CN1283041C - Method or controlling permanent magnet synchronous motor-air conditioner compressor system without speed sensor - Google Patents

Abstract

The present invention relates to a control method of a permanent magnet synchronous motor namely an air-condition compressor system of a speed-less sensor, which belongs to the technical field of frequency conversion air conditioners. The present invention is characterized in that the vector control of the speed-less sensor is used for the permanent magnet synchronous motor namely the air condition compressor system, and the defect of large rotational speed pulsation is overcome. Simultaneously, a composite control method of torque moment instruction current is used for further decreasing the pulsation of motor speed. The control method effectively overcomes contradiction between system stability and response speed appearing in the process that the traditional vector control system sets the parameters of a PI adjustor, and the control method also enhances the control performance of the permanent magnet synchronous motor in the frequency conversion air conditioners.

Description

The control method of Speedless sensor permagnetic synchronous motor-air conditioner compressor system

Technical field

The invention belongs to electric machines control technology, especially the technical field of in convertible frequency air-conditioner, using of permagnetic synchronous motor.

Background technology

Along with the arrival of worldwide energy crisis, national governments are all promoting energy-saving and cost-reducing technology energetically for the purpose of sustainable economic development.As one of capital equipment of household electricity, traditional fixed frequency air conditioner device since its operational efficiency lowly withdraw from the market gradually.The transducer air conditioning of a new generation because have that energy-saving effect is obvious, adjustment is steady, series of advantages such as running noises is low in the whole frequency range, thereby receive the concern in market.

Along with permanent magnetic material performance improve constantly and perfect, and magneto research and development experience is progressively ripe, magneto obtains application more and more widely at aspects such as national defence, industrial and agricultural production and daily lifes, forward high-power, multifunction and the development of microminiaturized aspect.That permanent magnet motor has is simple in structure, volume is little, in light weight, characteristics such as loss is little, efficient height, is subjected to the attention of industrial quarters gradually.

Traditionally, according to the difference of the geometry of permanent magnet motor rotor magnetic steel, can be divided into permagnetic synchronous motor (Permanent-Magnet Synchronous Motor, PMSM) and brshless DC motor (Brush-less DC Motor, BLDCM) two kinds.Both have a lot of similarities, and difference maximum between them is: when rotor rotates, and the back emf waveform difference that on stator, produces, the back-emf of permanent magnet synchronous motor is sinusoidal wave, and brushless DC motor is a trapezoidal wave.Therefore, two kinds of motor are all different on principle, model and control method.What use in a large number on air-conditioning at present is the control system of brshless DC motor.

The vector control system of permagnetic synchronous motor can be realized high accuracy, high dynamic performance, speed and Position Control on a large scale.Compare with direct current machine, it does not have mechanical commutator and brush; Compare with asynchronous motor, it does not need idle exciting current, thereby the power factor height, and the stator resistance loss is little, and rotor parameter can be surveyed, control performance is good.At present, the servo system of being made up of permagnetic synchronous motor has been widely used in fields such as flexible manufacturing system, robot, office automation, Digit Control Machine Tool.

For the compressor load in the air-conditioning, in the process that rotor whenever rotates a circle, because the variation of compresser cylinder internal pressure, the load torque of motor can periodic fluctuation.In the control system of traditional employing polyphase brushless dc motor, the brushless DC motor stator electric current is a square wave, whenever opens 120 ° of electrical degrees mutually, turn-offs 60 ° of electrical degrees then.Per 60 ° of electrical degrees have a switch to change state, so the brushless DC motor rotor position detector is every pulse of 60 ° of electrical degree output.General polyphase brushless dc motor its rotor position angle in each electrical degree of 60 ° is unknown, so polyphase brushless dc motor also exports the mode approximate load torque of six torque reference values with six ladders accordingly, and can not follow the variation of load torque accurately.Fig. 1 is compressor load and polyphase brushless dc motor method for controlling torque schematic diagram, and visible setpoint torque can not well be followed the tracks of the compressor load torque.So just inevitably produce the big fluctuation of speed, influenced the performance of compressor operating.

The permanent-magnetic synchronous motor stator electric current is sinusoidal wave, needs detection rotor position continuously.The vector control system of permagnetic synchronous motor can be realized high accuracy, high dynamic performance, and on a large scale speed and Position Control, thus reach the effect of the high performance control in the convertible frequency air-conditioner.Adopt permanent magnet synchronous motor vector control system can replace initial polyphase brushless dc motor, reduce the fluctuation of rotating speed, realize high performance control to follow the tracks of load torque fast and accurately.Simultaneously, because can't installation rate on the compressor and the mechanical pick-up device of position, use Speedless sensor method realizes being controlled to for a kind of novel compressor of air conditioner control mode of high-performance compressor load.

Permanent magnet synchronous motor vector control system

1971, the vector control method of the alternating current machine that people such as German Blaschke propose solved the high performance control problem of alternating current motor torque theoretically.The basic thought of vector control comes from the strictness simulation to direct current machine.Direct current machine itself has good decoupling, and it can be respectively by controlling the purpose that its armature supply and exciting current reach the control motor torque.The final purpose of vector control is to improve the torque control performance of motor, and implements still to implement in the control to stator current.Vector control is divided into excitation component and torque component by the motor-field orientation with stator current, is controlled respectively, thereby obtains good decoupling zero characteristic.This control method at first is applied on the asynchronous machine, is transplanted to synchronous machine very soon.In fact, easier realization vector control on permagnetic synchronous motor.Because this motor does not have slip frequency electric current and the control in the induction machine to be subjected to the influence of parameter (mainly being rotor parameter) little in the vector control process yet.At present, vector control technology has obtained using widely in permagnetic synchronous motor.

Speedless sensor control

The vector control of PMSM is generally controlled stator current or voltage by the position and the amplitude of detection or estimation rotor magnetic flux.Like this, the torque of motor only and magnetic flux, current related, and is similar to the control of DC method, can obtain very high control performance.In traditional Control of PMSM, for exact position and the speed that obtains rotor, general speed and the position that need on the axle of rotor, mechanical type sensor measurement motor be installed.These mechanical pick-up devices often are encoder, solver and tachogenerator.Mechanical pick-up device provides motor needed rotor signal, but has also brought some problems to governing system:

(1) mechanical pick-up device has increased the moment of inertia on the rotor axis of electric, has strengthened motor bulk and volume, and the use of mechanical pick-up device has increased connecting line and the interface circuit between motor and the control system, makes system be subject to disturb, and has reduced reliability.

(2) be subjected to the restriction of conditions such as mechanical pick-up device service condition such as temperature, humidity and vibrations, make this system can not be widely used in various occasions.

