Method and device for calibrating a regulator of an electric machine

Abstract

A method (400) for calibrating a controller of an electric motor (120) to obtain a predeterminable torque value (T_Des), wherein the electric motor (120) is operated using field-oriented control. The method comprises the steps of: a.) predetermining a current vector (Ix_v) (410) for generating the predeterminable torque value (T_Des) by means of a connectable electric motor (120), b.) predetermining a test signal (Sx_Test) (420) and superimposing the current vector (Ix_v) on the test signal (Sx_Test), c.) detecting (430) a response signal (Sx_Antw) resulting from the superposition by means of a sensor (130), and e.) determining (450) a calibrated current vector (I_Vk) based on an evaluation of the response signal (Sx_Antw).

Description

Technical Field

[0001] The invention relates to a method and a device for calibrating a controller of an electric machine to obtain a predeterminable torque value. The invention also relates to an electric drive system with a corresponding device and a vehicle with an electric drive system as well as a computer program and a computer-readable storage medium. Background Art

[0002] In order to generate the desired torque by means of a suitable combination of direct torque and reluctance torque, an induction motor, in particular a permanent magnet synchronous motor with embedded magnets, is controlled. In the case of field-oriented control, the direct torque and the reluctance torque are set by means of corresponding selection of the d and q current operating points (id, iq) in the rotor-fixed coordinate system. In order to carry out current control in field-oriented control (FOC), in particular in the basic speed range, the associated current of the desired torque is determined from the position curve, the so-called MTPC (maximum torque per current). Under the assumption of an ideal machine with known inductances Ld and Lq, these MTPC position curves can be determined, for example, analytically. For real machines with saturation effects, another approach is preferably to numerically generate the MTPC position curve from simulation data (for example finite element simulation). Other correlations such as temperature dependence, internal motor losses, dispersion of material parameters and flow, in many cases, seem to be the most feasible approach to experimentally determine the MTPC position curve on a test bench, preferably on a gold prototype as an example. Here, sample dispersion, tolerances, etc. are ignored. For example, the torque characteristic diagram is scanned by id / iq, current and torque are measured, and the point of the shortest current vector for the desired torque is determined and appropriately stored. This process is usually complex and slow and therefore cost-intensive.

[0003] Therefore, there is a need for alternative methods and devices for calibrating a controller of an electric machine to obtain a predefinable torque value. Summary of the invention

[0004] A method is provided for calibrating a controller, preferably a current controller, of an electric machine to obtain a predeterminable torque value. The electric machine is operated using field-oriented control. The method comprises the following steps:

[0005] a.) A current vector is predefined by means of a connectable electric machine for generating the predefinable torque value. The current vector has a length and a direction as parameters.

[0006] b.) A test signal is specified and the current vector is superimposed on the test signal.

[0007] c.) detecting a response signal resulting from the superposition by means of a sensor, preferably detecting the amplitude of the resulting response signal.

[0008] d.) evaluating the response signal.

[0009] e.) Determining a calibrated current vector based on an evaluation of the response signal.

[0010] f.) Operating a controller of the electric machine by specifying the calibrated current vector in order to obtain the predeterminable torque value.

