METHOD AND DEVICE FOR PROVIDING A COMMUTING INTERVAL

Description

[0001] The present invention relates to a method and a device for determining magnetization parameters of a three-phase electric motor. The invention thus lies particularly in the field of electric motors and the control of electric motors.

[0002] For the reliable operation of a three-phase electric motor, the detection of the rotor position is often advantageous, especially when the motor is stationary or running at low speeds. In particular, knowledge of the rotor position is advantageous for accurate and reliable commutation. Various different methods are known in the state of the art for this purpose, which can be divided into sensor-based and sensorless methods. Sensor-based methods have the disadvantage that they require suitable sensors and, accordingly, the hardware complexity and associated costs are higher than with sensorless methods. Sensorless methods typically rely on the injection of test pulses or measurement pulses into the electric motor, which can lead to undesirable noise. Sensorless methods are often based on the position and current dependence of the stator inductances of the electric motor.The various sensorless process types are described, for example, in the published patent application DE 10 2019 127 051 A1.

[0003] The application limitations of sensorless rotor position detection methods are still the subject of research and development and can often only be determined experimentally with considerable effort. This can also lead to misinterpretations that may only become apparent at a late stage of product development, leading to corresponding problems and / or requiring significant effort and expense to correct.

[0004] A simulation-based prediction of the properties of the electric motor or the application is, in principle, convenient, but traditionally not readily possible, as this requires precise knowledge of the electric motor's magnetization parameters. The magnetization parameters include the parameter k 1 , which characterizes an inductance variation caused by the rotor magnets, and the parameter k 2 , which characterizes an inductance variation caused by the electric motor's current supply.

[0005] Traditionally, the determination of magnetization parameters, especially k 2 , is only possible empirically. Simple setups for direct determination from measurements have so far failed, particularly due to the influence of parasitic resistances in the overall setup. The parameter k 1 can in principle be determined directly by measuring the inductances LD and LQ on a single motor winding; however, it is usually not possible to directly measure a single motor winding. This is because typically only the outer three terminals of the three-phase electric motor are accessible. Internally, the motor is then wired either in a star or delta connection. As a result, the effects on at least two of the three windings overlap, making direct calculation impossible.

[0006] Sensorless methods for determining the initial rotor position typically rely on measuring the voltage at the inductive voltage divider, i.e., at the terminal of the de-energized phase. However, these methods have the disadvantage that, in some cases, reliable north / south detection is not possible. This can be particularly the case with high-current motors.

[0007] NARAYANAN GEETHA ET AL, "Realistic Experience without Risk Using Virtual Electrical Machines Laboratory", 2016 INTERNATIONAL CONFERENCE ON LEARNING AND TEACHING IN COMPUTING AND ENGINEERING (LATICE), IEEE, March 31, 2016 (2016-03-31), pages 132-133, describes a virtual machine laboratory in which experiments can be conducted on virtual machines.

[0008] It is an object of the present invention to provide a method and a device by means of which the magnetization parameters k 1 and k 2 can be determined reliably and optionally automatically.

[0009] In one aspect, the invention therefore relates to a method for determining magnetization parameters of a three-phase electric motor having a rotor and a stator. The method comprises feeding current pulses into a first and second phase of the three-phase electric motor, wherein the feeding of the current pulses occurs over a measuring period, and rotating the rotor by at least part of an electrical revolution relative to the stator during the measuring period, wherein the rotation occurs uniformly over the entire measuring period. Furthermore, the method comprises measuring a voltage at an inductive voltage divider of the three-phase electric motor during the measuring period and determining measured values ​​of the measured voltage at predetermined angles of rotation of the rotor relative to the stator.The method further comprises calculating corresponding simulation values ​​of the voltage at the voltage divider at the predetermined rotation angles using predetermined estimated values ​​for the magnetization parameters and adjusting the predetermined estimated values ​​for the magnetization parameters such that a deviation of the simulation values ​​of the voltage from the measured values ​​of the measured voltage is minimized. Furthermore, the method comprises determining at least one or more of the magnetization parameters of the three-phase motor based on the estimated values ​​of the magnetization parameters adjusted to minimize the deviation of the simulation values ​​from the measured values ​​of the measured voltage.

