Parameter static self-learning method for high-speed dual three-phase permanent magnet synchronous motor driving system

Abstract

The application relates to a parameter static self-learning method of a high-speed double three-phase permanent magnet synchronous motor driving system and relates to the technical field of motor control. The application is aimed at solving the problem that, in the field of motor parameter identification, the non-linear compensation of an inverter cannot accurately compensate for the non-linear voltage of the inverter, thereby leading to low parameter identification accuracy. In the small current area, the application injects a circulating current, so that all phase currents are far away from the small current area of the non-linear inverter, thereby avoiding the influence of the non-linear voltage error of the small current area of the inverter; in the large current area, the voltage errors of the d, q, x and y axes of the motor in the static state are constants, and the voltage error related to the saturation value of the non-linear voltage error of the inverter is eliminated through the difference of different current injection states. In summary, through the special circulating current injection mode and the difference of different injection states, the influence of the non-linear voltage error of the inverter is completely eliminated, and the voltage accuracy in small resistance and inductance identification is ensured.

Description

Technical Field

[0001] This invention belongs to the field of motor control technology. Background Technology

[0002] The stability of current control and the accuracy of position observation in a high-speed dual three-phase permanent magnet synchronous motor drive system depend on the motor's resistance and inductance parameters. Therefore, accurately identifying these parameters is crucial. Compared to conventional three-phase permanent magnet synchronous motors, identifying the resistance and inductance parameters of high-speed dual three-phase permanent magnet synchronous motors is more challenging for two reasons:

[0003] 1. Constrained by rated voltage and line load, high-speed motors have fewer winding turns, larger current, larger wire cross-sectional area, and smaller resistance and inductance parameters.

[0004] 2. Control of dual three-phase motors requires not only the resistance and inductance parameters of the fundamental plane but also the resistance and inductance parameters of the harmonic plane. The harmonic plane parameter is the stator winding leakage inductance, which is smaller than the inductance of the fundamental plane.

[0005] Therefore, the identification of resistance and inductance parameters in high-speed dual three-phase permanent magnet synchronous motors is a problem of identifying small resistance and inductance parameters. The accuracy of small resistance and inductance parameter identification is mainly constrained by the accuracy of voltage. Inverter nonlinear voltage errors, especially those in the low-current region, severely affect the feasibility and accuracy of small resistance and inductance parameter identification methods.

[0006] In the field of motor parameter identification, most existing solutions aim to improve parameter identification accuracy by compensating for inverter nonlinearity. Inverter nonlinearity compensation methods include the trapezoidal method, the sigmoid function method, and high-dimensional lookup table methods. The trapezoidal method and the sigmoid function method fit the voltage error into a zero-crossing piecewise linear curve and an S-curve, respectively. These two methods can characterize the trend of nonlinear voltage, but they cannot accurately compensate for the details of nonlinear voltage error in the low-current region. The high-dimensional lookup table method can accurately compensate for voltage error in the low-current region, but the table creation requires extensive inverter testing, and the implementation process requires storing a large amount of data and using high-dimensional interpolation algorithms, consuming significant resources. Therefore, among existing methods, the trapezoidal method and the sigmoid function method cannot accurately compensate for inverter nonlinear voltage, while the high-dimensional lookup table method has high application costs and poor universality. Summary of the Invention

[0007] This invention addresses the problem that in the field of motor parameter identification, when using inverter nonlinearity compensation to improve parameter identification accuracy, it is impossible to accurately compensate for the inverter nonlinear voltage, resulting in low parameter identification accuracy. The invention provides a static self-learning method for the resistance and inductance parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system based on circulating current injection and state differential.

[0008] A static self-learning method for parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system is described below.

[0009] Step 1: Construct a four-dimensional current vector control structure for a dual three-phase permanent magnet synchronous motor based on a proportional-integral-resonant current controller, and use this four-dimensional current vector control structure as the current controller for the high-speed dual three-phase permanent magnet synchronous motor drive system.

