Vector control method for a synchronous machine and control unit

Abstract

Vector control method for a synchronous machine, in particular a permanent magnet synchronous machine, in which Q-axis and D-axis setpoint currents (Iq, Id) are determined by respective assignments (5, 5A) that define a flux linkage torque (M&PSgr;) and a setpoint flux strength (&PSgr;Tgt_MTPL) with the Q-axis setpoint current and D-axis setpoint current, respectively.The D-axis target current (Iq, Id) is determined by linking the flux linkage torque (M) to a difference between a target torque (MTgt) and an iron loss torque (MIron), wherein the assignments (5, 5A) for determining the Q-axis target current (Iq) and the D-axis target current (Id) disregard a voltage variation of a voltage to be applied to the stator windings of the synchronous machine and the voltage variation is previously regulated or compensated by adjusting the target flux strength (Tgt_MTPL), characterized in that the voltage variation is regulated or compensated on the basis of a flux strength difference (Error) between a maximum flux strength (Max) and a current flux strength (ctrl_req).

Description

The invention relates to a vector control method for a synchronous machine, in particular a permanent magnet synchronous machine, and a control unit which is designed and programmed to execute the vector control method. The document RUF, Andreas [et al.]: Loss minimizing control strategy for electrical machines considering iron loss distributions. In: 2015 IEEE International Electric Machines & Drives Conference (IEMDC): 10-13 May 2015, Coeur d'Alene, Idaho, USA. New York, NY: IEEE, 2015. pp. 974-980. ISBN 978-1-4799-7941-7 discloses a control unit on which a vector control method can be executed. For optimal efficiency in controlling a permanent magnet synchronous machine (PMSM), losses must be taken into account with sufficient accuracy. In a PMSM, losses can typically be divided into stator resistance losses and iron losses. While the stator resistance behaves essentially like an ohmic resistor, causing an additional voltage drop and consequently heat losses, the iron losses result from rotating magnetic fields within the synchronous machine. Since iron losses have a braking effect, they should be compensated for to accurately adjust the torque and determine appropriately adapted target currents to minimize the overall losses of the synchronous machine. Typically, the efficiency-optimized currents for operating the synchronous machine are stored in lookup tables. One table is used to determine the D-axis current and another to determine the Q-axis current. These tables contain the currents as a function of a target torque (positive and negative), a rotational speed, a voltage (especially a DC voltage) that can be applied to the stator windings by the power electronics unit, and a rotor temperature. Such tables can take the iron loss torque into account to enable optimized operation of the synchronous machine, but require a significant amount of memory due to the four input variables and are also complex to calibrate. Alternatively, the currents can be defined as a function of a target torque, a rotational speed, and a rotor temperature. In this case, the rotational speed is scaled according to the ratio of the voltage that can be applied to the stator windings by the power electronics to the nominal voltage of the power electronics, i.e., the maximum voltage that can be applied by the power electronics. For example, a ratio of 0.5 doubles the rotational speed. In this way, the iron loss torque is not correctly compensated at all voltages, since the synchronous machine is not always operated at its optimal operating point, and there is no longer a separate speed dependency. It is therefore an object of the present invention to provide an improved vector control method and a control unit. This object is achieved by the vector control method and the control unit with the features according to the independent claims. Advantageous embodiments are the subject of the dependent claims. In a vector control method according to the invention for a synchronous machine, in particular a permanent magnet synchronous machine, Q-axis and D-axis target currents are determined by respective assignments that link a flux linkage torque and a target flux strength to the Q-axis target current and the D-axis target current, respectively. The flux linkage torque corresponds to a difference between a target torque to be provided by the synchronous machine or required by the driver of an electric vehicle, and an iron loss torque that opposes the target torque. The assignments can be stored, in particular, as a lookup table. It should be noted that the flux strength is the quotient of the