CN106357188A - Unified single/double-vector model prediction control method and device for permanent magnet motors - Google Patents

Abstract

The invention discloses a unified single/double-vector model prediction control method and device for permanent magnet motors and belongs to the technical field of transmission control of the permanent magnet motors. The unified single/double-vector model prediction control method comprises the steps of: differential equation discretization, solution of counter electromotive force, solution of three-phase duty ratio and prediction for control vector and acting time and the like. The unified single/double-vector model prediction control method and device disclosed by the invention have the advantages that a single-vector model prediction control algorithm and a double-vector model prediction control algorithm are creatively integrated into a single algorithm, and pulse width modulation strategies for the single-vector model prediction control algorithm and the double-vector model prediction control algorithm are unified, so that the defect that different numbers of vectors need to be differently treated when the traditional model prediction control method is used is overcome, the universality and practicability of the model prediction control method are improved, and an important improvement on the prior art is achieved.

Description

A kind of unified magneto mono-/bis-vector model forecast Control Algorithm and device

Technical field

The present invention relates to magneto transmission control technology field, particularly relate to a kind of unified magneto mono-/bis-arrow Amount model predictive control method and device.

Background technology

Model Predictive Control is a kind of very simple control algolithm of structure, because it has that principle is simple and clear, implements Execute convenience, the advantages of controller parameter adjusts simple, attract a large amount of scholars both at home and abroad that it is led in Electric Drive in recent years The application in domain is studied.

Traditional many vector models predictive control algorithm mainly comprises single vector model predictive control algorithm and double Vector Mode Type predictive control algorithm, both algorithms are relatively independent, can only apply one of which for different vector numbers.

For example, double vector model predictive control algorithms turn typically by vector needed for object function determination, then basis The minimum principle of square pulsation obtains action time, and this method may not apply to single vector model PREDICTIVE CONTROL；Single vector model Predictive control algorithm is typically with acting on the method for same vector within the whole sampling period, and to solve single vector model pre- The modulation problems of observing and controlling, this method also cannot be applied to double vector model PREDICTIVE CONTROL.

It can be seen that, model predictive control method of the prior art is due to its relative independentability, thus has that versatility is low, makes With the loaded down with trivial details, defect such as algorithm is complicated.

Content of the invention

In view of this, it is an object of the invention to proposing a kind of unified magneto mono-/bis-vector model PREDICTIVE CONTROL Method and device, it can unify single vector model predictive control algorithm and the pulse width of double vector model predictive control algorithm Modulation strategy, two kinds of predictive control algorithms are fused to a kind of algorithm, improve versatility and the practicality of model predictive control method Property.

Based on above-mentioned purpose, present invention provide the technical scheme that

A kind of unified magneto mono-/bis-vector model forecast Control Algorithm, it comprises the steps:

According to the mathematical model of magneto, adopt discretization method by the stator current differential equation obtain k-1, k-2 with And the counter electromotive force e in k-3 moment^k-1、e^k-2And e^k-3；

Negate electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Counter electromotive force e according to the k moment^k, torque reference valueAnd the permanent magnet flux linkage ψ of motor itself_fPrediction obtains Target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component obtains three-phase dutycycle d_a:d_b:d_c；

According to the relative size relation between control algolithm and three-phase dutycycle, obtain the control vector effect predicted Time.

Preferably, the stator current differential equation can be:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Represent stator current i_sDifferential.

Preferably, discretization method can adopt Bilinear transformation method, then counter electromotive force e^k-1、e^k-2And e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, It is the stator current in k, k-1, k-2, k-3 moment respectively, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period.

Preferably, the counter electromotive force e in k moment^kCounter electromotive force e can be adopted^k-1、e^k-2And e^k-3Arithmetic average.

Preferably, target voltage vectorCan be:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor Sampling period,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, e For natural constant.

Preferably, three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cCan be respectively as follows:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, obtained from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b, u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three.

Preferably, according to the relative size relation between control algolithm and three-phase dutycycle, obtain the control arrow predicted The concrete mode of amount and action time can be:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 synthesized Target voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁ With t ' action time₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v ′₂₁,v′₂₀With t ' action time₂₁,t′₂₀；

	t1+t0＞ t2+t0	t1+t0<t2+t0
			Selected vector [' v21,v′20]	[v1,v0]	[v2,v0]
Action time [t '21,t′20]	[t1+0.5t2t0+0.5t2]	[t2+0.5t1t0+0.5t1]

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by discretization method.

