Method for controlling an electric machine, control device for an electric machine and electric drive system

Abstract

Method (100) for controlling an electrical machine (4), comprising the steps: Acquire (110) at least one manipulated or measured variable (P); Determining (120) proportions of a predetermined frequency in the detected manipulated or measured variable (P); Calculating (130) a control variable (R) using the determined components of the predetermined frequency of the detected manipulated or measured variable (P) and a predetermined control matrix; and Controlling (140) the electric machine (4) with the calculated control variable (R), wherein the method is characterized by the following steps: Providing several predetermined control matrices and selecting one control matrix from the several predetermined control matrices to calculate the controlled variable (R), wherein the control matrix is ​​selected depending on a given operating mode for the electrical machine (4).

Description

[0001] The present invention relates to a control device for an electric machine, a method for controlling an electric machine and an electric drive system. State of the art

[0002] German patent applications DE 10 2009 000 930 A1 and DE 10 2016 220 736 A1 disclose a method and an arrangement for reducing torque ripple in a permanent magnet motor system. The motor system comprises a permanent magnet motor coupled to an inverter. The method described in these documents includes a step for modifying operating control signals to generate ripple-reducing operating control signals. These modified operating control signals are supplied to an inverter for controlling the permanent magnet motor.

[0003] Rotating field machines, such as squirrel-cage induction motors or permanent magnet synchronous motors, do not possess ideal sinusoidal flux distributions in the air gap due to their design. During operation, this leads to uneven torques containing harmonics when controlled with sinusoidal currents. Two fundamental approaches are possible for minimizing these harmonics. One approach involves the targeted, controlled application of a superimposed exogenous compensation voltage to the motor voltage. Alternative approaches are based on an additional control loop. For example, in a cascade structure, a target torque can first be specified in an inner control loop, which is then superimposed from the outside using a compensation method for higher harmonic oscillations. Disclosure of the invention

[0004] The present invention discloses a method for controlling an electric machine according to claim 1, a control device for an electric machine according to claim 7 and an electric drive system according to claim 10. Accordingly, the following is planned:

[0005] A method for controlling an electric machine. The method comprises the steps of acquiring at least one manipulated or measured variable and determining components of a predetermined frequency in the acquired at least one manipulated or measured variable. Furthermore, the method comprises a step for calculating a control variable using a predetermined control matrix and the determined components of the predetermined frequency of the acquired manipulated or measured variable. Finally, the method may include a step for controlling the electric machine with the calculated control variable. According to the invention, the method includes a step for Providing several predetermined control matrices and selecting a control matrix from the several predetermined control matrices to calculate the controlled variable, wherein the control matrix is ​​selected depending on a predetermined operating mode for the electrical machine. Furthermore, the following is planned:

[0006] A control device for an electric machine comprising a first transformation unit and a first computing unit. The first transformation unit is designed to determine components of a predetermined frequency from at least one manipulated or measured variable. The first computing unit is designed to calculate a control variable using a predetermined control matrix and the determined components of the predetermined frequency of the manipulated or measured variable. According to the invention, the control device is designed to to provide several predetermined control matrices and to select one control matrix from the several predetermined control matrices to calculate the controlled variable, wherein the control matrix is ​​selected depending on a predetermined operating mode for the electrical machine. Furthermore, the following is planned:

[0007] An electric drive system comprising an electric machine and a control device according to the invention. Advantages of the invention

[0008] The present invention is based on the understanding that disturbances such as ripple or harmonic overtones can occur in electrical drive systems. Such disturbances can, for example, lead to torque with harmonic overtones due to the design. Furthermore, it is possible for electrical parameters, such as the phase currents supplying the electric machine, to be affected, or for higher-frequency disturbances to propagate via an electrical energy input point for a drive system. Compensating for or minimizing such disturbances requires a complex and, in some cases, costly control system.

[0009] The present invention is therefore based on the idea of ​​taking this knowledge into account and providing a simple, efficient, and robust control system for an electric machine. In particular, such a control system for the electric machine is intended to enable efficient control for minimizing or optimizing disturbances in an electric drive system. A particular focus here is on minimizing or compensating disturbances with a frequency corresponding to a harmonic overtone of the motor frequency.

[0010] The aim is to extract predetermined frequency components from one or more manipulated or measured variables of an electric drive system and to calculate a control variable using a previously calculated control matrix and the determined frequency components of the manipulated or measured variable. Since the control matrix for calculating the control variable only needs to be determined once, the control variable can be easily determined during operation. This enables very fast control of an electric drive system. In particular, such control requires a low computational load, allowing for the use of relatively simple and cost-effective hardware. This reduces the implementation costs.