(3) mechanical pick-up device and auxiliary circuit thereof have increased the cost of governing system, the price of some high-precision sensor even can compare with the price of motor own.

In order to overcome the shortcoming of using transducer to bring to system, a lot of scholars have carried out the research of no transducer control system for permanent-magnet synchronous motor.Do not have mechanical transducer Alternating Current Governor System and be meant the relevant signal of telecommunication that utilizes in the motor windings, the position and the speed that estimate rotor by suitable method replace mechanical pick-up device, to realize the motor closed-loop control.

The control of permagnetic synchronous motor-air conditioner compressor system

For the permagnetic synchronous motor-compressor of air conditioner load system of Speedless sensor vector control, because the periodicity wide variation of load torque has caused the fluctuation of speed.In the motor operation course of the system of reality, because the variation of the operation conditions (as electric current, temperature, humidity etc.) of motor, make the parameter of motor, comprise that resistance, inductance, back emf coefficient etc. all can change, error during simultaneously owing to measurement, the parameter of true motor can't obtain accurately.In addition, in order to reduce cost, also wish to increase the length of control cycle time as far as possible.Studies show that during to the parameter of electric machine and control cycle variation: for the compressor load in the air-conditioning, in the process that rotor whenever rotates a circle, because the variation of compresser cylinder internal pressure, the load torque of motor can periodically fluctuate, the identification of rotating speed and position and the runnability of system very easily have been subjected to the parameter variation, when similar problem also occurs in system's control cycle increase.

For the problems referred to above, generally be the response of system to be met the demands in traditional vector control by the parameter of regulating the pi regulator in the control block diagram.And in the compressor load system, because the periodically pulsing of compressor load, the stability in the method for PI parameter tuning and the contradiction of response speed just show especially out: PI gains when too small, and the rotating speed that picks out is difficult to the true rotating speed of tracking; PI gain is excessive, and system is unstable and vibration occurs.By adjusting of regulator parameter, the tracking actual speed that the speed of estimation can be stable, however because the minimizing of speed regulator gain, the response speed of system also just decreases.The output of speed regulator, just the torque instruction electric current can't be followed the tracks of the variation of load torque well, causes the motor electromagnetic torque obviously to lag behind load torque.Fig. 2. in when, for q axle inductance 10% fluctuation taking place, the simulation result after the speed regulator PI parameter of adjusting.As seen, each rotor mechanical is in the cycle, and speed can't guarantee steady state value, bigger fluctuation occurred, has influenced the performance of compressor operating.

Summary of the invention

The object of the present invention is to provide a kind of control method of permagnetic synchronous motor-air conditioner compressor system of the less Speedless sensor that fluctuates.

With respect to the brushless DC motor control systems that in air-conditioning system, use in a large number at present, characteristics of the present invention are that the Speedless sensor vector control has been applied in permagnetic synchronous motor-compressor of air conditioner load system, export the variation that torque can not be followed load torque accurately thereby overcome brshless DC motor, cause the defective of bigger speed ripple.Simultaneously, according to the characteristics of compressor cycle fluctuating load torque, the present invention proposes and in permagnetic synchronous motor-compressor of air conditioner load system, reduce improving one's methods of motor speed pulsation---torque instruction electric current composite control method.This method can overcome the stability of a system that traditional vector control system occurred and the contradiction of response speed effectively in pi regulator parameter tuning process, improved Control of PMSM performance in the convertible frequency air-conditioner.

Speedless sensor permagnetic synchronous motor-air conditioner compressor system

The present invention proposes permagnetic synchronous motor-air conditioner compressor system, the control principle block diagram of its system as shown in Figure 3.In this system, the stator phase current of permagnetic synchronous motor at first is input to the position estimation link through sampling and A/D conversion, pick out the position and the speed of motor by the Speedless sensor algorithm, utilize existing vector control method then, stator current is divided into excitation component and torque component, is controlled respectively.Wherein, exciting current i _dReference value i _Dref=0, inverter adopts the method for space vector PWM modulation, and the position estimation link has adopted built-in type permagnetic synchronous motor Speedless sensor rotating speed, the position identification algorithm based on back-emf.The innovative point of native system is to adopt permanent magnet synchronous motor vector control system to replace the polyphase brushless dc motor system, and adopt the Speedless sensor method to estimate rotating speed of motor and position, thereby follow the tracks of load torque fast and accurately, reduce the fluctuation of rotating speed, realize high performance control.

Torque instruction electric current composite control method

Torque instruction electric current composite control method, the pi regulator parameter tuning problem when changing at parameter in permagnetic synchronous motor-compressor of air conditioner load system proposes.With respect to traditional vector control strategy, this method can effectively reduce speed ripple, improves the permanent magnet synchronous motor control performance in the air-conditioning system.

Owing between the torque of compressor load and the position angle comparatively strict corresponding relationship is arranged, therefore can add feedforward compensation according to the position of estimation to torque current.The i that native system adopts _d=0 control method, thereby when stable state, can get:

T _em＝p _n(i _qψ _d-i _dψ _q)＝p _n(i _qψ _r+(L _dx-L _q)i _di _q)＝p _ni _qψ _r

In the formula, T _EmBe electromagnetic torque, p _nBe number of pole-pairs, i _d, i _qBe stator winding d-q shaft current, ψ _d, ψ _qBe d-q axle magnetic linkage, ψ _rBe rotor permanent magnetism magnetic linkage, L _d, L _qBe d-q axle inductance.

So, according to the corresponding relation between compressor torque and the motor rotor position, form position-torque indicator.At each control cycle, table look-up and find the level of torque of load according to the position of estimation, also just obtained the electromagnetic torque size that needs, directly calculate the torque current size that needs thus and be:

${i_{q\_\mathrm{ref}}}^{\′}=\frac{T_{\mathrm{em}}}{p_n{\mathrm{\ψ}}_r}$

Fig. 4. be torque instruction electric current composite control method theory diagram.As shown in the figure, this method with the estimation angle introduced the torque current controlling unit, by compressor load torque reference table to a torque current i _{Q_ref}', again with the torque current i that calculates _{Q_ref}' with the output sum of speed pi regulator as torque instruction current value (i _{Q_ref}+ i _{Q_ref}'), other parts of system are identical with existing vector control system.Fig. 3. for improved system after having introduced torque instruction electric current composite control method controls block diagram.

