Electrical machine and method for estimating the rotor current of the electrical machine

Abstract

The present invention relates to an electric machine (1) for electric or hybrid vehicles, comprising a rotating part, a transformer (6) and a converter (16) which is operated with phase shifts, wherein the converter (16) comprises a commutation cell (15) comprising a capacitor (17) on an input side of the converter (16) and several transistors (18) on an output side of the converter (16), wherein the commutation cell (15) comprises a current sensor (S) which is located between the capacitor (17) and the transistors (18).

Description

[0001] The present invention relates to the fields of electronics and electrical engineering, and in particular to the field of electrical machines.

[0002] Electric and hybrid vehicles can use electric motors to supply torque to the vehicle's wheels. The motor can be, for example, an electrically excited synchronous machine (EESM). The EESM is an alternative to permanent magnet synchronous machines, as the latter require the use of rare earth materials.

[0003] In an EESM (electrical energy storage system), energy must be transferred from a non-rotating part, i.e., the stator of the electric machine, to a rotating part, i.e., the rotor. Such power transfer can be either conductive or inductive.

[0004] In conductive power transmission, electrical contact is established by brushes on the non-rotating part pressing against slip rings on the rotating part. While the slip ring system of conductive power transmission allows for direct measurement of the rotor current, which is necessary for current and torque control, it causes both mechanical and electrical losses and requires regular maintenance.

[0005] In inductive power transfer, power is transmitted without contact and uses a rotating high-frequency transformer fed by a phase-shift full-bridge converter (PSFBC). A rectifier is connected between the transformer and the rotor winding of the electric machine. Inductive power transfer involves no mechanical losses and is maintenance-free. However, due to complexity and cost, placing a current sensor on the rotating part is highly undesirable, as this would require a sensor power supply at the rotating part of the machine and contactless signal transmission to the non-rotating part. Therefore, direct measurement of the rotor current is not feasible with current technology at a reasonable cost.

[0006] To accurately determine the rotor current required for efficient operation of the EESM, sensors can be used on the non-rotating part of the inductive power transfer system, and a predictive model is provided to estimate the rotor current. In other words, the sensors are positioned on a primary side of the non-rotating part to further estimate the current supplied on a secondary side, i.e., the rotating part.

[0007] For example, it is known to sample the primary-side transformer current at a time corresponding to the midpoint of a positive pulse in the transformer voltage and then apply a transformer transformation ratio. It is also known to use the input current, input voltage, and duty cycle of the converter and then apply a transformer transformation ratio.

[0008] However, errors in sampling the primary transformer current of a few hundred nanoseconds can lead to a significant degradation of the measurement. Additionally, when sampling in the primary transformer current, it is not possible to eliminate interference, as filtering is difficult to implement.

[0009] The present invention fits into this context by providing a more robust current estimation method that is less sensitive to timing errors and is accurate even under dynamic conditions. The invention also relates to a machine capable of implementing such a current estimation method based on a specific location of its current sensor.

[0010] In this context, the present invention relates to an electric drive system for electric or hybrid vehicles, comprising an electric machine comprising a rotating part, a transformer and a converter operated with phase shifts, wherein the converter comprises a commutation cell comprising a capacitor on an input side of the converter and several transistors on an output side of the converter, wherein the commutation cell comprises at least one current sensor located between the capacitor and the transistors.

[0011] According to an optional feature of the invention, the electrical machine according to the invention is an inductively excited synchronous machine or iEESM.

[0012] It includes, among other things, a rotating part or rotor, a DC / AC converter, a transformer and a rectifier.

[0013] According to an optional feature of the invention, the converter is a phase-shift full-bridge converter or PSFBC.

[0014] In a known embodiment of the converter, the multiple transistors in the commutation cell are organized in multiple switching arms, wherein the switching arms comprise two transistors arranged in series between a positive branch or upper branch and a negative branch or lower branch of the commutation cell, wherein the common junction point of the two transistors defines a junction point of the second side of the commutation cell, and wherein the capacitor of the commutation cell is connected between the positive and negative branches on the first side of the cell.

[0015] According to an optional feature of the invention, the converter is operated with phase shifts to which a duty cycle is assigned.

[0016] According to an optional feature of the invention, the converter operates with a variable PWM frequency. It provides bipolar square wave voltages which are applied to the transformer, more precisely to a primary side of this transformer. This primary side of the transformer is located on a non-rotating part of the electrical machine, while a secondary side of the transformer is located on the rotating part of the electrical machine.

[0017] The converter comprises a commutation cell having a first side on an input side of the converter and a second side on an output side of the converter. The commutation cell includes, among other things, a capacitor and several transistors. According to the invention, a current sensor is located on the non-rotating part of the electrical drive system. The current sensor is located in the commutation cell between the input side of the converter and the output side of the converter, more precisely between the capacitor and the several transistors.

[0018] The current sensor is used to acquire current measurements at the non-rotating part of the electrical drive system in order to compare these with an estimated current for the rotating part of the electric machine, i.e., its rotor. Positioning the current sensor within the converter's commutation cell allows for measurements that are less sensitive to sensor delay variations. Furthermore, the current sensor's location between the capacitor and the transistors enables a precise rotor current estimate under both steady-state and dynamic conditions.

