A Symmetrical Generalized Double-Vector Model Predictive Current Control Method for Asynchronous Motors

Abstract

The present disclosure provides a dual-vector model predictive current control method for an induction motor, including: obtaining a stator current reference vector in a two-phase stationary coordinate system; deriving a predicted value of the stator current at time k + 2 considering one-beat delay compensation; calculating a stator voltage reference vector based on the principle of stator current deadbeat control; using a space vector pulse width modulation method to calculate the initial action times of two effective vectors and a zero vector; comparing the magnitude relationship of the vector initial action times to select two basic voltage vectors and their action times acting within a control period; determining symmetrically placed basic voltage vectors based on the principle of minimizing the number of switch state transitions and optimizing the vector action sequence; optimizing the vector action time with the goal of minimizing the effective value of the current error within a control period. The dual-vector model predictive current control method for an induction motor provided by the present disclosure effectively reduces current harmonics, has low computational complexity, is simple to implement, and has strong practicability.

Description

Technical Field

[0001] The present disclosure relates to the technical field of asynchronous motor drive control, and more specifically, to a symmetric generalized double-vector model predictive current control method for asynchronous motors. Background Art

[0002] The main control idea of model predictive control is to apply one or more voltage vectors within a control period to better approximate its control target. Traditional model predictive control is single-vector model predictive control, that is, only one voltage vector is applied within a control period, so its steady-state performance is poor. Various improved multi-vector model predictive controls apply multiple voltage vectors within a control period, effectively improving the steady-state performance of the system and have been widely studied by many scholars.

[0003] Existing double-vector model predictive controls can be divided into two categories: one is ordinary double-vector model predictive control, which applies one effective voltage vector and one zero vector within a control period; the other is generalized double-vector model predictive control, which can apply one effective voltage vector and one zero vector within a control period, or two effective voltage vectors. At medium and low speeds, the vector selection of generalized double-vector model predictive control is basically the same as that of ordinary double-vector model predictive control, that is, one effective voltage vector and one zero vector, so their control effects are almost the same; at high speeds, the vectors selected by generalized double-vector model predictive control are two effective voltage vectors, so its control effect is significantly better than that of ordinary double-vector model predictive control.

[0004] However, when determining the voltage vector application method in existing generalized double-vector model predictive control, the influence of vector application order and vector application time on the output current harmonics is not considered, so its steady-state performance is not optimal. Summary of the Invention

[0005] The purpose of the present disclosure is to provide a symmetric generalized double-vector model predictive current control method for asynchronous motors, which can optimize the vector application order and vector application time of traditional generalized double vectors, effectively improving the steady-state performance of the control system. The method of the present disclosure has simple voltage vector selection, low computational complexity, and strong practicability.

[0006] Based on the above purpose, the present disclosure provides a symmetric generalized double-vector model predictive current control method for asynchronous motors, which includes the following steps:

[0007] Step 1, obtain the stator current reference vector in the two-phase stationary coordinate system;

[0008] Step 2, considering one-beat delay compensation, derive the predicted value of the stator current at time k + 2;

[0009] Step 3: Based on the principle of stator current deadbeat control, calculate the reference vector of the stator voltage.

[0010] Step 4: Adopt the space vector pulse width modulation method to calculate the initial action time of two effective vectors and a zero vector.

[0011] Step 5: Compare the magnitude relationship of the vector initial action times, select two basic voltage vectors acting within a control period, and calculate their action times.

[0012] Step 6: Determine the symmetrically placed basic voltage vectors based on the principle of minimizing the number of switching state transitions, and optimize the vector action sequence.

[0013] Step 7: Optimize the vector action time with the goal of minimizing the effective value of the current error within a control period.

[0014] The present disclosure has the following features and advantages:

[0015] 1) Determine the symmetrically placed basic voltage vectors according to the principle of minimizing the number of switching state transitions, so as to reduce the switching frequency.

[0016] 2) Optimize the vector action time with the goal of minimizing the effective value of the current error within a control period.