[0010] Step 2: When the command currents for both the d-axis and q-axis are 0, inject a stepped circulating current with a circulating vector angle of 0 into the high-speed dual three-phase permanent magnet synchronous motor drive system. Iterate through each circulating current value of this stepped current and record the command voltages for the d-axis, q-axis, x-axis, and y-axis of the high-speed dual three-phase permanent magnet synchronous motor, as well as the six-phase feedback current, when the feedback current reaches a stable state at each circulating current value. Perform inverse rotational coordinate transformation and inverse vector space decoupling transformation on the command voltages for the d-axis, q-axis, x-axis, and y-axis sequentially to obtain the six-phase command voltages in the natural coordinate system. Calculate the inverter's nonlinear voltage error saturation value using the six-phase command voltage and six-phase feedback current corresponding to the last circulating current. Obtain the characteristic current I of the nonlinear voltage error using this inverter nonlinear voltage error saturation value and the recorded command voltages and six-phase feedback currents for the d-axis, q-axis, x-axis, and y-axis corresponding to each circulating current value. cri ;

[0011] Step 3: When the command currents for both the d-axis and q-axis are 0, inject a circulating current with an amplitude of [value missing] into the high-speed dual three-phase permanent magnet synchronous motor drive system. A DC circulating current with a circulating vector angle of γ1 is used to record the command voltages of the d-axis, q-axis, x-axis, and y-axis of the high-speed dual three-phase permanent magnet synchronous motor and the feedback currents of the d-axis, q-axis, x-axis, and y-axis when the feedback current reaches a stable state. This data is then used as the basic state dataset, and steps four, five, and six are executed respectively.

[0012] Step 4: With the command currents on the d-axis and q-axis and the angle of the injected circulating current vector remaining unchanged, a sinusoidal command current is superimposed on the DC circulating current described in Step 3, and the amplitude I of this sinusoidal command current is... HF satisfy Record the x-axis command voltage and x-axis feedback current of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Subtract these x-axis command voltage and x-axis feedback current from the corresponding data in the basic state dataset. Input the difference into the parameter solver to obtain the x-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor. x ;

[0013] Step 5: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, a sinusoidal command current is also injected into the d-axis, and the amplitude of the sinusoidal command current is I. HF satisfy Record the d-axis command voltage and d-axis feedback current of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Subtract these d-axis command voltage and d-axis feedback current from the corresponding data in the basic state dataset. Input the difference into the parameter solver to obtain the d-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor. d and stator resistance R s ;

[0014] Step Six: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, a sinusoidal command current is also injected into the q-axis, and the amplitude of the sinusoidal command current is I. HF satisfy Record the q-axis command voltage and q-axis feedback current of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Subtract these q-axis command voltage and q-axis feedback current from the corresponding data in the basic state dataset. Input the difference into the parameter solver to obtain the q-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor. q ;

[0015] The parameter solver includes a second-order generalized integrator, an orthogonal signal generator, and a recursive least squares stage. The second-order generalized integrator is used to filter the input signal, the orthogonal signal generator is used to differentiate and filter the input signal, and the recursive least squares stage performs least squares on the filtered input signal and the differential result of the input signal to obtain the output of the parameter solver.

[0016] Furthermore, the control flow of the above four-dimensional current vector control structure is as follows:

[0017] Vector space decoupling transformation is performed on the six-phase command current of the high-speed dual three-phase permanent magnet synchronous motor to obtain the α-axis and β-axis command currents. Rotational coordinate transformation is then performed on the α-axis and β-axis command currents to obtain the d-axis and q-axis command currents. The x-axis and y-axis command currents are calculated using the circulating current command amplitude and circulating current vector angle of the high-speed dual three-phase permanent magnet synchronous motor drive system. The difference between the command currents of the d-axis, q-axis, x-axis, and y-axis and the corresponding feedback currents is used to obtain the command voltages of the d-axis, q-axis, x-axis, and y-axis through a proportional-integral-resonant current controller. Vector space decoupling inverse transformation, rotational coordinate transformation, and zero-sequence injection modulation are then performed on the command voltages of the d-axis, q-axis, x-axis, and y-axis in sequence to obtain the inverter switching signal. The switching signal is applied to the inverter of the high-speed dual three-phase permanent magnet synchronous motor drive system to achieve four-dimensional current vector control of the high-speed dual three-phase permanent magnet synchronous motor.

[0018] Furthermore, the specific method for calculating the x-axis and y-axis command currents using the circulating current command amplitude and circulating current vector angle of the high-speed dual three-phase permanent magnet synchronous motor drive system is as follows:

[0019]

[0020] in, and These are the command currents for the x-axis and y-axis, respectively. γ represents the circulating current command amplitude of the high-speed dual three-phase permanent magnet synchronous motor drive system, and γ is the circulating current vector angle.

[0021] Furthermore, the transfer function PIR(z) of the above proportional-integral-resonant current controller is expressed as follows:

[0022]

[0023] Where z is a complex variable of the transfer function PIR(z), and K P K I and K R These are the proportional, integral, and resonant coefficients, respectively, and K. d T is the delay compensation coefficient. s ω is the sampling time. R It is the resonant frequency.