[0010] In a further aspect, the invention relates to a device for determining magnetization parameters of a three-phase electric motor having a rotor and a stator. The device comprises a control unit for energizing the three-phase electric motor and for feeding current pulses into a first and second phase of the three-phase electric motor over a measuring period. The device further comprises a rotating unit for rotating the rotor relative to the stator, wherein the device is configured to rotate the rotor relative to the stator uniformly by one electrical revolution by means of the rotating unit during the entire measuring period, as well as a measuring element for measuring a voltage at an inductive voltage divider of the three-phase electric motor during the measuring period.The device also comprises a control unit configured to determine measured values ​​of the voltage measured by the measuring element at predetermined rotation angles of the rotor relative to the stator, as well as to calculate simulation values ​​of the voltage at the voltage divider at the predetermined rotation angles using predetermined estimated values ​​for the magnetization parameters, and to adapt the predetermined estimated values ​​for the magnetization parameters such that a deviation of the simulation values ​​of the voltage from the measured values ​​of the measured voltage is minimized. Furthermore, the control unit is configured to determine one or more of the magnetization parameters of the three-phase motor based on the estimated values ​​of the magnetization parameters adapted to minimize the deviation of the simulation values ​​from the measured values ​​of the measured voltage.

[0011] A three-phase electric motor is an electric motor with a stator and a rotor, which has three phases or three windings, each of which can be supplied with current via its own connection. The windings can be arranged in the electric motor such that they at least partially overlap each other. For example, the phases of the electric motor can be connected to each other in a delta connection or a star connection. The terms three-phase electric motor, electric motor, and motor are used synonymously in this disclosure text.

[0012] Current pulses can be injected, in particular, by applying voltage pulses to the respective phase terminals. The direction of the current flow can be influenced by the polarity of the applied voltage. However, due to the inductances of the windings or phases, the direction of the current flow does not always have to instantly follow the polarity of the voltage pulses. For example, a rapid reversal of the applied voltage pulse can reduce a previously existing current flow, but the direction of the current flow remains unchanged.

[0013] The inductive voltage divider can, in particular, correspond to the voltage between the terminal of the unenergized phase and ground. The voltage is determined by the ratio of the inductances L 1 and L 2 of the first and second energized phases, respectively, which in turn depends on the position of the rotor.

[0014] The fact that the voltage is measured at predetermined rotation angles means that the measured voltage values ​​are taken at least at specified rotation angles of the rotor. The predetermined rotation angles can represent relative angles with respect to the initial rotation angle position of the rotor. For example, specific measurement intervals can be spanned by the predetermined rotation angles. Alternatively or additionally, the predetermined rotation angles can be dependent on the measurement progress and correspond to angles at which the measured voltage has a minimum or maximum or another identifiable value.

[0015] Calculating simulated voltage values ​​in such a way that the deviation between the simulated voltage values ​​and the measured voltage values ​​is minimized means optimizing the simulation values ​​so that the simulated voltage values ​​correspond as closely as possible to the measured voltage values. In other words, the simulation values ​​are optimized by varying the parameters k 1 and k 2 .

[0016] The invention offers the advantage that the magnetization parameters k 1 and k 2 can be reliably determined. For example, the magnetization parameters can be determined ex works for each motor type and / or for each individual motor, and the parameters determined in this way can be used to reliably determine the rotor position during motor operation. The invention therefore offers the advantage that magnetization parameters determined according to the invention enable methods for sensorless determination of the rotor position that would not be applicable without magnetization parameters provided by a reliable determination. The invention therefore offers the advantage that even for motors that require reliable knowledge of the rotor angle, there is no need to resort to sensor-based determinations of the rotor angle, which involve additional costs.

[0017] Furthermore, the invention offers the advantage that the determination of the magnetization parameters can be partially or fully automated. This enables routine determination of the magnetization parameters of numerous electric motors. For example, the magnetization parameters of a large number of motors can be determined in this way, and the magnetization parameters calculated by averaging can be used for a specific series of electric motors.

