WO2020227891A1

WO2020227891A1 - Electric motor control method, controller, storage medium and electric motor driving system

Info

Publication number: WO2020227891A1
Application number: PCT/CN2019/086653
Authority: WO
Inventors: 孙天夫; 谢刚; 李慧云; 梁嘉宁
Original assignee: 中国科学院深圳先进技术研究院
Priority date: 2019-05-13
Filing date: 2019-05-13
Publication date: 2020-11-19

Abstract

Disclosed are an electric motor control method, a controller, a storage medium and an electric motor driving system. Controlled variable values, predicted when different alternative voltage vector groups act, can be obtained on the basis of a pre-constructed relationship model; then, by means of the relationship model and the alternative voltage vector groups, an action time group corresponding to each alternative voltage vector group, after one prediction control period elapses and when the predicted controlled variable value is the same as a control command value, is deduced inversely; afterwards, an optimal alternative voltage vector group and a corresponding action time group are screened out; and the optimal alternative voltage vector group and the action time group are then combined into a voltage vector command on the basis of an equivalent vector synthesis principle, and the voltage vector command is decomposed to a corresponding coordinate system in order to obtain a voltage command, so that the voltage command can act on an electric motor or an inverter to drive the electric motor to run. Therefore, no PI controller is needed to control an electric motor, thereby effectively improving the performance of an electric motor and a driving system therefor.

Description

Motor control method, controller, storage medium and motor drive system

Technical field

The invention belongs to the field of power electronics, and in particular relates to a motor control method, a controller, a storage medium and a motor drive system.

Background technique

Motors are widely used in aerospace, industrial automation, and electric vehicles, and the pros and cons of motor drive technology directly determine the overall reliability, stability, and efficiency of the motor drive system.

The currently widely used motor drive technology is mainly based on the Proportional Integral (PI) controller space vector control (Filed Oriented Control, FOC). The basic control principle is shown in Figure 1. The existing FOC technology mainly adopts Two PI current controllers adjust the d-axis and q-axis currents of the motor respectively, and generate corresponding voltage commands to act on the Space Vector Pulse Width Modulation (SVPWM) module to generate modulation waves to drive the inverter , So as to realize the control of the motor. In order to avoid d-axis and q-axis current coupling, the existing FOC control technology also needs to realize the decoupling of the d-axis and q-axis current through the current decoupling module.

However, because the current control of the above FOC technology is generally based on PI controllers, there are problems such as integral saturation, dq axis current coupling, difficult to deal with system constraints, slow dynamic response speed, difficulty in PI parameter tuning, overshoot and overshoot, etc., which seriously affect the motor The performance of its drive system is further improved.

Summary of the invention

The purpose of the present invention is to provide a motor control method, a controller, a storage medium, and a motor drive system, which aims to solve the problem that the performance of the motor and its drive system cannot be effectively improved due to the use of PI controllers in the prior art. problem.

In one aspect, the present invention provides a motor control method, including:

Obtain the observed values of several current controlled quantities on the stator side of the motor at the current moment, the current rotor electrical angular velocity of the motor, and the motor parameters under the operating conditions of the motor in the predictive control period;

Input the predicted control period, the observed value of the current controlled quantity, the motor parameters, the current rotor electrical angular velocity, and a number of candidate voltage vector groups into a relational model to obtain the correspondence of each candidate voltage vector group Each of the current predicted controlled quantity value groups, the candidate voltage vector group includes the basic voltage vector and zero voltage vector output by the inverter, and the current predicted controlled quantity group includes several current controlled quantity predicted values, And make the predicted value of the current controlled quantity equal to the current control command value to obtain each action time group corresponding to each of the candidate voltage vector groups, the action time group including the basic voltage vector and the zero voltage vector Time of action;

Using a pre-established loss function conditioned on the action time group, selecting the action time group that causes the smallest loss function value and the corresponding candidate voltage vector group;

From the selected action time group and the candidate voltage vector group, the next voltage vector currently output by the inverter is obtained, and it is decomposed into the next d-axis voltage value and the next q-axis voltage value .

Further, the observed value of the current controlled quantity is: the current d-axis current value and the current q-axis current value transformed from the current measurement values of the current phases on the stator side of the motor, and the current controlled quantity predicted value Is: the current d-axis predicted current value and the current q-axis predicted current value,

Alternatively, the current controlled quantity observation value is: based on the current stator flux linkage amplitude and the current stator flux linkage angle value observed by the flux observer, the current controlled quantity predicted value is: current stator flux linkage amplitude prediction Value and the predicted value of the current stator flux linkage angle.

Further, the method further includes:

The predictive control period is adjusted.

Further, one of the candidate voltage vector groups corresponds to a stationary coordinate system sector, and the sector corresponds to two of the basic voltage vectors,

The loss function is constructed in the following manner: when the action time corresponding to each of the basic voltage vectors in the action time group is valid, use the candidate sector number corresponding to the candidate voltage vector group, and , The current sector number corresponding to the current voltage sum vector synthesized by the current d-axis voltage component and the current q-axis voltage component to determine the loss function value.

Further, the method further includes:

When the sum of the action time corresponding to each of the basic voltage vectors in the action time group is greater than the predictive control period, the action time corresponding to each of the basic voltage vectors in the action time group is equalized Scale down.

Further, from the selected action time group and the candidate voltage vector group, the next voltage vector currently output by the inverter is obtained, and it is decomposed into the next d-axis voltage value and the next q Shaft voltage value, including:

Obtain the next α-axis voltage value and the next β-axis voltage value currently output in the stationary coordinate system from the selected action time group and the candidate voltage vector group;

From the next α-axis voltage value and the next β-axis voltage value, the next d-axis voltage value and the next q-axis voltage value currently output by the inverter in the rotating coordinate system are obtained.