[0011] It is known to operate an electric motor with the aid of field-oriented regulation. In this process, the AC variables of the phase currents are respectively transmitted to a coordinate system that rotates at the frequency of the AC variables. In this rotating coordinate system, DC variables are then generated from the AC variables in the static operation of the electric motor, to which all common control technology methods can be applied. Due to the multi-phase, phase-shifted AC current applied to the stator, a rotating magnetic field is generated during the operation of the electric motor, which is composed of stator flux and rotor flux. The regulator of the electric motor predetermines the stator current according to a predeterminable torque value. In a rotating coordinate system, i.e., a d / q coordinate system, which rotates synchronously with the rotor flux and whose d-axis points in the direction of the rotor flux, the stator current is represented as a stator current phasor or stator current vector, which is characterized by its length and direction. This current phasor rotates synchronously with the rotating stator flux or rotor flux of the electric motor. In the d / q coordinate system, the current phasor can be represented by means of two mutually perpendicular components Id and Iq corresponding to its length and its direction, which are DC variables in the static state. In this coordinate system, machine-specific lines can be displayed along which the motor outputs a constant torque, the so-called isotorque lines. The controller of the motor can access the desired operating points on these isotorque lines by means of a family of characteristic curves, wherein these parameters may change during the operating time of the machine due to the above-mentioned dependence on each machine (example dispersion) and the rotor temperature, and therefore these parameters should be calibrated. For current vectors of the same length, there is only a completely specific direction in which the maximum torque is generated by the connected motor. In one step of the method, a current vector is predefined to generate a predefined torque value. In order to check whether this is the correct direction for generating the maximum torque here, a test signal is predefined and superimposed on the current vector. This superposition leads to oscillations of the output torque in the case of the connected motor. Such oscillations of the torque lead to mechanical and / or acoustic oscillations of the housing and the components connected thereto due to the mechanical coupling of the motor to the housing. The mechanical transmission behavior of the mechanical system leads to noise excitation of the connected motor and / or power electronics, which can be measured with suitable sensors, for example acoustically. The torque oscillations or the oscillations resulting from the torque oscillations are detected by means of sensors as a response signal from the superposition of the current vector and the test signal. The response signal is evaluated and a calibrated current vector, preferably the direction of the calibrated current vector, is determined based on the evaluation of the response signal. Thus, the direction parameter of the current vector for outputting a predeterminable torque value is calibrated and stored in a characteristic curve group. The regulator of the motor is then operated by means of the predetermining of the calibrated current vector to obtain a predetermined torque value.

[0012] A method for calibrating a regulator, preferably a current regulator, of an electric motor to obtain a predeterminable torque value is advantageously provided. Preferably, the method is provided for an integrated electric axle, which consists of an induction motor and a power electronics device mechanically attached (preferably mechanically strong or firmly coupled) or integrated to the motor. The method enables the method to be performed for each individual motor. The method can be performed at any time at the end of a production line and / or during the life of the electric drive, even during normal driving operation. The drive component (e.g., an electric axle) preferably includes a sensor and the method. The possibility of running the method at the start of operation, during an inspection at the end of the belt, or during driving operation is provided. Thereby, calibration parameters can be determined and readjusted or relearned in an example-specific, temperature-dependent and / or aging-dependent manner. A method for providing a self-learning / self-calibration regulation for an electric drive is created. A self-calibration of an MTPC position curve is achieved, which runs independently of a test bench and an external measurement sensor system.

[0013] In a further embodiment of the invention, the test signal has a length and a direction, wherein the direction is oriented orthogonally to the current vector, the test signal oscillating on both sides of the current vector and preferably being added vectorially to the current vector.

[0014] The test signal has a length and a direction as parameters corresponding to a predetermined current vector. The superposition is performed vectorially. The vector addition of the current vector and the test signal is obtained in a rotating coordinate system. The test signal is oriented orthogonally to the current vector and preferably oscillates at a predefined frequency. Due to the oscillation of the test signal, the connected motor generates different torques, in particular the torque of the harmonic oscillation component, depending on the current length and direction of the measurement signal. As a result, two different response signal types are obtained. The greater the deviation of the direction of the predetermined current vector from the maximum torque that can be achieved with its length, the greater the oscillation amplitude of the torque. The oscillation of the torque and the test signal have a common phase and frequency. The smaller the deviation of the direction of the predetermined current vector from the maximum torque that can be achieved with its length, the smaller the oscillation amplitude of the torque. When the direction of the current vector is very close to the direction belonging to the torque that can be achieved with this length, the torque oscillation contains a frequency component at twice the test signal frequency. The reason for the occurrence of this twice-frequency component is that due to the typical curvature of the isotorque line, the oscillation in the tangent direction of the isotorque line to the left and right of the output current vector leads to a minimum value of the torque, i.e. twice the frequency. Minimal oscillations of the torque or the disappearance of the torque pulsation or the appearance of a double frequency can be detected acoustically with a suitable harmonic frequency, i.e. when the corresponding frequency disappears or appears in the measured noise. A particularly favorable frequency of the test signal oscillation should be selected at which the connected machine has no or only a small intrinsic electromagnetic excitation, which for example for a 3-phase motor may preferably be the 5th or 7th order in the torque, since intrinsic excitation is expected here for the 6th and 12th order by way of example. Preferably, a favorable excitation order is selected for the method with respect to the set electrical frequency, so that in the absence of a predetermined test signal, no or almost no intrinsic electromagnetic order is generated in the torque oscillation for this order. As a result, the expected response signal will only be coupled to the method, while the disappeared response signal can be used as a target variable.

