Coupling inductance-based double three-phase permanent magnet synchronous motor phase current reconstruction error compensation method

Abstract

The application discloses a double three-phase permanent magnet synchronous motor phase current reconstruction error compensation method based on coupled inductance, and belongs to the field of motor driving and control. The application aims at solving the problems of large reconstruction current error and low control precision caused by single current sensor time-sharing sampling, obtains average voltage and average back electromotive force of each correction interval, solves six-phase average current change rate of each correction interval based on coupled inductance matrix containing self-inductance and mutual inductance, and uses the six-phase average current change rate to correct the original sampling current in time; during the correction, the current components belonging to different windings in the same sampling current are respectively corrected to the reference time corresponding to the respective windings, the corrected sampling current is obtained, six-phase current is reconstructed, and the d and q shaft feedback currents of the two sets of three-phase windings are obtained through coordinate transformation. The application introduces the coupled inductance matrix to take the double winding electromagnetic coupling into the current change rate calculation, effectively reduces the physical time deviation caused by time-sharing sampling, and improves the reconstruction precision and current closed-loop control precision.

Description

Technical Field

[0001] This invention relates to a method for compensating for phase current reconfiguration error in a dual three-phase permanent magnet synchronous motor based on coupled inductors, belonging to the field of motor drive and control. Background Technology

[0002] Dual Three-Phase Permanent Magnet Synchronous Motor (DTP-PMSM) features lower single-phase power ratings, stronger fault tolerance, smaller torque ripple, and suitability for medium- and low-voltage high-power drives. It has good application value in aerospace, electric vehicles, electric propulsion, and other applications with high requirements for system power density, reliability, and operating performance.

[0003] Traditional DTP-PMSM closed-loop drive systems typically require acquiring current information from two sets of three-phase windings for coordinate transformation and current closed-loop control. Using multiple current sensors increases the number of sampling circuits, conditioning circuits, and interfaces; furthermore, gain errors and zero-point biases between different sensors can impact control performance. Therefore, utilizing a single current sensor to acquire multiple sets of current information within a PWM cycle and reconstructing the six-phase current through corresponding current combination relationships has become a technical approach to reduce the number of current sensors in a DTP-PMSM drive system.

[0004] Existing research on single-current sensor phase current reconstruction and its accuracy improvement mainly focuses on traditional three-phase permanent magnet synchronous motors (PMSMs). For three-phase PMSMs, existing research has reduced time-sharing sampling errors through methods such as current rate of change estimation, current observation, and sampling time correction. However, DTP-PMSMs contain two sets of three-phase windings, and single-current sensor six-phase reconstruction requires acquiring multiple combined samples containing current information from both sets of windings within one PWM cycle. Different samples, and different winding current components within the same sample, may correspond to different sampling times. Existing sampling error compensation relationships established for single-set three-phase windings are difficult to directly apply to the six-phase time-sharing sampling scenario of DTP-PMSMs.

[0005] In terms of current rate of change estimation, existing methods are mostly based on a single three-phase PMSM voltage model, treating the two windings as independent systems and failing to account for the impact of electromagnetic coupling between the two windings through a common magnetic circuit on the current rate of change. When the mutual inductance between the two windings is large in the motor design, there is a deviation between the current rate of change obtained by existing calculation methods and the actual change process.

[0006] Furthermore, in a DTP-PMSM single-current sensor reconfiguration system employing carrier phase-shift modulation, the sampling states of the two windings are staggered in time, resulting in the reconfigured currents of the two windings corresponding to different reference times. Existing methods, when performing Clarke / Park coordinate transformation, uniformly apply the rotor electrical angle corresponding to the control calculation time or position acquisition time to both windings, without distinguishing the differences in the reconfiguration reference times between the two windings.

[0007] In summary, the existing single-current sensor dual three-phase permanent magnet synchronous motor phase current reconstruction method has a large reconstruction current error, resulting in low current closed-loop control accuracy. Summary of the Invention

[0008] To address the problem that existing dual three-phase permanent magnet synchronous motors with single current sensors have large reconfiguration current errors, which affect the accuracy of current closed-loop control, this invention provides a method for compensating for phase current reconfiguration errors in dual three-phase permanent magnet synchronous motors based on coupled inductors.

[0009] The present invention provides a method for compensating for phase current reconfiguration error in a dual three-phase permanent magnet synchronous motor based on coupled inductors, comprising:

[0010] Obtain the raw sampling current of a single current sensor within the current switching cycle;

[0011] Determine each correction interval, and calculate the average voltage and average back electromotive force for each correction interval;

[0012] The six-phase reconstructed current of the previous switching cycle is obtained as the current state. Based on the current state, the coupled inductance matrix including self-inductance and mutual inductance, the average voltage and average back electromotive force of each correction interval, the average current change rate of each correction interval in the dual αβ stationary coordinate system is solved. Then, the average current change rate of each correction interval is converted into the six-phase average current change rate of each correction interval through the three-phase stationary coordinate inverse transformation.

[0013] The original sampled current is time-corrected using the six-phase average current change rate of each correction interval to obtain the corrected sampled current, and the corrected six-phase reconstructed current is obtained based on the corrected sampled current.

[0014] The corrected six-phase reconstructed current is subjected to coordinate transformation to obtain the d-axis and q-axis feedback currents of two sets of three-phase windings.

[0015] Preferably, the six-phase average current change rate is:

[0016] ;

[0017] in, The average current change rate of the six phases. For the coupled inductor matrix, To correct the average voltage of the interval, For motor resistance, This is the six-phase reconfiguration current from the previous switching cycle. This is the average back electromotive force for the correction interval.