voltage and the electrical frequency.Consequently, a counter-torque generated due to iron losses is adequately taken into account when determining the Q-axis target current and the D-axis target current, thus enabling accurate and efficient control of the synchronous machine. According to one aspect of the present invention, the respective assignments for determining the Q-axis and D-axis setpoint currents can additionally link a property parameter relating to a variability of the magnetic and electrical properties of the synchronous machine to the Q-axis setpoint current or the D-axis setpoint current, respectively. This property parameter can preferably be a rotor temperature. Consequently, the Q-axis and D-axis setpoint currents can be determined even more precisely. According to a further aspect of the present invention, the iron loss torque can be determined based on the Q-axis and D-axis setpoint currents of a previous time step. Consequently, the iron loss torque can be easily taken into account by loading the previous Q-axis and D-axis setpoint currents from a memory. This enables simple and precise control of the synchronous machine. According to another aspect of the present invention, the iron loss torque can be determined from measured three-phase currents in the stator windings of the synchronous machine. These three-phase currents can be converted into a d / q coordinate system. Consequently, the iron losses can be taken into account when determining the Q-axis and D-axis setpoint currents based on actual current values, thus enabling precise and efficient control of the synchronous machine. According to the invention, the assignments for determining the Q-axis setpoint current and the D-axis setpoint current disregard voltage variations applied to the stator windings of the synchronous machine. These voltage variations primarily result from a voltage drop across a resistance in the stator windings. However, the voltage variation is also influenced by losses in a power electronics unit for controlling the synchronous machine and by voltage drops in the lines used. The method according to the invention therefore exploits the fact that the ideal Q-axis and D-axis currents are approximately linearly related to the flux strength. Furthermore, the assignments do not need to be extended. According to the invention, the voltage variation is regulated or compensated for beforehand by adjusting the target flux strength. Therefore, in the method according to the invention, the voltage variation can be regulated or compensated without expanding the assignment. According to the invention, the voltage variation is regulated or compensated based on a flux difference between a maximum flux, corresponding to the maximum voltage that can be applied by a power electronics unit, and a current flux, corresponding to the voltage currently applied to the stator windings. Consequently, the voltage variation can be adequately regulated. It should be noted that the maximum flux can be determined as the quotient of the maximum voltage that can be applied to the stator windings by the power electronics unit and the electrical frequency. The current flux can be determined as the quotient of the voltage currently applied to the stator windings and the electrical frequency. As an additional aspect, the flux strength difference can be input into an I-controller to determine a correction flux strength. Consequently, the correction flux strength can be determined in a simple manner. According to a further advantage, the sum of the maximum flow rate and the correction flow rate, or an efficiency-optimized flow rate if the sum exceeds the efficiency-optimized flow rate, can be used as the target flow rate. This ensures that the target flow rate is always limited to the efficiency-optimized flow rate. According to a preferred approach, the efficiency-optimized flux strength can be determined using the Maximum Torque per Loss (MTPL) method. The MTPL method takes into account the rotational speed of a rotor in the synchronous machine. Since the iron losses also change with the rotational speed, this method enables precise control of the synchronous machine. According to one aspect of the invention, the flux linkage torque can be normalized to a maximum torque, i.e., the maximum torque achievable at the target flux strength. The respective assignments for determining the Q-axis and D-axis target currents can then receive the normalized flux linkage torque as the flux linkage torque. In conventional lookup tables, the torque values ​​are stored as absolute values. That is, the currents for the individual flux strength values ​​are determined at predefined torque intervals, e.g., 10 Nm, and stored in the tables. However, in field