[0018] Optionally, measuring a voltage at the inductive voltage divider includes measuring a voltage difference between a first and a second voltage value. The first voltage value is generated by feeding in a current pulse using a voltage pulse of a first polarity, and the second voltage value is generated by feeding in a current pulse using a voltage pulse of a second polarity opposite to the first polarity. This offers the advantage that the reliability of the determination can be increased because interference can be reduced by using the voltage difference. In particular, interference that equally affects the first and second voltage values ​​or the resulting current pulses can be averaged out, so that it has no influence on the voltage difference used to determine the magnetization parameters.In particular, using the voltage difference can reduce a motor voltage caused by the rotation of the rotor relative to the stator, which would otherwise superimpose the measured voltage.

[0019] Optionally, the current pulses are designed as test pulses, with the method being performed at least once with the three-phase electric motor de-energized except for the test pulses. Determining the one or more magnetization parameters comprises determining a magnetization parameter k 1 that characterizes an inductance variation by one or more rotor magnets of the three-phase electric motor. By performing the method with the motor de-energized, the effects of the rotor magnet on the inductances of the electric motor, which are characterized by the parameter k 1, can be determined.

[0020] Optionally, the method is performed at least once with a three-phase electric motor energized in addition to the test pulses, wherein determining the one or more magnetization parameters comprises determining a magnetization parameter k 2 that characterizes an inductance variation due to energizing the three-phase electric motor. This offers the advantage that the influence of the inductances resulting from the energizing of the motor can also be determined reliably and easily. Optionally, the three-phase electric motor is energized in such a way that commutation of the energizing is maintained unchanged over the measurement period. In other words, the angle of rotation of the rotor is changed while the commutation of the energized motor remains unchanged.

[0021] In this way, the effects of the rotor rotation angle on the inductance changes caused by the current supply can be reliably determined.

[0022] Optionally, the method is performed at least once with the three-phase electric motor energized, with the current pulses being designed as block commutation pulses. This offers the advantage that no additional measurement pulses need to be provided for the method performed with the motor energized; instead, the regular commutation pulses can be used. This can reduce the effort and / or complexity of the measurement.

[0023] Optionally, the procedure is performed at least once with the three-phase electric motor de-energized and at least once with the motor energized. This offers the advantage that both magnetization parameters k 1 and k 2 can be reliably determined.

[0024] Optionally, the method is carried out iteratively, alternating between the de-energized and energized three-phase electric motors, with at least some of the magnetization parameters determined during one execution of the method being used as predetermined estimated values ​​for the magnetization parameters in the subsequent iterative execution of the method. This offers the advantage that the mutual dependence of the magnetization parameters k 1 and k 2 can also be taken into account, thus allowing a gradual improvement in the accuracy of the determination of both magnetization parameters. For example, a progressive refinement of the magnetization parameter k 1 during the iterative execution of the method can be taken into account in the further refinement of the magnetization parameter k 2, and vice versa. In this way, a reliable determination of both magnetization parameters can be achieved.

[0025] Optionally, such angles of rotation are predetermined as predetermined angles of rotation at which the respectively measured voltage has a local maximum and / or a local minimum. This offers the advantage that the magnetization parameters can be determined particularly reliably based on the measured values ​​of the measured voltages at the maxima and / or minima. This also offers the advantage that the magnetization parameters can be determined based on these predetermined angles of rotation with a small number of measured values ​​or angles of rotation. Optionally, at least one part of the electrical revolution by which the rotor is rotated relative to the stator has at least 90° and at least one predetermined angle of rotation at which the measured voltage has a local minimum and at least one further predetermined angle of rotation at which the measured voltage has a local maximum.This offers the possibility of reducing the angle of rotation and thus shortening the effort and time required for the measurement.

[0026] Optionally, the predetermined angles of rotation are predetermined in such a way that, based on the measured voltage at the predetermined angles of rotation, a continuous voltage profile can be determined over the entire electrical revolution. This allows for easy identification of those angles of rotation at which the measured voltage exhibits a local minimum or a local maximum. This may also be advantageous in that it facilitates the adaptation of the simulation values ​​to the measured values.

[0027] Optionally, the predetermined estimated values ​​for the magnetization parameters are incorporated into the calculation of the simulation values ​​via a predetermined mathematical model for the respective inductance of the first and second phases of the three-phase electric motor. In particular, the predetermined mathematical model can offer the possibility of automating the calculation of the simulation values, for example, by storing the predetermined mathematical model on a computing unit.