[0018] Preferably, the method for performing coordinate transformation on the corrected six-phase reconstructed current includes:

[0019] The Park transformation is performed on the current belonging to the first set of three-phase windings in the corrected six-phase reconstructed current using a first electrical angle, and on the Park transformation is performed on the current belonging to the second set of three-phase windings in the corrected six-phase reconstructed current using a second electrical angle.

[0020] The first electrical angle and the second electrical angle are calculated based on the rotor electrical angle, the rotor electrical angular velocity, the first correction reference time, and the second correction reference time, respectively.

[0021] Preferably, when the rotor electrical angle is acquired at [time]... hour:

[0022] ;

[0023] When the rotor electrical angle corresponds to the control calculation time hour:

[0024] ;

[0025] in, For the first electrical angle, For the second electrical angle, The rotor electrical angle, The rotor's electric angular velocity, This is the first correction reference time. This is the second correction reference time.

[0026] Preferably, the method for time-correcting the original sampled current using the six-phase average current change rate of each correction interval to obtain the corrected sampled current includes:

[0027] The current reconfiguration region is determined based on the reference voltage vector position or the switching states of the two inverters.

[0028] Based on the current reconstruction region, determine the correspondence between each original sampled current and each phase current component of the first set of three-phase windings and the second set of three-phase windings:

[0029] Based on the correspondence, determine the winding to which each phase current component in each original sampled current belongs and the corresponding correction reference time of the winding;

[0030] Based on the current reconstruction region and the corresponding relationship, the correction interval corresponding to each phase current component is determined, and the six-phase average current change rate corresponding to the correction interval is called.

[0031] Each phase current component is corrected from the actual sampling time to the correction reference time corresponding to its winding. If the correction reference time is later than the actual sampling time, forward correction is performed. If the correction reference time is earlier than the actual sampling time, reverse backtracking correction is performed through a negative time interval to obtain the corrected value of each phase current component.

[0032] Based on the correspondence, the corrected values ​​of each phase current component belonging to the same original sampling current are recombined to obtain the corrected sampling current.

[0033] The beneficial effects of this invention are as follows: Addressing the characteristic of time-division sampling of six-phase current in a dual three-phase permanent magnet synchronous motor with a single current sensor, this invention performs physical timing correction on the current components belonging to different windings within the same sampled current. This ensures that the reconstructed current of each phase corresponds to a unified reference time for its respective winding, avoiding reconstruction deviations caused by inconsistencies in the physical timing of the currents in each phase, and improving the accuracy of six-phase current reconstruction. Furthermore, based on a coupled inductance matrix including self-inductance and mutual inductance, a current change rate model in a stationary coordinate system is established. When calculating the current change rate within each correction interval, the electromagnetic coupling effect between the two sets of three-phase windings is simultaneously considered. This overcomes the deviations caused by existing methods that treat the two sets of windings as independent systems and calculate their change rates separately. This ensures that the calculated current change rate matches the actual physical change process, providing an accurate physical basis for sampling current time correction. Attached Figure Description

[0034] Figure 1 This is a control block diagram of the present invention;

[0035] Figure 2 This is a schematic diagram showing the PWM switching state, actual current change, sampling time, and current correction principle of a dual-reference time for a three-phase winding under a fixed carrier phase shift condition.

[0036] Figure 3 A schematic diagram showing the division of the phase current reconstruction region for a single current sensor in a dual three-phase permanent magnet synchronous motor.

[0037] Figure 4 This is a flowchart of the time-division sampling current error correction and control delay compensation method based on the dual-winding coupling model of the present invention;

[0038] Figure 5 The simulation comparison waveforms are shown in the following figures: (a) is the actual current of the six phases, (b) is the uncorrected reconfigured current, (c) is the reconfigured current without considering coupling correction, and (d) is the reconfigured current considering coupling correction.

[0039] Figure 6 The simulation comparison waveforms of the original sampling current and the time-corrected sampling current of the single current sensor at four time points are shown, where (a) is the original sampling current and (b) is the corrected sampling current.

[0040] Figure 7 The simulation comparison waveforms of the actual phase current and the corrected reconstructed value are shown in the figure, where (a) is the actual phase current and (b) is the corrected reconstructed value.

[0041] Figure 8 The simulation comparison waveforms of the reconstruction error before and after the representative phase current correction are shown in the figure. (a) is the reconstruction error before correction, and (b) is the reconstruction error after correction. Detailed Implementation

[0042] 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.

[0043] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0044] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0045] This embodiment provides a method for compensating for phase current reconstruction error of a dual three-phase permanent magnet synchronous motor based on coupled inductance, considering winding coupling. It is applicable to a dual three-phase permanent magnet synchronous motor drive system that uses a single current sensor for six-phase current reconstruction.

[0046] Existing single-current-sensor methods for reconstructing phase current in dual three-phase permanent magnet synchronous motors suffer from large reconstruction current errors, leading to low accuracy in current closed-loop control. To address this issue, this embodiment introduces a coupled inductance matrix, incorporating both self-inductance and mutual inductance, into the current rate of change calculation, ensuring that the calculated rate of change aligns with the actual physical change process. Furthermore, time correction is applied to current information obtained at different sampling times, correcting the current components belonging to different windings within the same sampled current to their respective reference times, thus eliminating reconstruction bias introduced by time-division sampling. Further, this invention calculates the corresponding rotor electrical angles based on the respective reconstruction reference times of the two windings, using these angles for coordinate transformation of the reconstruction currents of the two windings, thereby eliminating time mismatch between the reconstruction current time and the coordinate transformation electrical angles.

