WO2023173601A1 - 基于离散空间矢量调制的永磁同步电机预测转矩控制方法 - Google Patents

基于离散空间矢量调制的永磁同步电机预测转矩控制方法 Download PDF

Info

Publication number
WO2023173601A1
WO2023173601A1 PCT/CN2022/097284 CN2022097284W WO2023173601A1 WO 2023173601 A1 WO2023173601 A1 WO 2023173601A1 CN 2022097284 W CN2022097284 W CN 2022097284W WO 2023173601 A1 WO2023173601 A1 WO 2023173601A1
Authority
WO
WIPO (PCT)
Prior art keywords
vector
cost function
voltage vector
target
virtual voltage
Prior art date
Application number
PCT/CN2022/097284
Other languages
English (en)
French (fr)
Inventor
杨勇
顾明星
孙俊
樊明迪
谢门喜
肖扬
Original Assignee
苏州大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州大学 filed Critical 苏州大学
Publication of WO2023173601A1 publication Critical patent/WO2023173601A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P27/00—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage
    • H02P27/04—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage using variable-frequency supply voltage, e.g. inverter or converter supply voltage
    • H02P27/06—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage using variable-frequency supply voltage, e.g. inverter or converter supply voltage using dc to ac converters or inverters
    • H02P27/08—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage using variable-frequency supply voltage, e.g. inverter or converter supply voltage using dc to ac converters or inverters with pulse width modulation
    • H02P27/12—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage using variable-frequency supply voltage, e.g. inverter or converter supply voltage using dc to ac converters or inverters with pulse width modulation pulsing by guiding the flux vector, current vector or voltage vector on a circle or a closed curve, e.g. for direct torque control
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
    • H02P21/20—Estimation of torque
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P25/00—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details
    • H02P25/02—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details characterised by the kind of motor
    • H02P25/022—Synchronous motors
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P2207/00—Indexing scheme relating to controlling arrangements characterised by the type of motor
    • H02P2207/05—Synchronous machines, e.g. with permanent magnets or DC excitation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/64—Electric machine technologies in electromobility