[0028] Optionally, the method further comprises pre-positioning the rotor relative to the stator to a predetermined starting angle of rotation before feeding in the current pulses. This can be particularly advantageous in that a rotation angle range over which the rotor is rotated during the measurement encompasses a desired range and / or starts in a desired range. For example, the rotation angle range can be optimized to maximize the probability of a local maximum and / or minimum occurring in the predetermined rotation angle range. Furthermore, this can offer the advantage of achieving and / or increasing the comparability of several different measurements. Optionally, the pre-positioning is carried out by means of mechanical forced positioning and / or by applying current to the three-phase electric motor until the starting angle of rotation is reached. This offers the possibility of automating the pre-positioning.

[0029] The features and embodiments mentioned above and explained below are not only to be regarded as disclosed in the respective explicitly mentioned combinations, but are also encompassed by the disclosure content in other technically meaningful combinations and embodiments. In particular, the individual aspects of the present disclosure can be combined with one another. Thus, optionally, a method for providing a commutation interval can include the use of magnetization parameters determined according to a method according to a further aspect of the present disclosure. Alternatively or additionally, a method for determining an initial rotor position of a three-phase electric motor according to one aspect of the present disclosure can include the use of magnetization parameters determined according to a method according to a further aspect of the present disclosure.A method for providing a commutation interval according to one aspect of the present disclosure and a method for determining the initial rotor position of the three-phase electric motor according to one aspect of the present disclosure can also be implemented in a control system of a three-phase electric motor, which optionally use magnetization parameters of the three-phase electric motor determined according to a method according to another aspect of the present disclosure. In other words, the individual aspects of the invention can be combined with one another or used independently of one another.

[0030] Further details and advantages of the invention will now be explained in more detail with reference to the following examples and preferred embodiments with reference to the figures.

[0031] Short description of the characters: Fig. 1A and 1Bshows schematic representations of a three-phase electric motor in star connection ( Figure 1A ) and in delta connection ( Figure 1B ); Figure 2 shows an example of a time course of the voltages U 1 , U 2 and U 3 as well as the current course through the second phase 2212.2; Figures 3 to 6 shows graphs of the measured and simulated voltage U 3 versus the angle of rotation of the rotor relative to the stator.

[0032] In the following figures, identical or similar elements in the various embodiments are designated by identical reference numerals for the sake of simplicity.

[0033] The Figures 1A and 1B show, by way of example, in schematic representations, a three-phase electric motor 210 in star connection ( Figure 1A ) and in delta connection ( Figure 1B). The electric motor 210 has three connections 212.1, 212.2 and 212.3 for the three different phases 212. Each of the three phases 212.1, 212.2 and 212.3 is characterized, for example, by an associated inductance L 1 , L 2 or L 3 and an associated ohmic resistance R 1 , R 2 or R 3 . The supply voltage of the electric motor 210 is designated as US and corresponds to a potential difference with respect to a ground potential. Different voltages U 1 , U 2 or U 3 can be applied to the three connections 212.1, 212.2 and 212.3 of the three phases, which also represent a potential difference with respect to the ground potential.

[0034] Each of the three phases is connected at one end to a corresponding terminal 212.1, 212.2 and 212.3. In the case of the star connection ( Figure 1A ) the other end is connected to a star point 214 of the star connection. In the case of the delta connection ( Figure 1B) the other end is connected to terminal 212.1, 212.2 or 212.3 of the next phase.

[0035] In the following, a method for determining magnetization parameters of the electric motor 210 according to an optional embodiment is explained with reference to the figures, without, however, limiting the method to this optional embodiment. To illustrate the explanation, phases 212.1 and 212.2 of the electric motor 210 represent the first and second phases of the electric motor 210, into which current pulses are fed according to the method, while the third, non-energized phase is formed by phase 212.3. However, the assignment or sequence could also be chosen differently as desired.