[0047] The core implementation process of this invention is as follows: Obtain the original sampling current of a single current sensor within the current switching cycle; determine each correction interval and calculate the average voltage and average back electromotive force (EMF) of each correction interval; obtain the six-phase reconstructed current of the previous switching cycle as the current state; based on the current state, the coupled inductance matrix including self-inductance and mutual inductance, the average voltage and average back EMF of each correction interval, solve for the average current change rate of each correction interval in the dual αβ stationary coordinate system, and then convert it to the six-phase average current change rate of each correction interval through a three-phase stationary coordinate inverse transformation; use the six-phase average current change rate of each correction interval to perform time correction on the original sampling current to obtain the corrected sampling current; reconstruct the corrected six-phase reconstructed current based on the corrected sampling current; perform coordinate transformation on the corrected six-phase reconstructed current to obtain the d-axis and q-axis feedback currents of two sets of three-phase windings; send the d-axis and q-axis feedback currents into the current loop to achieve closed-loop control, and repeat the above process in the next switching cycle. Specifically, the method of this embodiment includes:

[0048] Step 1: Obtain the raw sampling current of a single current sensor within the current switching cycle;

[0049] A single current sensor is installed on the DC bus side or inverter output side of the dual three-phase permanent magnet synchronous motor drive system to collect current information time-sharingly within one PWM switching cycle. The dual three-phase permanent magnet synchronous motor comprises two sets of three-phase windings: the first set of three-phase windings ABC and the second set of three-phase windings XYZ, each powered by a separate three-phase inverter. The two inverters use a fixed... The carrier phase-shift modulation method causes the switching states of the two sets of windings to be staggered in time.

[0050] Within each switching cycle, the raw sampled current obtained by the single current sensor at multiple preset sampling times is acquired. ,in , The number of samples within one switching cycle; simultaneously acquiring the DC bus voltage. Rotor electrical angle Rotor electric angular velocity The PWM switching status of the two inverters and the motor parameters.

[0051] For a dual three-phase motor drive system using four-point sampling, the four sampling times can be represented as:

[0052] (1)

[0053] The four sampling times 0, ... , and Each single current sensor is controlled to sample, obtaining four raw sampled currents, which are denoted as follows: , , and ,in For the switching period. In the four sampling times, 0 and... The zero vector action times corresponding to the first set of three-phase windings ABC are ABC-000 and ABC-111 respectively. and The zero vector action times corresponding to the second set of three-phase windings XYZ (XYZ-000 and XYZ-111, respectively).

[0054] At the start of a switching cycle, the PWM switching states of the two inverters within the current switching cycle are read, and the current current reconfiguration region is determined based on the reference voltage vector position. ,in , refer to Figure 3 Based on the dual three-phase permanent magnet synchronous motor, the correspondence between the PWM switching state and the reconfiguration region can be obtained. The reconfiguration region is determined based on the reference voltage vector. The spatial position in the stationary coordinate system is determined, and each reconstruction region corresponds to a specific linear correspondence between the sampling current combination and the phase current component.

[0055] Table 1. Correspondence between four sampled values ​​and phase current components in different reconstruction regions.

[0056]

[0057] In Table 1, This represents the current in phase B of the first three-phase winding ABC. This represents the current in phase C of the first three-phase winding ABC. This represents the current in phase X of the second three-phase winding XYZ. This represents the current in the Y phase of the second three-phase winding XYZ.

[0058] Step 2: Determine each correction interval and calculate the average voltage and average back electromotive force for each correction interval; specifically:

[0059] Based on the current correction reference time of the first set of three-phase windings ABC Reference time for current correction of the second set of three-phase windings XYZ and four sampling times 0, , and Five correction intervals are determined. In this embodiment, the five correction intervals are as follows:

[0060] (2)

[0061] For any correction interval Based on the PWM switching states of the two inverters and the DC bus voltage during the current switching cycle Calculate the instantaneous voltage vectors of the two sets of three-phase windings in the stationary coordinate system within this interval. (The first...) Taking a three-phase winding as an example, its voltage in the stationary coordinate system is expressed as:

[0062] (3)

[0063] in, This is the three-phase stationary coordinate transformation matrix. Indicates the first Bridge arm switching function vector of three-phase winding, , and The first The switching states of the A, B, and C phase bridge arms in the three-phase winding are set, taking values ​​of 0 or 1, and the correction interval is calculated accordingly. Average voltage within:

[0064] (4)

[0065] This implementation does not require a correction interval to correspond to only a single voltage vector. Instead, it calculates the equivalent average voltage under the combined action of multiple voltage vectors based on the actual PWM switching state within that interval. When the correction interval contains multiple voltage vectors, a weighted calculation is performed based on the actual duration of each voltage vector.

[0066] (5)

[0067] in, This represents the number of voltage vectors contained within the correction interval. For the first The duration of action of each voltage vector;

[0068] Based on the obtained rotor electrical angle Rotor electric angular velocity and permanent magnet flux parameters Calculate the back electromotive force (EMF) of the two sets of three-phase windings in the stationary coordinate system. For the m-th winding, its back EMF can be expressed as:

[0069] (6)

[0070] in, This represents the spatial position offset angle of the corresponding winding. For any correction interval... Calculate the interval average back electromotive force:

[0071] (7)

[0072] When the electrical angular velocity is approximately constant within a switching cycle, the corresponding average back electromotive force can also be calculated based on the starting and ending electrical angles of the interval or the electrical angle at the midpoint of the interval. This step ensures that the average back electromotive force corresponds to the average voltage within the same physical time interval.

