Sequential model prediction control method for six-phase permanent magnet synchronous motor

Abstract

The invention provides a sequential model prediction control method for a six-phase permanent magnet synchronous motor, which comprises the following steps: expanding the six-phase permanent magnet synchronous motor into a four-dimensional orthogonal coordinate system through a transformation matrix, converting an original plane into an alpha-beta plane and an x-y plane, the alpha-beta plane being a main plane, and the x-y plane being a harmonic plane; a cost function of a model prediction control strategy is constructed through sequential model prediction, after optimization, an optimal switching sequence is selected to control on-off of an IGBT in a six-phase inverter, and the six-phase inverter is connected with the six-phase permanent magnet synchronous motor. The method has the advantages that the harmonic content (THD) of the output current can be greatly reduced by a sequential model prediction control strategy provided by the invention; the switching loss can be greatly reduced, and the service life of the power switch tube is prolonged; compared with a traditional space vector modulation strategy, the method has the advantage that the calculation time is shortened by about 64%.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, specifically to a sequential model predictive control method for a six-phase permanent magnet synchronous motor. Background Technology

[0002] With the rapid development of microelectronics and power electronics technologies, the foundation has been laid for motor control systems to break free from the limitations of traditional three-phase topologies. The combination of multiphase generators and multiphase converters provides a new approach for power generation applications with high power or higher reliability requirements. However, current research on multiphase motors still requires in-depth investigation into issues such as bus series connection, third harmonic distortion, and the poor robustness of traditional modulation methods.

[0003] To address harmonic and bus series connection issues in multiphase motors, several effective strategies have been proposed by academia and industry. Domestic research on multiphase motors began as early as 1995, primarily focusing on fault diagnosis, fault-tolerant control, and sensorless operation. Extensive research has also been conducted on analytical calculations of the electromagnetic performance of multiphase motors, as well as harmonic suppression techniques. Additionally, research has been carried out on PWM control and fault-tolerant control theory for five-phase permanent magnet synchronous motors. In summary, domestic research on multiphase motors mainly focuses on electromagnetic analysis and optimization design, primarily analyzing the performance of the motor itself. Analysis combining control technology is relatively limited; for example, issues such as harmonic injection and elimination, whether inductance changes truly decouple the motor, and how to suppress harmonic effects after fault-tolerant control are rarely addressed.

[0004] In international research on control strategies for multiphase motors, a vector control method was proposed in 1971. The main idea is to use the Park coordinate transformation to map a three-dimensional spatial vector in the natural coordinate system to a two-dimensional spatial vector in the rotor-oriented synchronous rotating coordinate system, based on the motor's dynamic mathematical model. This yields the dq-axis current component, decoupling the excitation and torque components. The electromagnetic torque and flux linkage components of the motor are then controlled separately, and finally, motor control is achieved through an inverter and a corresponding PWM modulation strategy. This vector control method can be further divided into current hysteresis-based vector control and space vector decoupling-based vector control. The current hysteresis-based vector control method, for multiphase motors, obtains the dq-axis current in the synchronous rotating coordinate system through the speed outer loop. Then, the inverse Park transformation is used to obtain the current command values ​​for each phase of the multiphase motor. By comparing these with the actual current values ​​obtained from current sensors, a PWM drive signal is obtained to control the multiphase inverter. This control method is applicable to multiphase motors with any number of phases, but it also introduces the drawback of an unstable current hysteresis switching frequency. The vector control method based on spatial vector decoupling was proposed in 1995. Its main idea is to map the six-dimensional space of a standard base of a 30° shifted dual-Y-phase asymmetrical six-phase motor to a new six-dimensional space composed of three new bases. Each new base forms a two-dimensional subspace that is pairwise orthogonal. Of the three subplanes formed, only the fundamental subplane participates in the motor's electromechanical energy conversion; the other two subplanes, one a harmonic subplane and the other a zero-sequence subplane, do not participate in the electromechanical energy conversion process. For multi-phase motors, this method becomes increasingly computationally intensive with the number of phases, as the control algorithm becomes quite complex due to the need to suppress the harmonic and zero-sequence planes.