The invention is characterized in that it contains following steps successively:

(1) system initialization

Digital signal processor in system is compressor load torque-rotor-position curve table that DSP input user provides;

In DSP, set:

The reference value ω of rotating speed _Ref=0, exciting current reference value i _{D_ref}=0;

Back-emf estimation constant K _e, the position estimation constant K _θBe set point;

Control cycle T is a set point;

The ratio of setting speed adjuster, integral constant K _P1, K _I1, the ratio of torque current regulator, integral constant K _P2, K _I2, the ratio of exciting current controller, integral constant K _P3, K _I3

The parameter of customer-furnished permagnetic synchronous motor: number of pole-pairs p _n, rotor permanent magnetism magnetic linkage ψ _r, dq axle inductance L _d, L _q, stator resistance R, back emf coefficient K _ESimultaneously, the dq shaft voltage initial value of setting above-mentioned motor is zero, i.e. u _d(0)=0, u _q(0)=0;

(2) DSP detects the dq shaft voltage of stator three-phase current and calculating motor

DSP detects current transformer, filter capacitor, the next stator three-phase current i of A/D converter of the above-mentioned motor stator side of process successively _a, i _b, i _c

U for n 〉=1 _d, u _qGet actual that a last digital control cycle T calculates among the DSP with reference to dq shaft voltage, i.e. u _d(n)=u _{D_ref}(n-1), u _q(n)=uq_ref (n-1);

(3) the DSP employing comes the rotating speed of identifying motor in the n cycle based on rotating speed, the position identifying method of the built-in type permagnetic synchronous motor Speedless sensor of back-emf

With position θ _M(n);

Set: have an identical molded motor of parameter to be based upon on the γ δ axle, described γ δ axle is the armature spindle of estimation, and the angle error of it and dq axle is Δ θ, Δ θ=θ-θ _M, θ is the angle of d axle and reference axis+A, θ _MAngle for γ axle and reference axis+A;

If initial time, rotating speed $\stackrel{\·}{\mathrm{\θ}}\left(0\right)=0,$ Position θ _M(0)=0,

Then:

${\mathrm{\θ}}_M\left(n\right)={\mathrm{\θ}}_M(n-1)+\frac{T}{K_E}{e}_M\left(n\right)+K_{\mathrm{\θ}}\mathrm{sgn}\left\{{\stackrel{\·}{\mathrm{\θ}}}_M(n-1)\right\}\mathrm{\Δ}{i}_{\mathrm{\γ}}\left(n\right)$

e _M(n)＝e _M(n-1)-K _eΔi _δ(n)

$\left[\begin{array}{c}\mathrm{\Δ}{i}_{\mathrm{\γ}}\left(n\right)\\ \mathrm{\Δ}{i}_{\mathrm{\δ}}\left(n\right)\end{array}\right]=\left[\begin{array}{c}{i}_{\mathrm{\γ}}\left(n\right)\\ i_{\mathrm{\δ}}\left(n\right)\end{array}\right]-\left[\begin{array}{c}{i}_{\mathrm{M\γ}}\left(n\right)\\ i_{\mathrm{M\δ}}\left(n\right)\end{array}\right]$

$\left[\begin{array}{c}{i}_{\mathrm{\γ}}\left(n\right)\\ i_{\mathrm{\δ}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\·}{\mathrm{\θ}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\stackrel{\·}{\mathrm{\θ}}}_M(n-1)\mathrm{\θ}\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\γ}}(n-1)\\ i_{\mathrm{\δ}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\γ}}(n-1)\\ L_d{u}_{\mathrm{\δ}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}e\left[\begin{array}{c}{L}_q\sin \left(\mathrm{\Δ\θ}\right)\\ -L_d\cos \left(\mathrm{\Δ\θ}\right)\end{array}\right]$

$\left[\begin{array}{c}{i}_{\mathrm{M\&gamma;}}\left(n\right)\\ i_{\mathrm{M\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}(n-1)\\ i_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\&gamma;}}(n-1)\\ L_d{u}_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_q}{e}_{\mathrm{<figure-callout\; id="0"\; label="M"\; filenames="C20041007814100192.png,C20041007814100232.png"\; state="\{\{state\}\}">M</figure-callout>}}\left[\begin{array}{c}0\\ 1\end{array}\right]$

$\left[\begin{array}{c}{u}_{\mathrm{\&gamma;}}\left(n\right)\\ u_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}R+{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&gamma;}}& -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&delta;}}-p{L}_{\mathrm{\&gamma;\&delta;}}\\ {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;}}-p{L}_{\mathrm{\&gamma;\&delta;}}& R-{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&delta;}}\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]+e\left[\begin{array}{c}-\sin \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\\ \cos \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right]$

Wherein,

Δθ(n)＝θ(n)-θ _M(n)

$\left.\begin{array}{c}{L}_{\mathrm{\&gamma;}}=\frac{1}{2}\{(L_d+L_q)-(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&delta;}}=\frac{1}{2}\{(L_d+L_q)+(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&gamma;\&delta;}}=\frac{1}{2}(L_q-L_d)\sin \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right\}$

P is a differential operator;

When estimation reaches stable state, Δ θ ≈ 0, so formula L _γ, L _δ, L _{γ δ}Be approximately equal to L respectively _d, L _q, 0, described L _γ, L _δ, L _{γ δ}Be respectively

The self-induction of γ axle, δ axle self-induction and γ, the mutual inductance of δ between centers;

Wherein,

N is the time cycle number after the discretization,

θ _MBe the rotor position angle of estimation,

e _MBe the back-emf of molded motor,

E is the back-emf of real electrical machinery,

ω is a rotor speed, promptly $\mathrm{\&omega;}={\mathrm{\&theta;}}_M^g,$

K _eBe back-emf estimation constant, K _θBe that the position estimation constant is set point;

u _γ, u _δFor real electrical machinery at γ, the voltage on the δ axle,

i _γ(n-1), i _δ(n-1) be true motor at γ, the current response on the δ axle,

i _{M γ}(n), i _{M δ}(n) be with γ, the δ axle is the current response of the molded motor of armature spindle foundation,

Δ i _γ(n), Δ _{I δ}(n) be in each cycle real current and molded motor at γ, the difference between currents on the δ axle;

(4) DSP adopts torque instruction electric current composite control method to calculate dq shaft voltage reference value, and it contains following steps successively;

(4.1) with the current i under the abc coordinate system _a, i _b, i _cMultiply by transformation matrix of coordinates and obtain current value i under the dq axle _d, i _q, i ₀:

$\left[\begin{array}{c}{i}_d\left(n\right)\\ i_q\left(n\right)\\ i_0\left(n\right)\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \left(\mathrm{\&theta;}\left(n\right)\right)& \cos (\mathrm{\&theta;}\left(n\right)-\frac{2}{3}\mathrm{\&pi;})& \cos (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \sin \left(\mathrm{\&theta;}\left(n\right)\right)& \sin (\mathrm{\&theta;}\left(n\right)-\frac{2}{3}\mathrm{\&pi;})& \sin (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]\left[\begin{array}{c}{i}_a\left(n\right)\\ i_b\left(n\right)\\ i_c\left(n\right)\end{array}\right]$