[0011] The predetermined frequency, or possibly several predetermined frequencies, for which the frequency components are extracted from the manipulated or measured variables, can be determined, in particular, based on the electrical frequency of the electric machine. Specifically, one or more harmonic overtones of the electric drive system can be used as the basis for determining the frequency components. In this way, influences and disturbances caused by these harmonic overtones in the electric machine and the entire drive system can be very effectively optimized and mitigated.

[0012] Control or measurement variables that can be used for controlling the electric machine can, for example, include variations, particularly harmonic overtones, in the torque curve of the electric machine. Furthermore, fluctuations, especially influences with frequencies of harmonic overtones, in currents or voltages of the electric machine or the electric drive system are also possible as control or measurement variables for a control system according to the invention. In addition, other parameters that are directly available as control or measurement variables, or quantities that can be indirectly derived from other control or measurement variables, can also be used for the control. For example, it is also possible to specify parameters for the desired noise generation of the electric machine.By selectively adjusting, emphasizing, or attenuating frequency components, particularly harmonic overtones, the noise level of the electric machine and thus of the entire drive system can be influenced. For example, it is also possible to deliberately increase the noise level of the electric machine using this method. This can, for instance, draw the attention of a user or a person in the vicinity of the electric machine to a potential hazard or event.

[0013] According to one embodiment, the at least one manipulated or measured variable comprises a torque or torque curve of the electric machine, an electric current and / or an electric voltage of the electric machine, and / or a desired noise level of the electric machine. Furthermore, any other variables can also be used as manipulated or measured variables. In particular, manipulated or measured variables from which further variables can be derived can also be considered. Thus, the control of the electric machine can also be adjusted to predetermined values ​​that cannot be directly measured.

[0014] According to one embodiment, the method includes a step for transforming the controlled variable into the time domain in the step for calculating the control variable. First, a calculation is performed in the frequency domain using a control matrix and frequency components of the manipulated or measured variables, and then the result is transformed into the time domain, so that the controlled variable is subsequently available for control in the time domain.

[0015] According to the invention, the method comprises a step for providing several predetermined control matrices and for selecting a control matrix from these several predetermined control matrices for calculating the controlled variable. In particular, different control matrices can be provided for different applications. These multiple control matrices can, for example, be pre-calculated so that no further calculation of control matrices is necessary during the control of the electrical machine. The provided control matrices can, for example, be stored in a memory. Alternatively, providing the control matrices can also include, for example, calculating the control matrices within the processing unit itself. In this case, the required control matrices can optionally be calculated online, i.e., during operation.Furthermore, the procedure can include a step to select a control matrix from several predefined control matrices. Subsequently, the controlled variable can be derived from the frequency components of the manipulated or measured variables using the selected control matrix. In this way, a suitable, adapted control matrix can be selected for different applications based on the several different control matrices. This allows for easy adaptation of the control system to different applications.

[0016] According to the invention, the control matrix is ​​selected depending on a predefined operating mode for the electric machine. Such operating modes can, for example, take into account different optimization of parameters. For instance, different control matrices can adjust different proportions of the harmonic overtones to varying degrees. Similarly, different noise levels of the electric machine are possible based on different control matrices. Furthermore, any other optimization goals can also be defined by different control matrices.

[0017] According to one embodiment, the method further comprises a step for calculating a target variable from the determined proportions of the predetermined frequencies of the acquired manipulated or measured variables. Such a target variable can, in particular, include a quantity that cannot be directly measured in the electric machine or drive system. By calculating, i.e., deriving, this target variable from other quantities that can be measured in the electric machine or drive system, it is thus possible to efficiently determine quantities that can only be determined indirectly. In particular, another matrix, which can also be created beforehand, can be used to calculate the target variable from manipulated or measured variables.

[0018] According to one embodiment of the electric drive system, the electric machine comprises an asynchronous machine or a permanent magnet synchronous machine.

[0019] The above embodiments and further developments can be combined with one another as appropriate. Further embodiments, developments, and implementations of the invention also include combinations of features of the invention described previously or subsequently with respect to the exemplary embodiments, even if not explicitly mentioned. In particular, those skilled in the art will also add individual aspects as improvements or additions to the respective basic forms of the present invention. Brief description of the drawings

[0020] The present invention will be explained in more detail below with reference to the exemplary embodiments shown in the schematic figures of the drawing.