[0015] Furthermore, preferably when selecting the excitation frequency or order, preferably a frequency / order is selected which can be easily detected by the sensor, which has regard to sensor sensitivity and in particular to the transfer behavior from this frequency of the torque ripple to the sensor.

[0016] Advantageously, a possible test signal is provided which allows the direction of the predefined current vector to be evaluated with regard to the maximum achievable torque.

[0017] In another embodiment of the invention, steps a.) to d.) are repeated at least twice, wherein the direction of the current vector is changed by a predetermined absolute value. When evaluating the response signal according to step d.), the detected response signals are compared. In this case, the gradient or the minimum value of the detected response signal is determined.

[0018] The pre-determination of the current vector and the test signal is repeated at least twice with different directions of the current vector. The response signal is evaluated by comparing the response signal, preferably the amplitude of the response signal. Preferably, based on the size and change of the amplitude of the response signal, preferably based on the gradient and / or the minimum value, it is derived whether the pre-determined current vector generates the maximum torque in the connected motor, or in which direction the current vector must be changed during further repetitions in order to get closer to the direction in which the connected motor generates the maximum torque. The amplitude of the resulting torque oscillation increases with increasing length of the test signal or increasing absolute value of the test signal. Therefore, in the case of an excessively large amplitude, the length of the test signal should be reduced.

[0019] Advantageously, a method is provided for iteratively approaching a current vector direction that produces maximum torque for a connected electric machine.

[0020] In a further embodiment of the invention, the direction of the current vector changes in each repetition of the steps in the positive and negative direction of the last predefined current vector by a predefined absolute value or is predefined in each iteration step in the positive or negative direction of the last predefined current vector. The steps are preferably repeated until the gradient between the last three response signals is below a first predefined limit value or the response signal is below a second predefined limit value.

[0021] Advantageously, different variants are provided for iteratively approaching the current vector direction which generates the maximum torque of the connected electric machine.

[0022] In a further embodiment of the invention, according to step e.), the calibrated current vector is predefined in that parameters of the predefined current vector are predefined for the calibrated current vector, wherein the detected response signal of the predefined current vector is minimal.

[0023] If a current vector direction is reached that allows the connected electric machine to generate a maximum torque, a minimum response signal, preferably a minimum amplitude of the response signal, is obtained. The current parameters of the current vector, the length of the vector and the direction of the vector are predefined as parameters of the calibrated current vector.

[0024] Advantageously, a method is provided for determining a calibrated current vector.

[0025] Furthermore, the invention relates to a computer program comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the method described hereinabove.

[0026] Furthermore, the invention relates to a computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to perform the steps of the method described hereinabove.

[0027] The invention also relates to a device for calibrating a regulator of an electric machine. The device comprises a sensor, preferably a mechanical sensor. The device also comprises a circuit carrier, wherein the circuit carrier has a test signal generator and a computing unit. The device is configured to perform the steps of the method.

[0028] Advantageously, a device for calibrating a regulator of an electric machine is provided. The device comprises a preferably mechanical sensor for detecting a response signal resulting from a superposition of a current vector and a test signal. The device also comprises a test signal generator for predetermining the test signal and a calculation unit for executing the method.

[0029] In another embodiment of the invention, the sensor is mechanically fixedly or substantially rigidly connected to the electric motor. Alternatively, the sensor is fixedly mounted on the circuit carrier and the circuit carrier is fixedly integrated on or in the electric motor.

[0030] For high-resolution and undisturbed detection of the response signal, a mechanically fixed connection to the motor or via a circuit carrier attached to or in the motor is provided. Alternatively, it is of course also possible to use sensors or power electronics external to the device, such as microphones on or next to the motor, or by means of structure-borne sound sensors, for example in the form of acceleration sensors mounted on a surface, preferably the surface of the motor or a control unit or an inverter.

[0031] Advantageously, a location is provided for fixing the sensor for good signal transmission.

[0032] In a further embodiment of the invention, the mechanical sensor is a microphone, an acceleration sensor or a structure-borne sound sensor or a rotational speed sensor.