weakening operation, i.e., at high speed, the maximum available torque can decrease significantly, rendering some of the lookup tables unusable and reducing the resolution in an effectively usable range. Furthermore, field weakening in the region of the synchronous machine's maximum torque can lead to significant interpolation errors that are difficult to correct.By normalizing the flux linkage torque to the maximum torque and determining the current values ​​at predetermined intervals of the proportion of the maximum torque, e.g. 0.05 or 0.1, the characteristic curves of the currents in the range of a maximum torque can be interpolated with sufficient accuracy, and a high resolution of the assignment can be ensured in the entire operating range of the synchronous machine. According to another aspect, the maximum moment can be determined based on the target flow rate and the property parameter. Therefore, the maximum moment can be determined in a simple way. According to a preferred approach, the respective mappings for determining the Q-axis and D-axis setpoint currents for motor operation of the synchronous machine can be parameterized, and generator operation of the synchronous machine can be accounted for by inverting the Q-axis setpoint current. Consequently, in contrast to parameterizing the mappings for motor and generator operation, the memory space required to store the respective mappings can be reduced by half. In contrast to conventional lookup tables, the required memory space can even be reduced to approximately one-sixth. A control unit according to the invention is designed and programmed to execute the vector control method according to one of the preceding aspects. For this purpose, the control unit can receive the required input variables either directly, e.g., from sensors, or indirectly from other control units and perform any necessary transformations, e.g., Clarke and Park transformations. The control unit can also output parameters to the power electronics unit for operating the power electronics unit that drives the synchronous machine. An embodiment of the invention is described in detail below with reference to the figures. Figure 1 shows a schematic view of a control unit; Figure 2 shows a diagram indicating current points for current tables for determining a Q-axis setpoint current and a D-axis setpoint current, wherein the diagram shows an MTPA line and several MTPL lines; Figure 3 shows a diagram indicating flux strength, achievable torque, and efficiency of the synchronous machine for different combinations of the Q-axis setpoint current and the D-axis setpoint current, wherein the diagram additionally shows an MTPA line and an MTPL line; Figure 4 shows a schematic view of a current determination unit according to a modification of the present invention.Figure 5 shows a diagram showing current points determined for flux linkage torque values ​​normalized to the maximum torque as a function of the target flux strength and the flux linkage torque, the diagram additionally showing an MTPL line, and Figure 6 shows a schematic view of a target flux strength adjustment section of the control unit. The figures are purely schematic and serve solely to illustrate the invention. The same elements are identified by the same reference symbols. Fig. 1 shows a schematic view of a control unit 1 for controlling the operation of a synchronous machine or a power electronics unit, which in turn controls the synchronous machine, according to an embodiment of the invention. The control unit 1 comprises functional sections for executing processes of a method according to the invention. The functional sections are implemented by executing software stored in the control unit 1 or by calling up data stored in the control unit 1. As shown in Fig. 1, the control unit 1 comprises a current determination section 2, a target flux adjustment section 3, and an iron loss torque determination section 4. The current determination section 2, in turn, contains current tables 5 as mappings for determining the Q-axis current Iq and the D-axis current Id. The current tables 5 have at least one target flux ΨTgt_TPL and one flux linkage torque MΨ as input parameters. That is, the current tables 5 each link the target flux ΨTgt_MTPL and the flux linkage torque MΨ with the Q-axis current Iq and the D-axis current Id, respectively. It should be noted that the flux linkage torque MΨ is the difference between a target torque MTgt to be provided by the synchronous machine and an iron loss torque MIron, which opposes the target torque MTgt. The assignments can be stored in a common table or in different tables.When using different tables, the tables may have different resolutions with respect to their input parameters. Particularly