Further, the motor parameters are called from a data table or obtained through online parameter identification technology, and the motor parameters include one or a combination of the following parameters: d-axis inductance, q-axis inductance, permanent magnet flux, stator Resistance and number of motor pole pairs.

On the other hand, the present invention also provides a motor controller including a memory and a processor, and the processor implements the steps in the above method when the processor executes the program stored in the memory.

On the other hand, the present invention also provides a readable storage medium, the readable storage medium stores a program, and the program is executed by a processor to implement the steps in the above method.

On the other hand, the present invention also provides a motor drive system, including: an inverter, a space vector pulse width modulation module, and the motor controller according to claim 8, wherein the space vector pulse width modulation module combines the The next d-axis voltage value and the next q-axis voltage value are converted into state control commands of the inverter, so as to realize drive control of the motor.

According to the predicted control period to be set, the observed value of the current controlled variable, the motor parameters and the electrical angular velocity of the rotor, the present invention constructs the predicted controlled variable value and the aforementioned candidate voltage vector when different candidate voltage vector groups act The relationship model between the action time of the voltage vector in the group, the combination of the above model and the candidate voltage vector, inversely deduces that after a predictive control cycle, if the predicted controlled value is the same as the control command value, each candidate The voltage vector group corresponds to each action time group, and then the optimal candidate voltage vector group and the corresponding action time group are screened out, and then based on the principle of equivalent vector synthesis, the selected candidate voltage vector group is selected according to each voltage The action time corresponding to the vector synthesizes the voltage vector command and decomposes it into the corresponding coordinate system to obtain the voltage command. Finally, the space pulse width modulation technology is used to apply the above voltage command to the inverter and the motor to drive the motor to run. In this way, no PI controller is needed to control the motor, avoid the current coupling, difficult parameter tuning, integral saturation, fast response and overshoot contradictions in the existing motor control technology, and effectively improve the performance of the motor and its drive system .

Description of the drawings

Figure 1 is a schematic diagram of FOC in the prior art;

FIG. 2 is an implementation flowchart of a motor control method provided by Embodiment 1 of the present invention;

Fig. 3 is a schematic structural diagram of a motor controller provided by a fourth embodiment of the present invention;

4 is a schematic diagram of the structure of a motor drive system provided by the sixth embodiment of the present invention;

Fig. 5 is a schematic diagram of a PI-free motor control strategy provided by a specific application example 1 of the present invention;

6 is a schematic diagram of the circuit topology of the three-phase two-level inverter in specific application example 1 of the present invention;

7 is a schematic diagram of six effective basic voltage vectors and two zero voltage vectors in specific application example 1 of the present invention;

FIG. 8 is a schematic diagram of synthesizing arbitrary voltage vectors through basic vectors in specific application example 1 of the present invention;

Fig. 9 is an experimental test result 1 of the motor drive technology invented in the corresponding experiment and simulation data of specific application example 1 of the present invention;

FIG. 10 is a schematic diagram of the three-phase current of the motor corresponding to the working condition shown in FIG. 9 in the corresponding experiment and simulation data of the specific application example 1 of the present invention;

11 is a schematic diagram of computer simulation and comparison between the motor drive technology invented in the corresponding experiment and simulation data of specific application example 1 of the present invention and the existing space vector control technology based on PI controller;

Fig. 12 is a computer simulation comparison diagram of the response effect of the motor drive technology on the torque step command when T _s =0.01 second, 0.001 second, and 0.0001 second in the corresponding experiment and simulation data of specific application example 1 of the present invention.

FIG. 13 is a schematic diagram of a PI-free motor control strategy provided by the second specific application example of the present invention;

FIG. 14 is the control result of the flux linkage angle of the motor drive technology invented in the corresponding experiment and simulation data of the specific application example 2 of the present invention;

15 is a schematic diagram of the motor flux amplitude control results corresponding to the operating conditions shown in FIG. 14 in the corresponding experiment and simulation data of the specific application example 2 of the present invention;

16 is a schematic diagram of the motor torque control results corresponding to the operating conditions shown in FIG. 14 in the corresponding experiment and simulation data of the specific application example 2 of the present invention;

FIG. 17 is a computer simulation comparison diagram of the response effect of the motor drive technology on the torque step command in the corresponding experiment and simulation data of the specific application example 2 of the present invention at T _s =0.001 second and 0.0001 second.

Detailed ways

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the present invention in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not to limit the present invention.

The specific implementation of the present invention will be described in detail below in conjunction with specific embodiments:

Example one:

FIG. 2 shows the implementation process of the motor control method provided in the first embodiment of the present invention. For ease of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

In step S201, obtain the observed values of several current controlled quantities on the stator side of the motor at the current moment, the current rotor electrical angular velocity of the motor, and the motor parameters under the motor operating conditions in the predicted control period.

[Corrected according to Rule 91 14.05.2019]
In this embodiment, the inverter controls the variable a, b, and c three-phase currents at the stator side of the motor through the state change of its switch tube, thereby controlling the operation of the motor.

The three-phase current can be collected by Hall sensors or resistors. The three-phase current can be obtained through clark transformation and park transformation to obtain the corresponding direct-axis current in the rotating coordinate system, namely the d-axis current, and the quadrature-axis current, namely the q-axis current. In addition, the motor flux observer can also be used to observe the stator side of the motor to obtain the corresponding stator flux linkage amplitude and stator flux angle value in the f-t coordinate system (or called f-m coordinate system). The above-mentioned d and q-axis current values, or the stator flux linkage amplitude and angle values are the observed values of the above-mentioned current controlled variables. The current observed value of the controlled quantity can also be other types of values in other coordinate systems.

The electrical angular velocity can be acquired through position sensor or non-position sensing technology.

In the subsequent relational model, the corresponding motor parameters are also needed, such as: d-axis inductance, q-axis inductance, permanent magnet flux linkage, stator resistance, and motor pole pairs. All or part of these motor parameters can be fixed in advance, or called from a data sheet, or obtained through online parameter identification technology.