Additionally, the present invention also provides a kind of unified magneto mono-/bis-vector model prediction control device, comprising:

Descretization module, for the mathematical model according to magneto, using discretization method by stator current differential side Journey obtains the counter electromotive force e in k-1, k-2 and k-3 moment^k-1、e^k-2And e^k-3；

Averaging module, is used for negating electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Dutycycle asks for module, for the counter electromotive force e according to the k moment^k, torque reference valueAnd motor itself Permanent magnet flux linkage ψ_fPrediction obtains target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component obtains To three-phase dutycycle d_a:d_b:d_c；

Object module, for according to the relative size relation between control algolithm and three-phase dutycycle, obtaining prediction Control vector action time.

Preferably, this device can also comprise:

Control module, for control vector action time of being obtained according to object module of each switching tube to inverter Output drive signal.

Preferably, in said apparatus:

The stator current differential equation can be:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Represent stator current i_sDifferential；

Discretization method can adopt Bilinear transformation method, counter electromotive force e^k-1、e^k-2And e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, It is the stator current in k, k-1, k-2, k-3 moment respectively, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period；

The counter electromotive force e in k moment^kCan be counter electromotive force e^k-1、e^k-2And e^k-3Arithmetic average；

Target voltage vectorFor:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor Sampling period,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, e For natural constant；

Three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cCan be respectively:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) for noting The zero-sequence component entering；Wherein, max (u_a,u_b,u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a, u_b,u_cMinimum one in three；

According to the relative size relation between control algolithm and three-phase dutycycle, obtain the control vector effect predicted The concrete mode of time can be:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 synthesized Target voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁ With t ' action time₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v ′₂₁,v′₂₀With t ' action time₂₁,t′₂₀；

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by discretization method.

From the above it can be seen that the beneficial effects of the present invention is:

(1) present invention to obtain single vector, double vector using the method comparing relative size relation between three-phase dutycycle Required control vector corresponding action time in control algolithm, solve single vector and double pre- observing and controlling of vector model well The skimble-scamble problem of algorithm modulation strategy processed；

(2) the relatively conventional scheme of the present invention, repeatedly enumerating and complicated double counting without voltage vector, substantially reduce The complexity of algorithm.

In a word, single vector model predictive control algorithm and double vector model predictive control algorithm are creatively melted by the present invention It is combined into a kind of single algorithm, the pulse width of unified single vector model predictive control algorithm and double vector model predictive control algorithm Degree modulation strategy, overcomes the defect that conventional model forecast Control Algorithm need to be treated with a certain discrimination to different vector numbers when using, Improve versatility and the practicality of model predictive control method, be the important improvement to prior art.

Brief description

In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing Have technology description in required use accompanying drawing be briefly described it should be apparent that, drawings in the following description be only this Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, acceptable Other accompanying drawings are obtained according to these accompanying drawings.

Fig. 1 is the hardware structure diagram of magneto speed-adjusting and control system；

Fig. 2 is the control principle drawing of the embodiment of the present invention；

Fig. 3 is a kind of flow chart of present invention method；

Fig. 4 is a kind of structured flowchart of embodiment of the present invention device；

Fig. 5 be magneto single vector model PREDICTIVE CONTROL under 10khz sample rate, motor operation carries in 1500rpm The experimental result of nominal load；

Fig. 6 be the double vector model PREDICTIVE CONTROL of magneto under 10khz sample rate, motor operation carries in 1500rpm The experimental result of nominal load；

Fig. 7 be magneto single vector model PREDICTIVE CONTROL under 10khz sample rate, motor operation carries volume in 150rpm The experimental result of fixed load；

Fig. 8 be the double vector model PREDICTIVE CONTROL of magneto under 10khz sample rate, motor operation carries volume in 150rpm The experimental result of fixed load；

Fig. 9 be magneto single vector model PREDICTIVE CONTROL under 10khz sample rate, motor is started to by static Experimental result during 1500rpm；

Figure 10 be the double vector model PREDICTIVE CONTROL of magneto under 10khz sample rate, motor is started to by static Experimental result during 1500rpm；

Figure 11 be magneto single vector model PREDICTIVE CONTROL under 10khz sample rate, carry out during 1500rpm rotating Experimental result；

Figure 12 be the double vector model PREDICTIVE CONTROL of magneto under 10khz sample rate, carry out during 1500rpm rotating Experimental result.