preferably, the current tables 5 can also include a property parameter relating to a variability of the electrical or magnetic properties of the synchronous machine or its components as an input parameter. In the present embodiment, the current tables 5 include a rotor temperature Trotor as a further input parameter. Consequently, the currents Iq and Id can be determined even more precisely. Current tables 5 only contain steady-state operating points of the synchronous machine, thus precluding consideration of terms relating to transient changes in the currents Iq and Id. Preferably, current tables 5 only consider the flux responsible for generating the stator field. Voltage variations, such as those caused by voltage drops across stator windings, losses in the control electronics unit, and voltage drops in the conductors, can be neglected in current tables 5. Using this approach, current tables 5 cover all voltage levels and rotational speeds. The diagram in Fig. 2 shows current points with their corresponding Iq and Id values ​​at a rotor temperature of 20 °C, as well as the resulting flux and achievable torque. The diagram in Fig. 2 therefore illustrates the values ​​stored in current tables 5. It should be noted that the flux strength corresponds to the quotient of the voltage applied to the stator windings and the electrical frequency. Voltage variation thus corresponds to a reduction in flux strength due to losses. Therefore, voltage variation can also be considered a flux strength variation. The electrical frequency can be calculated from the mechanical rotational speed of a synchronous machine's rotor and the number of poles. As already mentioned, the flux linkage torque MΨ is calculated as a difference between the target torque MTgt and the iron loss torque MIron before determining the currents Iq and Idals. For this, it is first necessary to determine the iron loss torque MIron. It is known that the iron losses consist of hysteresis losses and eddy current losses. The hysteresis losses increase linearly and the eddy current losses increase quadratically with the rotational speed in the iron loss power PIron (see equation (1)): Here, kHyst and kEddy are temperature-dependent factors of the hysteresis losses and eddy current losses, respectively. Consequently, the iron loss torque MIron, as shown in Fig. 1, depends not only on the currents Id and Iq and the rotational speed NMech, but also on the rotor temperature TRotorab. Since the iron loss torque MIron corresponds to the quotient of the iron loss power PIron and the rotational speed Nmech, the eddy current losses are linearly related to the rotational speed NMech in relation to the iron loss torque MIron (see equation (2)): The values ​​for the iron loss torque MIron can be calculated beforehand and stored as a mapping that links the current values ​​Iq and Id, the rotational speed NMech, and the rotor temperature TRotor as input parameters to the iron loss torque MIron. Alternatively, the iron loss torque MIron can be determined experimentally on an electric motor test bench, taking into account the measured torque at the same magnitude Q-current, once with a positive and once with a negative Q-current. The mapping can be a lookup table. Since the current values ​​for the target currents Id and Iq cannot be used to determine the iron loss torque MIron, the following two methods can be applied to determine the iron loss torque MIron. Firstly, the iron loss torque MIron, as shown in Fig. 1, can be determined using a Q-axis setpoint current Iq_t-1 and a D-axis setpoint current Id_t-1 from a previous time step. For this purpose, the two currents Iq_t-1 and Id_t-1 are stored in a memory of the control unit 1 after their determination and loaded from the memory as needed. Therefore, the iron loss torque MIron can be determined easily without having to input additional values ​​into the control unit 1. On the other hand, the iron loss torque MIron can be determined from measured three-phase currents in the stator windings of the synchronous machine. Since current values ​​are used, an accurate determination of the iron loss torque MIron can be performed. It should be noted that the measured three-phase currents can be transformed into a d / q coordinate system by the control unit 1 using Clarke and Park transformations before they can be used to determine the iron loss torque MIron. Fig. 3 shows the flux strength, achievable torque, and efficiency of the synchronous machine for different combinations of Iq and Id. Additionally, a Maximum Torque per Ampere (MTPA) curve and a Maximum Torque per Loss (MTPL) curve are shown. It can be seen that the maximum efficiency is not achieved at the MTPA curve, but at the MTPL curve, where the field in the