(4.2) DSP is the reference value ω of rotating speed _RefThe estimation rotating speed that obtains with step (3) $\mathrm{\&omega;}\left(n\right)={\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)$ Difference Δ ω (n)=ω _Ref-ω (n) input speed adjuster is tried to achieve q axle reference current i _{Q_ref}(n), $i_{q\_\mathrm{ref}}\left(n\right)=K_{p1}\mathrm{\&Delta;\&omega;}\left(n\right)+K_{i1}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;\&omega;}\left(i\right)T,$ Wherein, K _P1, K _I1Ratio, integral constant for speed regulator;

(4.3) rotor position that estimation obtains according to step 3 _M(n), in each control cycle, look into rotor-position-load torque curve table according to the position of estimation, the electromagnetic torque of motor equals load torque simultaneously, therefore obtains electromagnetic torque T _Em, calculate the torque current i that needs thus _{Q_ref}':

${i_{q-\mathrm{ref}}}^{\&prime;}=\frac{T_{\mathrm{em}}}{p_n{\mathrm{\&psi;}}_r};$

(4.4) DSP is i _{Q_ref}'+i _{Q_ref}With actual q shaft current i _qDifference DELTA i _q(n)=i _{Q_ref}' (n)+i _{Q_ref}(n)-i _q(n) input torque current regulator obtains q axle reference voltage u _{Q_ref}(n),

$u_{q\_\mathrm{ref}}\left(n\right)=k_{p2}\mathrm{\&Delta;}{i}_q\left(n\right)+k_{i2}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{i}_q\left(i\right)T$

Wherein, K _P2, K _I2Ratio, integral constant for torque current regulator;

(4.5) DSP is i _{D_ref}=0 and i _dDifference DELTA i _d(n)=i _{D_ref}(n)-i _d(n) the input exciting current controller obtains d axle reference voltage u _{D_ref}(n), $u_{d\_\mathrm{ref}}\left(n\right)=K_{p3}\mathrm{\&Delta;}{i}_d\left(n\right)+K_{i3}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{i}_d\left(i\right)T$ , wherein, K _P3, K _I3Ratio, integral constant for exciting current controller;

(5) DSP is according to u _{D_ref}, u _{Q_ref}Calculate the voltage vector under α β 0 coordinate system,

$\left[\begin{array}{c}{u}_{\mathrm{\&alpha;}}\left(n\right)\\ u_{\mathrm{\&beta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}\cos \mathrm{\&theta;}\left(n\right)& -\sin \mathrm{\&theta;}\left(n\right)\\ \sin \mathrm{\&theta;}\left(n\right)& \cos \mathrm{\&theta;}\left(n\right)\end{array}\right]\left[\begin{array}{c}{u}_{d\_\mathrm{ref}}\left(n\right)\\ u_{q\_\mathrm{ref}}\left(n\right)\end{array}\right]$

(6) DSP obtains the switching signal of inverter with the space vector PWM method, obtains opening the turn-off time of each switching tube of inverter, sends the output voltage that corresponding pwm pulse removes control inverter, realizes Control of PMSM:

In each control cycle T, each switching tube is opened turn-off time T in three brachium pontis of inverter _a, T _b, T ₀:

T ₀＝T-T _a-T _b

$V_s=\sqrt{{u_{\mathrm{\&alpha;}}}^2+{u_{\mathrm{\&beta;}}}^2},\mathrm{\&gamma;}={\tan }^{-1}\frac{u_{\mathrm{\&alpha;}}}{u_{\mathrm{\&beta;}}}$

The space that voltage vector distributes is divided into 6 zones, voltage u in each control cycle _α, u _βSynthetic space vector amplitude under α β axle is V _s, argument is γ; And the action time of switch vector by each the zone in by V _sT, V _aT _a, V _bT _bThe triangle decision that constitutes, wherein V _a=V _b=E _Dc, T is a control cycle, E _DcBe DC bus-bar voltage; γ has determined that also the switch vector is the zone at space vector place simultaneously, i.e. the current residing running status of inverter.

Application of torque instruction current composite control method, owing to added the feedfoward control of torque instruction electric current, according to the rotor estimated position, the required torque current of wide variation torque can precompute, simultaneously traditional PI controller can also be regulated the rate signal of feedback, and the electromagnetic torque size that obtains like this is with regard to the goodish load torque of having followed the trail of.Fig. 9. for 10% fluctuation takes place q axle inductance, control cycle increases to 400us from 100us, system emulation result behind the employing torque instruction electric current composite control method.As seen, the fluctuation of speed significantly reduces (in ± 5%) during the motor stable state, system's estimating algorithm is not subjected to the influence of parameter and can continues stable carrying out, thereby has proved the validity of this method Control of PMSM aspect of performance in improving convertible frequency air-conditioner.

Description of drawings

Fig. 1. compressor load and polyphase brushless dc motor method for controlling torque schematic diagram;

When 10% fluctuation takes place in Fig. 2 .q axle inductance, the simulation result after the speed regulator PI parameter of adjusting;

Fig. 3. permagnetic synchronous motor-compressor of air conditioner control system structure chart;

Fig. 4. torque instruction electric current composite control method;

Fig. 5 real electrical machinery d-q axle and molded motor γ-δ axle;

Fig. 6 software systems control flow;

Fig. 7. the computed in software flow chart;

Fig. 8. space vector PWM method schematic diagram figure: a. contravarianter voltage vector and sector schematic diagram; The b vector synthesizes schematic diagram;

10% fluctuation takes place in Fig. 9 .q axle inductance, and control cycle increases to 400us from 100us, system emulation result behind the employing torque instruction electric current composite control method;

Figure 10. the hardware system structure block diagram.

Symbol and variable declaration are as follows among the figure:

The PMSM permagnetic synchronous motor;

The rotor position angle of θ estimation;

The motor speed of ω estimation;

ω _RefThe motor speed of setting;

i _dThe d shaft current of motor under the dq0 coordinate system, i.e. exciting current;

i _qThe q shaft current of motor under the dq0 coordinate system, i.e. torque current;

i _{Q_ref}The torque reference electric current;

i _{Q_ref}' torque reference the electric current that obtains through torque instruction electric current composite control method;

i _{D_ref}The exciting current reference value;

u _{D_ref}, u _{Q_ref}Be respectively d axle reference voltage and the q axle reference voltage of motor under the dq0 coordinate system;

Abc-dq abc coordinate is tied to the modular converter under the dq0 coordinate system;

Dq-α β dq0 coordinate is tied to the modular converter under α β 0 coordinate system;

SVPWM is a space vector of voltage PWM modulation module, and PI is a proportional integral device module;

u _d, u _qStator winding d-q shaft voltage;

i _d, iq stator winding d-q shaft current;

ψ _d, ψ _qD-q axle magnetic linkage;

ψ _rRotor permanent magnetism magnetic linkage;

R stator armature resistance;

L _d, L _qD-q axle inductance;

T _Em, T _LElectromagnetic torque and load torque;

p _nNumber of pole-pairs;

The J rotor moment of inertia

The p differential operator

ω rotor speed (also is designated as

θ is the estimation rotor position angle);

e _MIt is the back-emf of molded motor;

E is the back-emf of real electrical machinery.