It is necessary to construct in advance or in real time a model of the relationship between the action time of the basic voltage vector and the zero voltage vector output by the inverter and the predicted controlled value, that is, the prediction model. In this embodiment, a prediction model under the corresponding coordinate system (ie, dq coordinate system or ft coordinate system, etc.) is established, and the predicted control period, the current observed value of the controlled variable, the motor parameters, and the current rotor electrical angular velocity are used as inputs to obtain each voltage The relationship between the vector group and its action time and the predicted controlled value.

In step S202, the predicted control period, the observed value of the current controlled variable, the motor parameters, the current rotor electrical angular velocity, and a number of candidate voltage vector groups are input into a relational model to obtain each current corresponding to each candidate voltage vector group. Predict the controlled quantity value group. The candidate voltage vector group includes the basic voltage vector and zero voltage vector output by the inverter. The current predicted controlled quantity group includes a number of current controlled quantity predicted values, and the current controlled quantity is predicted The value is equal to the current control command value, and each action time group corresponding to each candidate voltage vector group is obtained. The action time group includes the action time of the basic voltage vector and the zero voltage vector.

In this embodiment, for the inverter, the switching state of the switch tube of the inverter can correspond to a number of basic voltage vectors and zero voltage vectors. The candidate voltage vector group can be composed of basic voltage vector and zero voltage vector. These basic voltage vectors and zero voltage vectors can be projected onto the space voltage vector distribution diagram in the stationary coordinate system (that is, the α-β coordinate system). There are several sectors on the space voltage vector distribution diagram, and the basic voltage vector corresponds to the sector. At the edge position of, the zero-voltage vector corresponds to the origin of the stationary coordinate system, then each sector corresponds to a zero-voltage vector and two basic voltage vectors, and a candidate voltage vector group corresponds to one sector.

Each candidate voltage vector group corresponds to an action time group, which includes the action time of the corresponding basic voltage vector and the action time of the zero voltage vector.

In step S203, the pre-established loss function based on the action time group is used to select the action time group that causes the smallest loss function value and the corresponding candidate voltage vector group.

In this embodiment, each candidate voltage vector group corresponding to each sector and each action time group corresponding to each candidate voltage vector group are screened, and the one that can achieve fast transient response and minimize the fluctuation of the controlled quantity is selected. Alternative voltage vector group and action time group.

The loss function can be judged based on whether the calculated action time is valid or not. Each candidate voltage vector group and each action time group can correspond to a loss function value. According to the comparison of the loss function values, the minimum loss function value is obtained, and the minimum loss is determined The candidate voltage vector group and action time group corresponding to the function value.

In step S204, from the selected action time group and the corresponding candidate voltage vector group, the next voltage vector currently output by the inverter is obtained, and it is decomposed into the next d-axis voltage value and the next q-axis voltage value.

In this embodiment, after the candidate voltage vector group and the action time group corresponding to the minimum loss function value are determined, the candidate voltage vector group and the action time group can be used to obtain the α-axis voltage component and the voltage component in the stationary coordinate system. The β-axis voltage component is further converted from the α-axis voltage component and the β-axis voltage component to obtain the d-axis voltage component and the q-axis voltage component. The obtained d-axis voltage component and q-axis voltage component are output to the inverter for the next stage of motor control.

In the implementation of this embodiment, according to the predicted control period to be set, the current observed value of the controlled variable, the motor parameters and the electrical angular velocity of the rotor, when different candidate voltage vector groups are used, the predicted controlled value and the aforementioned candidate voltage are constructed. The relationship model between the action time of the voltage vector in the vector group, the combination of the above model and the alternative voltage vector, inversely deduces that after a predictive control period, if the predicted controlled value is the same as the control command value, each standby Select each action time group corresponding to the voltage vector group, and then filter out the optimal candidate voltage vector group and the corresponding action time group, and then based on the principle of equivalent vector synthesis, the selected candidate voltage vector group is based on each of them The action time corresponding to the voltage vector synthesizes the voltage vector command and decomposes it into the corresponding coordinate system to obtain the voltage command. Finally, the space pulse width modulation technology is used to apply the above voltage command to the motor to drive the motor to run. In this way, no PI controller is needed to control the motor, avoid the current coupling, difficult parameter tuning, integral saturation, fast response and overshoot contradictions in the existing motor control technology, and effectively improve the performance of the motor and its drive system .

Embodiment two:

Based on the first embodiment, this embodiment further provides the following content:

In this embodiment, the loss function involved in step S203 is constructed in the following manner:

When the action time corresponding to each basic voltage vector in the action time group is valid, the current voltage combined vector composed of the candidate sector number corresponding to the candidate voltage vector group, the current d-axis voltage component and the current q-axis voltage component The corresponding current sector number determines the loss function value.

In step S203, after calculating the candidate voltage vector groups and the action time groups corresponding to each sector, if the action time corresponding to a certain basic voltage vector in the action time group corresponding to a certain sector is less than 0, the action time is considered Invalid, the candidate voltage vector group and action time group will not be selected, but only when the action time corresponding to all the basic voltage vectors in the action time group corresponding to the sector is greater than 0, the action time is valid. Only the voltage vector group and the action time group can be selected.

When there are multiple sets of candidate voltage vector groups and action time groups that are valid for the action time, the loss function value can be calculated by the candidate sector number corresponding to the candidate voltage vector group and the above current sector number, and then proceed Comparison of loss function values.

Of course, in other embodiments, other forms of loss functions that meet the requirements and constraints of fast transient response can also be constructed.

Example three:

Based on the first or second embodiment, this embodiment further provides the following content:

In this embodiment, after step S202 and before step S203, the following processing may be performed:

When the sum of the action time corresponding to each basic voltage vector in the action time group is greater than the predictive control period, the action time corresponding to each basic voltage vector in the action time group is reduced proportionally.