Specific embodiment

For making the object, technical solutions and advantages of the present invention become more apparent, below in conjunction with specific embodiment, and reference Accompanying drawing, the present invention is described in more detail.

Fig. 1 is the hardware circuit figure of the embodiment of the present invention, includes three-phase voltage source, three-phase diode rectifier bridge, directly Flow lateral capacitance, magneto, voltage x current sample circuit, (digital signal processing, at digital signal for dsp Reason) controller and drive circuit.Voltage x current sample circuit is adopted respectively using voltage hall sensor and current Hall sensor Collection DC voltage and magneto a, b phase current, sampled signal enters the conversion of dsp controller after signal conditioning circuit For digital signal.Dsp controller completes the computing of method proposed by the invention, exports six way switch pulses, is then passed through driving The final drive signal of six switching tubes of inverter is obtained after circuit.

Fig. 2 is the control principle drawing of the embodiment of the present invention, and the control method in the embodiment of the present invention controls in the dsp of Fig. 2 Execute on device.

As shown in figure 3, a kind of unified magneto mono-/bis-vector model forecast Control Algorithm, it comprises the steps:

Step 101, according to the mathematical model of magneto, obtains k- using discretization method by the stator current differential equation 1st, the counter electromotive force e in k-2 and k-3 moment^k-1、e^k-2And e^k-3；

Step 102, negates electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Step 103, according to the counter electromotive force e in k moment^k, torque reference value t_e ^refAnd the permanent magnet flux linkage ψ of motor itself_f Prediction obtains target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component obtains three-phase dutycycle d_a:d_b:d_c；

Step 104, according to the relative size relation between control algolithm and three-phase dutycycle, obtains the control arrow predicted Amount and action time.

Single vector model predictive control algorithm and double vector model predictive control algorithm are creatively fused to by the method A kind of single algorithm, the pulse width of unified single vector model predictive control algorithm and double vector model predictive control algorithm is adjusted System strategy, overcomes the defect that conventional model forecast Control Algorithm need to be treated with a certain discrimination to different vector numbers when using, and improves The versatility of model predictive control method and practicality, are the important improvement to prior art.

Based on the above method, the stator current differential equation can be:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Represent stator current i_sDifferential.

Based on the above method, discretization method can adopt Bilinear transformation method, then counter electromotive force e^k-1、e^k-2With e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, It is the stator current in k, k-1, k-2, k-3 moment respectively, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period.

Based on the above method, the counter electromotive force e in k moment^kCounter electromotive force e can be adopted^k-1、e^k-2And e^k-3Calculation Number meansigma methodss.

Based on the above method, target voltage vectorCan be:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor Sampling period,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, e For natural constant.

Based on the above method, three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cCan be respectively as follows:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, its value obtains from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b, u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three.

Based on the above method, according to the relative size relation between control algolithm and three-phase dutycycle, obtain The concrete mode of predictive vector and predicting function time can be:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 synthesized Target voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁ With t ' action time₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v ′₂₁,v′₂₀With t ' action time₂₁,t′₂₀；

	t1+t0＞ t2+t0	t1+t0<t2+t0
			Selected vector [v '21,v′20]	[v1,v0]	[v2,v0]
Action time [t '21,t′20]	[t1+0.5t2t0+0.5t2]	[t2+0.5t1t0+0.5t1]

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by discretization method.

Comprehensive above method, can obtain one kind more specifically control method, it comprises the steps of

Step 1: the torque reference value being obtained according to outer shroud rotating speed pi (proportional integral) actuatorIt is embodied as:

$t_e^{ref}=(k_p+\frac{k_i}{s})({\mathrm{\ω}}_r^{ref}-{\mathrm{\ω}}_r),$

Wherein, k_pAnd k_iIt is respectively the proportional gain in pi actuator and storage gain,And ω_rIt is respectively rotating speed reference Value and rotary speed actual value,Represent integration.

Step 2: according to the mathematical model of magneto, the differential equation of stator current can be obtained:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

Wherein, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Represent stator current i_sDifferential.

Step 3: stator current differential equation discretization step 2 being obtained using the method for bilinear transformation, thus Counter electromotive force e to k-1, k-2 and k-3 moment^k-1、e^k-2And e^k-3.