[0036] In Figure 2is an exemplary temporal progression of the voltages U 1 , U 2 and U 3 as well as the current progression i L2 through the second phase 212.2 versus time t. As can be seen in the two upper graphs, which represent U 1 and U 2, the first and the second phase are alternately provided with a voltage pulse, so that correspondingly in the second phase 212.2 alternating rising and falling current pulses are fed in (i L2 , bottom graph in Figure 2 ). Furthermore, Figure 2The third graph also shows the time course of the voltage in the third phase or at the third phase connection 212.3, which assumes alternating values ​​that are above or below half of the supply voltage US, depending on whether the voltage is applied to the first or second phase 212.1 or 212.2. Furthermore, in the respective alternating intervals of the voltage U3, a voltage increase can be seen, the amplitude of which is small compared to the total voltage U3, which is caused by the current flow changing over time.

[0037] The current flow i L2 through the second phase 212.2 follows a sawtooth-like curve, the gradient of which changes sign with each change of the current pulses. The time intervals for which a voltage pulse is applied and the corresponding gradient of the current i L2 is maintained are shown in the graphs in Figure 2referred to as t 1.

[0038] When measuring the voltage U 3 on the third, de-energized phase, a voltage difference of the voltage at different points in time can also be evaluated, in which the difference between the voltage U 3 at a point in time at which the voltage pulse is fed in the first phase and the voltage U 3 at a second point in time at which the voltage pulse is fed in the second phase 212.2 is formed. Thus, the voltage U 3 or also the mentioned voltage difference can be used to carry out the method. Within the scope of this disclosure, explanations regarding the voltage U 3 also include the corresponding explanation using the voltage difference.

[0039] To determine the magnetization parameters of the three-phase electric motor, the voltage or the voltage difference at the third phase connection 212.3, which represents the tap of an inductive voltage divider, is measured over a measurement period. A plurality of measurements are taken at regular or irregular intervals during the measurement period. During the measurement period, the rotor is also rotated uniformly by at least part of an electrical revolution relative to the stator. The voltage is measured during the measurement period in such a way that the voltage at the inductive voltage divider is measured at least when the rotor assumes one of several predetermined angles of rotation relative to the stator.

[0040] The rotation is performed by at least a portion of a full electrical revolution such that at least one maximum and at least one minimum of the measured voltage values ​​U3 are measured at the inductive voltage divider within the covered rotation angle range. Optionally, the measurements can be performed at much shorter time intervals or rotation angle intervals, allowing a curve of the voltage U3 as a function of the rotation angle to be reconstructed, although this is not mandatory.

[0041] To determine the magnetization parameter k 1 , a measurement of the voltage U 3 at the inductive voltage divider is carried out as a function of the angle of rotation with the motor de-energized, i.e. when the motor is not energized apart from the voltage pulses for generating the current pulses for determining the magnetization parameters. To determine the magnetization parameter k 2 , a corresponding measurement is carried out with the motor 210 energized, wherein in addition to the current pulses, a commutated current is also supplied to the motor 210, wherein the commutation is maintained unchanged during the measurement period and is not adapted to the changed angle of rotation. Both measurements can be carried out several times, in particular alternately, in order to determine both magnetization parameters, for example within the framework of an iterative process.

[0042] The following is an explanation of the determination of the magnetization parameter k 1 . For this purpose, current pulses are periodically fed into the first and second phases 212.1 and 212.2 respectively by a control of the electric motor 210 over a measuring period, while the rotor is either rotated slowly and evenly relative to the stator over the measuring period or is rotated further incrementally between successive measurements. The rotation can be carried out manually by a user or automatically. In addition, at the same time as the current pulses are fed in, the voltage values ​​of the voltage U 3 are measured at the inductive voltage divider and the measured values ​​are saved. It may be sufficient to only save the voltage values ​​of the voltage U 3 at those angles of rotation at which the voltage U 3 has a local minimum or maximum, which are then assigned to the respective angle values.Alternatively, the voltage values ​​can be recorded more closely so that the course of the voltage U 3 over the angle of rotation range can be recognized and / or reconstructed.