[0024] Furthermore, in step two above, the nth order circulation... The expression is:

[0025]

[0026] Where n = 1, 2, ..., N, N is the total number of steps in the stepped circulation, I N This is the rated current of a high-speed dual three-phase permanent magnet synchronous motor.

[0027] Furthermore, in step two above, the six-phase command voltage in the natural coordinate system is obtained according to the following formula:

[0028]

[0029] in, Let O2 and E2 be the six-phase command voltages corresponding to the nth order circulating current in the natural coordinate system, respectively. O2 is a 2×2 all-zero matrix, and E2 is a 2×2 identity matrix. and These represent the d-axis, q-axis, x-axis, and y-axis command voltages corresponding to the nth-order circulating current, where n = 1, 2, ..., N, and N is the total number of steps in the stepped circulating current.

[0030] T VSD The vector space decoupling transformation matrix has

[0031] T Park Let be the rotation coordinate transformation matrix, and have θ eIt is the angle between the d-axis and the axis of the A-phase winding.

[0032] Furthermore, the inverter nonlinear voltage error saturation value is obtained according to the following formula:

[0033]

[0034] Where ΔU is the inverter's nonlinear voltage error saturation value. and These are the command voltages for phases A, B, D, and E corresponding to the Nth order circulating current, i A-N and i D-N These are the feedback currents of phases A and D corresponding to the Nth order circulating current, respectively.

[0035] Furthermore, the specific method for obtaining the characteristic current of the nonlinear voltage error using the inverter's nonlinear voltage error saturation value and the recorded circulating current values ​​of each order corresponding to the d-axis, q-axis, x-axis, and y-axis command voltages and six-phase feedback currents is as follows:

[0036] S1: Determine if the following expression is true. If yes, execute S2; otherwise, execute S3.

[0037]

[0038] in, and These are the command voltages for phases A, B, D, and E corresponding to the nth order circulating current, i. A-n and i D-n These are the A-phase and D-phase feedback currents corresponding to the nth-order circulating current, respectively.

[0039] S2: The E-phase feedback current i corresponding to the nth order circulating current E-n As characteristic current I cri ,

[0040] S3: Set n = n + 1 and return to S1.

[0041] Furthermore, in step three above,

[0042] The amplitude of the DC circulating current

[0043] The circulation vector angle of the DC circulating current

[0044] The basic state dataset

[0045] The above, I N This refers to the rated current of a high-speed dual three-phase permanent magnet synchronous motor.

[0046] and These represent the d-axis, q-axis, x-axis, and y-axis command voltages of the high-speed dual three-phase permanent magnet synchronous motor when a DC circulating current is injected.

[0047] i d0 i q0 i x0 and i y0 These are the d-axis, q-axis, x-axis, and y-axis feedback currents of the high-speed dual three-phase permanent magnet synchronous motor when a DC circulating current is injected;

[0048] The sinusoidal command current superimposed on the DC circulating current in step four, the sinusoidal command current injected into the d-axis in step five, and the sinusoidal command current injected into the q-axis in step six are all I. HF cos(ω HF t),

[0049] I HF ω represents the amplitude of the superimposed sinusoidal command current. HF t represents the frequency of the sinusoidal command current and t represents time.

[0050] Furthermore, the transfer function SOGI(z1) of the second-order generalized integrator in the above parameter solver is expressed as:

[0051]

[0052] The transfer function QSG(z2) of the orthogonal signal generator in the above parameter solver is expressed as follows:

[0053]

[0054] The identification equation for the recursive least squares stage in the above parameter solver is:

[0055]

[0056] Where z1 is the variable of the transfer function SOGI(z1), z2 is the variable of the transfer function QSG(z2), and K GI ω is the integral coefficient. HF T is the frequency of the injected signal. s Sampling time, Δi is the difference between the r-axis command voltage and the corresponding data in the basic state dataset. r L is the difference between the r-axis feedback current and the corresponding data in the basic state dataset. r R is the r-axis inductance. s For the stator resistance, D(Δi) r ) is Δi r The differential value, and D(Δi) r )=ω HF *QSG(Δi r), r = d, q, x, y.

[0057] The present invention achieves the following beneficial effects through the parameter static self-learning method of the above-mentioned high-speed dual three-phase permanent magnet synchronous motor drive system:

[0058] 1. This invention eliminates the need for inverter nonlinearity identification, simplifies the static self-learning process of motor parameters, and has greater versatility.

[0059] 2. This invention completely eliminates the influence of inverter nonlinear voltage error and can achieve accurate identification of small resistance and inductance parameters.