In the action time group calculated in step S202, when the action time corresponding to each basic voltage vector is greater than or equal to 0 and the sum value is greater than the predictive control period, the operability requirement cannot actually be met. At this time, the action The action time corresponding to each basic voltage vector in the time group is reduced proportionally, so that the sum of the action time corresponding to each basic voltage vector in the action time group is not greater than the predictive control period.

Embodiment four:

FIG. 3 shows the structure of the motor controller provided in the fourth embodiment of the present invention. For ease of description, only the parts related to the embodiment of the present invention are shown.

The motor controller of the embodiment of the present invention includes a processor 301 and a memory 302 (including the integration of the two, for example: some special chips have both storage and processing functions. The control program is solidified into a logic circuit of hardware in these chips. These logic circuits themselves record the control program, and no additional memory is needed. When the processor 301 executes the program 303 stored in the memory 302, the steps in the foregoing method embodiments are implemented, such as steps S201 to S204 shown in FIG. .

The program is generally an embedded system program, of course, non-embedded system programs can also be used in some cases.

The inverter and the motor controller of the embodiment of the present invention may be a Micro Controller Unit (MCU), a Digital Signal Processor (Digital Signal Processing, DSP), a Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA) ), programmable logic controllers (Programmable Logic Controller, PLC), chipsets, application specific integrated circuits (Application Specific Integrated Circuits, ASICs) and other devices with signal processing capabilities, or even separate computers or computer networking.

The motor current controller in the embodiment of the present invention is not a PI controller. For the steps implemented when the processor 301 executes the program 303 in the motor controller to implement the above methods, please refer to the description of the above method embodiments, which will not be repeated here.

Embodiment five:

In an embodiment of the present invention, a readable storage medium is provided, and the readable storage medium stores a program, and when the program is executed by a processor, the steps in the foregoing method embodiments are implemented, for example, the steps shown in FIG. 2 S201 to S204.

The readable storage medium in the embodiment of the present invention may include any entity or device or recording medium capable of carrying and storing program code, such as flash memory chips, field programmable gate arrays, ROM/RAM, magnetic disks, optical disks and other memories.

Embodiment 6:

Fig. 4 shows the structure of the motor drive system provided by the sixth embodiment of the present invention. For ease of description, only the parts related to the embodiment of the present invention are shown.

The motor drive system of the embodiment of the present invention includes the following devices that control the motor 401: an inverter 402, a space vector pulse width modulation module 403, and the motor controller 404 described in the fourth embodiment, wherein the space vector pulse width modulation The module 403 converts the next d-axis voltage value and the next q-axis voltage value obtained by the above processing into the state control command of the inverter 402 to realize the drive control of the motor 401.

The space vector pulse width modulation module 403 and the motor controller 404 can be integrated in the same chip or embedded system.

The motor drive system may also include other electronic circuits, such as current acquisition circuits, rotor position (rotation speed) acquisition circuits, protection circuits, etc.

In the motor drive system, the motor can be any motor, such as: built-in permanent magnet synchronous motor, surface mount permanent magnet synchronous motor, switched reluctance motor, linear motor, magnetic field memory motor, rotor excitation motor, induction motor, etc. The three-phase current output device is a two-stage three-phase inverter. The power electronic devices in the inverter can be insulated gate bipolar transistors (IGBT), metal-oxide semiconductor field effect transistors (MOSFET), silicon carbide, etc. Electronic devices.

Specific application example 1:

In the following, the content of each of the above-mentioned embodiments will be exemplified by specific application example 1.

The disadvantages of the motor drive technology based on the PI current controller in the prior art are mainly listed as follows:

1. The motor drive control technology based on PI current controller has the contradiction between fast response and overshoot.

2. The motor drive control technology based on the PI current controller has the problem of AC-DC axis current coupling.

3. Since the motor parameters change with the operating conditions, it is difficult to set the PI controller parameters based on the PI current controller-based motor drive control technology.

4. Traditional PI current control is not easy to deal with system constraints, and there is a phenomenon of integral saturation.

In view of the above-mentioned shortcomings of the existing PI current controller-based motor drive technology, this example proposes a PI-free motor control strategy to realize the control and regulation of the motor current and avoid the current coupling existing in the existing motor drive technology , Parameter is difficult to set, integral saturation, and the contradiction between fast response and overshoot.

This example first derives the influence of the switching state of each leg of the inverter on the dq-axis current according to the formula, and builds a prediction model based on this, and then calculates based on the prediction model within the specified prediction time step (T _s ), If the predicted d-axis and q-axis currents are equal to the d-axis and q-axis current command values, the action time of the basic voltage vector corresponding to the different switch state combinations of the inverter, and then combined with the current motor voltage vector command position, select one The optimal basic vector combination and its corresponding action time are used to calculate the synthesized new voltage vector command based on the selected optimal basic vector combination. Through space pulse width modulation technology (SVPWM), the synthesized motor vector command is applied to the motor to realize the drive control of the motor. This technology can easily adjust the bandwidth of the current controller by changing the predictive control step length (that is, the above-mentioned T _s ) without overshoot and overshoot, and at the same time avoids the problem of the traditional method of d-axis and q-axis current coupling. Finally, the controlled current is used to control the motor torque, and then the motor speed is controlled by controlling the motor torque. The d-axis inductance, q-axis inductance and permanent magnet flux linkage of the motor used in the proposed control technology can be obtained through table lookup or online parameter identification. The online parameter identification strategies involved include but are not limited to: least square method, segmented affine projection, particle swarm and other parameter identification strategies.

This example proposes a PI-free permanent magnet synchronous motor control strategy. The control principle block diagram is shown in Figure 5. The specific details are as follows:

As shown in Figure 6, the three bridge arms of the three-phase two-level inverter have a total of eight switching states, which correspond to the six effective basic voltage vectors and two zero voltage vectors as shown in Figure 7, and can be reversed The clockwise rotation direction divides it into six sectors (sector I-sector VI).