Definitions

  • the invention relates to the field of automation control, and in particular to a predictive torque control method, system, device and computer-readable storage medium for a permanent magnet synchronous motor based on discrete space vector modulation.
  • permanent magnet synchronous motors Compared with asynchronous motors, permanent magnet synchronous motors have their unique advantages such as small size, light weight, high power factor and high efficiency. This makes permanent magnet synchronous motors popular in various fields.
  • the pursuit of high-performance control of permanent magnet synchronous motors has always been a research hotspot and difficulty.
  • the high-performance control methods applied to permanent magnet synchronous motors are nothing more than the following: direct torque control, vector control and deadbeat control.
  • Direct torque control based on hysteretic comparators cannot select a more accurate voltage vector, resulting in high torque ripple and high current harmonics.
  • Vector control based on PI controller requires parameter tuning, has weak dynamic adjustment performance and slow dynamic response.
  • Deadbeat control is a method based on system parameters, and the parameters are weak in robustness.
  • Model predictive control can predict the performance of the system in the future in real time, select the optimal control action, and perform rolling optimization. Moreover, using the form of cost function, multiple constraints can be included, enabling simultaneous control of multiple objectives. Model predictive control can be divided into finite set model predictive control and continuous set model predictive control. Compared with continuous set model predictive control, finite set model predictive control can better utilize the discrete characteristics of the inverter and does not require the calculation of space vector modulation, making it easy to implement. Applying finite set model predictive control to permanent magnet synchronous motors has fast dynamic response and strong parameter robustness, which greatly improves the control performance.
  • the traditional finite set model predictive control only obtains the optimal voltage vector in the discrete switching state of the inverter, which results in large torque ripple and flux ripple of the motor accompanied by large current harmonics. Improving steady-state performance has become a major problem in traditional finite set model predictive control.
  • the concept of virtual voltage vector began to appear.
  • the so-called virtual voltage vector is a voltage vector with different amplitudes and different phase angles synthesized from the real voltage vector.
  • Experimental results show that the virtual voltage vector can improve the steady-state performance of model predictive control: reducing the torque ripple and current harmonics of the three-phase permanent magnet synchronous motor.
  • the proposed idea of discrete space vector modulation gives a good explanation for the virtual voltage vector.
  • Discrete space vector modulation is similar to space vector modulation. Within a control cycle, the discrete space vector arranges and combines multiple discrete voltage vectors to obtain a discrete virtual voltage vector located in the inverter control area. The number of these virtual voltage vectors is limited.
  • model predictive control Like traditional model predictive control, the optimal voltage vector in the control set is obtained by enumerating all candidate voltage vectors and minimizing the cost function.
  • Model predictive control based on discrete space vector modulation not only helps to improve steady-state performance, but also solves another shortcoming that the switching frequency is not fixed.
  • discrete space vector modulation is not without its shortcomings. The large number of virtual voltage vectors imposes a huge computational burden on the implementation of prediction equations and cost function minimization.
  • the purpose of the present invention is to provide a predictive torque control method, system, device and computer-readable storage medium for a permanent magnet synchronous motor based on discrete space vector modulation to improve calculation efficiency.
  • the specific plan is as follows:
  • a predictive torque control method for permanent magnet synchronous motors based on discrete space vector modulation including:
  • all small virtual voltage vectors closest to the zero voltage vector are selected from the 8 real voltage vectors output by the inverter and multiple virtual voltage vectors synthesized based on the real voltage vectors;
  • the voltage vectors with uncalculated cost function values within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the virtual voltage vector in the target are substituted into the cost respectively. function to select the target voltage vector with the smallest cost function value.
  • the cost function is:
  • ⁇ represents the preset weight coefficient
  • T e represents the electromagnetic torque
  • ⁇ s represents the stator flux linkage
  • k represents k moment.
  • the closest to the zero voltage vector is selected from the 8 real voltage vectors output by the inverter and the multiple virtual voltage vectors synthesized based on the real voltage vectors.
  • All small virtual voltage vector processes include:
  • the 6 small virtual voltage vectors closest to the zero voltage vector are selected from the 8 real voltage vectors output by the inverter and the 30 virtual voltage vectors synthesized from the real voltage vectors. .
  • the process of substituting each small virtual voltage vector into the cost function generated by the torque prediction model based on discrete space vector modulation and selecting the target small virtual voltage vector with the smallest cost function value includes:
  • the six small virtual voltage vectors are substituted into the cost function generated by the torque prediction model based on discrete space vector modulation, and the target small virtual voltage vector with the smallest cost function value is selected.
  • the voltage whose cost function value has not been calculated is within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the target virtual voltage vector.
  • the vectors are substituted into the cost functions respectively, and the process of selecting the target voltage vector with the smallest cost function value includes:
  • the vector direction of the target small virtual voltage vector is the hypotenuse
  • the real voltage vector is the end point of the hypotenuse
  • the direction of the virtual voltage vector in the target is a right-angled side to construct a right-angled triangle
  • the voltage vectors for which the cost function value has not been calculated and are included in the right triangle are respectively substituted into the cost function, and the target voltage vector with the smallest cost function value is selected.
  • the invention also discloses a permanent magnet synchronous motor predictive torque control system based on discrete space vector modulation, including:
  • the first virtual voltage selection module is used to select the distance zero voltage from 8 real voltage vectors output by the inverter and multiple virtual voltage vectors synthesized from the real voltage vector according to the torque prediction model based on discrete space vector modulation. All small virtual voltage vectors closest to the vector;
  • the cost function calculation module is used to substitute each small virtual voltage vector into the cost function generated by the torque prediction model based on discrete space vector modulation, and select the target small virtual voltage vector with the smallest cost function value;
  • the second virtual voltage selection module is used to substitute two medium virtual voltage vectors adjacent to the target small virtual voltage vector that are not in the same vector direction into the cost function, and select the target medium virtual voltage with the smallest cost function value. vector;
  • the target voltage vector selection module is used to take the zero voltage vector as the origin and the uncalculated cost function within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the virtual voltage vector in the target.
  • the voltage vectors of the values are respectively substituted into the cost function, and the target voltage vector with the smallest cost function value is selected.
  • the cost function is:
  • ⁇ represents the preset weight coefficient
  • T e represents the electromagnetic torque
  • ⁇ s represents the stator flux linkage
  • k represents k moment.
  • the target voltage vector selection module includes:
  • a range determination unit configured to take the zero voltage vector as the origin, the vector direction of the target small virtual voltage vector as the hypotenuse, the real voltage vector as the end point of the hypotenuse, and the virtual voltage vector direction in the target as the right-angled side. Construct a right triangle;
  • the target voltage vector selection unit is used to substitute the voltage vectors included in the right triangle for which the cost function value has not been calculated into the cost function respectively, and select the target voltage vector with the smallest cost function value.