[0005] In summary, the shortcomings of existing technologies are: traditional space vector modulation strategies require independent design of the main plane and harmonic plane for separate control, the decoupling capability of multiple harmonic planes is insufficient, the dynamic response is lagging, and the control algorithm is complex. Summary of the Invention

[0006] The objective of this invention is achieved through the following technical solutions.

[0007] According to one aspect of the present invention, a sequential model predictive control method for a six-phase permanent magnet synchronous motor is provided, comprising:

[0008] The six-phase permanent magnet synchronous motor is extended into a four-dimensional orthogonal coordinate system through a transformation matrix, and the original plane is transformed into the α-β plane and the xy plane, where the α-β plane is the principal plane and the xy plane is the harmonic plane.

[0009] By using sequential model prediction, a cost function for the model predictive control strategy is constructed. After optimization, the optimal switching sequence is selected to control the on / off state of the IGBTs in the six-phase inverter, which is connected to the six-phase permanent magnet synchronous motor.

[0010] According to another aspect of the present invention, a sequential model predictive control system for a six-phase permanent magnet synchronous motor is also provided, comprising:

[0011] The coordinate transformation module is used to expand the six-phase permanent magnet synchronous motor into a four-dimensional orthogonal coordinate system through the transformation matrix, transforming the original plane into the α-β plane and the xy plane, where the α-β plane is the principal plane and the xy plane is the harmonic plane.

[0012] The model prediction module is used to construct the cost function of the model predictive control strategy through sequential model prediction. After optimization, the optimal switching sequence is selected to control the on and off of the IGBTs in the six-phase inverter, which is connected to the six-phase permanent magnet synchronous motor.

[0013] The advantages of this invention are:

[0014] 1. The sequential model predictive control strategy proposed in this invention can significantly reduce the output current harmonic content (THD).

[0015] 2. The sequential model predictive control strategy proposed in this invention can greatly reduce switching losses and extend the service life of power switching transistors.

[0016] 3. The sequential model predictive control strategy proposed in this invention significantly simplifies the computational load, reducing the computation time by approximately 64% compared to the traditional space vector modulation strategy. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0018] Figure 1 A structural diagram of a six-phase PMSG power generation system according to an embodiment of the present invention is shown.

[0019] Figure 2 A diagram showing the vector space decoupled coordinate system relationship of a six-phase permanent magnet synchronous motor according to an embodiment of the present invention is provided.

[0020] Figure 3 A vector control diagram of a six-phase permanent magnet synchronous motor according to an embodiment of the present invention is shown.

[0021] Figure 4 A spatial structure diagram of a six-phase winding according to an embodiment of the present invention is shown.

[0022] Figure 5 A block diagram of a six-phase permanent magnet synchronous motor control based on sequential model predictive torque control according to an embodiment of the present invention is shown.

[0023] Figure 6 A torque error tracking curve according to an embodiment of the present invention is shown.

[0024] Figure 7 A flux error tracking curve according to an embodiment of the present invention is shown.

[0025] Figure 8 An angular velocity tracking curve according to an embodiment of the present invention is shown.

[0026] Figure 9 The diagram shows the αβ current output under conventional vector control.

[0027] Figure 10 The diagram shows the αβ current output under SMPTC according to an embodiment of the present invention.

[0028] Figure 11 The xy current output diagram under conventional vector control is shown.

[0029] Figure 12 The xy current output diagram under SMPTC according to an embodiment of the present invention is shown. Detailed Implementation

[0030] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0031] Terminology Explanation:

[0032] Six-phase permanent magnet synchronous motors (SP-PMSGs) utilize two sets of three-phase windings with a 30° spatial displacement to form a six-phase stator structure, achieving synchronous operation using a permanent magnet rotor. Their core advantage lies in achieving high power output with low-voltage components, significantly reducing torque ripple, and exhibiting good fault tolerance and reliability. However, six-phase permanent magnet synchronous motors suffer from strong coupling effects across multiple harmonic planes; achieving harmonic current decoupling control becomes a key challenge for optimizing their dynamic performance.