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

Wherein,It is the stator voltage in k-1, k-2, k-3 moment respectively, It is the stator current in k, k-1, k-2, k-3 moment respectively；t_sFor the sampling period.

Step 4: the counter electromotive force e being obtained according to step 3^k-1、e^k-2And e^k-3The k moment is obtained by the method averaged Counter electromotive force e^k:

$e^k=\frac{1}{3}(e^{k-1}+e^{k-2}+e^{k-3}).$

Step 5: the torque reference value being obtained using step 1The counter electromotive force e in the k moment that step 4 obtains^kAnd electricity The permanent magnet flux linkage ψ of machine itself_fPrediction obtains target voltage vector

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

Wherein,For stator current reference value, p is the number of pole-pairs of motor, and θ is rotor electric angle Degree, j is imaginary unit, and e is natural logrithm.

Step 6: the target voltage vector being obtained according to step 5Carrier wave pwm modulation skill using injection zero-sequence component Art, obtains corresponding three-phase dutycycle d_a:d_b:d_c:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, its value obtains from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b, u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three.

Step 7: according to three-phase dutycycle d_a:d_b:d_cObtain synthesizing target voltage vector with table 1Three required electricity Pressure vector v₀,v₁,v₂And their t action time₀,t₁,t₂.

Step 8: obtain the single control vector v ' of corresponding single vector control method according to table 2₁₁With t ' action time₁₁, root Obtain two control vector v ' of double vector control methods according to table 3₂₁,v′₂₀With t ' action time₂₁,t′₂₀.

Step 9: the voltage vector combination being obtained according to step 8 and information action time can build and obtain driving inverter The drive signal of switching tube.

The effectiveness of said method can be drawn by contrasting the experimental result shown in Fig. 5～Figure 12.All many Vector Modes The experiment of type PREDICTIVE CONTROL is all carried out under 10khz sample rate.Each in figure of Fig. 5～Figure 12, waveform is successively from top to bottom For rotating speed, electromagnetic torque, stator magnetic linkage amplitude and motor stator end a phase current.Fig. 5～Fig. 8 is motor Steady Experimental result, Fig. 5 and 6 corresponds to the experimental result carrying out full-load run using motor during single, double vector control method in 1500rpm, Fig. 7 respectively Correspond to the experimental result carrying out full-load run using motor during single, double vector control method in 150rpm with 8 respectively.From Fig. 5～ It is found that the method implemented used by the embodiment of the present invention has obtained steady state effect well in the contrast of Fig. 8, simultaneously acceptable Find out, lower torque, magnetic linkage pulsation and more sinusoidal stator current can be obtained using double vector control methods.Fig. 9～ Figure 12 is motor dynamics experimental result, and correspondence carries out opening from static Fig. 9 and 10 using motor during single, double vector control method respectively Move the experimental result of 1500rpm operation, Figure 11 and 12 is corresponding respectively to be existed using motor during single, double vector control method 1500rpm carries out the experimental result of rotating operation.It is found that implementing the embodiment of the present invention from the contrast of Fig. 9～Figure 12 Method used has obtained dynamic response well, simultaneously it can also be seen that during Larger Dynamic, described in the embodiment of the present invention Method there is quick dynamic property, simultaneously compared to single vector control method, double vector control methods have smoother turning Square, magnetic linkage waveform and more sinusoidal stator current.

Fig. 4 show a kind of unified magneto mono-/bis-vector model prediction control device, comprising:

Descretization module 401, for the mathematical model according to magneto, using discretization method by stator current differential Equation obtains the counter electromotive force e in k-1, k-2 and k-3 moment^k-1、e^k-2And e^k-3；

Averaging module 402, is used for negating electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Dutycycle asks for module 403, for the counter electromotive force e according to the k moment^k, torque reference valueAnd motor itself Permanent magnet flux linkage ψ_fPrediction obtains target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component Obtain three-phase dutycycle d_a:d_b:d_c；

Object module 404, for according to the relative size relation between control algolithm and three-phase dutycycle, being predicted Control vector action time.

Additionally, still seeing Fig. 4, said apparatus can also comprise:

Control module 405, opened to each of inverter for control vector action time of being obtained according to object module Close pipe output drive signal.