Motor d, q, 0, α, β, the introduction of 0 coordinate system:

D, q, 0 coordinate system is to be placed on epitrochanterian coordinate system of rotating with rotor, and the d axle is on the direction of rotor longitudinal axis, and the q axle takes the lead 90 ° of electrical degrees of d axle, and 0 is in order to guarantee the reversible imaginary axis of conversion.α, β, 0 coordinate system is the static coordinate system that is based upon on the stator.α, β, 0 can to regard rotating speed as be zero d, q, 0 coordinate system.

D, q, 0 coordinate system and a, b, the transformational relation of the voltage of c coordinate system, electric current, magnetic linkage is:

$\left[\begin{array}{c}{i}_d\\ i_{\mathrm{<figure-callout\; id="0"\; label="q\; i"\; filenames="C20041007814100192.png,C20041007814100232.png"\; state="\{\{state\}\}">q</figure-callout>}}& \\ i_{}{}_0\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \left(\mathrm{\&theta;}\right)& \cos (\mathrm{\&theta;}-\frac{2}{3}\mathrm{\&pi;})& \cos (\mathrm{\&theta;}+\frac{2}{3}\mathrm{\&pi;})\\ \sin \left(\mathrm{\&theta;}\right)& \sin (\mathrm{\&theta;}-\frac{2}{3}\mathrm{\&pi;})& \sin (\mathrm{\&theta;}+\frac{2}{3}\mathrm{\&pi;})\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]\left[\begin{array}{c}{i}_a\\ i_b\\ i_c\end{array}\right]$

Wherein, θ is a rotor position angle.

α, β, 0 coordinate system and a, b, the transformational relation of the voltage of c coordinate system, electric current, magnetic linkage is: $\left[\begin{array}{c}{i}_{\mathrm{\&alpha;}}\\ i_{\mathrm{\&beta;}}\\ i_0\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \left(0\mathrm{\&theta;}\right)& \cos \left(\frac{2}{3}\mathrm{\&pi;}\right)& \cos \left(\frac{2}{3}\mathrm{\&pi;}\right)\\ \sin \left(0\mathrm{\&theta;}\right)& \sin (-\frac{2}{3}\mathrm{\&pi;})& \sin \left(\frac{2}{3}\mathrm{\&pi;}\right)\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]\left[\begin{array}{c}{i}_a\\ i_b\\ i_c\end{array}\right].$

Embodiment

Figure 10 is a hardware system structure block diagram of the present invention.The Electric Machine Control development system PE-Expert of the Japanese Myway of experiment hardware using of the present invention company, this platform has utilized the dsp chip TMS320C32 of TI company, adopts the C Programming with Pascal Language.Hardware system is mainly by PC, dsp board, A/D, and D/A converter, PWM generator and the two level voltage type inverters that aim at the design of alternating current machine vector control are formed.System of the present invention is by the electric current in sensor PMSM stator loop, the DC bus-bar voltage of inverter, utilize development system PE-Expert to carry out the AD conversion, and in its DSP, carry out the Speedless sensor vector control, and module such as related coordinate transform in Fig. 6, PI adjusting.Utilize space vector modulating method to form pwm pulse, control inverter, thus realization is to the high performance control of the permagnetic synchronous motor in the convertible frequency air-conditioner.

System of the present invention control flow can be divided into following step as shown in Figure 6:

1. at first to the software systems initialization, set the reference value ω of rotating speed _Ref, the turn count constant K _e, K _θ, control cycle T, the ratio of each adjuster, integral constant K _p, K _i, compressor load torque-position curve table that the input user provides.The initialization parameter of electric machine: number of pole-pairs p _n, rotor permanent magnetism magnetic linkage ψ _r, d-q axle inductance L _d, L _q, stator resistance R.

2. detect the dq shaft voltage of stator three-phase current and calculating motor.

In the system of reality, stator current can pass through current sensor senses, and by carrying out A/D conversion supplied with digital signal processor (DSP) utilization after the filtering conditioning.The dq shaft voltage initial value of motor is zero, i.e. u _d(0)=0, u _q(0)=0; U for n 〉=1 _d, u _qFor a last digital control cycle among the DSP calculate actual in dq shaft voltage, i.e. u _d(n)=u _{D_ref}(n-1), u _q(n)=u _{Q_ref}(n-1), n 〉=1.

3. pass through the rotating speed and the position of position estimation link identifying motor.

System of the present invention has adopted a kind of built-in type permagnetic synchronous motor Speedless sensor rotating speed, position identification algorithm based on back-emf.

Because rotor position angle θ is unknown, so the accurate d-q axle of motor also can't obtain, and all estimating algorithms can only be based on the angular position theta of estimation _M, on the armature spindle of just supposing to have an identical molded motor of parameter to be based upon estimation (being made as γ-δ axle), as shown in Figure 5, initial time θ=0, ${\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M=0.$

Because the existence of rotor angle error delta θ, then the equation of real electrical machinery on γ-δ axle is changed to:

$\left[\begin{array}{c}{u}_{\mathrm{\&gamma;}}\left(n\right)\\ u_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}R+{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&gamma;}}& -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&delta;}}-p{L}_{\mathrm{\&gamma;\&delta;}}\\ {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;}}-p{L}_{\mathrm{\&gamma;\&delta;}}& R-{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&delta;}}\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]+e\left[\begin{array}{c}-\sin \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\\ \cos \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right]$

Wherein,

Δθ(n)＝θ(n)-θ _M(n)

$\left.\begin{array}{c}{L}_{\mathrm{\&gamma;}}=\frac{1}{2}\{(L_d+L_q)-(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&delta;}}=\frac{1}{2}\{(L_d+L_q)+(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&gamma;\&delta;}}=\frac{1}{2}(L_q-L_d)\sin \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right\}$

When estimation reached stable state, Δ θ approached 0, so the L in the formula (3) _γ, L _δ, L _{γ δ}Be approximately equal to L respectively _d, L _q, 0.