[0033] Vector control strategy (Field-Oriented Control, FOC): For a six-phase permanent magnet synchronous motor, the stator current I in the original six-dimensional coordinate system... a ,I b ,I c ,I u ,I v ,I w The Clark transformation can be used to convert a six-dimensional vector in the natural coordinate system into an equivalent vector in two phase stationary orthogonal coordinate systems. α and I β After undergoing synchronous rotating coordinate transformation, the DC current I in the synchronous rotating coordinate system is obtained. d and I q In this way, the originally complex AC motor model can be equivalent to a DC motor model for control.

[0034] Sequential Model Predictive Torque Control (SMPTC): The sequential model predictive control strategy aims to progressively optimize the voltage vector selection process. By minimizing the cost function in the model predictive control strategy through specific weights and sequences, the optimal voltage vector is selected, thereby determining the switching sequence of the power switches.

[0035] A. System Model

[0036] The overall structure of the six-phase generator system studied in this invention is as follows: Figure 1 As shown in the figure. This design adopts a two-level topology, with a two-level three-phase inverter for grid connection and an LC filter for grid connection.

[0037] Compared to traditional three-phase motors, the winding coupling relationship is more complex. For ease of analysis, the following assumptions are made for a six-phase PMSG:

[0038] (1) Ignore the effects of stator core magnetic saturation and eddy current losses;

[0039] (2) The current is a sinusoidal current and satisfies the three-phase symmetry of A, B, and C, and D, E, and F lag behind A, B, and C by 30°.

[0040] (3) The rotor has no damping winding, and the eddy current effect in the permanent magnet is ignored.

[0041] (4) Ignore the skin effect and temperature effect of the conductor.

[0042] Assuming the current flows positively from inside the generator to the generator terminals, the voltage equations for a six-phase PMSG are as follows:

[0043]

[0044] In the formula, U s —Generator winding terminal voltage vector, U s =[U A U B U C U D U E U F ] T ;

[0045] R s — Resistance matrix, Rs = RsI 6x6 R s The resistance of each phase winding;

[0046] I s — Current vector, I s =[I A I B I C I D I E I F ] T ;

[0047] ψ s —Magnetic flux linkage matrix, ψ s =[ψ A ψ B ψ C ψ D ψ E ψ F ] T ;

[0048] magnetic flux ψ s It can be represented as

[0049] ψ s =L s I s +ψ m #(2)

[0050] Where, ψ m It refers to the magnetic flux linkage between the stator winding and the permanent magnet.

[0051] ψ m =[ψ mA ψ mB ψ mC ψ mD ψ mE ψ mF ] T ;

[0052] L s This is the stator inductance matrix.

[0053] The electromagnetic torque T can be obtained by taking the partial derivative of the magnetic co-energy with respect to the mechanical angle. e ,

[0054]

[0055] In the formula, W c For magnetic co-energy, θ m For mechanical angles, P n This represents the number of pole pairs and the motor power.

[0056] B. Traditional vector control based on vector space decoupling

[0057] Traditional vector control based on vector space decoupling mainly utilizes matrix theory to map a six-dimensional vector formed by six standard basis vectors to a new coordinate system. This coordinate system uses six new vectors as its basis: the dq plane, the z1-z2 plane, and the o1-o2 plane. In other words, it maps the motor current and voltage vectors in the original six-dimensional vector space to three two-dimensional orthogonal subspaces, obtaining a unified transformation matrix. Thus, the mathematical model of the motor can be represented by three independent sets of decoupling equations. For example... Figure 2 The diagram shows the vector space decoupled coordinate system relationship of a six-phase permanent magnet synchronous motor.