Further, in said apparatus:

The stator current differential equation can be:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Represent stator current i_sDifferential；

Discretization method can adopt Bilinear transformation method, counter electromotive force e^k-1、e^k-2And e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, It is the stator current in k, k-1, k-2, k-3 moment respectively, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period；

The counter electromotive force e in k moment^kCan be counter electromotive force e^k-1、e^k-2And e^k-3Arithmetic average；

Target voltage vectorFor:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor Sampling period,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, e For natural constant；

Three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cCan be respectively:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, its value obtains from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b, u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three；

According to the relative size relation between control algolithm and three-phase dutycycle, obtain the control vector effect predicted The concrete mode of time can be:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 synthesized Target voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁ With t ' action time₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v ′₂₁,v′₂₀With t ' action time₂₁,t′₂₀；

	t1+t0＞ t2+t0	t1+t0<t2+t0
			Selected vector [v '21,v′20]	[v1,v0]	[v2,v0]
Action time [t '21,t′20]	[t1+0.5t2t0+0.5t2]	[t2+0.5t1t0+0.5t1]

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by discretization method.

The device of above-described embodiment is used for realizing corresponding method in previous embodiment, and has corresponding method enforcement The beneficial effect of example, will not be described here.

Those of ordinary skill in the art are it is understood that the discussion of any of the above embodiment is exemplary only, not It is intended to imply that the scope of the present disclosure (inclusion claim) is limited to these examples；Under the thinking of the present invention, above example Or can also be combined between the technical characteristic in different embodiments, step can be realized with random order, and exists such as The other change of many of the upper described different aspect of the present invention, for their not offers in details simple and clear.

Embodiments of the invention be intended to fall into all such replacement within the broad range of claims, Modification and modification.Therefore, all any omissions within the spirit and principles in the present invention, made, modification, equivalent, improvement Deng should be included within the scope of the present invention.

Claims (10)

1. a kind of unified magneto mono-/bis-vector model forecast Control Algorithm is it is characterised in that comprise the steps:

According to the mathematical model of magneto, k-1, k-2 and k-3 are obtained by the stator current differential equation using discretization method The counter electromotive force e in moment^k-1、e^k-2And e^k-3；

Take described counter electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Counter electromotive force e according to the described k moment^k, torque reference valueAnd the permanent magnet flux linkage ψ of motor itself_fPrediction obtains Target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component obtains three-phase dutycycle d_a:d_b:d_c；

According to the relative size relation between control algolithm and described three-phase dutycycle, obtain the control vector effect predicted Time.

2. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that The described stator current differential equation is:

$\frac{{\mathrm{di}}_s}{dt}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Table Show stator current i_sDifferential.

3. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that Described discretization method is Bilinear transformation method, described counter electromotive force e^k-1、e^k-2And e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, Respectively It is the stator current in k, k-1, k-2, k-3 moment, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period.

4. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that The counter electromotive force e in described k moment^kFor described counter electromotive force e^k-1、e^k-2And e^k-3Arithmetic average.

5. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that Described target voltage vectorFor:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor sampling Cycle,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, and e is nature Constant.

6. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that Described three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cIt is respectively as follows:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, its value obtains from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b,u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three.

7. unified magneto mono-/bis-vector model forecast Control Algorithm according to claim 1 it is characterised in that Described according to the relative size relation between control algolithm and described three-phase dutycycle, obtain the control vector effect predicted The concrete mode of time is:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 obtains synthesizing target Voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁And work With time t '₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v '₂₁,v′₂₀ With t ' action time₂₁,t′₂₀；

t₁+t₀＞ t₂+t₀ t₁+t₀<t₂+t₀ Selected vector [v '₂₁,v′₂₀] [v₁,v₀] [v₂,v₀] Action time [t '₂₁,t′₂₀] [t₁+0.5t₂t₀+0.5t₂] [t₂+0.5t₁t₀+0.5t₁]

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by described discretization method.

8. a kind of unified magneto mono-/bis-vector model prediction control device is it is characterised in that include:

Descretization module, for the mathematical model according to magneto, is obtained by the stator current differential equation using discretization method Counter electromotive force e to k-1, k-2 and k-3 moment^k-1、e^k-2And e^k-3；

Averaging module, is used for taking described counter electromotive force e^k-1、e^k-2And e^k-3Meansigma methodss as the k moment counter electromotive force e^k；

Dutycycle asks for module, for the counter electromotive force e according to the described k moment^k, torque reference valueAnd motor itself Permanent magnet flux linkage ψ_fPrediction obtains target voltage vectorCarrier pulse width modulation technology using injection zero-sequence component obtains To three-phase dutycycle d_a:d_b:d_c；

Object module, for according to the relative size relation between control algolithm and described three-phase dutycycle, obtaining prediction Control vector action time.