[0073] Step 3: Obtain the six-phase reconstructed current from the previous switching cycle as the current state. Based on the current state, the coupled inductance matrix including self-inductance and mutual inductance, the average voltage and average back electromotive force of each correction interval, solve for the average current change rate of each correction interval in the dual αβ stationary coordinate system. Then, convert it into the six-phase average current change rate of each correction interval through a three-phase stationary coordinate inverse transformation. Step 3 is the core of this implementation method. The concept is that, in the process of solving the current change rate, a coupled model of the current change rate in the stationary coordinate system is established based on the coupled inductance matrix including self-inductance and mutual inductance. The electromagnetic coupling effect between the two sets of three-phase windings is included in the calculation, overcoming the deviation caused by the existing method of treating the two sets of windings as independent systems and calculating the change rate separately. After obtaining the six-phase average current change rate of each correction interval based on the coupled model, it is used for the time correction of the sampled current and subsequent reconstruction and coordinate transformation, constituting a complete phase current reconstruction error compensation method. Specifically, it includes:

[0074] The corrected six-phase reconstructed current obtained in the previous switching cycle is transformed into a dual-phase current. Using a stationary coordinate system, we obtain the current state used for calculating the rate of change of current during the current switching cycle:

[0075] (8)

[0076] For the first execution, there is no six-phase reconfiguration current after correction in the previous switching cycle. At this time, a preset initial current is used as the current state. The preset initial current can be set to zero or the current estimate read from the initial state of the motor.

[0077] For any correction interval Based on the average voltage, average back electromotive force, previous cycle correction and reconstruction current, and motor resistance within this interval. and coupling inductance parameters Calculate the rate of change of the interval average current in the stationary coordinate system of the two windings:

[0078] (9)

[0079] The interval average voltage and back electromotive force vector are:

[0080] (10)

[0081] in, This is a coupled inductance matrix containing the self-inductance and mutual inductance of two sets of three-phase windings. Its general form can be written as:

[0082] (11)

[0083] For two sets of windings with identical parameters, the inductance matrix can be simplified accordingly as follows:

[0084] (12)

[0085] When the algorithm is first put into operation and there is no previous cycle correction reconfiguration current, initialization can be completed using the uncorrected reconfiguration current, the preset initial current, or other available current estimates; subsequent switching cycles use the corrected reconfiguration current for rolling updates.

[0086] Based on the inverse transformation relationship of the three-phase stationary coordinates, the double coordinates obtained in each correction interval are... The rate of change of current is converted into the six-phase current rates of change of ABC and XYZ. Its general form can be expressed as:

[0087] (13)

[0088] Indicates double Transformation matrix from stationary coordinate system to six-phase stationary coordinate system;

[0089] In this step, formulas (8) to (13) constitute a coupled model, thereby obtaining the average rate of change of each phase current within different correction time intervals.

[0090] Step 4: Use the six-phase average current change rate of each correction interval to perform time correction on the original sampled current to obtain the corrected sampled current.

[0091] This step, based on the correspondence between the original sampled currents and the current components of each phase determined in the current reconstruction region, uses the six-phase average current change rate of each correction interval calculated in step three to perform time correction on the current components belonging to the first set of three-phase windings and the current components belonging to the second set of three-phase windings in the original sampled current. This ensures that the current components belonging to different windings in the same sampled current are corrected from their actual sampling time to the correction reference time corresponding to their respective windings, thus obtaining the corrected sampled current. Specifically:

[0092] First, determine the current reconstruction region based on the reference voltage vector position. Based on the correspondence between each original sampled current and each phase current component determined in step 1, the winding to which each phase current component in each original sampled current belongs is determined. Specifically:

[0093] The current components belonging to the first set of three-phase windings ABC include , , ;

[0094] The current components belonging to the second set of three-phase windings XYZ include , , .

[0095] Determine the corresponding correction reference time based on the winding to which each phase current component belongs:

[0096] The current components of the first set of three-phase windings ABC are corrected to the first correction reference time. ;

[0097] The current components of the second set of three-phase windings XYZ are corrected to the second correction reference time. .

[0098] Based on the current reconstruction area The correction interval for each phase current component is determined by identifying the winding to which each phase current component belongs. The principle for determining the correction interval is: the time interval between the actual sampling time of each phase current component and the correction reference time corresponding to the winding to which the component belongs is the correction interval for that component.

[0099] Specifically, for the original sampled current at sampling time 0 The B-phase current component (belonging to the ABC winding) has an actual sampling time of 0 and a correction reference time of 0. Therefore, its correction interval is For sampling time is The original sampling current The B-phase current component (belonging to the ABC winding) is sampled at the following time: The correction reference time is Therefore, its correction interval is For sampling time is The original sampling current The B-phase current component (belonging to the ABC winding) is sampled at the following time: The correction reference time is Because the target reference time is earlier than the actual sampling time, its correction interval is... Furthermore, reverse backtracking correction is required.

[0100] The six-phase average current change rate corresponding to each correction interval is used to perform time correction on each phase current component in each original sampled current. The basic correction formula is:

[0101] (14)

[0102] in, and The combination coefficient of the corresponding phase current in the j-th sample under the current reconstruction region; and These represent the correction intervals for the phase current from the actual sampling time to the reference time of its corresponding winding. Forward correction is performed when the target reference time is later than the actual sampling time; reverse backtracking is achieved through a negative time interval when the target reference time is earlier than the actual sampling time.

[0103] For different reconstruction regions, a correspondence of "reconstruction region - sampling current combination - correction interval - current change rate" is pre-established. The corresponding sampling correction expression is selected based on the current region. The specific correspondence is shown in Table 2, thereby obtaining multiple corrected sampling values. .