[0019] According to an optional feature of the invention, the current sensor of the commutation cell is located on a positive branch of the commutation cell.

[0020] According to an optional feature of the invention, the current sensor of the commutation cell is located on a negative branch of the commutation cell.

[0021] The current sensor of the commutation cell can alternatively be located on either a positive branch (upper branch) or a negative branch (lower branch) of the commutation cell without affecting the current measurements. In some embodiments, multiple current sensors could be present in the commutation cell, for example, for safety reasons. For instance, at least two current sensors could be located on a positive branch of the commutation cell. In another example, at least one current sensor could be located on a positive branch and at least one sensor on a negative branch.

[0022] The present invention also relates to a method for estimating the current for the rotating part of the electric machine for electric or hybrid vehicles of the electric drive system, as previously described, comprising: - a measurement step in which current measurements are taken by the commutation cell current sensor within at least one PWM half-period, wherein the current measurements are taken for a duration that is at least as long as the duration of an active pulse of a voltage waveform generated by the converter between the converter and the transformer during the same PWM half-period; - an estimation step in which an absolute value of an average primary transformer current is estimated based on the current measurements taken during the measurement step; - A calculation step in which a rotor current estimate is calculated based on the estimate of the mean primary transformer current obtained in the estimation step.

[0023] The current estimation method according to the invention is intended to be used to determine the output current of the rotating part of the electric machine, for example, the rotor of the motor of an electric or hybrid vehicle. To estimate the rotor current, the current estimation method comprises several steps that are performed sequentially, including a measurement step, an estimation step, and a calculation step.

[0024] During the measurement step, the current sensor, located inside the converter between the capacitor and the transistors, measures the current. Such measurements are taken during at least one active pulse, be it a positive pulse and / or a negative pulse.

[0025] The current measurements obtained via the current sensor are then used during the estimation step to determine an absolute value of the average.

[0026] The primary transformer current is estimated. A rotor current estimate is then calculated using a mathematical formula based on the previously estimated average primary transformer current during an active pulse.

[0027] Once all steps of the current estimation method are complete, the rotor current value is known, and this value can be used to control the torque delivered to the wheels of the electric or hybrid vehicle. Unlike other methods that use current measurements taken in the converter input current or the primary transformer current, the present method is insensitive to sensor delays and can easily implement filtering to remove potential interference. Furthermore, the accuracy of this current estimation method is improved under both steady-state and dynamic conditions. In particular, the current sensor located in the commutation cell allows only the portions of the primary transformer current waveform relevant to the calculation step to be extracted.

[0028] According to an optional feature of the invention, the measurements are obtained by oversampling.

[0029] Using a single measurement for the current value can be considered a single-shot sampling method, while using multiple measurements can be considered an oversampling method. Single-shot sampling is easier to implement, while oversampling improves both accuracy and signal quality.

[0030] According to an optional feature of the invention, the PWM half-period in which the current measurements are performed comprises at least half of an active pulse of the voltage waveform.

[0031] The aim of the measurement step is to obtain measurements that correspond to the mean of the primary transformer current during the active pulse.

[0032] According to an optional feature of the invention, the PWM half-period begins in the middle of an active pulse of the voltage waveform.

[0033] Each active pulse is thus centered around the beginning of the PWM half-period. The duty cycle, for example, is updated in the middle of the active pulse.

[0034] According to an optional feature of the invention, the PWM half-period begins in a zero-voltage output phase in the middle between two successive active pulses of the voltage waveform.

[0035] Each active pulse is thus centered on the middle of a PWM half-period.

[0036] According to an optional feature of the invention, current measurements are taken during several PWM half-periods during the measurement step.

[0037] Using multiple PWM half-periods for current measurements ensures a more accurate measurement step.

[0038] According to an optional feature of the invention, current measurements are taken over the entire PWM half-period during the measurement step.

[0039] Such current measurements correspond to a first embodiment of the current estimation method according to the invention. Taking current measurements throughout the entire duration of the PWM half-period enables more robust acquisition, as it provides the average value over the entire time span as measured by the commutation cell current sensor. Taking current measurements over the entire PWM half-period requires specific hardware, such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or specialized microcontrollers with advanced analog-to-digital converter (ADC) peripherals.

[0040] According to an optional feature of the invention, during the measurement step the current measurements are taken for a duration that is shorter than the PWM half-period.

[0041] Such current measurements correspond to a second embodiment of the current estimation method according to the invention. Taking current measurements for a shorter duration than the PWM half-period is less expensive than taking measurements over the entire PWM half-period. Consequently, the second embodiment can be implemented when ASIC- or FPGA-based acquisitions or solutions using specialized microcontrollers with advanced ADC peripherals are considered too expensive.

[0042] According to an optional feature of the invention, the current measurements are performed for a period of time that is longer than the active pulse of the voltage waveform.

[0043] This ensures that the current measurements robustly correspond to the average primary transformer current during the active pulse.

[0044] According to an optional feature of the invention, the current estimation method comprises a detection step in which a hardware circuit detects active pulse points.

[0045] Such current measurements correspond to a third embodiment of the current estimation method according to the invention, in which an average current value is obtained. This third embodiment comprises a hardware circuit that uses the PWM signals or an edge detection circuit to detect the active pulses. This results in a highly accurate reconstruction of the primary transformer current mean during a specific active pulse. This third embodiment is accurate under steady-state conditions and can be easily implemented.