synchronous machine is weaker. Furthermore, it was found to be advantageous if the flux linkage torque MΨ is not entered as an absolute value in the current tables 5, but is first normalized to a maximum torque Mmax_MTPL. This is because, with increasing rotational speed and consequently decreasing achievable flux strength (which results from the maximum voltage achievable in field weakening operation), the maximum adjustable torque decreases. Since the individual current points are determined at a predefined torque interval, e.g., 10 Nm, there are fewer current points in regions with lower achievable torques. Consequently, when using absolute torque values ​​for low flux strengths, fewer current points along the torque axis are available, meaning that only a reduced portion of the current tables 5 is actually usable. To overcome this disadvantage, a current determination section 2A according to a modification of the invention, as shown in Fig. 4, can have a maximum torque determination section 9 and a normalization section 10. The maximum torque determination section 9 can be designed as a lookup table and can determine a correspondingly achievable maximum torque Mmax_MTPL based on the target flux ΨTgt_MTPL and the rotor temperature TRotor. The flux linkage torque MΨ is then normalized to the determined maximum torque Mmax_MTPL and entered into current tables 5A, which accordingly receive the normalized target torque MRelativ as an input parameter. A diagram in Fig. 5 shows current points for different combinations of target flux ΨTgt_MTPL and flux linkage torque MΨ, where the current points were determined at predetermined intervals of the normalized target torque MRelativ, e.g., 0.05 or 0.1. It should be noted that the torque axis in the diagram indicates the actual torque value, and the maximum torque Mmax_MTPL corresponds to the uppermost current point. The diagram shows that normalizing the flux linkage torque MΨ to the maximum torque Mmax_MTPL significantly increases the density of current points for low flux levels, thus enabling more efficient control of the synchronous machine in this range. As already mentioned, the current tables 5 neglect voltage variations in the voltage applied to the stator windings. Therefore, an adjustment of the target flux ΨTgt_MTPA in the superimposed or upstream target flux adjustment section 3 is necessary to regulate or compensate for the voltage variation. Fig. 6 shows a schematic view of the target flux adjustment section 3, which in turn includes a section 6 for determining the efficiency-optimal flux and a voltage control section 7. The target flux adjustment section 3 receives the flux linkage torque MΨ, the rotor temperature TRotor, a maximum voltage UMax, a voltage currently applied to the stator windings Uctrl_req, the current electrical frequency ωel, and a mechanical rotational speed NMech of the synchronous machine's rotor. It should be noted that the maximum voltage UMax is limited by the power electronics unit that drives the synchronous machine, or by the DC supply voltage available at the power electronics. First, the voltage control section 7 calculates the corresponding flux strengths from the maximum voltage UMax and the currently applied voltage Uctrl_req, i.e., a maximum flux strength ΨMax and a current flux strength Ψctrl_req, by dividing by the electrical frequency ωel. From these values, a flux difference ΨErro, the difference between the maximum flux strength ΨMax and the current flux strength Ψctrl_req, is then calculated. The flux difference ΨErro is then fed into an I-controller 11 to determine a correction flux strength Ψcorrection. This correction flux strength is then added to the maximum flux strength ΨMax. Furthermore, Section 6 determines an efficiency-optimal flux strength ΨMax_MTPL for the flux linkage torque MΨ at rotor temperature TRotor and mechanical speed NMech using the MTPL method and inputs this into the voltage control section 7. Determining the flux linkage torque MΨ as the difference between the target torque MTgt and the iron loss torque MIron has already been described previously. Since the MTPL curve, like the iron losses, changes depending on the rotational speed, precise control of the synchronous machine can be achieved in this way. The comparison section 12 in the voltage control section 7 then compares the efficiency-optimized flux ΨMax_MTPL with the sum of the maximum flux ΨMax and the correction flux Ψcorrection and outputs the minimum of these two values. Consequently, the target flux ΨTgt_MTPL is limited to the efficiency-optimized flux ΨMax_MTPL. This means that a region located to the right of the MTPL line in the diagram from Fig. 5 is disregarded during the control of the synchronous machine. Finally, the