According to the coordinate transformation, the relationship between the switch state and the d and q axis voltages in Figure 6 is:

Among them, S _a (k), S _b (k), S _c (k) are the switching states of the upper arm of the inverter (the upper arm is turned on as 1, otherwise it is 0), and θ _r (k) is the motor Rotor position, V _dc is the DC bus voltage,

Is the d-axis voltage,

Is the q-axis voltage, and k is the sampling time.

The mathematical model of the permanent magnet synchronous motor under the d-q axis is:

Among them, T _s is the time step of the voltage vector action,

They are the projections of a certain voltage vector on the dq coordinate axis in Fig. 7, L _d , L _q , and Ψ _m are the d-axis inductance, q-axis inductance and permanent magnet flux of the motor. R or the following R _s is the stator Resistance, p is the number of pole pairs of the motor,

Is the mechanical angular velocity of the motor rotor, from p and

The electrical angular velocity of the motor rotor can be obtained.

Based on current d and q axis current

You can use (2) and (3) to predict the d and q axis currents of the next time step

Through repeated iterations (2) and (3), the prediction of the d and q axis currents after N _p time steps in the future can be realized.

As shown in Figure 8, according to the principle of space vector pulse width modulation (SVPWM), the two adjacent basic voltage vectors (U _x , U _x+1 ) and the zero vector in Figure 8 are quickly transformed and their action time is adjusted (t ₁ , t ₂ , t ₀ ), any voltage vector U _a within the voltage limit can be synthesized, where x is the number of the sector shown in Figure 7 where the two voltage vectors are located (when x = 6, x +1 is assigned the value 1).

U _x and U _x+1 in Fig. 8 are two adjacent basic voltage vectors belonging to the sector numbered x, where x=1, 2,...,6. By switching between U _x , U _x+1 and the zero vector, the template voltage vector U _a can be synthesized. If the total action time of U _x , U _x+1 and the zero vector are respectively t ₁ , t ₂ and t ₀ , then based on formulas (2) and (3), the predicted dq axis current

With t ₁ , t ₂ and t ₀ and the current dq axis current

The relationship can be expressed as:

Among them, C _d0 , C _q0 , C _{d_x} , C _{q_x} , C _{d_x+1} , C _{q_x+1} are the changes of the d and q axis currents when the zero vector, voltage vector U _x , and voltage vector U _x+1 act, respectively These rates of change can be given by the following formulas according to (2) and (3):

Among them, u _{d_x} , u _{q_x} , u _{d_x+1} , u _{q_x+1} are the components of the first voltage vector (U _x ) and the second voltage vector (U _x+1 ) on the d and q axes, respectively. The coordinate transformation is obtained by (1), where S _a (k), S _b (k), and S _c (k) in (1) correspond to the corresponding U _x and U _x+1 . (4)-(11) is the relationship model between the action time of the inverter output vector and the predicted d-axis and q-axis current values, that is, the prediction model.

Through repeated iterations (4) and (5), the d and q axis current values at the k+N _p time step can be predicted

In order to eliminate the steady-state error of current control, the d and q axis current values at the k+N _p time step are equal to the current command

Have:

For the convenience of description, only the case where the prediction step length is 1 is selected for description in this example description, that is, N _p =1. The case where N _p is equal to other values is similarly derived. Combining (4)-(12) can obtain the action time of the three vectors (two basic voltage vectors and zero voltage vector):

among them,

Q=C _q0 C _{d_x+1} +C _{q_x} C _d0 +C _{q_x+1} C _{d_x} -C _{q_x} C _{d_x+1} -C _{q_x+1} C _d0 -C _q0 k _{d_x}

(15)

If the sum of t ₁ , t ₂ calculated based on the selected U _x , U _x+1 basic voltage vector combination is greater than T _s , then t ₁ , t ₂ are scaled proportionally according to (16) and (17) :

Substituting U _x and U _x+1 of different sectors into (13)-(15) in turn, several groups (t ₁ , t ₂ , t ₀ ) can be obtained, where t ₀ =T _s -t ₁ -t ₂ . For ease of presentation, write them

Where x is 1, 2, ..., 6 or the number of two sectors adjacent to the sector where the current voltage vector is located. Then add the above

Substitute into the loss function (18), where x is

The corresponding voltage vector sector number, x _present is the sector number corresponding to the current voltage vector command. M is a large value to ensure that when

or

When it is less than 0, J ^x will be greater than any

And

The corresponding loss function value when it is greater than or equal to 0.

By choosing the one that minimizes the loss function (J ^x )

The voltage vector command can be synthesized based on its corresponding U _x and U _x+1 for controlling the motor. Will minimize the loss function value (J ^x )

Mark as

Is based on

And its corresponding U _x , U _x+1 , the synthesized voltage vector command can be calculated through (19)-(22)

Projection on the d-axis and q-axis.

Among them, θ _x and θ _x+1 are the spatial phase angles between U _x and U _x+1 and the α axis, respectively, and θ _r is the electrical angle between the d axis and the α axis.

Then the calculated

As d-axis and q-axis voltage commands are sent to the SVPWM module in Figure 5 respectively, the inverter switching state control commands can be generated to realize the drive control of the motor. The content shown in Figure 5 can realize the control of the motor d-axis and q-axis current, and then realize the control of the motor torque. And by controlling the motor torque, the speed of the motor can be further controlled. In addition, by adjusting the size of T _s , the current control bandwidth can be easily adjusted.

Compared with the existing widely used space vector control strategy, the control strategy of this example has the following advantages:

1. Traditional control strategies are designed based on PI controllers, so there are low dynamic response speed, saturation of integrator, direct-axis quadrature-axis current coupling, system constraints are difficult to handle, and there are contradictions between rapid response and overshoot and overshoot, etc. The problem seriously affected the performance of the existing motor control technology. In view of the above shortcomings of the existing motor control technology, the PI-free permanent magnet synchronous motor control technology proposed in this example solves the integrator saturation, direct-axis quadrature-axis current coupling, system constraints that are difficult to handle and the existence of the traditional motor control technology. The contradiction between fast response and overshoot has the advantages of no overshoot, fast current response, etc., and avoids the problem of AC-DC axis current coupling.