[0017] 3) Small current harmonics are generated at a similar switching frequency, especially at high speeds, and the current harmonics can be reduced by 22.3% compared with the traditional method. Description of the Drawings

[0018] Figure 1 is a schematic flow chart of a symmetric generalized dual-vector model predictive current control method for an asynchronous motor;

[0019] Figure 2 is a control block diagram of the symmetric generalized dual-vector model predictive current control method for an asynchronous motor proposed by the present disclosure;

[0020] Figure 3 is a voltage space vector diagram of a two-level voltage source pulse width modulation (PWM) converter in an embodiment of the present disclosure;

[0021] Figure 4 is a schematic distribution diagram of the vector combinations selected by the generalized dual-vector model predictive control in the entire complex plane in an embodiment of the present disclosure;

[0022] Figure 5 is a schematic diagram of the vector action mode of the symmetric generalized dual-vector model predictive current control method for an asynchronous motor proposed by the present disclosure within a control period;

[0023] Figure 6It is a schematic diagram of harmonic current of the asynchronous motor symmetric generalized double-vector model predictive current control method proposed in this disclosure within one control period;

[0024] Figure 7 It is the steady-state experimental waveform of the traditional generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 300 r / min;

[0025] Figure 8 It is the steady-state experimental waveform of the symmetric generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 300 r / min;

[0026] Figure 9 It is the steady-state experimental waveform of the traditional generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 900 r / min;

[0027] Figure 10 It is the steady-state experimental waveform of the symmetric generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 900 r / min;

[0028] Figure 11 It is the steady-state experimental waveform of the traditional generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 1500 r / min;

[0029] Figure 12 It is the steady-state experimental waveform of the symmetric generalized double-vector model predictive current control method in the embodiment of this disclosure under rated load at a speed of 1500 r / min;

[0030] Figure 13 It is the comparison of switching frequencies of the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control in the full speed range in the embodiment of this disclosure. Among the two columns in the same group, the left side corresponds to the traditional method and the right side corresponds to the improved method;

[0031] Figure 14 It is the comparison of current THD of the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control in the full speed range in the embodiment of this disclosure. Among the two columns in the same group, the left side corresponds to the traditional method and the right side corresponds to the improved method. Detailed implementation manners

[0032] The following embodiments can enable those skilled in the art to understand this disclosure more comprehensively, but do not limit this disclosure in any way.

[0033] Figure 1 It is a schematic flowchart of an asynchronous motor symmetric generalized double-vector model predictive current control method, and this method is inFigure 2 In the control block diagram of

[0034] Step 1: Obtain the stator current reference vector in the two-phase stationary coordinate system:

[0035]

[0036] where is the reference value of the d-axis current, is the reference value of the q-axis current, θ e is the rotor flux angle of the motor, and j is the imaginary unit.

[0037] Step 2: Considering one-beat delay compensation, derive the predicted value of the stator current at time k + 2.

[0038] According to the mathematical model of the induction motor and considering one-beat delay compensation, the predicted value of the stator current at time k + 1 is:

[0039]

[0040] where is the control period from time k to k + 1, is the voltage vector applied from time k to k + 1, and are the stator current vector and the rotor flux vector at time k, respectively, R s is the stator resistance, R r is the rotor resistance, L m is the mutual inductance between the stator and rotor windings, L s is the stator inductance, L r is the rotor inductance, T r = L r / R r is the rotor time constant, is the leakage coefficient, ω r is the angular velocity of the motor.

[0041] The predicted value of the stator current at time k + 2 can be expressed as:

[0042]

[0043] where is the control period from time k + 1 to k + 2, is the voltage vector applied from time k + 1 to k + 2, and are the stator current vector and the rotor flux vector at time k + 1, respectively.

[0044] Step 3: Based on the principle of stator current deadbeat control, calculate the stator voltage reference vector:

[0045]

[0046] Among them, is the stator current reference vector at the (k + 2)th moment, and ω e is the synchronous angular velocity.

[0047] Step 4: Using the space vector pulse width modulation method, calculate the initial action times of two effective vectors and a zero vector.

[0048] Referring to Figure 3 , the initial vector action times T1, T2, and T0 corresponding to the two effective vectors u1, u2, and a zero vector u0 / u7 are respectively:

[0049]

[0050] T0 = T sc -(T1 + T2)

[0051] Among them, U dc is the DC bus voltage, θ is the angle between and u1, and T sc is the control period.

[0052] Step 5: Compare the magnitude relationships of the vector initial action times, select two basic voltage vectors acting within one control period, and calculate their action times.

[0053] The two basic voltage vectors acting within one control period and their action times are shown in Table 1. The distribution of the selected vector combinations in the entire complex plane refers to Figure 4 .

[0054] Table 1

[0055]

[0056] Step 6: Determine the symmetrically placed basic voltage vectors based on the principle of minimizing the number of switching states, and optimize the vector action sequence;

[0057] Select one voltage vector from the two voltage vectors determined in Step 5 and place it on both sides of the other voltage vector to optimize the vector action sequence in the double-vector model predictive control. The optimized vector action sequence and vector action times are shown in Table 2. Among them, k0 represents the distribution ratio of the action time of the voltage vector placed on both sides.