[0104] Table 2. Time Correction Relationship of Four Sampling Currents in Different Reconstruction Regions

[0105]

[0106] in, Indicates phase B in The average rate of change of current within the interval; the same applies to other phases and other intervals. The subscripts 0050, 0075, 2550, 2575, and 5075 represent... , , , and In I4, the component of the ABC winding is varied within a 5075 interval using a rate of change from... Backtracking .

[0107] To reconstruct the region For example, the correction process for each original sampled current is as follows:

[0108] For the original sampling current It contains only the B-phase current component, with the actual sampling time being 0 and the correction reference time being [missing information]. Using the correction interval The average current change rate of phase B The corrected sampling current is obtained by performing correction. :

[0109] (15)

[0110] For sampling current It contains only the C-phase current component, and the actual sampling time is It is exactly equal to the first correction reference time. No time correction is required:

[0111]

[0112] For sampling current This includes the B-phase current component and the X-phase current component. Corrections are applied to both components separately: the actual sampling time for the B-phase component is... Corrected to the reference time Using the correction interval The average current change rate of phase B The actual sampling time of the X-phase component is Corrected to the reference time Using the correction interval The average current change rate of phase X After correction, the two components are recombined to obtain:

[0113] (16)

[0114] For sampling current This includes both B-phase and Y-phase current components. The actual sampling time for the B-phase component is... It is necessary to backtrack to the reference time. Using the correction interval The average current change rate of phase B Perform reverse correction; the actual sampling time of the Y-phase component is It is exactly equal to the second correction reference time. No correction is needed. After correction, we get:

[0115] (17)

[0116] Similarly, for the reconstructed region to Based on the correspondence shown in Table 2, the correction intervals corresponding to each phase current component in each original sampled current are determined, and the corresponding six-phase average current change rate is called to perform correction according to the above correction method.

[0117] Step 5: Reconstruct the corrected six-phase reconstructed current based on the corrected sampled current; this step uses the four corrected sampled currents obtained in Step 4. Algebraic operations are performed using the six-phase current reconstruction expression corresponding to the current reconstruction region to reconstruct the six-phase current values, which are the corrected six-phase reconstructed currents.

[0118] Based on the current reconstruction area Select the corresponding six-phase current reconstruction expression, and convert the four corrected sampled currents... Substituting into the six-phase current reconstruction expression, the corrected six-phase reconstructed currents are obtained, including the corrected phase currents of the first set of three-phase windings ABC. Corrected phase currents of the second set of three-phase windings XYZ The six-phase current reconstruction expressions corresponding to different reconstruction regions are shown in Table 3.

[0119] Table 3. Expressions for the corrected six-phase current reconstruction under different reconstruction regions.

[0120]

[0121] To reconstruct the region For example, the chosen six-phase current reconstruction expression is:

[0122] (18)

[0123] For the reconstructed region to The calculations are performed according to the six-phase current reconstruction expressions for each region shown in Table 3.

[0124] Step 6: Perform coordinate transformation on the corrected six-phase reconstructed current to obtain the d and q axis feedback currents of the two sets of three-phase windings. In this step, the corrected six-phase reconstructed current obtained in step 5 is transformed to a rotating coordinate system through Clarke / Park coordinate transformation to obtain the current feedback values ​​of the two sets of three-phase windings on their respective d and q axes, which are used for current closed-loop control.

[0125] The first and second windings respectively adopt rotor electrical angles corresponding to their respective current reconstruction reference times.

[0126] Rotor electrical angle obtained from rotor position detection Rotor electric angular velocity The time for acquiring the rotor electrical angle is and the correction reference time for the two sets of windings , Calculate the corrected electrical angles for the Park transformation of the first set of three-phase windings ABC and the second set of three-phase windings XYZ respectively:

[0127] (19)

[0128] When the rotor electrical angle Corresponding to the PWM update or control calculation time When the above formula is used, it can be expressed as:

[0129] (20)

[0130] When using fixed For carrier phase shifting and symmetrical seven-segment SVPWM, the following can be selected:

[0131] (twenty one)

[0132] Using the corrected electrical angle Reconfiguration current of the first set of windings Perform Clarke / Park coordinate transformation to obtain the d-axis and q-axis feedback currents of the first set of windings:

[0133] (twenty two)

[0134] Using the corrected electrical angle Reconfiguration current of the second set of windings Perform Clarke / Park coordinate transformation to obtain the d-axis and q-axis feedback currents of the second set of windings:

[0135] (twenty three)

[0136] The result As feedback quantities for the two sets of winding current loops, they are compared with the corresponding d-axis and q-axis current setpoints, and the difference is used by the current regulator to obtain the reference voltage command for the two sets of windings.

[0137] PWM modulation is executed according to the reference voltage command to generate switching signals for two inverters and update the PWM duty cycle. After entering the next switching cycle, the above steps are repeated to achieve multi-time sampling of a single current sensor, sampling current correction based on a coupling model, six-phase current reconstruction at dual reference times, time-sharing electrical angle compensation, and current closed-loop control.

[0138] This embodiment addresses the six-phase current time-division sampling characteristics of the DTP-PMSM single current sensor by performing time correction on the current information obtained at different sampling times, thereby improving the accuracy of six-phase current reconstruction. Based on considering the coupling current change rate of the self-inductance and mutual inductance of the two sets of three-phase windings, this invention can more accurately calculate the current change within each correction interval. According to different reconstruction regions and sampling current combination relationships, this embodiment corrects the sampling currents of the ABC winding and the XYZ winding to their corresponding reference times, achieving unified reconstruction of the sampling current at multiple times. Furthermore, this embodiment corrects the Park transformation electrical angle according to the different current reconstruction reference times of the two sets of windings, reducing the time deviation between the current reconstruction time and the coordinate transformation electrical angle, thus improving the accuracy of current closed-loop control.