[0043] Figure 3 shows in graph 2100 as an example the measured course of the voltage U 3 at the inductive voltage divider, ie at the connection of the third phase 212.3, over the course of the angle of rotation ϕwith a de-energized motor, i.e. when the motor is not energized to determine the magnetization parameter k 1 apart from the voltage pulses for generating the current pulses for determining the magnetization parameters. The horizontal axis shows the angle of rotation of the rotor relative to the stator in degrees from 0° to 360° and the vertical axis shows the measured voltage difference between the measured voltage values ​​for measuring pulses with different polarity in volts. Graph 2100 shows that the measured voltage U 3 or voltage difference has a sinusoidal curve with a periodicity of approximately 180° and an amplitude of approximately 1.3 V, with the sine curve oscillating around the zero line.

[0044] The method further comprises calculating corresponding simulation values ​​of the voltage at the voltage divider at the predetermined rotation angles using predetermined estimated values ​​for the magnetization parameters and adjusting the predetermined estimated values ​​for the magnetization parameters such that a deviation of the simulation values ​​of the voltage from the measured values ​​of the measured voltage U 3 is minimized. The simulation values ​​are calculated using a predetermined mathematical model. An exemplary mathematical model is explained below.

[0045] The mathematical model is described for the example case mentioned above, in which the first and second phases 212.1 and 212.2 are supplied with current pulses, while the voltage U3 or voltage difference is measured on the third, de-energized phase 212.3. The mathematical model is based on the following system of differential equations. di L 2 dt − U S − i L 2 R 1 + R 2 L 1 + L 2 U 3 t = i L 2 ⋅ R 2 + L 2 ⋅ di L 2 dt where L 1 = L max ⋅ 1 − k 1 1 + cos 2 φ + k 2 i L 1 cos φ + cos 2 φ L 2 = L max ⋅ 1 − k 1 1 + cos 2 φ − 120 ° + k 2 i L 2 cos φ − 120 ° + cos 2 φ − 120 ° i L 1 = − i L 2

[0046] Here, L 1 and L 2 indicate the inductance of the first and second phases 212.1 and 212.2 of the three-phase electric motor, US the supply voltage, R 1 and R 2 the ohmic resistance of the first and second phases 212.1 and 212.2, t the time, and U 3 the voltage or voltage difference measured at the connection of the third phase 212.3. The maximum inductance L max is assumed to be the same for all phases of the electric motor and can be determined, for example, by measuring the inductance for a single phase. If this is not readily possible due to a spatial overlap of two of the phases, the maximum inductances of two phases 212.1 and 212.2 can be measured together, and the value L max of an individual inductance can be determined by scaling down accordingly. The angle of rotation ϕis the angle of rotation of the rotor relative to the stator. Parameter k 1 corresponds to the current-independent magnetization parameter, which characterizes the inductance variation caused by the rotor magnets, and k 2 corresponds to the current-dependent magnetization parameter, which characterizes the inductance variation due to the current supply to the respective phase.

[0047] Using equations (10) to (14) presented above, simulation values ​​for the voltage U 3 can then be calculated. These simulation values ​​correspond to the expected measured voltage values ​​at the predetermined rotation angles of the rotor during the measurement period. The voltage values ​​are therefore calculated in particular for those rotation angles at which the voltage U 3 was also measured. These calculated voltage values ​​can then be compared with the measured voltage values. Based on this, the magnetization parameter k 1 can then be varied and optimized using a regression process, and the influence on the calculated simulation values ​​can be checked. For the initial simulation and optimization of k 1 with the motor de-energized, a predetermined value of k 2 can initially be assumed. If k 2 has already been determined and / or optimized in a previous measurement, this value can optionally be used. Alternatively, for example, the simulation orOptimization first with . k 2 = 0 can be started and / or carried out.

[0048] In this way, the magnetization parameter k 1 can be optimized with the goal of minimizing the deviation of the calculated simulation voltage values ​​from the corresponding measured voltage values. The value of the magnetization parameter k 1 determined in this way, for which the deviation is smallest, can then be assigned as the actual magnetization parameter k 1 to the measured three-phase electric motor and / or used to determine the parameter k 2.

[0049] Figure 4 shows an example of the measured voltage U 3 at the inductive voltage divider from Figure 3 (Graph 2100) compared to the simulated and with regard to the magnetization parameter k 1 optimized curve of the voltage U 3 over the angle of rotation ϕ(Graph 2200). It can be seen that graphs 2100 and 2200 show only a slight deviation, and thus, by varying k 1, a very precise fit of the simulation to the measured values ​​can be achieved.