[0058] Table 2

[0059]

[0060]

[0061] Meanwhile, to reduce the switching frequency, the voltage vectors placed on both sides are selected according to the principle of the minimum number of switching states based on the last voltage vector acting in the previous control period.

[0062] Step 7: Optimize the vector action time with the goal of minimizing the effective value of the current error within one control period.

[0063] The instantaneous value of the current harmonics within one control period can be expressed as:

[0064]

[0065] where u(t) represents the three-phase output voltage; t0 represents the starting moment of the control period.

[0066] The square of the effective value of the harmonic current within one control period can be expressed as:

[0067]

[0068] The vector action mode of the method proposed in this disclosure within one control period refers to Figure 5 , and its current harmonics within one control period refer to Figure 6 .

[0069] Assume that the initial harmonic current and the terminal harmonic current of one control period are i h0 and i h3 respectively, then there is:

[0070]

[0071] where respectively represent the stator current reference vectors at the initial moment and the end moment of one control period.

[0072] The slope of the current reference vector of one control period is:

[0073]

[0074] Assume that the selected optimal voltage vector combination is (u x , u y ), then the current change rates corresponding to the two basic voltage vectors u x , u y are respectively:

[0075]

[0076] Define the differences between k ref and k x and k y as:

[0077] Δkx = k ref -k x

[0078] Δk y = k ref -k y

[0079] Assume that the vector application time T = [k0T x T y T x corresponding to the basic voltage vector sequence U = [u x T y (1 - k0)T x , then the harmonic currents corresponding to the two middle endpoints of the three-segment broken line are respectively:

[0080] i h1 = i h0 + Δk x k0T x

[0081] i h2 = i h1 + Δk y T y

[0082] Substitute the harmonic currents i h = [i h0 i h1 i h2 i h3 corresponding to the four endpoints of the three-segment broken line into Equation to obtain an expression of Ih rms 2 with respect to k0. It is found that Ih rms 2 is a quadratic function of k0. Due to space limitations, only the simplified calculation formula of k0 is given:

[0083]

[0084] Substitute k0 into T = [k0T x T y (1 - k0)T x to obtain the optimized vector application time.

[0085] The effectiveness of the method proposed in this disclosure can be obtained by comparing the Figures 7 to 13 shown experimental results. Figure 7 and Figure 8They are the steady-state experimental waveforms of the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control under the rated load at a speed of 300 r / min. It can be seen that at low speeds, the actual speed fluctuation of the traditional method is relatively obvious, and the dq-axis current ripple is relatively large. Figure 9 and Figure 10 They are the steady-state experimental waveforms of the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control under the rated load at a speed of 900 r / min. It can be seen that both the traditional method and the improved method can accurately track the reference speed of the motor, and the current waveforms are sinusoidal. Figure 11 and Figure 12 They are the steady-state experimental waveforms of the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control under the rated load at a speed of 1500 r / min. It can be seen that the current waveforms of both the traditional method and the improved method are very sinusoidal, and compared with low and medium speeds, the q-axis current ripple is significantly reduced. Figure 13 and Figure 14 They are the comparison of the switching frequency and the total harmonic distortion (THD) of the current between the traditional generalized double-vector model predictive current control and the symmetric generalized double-vector model predictive current control in the full speed range. It can be seen that in the full speed range, the switching frequencies of both the traditional method and the improved method are not fixed and both decrease with the increase of the motor speed. The switching frequency of the improved method is similar to that of the traditional method, but the current harmonics are significantly reduced. Especially at high speeds, compared with the traditional method, the improved method can reduce the current harmonics by up to 22.3%.

[0086] The asynchronous motor double-vector model predictive current control method provided by the present disclosure effectively reduces current harmonics, has a small computational complexity, is simple to implement, and has strong practicability.

[0087] Those skilled in the art should understand that the above embodiments are only exemplary embodiments, and various changes, substitutions, and alterations can be made without departing from the spirit and scope of the present disclosure.

