15\right)\end{array}$
here, introducing a new variable:
η=d−k.sub.2∫.sub.0.sup.tsign(s)  (16)
conducting state transformation on equation (15), obtaining:

(23) $\begin{array}{cc}\{\begin{array}{c}\stackrel{.}{s}=\eta -k_1|s{|}^{\frac{1}{2}}\mathrm{sign}\left(s\right)\\ \stackrel{.}{\eta }=\stackrel{.}{d}-k_2\mathrm{sign}\left(s\right)\end{array}& \left(17\right)\end{array}$
then, by combining with the control gains condition given in equation (14) and using theorem 1 and lemma 5 in the reference [A. Polyakov and A. Poznyak, Reaching time estimation for “super-twisting” second order sliding mode controller via Lyapunov function designing, IEEE Transactions on Automatic Control, 2009, 54(8):1951-1955.], it can be deduced that within finite time:
s.fwdarw.0,{dot over (s)}.fwdarw.0  (18)
if s=0, {dot over (s)}=0, by combining with equations (11) and (12), it can be known that:
ε.fwdarw.0ƒ.fwdarw.d  (19)
here, if defining the disturbance estimation error {tilde over (d)}=d(t)−ƒ(t), it can be known that:

(24) $\begin{array}{cc}\underset{t.fwdarw.\infty }{\lim }\tilde{d}\left(t\right)=d\left(t\right)-f\left(t\right)=0& \left(20\right)\end{array}$

(25) The above theoretical analysis indicates that the super-twisting disturbance observer designed by the present invention can accurately estimate lumped disturbances of the system. In order to ensure the rigor of the proposed technical solution, the present invention will consider the existence of a disturbance estimation error, which meets the following bounded condition:
|{tilde over (d)}|=|d(t)−ƒ(t)|≤μ,∀t>0  (21)
where μ represent an upper limit of the disturbance estimation error, which is a small positive constant.
(IV) Constructing a permanent magnet synchronous motor speed controller based on a system model considering system parameters uncertainty and unknown load torque influence:
defining a motor speed tracking error:
e=w*−w  (22)
where w* represents a given speed of the motor;
in order to achieve a control objective of current constraint, introducing the following saturation function:

(26) $\begin{array}{cc}\Psi \left(\right)=\frac{}{\sqrt{n+{}^2}}& \left(23\right)\end{array}$
where n is a constant greater than or equal to 0; it is easy to know from the form of the saturation function that:
|Ψ(•)≤1  (24)
further, on the premise of avoiding introducing unavailable state variables such as motor acceleration signal and the like, constructing the quasi-integral signal of the following type:
E=Λ+k.sub.i∫.sub.0.sup.tedt,{dot over (Λ)}=−k.sub.i(Λ+k.sub.i∫.sub.0.sup.tedt)  (25)
where Λ represents an auxiliary signal, k.sub.i represents a positive constant and is also a control gain of a permanent magnet synchronous motor speed controller subsequently given, so it is easy to know according to equation (25) that a dynamic equation of the quasi-integral signal is:
Ė=−k.sub.iĖ+k.sub.ie  (26)
in combination with the designed super-twisting disturbance observer, designing a permanent magnet synchronous motor speed controller:

(27) $\begin{array}{cc}{i}_q=\frac{J_o}{K_{\mathrm{t0}}}\left[f+\frac{B_o}{J_o}{w}^{*}+k_p\psi \left(e\right)+k_i\psi \left(E\right)+k_l\mathrm{sign}\left(e\right)\right]& \left(27\right)\end{array}$
where k.sub.p, k.sub.i, k.sub.l represent positive adjustable control gains, and k.sub.l satisfies
k.sub.l>μ  (28)
closed loop system stability analysis:
constructing a Lyapunov function candidate of the following form:

(28) $\begin{array}{cc}V=\frac{1}{2}{e}^2+{\int }_0^E\psi \left(z\right)\mathrm{dz}& \left(29\right)\end{array}$
taking the derivative of the above equation with respect to time and combining with equations (5), (21), (22) and (26), so the following conclusion can be drawn:

(29) $\begin{array}{cc}\begin{array}{c}\stackrel{.}{V}=e\stackrel{.}{e}+\stackrel{.}{E}\psi \left(E\right)\\ =e\left({\stackrel{.}{w}}^{*}-\frac{K_{\mathrm{to}}}{J_o}{i}_q+\frac{B_o}{J_o}w+d\left(t\right)\right)+k_{i}e\psi \left(E\right)-k_{i}E\psi \left(E\right)\\ =e\left(-\frac{K_{\mathrm{to}}}{J_o}{i}_q+\frac{B_o}{J_o}w+d\left(t\right)+k_i\psi \left(E\right)\right)-k_{i}E\psi \left(E\right)\\ =e\left(-\frac{B_o}{J_o}e-k_p\psi \left(e\right)+d\left(t\right)-f-k_l\mathrm{sign}\left(e\right)\right)-k_{i}E\psi \left(E\right)\\ =-\frac{B_o}{J_o}{e}^2-k_{p}e\psi \left(e\right)-k_{i}E\psi \left(E\right)+e\left(d\left(t\right)-f-k_l\mathrm{sign}\left(e\right)\right)\\ =-\frac{B_o}{J_o}{e}^2-k_{p}e\psi \left(e\right)-k_{i}E\psi \left(E\right)+e\left(\tilde{d}-k_l\mathrm{sign}\left(e\right)\right)\le \\ -\frac{B_o}{J_o}{e}^2-k_{p}e\psi \left(e\right)-k_{i}E\psi \left(E\right)+e\mu -k_l.Math.e.Math.\le \\ -\frac{B_o}{J_o}{e}^2-k_{p}e\psi \left(e\right)-k_{i}E\psi \left(E\right)-\left(k_l-\mu \right).Math.e.Math.\end{array}& \left(30\right)\end{array}$
when the control gain k.sub.l is selected to satisfy equation (28), equation (30) can be further collated as
{dot over (V)}≤0  (31)
next, by combining with equation (29) and equation (31), it can be known that
0≤V(t)<V(0)<<+∞
the above result shows that v(t) is bounded, i.e. v(t)∈ζ.sub.∞, further, according to the form of v(t), it can be obtained that
e,E∈ζ.sub.∞.Math.i.sub.q,ė,Ė∈ζ.sub.∞  (32)
according to the result of equation (30), it can be known that

(30) $\begin{array}{cc}\stackrel{.}{V}\le -\frac{B_o}{J_o}{e}^2& \left(33\right)\end{array}$
by integrating both sides of the above equation, it is easy to obtain that:

(31) 0$\begin{array}{cc}V\left(t\right)-V\left(0\right)\le -\frac{B_o}{J_o}{\int }_0^t{e}^2\mathrm{dt}& \left(34\right)\end{array}$
further, it is easy to obtain that

(32) $\begin{array}{cc}0\le V\left(t\right)\le V\left(0\right)-\frac{B_o}{J_o}{\int }_0^t{e}^2\mathrm{dt}.Math.{\int }_0^t{e}^2\mathrm{dt}\le \frac{J_o}{B_o}V\left(0\right)& \left(35\right)\end{array}$
the above result means that the error signal e is square-integrable, i.e.
e∈ζ.sub.2  (36)
and because it has been proved that
e,ė∈ζ.sub.∞  (37)
according to Barbalat lemma, it can be known that the speed tracking error e asymptotically converges to 0, i.e.

(33) $\begin{array}{cc}\underset{t.fwdarw.\infty }{\lim }e=0& \left(38\right)\end{array}$

(34) The analysis proves that the present invention can control the high performance speed of the permanent magnet synchronous motor system, that is, control the speed of the motor to rapidly and accurately reach a given value.

(35) (V). Determining control gains of the permanent magnet synchronous motor speed controller:

(36) next, the present invention will give a control gains selection method of the designed controller to ensure that the current of the motor is always within a given safety range during motor operation, i.e. satisfies
|i.sub.q(t)|<I.sub.max,∀t≥0  (7)
further, according to the expression (27) of the controller i.sub.g(t), the following relation is required to hold:

(37) $\begin{array}{cc}.Math.i_q.Math.=.Math.\frac{J_o}{K_{to}}\left[f+\frac{B_o}{J_o}{w}^{*}+k_p\psi \left(e\right)+k_i\psi \left(E\right)+k_l\mathrm{sign}\left(e\right)\right].Math.\le I_{\max }& \left(39\right)\end{array}$
according to equations (4) and (21), it can be obtained that:
|ƒ|≤l.sub.1+μ  (40)
meanwhile, the signum function satisfies |sign(•)|≤1, and according to the property |Ψ(•)|≤1 of the saturation function in equation (23), it can be known that if equation (39) holds, the control gains of the proposed controller shall be determined according to the following method:

(38) $\begin{array}{cc}{k}_p+k_i+k_l\le \frac{K_{\mathrm{to}}}{J_o}{I}_{\max }-l_1-\mu -\frac{B_o}{J_o}{w}^{*}& \left(41\right)\end{array}$
in addition, in combination with equation (28), it can be known that the complete control gains selection condition of the proposed permanent magnet synchronous motor speed controller is

(39) $\begin{array}{cc}\{\begin{array}{c}{k}_p+k_i+k_l\le \frac{K_{\mathrm{to}}}{J_o}{I}_{\max }-l_1-\mu -\frac{B_o}{J_o}{w}^{*}\\ k_l>\mu \end{array}& \left(42\right)\end{array}$
(VI) Achieving a control method:

(40) the speed IN of the motor is measured by a speed sensor installed in the permanent magnet synchronous motor system, the control method (27) is used as a speed loop controller to control the motor, so the speed of the motor can be accurately tracked, and the current of the motor can be constrained within a given range, and the disturbance influence can be effectively suppressed to achieve the control objective of the speed-governing system of the permanent magnet synchronous motor.

(41) Simulation Result Description:

(42) In order to verify the performance of the speed control technology of the permanent magnet synchronous motor proposed by the present invention, the results of simulation comparison between the proposed control method and the traditional PI control method are given in this part. In simulation, the values of system parameters are selected as follows:
J.sub.o=0.089 kg.Math.m.sup.2,B.sub.o=0.005N.Math.m.Math.s/rad,K.sub.to=6.219N.Math.m/A,
L=7.8×10.sup.−3 H,R=0.346Ω,Ψ.sub.ƒ=0.51825Wb,n.sub.p=2

(43) It is noteworthy that J.sub.o, B.sub.o, K.sub.to in the above equation are nominal values of the system, in actual simulation, in order to simulate system parameters uncertainty, the actual system parameters are adjusted as:
J=1.2J.sub.o,B=2B.sub.o,K.sub.t=K.sub.to

(44) Meanwhile, in order to test the control performance of the method proposed by the present invention in the face of unknown load torque disturbance, the suddenly added load torque T.sub.L=5.5N.Math.m at 0.8 s and the suddenly reduced load at is are simulated. In addition, the maximum allowable current during motor operation is set to I.sub.max=10 A.

(45) The double closed loop vector control framework of the permanent magnet synchronous motor shown in FIG. 2 is adopted in the simulation. The control gains settings of the proposed control method and the traditional PI control method are given below:

(46) 1) The traditional PI control method:

(47) the control gains of the traditional PI controller adopted in the speed loop are set to:
k.sub.ps=11,k.sub.is=0.5.
2) The method proposed by the present invention:
the control gains of the speed controller in (27) designed by the present invention, the saturation function in (23) and the super-twisting disturbance observer in (13) adopted in the speed loop are set to:
k.sub.p=480,k.sub.i=27,k.sub.l=0.01,n=75,
λ=900,k.sub.1=11,k.sub.2=3.5

(48) FIG. 3 and FIG. 4 respectively show the simulation results of the method proposed by the present invention and the traditional PI control method. The given speed of the system is set to w*=1000 r/min. By comparing FIG. 3 and FIG. 4, it can be known that the method proposed by the present invention has a faster response speed and higher tracking accuracy as compared with the traditional PI control method. When the load torque disturbance is suddenly added or suddenly reduced, by estimating and compensating the disturbance by means of the designed super-twisting disturbance observer, the method proposed by the present invention can make the system have strong anti-disturbance performance, thus the speed tracking performance of the motor is almost not affected by disturbance. While for the compared traditional PI control method, the speed of the motor fluctuates obviously when the load torque disturbance is suddenly added and suddenly reduced. In addition, it is more noteworthy that the method proposed by the present invention can ensure that the current of the motor is always within a given range during the whole motor operation process, while for the PI control method, the initial current reaches 16 A at the motor starting stage, which far exceeds the given upper limit of the current.

(49) In summary, the method of the present invention can suppress disturbance influence, effectively realize the rapid tracking of the speed of the motor, meanwhile, compared with the prior art, this technology can ensure that the current of the motor is always within a given range, plays a role of prevention of current saturation and safety protection, and can be applied to an actual system.