[0050] Corresponding process steps can then be performed to determine the magnetization parameter k 2 . Unlike the method described above for determining the magnetization parameter k 1 , the process steps are performed with the motor energized. The motor is energized with a defined, predefined current and a predefined, unchanged commutation.

[0051] Figure 5 shows in graph 2300 the measured voltage U 3 (vertical axis, in volts) at the inductive voltage divider as a function of the angle of rotation ϕ(in degrees). Graph 2300 corresponds to the curve of the voltage U 3 , which is caused by periodically repeated current pulses fed into the first and second phases 212.1 and 212.2, while the rotor is rotated slowly and evenly relative to the stator over the measuring period. The measured voltage values ​​U 3 are then stored in a memory, whereby the memory can be part of the control of the electric motor or can be designed separately from it. In graph 2300 it can also be observed that the voltage U 3 has local minima and maxima over one electrical revolution, which can be used to determine the magnetization parameter k 2.

[0052] When determining the magnetization parameter k 2, the corresponding simulation values ​​are calculated and optimized and compared with the measured values, which are shown in Figure 6where the optimization is carried out by varying the magnetization parameter k 2. The graph 2300 corresponds to the one already shown in Figure 5 shown graph of the measured voltage and graph 2400 the simulated and optimized curve of the voltage U 3 . When optimizing the magnetization parameter k 2 , a predetermined value for k 1 can be assumed. If the magnetization parameter k 1 has already been determined and / or optimized in a previous measurement, this value can optionally also be used when determining the magnetization parameter k 2 . By repeatedly iterating the magnetization parameters k 1 and k 2 , where the results of the previous iteration step are used in each iteration step, the accuracy of the determination of the magnetization parameters k 1 and k 2 can be improved. As in Figure 6As can be seen, a very good agreement between the simulated and measured voltage values ​​U 3 can be achieved for the determination of the magnetization parameter k 2. The first initial values ​​can, for example, be k 1 = k 2 = 0 can be selected and after each iteration step, the last determined optimized value can then be used for the other parameter. List of reference symbols

[0053] 210Three-phase electric motor 212Phase of the three-phase electric motor 212.1, 212.2, 212.3First, second or third phase of the three-phase electric motor 214Star point U 1 , U 2 , U 3Voltage at the connection of the first, second or third phase L 1 , L 2 , L 3Inductance of the first, second or third phase R 1 , R 2 , R 3Ohmic resistance of the first, second or third phase i L1 , i L2 , i L3Current flow in the first, second or third phase USSupply voltage 2100Graph of the measured voltage difference in the de-energized case 2200Graph of the simulated voltage difference in the de-energized case 2300Graph of the measured voltage difference in the energized case 2400Graph of the simulated voltage difference in the energized case

Claims

1. Method for determining magnetization parameters of a three-phase electric motor (210) having a rotor and a stator, the method comprising and being characterized by: - feeding current pulses into a first and second phase (2212.1, 2212.2) of the three-phase electric motor (210), wherein the feeding-in of the current pulses takes place over a measurement period; - rotating the rotor by at least one part of an electrical revolution relative to the stator during the measurement period, wherein the rotation takes place evenly over the total measurement period; - measuring a voltage (U3) across an inductive voltage divider of the three-phase electric motor (210) during the measurement period and ascertaining measured values (2300) of the measured voltage at predetermined angles of rotation (ϕ) of the rotor relative to the stator; - calculating corresponding simulation values (2400) of the voltage across the voltage divider at the predetermined angles of rotation (ϕ) by using predetermined estimated values for the magnetization parameters and adjusting the predetermined estimated values for the magnetization parameters in such a way that a deviation of the simulation values (2400) of the voltage from the measured values (2300) of the measured voltage (U3) is minimized; - determining at least one or more of the magnetization parameters of the three-phase motor (210) with the aid of the estimated values of the magnetization parameters which were adjusted for the minimizing of the deviation of the simulation values from the measured values of the measured voltage (U3).

2. Method according to Claim 1, wherein the measurement of a voltage (U3) across the inductive voltage divider comprises a measurement of a voltage difference between a first and a second voltage value, wherein the first voltage value is produced by the feeding-in of a current pulse by means of a voltage pulse of a first polarity and the second voltage value is produced by the feeding-in of a current pulse by means of a voltage pulse of a second polarity opposite to the first polarity.