2. Since the motor parameters will change with the operating conditions, the existing PI controller-based motor control technology has the problem of difficulty in tuning the PI controller parameters. The PI-free control technology proposed in this example has only one parameter that needs to be adjusted, that is, the predicted time step (T _s ), and the motor current control bandwidth can be adjusted by adjusting the size of T _s , and there is no overshoot and overshoot. . Therefore, the problem of difficulty in tuning the technical parameters of the existing motor control based on the PI controller is avoided.

3. Since the motor control method proposed in this example has only one parameter that needs to be tuned, it is easier to implement and has stronger adaptability to motors with different parameters, and there is no need to re-tune the controller parameters according to different motor parameters. .

Experimental and simulation data corresponding to specific application case 1:

Experiment and simulation data one:

In order to demonstrate the effectiveness and correctness of the control strategy proposed by this patent, the inventor has carried out experimental verification on the control technology proposed by the present invention based on a permanent magnet synchronous servo motor drive system. The measured motor torque, d-axis current, q-axis current and their command values are shown in Figure 9. It can be found that the proposed control method can accurately and quickly control the motor current. The torque measured in Figure 9 has a time delay of about 0.5 seconds relative to the q-axis current. This is because the sampling time of the motor torque sensor is 0.5 seconds, so the measured motor torque is comparable to the actual torque produced by the motor There is a delay of about 0.5 seconds.

The three-phase current and q-axis current of the motor corresponding to the working conditions shown in Fig. 9 are shown in Fig. 10. It can be seen from Figure 10 that the proposed method can quickly and accurately control the motor current with small harmonics, no overshoot and overshoot.

Experimental and simulation data two:

In order to verify that the invented motor drive control method has a faster current response rate and no overshoot and overshoot than the existing method, the inventors respectively compared the invented motor drive technology with the existing PI based on the nonlinear motor system model. The space vector control technology of the controller is simulated and compared by computer. The response of the torque generated by the two control techniques to the step torque command is shown in Figure 11.

As shown in Figure 11, the torque response speed generated by the traditional motor control technology is slow, and there are obvious overshoot and overshoot. The invented motor drive technology produces a fast torque response without overshoot and overshoot.

Experiment and simulation data three:

In order to prove that the invented motor control technology can control the response bandwidth by adjusting the predicted time step (T _s ), the inventor performed computer simulations for 0.01 second, 0.001 second, and 0.0001 second based on the nonlinear motor system model. The response effect of the motor control technology to the torque step command is shown in Figure 12.

As shown in Figure 12, as the predicted time step (T _s ) changes, the torque response bandwidth of the motor also changes. But the generated torque has no overshoot and overshoot. Therefore, the response bandwidth of the motor torque or current can be conveniently controlled by adjusting T _s . Therefore, it is very easy to set the parameters of the motor control technology, and only need to adjust T _s .

Specific application example 2:

The content of each of the above-mentioned embodiments will be exemplarily described below through specific application example two.

The direct flux linkage vector (DFVC) control technology in the prior art mainly controls the motor flux linkage amplitude and the t-axis current through two PI controllers, and generates the f-axis and t-axis voltage commands, and generates the corresponding The voltage commands of the d-axis and q-axis are input to the inverter to generate a control signal to act on the inverter, thereby realizing the control of the motor. The disadvantages are mainly listed as follows:

1. The existing PI controller-based motor flux linkage amplitude and t-axis current control technology has the contradiction between fast response and overshoot.

2. Because the motor parameters change with the operating conditions, the existing PI controller parameter tuning based on the PI controller's motor flux amplitude and t-axis current control technology is difficult.

3. The traditional PI controller-based motor flux linkage amplitude and t-axis current control technology is not easy to deal with system constraints, and there is a phenomenon of integral saturation.

In view of the above-mentioned problems in the existing PI controller-based motor drive control technology, this example proposes a PI-free direct flux vector control strategy for the motor to realize the control and regulation of the motor torque and flux linkage, avoiding the existing motor The contradiction between PI parameters difficult to tune, integral saturation, fast response and overshoot in control technology.

This example first calculates the influence of the switching state of each bridge arm of the inverter on the motor stator flux linkage amplitude (ψ _s ) and the flux linkage angle, that is, the angle between the f-axis and the d-axis, δ, according to the formula. Model, and then calculate based on the prediction model that within the specified prediction time step (T _s ), when the predicted stator flux linkage amplitude and angle are respectively equal to their command values, the inverter's different alternative switch state combinations The corresponding basic voltage vector action time group, combined with the current voltage vector command position of the motor, selects an optimal basic voltage vector combination and its corresponding action time group, and based on the selected optimal basic voltage vector combination and its The corresponding action time group calculates the synthesized new voltage vector command, and uses the space pulse width modulation technology (SVPWM) to apply the synthesized motor vector command to the inverter to realize the control of the motor torque. This technology can easily adjust the bandwidth of torque control by changing T _s without overshoot and overshoot. Finally, control the motor speed by controlling the motor torque. The current motor stator flux linkage amplitude and the feedback value of the flux linkage angle used by the proposed control technology can be obtained by the flux linkage observer.

This example proposes a direct flux vector control technology without PI for a motor. The control principle block diagram is shown in Figure 13, and the details are as follows:

As shown in Figure 6, the three legs of the three-phase two-level inverter have a total of eight switching states, corresponding to the six effective basic voltage vectors and two zero voltage vectors as shown in Figure 7, and can be The counterclockwise rotation direction divides it into six sectors (sector I-sector VI).