[0058] By employing the vector space decoupling method, the Clark transformation matrix for transforming from the natural coordinate system to the stationary coordinate system can be obtained, where the Clark transformation matrix T... αβ for:

[0059]

[0060] For the Clark transformation matrix, the coefficients... This ensures that the magnitude of the variable remains unchanged before and after the transformation.

[0061] After T αβ After the transformation, the voltage, flux linkage, and torque of the six-phase permanent magnet synchronous motor in the natural coordinate system can be converted into three mutually orthogonal planar subspaces, where:

[0062] (1) The fundamental component and the 12k±1 (k=1, 2, 3…) harmonics in the motor are mapped into the dq subspace. The rotating magnetomotive force in the air gap of the motor is mainly formed by this component and is related to the electromechanical energy conversion process of the motor.

[0063] (2) 6k±1 (k=1, 2, 3…) harmonics, such as the fifth and seventh harmonics, are mapped to the z1-z2 subspace.

[0064] (3) The 6k±3 (k=1, 2, 3…) harmonics are mapped to the o1-o2 subspace. They do not participate in the electromechanical energy conversion of the motor and belong to the zero-sequence component.

[0065] After T αβ The transformed spatial components are AC variables, requiring a rotational transformation to convert them into DC components. Since electromechanical energy conversion only occurs in the dq subspace, only the components in this subspace require rotational transformation; therefore, the rotation coordinate matrix T... dq as follows:

[0066]

[0067] Transformation matrix T = T αβ T dq This allows us to obtain the overall conversion formula.

[0068]

[0069] The vector control structure block diagram of a six-phase permanent magnet synchronous motor is as follows: Figure 3 As shown, Figure 3 In this context, ω* is the reference angular velocity, ud* and uq* are the reference voltages of the dq coordinate axis, id* and iq* are the reference currents of the dq coordinate axis, and uα* and uβ* are the reference voltages of the αβ axis.

[0070] For a six-phase permanent magnet synchronous motor, the stator current I in the original six-dimensional coordinate system a ,I b ,I c ,I u ,I v ,I w The Clark transformation can be used to convert a six-dimensional vector in the natural coordinate system into an equivalent vector in two phase stationary orthogonal coordinate systems. α and I β After undergoing synchronous rotating coordinate transformation, the DC current I in the synchronous rotating coordinate system is obtained. d and I q In this way, the originally complex AC motor model can be equivalent to a DC motor model for control.

[0071] C. Sequential model predictive torque control

[0072] Traditional vector control requires separate control of the harmonic plane and the zero-sequence plane, resulting in high computational complexity, poor dynamic response, and low efficiency. Therefore, this invention proposes a novel sequential model predictive torque control strategy. First, the six-phase permanent magnet synchronous motor's six-dimensional system needs to be expanded into a four-dimensional orthogonal coordinate system (α, β, x, y). The spatial structure of the six-phase windings is as follows: Figure 4 As shown.

[0073] By transforming matrix T 6-4 These six vectors can be converted into a four-dimensional orthogonal coordinate system:

[0074]

[0075] The generated αβ plane is the principal plane, primarily responsible for the conversion of fundamental magnetic field energy. This plane is mainly used to regulate torque and flux, while the generated xy plane is used to eliminate harmonic losses. To implement the SMPTC strategy, prediction equations are used to estimate the torque, flux linkage, and harmonic current at prediction time k+2. A two-step prediction time domain is chosen to account for delays caused by sampling, actuation, and calculation, ensuring that the control action applied at time k is aligned with the actual system state during execution, thereby improving the accuracy and dynamic performance of the predictive control strategy. The predicted electromagnetic torque is calculated as follows:

[0076] T e (k+2)=3P(ψ sα(k+2) i sβ(k+2) -ψ sβ(k+2) i sα(k+2) )#(8)

[0077] The predicted stator flux components are given by the following formula:

[0078] ψ sαβ (k+2)=σL s i sαβ(k+2) +K r ψ rαβ (k+2)#(9)

[0079] In equation (9), the rotor flux prediction expression is:

[0080]

[0081] In equations (8), (9), and (10), i s This represents the predicted current value, where P is the generator power and ω is the generator power. r T is the rotor angular velocity. s λ is the system sampling period. r For coefficients, in addition:

[0082] Among them, R r L is the rotor resistance. S For stator leakage inductance, L r For rotor leakage inductance, L m They are mutually inductive. Therefore, a cost function can be constructed.