[0060] 3. This invention identifies the AC resistance and inductance parameters of a motor, and can be used for self-learning of motor parameters at different excitation frequencies. Attached Figure Description

[0061] Figure 1 A flowchart of a static self-learning method for parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system;

[0062] Figure 2 A block diagram illustrating the principle of the static self-learning method for parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system.

[0063] Figure 3 This is a block diagram of a resistor and inductor parameter solver based on a generalized integrator and recursive least squares.

[0064] Figure 4 The d, q, x, and y axis current waveforms of a high-speed dual three-phase permanent magnet synchronous motor drive system using a static self-learning method for parameter parameters.

[0065] Figure 5 The voltage command waveforms of d, q, x, and y for the parameter static self-learning method of the high-speed dual three-phase permanent magnet synchronous motor drive system are shown.

[0066] Figure 6 The phase current waveform is shown in the static self-learning method for parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0068] Static self-learning of motor parameters is a special type of offline motor parameter identification method with more constraints. Static self-learning of motor parameters requires ensuring parameter identification can be completed at any initial rotor position while maintaining rotor stillness throughout the identification process. If the inverter nonlinearity identification stage is included as part of the static self-learning method, then the inverter nonlinearity identification stage must also guarantee the conditions of identification at any initial position and rotor stillness during the identification process. However, most existing accurate inverter nonlinearity identification and compensation methods fail to meet the conditions of static self-learning. To address these issues, [the following text is missing from the original] Figures 1 to 6 This embodiment specifically describes the parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system. The specific method is as follows:

[0069] Step 1: Construct a four-dimensional current vector control structure for a dual three-phase permanent magnet synchronous motor based on a proportional-integral-resonant current controller. The control flow of the four-dimensional current vector control structure is as follows:

[0070] Vector space decoupling transformation is performed on the six-phase command current of the high-speed dual three-phase permanent magnet synchronous motor to obtain the α-axis and β-axis command currents. Rotation coordinate transformation is then performed on the α-axis and β-axis command currents to obtain the d-axis and q-axis command currents.

[0071] The circulating current command amplitude of the high-speed dual three-phase permanent magnet synchronous motor drive system The command currents for the x-axis and y-axis are calculated using the circulating current vector angle γ, respectively. and

[0072]

[0073] The difference between the command current and the corresponding feedback current for the d-axis, q-axis, x-axis, and y-axis is used to obtain the command voltages for the d-axis, q-axis, x-axis, and y-axis, respectively, through a proportional-integral-resonant current controller. The transfer function PIR(z) of the proportional-integral-resonant current controller is expressed as follows:

[0074]

[0075] Where z is a variable in the transfer function PIR(z), and K P K I and K R These are the proportional, integral, and resonant coefficients, respectively, and K. d T is the delay compensation coefficient. s ω is the sampling time. R It is the resonant frequency.

[0076] The command voltages of the d-axis, q-axis, x-axis, and y-axis are sequentially subjected to inverse vector space decoupling transformation, rotating coordinate transformation, and zero-sequence injection modulation to obtain the switching signals of the inverter.

[0077] The switching signal is applied to the inverter of the high-speed dual three-phase permanent magnet synchronous motor drive system to achieve four-dimensional current vector control of the high-speed dual three-phase permanent magnet synchronous motor.

[0078] After the structure is completed, the four-dimensional current vector control structure will be used as the current controller for the high-speed dual three-phase permanent magnet synchronous motor drive system.

[0079] Step 2: When the command currents for both the d-axis and q-axis are 0, inject a stepped circulating current with a circulating vector angle of 0 into the high-speed dual three-phase permanent magnet synchronous motor drive system. The nth order circulating current... N is the total number of steps in the stepped circulation, I N This is the rated current of a high-speed dual three-phase permanent magnet synchronous motor.

[0080] Iterate through each circulating current value of the stepped circulating current, and record the command voltages of the d-axis, q-axis, x-axis and y-axis of the high-speed dual three-phase permanent magnet synchronous motor and the six-phase feedback current when the feedback current reaches a stable value under each circulating current value.

[0081] By sequentially performing inverse rotational coordinate transformation and inverse vector space decoupling transformation on the command voltages along the d-axis, q-axis, x-axis, and y-axis, the six-phase command voltages in the natural coordinate system are obtained, as shown in the following formula:

[0082]

[0083] in, Let O2 and E2 be the command voltages of phases A, B, C, D, E, and F corresponding to the nth order circulating current in the natural coordinate system, respectively. O2 is a 2×2 all-zero matrix, and E2 is a 2×2 identity matrix. and These are the command voltages for the nth order circulating current along the d-axis, q-axis, x-axis, and y-axis, respectively.