9. unified magneto mono-/bis-vector model prediction control device according to claim 8 it is characterised in that Also comprise:

Control module, for control vector action time of being obtained according to described object module of each switching tube to inverter Output drive signal.

10. unified magneto mono-/bis-vector model prediction control device according to claim 8 it is characterised in that:

The described stator current differential equation is:

$\frac{{\mathrm{di}}_s}{\mathrm{dt}}=\frac{1}{l_s}(u_s-r_s{i}_s-e),$

In formula, i_sFor stator current, l_sFor stator inductance, u_sFor stator voltage, r_sFor stator resistance, e is counter electromotive force,Table Show stator current i_sDifferential；

Described discretization method is Bilinear transformation method, described counter electromotive force e^k-1、e^k-2And e^k-3It is respectively as follows:

$e^{k-1}=u_s^{k-1}-0.5r_s(i_s^k+i_s^{k-1})-l_s(i_s^k-i_s^{k-1})/t_s$

$e^{k-2}=u_s^{k-2}-0.5r_s(i_s^{k-1}+i_s^{k-2})-l_s(i_s^{k-1}-i_s^{k-2})/t_s,$

$e^{k-3}=u_s^{k-3}-0.5r_s(i_s^{k-2}+i_s^{k-3})-l_s(i_s^{k-2}-i_s^{k-3})/t_s$

In formula,It is the stator voltage in k-1, k-2, k-3 moment respectively, Respectively It is the stator current in k, k-1, k-2, k-3 moment, l_sFor stator inductance, r_sFor stator resistance, t_sFor the sampling period；

The counter electromotive force e in described k moment^kFor described counter electromotive force e^k-1、e^k-2And e^k-3Arithmetic average；

Described target voltage vectorFor:

$u_s^{ref}=0.5r_s(i_s^{ref}+i_s^k)+l_s(i_s^{ref}-i_s^k)/t_s+e^k,$

In formula,For stator current reference value, r_sFor stator resistance, l_sFor stator inductance, t_sFor sampling Cycle,Stator current for the k moment；

In formulaIn, p is the number of pole-pairs of motor, and θ is rotor electrical angle, and j is imaginary unit, and e is nature Constant；

Described three-phase dutycycle d_a:d_b:d_cIn d_a、d_b、d_cIt is respectively as follows:

d_a=0.5* (u_a+u_z+1)

d_b=0.5* (u_b+u_z+ 1),

d_c=0.5* (u_c+u_z+1)

In formula, u_a,u_b,u_cFor original three-phase modulations ripple, its value obtains from target voltage vector according to following formula:

Wherein u_dcFor inverter DC bus-bar voltage；

u_z=-0.5* (max (u_a,u_b,u_c)+min(u_a,u_b,u_c)) it is the zero-sequence component injected, wherein, max (u_a,u_b,u_c) represent u_a,u_b,u_cMaximum one in three, min (u_a,u_b,u_c) represent u_a,u_b,u_cMinimum one in three；

Described according to the relative size relation between control algolithm and described three-phase dutycycle, obtain the control vector predicted The concrete mode of action time is:

According to three-phase dutycycle d_a:d_b:d_cMiddle d_a、d_b、d_cMagnitude relationship, corresponding relation as shown in Table 1 obtains synthesizing target Voltage vectorThree required voltage vector v₀,v₁,v₂And their t action time₀,t₁,t₂；

Table 1

For single vector model, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 2 obtains controlling vector v '₁₁And work With time t '₁₁；

Table 2

For double vector models, according to t action time₀,t₁,t₂, corresponding relation as shown in Table 3 obtains controlling vector v '₂₁,v′₂₀ With t ' action time₂₁,t′₂₀；

t₁+t₀＞ t₂+t₀ t₁+t₀<t₂+t₀ Selected vector [v '₂₁,v′₂₀] [v₁,v₀] [v₂,v₀] Action time [t '₂₁,t′₂₀] [t₁+0.5t₂t₀+0.5t₂] [t₂+0.5t₁t₀+0.5t₁]

Table 3

In Tables 1 and 2, t_sThe sampling period being adopted by described discretization method.