  • the invention also discloses a permanent magnet synchronous motor predictive torque control device based on discrete space vector modulation, which includes:
  • Memory used to store computer programs
  • a processor configured to execute the computer program to implement the aforementioned predictive torque control method of a permanent magnet synchronous motor based on discrete space vector modulation.
  • the invention also discloses a computer-readable storage medium.
  • a computer program is stored on the computer-readable storage medium.
  • the computer program is executed by a processor, the prediction of the permanent magnet synchronous motor based on discrete space vector modulation is realized as mentioned above. Torque control method.
  • the permanent magnet synchronous motor predictive torque control method based on discrete space vector modulation includes: based on the torque prediction model based on discrete space vector modulation, from 8 real voltage vectors output by the inverter and based on the real voltage Select all the small virtual voltage vectors closest to the zero voltage vector among the multiple virtual voltage vectors synthesized by vectors; substitute each small virtual voltage vector into the cost function generated by the torque prediction model based on discrete space vector modulation, and select the cost function The target small virtual voltage vector with the smallest value; substitute the two medium virtual voltage vectors adjacent to the target small virtual voltage vector that are not in the same vector direction into the cost function, and select the target medium virtual voltage with the smallest cost function value.
  • the voltage vectors with uncalculated cost function values within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the virtual voltage vector in the target are substituted into the Describe the cost function and select the target voltage vector with the smallest cost function value.
  • the present invention first selects the target small virtual voltage vector with the lowest cost function value from all small virtual voltage vectors, and then selects the target virtual voltage vector accordingly to limit the final target voltage vector selection range, and finally selects the target small virtual voltage vector and Within the range limited by the virtual voltage vector in the target, the target voltage vector with the smallest cost function value within the range is selected, which greatly reduces the number of calculations for calculating the target voltage vector. There is no need to calculate each voltage vector, which greatly reduces It is time to implement the model predictive torque control method based on space vector modulation.
  • Figure 1 is a schematic flow chart of a predictive torque control method for a permanent magnet synchronous motor based on discrete space vector modulation disclosed in an embodiment of the present invention
  • Figure 2 is a schematic diagram of a two-level voltage source power inverter topology disclosed in an embodiment of the present invention
  • Figure 3 is a schematic diagram of the switching state and the corresponding voltage vector of a two-level voltage source power inverter disclosed in an embodiment of the present invention
  • Figure 4 is a schematic diagram of the spatial distribution of a virtual voltage vector disclosed in an embodiment of the present invention.
  • Figure 5 is a schematic diagram of small virtual voltage vector selection disclosed in an embodiment of the present invention.
  • Figure 6 is a schematic diagram of selecting a medium virtual voltage vector disclosed in an embodiment of the present invention.
  • Figure 7 is a schematic diagram of target virtual voltage vector selection disclosed in an embodiment of the present invention.
  • Figure 8 is a control block diagram of a discrete space vector modulation model predictive torque control disclosed in an embodiment of the present invention.
  • Figure 9 is a schematic structural diagram of a permanent magnet synchronous motor predictive torque control system based on discrete space vector modulation disclosed in an embodiment of the present invention.
  • An embodiment of the present invention discloses a predictive torque control method for a permanent magnet synchronous motor based on discrete space vector modulation. As shown in Figure 1, the method includes:
  • u d and u q are the stator voltage components of the dq axis
  • i d and i q are the stator current components of the dq axis
  • ⁇ d and ⁇ q are the flux linkage components of the dq axis
  • ⁇ f is the permanent magnet flux linkage
  • R s is the stator resistance
  • ⁇ e is the electrical angular velocity
  • p is the number of pole pairs of the motor
  • Te is the electromagnetic torque.
  • T s is the sampling period
  • u d (k+1) are the dq-axis voltage at k+1 moment
  • i d (k+1) are The dq-axis current at time k+1.
  • a torque prediction model based on discrete space vector modulation is obtained based on formula (5) and formula (6).
  • the cost function is the criterion used to select the optimal action in the next control cycle.
  • Reasonable selection of the constraints of the cost function is extremely important for model predictive control. Torque and flux linkage are the two variables most closely related to motor performance, and torque ripple and flux linkage ripple are important indicators for measuring motor performance. Therefore, it is most suitable to select torque and flux linkage as the constraints of the cost function.
  • the cost function g of this application can be defined as:
  • ⁇ represents the preset weight coefficient
  • T e represents the electromagnetic torque
  • ⁇ s represents the stator flux linkage
  • k represents k moment.
  • the inverter that drives a three-phase permanent magnet synchronous motor affects the distribution of the real voltage vector.
  • a two-level voltage source power inverter is used to drive a three-phase permanent magnet synchronous motor.
  • the topology of the motor, in which the inverter is shown in Figure 2.
  • each group of bridge arms has two switching states, and the three groups of bridge arms have a total of 8 switching states.
  • a switching state produces a voltage vector, and the voltage vector output by the two-level inverter can be given by the following expressions (8) and (9):
  • V j U dc (S a +e j2 ⁇ /3 S b +e j4 ⁇ /3 S c ) (8)
  • U dc is the DC bus voltage
  • the corresponding relationship between switching states and voltage vectors is shown in Figure 3.
  • V j (S a , S b , S c ) represents the voltage vector corresponding to each switching state of the two-level voltage source inverter.
  • the core idea of discrete space vector modulation is to make more use of the control area of the inverter, and synthesize additional voltage vectors in the control area through the existing actual voltage vectors.
  • the synthesized voltage vector is called a virtual voltage vector.
  • a sampling period is artificially divided into N parts, and only one real voltage vector is applied to each part.
  • Any virtual voltage vector can be expressed as shown in Equation (10):
  • V real ⁇ V 0 ,V 1 ,V 2 ,V 3 ,V 4 ,V 5 ,V 6 ,V 7 ⁇ (12);
  • V vir represents the virtual voltage vector
  • V real represents the real voltage vector
  • the total number of synthesized virtual voltage vectors n vir (excluding 8 real voltage vectors) is determined by the number N.
  • nvir 3N2 +3N-6 (13)
  • 8 real voltage vectors divide the control area into 6 sectors.
  • the virtual voltage vector can be synthesized from two effective voltage vectors and a zero voltage vector.
  • the 30 synthesized virtual voltage vectors can be divided into three categories: 6 small virtual voltage vectors (V 8 ⁇ V 13 ); 12 medium Virtual voltage vector (V 14 ⁇ V 25 ); 12 large virtual voltage vectors (V 26 ⁇ V 37 ). If the enumeration method is used to select the optimal voltage vector, the prediction equation and the minimization of the cost function need to be calculated 38 times respectively, which takes a lot of time. For this reason, this application first selects all the small virtual voltage vectors for judgment and screening. , in order to reduce the number of subsequent calculations.
  • the small virtual voltage vector is the circle of virtual voltage vectors closest to the zero voltage vector (V 0 , V 7 ), and the subsequent medium virtual voltage vectors and large virtual voltage vectors are in turn, and so on.
  • the absolute value is greater than the virtual voltage vector of the previous stage, the absolute value of the medium virtual voltage vector is greater than the small virtual voltage vector, the absolute value of the large virtual voltage vector is greater than the medium virtual voltage vector, the specific relative position relationship of the large, medium and small virtual voltage vectors can be seen in the figure 5 shown.
  • each small virtual voltage vector is substituted into the cost function generated by the torque prediction model based on discrete space vector modulation, the cost function value of each small virtual voltage vector is calculated, and the small virtual voltage vector with the smallest cost function value is selected.