[0084] T VSD The vector space decoupling transformation matrix has

[0085] T Park Let be the rotation coordinate transformation matrix, and have θ e It is the angle between the d-axis and the axis of the A-phase winding.

[0086] The inverter nonlinear voltage error saturation value ΔU is calculated using the six-phase command voltage and six-phase feedback current corresponding to the last circulating current:

[0087]

[0088] i A-N and i D-N These are the feedback currents of phases A and D corresponding to the Nth order circulating current, respectively.

[0089] The characteristic current I of the nonlinear voltage error is obtained by using the inverter's nonlinear voltage error saturation value and the recorded circulating current values ​​of each order, corresponding to the d-axis, q-axis, x-axis, and y-axis command voltages and the six-phase feedback current. cri Characteristic current I cri This refers to the minimum current at which the inverter's nonlinear voltage error reaches its saturation value, and it also marks the boundary between the inverter's nonlinear low-current region and the saturation region. Therefore, the specific method for obtaining the characteristic current of the nonlinear voltage error is as follows:

[0090] S1: Determine if the following expression is true. If yes, execute S2; otherwise, execute S3.

[0091]

[0092] S2: The E-phase feedback current i corresponding to the nth order circulating current E-n As characteristic current I cri ,

[0093] S3: Set n = n + 1 and return to S1.

[0094] Step 3: When the command currents for both the d-axis and q-axis are 0, inject a circulating current with an amplitude of [value missing] into the high-speed dual three-phase permanent magnet synchronous motor drive system. The circulation vector angle is DC circulating current, I N This is the rated current of a high-speed dual three-phase permanent magnet synchronous motor.

[0095] Record the command voltages and feedback currents along the d, q, x, and y axes of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state, and use this data as the basic state dataset. Then perform steps four, five, and six respectively. and These represent the d-axis, q-axis, x-axis, and y-axis command voltages of the high-speed dual three-phase permanent magnet synchronous motor when a DC circulating current is injected. d0 i q0 i x0 and i y0 These are the d-axis, q-axis, x-axis, and y-axis feedback currents of the high-speed dual three-phase permanent magnet synchronous motor when a DC circulating current is injected.

[0096] Step 4: With the command currents and injected circulating current vector angles on the d-axis and q-axis remaining unchanged, superimpose a sinusoidal command current I onto the DC circulating current described in Step 3.HF cos(ω HF t), and the amplitude I of the sinusoidal command current. HF satisfy Record the x-axis command voltage of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. and x-axis feedback current i x The x-axis command voltage and x-axis feedback current i x The difference is calculated by subtracting the corresponding data from the base state dataset, and the resulting difference is... and i x -i x0 The x-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor is obtained by inputting the parameters into the parameter solver. x L x =L y =L z .

[0097] Step 5: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, a sinusoidal command current I is also injected into the d-axis. HF cos(ω HF t), and the amplitude I of the sinusoidal command current. HF satisfy Record the d-axis command voltage of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. and d-axis feedback current i d , the d-axis command voltage and d-axis feedback current i d The difference is calculated by subtracting the corresponding data from the base state dataset, and the resulting difference is... and i d -i d0 The input is fed into the parameter solver to obtain the d-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor. d and stator resistance R s .

[0098] Step Six: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, a sinusoidal command current I is also injected into the q-axis. HF cos(ω HF t), and the amplitude I of the sinusoidal command current. HF satisfy Record the q-axis command voltage of the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. and q-axis feedback current i q , the q-axis command voltage and q-axis feedback current i q Subtract the corresponding data from the base state dataset, and obtain the difference value. and i q -i q0 The input is fed into the parameter solver to obtain the q-axis inductance L of the high-speed dual three-phase permanent magnet synchronous motor. q .

[0099] The above, I HF ω represents the amplitude of the superimposed sinusoidal command current. HF t represents the frequency of the sinusoidal command current and t represents time.

[0100] The parameter solver mentioned above includes a second-order generalized integrator, an orthogonal signal generator, and a recursive least squares stage.

[0101] A second-order generalized integrator is used to filter the input signal. Its transfer function SOGI(z1) is expressed as follows:

[0102]

[0103] The quadrature signal generator is used to differentiate and filter the input signal. Its transfer function QSG(z2) is expressed as follows:

[0104]

[0105] The recursive least squares stage performs least squares on the filtered input signal and its derivative to obtain the output of the parameter solver. The identification equation for the recursive least squares stage is:

[0106]

[0107] z1 is a variable in the transfer function SOGI(z1), z2 is a variable in the transfer function QSG(z2), and K GI ω is the integral coefficient. HF T is the frequency of the injected signal. s Sampling time, Δi is the difference between the r-axis command voltage and the corresponding data in the basic state dataset. r L is the difference between the r-axis feedback current and the corresponding data in the basic state dataset. r For the r-axis inductance, D(Δi) r ) is Δi r The differential value, and D(Δi) r )=ω HF *QSG(Δi r ), r = d, q, x, y.