[0046] According to an optional feature of the invention, the measuring step begins at the beginning of a specific active pulse and ends at the end of the active pulse.

[0047] In other words, the measuring points are chosen so that they surround a specific pulse of the voltage waveform without any delay before or after the pulse.

[0048] According to an optional feature of the invention, the current measurements are shifted from a midpoint of the active pulse.

[0049] For example, the measurement step can begin at the start of a specific active pulse and end after the active pulse has finished. In other words, the current measurements take into account the active pulse plus some delays. This accounts for potential sensor delays and / or delays in PWM generation.

[0050] According to an optional feature of the invention, the measurement step begins before a certain active pulse in a zero output voltage phase and ends after the end of the active pulse.

[0051] According to an optional feature of the invention, the calculation step takes into account a transformer conversion ratio.

[0052] The transformer transformation ratio is a ratio between the inductance and mutual inductance of the transformer.

[0053] According to an optional feature of the invention, the calculation step takes into account an effective translation ratio.

[0054] Such an effective conversion ratio is required when the inductance values ​​are not constant but current-dependent, for example due to magnetic saturation.

[0055] According to an optional feature of the invention, the current estimation method includes a calibration step in which measurements are taken under zero current conditions.

[0056] In other words, measurements are taken between active pulses. This calibration step helps to learn about any sensor offset that might result from temperature drift or lifetime drift, thus improving the accuracy of the rotor current estimation. Such sensor offset measurements can be performed using successive approximation register (SAR) or delta-sigma (DS) analog-to-digital converters (ADCs). Using multiple measurements under zero-current conditions helps improve the quality of the offset learning. The number of such measurements can vary depending on the specific implementation. For example, the number of measurements may depend on the duty cycle and / or switching frequency of the converter.

[0057] Alternatively, the sensor offset can be learned by interrupting PWM generation for a predetermined time or by performing measurements when no duty cycle is requested. In either case, the calibration step can be performed before the measurement step during a period without active pulses, e.g., during the start-up of the electronic control unit of the electric machine.

[0058] According to an optional feature of the invention, the current estimation method comprises an averaging step in which the measurements taken in a PWM half-period corresponding to a positive pulse and the measurements taken in a PWM half-period corresponding to a negative pulse are averaged.

[0059] The averaging step helps to improve the accuracy of the rotor current estimation under dynamic conditions.

[0060] According to an optional feature of the invention, the current estimation method comprises a correction step in which the mean primary transformer current estimated during the estimation step is corrected by primary transformer current measurements at zero voltage.

[0061] Compared to the averaging step, the correction step is another way to improve the accuracy of rotor current estimation under dynamic conditions. The correction step allows for waveform correction when an offset exists due to a transformer flux imbalance. The primary transformer current measurements at zero voltage are taken around a specific active voltage pulse.

[0062] According to an optional feature of the invention, the current estimation method is combined with a machine model by embedding the machine model in an observer structure.

[0063] Combining the current estimation method with a machine model improves the accuracy of the rotor current estimation, particularly under dynamic conditions and when the converter is operating at a near-zero duty cycle. The machine model can, for example, help to reduce the deviations that commonly occur under dynamic conditions. The observer structure could be, for example, a Luenberger observer, an extended Luenberger observer, or a Kalman filter.

[0064] Further features, details and advantages of the invention will become clearer on the one hand by reading the following description and on the other hand by several examples of implementation, which are given as guidelines and without limitation with reference to the attached schematic drawings: [ Fig. 1] is a schematic representation of an electric drive system for an electric or hybrid vehicle, comprising an electric machine comprising a rotor, a stator and a transformer comprising a rectifier and a transformer; [ Fig. 2] is another schematic representation of a part of the electrical machine made of Fig. 1, which further includes a converter; [ Fig. Figure 3] is a schematic representation of a first general embodiment of a current estimation method according to the invention, wherein such a method is used to determine a current value for the rotor of Fig. 1 to determine; [ Fig. Figure 4] is a schematic representation of a second general embodiment of the current estimation method according to the invention; [ Fig. Figure 5] is a schematic representation of a first specific embodiment of the current estimation method according to the invention; [ Fig. Figure 6] is a schematic representation of a second specific embodiment of the current estimation method according to the invention; [ Fig. Figure 7] is a schematic representation of a third specific embodiment of the current estimation method according to the invention.

[0065] The features, variants, and different ways of implementing the invention can be combined in various ways, provided they are not incompatible or mutually exclusive. In particular, variants of the invention are conceivable that include only a selection of the features described below from the other described features, if this selection of features is sufficient to achieve a technical advantage and / or to differentiate the invention from the prior art.

[0066] In the drawings, identical numbers refer to identical elements.

[0067] Fig. 1 and Fig. Figure 2 shows schematic representations of an electric drive system comprising an electric machine 1, or a part thereof. The electric drive system and its electric machine 1 are intended for installation in a vehicle, for example, an electric or hybrid vehicle, where they can be used to provide torque to the vehicle's wheels. Here, the electric machine 1 is an inductively excited synchronous machine, also known by its abbreviation iEESM.