According to the coordinate transformation, the relationship between the switch state in Figure 6 and the voltage on the f and t axis is:

Among them, S _a (k), S _b (k), and _Sc (k) are the switching states of the upper arm of the inverter (the upper arm is turned on as 1, otherwise it is 0), and θ _f (k) is the motor The angle between the stator flux linkage and the α axis is θ _f (k) = δ(k)+θ _e (k), where δ(k) is the angle between the f axis and the d axis, and θ _e (k) is the motor Rotor electrical angle. V _dc is the DC bus voltage,

Is the projection of the voltage vector on the f-axis, that is, the f-axis voltage,

Is the projection of the voltage vector on the t-axis, that is, the t-axis voltage, and k is the sampling time.

The mathematical model of the motor under the f-t axis is:

Among them, t _s is the time step of the voltage vector action,

They are the t-axis current and the f-axis current, R is the stator resistance, p is the number of motor pole pairs, and ω _m is the mechanical angular velocity of the motor rotor. From p and ω _{m, the} electrical angular velocity of the motor rotor can be obtained. Based on current f, t axis stator flux linkage amplitude

And the flux linkage angle δ ^k , we can use (2) and (3) to predict the stator flux linkage amplitude and angle after the next time step, namely

And δ ^k+1 . Through repeated iterations (2) and (3), the prediction of the stator flux linkage amplitude and the flux linkage angle after N _p time steps in the future can be achieved.

U _x and U _x+1 in Fig. 8 are two adjacent basic voltage vectors belonging to the sector numbered x, where x=1, 2,...,6. By switching between U _x , U _x+1 and the zero vector, the template voltage vector U _a can be synthesized. If the total action time of U _x , U _x+1 and the zero vector are respectively t ₁ , t ₂ and t ₀ , based on formulas (2) and (3), the stator flux linkage amplitude ψ _s and the flux linkage angle δ The relationship with t ₁ , t ₂ and t ₀ can be expressed as:

δ ^k+1 =δ ^k +C _t0 t ₀ +C _{t_x} t ₁ +C _{t_x+1} t ₂ (5)

among them,

And δ ^k are the current stator flux amplitude and angle of the motor observed by the flux observer. _{_{_{C f0, C t0, C f_x}}} , C t_x, C f_x + 1, C t_x + 1 respectively, when the zero vector, the voltage vector U _x, the voltage vector U _{x + 1} action, f, t axis coordinates lower stator magnetic The chain amplitude and the rate of change of the magnetic chain angle. These rates of change can be given by the following formulas according to (2) and (3):

_{_{Wherein, u f_x, u t_x, u}} f_x + 1, u t_x + 1 are a first voltage vector (U _x) and a second voltage vector (U _{x + 1)} in F, the t-axis component, according to The coordinate transformation is obtained by (1), where S _a (k), S _b (k), and S _c (k) in (1) correspond to the corresponding U _x and U _x+1 .

[Corrected according to Rule 91 14.05.2019]
Through repeated iterations (4) and (5), the stator flux linkage at the k+N _p time step can be predicted

And the flux angle

In order to eliminate the steady-state error of current control, let the stator flux linkage at the k+N _p time step

And the flux angle

Equal to the reference command value (ψ* _s , δ*), there are:

among them,

_{_{Q = C t0 C f_x + 1}} + C t_x C f0 + C t_x + 1 C f_x -C t_x C f_x + 1 -C t_x + 1 C f0 -C t0 C f_x

(15)

Substitute into the loss function (18), where x is

or

When it is less than 0, J ^x will be greater than any

And

The corresponding loss function value when it is greater than or equal to 0.

By choosing the one that minimizes the loss function (J ^x )

The voltage vector command can be synthesized based on the corresponding U _x and U _x+1 to control the inverter to control the motor. Will minimize the loss function value (J ^x )

Mark as

Is based on

Projection on the d-axis and q-axis.

Among them, θ _x and θ _x+1 are the spatial phase angles between U _x and U _x+1 and the α axis, respectively, and θ _e is the electrical angle of the motor rotor.

Then the calculated

As d-axis and q-axis voltage commands are sent to the SVPWM module in Figure 13, the inverter switching state control commands can be generated to realize the drive control of the inverter and the motor. The content shown in Figure 13 can realize the control of the amplitude and angle of the motor stator flux linkage, thereby realizing the control of the motor torque. And by controlling the motor torque, the speed of the motor can be further controlled. In addition, by adjusting the size of T _s , the current control bandwidth can be easily adjusted.

The control strategy of this example has the following advantages compared with the existing widely used DFVC control strategy:

1. The traditional direct flux vector control strategy based on the ft coordinate system is designed based on the PI controller. There are low dynamic response speed, saturation of the integrator, difficult system constraints, and fast response and overshoot and overshoot. Problems such as contradictions have seriously affected the performance of the existing motor control technology. In view of the above shortcomings of the existing motor control technology, the PI-free direct flux vector control technology of the motor proposed in this example solves the integrator saturation, direct-axis quadrature-axis current coupling, system constraints that are not easy to handle and the traditional motor control technology mentioned above. There are problems such as the contradiction between fast response and overshoot. It has the advantages of no overshoot, fast current response, etc., and avoids the problem of AC-DC axis current coupling.

2. Since the motor parameters will change with the operating conditions, the existing PI controller-based motor control technology has the problem of difficulty in tuning the PI controller parameters. The PI-free direct flux vector control technology proposed in this example has only one parameter that needs to be adjusted, that is, the predicted time step (T _s ), and the motor current control bandwidth can be adjusted by adjusting the size of T _s , and there is no Overshoot and overshoot. Therefore, the problem of difficulty in tuning the technical parameters of the existing motor control based on the PI controller is avoided.