adjusted target flux ΨTgt_MTPL is output to the current determination section 2, which then determines the Q-axis target current Iq and the D-axis target current Id using the current tables 5. Additionally, the voltage control section 7, as shown in Fig. 6, can include a further comparison section 13, which compares a maximum torque flux ΨAtMaxTorque, corresponding to the maximum applicable flux at the current operating point, with the sum of the maximum flux ΨMax and the correction flux Ψcorrection, and outputs the minimum of these two values ​​as the theoretical maximum setpoint flux ΨTgt_Max. Subsequently, a section 9 for determining the theoretical maximum torque (see Fig. 1) can output a theoretical maximum torque MMax_Ψ based on the theoretical maximum setpoint flux ΨTgt_Max and the rotor temperature TRotor. The theoretical maximum torque MMax_Ψ can then be added to the previously determined iron loss torque MIron to derive a theoretical maximum torque Mmax_Iron_corrected.The theoretical maximum torque Mmax_Iron_corrected is, for example, an important parameter with regard to the utilization of the synchronous machine. It has also proven particularly advantageous if the current tables 5 are created and stored in the control unit 1 solely for motor operation of the synchronous machine. In this case, generator operation of the synchronous machine is accounted for by inverting the Q-axis setpoint current. As shown in Fig. 6, this is achieved by ensuring that the current voltage applied to the stator windings, Ucrtl_req, has the opposite sign to the maximum voltage, UMax. This allows the current tables 5 to be smaller, thus saving memory space in the control unit 1. Reference symbol list 1 Control unit 2 Current determination section 2A Current determination section 3 Target flux adjustment section 4 Iron loss torque determination section 5 Current tables 5A Current tables 6 Section for determining the efficiency-optimal flux 7 Voltage regulation section 8 Section for determining the theoretical maximum torque 9 Maximum torque determination section 10 Normalization section 11 I-controller 12 Comparison section 13 Comparison section

Claims

Vector control method for a synchronous machine, in particular a permanent magnet synchronous machine, in which Q-axis and D-axis setpoint currents (Iq, Id) are determined by respective assignments (5, 5A) that define a flux linkage torque (MΨ) and a setpoint flux strength (ΨTgt_MTPL) with the Q-axis setpoint current and D-axis setpoint current, respectively.the D-axis target current (Iq, Id) are determined, wherein the flux linkage torque (MΨ) corresponds to a difference between a target torque (MTgt) and an iron loss torque (MIron), wherein the assignments (5, 5A) for determining the Q-axis target current (Iq) and the D-axis target current (Id) disregard a voltage variation of a voltage to be applied to stator windings of the synchronous machine and the voltage variation is previously regulated or compensated by adjusting the target flux strength (ΨTgt_MTPL), characterized in that the voltage variation is regulated or compensated on the basis of a flux strength difference (ΨError) between a maximum flux strength (ΨMax) and a current flux strength (Ψctrl_req). Method according to claim 1, wherein the assignments (5, 5A) for determining the Q-axis set current (Iq) and the D-axis set current (Id) additionally link a property parameter relating to a variability of magnetic and electrical properties of the synchronous machine, in particular a rotor temperature (TRotor), with the Q-axis set current (Iq) or the D-axis set current (Id). Method according to claim 1 or 2, wherein the iron loss moment (MIron) is determined based on Q-axis and D-axis target currents (Iq_t-1, Id_t-1) of a previous time step. Method according to claim 1 or 2, wherein the iron loss moment (MIron) is determined based on measured three-phase currents in stator windings of the synchronous machine. Method according to any one of claims 1 to 4, wherein the flux strength difference (ΨError) is input into an I-controller (11) to determine a correction flux strength (Ψcorrection). Method according to claim 5, wherein a sum of the maximum flow rate (ΨMax) and the correction flow rate (Ψcorrection) or an efficiency-optimal flow rate (ΨMax_MTPL) is used as the target flow rate (ΨTgt_MTPL) in the case that the sum exceeds the efficiency-optimal flow rate (ΨMax_MTPL). Method according to claim 6, wherein the efficiency-optimal flux strength (ΨMax_MTPL) is determined using the Maximum Torque per Loss (MTPL) method, which takes into account a rotational speed (NMech) of a rotor of the synchronous machine. Control unit that is trained and programmed to execute the method according to any one of claims 1 to 7.

Images