[0021] This shows: Fig. 1: a schematic representation of an electric drive system with a control device for an electric machine according to one embodiment; Fig. 2: a schematic representation of an electric drive system with a control device for an electric machine according to a further embodiment; and Fig. 3: a schematic representation of a flowchart as it underlies a method for controlling an electrical machine according to one embodiment. Embodiments of the invention

[0022] Fig. Figure 1 shows a schematic representation of an electric drive system with a control device for an electric machine 4. The control device comprises a first transformation unit 1 and a first computing unit 2. The control variable R calculated in the control device can be provided to a control unit 3, which uses the calculated control variable R to control the electric machine 4.

[0023] Any manipulated or measured variables P can be used as input signals for the control device. For example, a control voltage from the electric drive system can be used as an input signal. Furthermore, an electric current of the electric machine 4, such as the electric currents in the phase terminals of the electric machine 4, can be detected and provided as measured variables. Additionally, the torque curve of the electric machine 4 can be detected or calculated and used as an input variable P for the control device. Other manipulated or measured variables are also possible. In particular, a specified noise level for the electric machine 4 can also be provided as an input variable P.For example, the noise level of the electric machine 4 can be influenced by vibrations in its torque curve. Furthermore, it is possible to measure parameters from which other, potentially non-directly measurable, parameters of the electric machine 4 or the entire electric drive system can be derived. Additionally, the voltage or current at the input of a power converter (not shown here) of the electric drive system can be evaluated and incorporated into the control of the electric machine 4. In this way, for example, feedback from the electric drive system into the power supply network feeding the drive system can be adjusted or minimized.

[0024] The manipulated or measured variables P provided at the input of the control device are first split into predetermined frequency components at the first transformation unit 1. For each manipulated or measured variable P, the component of a predetermined frequency or the components of several predetermined frequencies can be extracted. The predetermined frequencies whose components are extracted from the manipulated or measured variables can, in particular, be frequencies of harmonic overtones of the electrical frequency of the electrical machine 4. For example, the cosine and sine components of a harmonic to be controlled can be extracted from the time signals of the manipulated or measured variables P by demodulating, i.e., multiplying, the signals with the respective basis functions. This can be done, for example, using the multipliers 11 and 12 in the first transformation unit 1.

[0025] For a permanent magnet synchronous machine, the current reference angle for the respective harmonic overtone is given by ϕ = m · N · ϕ_mech. Here, ϕ_mech represents the mechanical rotor angle, N the number of pole pairs, and m the harmonic order. From the sine component Vs and the cosine component Vc calculated in this way, a complex vector V = Vs + jVc can be formed. However, a purely real-valued implementation is also possible without further restrictions. The complex vectors V or the individual sine or cosine components can be combined in a weighting matrix H. A quality factor J can then be derived for subsequent control. J=V*H*H V.

[0026] Here, an asterisk (*) denotes the transposed complex conjugate element. Minimizing this quality function J corresponds to minimizing the harmonic components in the manipulated or measured variables P. To define different weightings and thus different priorities for individual objectives, the weighting matrix H mentioned above is used. Considering a steady-state transfer characteristic in the electric machine 4, the quality function J can be interpreted as a function of the harmonics in the manipulated or measured variables P of the electric machine 4. Based on gradient-based optimization, the coefficients of the manipulated or measured variables P for the electric drive system can then be iteratively adjusted to minimize the quality function J.This ultimately results in the coefficients of the manipulated variables being calculated by simply multiplying an operating-point-dependent matrix G with the complex vectors of the frequency components of the manipulated or measured variables. This can be done, for example, in the multiplication device 20. Here, the complex operating-point-dependent matrix G can be calculated in advance and therefore does not burden the computing power during the control process.

[0027] The corresponding multiplication of the operating point-dependent matrix G with the coefficient vectors V can, for example, be carried out in the first computing unit 2.

[0028] Finally, a control variable R can be generated by modulation with the respective basic functions in modulators 21 and 22, which is then fed to the control unit 3. The control unit 3 then controls the electric machine 4 with the generated control variable R.

[0029] Fig. Figure 2 shows a schematic representation of a control device for an electric machine 4 according to a further embodiment. This embodiment essentially corresponds to the embodiment described above and has been further extended by a second computing unit 24. In this second computing unit 24, parameters of the electric machine 4 or the entire electric drive system that cannot be directly measured can be derived from the measured quantities. Thus, it is also possible to implement control for target values ​​that cannot be directly determined in the electric drive system. This calculation of the further parameters of the electric drive system can also be carried out, for example, by a simple multiplication of the measured values ​​with a previously created matrix.This enables particularly simple control of the electric drive system and especially the electric machine 4, even for target values ​​that are not measured or may not be directly measurable. To set predefined setpoint values ​​Sa and Sb, these setpoint values ​​can also be provided as setpoint specifications in the control device.