[0139] Example:

[0140] This embodiment presents a method for compensating for phase current reconstruction errors using a single current sensor in a dual three-phase permanent magnet synchronous motor, considering winding coupling, and its control implementation. For example... Figure 1As shown, the dual three-phase permanent magnet synchronous motor is powered by two sets of three-phase inverters. The two sets of three-phase windings adopt independent current closed-loop control and SVPWM modulation. The system obtains the combined current information required for six-phase current reconstruction through a single current sensor SCS, and obtains the rotor electrical angle through a rotary transformer. and rotor electric angular velocity The current reconstruction and compensation process sequentially completes the following steps: reconstruction region determination, calculation of the interval current change rate based on the dual-winding coupling model, sampling current time correction, and six-phase current reconstruction. The reconstructed currents of the two windings are respectively compared to 0.5T... s and 0.75T s The rotor electrical angle corresponding to the reference time is transformed into Clarke / Park coordinates and used as feedback for the two sets of winding current loops. Finally, the PWM modulation module generates the switching drive signals for the two inverters.

[0141] In this embodiment, two sets of dual three-phase permanent magnet synchronous motor drive systems with a winding spatial displacement angle of 0 are established in PLECS to simulate and verify the error compensation method described in this invention. The main parameters used in the simulation are shown in Table 4.

[0142] Table 4 Main simulation parameters of this embodiment

[0143]

[0144] In this embodiment The self-inductance of the two sets of windings The mutual inductance of the two windings used for calculating the rate of change of coupling current is 7.26mH on both shafts and 2mH for leakage inductance.

[0145]

[0146] Under steady-state operation at 1000 r / min, the mechanical angular velocity and the rotor electrical angular velocity are respectively:

[0147]

[0148] The corresponding electrical frequency is approximately 83.33 Hz. The PWM switching frequency is 10 kHz, therefore .

[0149] like Figure 2 As shown, the first set of ABC windings and the second set of XYZ windings use a fixed π / 2 carrier phase shift, and the two sets of PWM switching sequences are staggered within one switching cycle. Four single current sensor sampling times are generated. Figure 2 The actual sampling points in the data represent the combined currents directly obtained by the SCS at each moment, while the compensated sampling points represent the current information after different phase current components have been corrected to the corresponding reference time according to their respective windings.

[0150] In this embodiment, the current reconstruction reference time for the ABC windings is set to... The current reconstruction reference time of the XYZ winding is set to The specific value is:

[0151]

[0152] Depend on Figure 2 The four sampling times and two reconstruction reference times shown in this embodiment actually employ five types of correction intervals:

[0153]

[0154] Within each PWM cycle, At any given time, sample values ​​from four single current sensors are collected. Simultaneously obtain bus voltage Rotor electrical angle Rotor electric angular velocity And the switching states of the two inverters in the previous PWM cycle. For example... Figure 3 As shown, according to the reference voltage vector The position divides the voltage vector plane into Six reconstructed regions, based on the region to which the current reference voltage vector belongs. Determine the combination relationship between the four sampled values ​​and the six-phase currents ABC and XYZ.

[0155] For example, when the reference voltage vector is located Figure 3 When looking up the G1 region as shown, we get the following from Table 1:

[0156]

[0157] according to Figure 4 The process shown involves determining the relationship between the reconstruction region and the sampling components, and then calculating the average voltage of each correction interval based on the measured bus voltage and the switching function of the previous PWM cycle. For any correction interval... When the interval contains When there are multiple voltage vectors, a weighted average is calculated based on the actual duration of each voltage vector:

[0158]

[0159] by For example, its interval length is ,therefore:

[0160]

[0161] The interval lengths are respectively The average voltage of each interval is calculated based on the actual PWM switching state and 550V bus voltage within that interval, and it is not required that a correction interval correspond to only a single voltage vector.

[0162] At the same time, according to the rotor electrical angle Rotor electric angular velocity and permanent magnet chain Calculate the average back electromotive force corresponding to the five correction intervals mentioned above. When a correction interval... Inside When approximately constant, let , Then we have:

[0163]

[0164] Under steady-state conditions of 1000 r / min For example, in this embodiment If the electrical angle starts at a certain PWM cycle As an example of numerical calculation, then:

[0165]

[0166] the remaining The intervals are calculated in the same way so that the average back electromotive force and the corresponding average voltage belong to the same physical time interval.

[0167] The rate of change of current within each correction interval is then calculated. The six-phase reconstructed current after correction in the previous PWM cycle is then subjected to a Clarke transform to obtain... The average voltage, average back electromotive force, motor resistance, and coupled inductance parameters are then substituted into the current rate of change model.

[0168] This embodiment is in The self-inductance on the shaft is H, mutual inductance are both H, therefore:

[0169]

[0170] For any correction interval W, define:

[0171]

[0172] Substituting the inductance parameters into the coupling model, we obtain the dual-inductance model actually used in the five types of correction ranges in this embodiment. Formula for calculating the average rate of change of current:

[0173]

[0174] Calculate separately The rate of change of the double αβ average current within the five correction intervals is denoted as follows: , , and .

[0175] Then, according to the inverse transformation relationship of the three-phase stationary coordinates, the double... The rate of change of current is converted into the six-phase current rates of change of ABC and XYZ. For any correction interval W, we have:

[0176]

[0177] Therefore, we obtain the following respectively , , and Average current change rate of five groups of six phases.