[0068] As in Fig. As can be seen in Figure 1, the electric machine 1 comprises a rotor 2 and a stator 4. In this example, the stator 4 is positioned around the rotor 2; however, in other embodiments, the rotor 2 could be positioned around the stator 4. In inductively excited synchronous machines like this electric machine 1, the power is transferred to a winding attached to the rotor 2 by means of a transformer 6. More precisely, the transformer 6 is a rotating high-frequency transformer. The transformer 6 comprises a primary side 8 and a secondary side 10, these two sides 8 and 10 differing in that the primary side 8 is located on a non-rotating part of the electric machine 1, i.e., the stator, while the secondary side 10 is located on a rotating part of the electric machine 1, i.e., the rotor. Fig. 1 is a separation between the non-rotating part and the rotating part, represented as a dashed line.

[0069] A rectifier 12 is associated with the transformer 6. The rectifier 12 is a rotating rectifier and, as such, is located on the secondary side 10 of the transformer 6, i.e., on the rotating part of the electric machine 1. The function of the rectifier 12 is to supply only positive voltage to the rotor 2. This rectifier 12 is implemented together with the transformer 6 in a transformer 14, in this case an inductive transformer.

[0070] As in Fig. As can be seen in Figure 2, a converter 16 is connected to the primary side 8 of the transformer 6. The converter 16 is a phase-shift full-bridge converter, or PSFBC. It operates with a variable pulse-width modulation frequency, or PWM frequency. The PWM frequency of the converter 16 can indeed vary over a wide range, for example from 5 to 100 kHz, depending on the design of the converter 16, the transformer 6, and the current required in the rotor 2.

[0071] The converter 16 comprises a commutation cell 15, which includes an input capacitor 17 and several transistors on an output side of the converter 16. According to the invention, a current sensor S is placed within the commutation cell 15 of the converter 16. More precisely, the current sensor S is located between the input capacitor 17 and the transistors 18. As shown in Fig. Figure 2 illustrates that the current sensor S is shown placed on a positive branch of the commutation cell 15. In other embodiments, however, the current sensor S could alternatively be positioned on a negative branch of the cell 16.

[0072] The converter 16 is controlled by an electronic control unit of the electric machine, such as a microcontroller, a field-programmable gate array (FPGA), or a digital signal processor (DSP), to convert direct current supplied by a vehicle battery into alternating current and thus apply a corresponding voltage to the primary side 8 of the transformer 6. As shown here, the converter 16 comprises four transistors 18, with two upper transistors 18A and two lower transistors 18B. Each upper transistor 18A is associated with one lower transistor 18B, forming a respective half-bridge or phase branch. The upper 18A and the lower transistor 18B of each half-bridge are driven inversely. To prevent a short circuit at the converter input, an additional latching time is introduced in the command between the upper 18A and the lower 18B transistor of each phase branch.This additional locking time results in a delayed turn-on command for the upper transistor 18A or the lower transistor 18B of a half-bridge compared to the turn-off command for the upper transistor 18A or the lower transistor 18B of a half-bridge.

[0073] The converter 16 is operated by phase shifts, with each phase shift ranging from 0 to 180°. Additionally, each phase shift of the converter 16 is assigned a duty cycle, which ranges from 0 to 1. For example, a phase shift of 0° corresponds to a duty cycle of 0, a phase shift of 90° to a duty cycle of 0.5, and a phase shift of 180° to a duty cycle of 1.

[0074] The converter 16 generates a voltage waveform 20; this waveform 20 is in Fig. 3, Fig. 4, Fig. 5, Fig. 6 to Fig. 7 visible. The voltage waveform 20 can be divided into a plurality of PWM periods 22, wherein in the waveform 20, for each PWM period 22, both a positive pulse 24 and a negative pulse 26 occur. With a duty cycle of 1, positive or negative voltages are applied successively during all given PWM periods, without zero voltage values, with each positive or negative voltage being applied for half of the PWM period, whereas with a duty cycle of, for example, 0.5, positive and negative voltages are applied for a shorter duration during the given PWM period, with zero voltage values ​​in between, with each positive or negative voltage being applied for half the time as is the case with a duty cycle of 1.

[0075] As mentioned previously, the electric machine 1 can be used to supply torque to the wheels of the hybrid or electric vehicle. To control the torque and, more generally, to ensure efficient operation of the electric machine 1, it is necessary to accurately determine the current of the rotor 4.

[0076] For this purpose, a rotor current estimation method according to the invention can be used. This current estimation method, which comprises several embodiments, is now described in relation to Fig. 3, Fig. 4, Fig. 5, Fig. 6 to Fig. Figure 7 describes various waveforms. These waveforms correspond to the ideal voltage waveform 20 generated by the transducer 16, an idealized input current waveform 28 in the commutation cell 15 of the transducer 16, and a sensor signal waveform 30 obtained by the current sensor S. The voltage waveform 20 comprises several active pulses, with at least one positive pulse 24 and at least one negative pulse 26.

[0077] Fig. 3 and Fig. Figure 4 each corresponds to a first general embodiment and a second general embodiment of the current estimation method. In both the first and the second general embodiment, the rotor current estimation method can include a preliminary step, which is a calibration step. This calibration step serves to detect potential offsets in the function of the current sensor S. The calibration step is performed by taking measurements in the voltage waveform 20 under zero-current conditions. The calibration step can be performed, for example, when the electronic control unit of the converter 16, i.e., its microcontroller, is started, before active pulses are generated. Alternatively, the calibration step can be performed while active pulses are being generated, in which case the measurements are taken under zero-current conditions between active pulses.The calibration step can then be performed simultaneously with other steps of the rotor current estimation method.