  • the voltage vector serves as the target small virtual voltage vector and serves as one of the benchmarks.
  • V 8 , V 9 , V 10 , V 11 , V 12 , V 13 are used as the control set for the first step of optimization.
  • selecting two medium virtual voltage vectors that are inconsistent with the vector direction of the target small virtual voltage vector in the dimension of the medium virtual vector is equivalent to selecting two adjacent medium virtual voltage vectors to the left and right of the target small virtual voltage vector, for example, as As shown in Figure 6, assuming that the target small virtual voltage vector V opt1 is V 8 , then the two adjacent medium virtual voltage vectors (V 15 , V 25 ) are also substituted into the cost function (7), and the cost function value is selected
  • the minimum target virtual voltage vector V opt2 is for example V 15 .
  • the vector square of the target small virtual voltage vector is the voltage within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the virtual voltage vector in the target without calculating the cost function value.
  • the vectors are substituted into the cost function respectively, and the target voltage vector with the smallest cost function value is selected. For example, as shown in Figure 7, from all voltage vectors (V 0 , V 1 , V 7 , V 8 , V 14 , V 15 , V 26 ), calculate all the voltage vectors for which the cost function value has not been calculated before.
  • the voltage vectors that need to be calculated this time include V 0 , V 1 , V 7 , V 14 , and V 26 , and select the target voltage vector and substitute it into the next control cycle, thereby predicting the torque of the motor at the next moment and reducing the number of calculations.
  • the number of calculations is reduced from The number of calculations is reduced from 38 times to 13 times, which greatly improves the calculation efficiency.
  • the embodiment of the present invention first selects the target small virtual voltage vector with the lowest cost function value from all small virtual voltage vectors, and then selects the target virtual voltage vector accordingly to limit the final target voltage vector selection range, and finally selects the target small virtual voltage vector from the target small virtual voltage vector.
  • the target voltage vector with the smallest cost function value within the range is selected, which greatly reduces the number of calculations for calculating the target voltage vector and does not need to be performed on each voltage vector. calculation, greatly reducing the time to implement the model predictive torque control method based on space vector modulation.
  • a right-angled triangle can be formed based on the zero voltage vector, the target small virtual voltage vector and the target medium virtual voltage vector to further clarify the calculation range, for example, as shown in Figures 6 and 7, specifically , take the zero voltage vector as the origin, the vector direction of the target small virtual voltage vector is the hypotenuse, the real voltage vector is the end point of the hypotenuse, and the direction of the virtual voltage vector in the target is the right-angled side to construct a right-angled triangle; the uncalculated parts included in the right-angled triangle are The voltage vectors that give the cost function value are substituted into the cost function respectively, and the target voltage vector with the smallest cost function value is selected.
  • the real voltage vector divides the control area into six sectors, and each sector can be further divided into two small right-angled triangle areas, that is, a total of twelve right-angled triangles.
  • the optimal method realizes that the optimal target voltage vector can be obtained by calculating only a right triangle.
  • control block diagram of the proposed model predictive torque control method based on discrete space vector modulation is shown in Figure 8.
  • the motor speed outer loop is controlled by a PI controller.
  • the current control cycle the three-phase stator current is sampled, the current component of the rotating reference system is obtained through coordinate transformation, and the torque and flux linkage are predicted by substituting into the prediction equation.
  • the cost function is used as the evaluation criterion to select a more appropriate voltage vector to apply to the next control cycle.
  • an embodiment of the present invention also discloses a permanent magnet synchronous motor predictive torque control system based on discrete space vector modulation. See Figure 9.
  • the system includes:
  • the first virtual voltage selection module 11 is used to select a distance from zero from 8 real voltage vectors output by the inverter and multiple virtual voltage vectors synthesized from the real voltage vectors according to the torque prediction model based on discrete space vector modulation. All small virtual voltage vectors closest to the voltage vector;
  • the cost function calculation module 12 is used to substitute each small virtual voltage vector into the cost function generated by the torque prediction model based on discrete space vector modulation, and select the target small virtual voltage vector with the smallest cost function value;
  • the second virtual voltage selection module 13 is used to substitute two medium virtual voltage vectors adjacent to the target small virtual voltage vector that are not in the same vector direction into the cost function, and select the target medium virtual voltage vector with the smallest cost function value;
  • the target voltage vector selection module 14 is used to take the zero voltage vector as the origin and the voltage vectors with uncalculated cost function values within the angle range formed by the vector direction of the target small virtual voltage vector and the direction of the virtual voltage vector in the target. Substitute the cost function and select the target voltage vector with the smallest cost function value.
  • the embodiment of the present invention first selects the target small virtual voltage vector with the lowest cost function value from all small virtual voltage vectors, and then selects the target virtual voltage vector accordingly to limit the final target voltage vector selection range, and finally selects the target small virtual voltage vector from the target small virtual voltage vector.
  • the target voltage vector with the smallest cost function value within the range is selected, which greatly reduces the number of calculations for calculating the target voltage vector and does not need to be performed on each voltage vector. calculation, greatly reducing the time to implement the model predictive torque control method based on space vector modulation.
  • the cost function is:
  • ⁇ represents the preset weight coefficient
  • T e represents the electromagnetic torque
  • ⁇ s represents the stator flux linkage
  • k represents k moment.
  • the first virtual voltage selection module 11 is specifically used to synthesize 30 virtual voltage vectors from 8 real voltage vectors output by the inverter and 30 virtual voltage vectors based on the real voltage vector according to the torque prediction model based on discrete space vector modulation. Select the 6 small virtual voltage vectors closest to the zero voltage vector.
  • the cost function calculation module 12 is specifically used to substitute six small virtual voltage vectors into the cost function generated by the torque prediction model based on discrete space vector modulation, and select the target small virtual voltage vector with the smallest cost function value.
  • the target voltage vector selection module 14 may include a range determination unit and a target voltage vector selection unit; where,
  • the range determination unit is used to construct a right-angled triangle with the zero voltage vector as the origin, the vector direction of the target small virtual voltage vector as the hypotenuse, the real voltage vector as the end point of the hypotenuse, and the direction of the virtual voltage vector in the target as the right-angled side;
  • the target voltage vector selection unit is used to substitute the voltage vectors included in the right triangle for which the cost function value has not been calculated into the cost function respectively, and select the target voltage vector with the smallest cost function value.
  • the embodiment of the present invention also discloses a permanent magnet synchronous motor predictive torque control device based on discrete space vector modulation, including:
  • Memory used to store computer programs
  • a processor configured to execute a computer program to implement the aforementioned predictive torque control method of a permanent magnet synchronous motor based on discrete space vector modulation.
  • embodiments of the present invention also disclose a computer-readable storage medium.
  • a computer program is stored on the computer-readable storage medium.
  • the prediction of permanent magnet synchronous motor based on discrete space vector modulation is implemented as mentioned above. Torque control method.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Control Of Ac Motors In General (AREA)