In digital control, need to use the discrete model of motor, the note control cycle is T, then releases the current response of the true motor of each control cycle on γ-δ axle to be:

$\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}(n-1)\\ i_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\&gamma;}}(n-1)\\ L_d{u}_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}e\left[\begin{array}{c}{L}_q\sin \left(\mathrm{\&Delta;\&theta;}\right)\\ -L_d\cos \left(\mathrm{\&Delta;\&theta;}\right)\end{array}\right]$

Simultaneously, be that the molded motor current response that armature spindle is set up is with γ-δ axle,

$\left[\begin{array}{c}{i}_{\mathrm{M\&gamma;}}\left(n\right)\\ i_{\mathrm{M\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}(n-1)\\ i_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\&gamma;}}(n-1)\\ L_d{u}_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_{\mathrm{<figure-callout\; id="0"\; label="q\; e\; M"\; filenames="C20041007814100192.png,C20041007814100232.png"\; state="\{\{state\}\}">q</figure-callout>}}}{e}_M{}_{}\left[\begin{array}{c}0\\ 1\end{array}\right]$

Thus, can obtain that the difference between currents of true motor and molded motor is in each cycle,

$\left[\begin{array}{c}\mathrm{\&Delta;}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ \mathrm{\&Delta;}{i}_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]-\left[\begin{array}{c}{i}_{\mathrm{M\&gamma;}}\left(n\right)\\ i_{\mathrm{M\&delta;}}\left(n\right)\end{array}\right]$

At each control cycle, given constant K _eAnd K _θ, back-emf is carried out recursion, and rotor position angle is proofreaied and correct, as follows.

e _M(n)＝e _M(n-1)-K _eΔi _δ(n)

${\mathrm{\&theta;}}_M\left(n\right)={\mathrm{\&theta;}}_M\left(n-1\right)+\frac{T}{K_E}{e}_M\left(n\right)+K_{\mathrm{\&theta;}}\mathrm{sgn}\left\{{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\right\}\mathrm{\&Delta;}{i}_{\mathrm{\&gamma;}}\left(n\right)$

Wherein

Motor speed can be estimated by following formula:

${\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)=\frac{1}{T}\{{\mathrm{\&theta;}}_M\left(n\right)-{\mathrm{\&theta;}}_M(n-1)\}=\frac{e_M\left(n\right)}{K_E}+\frac{K_{\mathrm{\&theta;}}}{T}\mathrm{sgn}\left\{{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\right\}\mathrm{\&Delta;}{i}_{\mathrm{\&gamma;}}\left(n\right)$

4. calculate dq shaft voltage reference value by torque instruction electric current composite control method.

The rotating speed and the position that obtain estimating according to step 3 are with the current i under the abc coordinate system _a, i _b, i _cMultiply by transformation matrix of coordinates and obtain current value i under the dq axle _d, i _q, i ₀

$\left[\begin{array}{c}{i}_d\left(n\right)\\ i_q\left(n\right)\\ i_0\left(n\right)\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \left(\mathrm{\&theta;}\left(n\right)\right)& \cos (\mathrm{\&theta;}\left(n\right)-\frac{2}{3}\mathrm{\&pi;})& \cos (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \sin \left(\mathrm{\&theta;}\left(n\right)\right)& \sin (\mathrm{\&theta;}\left(n\right)-\frac{2}{3}\mathrm{\&pi;})& \sin (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]\left[\begin{array}{c}{i}_a\left(n\right)\\ i_b\left(n\right)\\ i_c\left(n\right)\end{array}\right]$

In native system, electric current, speed regulator have all adopted typical amplitude limit to add the form of PI (proportional integral device) adjuster.

$y\left(n\right)=K_{p}e\left(n\right)+K_i\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}e\left(i\right)T$

E (n) is the adjuster input, and y (n) is adjuster output, K _p, K _iBe respectively ratio, integral constant.

The reference value ω of rotating speed _RefDifference DELTA ω (n)=ω with the estimation rotating speed _Ref-ω (n) input speed adjuster, the passing ratio integral operation obtains q axle reference current i _{Q_ref}(n), $i_{q\_\mathrm{ref}}\left(n\right)=K_{p1}\mathrm{\&Delta;\&omega;}\left(n\right)+K_{i1}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;\&omega;}\left(i\right)T.$

The rotor-position that is obtained by step 3 estimation in each control cycle, is tabled look-up according to the position of estimation and to be found the level of torque of load, has also just obtained the electromagnetic torque size that needs, and directly calculates the torque current size that needs thus and is:

${i_{q\_\mathrm{ref}}}^{\&prime;}=\frac{T_{\mathrm{em}}}{p_n{\mathrm{\&psi;}}_r}$

With i _{Q_ref}'+i _{Q_ref}With i _qDifference DELTA i _q(n)=i _{Q_ref}' (n)+i _{Q_ref}(n)-i _q(n) input torque current regulator, passing ratio integral operation are output as q axle reference voltage u _{Q_ref}(n), promptly $u_{q\_\mathrm{ref}}\left(n\right)=K_{p2}\mathrm{\&Delta;}{i}_q\left(n\right)+K_{i2}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{i}_q\left(i\right)T.$ i _{D_ref}=0 and i _dDifference DELTA i _d(n)=i _{D_ref}(n)-i _d(n) input exciting current controller, the passing ratio integral operation is output as d axle reference voltage u _{D_ref}(n), $u_{d\_\mathrm{ref}}\left(n\right)=K_{p3}\mathrm{\&Delta;}{i}_d\left(n\right)+K_{i3}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{i}_d\left(i\right)T.$

5. calculate the voltage vector under the α β axle.Be tied to the transformational relation of α β 0 coordinate system by the dq0 coordinate, according to u _{D_ref}, u _{Q_ref}Calculate the voltage vector under α β 0 coordinate system.

$\left[\begin{array}{c}{u}_{\mathrm{\&alpha;}}\left(n\right)\\ u_{\mathrm{\&beta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}\cos \mathrm{\&theta;}\left(n\right)& -\sin \mathrm{\&theta;}\left(n\right)\\ \sin \mathrm{\&theta;}\left(n\right)& \cos \mathrm{\&theta;}\left(n\right)\end{array}\right]\left[\begin{array}{c}{u}_{d\_\mathrm{ref}}\left(n\right)\\ u_{q\_\mathrm{ref}}\left(n\right)\end{array}\right]$

6. utilize the space vector PWM method to obtain the switching signal of inverter.

The on off state of power device is used three on off state S respectively in three brachium pontis of inverter _A, S _B, S _CExpression.If S _i(i=A, B C)=1 represent inverter A, B, the power device conducting of C arm top, S _i(i=A, B C)=0 represent inverter A, B, the power device conducting of C brachium pontis bottom.According to S _A, S _B, S _CDifferent conditions combination, inverter has 8 running statuses.8 corresponding running statuses have 8 space vector of voltage V0～V7, and wherein V1～V6 is non-zero vector, and the amplitude size is E _Dc(DC bus-bar voltage), V0, V7 are zero vector, amplitude is 0.The spatial distribution of V0～V7 is shown in Fig. 8 .a.