[0078] The rotor current estimation method comprises one measurement step. During this measurement step, the current sensor S of the commutation cell 15 of the converter 16 takes current measurements for a duration at least as long as the duration of an active pulse of the voltage waveform 20. In other words, a detection period 32 in the voltage waveform 20 is at least as long as an active pulse of the voltage waveform 20. The active pulse is either a positive pulse 24 or a negative pulse 26. It should be noted that such current measurements can be acquired either by single-shot sampling, i.e., by sampling a single value, or by oversampling, i.e., by sampling a multitude of values ​​within the same active pulse.

[0079] In the general embodiment of Fig. 3, i.e., in the first general embodiment, the acquisition period 32, during which the current measurements are taken, covers at least one full PWM half-period. There may be multiple acquisition periods 32, covering multiple full PWM half-periods. For this purpose, the converter 16 includes specific hardware designed to perform measurements for the entire duration of the PWM half-period. This specific hardware may be, for example, an application-specific integrated circuit, a field-programmable gate array, or specialized microcontrollers with advanced ADC (analog-to-digital) peripherals. In the general embodiment of Fig. In the second general embodiment, the sensing period 32, during which the current measurements are taken, covers a duration shorter than a full PWM half-period. The second general embodiment therefore does not require specific hardware designed to perform measurements for the entire duration of the PWM half-period, as was the case in the first general embodiment. Although the current measurements taken by the current sensor S in the second general embodiment do not cover a full PWM half-period, as in the first general embodiment, it is also possible to cover multiple PWM half-periods with multiple sensing periods 32. In any case, in this second general embodiment, the sensing period 32 is taken for a duration longer than a specific active pulse.For example, the detection period 32 can cover a duration that begins slightly before a particular active pulse and ends slightly after the active pulse.

[0080] In the first general embodiment of Fig. 3, which represents steady-state conditions with a constant duty cycle in the relevant PWM period and the preceding PWM period, the PWM half-period begins in the middle of the positive pulse 24 and ends in the middle of the negative pulse 26 of the voltage waveform 20. In the case of a constant duty cycle, the PWM half-period is thus aligned with a midpoint of the voltage waveform 20 between two successive active pulses. Therefore, each PWM half-period in this sequence comprises half an active pulse, a zero-current period, and half an active pulse. In the second general embodiment of Fig. 4. Each PWM half-period is centered on the middle of one of the active pulses, meaning that the PWM half-period begins midway between two consecutive active pulses of the voltage waveform 20. The PWM half-period is thus aligned with the middle of an active pulse, i.e., centered. Consequently, each PWM half-period comprises a zero-current period, an entire active pulse, and another zero-current period. However, there may be some embodiments where no zero-current periods are present within the complete PWM half-period, for example, embodiments with a duty cycle of 1.

[0081] It should be noted that, although previously the first general embodiment with a PWM half-period aligned with the midpoint between two successive active pulses was described, and the second general embodiment with a PWM half-period aligned with the midpoint of an active pulse was described, the rotor current estimation method according to the invention also intends to cover other embodiments, in which the first general embodiment, in which the detection period 32 covers at least one full PWM half-period, would cover a PWM half-period aligned with the midpoint of an active pulse, and the second general embodiment, in which the detection period 32 covers a duration shorter than one full PWM half-period, would cover a PWM half-period aligned with the midpoint between two successive active pulses.

[0082] Additionally, the measurement step of the rotor current estimation method can be implemented differently than previously described if it is performed according to a third general embodiment not shown here. In this third general embodiment, the measurement step includes a sub-step that is a detection step. During this detection step, a hardware circuit of the converter 16 detects active pulse points in the voltage waveform 20 in order to then perform the current measurements of the measurement step. More precisely, the hardware circuit is used to analyze the output voltage of the converter 16. This analysis includes the detection of voltage edges by a comparator stage. The edges are then used to successively start and reset a function for calculating an average, which is performed, for example, by an integrator stage.The resulting calculated average signal is then temporarily stored, and an average value is fed to the converter.

[0083] In the third general embodiment, the detection period 32 can either begin at the beginning of a specific active pulse and end at the end of the active pulse, or it can begin at the beginning of a specific active pulse and end after the end of the active pulse, thus taking into account a delay after the active pulse.

[0084] For each of the three general embodiments, the rotor current estimation method can include an additional averaging step. In the averaging step, the measurements taken during the measurement step in a detection period 32 covering a positive pulse 24 and the measurements taken during the measurement step in a detection period 32 covering a negative pulse 26 are averaged together.

[0085] After the current measurements are taken, the rotor current estimation method includes an estimation step. During this estimation step, the current measurements are used to estimate an absolute value of the mean primary transformer current. More precisely, the current measurements are used to estimate an absolute value of the mean primary transformer current in a given active pulse. Such an estimate is based on the following formula: i1,abs,PwrBaiCommCell≈ηICommCell,avg,HalfPeriodδHalfPeriod where: “i 1,abs,PwrBalCommCel “ the absolute value of the average primary transformer current to be estimated; “η” is an efficiency parameter of the converter 16 and the transformer 6; “I Commcell,avg,HalfPeriod “ is an average input current of the commutation cell for the PWM half-period under consideration; “δ HalfPeriod “ the duty cycle of converter 16 for the PWM half-period under consideration is.