Abstract

提供一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法、系统、装置及计算机可读存储介质，首先从全部小虚拟电压矢量中挑选代价函数值最低的目标小虚拟电压矢量，再依此挑选目标中虚拟电压矢量，限定最终的目标电压矢量挑选范围，最后从目标小虚拟电压矢量和目标中虚拟电压矢量所限定的范围内，挑选出代价函数值最小的目标电压矢量，极大的减少了计算目标电压矢量的计算次数，不需要对每个电压矢量进行计算，大大减少了实现基于空间矢量调制的模型预测转矩控制方法的时间。

Description

基于离散空间矢量调制的永磁同步电机预测转矩控制方法
本申请要求于2022年03月16日提交中国专利局、申请号为202210258167.2、发明名称为“基于离散空间矢量调制的永磁同步电机预测转矩控制方法”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及自动化控制领域，特别涉及一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法、系统、装置及计算机可读存储介质。
背景技术
与异步电机相比，永磁同步电机有其独特的优势如体积小，重量轻，功率因数高，效率高。这使得永磁同步电机在各个领域都饱受青睐。追求永磁同步电机的高性能控制一直都是研究热点和难点。目前，应用于永磁同步电机的高性能控制方法不外乎以下几种：直接转矩控制、矢量控制和无差拍控制。基于滞后比较器的直接转矩控制无法选出较为准确的电压矢量，结果是高转矩脉动和高电流谐波。基于PI控制器的矢量控制需要参数整定，动态调节性能较弱，动态响应慢。无差拍控制是基于系统参数的方法，参数鲁棒性弱。
模型预测控制能够实现实时预测系统未来时刻的表现，选择最优的控制动作，进行滚动优化。而且，使用代价函数的形式，可以包含多个约束条件，能够实现多个目标同时控制。模型预测控制可以分为有限集模型预测控制和连续集模型预测控制。与连续集模型预测控制相比，有限集模型预测控制能够更好地利用逆变器的离散特性，不需要进行空间矢量调制的计算，便于实现。将有限集模型预测控制应用在永磁同步电机具有快速的动态响应和较强的参数鲁棒性，使得控制性能有了较大的提升。传统的有限集模型预测控制只在逆变器的离散开关状态中获取最优电压矢量，这导致电机的转矩脉动和磁链脉动很大同时伴随着较大的电流谐波。提高稳态性能成为传统有限集模型预测控制的一大难题。
虚拟电压矢量的概念开始出现，所谓虚拟电压矢量就是由真实电压矢量合成的具有不同幅值不同相角的电压矢量。实验结果表明虚拟电压矢量能够提高模型预测控制的稳态性能：减小三相永磁同步电机的转矩脉动和电流谐波。提出的离散空间矢量调制的思想对于虚拟电压矢量给出很好的解释。离散空间矢量调制和空间矢量调制类似，在一个控制周期内，离散空间矢量将多个离散电压矢量进行排列组合，得到位于逆变器控制区域的离散虚拟电压矢量。这些虚拟电压矢量的个数是有限的，和传统模型预测控制一样，通过枚举所有候选电压矢量最小化代价函数得到控制集中最优电压矢量。基于离散空间矢量调制的模型预测控制不仅有助于稳态性能的提高，还解决另一个缺点即开关频率不固定。然而，离散空间矢量调制也不是没有缺点，大量的虚拟电压矢量对预测方程和代价函数最小化的实现造成了极大的计算负担。
为此，需要一种计算量更小、计算效率更高的电机控制方法。
发明内容
有鉴于此，本发明的目的在于提供一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法、系统、装置及计算机可读存储介质，提高计算效率。其具体方案如下：
一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法，包括：
根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；
将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；
将与所述目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；
以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的 电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量。
可选的，所述代价函数为：
Figure PCTCN2022097284-appb-000001
式中，λ表示预设的权重系数，T e表示电磁转矩，
Figure PCTCN2022097284-appb-000002
表示参考电磁转矩，ψ s表示定子磁链，
Figure PCTCN2022097284-appb-000003
表示参考定子磁链，k表示k时刻。
可选的，所述根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量的过程，包括：
根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的30个虚拟电压矢量中选取距离零电压矢量最近的6个小虚拟电压矢量。
可选的，所述将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量的过程，包括：
将6个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量。
可选的，所述以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量的过程，包括：
以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为所述斜边终点，所述目标中虚拟电压矢量方向为直角边构建直角三角形；
将所述直角三角形所包括的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的所述目标电压矢量。
本发明还公开了一种基于离散空间矢量调制的永磁同步电机预测转矩控制系统，包括：
第一虚拟电压挑选模块，用于根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成 的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；
代价函数计算模块，用于将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；
第二虚拟电压挑选模块，用于将与所述目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；
目标电压矢量选定模块，用于以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量。
可选的，所述代价函数为：
Figure PCTCN2022097284-appb-000004
式中，λ表示预设的权重系数，T e表示电磁转矩，
Figure PCTCN2022097284-appb-000005
表示参考电磁转矩，ψ s表示定子磁链，
Figure PCTCN2022097284-appb-000006
表示参考定子磁链，k表示k时刻。
可选的，所述目标电压矢量选定模块，包括：
范围确定单元，用于以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为所述斜边终点，所述目标中虚拟电压矢量方向为直角边构建直角三角形；
目标电压矢量选定单元，用于将所述直角三角形所包括的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的所述目标电压矢量。
本发明还公开了一种基于离散空间矢量调制的永磁同步电机预测转矩控制装置，包括：
存储器，用于存储计算机程序；
处理器，用于执行所述计算机程序以实现如前述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法。
本发明还公开了一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机程序，所述计算机程序被处理器执行时实现如前述的基于 离散空间矢量调制的永磁同步电机预测转矩控制方法。
本发明中，基于离散空间矢量调制的永磁同步电机预测转矩控制方法，包括：根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；将与所述目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量。
本发明通过首先从全部小虚拟电压矢量中挑选代价函数值最低的目标小虚拟电压矢量，再依此挑选目标中虚拟电压矢量，限定最终的目标电压矢量挑选范围，最后从目标小虚拟电压矢量和目标中虚拟电压矢量所限定的范围内，挑选出该范围内代价函数值最小的目标电压矢量，极大的减少了计算目标电压矢量的计算次数，不需要对每个电压矢量进行计算，大大减少了实现基于空间矢量调制的模型预测转矩控制方法的时间。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据提供的附图获得其他的附图。
图1为本发明实施例公开的一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法流程示意图；
图2为本发明实施例公开的一种两电平电压源功率逆变器拓扑结构示意图；
图3为本发明实施例公开的一种两电平电压源功率逆变器的开关状态 和对应的电压矢量示意图；
图4为本发明实施例公开的一种虚拟电压矢量的空间分布示意图；
图5为本发明实施例公开的一种小虚拟电压矢量挑选示意图；
图6为本发明实施例公开的一种中虚拟电压矢量挑选示意图；
图7为本发明实施例公开的一种目标虚拟电压矢量挑选示意图；
图8为本发明实施例公开的一种离散空间矢量调制模型预测转矩控制控制框图；
图9为本发明实施例公开的一种基于离散空间矢量调制的永磁同步电机预测转矩控制系统结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例公开了一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法，参见图1所示，该方法包括：
S11：根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量。
具体的，三相永磁同步电机在旋转坐标系(d-q)下的数学模型可由如下方程(1)和(2)表示：
Figure PCTCN2022097284-appb-000007
Figure PCTCN2022097284-appb-000008
式中，u d和u q是d-q轴的定子电压分量，i d和i q是d-q轴的定子电流分量，ψ d和ψ q是d-q轴的磁链分量，ψ f是永磁体磁链，R s是定子电阻，ω e是电角速度，p为电机的极对数，T e是电磁转矩。
对于表贴式永磁同步电机，有L d＝L q＝L s，L d是电机定子电感在d轴的分量，L q是电机定子电感在q轴的分量，L s是电机定子电感，那么磁链方程(3)为：
Figure PCTCN2022097284-appb-000009
具体的，建立预测模型方程时需要将连续状态下的数学模型进行离散化，采用前向欧拉法对(1)进行离散化，可化为如下方程(4)所示：
Figure PCTCN2022097284-appb-000010