The voltage vector distribution space is divided into 6 zones, in each control cycle according to voltage u _α, u _β(amplitude is to obtain synthetic space vector under the α β axle $V_s=\sqrt{{u_{\mathrm{\&alpha;}}}^2+{u_{\mathrm{\&beta;}}}^2},$ Argument is $\mathrm{\&gamma;},(\tan \mathrm{\&gamma;}=\frac{u_{\mathrm{\&alpha;}}}{u_{\mathrm{\&beta;}}}))$ Suitable switch vector is selected in the sector at place.V in each zone _sCan utilize two non-zero vector V of zone boundary _a, V _bWith zero vector V0 or V7, by their T action time of reasonable control _a, T _b, T ₀Obtain the voltage vector of reference.Shown in Fig. 8 .b, can be in each zone by V _sT, V _aT _a, V _bT _bThe triangle that constitutes obtains T action time according to the triangle sine _a, T _b, T ₀:

T ₀＝T-T _a-T _b

In the formula, T is a control cycle, E _DcBe DC bus-bar voltage.

So just obtain opening the turn-off time of each switching tube, send the output voltage of corresponding pwm pulse control inverter, also just realized Control of PMSM.

7. judge whether to end control program., then not coming back to for the 2nd step begins to carry out; Be that then the record data display waveform is ended control program.

The computed in software flow process of said system as shown in Figure 6.

Claims (1)

1. the control method of Speedless sensor permanent magnet synchronous motor-air conditioner compressor system is characterized in that, is used for the compressor load of the periodicity wide fluctuations relevant with rotor-position, and this method contains following steps successively:

(1) system initialization

Digital signal processor in system is compressor load torque-rotor-position curve table that DSP input user provides; In DSP, set:

The reference value ω of rotating speed _Ref=0, exciting current reference value i _{D_ref}=0;

Back-emf estimation constant K _e, the position estimation constant K _θBe set point;

Control cycle T is a set point;

The ratio of setting speed adjuster, integral constant K _P1, K _I1, the ratio of torque current regulator, integral constant K _P2, K _I2, the ratio of exciting current controller, integral constant K _P3, K _I3

The parameter of customer-furnished permagnetic synchronous motor: number of pole-pairs p _n, rotor permanent magnetism magnetic linkage ψ _r, dq axle inductance L _d, L _q, stator resistance R, back emf coefficient K _ESimultaneously, the dq shaft voltage initial value of setting above-mentioned motor is zero, i.e. u _d(0)=0, u _q(0)=0;

(2) DSP detects the dq shaft voltage of stator three-phase current and calculating motor

DSP detects current transformer, filter capacitor, the next stator three-phase current i of A/D converter of the above-mentioned motor stator side of process successively _a, i _b, i _c

U for n 〉=1 _d, u _qGet actual that a last digital control cycle T calculates among the DSP with reference to dq shaft voltage, i.e. u _d(n)=u _{D_ref}(n-1), u _q(n)=u _{Q_ref}(n-1);

(3) the DSP employing comes the rotating speed of identifying motor in the n cycle based on rotating speed, the position identifying method of the built-in type permagnetic synchronous motor Speedless sensor of back-emf

With position θ _M(n);

Set: have an identical molded motor of parameter to be based upon on the γ δ axle, described γ δ axle is the armature spindle of estimation, and the angle error of it and dq axle is Δ θ, Δ θ=θ-θ _M, θ is the angle of d axle and reference axis+A, θ _MAngle for γ axle and reference axis+A;

If initial time, rotating speed $\stackrel{\&CenterDot;}{\mathrm{\&theta;}}\left(0\right)=0,$ Position θ _M(0)=0, then:

${\mathrm{\&theta;}}_M\left(n\right)={\mathrm{\&theta;}}_M(n-1)+\frac{T}{K_E}{e}_M\left(n\right)+K_{\mathrm{\&theta;}}\mathrm{sgn}\left\{{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\right\}{\mathrm{\&Delta;i}}_{\mathrm{\&gamma;}}\left(n\right)$

e _M(n)＝e _M(n-1)-K _eΔi _δ(n)

$\left[\begin{array}{c}{\mathrm{\&Delta;i}}_{\mathrm{\&gamma;}}\left(n\right)\\ {\mathrm{\&Delta;i}}_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]-\left[\begin{array}{c}{i}_{\mathrm{M\&gamma;}}\left(n\right)\\ i_{\mathrm{M\&delta;}}\left(n\right)\end{array}\right]$

$\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\mathrm{\&theta;}}_M(n-1)\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}(n-1)\\ i_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\&gamma;}}(n-1)\\ L_d{u}_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}e\left[\begin{array}{c}{L}_q\sin \left(\mathrm{\&Delta;\&theta;}\right)\\ -L_d\cos \left(\mathrm{\&Delta;\&theta;}\right)\end{array}\right]$

$\left[\begin{array}{c}{i}_{\mathrm{M\&gamma;}}\left(n\right)\\ i_{\mathrm{M\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R}{L_d}T& {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_q}{L_d}T\\ -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M(n-1)\frac{L_d}{L_q}T& 1-\frac{R}{L_q}T\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}(n-1)\\ i_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_d{L}_q}\left[\begin{array}{c}{L}_q{u}_{\mathrm{\&gamma;}}(n-1)\\ L_d{u}_{\mathrm{\&delta;}}(n-1)\end{array}\right]+\frac{T}{L_q}{e}_M\left[\begin{array}{c}0\\ 1\end{array}\right]$

$\left[\begin{array}{c}{u}_{\mathrm{\&gamma;}}\left(n\right)\\ u_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}R+{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+{\mathrm{pL}}_{\mathrm{\&gamma;}}& -{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&delta;}}-{\mathrm{pL}}_{\mathrm{\&gamma;\&delta;}}\\ {\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;}}-{\mathrm{pL}}_{\mathrm{\&gamma;\&delta;}}& R-{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)L_{\mathrm{\&gamma;\&delta;}}+{\mathrm{pL}}_{\mathrm{\&delta;}}\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\&gamma;}}\left(n\right)\\ i_{\mathrm{\&delta;}}\left(n\right)\end{array}\right]+e\left[\begin{array}{c}-\sin \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\\ \cos \left(\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right]$

Wherein,

Δθ(n)＝θ(n)-θ _M(n)

$\left.\begin{array}{c}{L}_{\mathrm{\&gamma;}}=\frac{1}{2}\{(L_d+L_q)-(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&delta;}}=\frac{1}{2}\{(L_d+L_q)+(L_q-L_d)\cos \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\}\\ L_{\mathrm{\&gamma;\&delta;}}=\frac{1}{2}(L_q-L_d)\sin \left(2\mathrm{\&Delta;\&theta;}\left(n\right)\right)\end{array}\right\}$