[0033] Advantageously, a sensor is provided for detecting a response signal resulting from the torque oscillations. The torque oscillations can be detected acoustically by means of acceleration measurement (preferably on a mechanically rigid unit with the electric machine) or by means of structure-borne sound. Likewise, a change in the speed of the electric machine results from the torque oscillations, so that the response signal can also be detected by means of a speed sensor.

[0034] The invention also relates to an electric drive system having an electric machine and the described device. Such an electric drive system is used, for example, to drive an electric vehicle. With the aid of the method and the device, a properly controlled operation of a drive train can be achieved.

[0035] Furthermore, the invention relates to a vehicle having the described drive system.Therefore, advantageously a vehicle is provided which comprises a device which can be used to calibrate a regulator of an electric machine.

[0036] It will be appreciated that the features, characteristics and advantages of the method according to the invention are correspondingly applicable or applicable to the device or the drive system and the vehicle, and vice versa.

[0037] Other features and advantages of the embodiments of the present invention are apparent from the following description with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The invention will be explained in more detail below based on some drawings, for which:

[0039] Figure 1 A schematic diagram showing an apparatus for calibrating a regulator of an electric motor,

[0040] Figure 2 shows a diagram of the dq current plane, in which isotorque lines are entered to apply field oriented regulation,

[0041] Figure 3 A schematically illustrated vehicle with a drive train is shown,

[0042] Figure 4 A schematically illustrated flow chart of a method for calibrating an offset angle of a field-oriented control of an electrical machine is shown. DETAILED DESCRIPTION

[0043] Figure 1 A device 100 is shown for calibrating a regulator 110 of an electric machine 120. The device comprises a sensor 130, preferably a mechanical sensor having a mechanically rigid or secure direct or indirect connection to the electric machine 120. The device also comprises a circuit carrier 150, wherein the circuit carrier has a test signal generator 160 and a calculation unit 170. The regulator 110 is preferably integrated in an inverter 140, wherein the inverter comprises power electronics 145, preferably a B6 bridge, for feeding power from a battery 155 to the connectable machine 120. In addition, in Figure 1 2 shows an electric drive system 200 having a device 100 and an electric machine 120 .

[0044] Figure 2A diagram of the dq current plane is shown, in which equal torque lines are entered for applying field-oriented regulation. In this rotating coordinate system, DC variables are generated from AC variables such as phase currents in the static operation of the motor. In the d / q coordinate system, which rotates synchronously with the rotor flux and whose d axis points in the direction of the rotor flux, the stator current is represented as a current vector Ix_v, which is characterized by its absolute value or its length I_s and its direction Ix_a. The current vector Ix_v rotates synchronously with the rotating stator flux or rotor flux of the motor. In this coordinate system, motor-specific lines T1, T2, T3, T_Des can be represented, along which the motor outputs a constant torque. The regulator of the motor can access the parameters of these lines by means of a family of characteristic curves or parameterizable data. Different operating points on these lines can be set by changing the direction Ix_a of the current vector and thus by means of different id and iq components. I1_v, I2_v and I3_v are used to represent three current vectors with the same length Is, whose directions Ix_a differ by a predeterminable absolute value Ix_a_Delta. Predeterminable test signals S1_Test, S2_Test, S3_Test are shown orthogonally to these current vectors. As can be seen from the diagram, the oscillating test signals S1_Test and S3_Test intersect more equal torque lines than the test signal S2_Test. Therefore, using the connected motor, a larger torque fluctuation is generated from the superposition of the current vector and the test signals S1_Test and S3_Test than in the case of the superposition of the current vector I2_v and the test signal S2_Test. Correspondingly, in this example, the parameter of the current vector I2_v, preferably the direction, is used for the calibrated current vector I_Vk.

[0045] Figure 3 A schematically illustrated vehicle 300 is shown with an electric drive system 200. The drive system 200 comprises an apparatus 100 for calibrating a regulator 110 of an electric machine 120 in an inverter 140 and an electric machine 210. The electric drive system preferably comprises a battery 150.