[0086] The average input current of the commutation cell is calculated according to the following formula: ICommCell,avg,HalfPeriod=1Tacq∫tStarttEndiSensCommCell(τ)dτ

[0087] To implement such an integral, the beginning of the recording period 32 is used to define "t start “to define, and the end of the recording period 32 is used to “t End“to define. The beginning and end of the recording period 32 may differ according to the general embodiment used, as previously described in detail.

[0088] The primary transformer current estimate may need to be corrected. For this purpose, a correction step is performed in the rotor current estimation method. The correction step is described below with respect to... Fig. This correction step is described further in section 7. This step helps to avoid an offset due to the flux imbalance of transformer 6. To perform the correction step, primary transformer current measurements are taken in the primary transformer current waveform at zero voltage. More precisely, the zero-voltage measurements are taken so that they surround an active pulse of the primary transformer current waveform. The zero-voltage measurements are taken either at intervals of half a PWM period, or oversampling is performed. After acquiring the zero-voltage measurements, they are averaged to obtain a correction value. The correction value is then subtracted from the absolute value of the mean primary transformer current obtained during the estimation step to obtain a corrected mean primary transformer current value.

[0089] After the absolute value of the mean primary transformer current has been estimated and possibly corrected, the rotor current estimation method according to the invention comprises a calculation step. During such a calculation step, the result of the estimation step, i.e., the estimate of the mean primary transformer current, or optionally the result of the correction step, i.e., the corrected mean primary transformer current value, is used to calculate a rotor current estimate. For this calculation step, a transformer turns ratio is taken into account, which corresponds to a relationship between the inductance of transformer 6 and its mutual inductance. The rotor current estimate is calculated more accurately using the following formula: iF≈L11M12i1,abs,PwrBalCommCell where: “i F “the rotor current value to be estimated; “L 11 “the inductance of transformer 6 is; "M12 “the mutual inductance of transformer 6 is; and “i 1,abs,PwrBalCommCell “The mean primary transformer current value is the one obtained during the estimation step.

[0090] Alternatively, the value of the rotor current can also be estimated using the following formula: iF≈xi1,abs,PwrBalCommCell where “x” is an effective turns ratio that can be used instead of the transformer turns ratio when the inductance values ​​are current-dependent and not constant.

[0091] It should be noted that the inventive method for rotor current estimation can be combined with a machine model to improve the accuracy of the rotor current estimation. The machine model and the rotor current estimation method can be used in an observer structure, which is selected, for example, from a Luenberger observer, an extended Luenberger observer, or a Kalman filter.

[0092] Three specific embodiments of the rotor current estimation according to the invention are now described in relation to Fig. 5, Fig. 6 and Fig. Figure 7 describes, each representing a first specific embodiment, a second specific embodiment, and a third specific embodiment. These specific embodiments describe in detail possible acquisition strategies used to implement the measurement step. In these Fig. 5, Fig. 6 to Fig. Figure 7 shows the voltage waveform 20 generated by the converter 16, shown in solid lines; the idealized input current waveform 28 in the commutation cell 15 of the converter 16, shown in dashed lines; and the idealized sensor signal waveform 30 obtained by the current sensor S, shown in dashed lines. The PWM frequency within a given evaluation period can be constant or variable, depending on the embodiment. Here, the waveforms 20, 28, and 30 for different evaluation periods, corresponding to multiples of a PWM period 22, are shown. From top to bottom, the waveforms for evaluation periods corresponding to half a PWM period 34, a PWM period 22, 3 / 2 a PWM period 36, and two PWM periods 38 are shown. Fig. 5 and Fig. Figure 7 also shows waveforms 20, 28, 30 for an evaluation period corresponding to 2.5 PWM periods 40. In these specific embodiments, the evaluation period is measured between two duty cycle updates. Therefore, the duty cycle is constant within the evaluation period. However, other embodiments of the invention may exist in which the duty cycle is variable within the evaluation period.

[0093] In the first specific embodiment of Fig. 5 covers the acquisition period 32, in which the current measurements are taken, a duration that is shorter than a full PWM half-period, as is the case in the second general embodiment of Fig. 2 was the case, but the PWM half-period is aligned with the midpoint of an active pulse, as in the first general embodiment of Fig. 1 was the case. Here, the acquisition period 32 is synchronized with the PWM generation, whereby both the PWM generation and the duty cycle are constant for the duration of the acquisition period 32. As in Fig. As illustrated in Figure 5, due to technical limitations of the microcontroller of converter 16, there is no acquisition period 32 for the evaluation period of half a PWM period 32. Here, the acquisition period 32 for the other evaluation periods of Fig. 5 centered, meaning that it begins in the middle between two successive active pulses of the voltage waveform 20. The duration of the detection period 32 can be defined as follows: Tacq=(N−1)*TPWM / 2=(N−1) / N*Teval where “T acq “ the duration of the recording period is 32; “N” is the number of PWM half-periods in the evaluation period under consideration; “T PWM “the PWM period is; “T eval“the duration of the evaluation period is.”