[0083] The SMPTC strategy proposed in this invention employs a three-stage sequential selection process, progressively refining the voltage vector candidate options. The control stage includes:

[0084] The first step is to minimize the torque error:

[0085] g T =|T * -T e(k+2) |#(11)

[0086] The cost function is designed to minimize the torque prediction error in a six-phase permanent magnet synchronous motor. The formula follows standard predictive control, which tracks the reference variable by minimizing the absolute deviation of the reference variable.

[0087] The second step involves further filtering the voltage vectors obtained in the first step to minimize flux linkage error:

[0088]

[0089] The third step involves further filtering the voltage vectors obtained in the second step to minimize current harmonic errors:

[0090] g xy =|i sxy * -i sxy(k+2 )|#(13)

[0091] The data marked with * in Formulas 11-13 are reference values. The parameters marked with * in Formulas 11-13 are the torque reference value, flux reference value, and current reference value, respectively. The data in (k+2) is the data at time k+2. Here, k+2 means the last two times in model predictive control. A detailed explanation is given in Formula (7).

[0092] The control structure block diagram of the six-phase permanent magnet synchronous motor is as follows: Figure 5 As shown, this invention proposes a sequential model predictive control strategy for a six-phase permanent magnet synchronous motor in a wind power generation system. First, the six-phase permanent magnet synchronous motor is controlled by a transformation matrix T. 6-4 Expanding to a four-dimensional orthogonal coordinate system, the original plane is transformed into an α-β plane and an xy plane, with the α-β plane as the principal plane and the xy plane as the harmonic plane. Through sequential model prediction, a cost function for the model predictive control strategy is constructed. The control sequence can be divided into three steps: the first step is to minimize the torque error in the α-β plane to achieve torque balance; the second step is to minimize the current error in the α-β plane to reduce magnetic flux error; and the third step is to minimize the current error in the xy plane to reduce harmonics. After these three optimization steps, the optimal switching sequence can be selected.

[0093] Figure 5 In this equation, Te* is the torque reference value, ω* refers to the angular velocity reference value, ψs* is the reference magnetomotive force, and isxy* is the reference load current. The data marked with * in formulas 11-13 have the same meaning. N k(k = 1, 2) represents the filtered voltage vector, S opt This represents the selected switching sequence of the inverter's switching transistors. After three steps of optimization using sequential model predictive torque control, the optimal switching sequence is selected to control the on / off state of the IGBTs in the six-phase inverter, achieving the principle of minimizing the cost function. Specific implementation examples:

[0095] A simulation model was built on the Matlab platform to verify the correctness of the theory. The study found that the proposed strategy can stably reduce the harmonic content of the output current, and the control strategy proposed in this invention can accurately control and minimize torque error and flux linkage error.

[0096] Figure 6 and Figure 7 The figures show the torque error tracking curve and the flux error tracking curve, respectively. It can be seen that SMPTC has a good prediction effect and can achieve the optimization of the cost function.

[0097] Figure 8 The curves for angular velocity tracking in a six-phase permanent magnet synchronous motor show fast tracking speed and high accuracy, proving that the SMPTC strategy has good dynamic response.

[0098] To verify that the SMPTC strategy has lower harmonic content in the output current, the output current waveforms of traditional vector space decoupling control and SMPTC are compared. Figure 9-12 It can be seen that after using Sequential Model Predictive Torque Control (SMPTC), the harmonic content of the current output in the αβ plane and xy plane is greatly reduced, which significantly improves the system performance.