P is a differential operator;

When estimation reaches stable state, Δ θ ≈ 0, so formula L _γ, L _δ, L _{γ δ}Be approximately equal to L respectively _d, L _q, 0, described L _γ, L _δ, L _{γ δ}Be respectively

The self-induction of γ axle, δ axle self-induction and γ, the mutual inductance of δ between centers;

Wherein,

N is the time cycle number after the discretization,

θ _MBe the rotor position angle of estimation,

e _MBe the back-emf of molded motor,

E is the back-emf of real electrical machinery,

ω is a rotor speed, promptly $\mathrm{\&omega;}={\mathrm{\&theta;}}_M^g,$

K _eBe back-emf estimation constant, K _θBe that the position estimation constant is set point;

u _γ, u _δFor real electrical machinery at γ, the voltage on the δ axle,

i _γ(n-1), i _δ(n-1) be true motor at γ, the current response on the δ axle,

i _{M γ}(n), i _{M δ}(n) be with γ, the δ axle is the current response of the molded motor of armature spindle foundation,

Δ i _γ(n), Δ i _δ(n) be in each cycle real current and molded motor at γ, the difference between currents on the δ axle;

(4) DSP adopts torque instruction electric current composite control method to calculate dq shaft voltage reference value, and it contains following steps successively;

(4.1) with the current i under the abc coordinate system _a, i _b, i _cMultiply by transformation matrix of coordinates and obtain current value i under the dq axle _d, i _q, i ₀:

$\left[\begin{array}{c}{i}_d\left(n\right)\\ i_q\left(n\right)\\ i_0\left(n\right)\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \left(\mathrm{\&theta;}\left(n\right)\right)& \cos (\mathrm{\&theta;}\left(n\right)-\frac{2}{3}\mathrm{\&pi;})& \cos (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \sin \left(\mathrm{\&theta;}\left(n\right)\right)& \sin \left(\mathrm{\&theta;}\left(n\right)\frac{2}{3}\mathrm{\&pi;}\right)& \sin (\mathrm{\&theta;}\left(n\right)+\frac{2}{3}\mathrm{\&pi;})\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]\left[\begin{array}{c}{i}_a\left(n\right)\\ i_b\left(n\right)\\ i_c\left(n\right)\end{array}\right]$

(4.2) DSP is the reference value ω of rotating speed _RefThe estimation rotating speed that obtains with step (3) $\mathrm{\&omega;}\left(n\right)={\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_M\left(n\right)$ Difference Δ ω (n)=ω _Ref-ω (n) input speed adjuster is tried to achieve q axle reference current i _{Q_ref}(n), $i_{q\_\mathrm{ref}}\left(n\right)=K_{p1}\mathrm{\&Delta;\&omega;}\left(n\right)+K_{i1}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;\&omega;}\left(i\right)T,$ Wherein, K _P1, K _I1Ratio, integral constant for speed regulator;

(4.3) rotor position that estimation obtains according to step 3 _M(n), in each control cycle, look into rotor-position-load torque curve table according to the position of estimation, the electromagnetic torque of motor equals load torque simultaneously, therefore obtains electromagnetic torque T _Em, calculate the torque current i that needs thus _{Q_ref}':

${i_{q\_\mathrm{ref}}}^{\&prime;}=\frac{T_{\mathrm{em}}}{p_n{\mathrm{\&psi;}}_r};$

(4.4) DSP is i _{Q_rer}'+i _{Q_ref}With actual q shaft current i _qDifference DELTA i _q(n)=i _{Q_ref}' (n)+i _{Q_ref}(n)-i _q(n) input torque current regulator obtains q axle reference voltage u _{Q_ref}(n),

$u_{q\_\mathrm{ref}}\left(n\right)=K_{p2}{\mathrm{\&Delta;i}}_q\left(n\right)+K_{i2}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}{\mathrm{\&Delta;i}}_q\left(i\right)T$

Wherein, K _P2, K _I2Ratio, integral constant for torque current regulator;

(4.5) DSP is i _{D_ref}=0 and i _dDifference DELTA i _d(n)=i _{D_ref}(n)-i _d(n) the input exciting current controller obtains d axle reference voltage u _{D_ref}(n), $u_{d\_\mathrm{ref}}\left(n\right)=K_{p3}{\mathrm{\&Delta;i}}_d\left(n\right)+K_{i3}\underset{i=0}{\overset{i=n}{\mathrm{\&Sigma;}}}{\mathrm{\&Delta;i}}_d\left(i\right)T,$ Wherein, K _P3, K _I3Ratio, integral constant for exciting current controller;

(5) DSP is according to u _{D_ref}, u _{Q_ref}Calculate the voltage vector under α β 0 coordinate system,

$\left[\begin{array}{c}{u}_{\mathrm{\&alpha;}}\left(n\right)\\ u_{\mathrm{\&beta;}}\left(n\right)\end{array}\right]=\left[\begin{array}{cc}\cos \mathrm{\&theta;}\left(n\right)& -\sin \mathrm{\&theta;}\left(n\right)\\ \sin \mathrm{\&theta;}\left(n\right)& \cos \mathrm{\&theta;}\left(n\right)\end{array}\right]\left[\begin{array}{c}{u}_{d\_\mathrm{ref}}\left(n\right)\\ u_{q\_\mathrm{ref}}\left(n\right)\end{array}\right]$

(6) DSP obtains the switching signal of inverter with the space vector PWM method, obtain opening the turn-off time of each switching tube of inverter, send the output voltage that corresponding pwm pulse removes control inverter, realization is to Control of PMSM: in each control cycle T, each switching tube is opened turn-off time T in three brachium pontis of inverter _a, T _b, T ₀:

T ₀＝T-T _a-T _b

$V_s=\sqrt{{u_{\mathrm{\&alpha;}}}^2+{u_{\mathrm{\&beta;}}}^2},\mathrm{\&gamma;}={\tan }^{-1}\frac{u_{\mathrm{\&alpha;}}}{u_{\mathrm{\&beta;}}}$

The space that voltage vector distributes is divided into 6 zones, voltage u in each control cycle _α, u _βSynthetic space vector amplitude under α β axle is V _s, argument is γ; And the action time of switch vector by each the zone in by V _sT, V _aT _a, V _bT _bThe triangle decision that constitutes, wherein V _a=V _b=E _Dc, T is a control cycle, E _DcBe DC bus-bar voltage; γ has determined that also the switch vector is the zone at space vector place simultaneously, i.e. the current residing running status of inverter.

Images