3. Method according to Claim 1 or 2, wherein the current pulses are formed as test pulses and the method is performed at least once while the three-phase electric motor (210) is unenergized except for the test pulses, and wherein the determination of the one or more magnetization parameters comprises determining a magnetization parameter k1 which characterizes an inductance variation by one or more rotor magnets of the three-phase electric motor (210).

4. Method according to Claim 3, wherein the method is performed at least once while the three-phase electric motor (210) is energized in addition to the test pulses and wherein the determination of the one or more magnetization parameters comprises determining a magnetization parameter k2 which characterizes an inductance variation by energization of the three-phase electric motor (210).

5. Method according to Claim 4, wherein the energization of the three-phase electric motor (210) takes place in such a way that commutation of the energization is maintained unaltered over the measurement period.

6. Method according to one of Claims 1 to 5, wherein the method is performed at least once while the three-phase electric motor (210) is energized and the current pulses are formed as block commutation pulses.

7. Method according to one of Claims 1 to 6, wherein the method is performed at least once while the three-phase electric motor (210) is unenergized and at least once while the three-phase electric motor (210) is energized.

8. Method according to one of Claims 1 to 7, wherein the method is performed iteratively while the three-phase electric motor (210) alternately is unenergized and energized, wherein at least some of the magnetization parameters determined while the method was being performed are used as predetermined estimated values for the magnetization parameters in the subsequent iterative performance of the method.

9. Method according to one of Claims 135 to 8, wherein such angles of rotation (ϕ) are predetermined as the predetermined angles of rotation (ϕ) at which the respective measured voltage has a local maximum and / or a local minimum.

10. Method according to Claim 9, wherein the at least one part of the electrical revolution by which the rotor is rotated relative to the stator has at least 90° and has at least one predetermined angle of rotation (ϕ) at which the measured voltage (U3) has a local minimum and has at least one further predetermined angle of rotation (ϕ) at which the measured voltage (U3) has a local minimum has local maximum.

11. Method according to one of Claims 1 to 10, wherein the predetermined angles of rotation are predetermined in such a way that a continuous profile of the voltage over the electrical revolution can be determined with the aid of the measured voltage at the predetermined angles of rotation.

12. Method according to one of Claims 1 to 11, wherein the predetermined estimated values for the magnetization parameters are included in the calculation of the simulation values via a predetermined mathematical model for the respective inductance of the first and second phase of the three-phase electric motor.

13. Method according to one of Claims 1 to 12, further comprising a prepositioning of the rotor relative to the stator at a predetermined starting angle of rotation before the current pulses are fed in.

14. Method according to Claim 13, wherein the prepositioning takes place by means of a mechanical forced positioning and / or by injecting current into the three-phase electric motor until the starting angle of rotation is reached.

15. Device for determining magnetization parameters of a three-phase electric motor (210) having a rotor and a stator, the device comprising and being characterized by: - an activation unit for energizing the three-phase electric motor (210) and for feeding current pulses into a first and second phase (2212.1, 2212.2) of the three-phase electric motor over a measurement period; - a rotary unit for rotating the rotor relative to the stator, wherein the device is designed to rotate the rotor relative to the stator, by means of the rotary unit, evenly by one electrical revolution during the total measurement period; - a measurement element for measuring a voltage across an inductive voltage divider of the three-phase electric motor during the measurement period; and - a control unit which is designed to: +ascertain measured values (2300) of the voltage (U3), which is measured by means of the measurement element, at predetermined angles of rotation of the rotor relative to the stator; +calculate simulation values (2400) of the voltage across the voltage divider at the predetermined angles of rotation by using predetermined estimated values for the magnetization parameters and to adjust the predetermined estimated values for the magnetization parameters in such a way that a deviation of the simulation values (2400) of the voltage from the measured values (2300) of the measured voltage is minimized; +determine one or more of the magnetization parameters of the three-phase motor with the aid of the estimated values of the magnetization parameters which were adjusted for the minimizing of the deviation of the simulation values from the measured values of the measured voltage.