[0108] The inverter's nonlinear voltage error exhibits strong nonlinearity in the low-current region and saturation characteristics in the high-current region. To eliminate the adverse effects of the inverter's nonlinear voltage error on the identification of small impedance and inductance parameters, this embodiment injects γ1∈П in the low-current region.I The circulating current keeps all phase currents away from the inverter's nonlinear low-current region, thus avoiding the influence of inverter nonlinear voltage errors in the low-current region. In the high-current region, under static conditions, the voltage errors of the motor's d, q, x, and y axes are constant. This implementation eliminates the voltage error related to the inverter's nonlinear voltage error saturation value through differential injection of different current states. In summary, by using a special circulating current injection method and differential injection states, the influence of inverter nonlinear voltage errors is completely eliminated, ensuring voltage accuracy in low-resistance identification.

[0109] During the stationary self-learning process of a dual three-phase permanent magnet synchronous motor, the voltage equations of the motor's d, q, x, and y axes are the same as those of the RL load. To meet the circulating current injection conditions and effectively protect the inverter, this implementation method is based on a sinusoidal current injection parameter identification method. Through simple parameter tuning, the PIR current controller can track the injected current command. For the impedance identification under sinusoidal excitation conditions, the parameter solver in this implementation method consists of a second-order generalized integrator, an orthogonal signal generator, and recursive least squares. Under the condition that the injected current frequency is determined, the second-order generalized integrator is used to extract the sinusoidal component and filter out noise, the orthogonal signal generator is used to calculate the differential value of the sinusoidal signal and suppress differential noise, and recursive least squares eliminates white noise in the input signal, further ensuring the accuracy of the small impedance parameters.

[0110] Taking a high-speed dual three-phase permanent magnet synchronous motor drive system for aviation as an example, the drive system mainly consists of a six-phase inverter unit and a high-speed dual three-phase permanent magnet synchronous motor. The main parameters of the system are: bus voltage 80V, inverter dead time 450ns, switching frequency 40kHz, motor rated current 28A, and motor rated frequency 500Hz.

[0111] Using the method described in the first specific implementation method above, L can finally be calculated. z =15.4μH, L d =155.1μH, R s =46.7mΩ, L q =174.1μH.

[0112] The current waveforms and voltage command waveforms along the d, q, x, and y axes are plotted on [the graphs]. Figure 4 and Figure 5 In the process of self-learning, the phase current waveform is as follows: Figure 6 As shown. Figure 4 The complete current injection sequence during the self-learning process of motor resistance and inductance parameters is demonstrated. Figure 5In the diagram, the DC components of the voltage commands on each axis are the same. This phenomenon verifies that, through specific current injection, the inverter's nonlinear voltage error is constant in the d, q, x, and y axes. It also demonstrates the effectiveness of using differential current to eliminate the inverter's nonlinear voltage error. Figure 6 In the process, all six phase currents are far from zero, and the phase currents do not enter the nonlinear low-current region of the inverter during the entire injection process. This proves the effectiveness of avoiding phase currents from entering the nonlinear low-current region of the inverter through circulating bias injection.

[0113] To demonstrate the effectiveness of the resistance-inductance parameter identification results, four parameter self-learning experiments were conducted. The results of the four resistance-inductance parameter self-learning experiments are as follows:

[0114] 1 46.67 15.43 155.07 174.10 2 46.63 15.31 154.80 172.60 3 45.65 16.45 155.01 175.71 4 46.54 17.16 155.20 175.65

[0115] The power module datasheet was consulted, and the parameters of the motor and cables were measured using an LCR analyzer. The inverter on-state resistance was 6.5 mΩ; the motor phase resistance was 41.2 mΩ, zero-sequence inductance was 14.98 μH, d-axis inductance was 154.4 μH, and q-axis inductance was 166.0 μH; the cable resistance was 3 mΩ, and the cable inductance was 3.1 μH.