[0046] Figure 4 A schematic flow chart of a method 400 for calibrating a regulator of an electric motor 120 to obtain a predeterminable torque value T_Des is shown. The method starts with step 405. The electric motor 120 is operated using a field-oriented regulator. The method comprises the steps of:

[0047] a.) prescribing a current vector Ix_v for generating a predefinable torque value T_Des by means of a connectable electric machine 120 , 410 , wherein the current vector Ix_v has a length I_s and a direction Ix_a as parameters,

[0048] b.) predetermining a test signal Sx_Test, 420 and superimposing the current vector Ix_v on the test signal Sx_Test,

[0049] c.) detecting 430 a response signal Sx_Antw resulting from the superposition result by means of the sensor 130 ,

[0050] d.) evaluating 440 the response signal Sx_Antw,

[0051] e.) determining 450 a calibrated current vector I_Vk based on an evaluation of the response signal Sx_Antw,

[0052] f.) The controller of the electric machine 120 is operated 460 by means of a predefined calibrated current vector I_Vk to obtain a predefined torque value T_Des. Steps a.) to d.) 410 - 440 are preferably repeated at least twice to iteratively approach the direction of the current vector Ix_v that generates the maximum torque for the connected electric machine 120 . The method ends with step 470 .

Claims

1. A method (400) for calibrating a controller of an electric machine (120) to obtain a predeterminable torque value (T_Des), wherein the electric machine (120) is operated using field oriented regulation, The method comprises the steps of: a.) prescribing a current vector (Ix_v) (410) for generating the predeterminable torque value (T_Des) by means of a connectable electric machine (120), where the current vector (Ix_v) has length (I_s) and direction (Ix_a) as parameters, b.) predetermining a test signal (Sx_Test) ( 420 ) and superimposing the current vector (Ix_v) on the test signal (Sx_Test), c.) detecting (430) a response signal (Sx_Antw) obtained from the superposition result by means of a sensor (130), d.) evaluating (440) the response signal (Sx_Antw), e.) determining (450) a calibrated current vector (I_Vk) based on the evaluation of the response signal (Sx_Antw), f.) operating (460) a controller of the electric machine (120) by predefining the calibrated current vector (I_Vk) to obtain the predefinable torque value (T_Des), in, The test signal (Sx_Test) has a length (S_s) and a direction (Sx_a), wherein the direction (Sx_a) is oriented orthogonal to the current vector (Ix_V), and the test signal (Sx_Test) oscillates on both sides of the current vector (Ix_V).

2. The method according to claim 1, Steps a.) to d.) are repeated at least twice, The direction (Ix_a) of the current vector (Ix_V) is predefined as a change of a respectively predefined absolute value (Ix_a_Delta), wherein when evaluating the response signal (Sx_Antw) according to step d.), the detected response signal (Sx_Antw) is compared, and Determine the gradient or minimum of the detected response signal (Sx_Antw).

3. The method according to claim 2, in, The direction (Ix_a) of the current vector (Ix_V) is respectively predefined as a positive and negative direction towards the last predefined current vector (Ix_V). An absolute value (Ix_a_Delta) can be predefined or respectively predefined as a positive or negative direction towards the last predefined current vector (Ix_V).

4. The method according to claim 3, in, According to step e.), the calibrated current vector (I_Vk) is predefined by predefining parameters of the predefined current vector (Ix_V) for which the detected response signal (Sx_Antw) is minimized for the calibrated current vector (I_Vk).

5. Computer program product, comprising a computer program having instructions, which, when the computer program is executed by a computer, cause the computer to perform the steps of the method (400) according to any one of claims 1 to 4.

6. A computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to perform the steps of the method (400) of any one of claims 1 to 4.

7. A device (100) for calibrating a regulator (110) of an electric motor (120), Having a sensor (130), Having a circuit carrier (150), The circuit carrier comprises a test signal generator (160) and a computing unit (170), The device is configured to perform the steps of the method according to any one of claims 1 to 4.

8. The device according to claim 7, in, The sensor (130) is mechanically firmly connected to the electric motor (120), or the sensor (130) is firmly mounted on the circuit carrier (150) and the circuit carrier (150) is firmly integrated on or in the electric motor (120).

9. The device according to claim 8, in, The sensor (130) is a microphone, an acceleration sensor or a structure-borne sound sensor or a rotational speed sensor.

10. An electric drive system (200) comprising an electric machine (120) and a device (100) according to any one of claims 7 to 9.

11. A vehicle (300) comprising an electric drive system (200) according to claim 10.

Images