[0030] Fig.Figure 3 shows a schematic representation of a flowchart underlying a method 100 for controlling an electric machine 4 according to one embodiment. In step 110, at least one manipulated or measured variable is acquired. In step 120, predetermined frequency components of this at least one acquired manipulated or measured variable are determined. As previously described, these components can be generated by demodulation (multiplication) with corresponding basis functions. In principle, any other method for extracting frequency components of the manipulated or measured variables P is also possible. Preferably, but not necessarily, a complex vector can be generated from the extracted frequency components. The frequency components can, in particular, be components of frequencies from predetermined harmonic overtones of the electric machine 4.

[0031] In step 130, a control variable is calculated using a predetermined control matrix G and the frequency components of the manipulated or measured variables determined in step 120. This calculated control variable can then be used to control the electrical machine accordingly in step 140.

[0032] As previously explained, the control matrix G for the control of the electric machine 4 can be calculated offline in advance, thus avoiding any load on the control system's computing power during operation. Different control matrices G are possible for different operating modes of the electric drive system. If necessary, several different control matrices G can be calculated in advance and stored in a suitable memory device. This allows the control behavior of the electric machine to be adjusted during operation simply by changing the control matrix G. For example, different control matrices G can be created to pursue different objectives during the control of the electric machine 4.This makes it possible, on the one hand, to minimize harmonic overtones in the torque curve of the electric machine 4. Such a minimization of fluctuations in the torque curve of the electric machine leads to a particularly smooth operation of the electric machine.

[0033] In some cases, it may be desirable to deliberately deviate from such an optimized torque curve, for example, to increase the noise level of the electric motor 4. This can, for instance, draw the attention of a user or another person in the vicinity of the electric motor 4. This can, for example, alert the user to potential hazards. Furthermore, deliberately increasing the noise level of the drive system in an electric or hybrid vehicle can also be used to alert people in the vicinity of the vehicle.

[0034] Furthermore, the feedback effect of the electric drive system on the power supply network to which it is connected can be minimized, for example, through appropriate control. Moreover, by suitable adjustment of the respective control matrix, it is not only possible to pursue one or the other objective, but also to pursue a combination of several objectives and to weight the individual objectives accordingly.

[0035] In summary, the present invention relates to a control system for an electric machine. To control the electric machine, predetermined frequency components are extracted from the machine's manipulated or measured variables and multiplied by a previously calculated control matrix. This control matrix can be calculated in advance. Different control matrices can be generated for different applications. This enables simple, efficient, and robust control of the electric machine, particularly for optimizing and minimizing harmonic distortion.

Claims

[1] Method (100) for controlling an electrical machine (4), comprising the steps: Acquire (110) at least one manipulated or measured variable (P); Determining (120) proportions of a predetermined frequency in the detected manipulated or measured variable (P); Calculating (130) a control variable (R) using the determined components of the predetermined frequency of the detected manipulated or measured variable (P) and a predetermined control matrix; and Controlling (140) the electric machine (4) with the calculated control variable (R), wherein the method is characterized by the following steps: Providing several predetermined control matrices and selecting one control matrix from the several predetermined control matrices to calculate the controlled variable (R), wherein the control matrix is ​​selected depending on a given operating mode for the electrical machine (4). [2] Method (100) according to claim 1, wherein the at least one actuating or measuring variable (P) comprises a torque of the electric machine (4), an electric current in the electric machine (4), a noise generation of the electric machine (4) and / or a control voltage of the electric machine (4). [3] Method (100) according to claim 1 or 2, wherein the step to calculate the control variable (R) comprises a transformation of the control variable (R) into the time domain. [4] Method (100) according to any one of claims 1 to 3, comprising a step for calculating a target variable from the determined proportions of the predetermined frequency of the detected manipulated or measured variable (P). [5] Control device for an electric machine (4), comprising: a first transformation device (1) designed to determine components of a predetermined frequency from at least one manipulated or measured variable (P); and a first computing device (2) designed to calculate a control variable (R) using a predetermined control matrix and the determined components of the predetermined frequency of the manipulated or measured variable (P), characterized by that the control device is set up to to provide several predetermined control matrices and to select one control matrix from the several predetermined control matrices to calculate the controlled variable (R), wherein the control matrix is ​​selected depending on a given operating mode for the electrical machine (4). [6] Control device according to claim 5, comprising a control unit (3) designed to control the electrical machine (4) with the calculated control variable (R). [7] Control device according to claim 5 or 6, comprising a second computing device (24) designed to calculate a target variable from the determined proportions of the predetermined frequency of the detected manipulated or measured variable (P). [8] Electric drive system with: an electric machine (4); and a control device according to one of claims 5 to 7.

Images