[0099] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A sequential model predictive control method for a six-phase permanent magnet synchronous motor, characterized in that, include: The six-phase permanent magnet synchronous motor is extended into a four-dimensional orthogonal coordinate system through a transformation matrix, and the original plane is transformed into the α-β plane and the xy plane, where the α-β plane is the principal plane and the xy plane is the harmonic plane. By using sequential model prediction, a cost function for the model predictive control strategy is constructed. After optimization, the optimal switching sequence is selected to control the on / off state of the IGBTs in the six-phase inverter, which is connected to the six-phase permanent magnet synchronous motor.

2. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, The optimization consists of three steps: The first step is to minimize the torque error in the α-β plane to achieve torque balance; The second step is to minimize the current error in the α-β plane to reduce the magnetic flux error; The third step is to minimize the current error in the xy plane to reduce harmonics.

3. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, In the first step, the formula for minimizing the torque error is as follows: g T =|T * -T e(k+2) |#(11) Among them, g T T represents the torque error. * T represents the torque reference value. e(k+2) This represents the electromagnetic torque at time k+2.

4. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, In the second step, further filtering is performed on the voltage vectors after the first step to minimize flux linkage error: in, Indicates flux linkage error. Indicates the reference value for magnetic flux linkage, ψ sαβ(k+2) This represents the stator flux component at time k+2 in the α-β plane.

5. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, In the third step, further filtering is performed on the voltage vector after the second step to minimize current harmonic errors: g xy =|i sxy * -i sxy(k+2) #(13) Among them, g xy Indicates current harmonic error, i sxy * Indicates the current reference value, i sxy(k+2) This represents the current component in the xy plane at time k+2.

6. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, For the six-phase permanent magnet synchronous motor, the stator current I in the original six-dimensional coordinate system a ,I b ,I c ,I u ,I v ,I w The six-dimensional vector in the natural coordinate system is equivalent to I in two phase stationary orthogonal coordinate systems through Clark transformation. α and I β After undergoing synchronous rotating coordinate transformation, the DC current I in the synchronous rotating coordinate system is obtained. d and I q This allows the AC motor model to be equivalent to a DC motor model for control.

7. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, The αβ plane carries the conversion of the fundamental magnetic field energy and is used to regulate torque and magnetic flux, while the xy plane is used to eliminate harmonic losses.

8. The sequential model predictive control method for a six-phase permanent magnet synchronous motor according to claim 1, characterized in that, The predicted electromagnetic torque is calculated as follows: T e (k+2)=3P(ψ sα(k+2) and sβ(k+2) –ψ sβ(k+2) and sα(k+2) )#(8) The predicted stator flux components are given by the following formula: ψ sαβ (k+2)=σL s I sαβ(k+2) +K r ψ rαβ (k+2)#(9) In equation (9), the rotor flux prediction expression is: In equations (8), (9), and (10), i s This represents the predicted current value, where P is the generator power and ω is the generator power. r T is the rotor angular velocity. s λ is the system sampling period. r For coefficients, in addition: Among them, R r L is the rotor resistance. S For stator leakage inductance, L r For rotor leakage inductance, L m For mutual induction.

9. A sequential model predictive control system for a six-phase permanent magnet synchronous motor, characterized in that, include: The coordinate transformation module is used to expand the six-phase permanent magnet synchronous motor into a four-dimensional orthogonal coordinate system through the transformation matrix, transforming the original plane into the α-β plane and the xy plane, where the α-β plane is the principal plane and the xy plane is the harmonic plane. The model prediction module is used to construct the cost function of the model predictive control strategy through sequential model prediction. After optimization, the optimal switching sequence is selected to control the on and off of the IGBTs in the six-phase inverter, which is connected to the six-phase permanent magnet synchronous motor.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the method as described in any one of claims 1-8.

Images