[0094] To implement the measurement step of the rotor current estimation method, an idealized measurement window 42 is defined. It should be noted that in the specific embodiments of the Fig. 5, Fig. 6 to Fig. Figure 7 shows the evaluation period comprising only one measurement window 42. However, in other embodiments of the rotor current estimation method according to the invention, an evaluation period could be present that contains several measurement windows 42.

[0095] This measurement window 42 is shifted relative to the acquisition period 32 to account for any potential delays. Such delays can correspond, for example, to delays in PWM generation and / or delays caused by the current sensor S. Positioning the current sensor S in the commutation cell 15 according to the invention results in robust measurements. While shifting the measurement window 42 relative to the acquisition period 32 is not strictly necessary due to this positioning of the current sensor S, it is particularly useful under highly dynamic conditions, as it enables optimal results. In other embodiments, for example, under steady-state conditions, the measurement window 42 could be aligned with the acquisition period 32.

[0096] After the current measurements in the measuring window 42 have been performed, as described in the general embodiments mentioned above, the estimation step of the rotor current estimation method is carried out. The mean input current of the commutation cell is calculated according to the following formula, which was described in detail in the previous estimation step: ICommCell,avg=1 / Tacq∫tacq,starttacq,endiSensCommCell(τ)dτ where: “ CommCell,avg,HalfPeriod “ the average input current of the commutation cell for the considered acquisition period 42 is; “T acq “ the duration of the considered recording period is 32; “T acq,start “ the beginning of the considered recording period is 32; “T acq,end “ the end of the considered recording period is 32.

[0097] The calculated mean input current of the commutation cell is then used in the above-mentioned formula to calculate the absolute value of the primary transformer current, such a formula being described in the estimation step: i1,abs,PwrBalCommCell≈ηICommCell,avg,TacqδTacq

[0098] Once the absolute value of the primary transformer current is known, it is used in the calculation step to determine the rotor current according to the formula described above: iF≈L11M12i1,abs,PwrBalCommCell

[0099] The second specific embodiment is in Fig. Figure 6 illustrates this. It differs from the first specific embodiment in that the PWM generation is centrally aligned, since here the PWM generation is centered on the beginning of the evaluation period, as is the case with the second general embodiment of Fig. 4 was the case. Rather, in the first specific embodiment, the PWM generation was shifted, e.g., aligned to the left or to the right compared to the beginning of the evaluation period.

[0100] This second specific embodiment makes it possible, depending on the selected duty cycle, to completely cover the active pulses in the current waveform over the evaluation period. In this case, the duration of the detection period 32 corresponds to the duration of the evaluation period minus two gaps, with a first gap before the detection period 32 and a second gap after the detection period 32. Consequently, the formula used to calculate the average input current of the commutation cell is as follows: ICommCell,avg=1 / Teval∫tacq,starttacq,endiSensCommCell(τ)dτ where: “T eval “the duration of the evaluation period under consideration.”

[0101] The calculated mean input current of the commutation cell can then be used in the formula to determine the absolute value of the primary transformer current, which is used during the previously mentioned estimation step and the subsequent calculation step.

[0102] The third specific embodiment is now described with reference to Fig. 7 described. As was the case with the first specific embodiment, the PWM generation is offset here, e.g. left-aligned or right-aligned compared to the beginning of the evaluation period.

[0103] The third specific embodiment is more precise than the first specific embodiment because it illustrates the correction step mentioned above. After the current measurements in the voltage waveform 20 have been performed according to the measurement step and after the estimation step has been carried out, the correction step is implemented.

[0104] During this correction step, as previously described, the primary transformer current measurements are taken in the primary transformer current waveform at zero voltage. For the evaluation period, which corresponds to the PWM period 22, a first primary transformer current measurement M1 is taken before the active pulse, which here is a negative pulse 26, and a second primary transformer current measurement M2 is taken after the negative active pulse. The first primary transformer current measurement M1 and the second primary transformer current measurement M2 are taken at intervals corresponding to half a PWM period.

[0105] Similarly, for the evaluation period, which corresponds to two PWM periods 38, the first primary transformer current measurement M1 is taken before the first negative pulse 26 of the acquisition period 32, and the second primary transformer current measurement M2 is taken after the negative pulse 26. Additionally, a third primary transformer current measurement M3 is taken after the positive pulse 24, which is located in the middle of the acquisition period 32, and a fourth primary transformer current measurement M4 is taken after the second negative pulse 26 of the acquisition period 32.Consequently, the first primary transformer current measurement M1 and the second primary transformer current measurement M2 surround the first negative pulse 26 within the acquisition period 32, the second primary transformer current measurement M2 and the third primary transformer current measurement M3 surround the positive pulse 24 in the middle of the acquisition period 32, and the third primary transformer current measurement M3 and the fourth primary transformer current measurement M4 surround the second negative pulse 26 of the acquisition period 32. The same reasoning for sampling primary transformer current measurements can be applied to each evaluation period.

[0106] The correction value is then obtained from the primary transformer current measurements. Obtaining the correction value depends on the number of active pulses in the acquisition period 32. If there is only one active pulse in the acquisition period 32, the correction value corresponds to the average of the two surrounding primary transformer current measurements. For example, in the case of an evaluation period corresponding to two PWM periods 38, the first primary transformer current measurement M1 and the second primary transformer current measurement M2, which surround the negative pulse 26, are averaged together.