Claims

1. A method for predictive current control of an asynchronous motor using a symmetrical generalized dual-vector model, characterized in that: The steps include: Step 1, obtaining a stator current reference vector in a two-phase stationary coordinate system; Step 2, considering one-beat delay compensation, derive the predicted value of the stator current at time k+2; Step 3, based on the principle of deadbeat control of stator current, calculate and obtain the stator voltage reference vector; Step 4, using the space vector pulse width modulation method, calculate the initial action time of two effective vectors and one zero vector; Step 5, compare the magnitude relationship of the initial action time of the vectors, select two basic voltage vectors acting within a control cycle, and calculate their action time; Step 6, determining the basic voltage vectors placed symmetrically based on the principle of minimizing the number of switching state changes, and optimizing the order of vector action; Step 7, optimizing the vector action time with the goal of minimizing the effective value of the current error within a control cycle; Wherein, the step 2 comprises: According to the mathematical model of the asynchronous motor, considering one-beat delay compensation, the predicted value of the stator current at time k+1 is: in, is the control period from time k to time k+1, is the voltage vector acting from time k to k+1, and are the stator current vector and rotor flux vector at the kth moment, R s is the stator resistance, R r is the rotor resistance, L m is the mutual inductance between the stator and rotor windings, L s is the stator inductance, L r is the rotor inductance, T r =L r / R r is the rotor time constant, is the magnetic flux leakage coefficient, ω r is the motor angular velocity; The predicted value of stator current at time k+2 is expressed as: in, is the control period from k+1 to k+2, is the voltage vector acting from time k+1 to k+2, and They are the stator current vector and rotor flux vector at the k+1th moment respectively.

2. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 1, characterized in that: The stator current reference vector is: in, is the d-axis current reference value, is the q-axis current reference value, θ e is the rotor flux angle of the motor, and j is an imaginary unit.

3. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 2, characterized in that: In step 3, based on the principle of deadbeat control of stator current, the stator voltage reference vector calculated is: in, is the stator current reference vector at time k+2, ω e is the synchronous angular velocity.

4. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 3, characterized in that: In step 4, the initial vector action times T1, T2 and T0 corresponding to the two effective vectors u1, u2 and one zero vector u0 / u7 are: Among them, U dc is the DC bus voltage, θ is The angle between u1 and T sc To control the cycle.

5. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 4, characterized in that: In step 5, by comparing the magnitude relationship of the initial action time of the vectors, the two basic voltage vectors acting in one control cycle and their action time satisfy: When T0<min(T1,T2), the basic voltage vector is (u1 u2) and the vector action time is When T1<min(T0,T2), the basic voltage vector is (u2 u7) and the vector action time is When T2<min(T0,T1), the basic voltage vector is (u1 u0) and the vector action time is 6. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 5, characterized in that: The step 6 comprises: From the two voltage vectors determined in step 5, one voltage vector is selected and placed on both sides of the other voltage vector to optimize the vector action order in the dual-vector model predictive control. The optimized vector action order and vector action time are shown below, where k0 represents the distribution ratio of the voltage vector action time placed on both sides: When T0<min(T1,T2): the basic voltage vector is [u1u2u1], and the vector action time is The basic voltage vector is [u2 u1 u2], and the vector action time is When T1<min(T0,T2): the basic voltage vector is [u2 u7 u2], and the vector action time is The basic voltage vector is [u7 u2 u7], and the vector action time is When T2<min(T0,T1): the basic voltage vector is [u1 u0 u1], and the vector action time is The basic voltage vector is [u0 u1 u0], and the vector action time is 7. The asynchronous motor symmetric generalized dual vector model predictive current control method according to claim 6, characterized in that: The step 7 comprises: The instantaneous value of current harmonics within a control cycle is expressed as: Among them, u(t) represents the three-stage output voltage; t0 represents the starting time of the control cycle; The square of the effective value of harmonic current in one control cycle is expressed as: Assume that the initial harmonic current and the final harmonic current of a control cycle are i h0 and i h3 , then: in, Respectively represent the stator current reference vector at the initial moment and the end moment of a control cycle; The slope of the current reference vector in one control cycle is: Assume that the selected optimal voltage vector combination is (u x ,u y ), then the two basic voltage vectors u x 、u y The corresponding current change rates are: Define k ref With k x and k y The differences are: Δk x =k ref -k x Δk y =k ref -k y Assume that the basic voltage vector sequence U = [u x u y u x ] corresponds to the vector action time T = [k0T x T y (1-k0)T x ], then the harmonic currents corresponding to the two middle endpoints of the three-segment broken line are: i h1 =i h0 +Δk x k0T x i h2 =i h1 +Δk y T y The harmonic current i corresponding to the four endpoints of the three-segment broken line h =[i h0 i h1 i h2 i h3 ]Substitution That is, Ih is obtained within a control cycle rms 2 Regarding the expression of k0, Ih rms 2 It is a quadratic function about k0. The simplified calculation formula of k0 is: Substitute k0 into T = [k0T x T y (1-k0)T x ], that is, the optimized vector action time is obtained.

Images