上式(4)为定子电流的离散方程。式中，T s是采样周期，u d(k+1),u q(k+1)是k+1时刻的d-q轴电压，i d(k+1),i q(k+1)是k+1时刻的d-q轴电流。
具体的，得到d-q轴的电流之后自然可以得到k+1时刻的电磁转矩T e和定子磁链Ψ s：
Figure PCTCN2022097284-appb-000011
Figure PCTCN2022097284-appb-000012
具体的，基于公式(5)和公式(6)得到基于离散空间矢量调制的转矩预测模型，在有限集模型预测控制中，代价函数是用来选取下一控制周期最优动作的判断标准。合理选取代价函数的约束条件是对模型预测控制是极为重要的。转矩和磁链是和电机性能联系最紧密的两个变量，并且转矩脉动和磁链脉动是衡量电机性能的重要指标。因此，选择转矩和磁链作 为代价函数的约束条件最为适合，考虑到确保跟随参考的能力，本申请的代价函数g可以定义为：
Figure PCTCN2022097284-appb-000013
式中，λ表示预设的权重系数，T e表示电磁转矩，
Figure PCTCN2022097284-appb-000014
表示参考电磁转矩，ψ s表示定子磁链，
Figure PCTCN2022097284-appb-000015
表示参考定子磁链，k表示k时刻。
具体的，驱动三相永磁同步电机工作的逆变器影响着真实电压矢量的分布，例如，本申请实施例中使用一个两电平电压源功率逆变器去驱动一台三相永磁同步电机，其中逆变器的拓扑结构如图2所示。
具体的，在两电平电压源逆变器拓扑中，每组桥臂各有两种开关状态，三组桥臂共有8种开关状态。一个开关状态产生一个电压矢量，两电平逆变器输出的电压矢量可由如下表达式(8)和(9)给出：
V j＝U dc(S a+e j2π/3S b+e j4π/3S c)    (8)
Figure PCTCN2022097284-appb-000016
式中，U dc是直流母线电压，S j(j＝a,b,c)代表开关状态。开关状态和电压矢量的对应关系如图3所示，V j(S a,S b,S c)代表两电平电压源逆变器各开关状态对应的电压矢量。
具体的，离散空间矢量调制的核心思想是更多地利用逆变器的控制区，通过现有的实际电压矢量在控制区合成额外的电压矢量，合成的电压矢量称为虚拟电压矢量。具体来说，将一个采样周期人为地分为N个部分，每个部分只施加一个真实的电压矢量。任何一个虚拟电压矢量都可以表示为如式(10)所示：
Figure PCTCN2022097284-appb-000017
t 1+t 2+…+t N＝T s    (11)；
V real∈{V 0,V 1,V 2,V 3,V 4,V 5,V 6,V 7}   (12)；
式中，V vir表示虚拟电压矢量，V real表示真实电压矢量。
具体的，根据公式(13)所示，合成虚拟电压向量的总数n vir(不包括8个真实电压向量)由数量N确定。
n vir＝3N 2+3N-6   (13)
具体的，在具体的实施场景中，8个真实电压矢量将控制区域划分为6个扇区。在每个扇区中，虚拟电压矢量可以由两个有效电压矢量和一个零电压矢量合成。
具体的，图4展示了N＝3时所有虚拟电压矢量的分布情况，合成出的30个虚拟电压矢量可以分为三类：6个小虚拟电压矢量(V 8～V 13)；12个中虚拟电压矢量(V 14～V 25)；12个大虚拟电压矢量(V 26～V 37)。如果采用枚举法选取最优电压矢量，预测方程以及代价函数最小化需要分别进行38次计算，这需要耗费大量的时间，为此，本申请首先从中挑选出所有小虚拟电压矢量进行判断和筛选，以便后续减少计算次数。
具体的，如图5所示，小虚拟电压矢量便为相对距离零电压矢量(V 0、V 7)最近的一圈虚拟电压矢量，后续的中虚拟电压矢量与大虚拟电压矢量依次类推，分别绝对值大于前一级的虚拟电压矢量，中虚拟电压矢量的绝对值大于小虚拟电压矢量，大虚拟电压矢量的绝对值大于中虚拟电压矢量，大中小虚拟电压矢量的具体相对位置关系可以参见图5所示。
S12：将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量。
具体的，分别将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，计算每个小虚拟电压矢量代价函数值，并从中挑选出代价函数值最小的小虚拟电压矢量作为目标小虚拟电压矢量，作为基准之一。
例如，如图5所示，将6个小虚拟电压矢量(V 8,V 9,V 10,V 11,V 12,V 13)作为第一步寻优的控制集合。用代价函数(7)去评估这6个小虚拟电压矢量，获取代价函数值最小的那个目标小虚拟电压矢量。
S13：将与目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入代价函数，挑选出代价函数值最小的目标中虚拟电压矢 量。
具体的，在中虚拟矢量的维度上挑选两个与目标小虚拟电压矢量的矢量方向不一致的中虚拟电压矢量，相当于挑选目标小虚拟电压矢量左右相邻两个中虚拟电压矢量，例如，如图6所示，假设目标小虚拟电压矢量V opt1为V 8，则将与之相邻的两个中虚拟电压矢量(V 15,V 25)同样代入代价函数(7)，挑选出代价函数值最小的目标中虚拟电压矢量V opt2例如V 15。
S14：以零电压矢量为原点，目标小虚拟电压矢量的矢量方向和目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入代价函数，挑选出代价函数值最小的目标电压矢量。
具体的，最后对零电压矢量为原点，目标小虚拟电压矢量的矢量方目标小虚拟电压矢量的矢量方向和目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入代价函数，挑选出代价函数值最小的目标电压矢量，例如，图7所示，以枚举的形式从所有电压矢量中(V 0,V 1,V 7,V 8,V 14,V 15,V 26)，计算出之前未计算出代价函数值的所有电压矢量，以图5和图6为例，本次需要计算的电压矢量包括V 0、V 1、V 7、V 14、和V 26，并选取该目标电压矢量代入到下一控制周期，从而预测电机下一时刻的转矩，减少了计算次数，以图3所示的虚拟电压矢量分布图为例，将计算次数从38次缩减到了13次，大大提高了计算效率。
可见，本发明实施例通过首先从全部小虚拟电压矢量中挑选代价函数值最低的目标小虚拟电压矢量，再依此挑选目标中虚拟电压矢量，限定最终的目标电压矢量挑选范围，最后从目标小虚拟电压矢量和目标中虚拟电压矢量所限定的范围内，挑选出该范围内代价函数值最小的目标电压矢量，极大的减少了计算目标电压矢量的计算次数，不需要对每个电压矢量进行计算，大大减少了实现基于空间矢量调制的模型预测转矩控制方法的时间。
进一步的，在另一种具体的实施方式中，可以基于零电压矢量、目标小虚拟电压矢量和目标中虚拟电压矢量构成直角三角形，进一步明确计算范围，例如，图6和图7所示，具体的，以零电压矢量为原点，目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为斜边终点，目标中虚拟电压矢量方向为直角边构建直角三角形；将直角三角形所包括的未计算出代 价函数值的电压矢量分别代入代价函数，挑选出代价函数值最小的目标电压矢量。
具体的，参见图5所示，真实电压矢量将控制区域分为六个扇区，每个扇区又可以进一步划分为两个小直角三角形区域即共有十二个直角三角形，通过本申请的寻优方法，实现了仅对一个直角三角形的计算就可以得到最优的目标电压矢量。
具体的，所提的基于离散空间矢量调制的模型预测转矩控制方法的控制框图如图8所示。电机转速外环是由一个PI控制器控制。在当前控制周期，采样得到三相定子电流，通过坐标变换获得旋转参考系的电流分量，代入预测方程式预测得转矩和磁链。采用上述的寻优方法，用代价函数作为评估准则选择一个较为合适的电压矢量应用到下一控制周期。
相应的，本发明实施例还公开了一种基于离散空间矢量调制的永磁同步电机预测转矩控制系统，参见图9所示，该系统包括：
第一虚拟电压挑选模块11，用于根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；
代价函数计算模块12，用于将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；
第二虚拟电压挑选模块13，用于将与目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；
目标电压矢量选定模块14，用于以零电压矢量为原点，目标小虚拟电压矢量的矢量方向和目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入代价函数，挑选出代价函数值最小的 目标电压矢量。
可见，本发明实施例通过首先从全部小虚拟电压矢量中挑选代价函数值最低的目标小虚拟电压矢量，再依此挑选目标中虚拟电压矢量，限定最终的目标电压矢量挑选范围，最后从目标小虚拟电压矢量和目标中虚拟电压矢量所限定的范围内，挑选出该范围内代价函数值最小的目标电压矢量，极大的减少了计算目标电压矢量的计算次数，不需要对每个电压矢量进行计算，大大减少了实现基于空间矢量调制的模型预测转矩控制方法的时间。
其中，代价函数为：
Figure PCTCN2022097284-appb-000018
式中，λ表示预设的权重系数，T e表示电磁转矩，
Figure PCTCN2022097284-appb-000019
表示参考电磁转矩，ψ s表示定子磁链，
Figure PCTCN2022097284-appb-000020
表示参考定子磁链，k表示k时刻。
具体的，第一虚拟电压挑选模块11，具体用于根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的30个虚拟电压矢量中选取距离零电压矢量最近的6个小虚拟电压矢量。
具体的，代价函数计算模块12，具体用于将6个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量。
具体的，目标电压矢量选定模块14，可以包括范围确定单元和目标电压矢量选定单元；其中，
范围确定单元，用于以零电压矢量为原点，目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为斜边终点，目标中虚拟电压矢量方向为直角边构建直角三角形；
目标电压矢量选定单元，用于将直角三角形所包括的未计算出代价函数值的电压矢量分别代入代价函数，挑选出代价函数值最小的目标电压矢量。
此外，本发明实施例还公开了一种基于离散空间矢量调制的永磁同步电机预测转矩控制装置，包括：
存储器，用于存储计算机程序；
处理器，用于执行计算机程序以实现如前述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法。
另外，本发明实施例还公开了一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时实现如前述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法。
最后，还需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
专业人员还可以进一步意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
以上对本发明所提供的技术内容进行了详细介绍，本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (10)