3. Since the motor control method proposed in this example has only one parameter that needs to be tuned, it is easy to implement and has stronger adaptability to motors with different parameters. There is no need to re-tune the controller parameters according to the different motor parameters. At the same time, as the motor runs, the motor parameters will change, reducing the accuracy and accuracy of the motor control strategy. The PI-free direct flux vector control technology of the motor proposed in this example does not participate in the feedback calculation because of the motor parameters. Therefore, the influence of motor parameter changes on the motor control strategy is avoided, and the realization and requirements of higher-precision control strategies are easier to achieve.

Experimental and simulation data corresponding to specific application example 2:

Experiment and simulation data one:

In order to demonstrate the effectiveness and correctness of the control strategy proposed by this patent, the applicant conducted computer simulation verification on the control technology proposed by the present invention based on a nonlinear permanent magnet synchronous servo motor drive system model. The magnetic linkage angle command and the magnetic linkage angle control result are shown in Figure 14. It can be found that the proposed control method can accurately and quickly control the motor flux angle.

The control effect of the motor flux linkage amplitude corresponding to the operating conditions shown in Figure 14 is shown in Figure 15. It can be seen from Figure 15 that the proposed method can quickly and accurately control the motor flux linkage amplitude without overshoot and overshoot.

The motor magnetic torque control effect corresponding to the operating conditions shown in Figure 14 is shown in Figure 16. It can be seen that there is no overshoot and overshoot in the torque control of the motor.

Experimental and simulation data two:

In order to prove that the motor control technology invented in this patent can control the response bandwidth by adjusting the predicted time step (T _s ), the inventors respectively simulated 0.001 second and 0.0001 second based on the nonlinear motor system model. The response effect of the motor control technology to the torque step command is shown in Figure 17.

As shown in Figure 17, as the predicted time step (T _s ) changes, the torque response bandwidth of the motor also changes. But the generated torque has no overshoot and overshoot. Therefore, the response bandwidth of the motor torque or current can be conveniently controlled by adjusting T _s . Therefore, it is very easy to set the parameters of the motor control technology, and only need to adjust T _s . The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modification, equivalent replacement and improvement made within the spirit and principle of the present invention shall be included in the protection of the present invention. Within range.

Claims

A motor control method, characterized in that it comprises:

Obtain the observed values of several current controlled quantities on the stator side of the motor at the current moment, the current rotor electrical angular velocity of the motor, and the motor parameters under the operating conditions of the motor in the predictive control period;

Input the predicted control period, the observed value of the current controlled quantity, the motor parameters, the current rotor electrical angular velocity, and a number of candidate voltage vector groups into a relational model to obtain the correspondence of each candidate voltage vector group Each of the current predicted controlled quantity value groups, the candidate voltage vector group includes the basic voltage vector and zero voltage vector output by the inverter, and the current predicted controlled quantity group includes several current controlled quantity predicted values, And make the predicted value of the current controlled quantity equal to the current control command value to obtain each action time group corresponding to each of the candidate voltage vector groups, the action time group including the basic voltage vector and the zero voltage vector Time of action;

Using a pre-established loss function conditioned on the action time group, selecting the action time group that causes the smallest loss function value and the corresponding candidate voltage vector group;

From the selected action time group and the corresponding candidate voltage vector group, the next voltage vector currently output by the inverter is obtained, and it is decomposed into the next d-axis voltage value and the next q-axis Voltage value.
The method according to claim 1, wherein the observed value of the current controlled variable is: the current d-axis current value and the current q-axis current converted from the current measurement values of the current phases on the stator side of the motor The predicted value of the current controlled quantity is: the current d-axis predicted current value and the current q-axis predicted current value,

Alternatively, the current controlled quantity observation value is: based on the current stator flux linkage amplitude and the current stator flux linkage angle value observed by the flux observer, the current controlled quantity predicted value is: current stator flux linkage amplitude prediction Value and the predicted value of the current stator flux linkage angle.
The method of claim 1, wherein the method further comprises:

The predictive control period is adjusted.
The method according to claim 1, wherein one said candidate voltage vector group corresponds to a stationary coordinate system sector, and said sector corresponds to two said basic voltage vectors,

The loss function is constructed in the following manner: when the action time corresponding to each of the basic voltage vectors in the action time group is valid, the candidate sector number corresponding to the candidate voltage vector group is used, and , The current sector number corresponding to the current voltage sum vector synthesized by the current d-axis voltage component and the current q-axis voltage component to determine the loss function value.
The method of claim 1, wherein the method further comprises:

When the sum of the action time corresponding to each of the basic voltage vectors in the action time group is greater than the predictive control period, the action time corresponding to each of the basic voltage vectors in the action time group is equalized Scale down.
The method according to claim 1, characterized in that, from the selected action time group and the candidate voltage vector group, the next voltage vector currently output by the inverter is obtained and decomposed into The next d-axis voltage value and the next q-axis voltage value include:

Obtain the next α-axis voltage value and the next β-axis voltage value currently output in the stationary coordinate system from the selected action time group and the candidate voltage vector group;

From the next α-axis voltage value and the next β-axis voltage value, the next d-axis voltage value and the next q-axis voltage value currently output by the inverter in the rotating coordinate system are obtained.
The method of claim 1, wherein the motor parameters are called from a data table or obtained through online parameter identification technology, and the motor parameters include one or a combination of the following parameters: d-axis inductance, q-axis inductance, permanent magnet flux linkage, stator resistance and motor pole pair number.
A motor controller comprising a memory and a processor, wherein the processor implements the steps in the method according to any one of claims 1 to 7 when the processor executes the program stored in the memory.
A readable storage medium storing a program, wherein the program is executed by a processor to implement the steps in the method according to any one of claims 1 to 7.
A motor drive system, comprising: an inverter, a space vector pulse width modulation module, and the motor controller according to claim 8, wherein the space vector pulse width modulation module converts the next d-axis voltage The value and the next q-axis voltage value are converted into a state control command of the inverter, so as to realize drive control of the inverter and the motor.