[0116] The results of the four parameter self-learning processes are basically consistent, and the self-learned inductance parameters are approximately equal to the sum of the inverter parameters, transmission line parameters, and motor parameters. The data demonstrates the accuracy of the proposed method in inductance parameter identification. It should be noted that the inductance parameter identification method based on harmonic plane current injection and differential injection states proposed in this invention is not only applicable to dual three-phase permanent magnet synchronous motors, but also to other multiphase motors, including but not limited to dual-redundant motors, nine-phase motors, and fifteen-phase motors.

[0117] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A static self-learning method for parameters of a high-speed dual three-phase permanent magnet synchronous motor drive system, characterized in that, The specific method is as follows: Step 1: Construct a four-dimensional current vector control structure for a dual three-phase permanent magnet synchronous motor based on a proportional-integral-resonant current controller, and use this four-dimensional current vector control structure as the current controller for the high-speed dual three-phase permanent magnet synchronous motor drive system. Step Two: In shaft and When the command current of each shaft is 0, a stepped circulating current with a circulating vector angle of 0 is injected into the high-speed dual three-phase permanent magnet synchronous motor drive system. Each circulating current value of this stepped circulating current is traversed, and the feedback current of the high-speed dual three-phase permanent magnet synchronous motor at each circulating current value is recorded when it reaches a stable state. axis, axis, shaft and The shaft command voltage and the six-phase feedback current are sequentially applied to... axis, axis, shaft and The axis command voltage undergoes inverse rotational coordinate transformation and inverse vector space decoupling transformation to obtain the six-phase command voltage in the natural coordinate system. The inverter's nonlinear voltage error saturation value is calculated using the six-phase command voltage corresponding to the last circulating current and the six-phase feedback current. This nonlinear voltage error saturation value, along with the recorded circulating current values ​​for each order, is then used to calculate the inverter's nonlinear voltage error saturation value. axis, axis, shaft and The characteristic current of the nonlinear voltage error is obtained from the shaft command voltage and the six-phase feedback current. ; Step 3: In shaft and When the command current of each shaft is 0, a circulating current with an amplitude of [value missing] is injected into the high-speed dual three-phase permanent magnet synchronous motor drive system. The angle of the circulation vector is The DC circulating current of the high-speed dual three-phase permanent magnet synchronous motor was recorded when the feedback current reached a stable state. axis, axis, shaft and Shaft command voltage and axis, axis, shaft and The shaft feedback current is used as the basic state dataset, and then steps four, five and six are executed respectively; Step Four: In shaft and With the command current of the shaft and the angle of the injected circulating current vector unchanged, a sinusoidal command current is superimposed on the DC circulating current described in step three, and the amplitude of this sinusoidal command current is... satisfy Record the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Shaft command voltage and Shaft feedback current, which Shaft command voltage and The difference between the shaft feedback current and the corresponding data in the basic state dataset is calculated, and the difference is input into the parameter solver to obtain the high-speed dual three-phase permanent magnet synchronous motor. Shaft inductor ; Step 5: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, also... A sinusoidal command current is injected into the shaft, and the amplitude of this sinusoidal command current is... satisfy Record the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Shaft command voltage and Shaft feedback current, which Shaft command voltage and The difference between the shaft feedback current and the corresponding data in the basic state dataset is calculated, and the difference is input into the parameter solver to obtain the high-speed dual three-phase permanent magnet synchronous motor. Shaft inductor and stator resistance ; Step Six: While injecting the DC circulating current into the high-speed dual three-phase permanent magnet synchronous motor drive system, also... A sinusoidal command current is injected into the shaft, and the amplitude of this sinusoidal command current is... satisfy Record the high-speed dual three-phase permanent magnet synchronous motor when the feedback current reaches a stable state. Shaft command voltage and Shaft feedback current, which Shaft command voltage and The difference between the shaft feedback current and the corresponding data in the basic state dataset is calculated, and the difference is input into the parameter solver to obtain the high-speed dual three-phase permanent magnet synchronous motor. Shaft inductor ; The parameter solver includes a second-order generalized integrator, an orthogonal signal generator, and a recursive least squares stage. The second-order generalized integrator is used to filter the input signal, the orthogonal signal generator is used to differentiate and filter the input signal, and the recursive least squares stage performs least squares on the filtered input signal and the differential result of the input signal to obtain the output of the parameter solver. In step three, The amplitude of the DC circulating current , The circulation vector angle of the DC circulating current , The basic state dataset , The above, This refers to the rated current of a high-speed dual three-phase permanent magnet synchronous motor. , , and These are high-speed dual three-phase permanent magnet synchronous motors with injected DC circulating current. axis, axis, shaft and Shaft command voltage, , , and These are high-speed dual three-phase permanent magnet synchronous motors with injected DC circulating current. axis, axis, shaft and Shaft feedback current; The sinusoidal command current superimposed on the DC circulating current in step four, and the current directed to the DC circulating current in step five... The sinusoidal command current injected into the shaft, and the current injected into the shaft in step six. The sinusoidal command current injected into the shaft is all , The amplitude of the superimposed sinusoidal command current. The frequency of the sinusoidal command current. For time; The transfer function of the second-order generalized integrator in the parameter solver The expression is: ； The transfer function of the orthogonal signal generator in the parameter solver The expression is: ； The identification equation for the recursive least squares stage in the parameter solver is: ； in, For transfer function variables, For transfer function variables, The integral coefficient is... The frequency of the injected signal, Sampling time, for The difference between the shaft command voltage and the corresponding data in the basic state dataset. for The difference between the shaft feedback current and the corresponding data in the basic state dataset. for Shaft inductor, For stator resistance, for The differential value, and has , .

2. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 1, characterized in that, The control flow of the four-dimensional current vector control structure is as follows: Vector space decoupling transformation is performed on the six-phase command current of the high-speed dual three-phase permanent magnet synchronous motor to obtain... shaft and Shaft command current, for shaft and Obtained by rotating coordinate transformation using axis command current shaft and Shaft command current, The circulating current command amplitude and circulating current vector angle of the high-speed dual three-phase permanent magnet synchronous motor drive system are calculated respectively. shaft and Shaft command current, Each axis, axis, shaft and The difference between the command current and the corresponding feedback current of the shaft is obtained through a proportional-integral-resonant current controller. axis, axis, shaft and Command voltage of the shaft, In turn axis, axis, shaft and The command voltage of the axis undergoes vector space decoupling inverse transformation, rotating coordinate transformation, and zero-sequence injection modulation to obtain the switching signal of the inverter. The switching signal is applied to the inverter of the high-speed dual three-phase permanent magnet synchronous motor drive system to achieve four-dimensional current vector control of the high-speed dual three-phase permanent magnet synchronous motor.

3. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 2, characterized in that, The circulating current command amplitude and circulating current vector angle of the high-speed dual three-phase permanent magnet synchronous motor drive system are calculated respectively. shaft and The specific method for controlling the shaft current is as follows: ， in, and They are respectively shaft and Shaft command current, This refers to the circulating current command amplitude of the high-speed dual three-phase permanent magnet synchronous motor drive system. The angle is the circulation vector angle.

4. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 2 or 3, characterized in that, Transfer function of proportional-integral-resonant current controller The expression is as follows: ， in, For transfer function Complex variables, , and These are the proportional, integral, and resonance coefficients, respectively. This is the delay compensation coefficient. Sampling time, It is the resonant frequency.

5. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 1 or 2, characterized in that, In step two, the first Circulation The expression is: ， in, , The total number of steps in the stepped circulation. This is the rated current of a high-speed dual three-phase permanent magnet synchronous motor.

6. The parameter static self-learning method for a high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 1 or 2, characterized in that, In step two, the six-phase command voltage in the natural coordinate system is obtained according to the following formula: in, , , , , , They are respectively the first in the natural coordinate system The six-phase command voltage corresponding to the first-order circulating current. It is a 2×2 matrix containing all zeros. It is a 2×2 identity matrix. , , and The first The corresponding order circulation axis, axis, shaft and Shaft command voltage, , The total number of steps in the stepped circulation. The vector space decoupling transformation matrix has , Let be the rotation coordinate transformation matrix, and have , for The angle between the shaft and the axis of the A-phase winding.

7. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 6, characterized in that, The inverter nonlinear voltage error saturation value is obtained according to the following formula: ， in, This represents the saturation value of the inverter's nonlinear voltage error. , , and The first The command voltages for phases A, B, D, and E corresponding to the first-order circulating current. and The first The feedback currents of phases A and D corresponding to the circulating current.

8. The parameter static self-learning method for the high-speed dual three-phase permanent magnet synchronous motor drive system according to claim 7, characterized in that, Using the inverter's nonlinear voltage error saturation value and the recorded circulating current values ​​of each order... axis, axis, shaft and The specific method for obtaining the characteristic current of nonlinear voltage error from shaft command voltage and six-phase feedback current is as follows: S1: Determine if the following expression is true. If yes, execute S2; otherwise, execute S3. ， in, , , and The first The command voltages for phases A, B, D, and E corresponding to the first-order circulating current. and The first The feedback currents of phases A and D corresponding to the first-order circulating current. S2: The first E-phase feedback current corresponding to the first-order circulating current As characteristic current , S3: Make And return to S1.

Images