  1. 一种基于离散空间矢量调制的永磁同步电机预测转矩控制方法，其特征在于，包括：
    根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；
    将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；
    将与所述目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；
    以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量。
  2. 根据权利要求1所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法，其特征在于，所述代价函数为：
    Figure PCTCN2022097284-appb-100001
    式中，λ表示预设的权重系数，T e表示电磁转矩，
    Figure PCTCN2022097284-appb-100002
    表示参考电磁转矩，ψ s表示定子磁链，
    Figure PCTCN2022097284-appb-100003
    表示参考定子磁链，k表示k时刻。
  3. 根据权利要求1所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法，其特征在于，所述根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量的过程，包括：
    根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的30个虚拟电压矢量中选取距离零电压矢量最近的6个小虚拟电压矢量。
  4. 根据权利要求3所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法，其特征在于，所述将每个小虚拟电压矢量分别代入基于离 散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量的过程，包括：
    将6个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量。
  5. 根据权利要求1至4任一项所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法，其特征在于，所述以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量的过程，包括：
    以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为所述斜边终点，所述目标中虚拟电压矢量方向为直角边构建直角三角形；
    将所述直角三角形所包括的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的所述目标电压矢量。
  6. 一种基于离散空间矢量调制的永磁同步电机预测转矩控制系统，其特征在于，包括：
    第一虚拟电压挑选模块，用于根据基于离散空间矢量调制的转矩预测模型，从根据逆变器输出的8个真实电压矢量以及根据真实电压矢量合成的多个虚拟电压矢量中选取距离零电压矢量最近的全部小虚拟电压矢量；
    代价函数计算模块，用于将每个小虚拟电压矢量分别代入基于离散空间矢量调制的转矩预测模型生成的代价函数，挑选出代价函数值最小的目标小虚拟电压矢量；
    第二虚拟电压挑选模块，用于将与所述目标小虚拟电压矢量相邻两个不在同一矢量方向上的中虚拟电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标中虚拟电压矢量；
    目标电压矢量选定模块，用于以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向和所述目标中虚拟电压矢量方向形成的夹角范围之内的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的目标电压矢量。
  7. 根据权利要求6所述的基于离散空间矢量调制的永磁同步电机预测转矩控制系统，其特征在于，所述代价函数为：
    Figure PCTCN2022097284-appb-100004
    式中，λ表示预设的权重系数，T e表示电磁转矩，
    Figure PCTCN2022097284-appb-100005
    表示参考电磁转矩，ψ s表示定子磁链，
    Figure PCTCN2022097284-appb-100006
    表示参考定子磁链，k表示k时刻。
  8. 根据权利要求6或7所述的基于离散空间矢量调制的永磁同步电机预测转矩控制系统，其特征在于，所述目标电压矢量选定模块，包括：
    范围确定单元，用于以所述零电压矢量为原点，所述目标小虚拟电压矢量的矢量方向为斜边，真实电压矢量为所述斜边终点，所述目标中虚拟电压矢量方向为直角边构建直角三角形；
    目标电压矢量选定单元，用于将所述直角三角形所包括的未计算出代价函数值的电压矢量分别代入所述代价函数，挑选出代价函数值最小的所述目标电压矢量。
  9. 一种基于离散空间矢量调制的永磁同步电机预测转矩控制装置，其特征在于，包括：
    存储器，用于存储计算机程序；
    处理器，用于执行所述计算机程序以实现如权利要求1至5任一项所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法。
  10. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有计算机程序，所述计算机程序被处理器执行时实现如权利要求1至5任一项所述的基于离散空间矢量调制的永磁同步电机预测转矩控制方法。
PCT/CN2022/097284 2022-03-16 2022-06-07 基于离散空间矢量调制的永磁同步电机预测转矩控制方法 WO2023173601A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210258167.2A CN114584040B (zh) 2022-03-16 2022-03-16 基于离散空间矢量调制的永磁同步电机预测转矩控制方法
CN202210258167.2 2022-03-16