[0107] In the case of an evaluation period that includes several active impulses, for example a number of n active impulses, the correction value is obtained according to the following formula: iPrimTrafoCurr, correction=0.5 iPrimTrafoCurr,1+Σk=2n−1iPrimTrafCurr,k+0.5 iPrimTrafoCurr,n

[0108] The correction value is then subtracted from the absolute value of the mean primary transformer current obtained during the estimation step to obtain the corrected mean primary transformer current value. This corrected mean primary transformer current value can then be used, as described above, during the calculation step to obtain the rotor current estimate.

[0109] The present invention thus relates to an electric machine of an electric drive system in which a specific position of a current sensor enables an accurate rotor current estimation under both steady-state and dynamic conditions, and to a rotor current estimation method based on measurements taken by the current sensor.

[0110] Many modifications and other embodiments of the invention presented herein will be obvious to a person skilled in the art who is familiar with the teachings presented in the preceding descriptions and the accompanying figures. It is therefore understood that the invention is not limited to the specific embodiments disclosed and that modifications and further embodiments are included within the scope of the appended claims. Although certain terms are used herein, they are used only in a general and descriptive sense, and not for the purpose of limitation.

Claims

[1] Electric drive system for electric or hybrid vehicles comprising an electric machine (1) comprising a rotating part, a transformer (6) and a converter (16) operating with phase shifts, wherein the converter (16) comprises a commutation cell (15) comprising a capacitor (17) on an input side of the converter (16) and several transistors (18) on an output side of the converter (16), wherein the commutation cell (15) comprises at least one current sensor (S) located between the capacitor (17) and the transistors (18). [2] Electric drive system for electric or hybrid vehicles according to the preceding claim, wherein the current sensor (S) of the commutation cell (15) is located on a positive branch of the commutation cell (15). [3] Electric drive system for electric or hybrid vehicles according to claim 1, wherein the current sensor (S) of the commutation cell (15) is located on a negative branch of the commutation cell (15). [4] Method for estimating the current for the rotating part of the electric machine (1) for electric or hybrid vehicles of the electric drive system according to claims 1 to 3, comprising: - a measurement step in which current measurements are taken within at least one PWM half-period by the current sensor (S) of the commutation cell (15), wherein the current measurements are carried out for a duration that is at least as long as the duration of an active pulse of a voltage waveform (20) generated by the converter (16) between the converter (16) and the transformer (6) during the same PWM half-period; - an estimation step in which an absolute value of an average primary transformer current is estimated based on the current measurements taken during the measurement step; - a calculation step in which a rotor current estimate is calculated based on the estimate of the mean primary current obtained in the estimation step. [5] Method for current estimation according to claim 4, wherein the measurements are obtained by oversampling. [6] Method for current estimation according to one of claims 4 and 5, wherein the PWM half-period in which the current measurements are carried out comprises at least half of an active pulse of the voltage waveform (20). [7] Method for current estimation according to the preceding claim, wherein the PWM half-period begins in the middle of an active pulse of the voltage waveform (20). [8] Method for current estimation according to claim 6, wherein the PWM half-period begins in a zero-voltage output phase in the middle between two successive active pulses of the voltage waveform (20). [9] Method for current estimation according to any one of claims 4 to 8, wherein during the measurement step the current measurements are taken during several PWM half-periods. [10] Method for current estimation according to any one of claims 4 to 9, wherein during the measurement step the current measurements are carried out over the entire PWM half-period. [11] Method for current estimation according to any one of claims 4 to 9, wherein during the measurement step the current measurements are carried out for a duration which is shorter than the PWM half-period. [12] Method for current estimation according to claim 11, wherein the current measurements are carried out for a duration longer than the active pulse of the voltage waveform (20). [13] A method for current estimation according to any one of claims 4 to 12, comprising a detection step in which a hardware circuit detects active pulse points. [14] Method for current estimation according to claim 13, wherein the measurement step begins at the beginning of a certain active pulse and stops at the end of the active pulse. [15] Method for current estimation according to claim 13, wherein the current measurements are shifted from a center point of the active pulse. [16] Method for current estimation according to any one of claims 1 to 12, wherein the measurement step begins in a zero output voltage phase before a certain active pulse and stops after the end of the active pulse. [17] Method for estimating current according to any one of claims 4 to 16, wherein the calculation step takes into account a transformer conversion ratio. [18] Method for estimating current according to any one of claims 4 to 17, wherein the calculation step takes into account an effective transmission ratio. [19] A method for estimating current according to any one of claims 4 to 18, comprising a calibration step in which measurements are taken under zero current conditions. [20] A method for estimating current according to any one of claims 4 to 19, comprising an averaging step in which the measurements taken in a PWM half-period corresponding to a positive pulse and the measurements taken in a PWM half-period corresponding to a negative pulse are averaged. [21] Method for current estimation according to any one of claims 4 to 20, comprising a correction step in which the mean primary transformer current estimated during the estimation step is corrected by primary transformer current measurements (M1, M2, M3, M4) at zero voltage. [22] A method for estimating current according to any one of claims 4 to 21, which is combined with a machine model by embedding the machine model in an observer structure.

Images