[0178] Further according to Figure 4 The "Sampling Current Time Correction" process shown above involves looking up Table 2 in the current reconstruction region to... Time correction is performed on the different phase current components contained therein. (Still using...) Figure 3 In Taking a region as an example, Table 2 shows the correction relationships as follows:

[0179]

[0180] Due to this embodiment Substituting the actual time, we get:

[0181]

[0182] in, The B-phase component in 0T is from 0T s Correction to ABC windings Reference time; The C-phase component itself is located in No calibration required; The B phase component is composed of Correct to The X-phase component is composed of Correct to ;I sam4 The Y-phase component itself is located in The B phase component uses W 5075 The rate of change in the interval is Backtracking This process is related to Figure 2 This corresponds to the physical process of time correction from the actual sampling point to the compensated sampling point.

[0183] get Then, according to Figure 4 The "six-phase current reconstruction" step in the process is completed by referring to Table 3. (Still using...) Taking a region as an example, there are:

[0184]

[0185] From this, we obtain Unified correspondence Reference time, Unified correspondence Reference time. For For each region, the sampling current combination relationship is determined according to Table 1, the corresponding correction interval and current change rate are selected according to Table 2, and the six-phase current reconstruction is completed according to the corresponding expression in Table 3. The corrected six-phase reconstructed current is stored simultaneously and used as the current state of the previous cycle required for the coupling model calculation in the next PWM cycle.

[0186] like Figure 4 As shown, after completing the six-phase current reconstruction, the corrected reconstruction currents of the two windings are transformed using Clarke / Park coordinates. Since the reconstruction currents of windings ABC correspond to t... r1 =0.5T s The reconfiguration current of the XYZ winding corresponds to Therefore, the two windings respectively use the rotor electrical angle corresponding to their own reconstruction reference time for Park transformation. If the rotor electrical angle obtained by the resolver... Corresponding control operation or PWM update time ,but:

[0187]

[0188] In this embodiment Around 1000 r / min ,therefore:

[0189]

[0190] The Park transformation electrical angle difference between the two sets of windings due to the different reconstruction reference times is:

[0191]

[0192] use The Clarke / Park transformation is performed on the corrected reconfiguration current of the ABC winding. Clarke / Park transformation is performed on the XYZ winding correction and reconfiguration current:

[0193]

[0194]

[0195] The result As Figure 1 The feedback quantities of the two current loop ACRs shown are related to the corresponding... The shaft current setpoint is compared, and reference voltage commands for both windings are obtained via a PI current regulator. Subsequently, independent three-phase SVPWM is used to generate switching signals for both inverters, and the sampling, interval rate of change calculation, time correction, six-phase reconfiguration, and current closed-loop control processes are repeated in the next PWM cycle. The entire program execution relationship is as follows: Figure 4 The process shown corresponds to this.

[0196] To verify the effectiveness of the method, waveforms after the simulation entered stable operation were compared. Figure 5 Tables (a), (b), (c), and (d) respectively present the actual six-phase current, the six-phase reconfigured current without time correction, the reconfigured current after time correction without considering dual-winding coupling, and the reconfigured current after correction using the dual-winding coupling model of this invention. Figure 5 It is evident that time correction can reduce the reconstruction deviation caused by time-division sampling. Furthermore, considering the coupling of the two windings, the degree of agreement between the reconstructed waveform and the actual six-phase current is further improved.

[0197] Figure 6 Four raw sampled currents are given. With time correction This is used to illustrate that the combined current at different sampling times is corrected to the ABC winding after interval rate of change compensation. and XYZ windings Two reference times.

[0198] Figure 7 The representative actual current of phase A is given. and the actual current of phase Z With the reconfigured current considering coupling error compensation , The results show that the corrected reconstructed current can track the amplitude and phase changes of the actual phase current better. Figure 8 Further details are provided regarding the A-phase before correction. After phase A correction Before Z-phase correction After Z-phase correction The reconstruction error was significantly reduced after correction, thus verifying the effectiveness of the time-division sampling error compensation method based on the dual-winding coupling model.

[0199] This embodiment uses PI speed and PI current for verification. This invention mainly involves time-division sampling information acquisition from a single current sensor, calculation of average voltage and average back electromotive force within the PWM interval, solution of the rate of change of the coupled current in the dual windings, time correction of the sampling current, six-phase current reconstruction, and Park transformation electrical angle correction corresponding to the dual reference times. It does not impose limitations on the speed, current, or the specific form of the inverter topology.

[0200] 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 phase current reconfiguration error compensation method for a dual three-phase permanent magnet synchronous motor based on coupled inductors, characterized in that, include: Obtain the raw sampling current of a single current sensor within the current switching cycle; Determine each correction interval, and calculate the average voltage and average back electromotive force for each correction interval; The six-phase reconstructed current of the previous switching cycle is obtained as the current state. Based on the current state, the coupled inductance matrix including self-inductance and mutual inductance, the average voltage and average back electromotive force of each correction interval, the average current change rate of each correction interval in the dual αβ stationary coordinate system is solved. Then, the average current change rate of each correction interval is converted into the six-phase average current change rate of each correction interval through the three-phase stationary coordinate inverse transformation. The original sampled current is time-corrected using the six-phase average current change rate of each correction interval to obtain the corrected sampled current, and the corrected six-phase reconstructed current is obtained based on the corrected sampled current. The corrected six-phase reconstructed current is subjected to coordinate transformation to obtain the d-axis and q-axis feedback currents of two sets of three-phase windings.

2. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 1, characterized in that, The rate of change of the six-phase average current is: ； in, The average current change rate of the six phases. For the coupled inductor matrix, To correct the average voltage of the interval, For motor resistance, This is the six-phase reconfiguration current from the previous switching cycle. This is the average back electromotive force for the correction interval.

3. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 1, characterized in that, The method for performing coordinate transformation on the corrected six-phase reconstructed current includes: The Park transformation is performed on the current belonging to the first set of three-phase windings in the corrected six-phase reconstructed current using a first electrical angle, and on the Park transformation is performed on the current belonging to the second set of three-phase windings in the corrected six-phase reconstructed current using a second electrical angle. The first electrical angle and the second electrical angle are calculated based on the rotor electrical angle, the rotor electrical angular velocity, the first correction reference time, and the second correction reference time, respectively.

4. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 3, characterized in that, When the rotor electrical angle is acquired at the time of acquisition hour: ； When the rotor electrical angle corresponds to the control calculation time hour: ； in, For the first electrical angle, For the second electrical angle, The rotor electrical angle, The rotor's electric angular velocity, This is the first correction reference time. This is the second correction reference time.

5. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 1, characterized in that, A method for time-correcting the original sampled current by using the six-phase average current change rate of each correction interval to obtain the corrected sampled current includes: The current reconfiguration region is determined based on the reference voltage vector position or the switching states of the two inverters. Based on the current reconstruction region, determine the correspondence between each original sampled current and each phase current component of the first set of three-phase windings and the second set of three-phase windings: Based on the correspondence, determine the winding to which each phase current component in each original sampled current belongs and the corresponding correction reference time of the winding; Based on the current reconstruction region and the corresponding relationship, the correction interval corresponding to each phase current component is determined, and the six-phase average current change rate corresponding to the correction interval is called. Each phase current component is corrected from the actual sampling time to the correction reference time corresponding to its winding. If the correction reference time is later than the actual sampling time, forward correction is performed. If the correction reference time is earlier than the actual sampling time, reverse backtracking correction is performed through a negative time interval to obtain the corrected value of each phase current component. Based on the correspondence, the corrected values ​​of each phase current component belonging to the same original sampling current are recombined to obtain the corrected sampling current.

6. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 5, characterized in that, The sampling time of the original sampling current is , , and , The switching cycle.

7. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 6, characterized in that, Determining the correspondence between each original sampled current and the phase current components of the first and second three-phase windings includes: When the reconstructed region is hour, , , , ; When the reconstructed region is hour, , , , ; When the reconstructed region is hour, , , , ; When the reconstructed region is hour, , , , ; When the reconstructed region is hour, , , , ; When the reconstructed region is hour, , , , ; Among them, the four sampling times within the current switching cycle , , and Four raw sampling currents were obtained by sampling respectively. , , and ; This represents the current in phase B of the first three-phase winding ABC. This represents the current in phase C of the first three-phase winding ABC. This represents the current in phase X of the second three-phase winding XYZ. This represents the current in the Y phase of the second three-phase winding XYZ.

8. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 7, characterized in that, The current components of the first set of three-phase windings are corrected to the first correction reference time. The current components of the second set of three-phase windings are corrected to the second correction reference time. .

9. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 8, characterized in that, The method for determining the correction interval corresponding to each phase current component based on the current reconstruction region and the corresponding relationship includes: When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current The C-phase current component is not corrected; the original sampled current... The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The B-phase current component in the correction interval is used. Reverse backtracking correction to the first correction reference time ; When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current The Y-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The Y-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The C-phase current component is not corrected. When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current Both the X-phase current component and the Y-phase current component in the data use correction intervals. Corrected to the second correction reference time Original sampling current Both the X-phase current component and the Y-phase current component in the data use the correction interval. Corrected to the second correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The C-phase current component is not corrected. When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current Both the X-phase current component and the Y-phase current component in the data use the correction interval. Corrected to the second correction reference time Original sampling current Both the X-phase current component and the Y-phase current component in the data use the correction interval. Corrected to the second correction reference time Original sampling current The C-phase current component in the correction interval is used Corrected to the first correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The C-phase current component in the correction interval is used Reverse backtracking correction to the first correction reference time ; When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current Both the C-phase current component and the B-phase current component in the data use the correction interval. Corrected to the first correction reference time Sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current Both the C-phase current component and the B-phase current component in the data use the correction interval. Reverse backtracking correction to the first correction reference time ; When the reconstructed region is At that time, the original sampling current The B-phase current component in the correction interval is used. Corrected to the first correction reference time Original sampling current The C-phase current component is not corrected; the original sampled current... Both the C-phase current component and the B-phase current component in the data use the correction interval. Corrected to the first correction reference time Original sampling current The X-phase current component in the correction interval is used Corrected to the second correction reference time Original sampling current Both the C-phase current component and the B-phase current component in the data use the correction interval. Reverse backtracking correction to the first correction reference time .

10. The phase current reconstruction error compensation for a dual three-phase permanent magnet synchronous motor based on coupled inductors according to claim 9, characterized in that, The method for reconstructing the corrected six-phase reconstructed current based on the corrected sampled current includes: Select the corresponding six-phase current reconstruction expression according to the current reconstruction region, substitute the four corrected sampling currents into the six-phase current reconstruction expression to obtain the corrected six-phase reconstruction current; When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； When the reconstructed region is At that time, the expression for the six-phase current reconstruction is: ； in, This refers to the phase current after ABC correction for the first set of three-phase windings. This refers to the phase current after XYZ correction for the second set of three-phase windings. These are the four corrected sampling currents.

Images