Publications (1)

Publication Number Publication Date
WO2023173601A1 true WO2023173601A1 (zh) 2023-09-21

Family

ID=81780158

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/097284 WO2023173601A1 (zh) 2022-03-16 2022-06-07 基于离散空间矢量调制的永磁同步电机预测转矩控制方法

Country Status (2)

Country Link
CN (1) CN114584040B (zh)
WO (1) WO2023173601A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113206626B (zh) * 2021-06-04 2023-05-02 苏州大学 一种三相永磁同步电机的控制方法、装置、设备及介质
CN114584040B (zh) * 2022-03-16 2023-03-24 苏州大学 基于离散空间矢量调制的永磁同步电机预测转矩控制方法
CN115378324A (zh) * 2022-09-29 2022-11-22 山东大学 一种永磁同步电机控制方法、系统、介质及电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012244796A (ja) * 2011-05-20 2012-12-10 Denso Corp 回転機の制御装置
CN109039189A (zh) * 2018-07-17 2018-12-18 东南大学 基于几何法的永磁同步电机两矢量预测控制系统及方法
CN113258834A (zh) * 2021-05-31 2021-08-13 郑州大学 基于离散空间矢量调制的无差拍模型预测转矩控制方法
CN114584040A (zh) * 2022-03-16 2022-06-03 苏州大学 基于离散空间矢量调制的永磁同步电机预测转矩控制方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012244796A (ja) * 2011-05-20 2012-12-10 Denso Corp 回転機の制御装置
CN109039189A (zh) * 2018-07-17 2018-12-18 东南大学 基于几何法的永磁同步电机两矢量预测控制系统及方法
CN113258834A (zh) * 2021-05-31 2021-08-13 郑州大学 基于离散空间矢量调制的无差拍模型预测转矩控制方法
CN114584040A (zh) * 2022-03-16 2022-06-03 苏州大学 基于离散空间矢量调制的永磁同步电机预测转矩控制方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
OSMAN ILHAM; XIAO DAN; RAHMAN MUHAMMED F.; NORAMBUENA MARGARITA; RODRIGUEZ JOSE: "Discrete Space Vector Modulation Based Model Predictive Flux Control With Reduced Switching Frequency for IM Drive", IEEE TRANSACTIONS ON ENERGY CONVERSION., IEEE SERVICE CENTER, PISCATAWAY, NJ., US, vol. 36, no. 2, 23 October 2020 (2020-10-23), US , pages 1357 - 1367, XP011856175, ISSN: 0885-8969, DOI: 10.1109/TEC.2020.3033356 *

Also Published As

Publication number Publication date
CN114584040A (zh) 2022-06-03
CN114584040B (zh) 2023-03-24

Similar Documents

Publication Publication Date Title
WO2023173601A1 (zh) 基于离散空间矢量调制的永磁同步电机预测转矩控制方法
CN106936356B (zh) 矢量筛选和占空比结合的电机模型预测控制系统及方法
WO2021203495A1 (zh) 用于五相永磁同步电机的无权重系数的双虚拟电压矢量预测转矩控制方法
CN113659898B (zh) 一种双三相永磁同步电机模型预测转矩控制方法
CN109660160B (zh) 一种开关占空比预测转矩控制方法
CN110492821B (zh) 基于不固定矢量作用时间的永磁电机直接磁链控制方法
CN110929448B (zh) 一种永磁同步电机空载电流一致性仿真分析方法
CN109742991B (zh) 基于扇区划分算法的永磁同步电机多步预测控制方法
CN110176893B (zh) 一种单步预测的永磁同步电机模型预测电流控制方法
CN112803861B (zh) 一种永磁同步电机三矢量模型预测控制的无零矢量算法
CN109639192B (zh) 一种表贴式永磁同步电机分时控制系统及分时控制方法
CN107919829B (zh) 一种永磁同步电机模型预测转矩控制方法
CN111262491B (zh) 一种适用于永磁电机系统的增量式直接预测速度控制方法
CN111817627B (zh) 一种低开关频率下双三相感应电机的离散建模与控制方法
WO2022252369A1 (zh) 一种三相永磁同步电机的控制方法、装置、设备及介质
CN113098348B (zh) 一种双三相永磁同步电机预测转矩控制方法
CN111900907A (zh) 一种基于切换点优化的永磁电机模型预测磁链控制方法
Zhang et al. Predictive power control of induction motor drives with improved efficiency
CN112994565B (zh) 一种永磁同步电机三矢量五扇区模型预测电流控制算法
CN113285634B (zh) 基于多步零延迟模型预测的永磁同步电机高速弱磁控制方法及系统
CN115811264A (zh) 一种新型永磁同步电机模型预测直接转矩控制方法
CN115528975A (zh) 基于dsvm的smpmsm驱动系统双目标全局最优无模型预测控制方法
CN114244216A (zh) 一种永磁同步电机参数辨识方法、装置及系统
CN113691179A (zh) 一种基于固定时间的变幂次指数趋近律的永磁同步电机滑模控制方法
Zhou et al. Robust three-vector model predictive torque and stator flux control for PMSM drives with prediction error compensation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22931626

Country of ref document: EP

Kind code of ref document: A1