Method and apparatus for operating an electronically commutated electrical machine

Abstract

The invention relates to a method for operating an electronically commutated, multiphase electrical machine (3) using a power driver (2) with several power driver modules (21), comprising the following steps: - Operating the electrical machine (3) by setting phase voltages, in particular by means of pulse width modulated control; - Performing a procedure for asymmetric power loss compensation with the following further steps: ◯ Measuring phase temperatures (t PhaU , t PhaV , t PhaW ) for each of the power driver modules (21); ◯ Conversion of phase temperatures (t PhaU , t PhaV , t PhaW ) into a temperature space pointer with two pointer components (t TPtrα , t TPtrβ ) in a Cartesian coordinate system; ◯ Rules of the pointer components (t TPtrα , t TPtrβ) to a predetermined setpoint, in order to determine for each pointer component (t TPtrα , t TPtrβ ) to obtain a power loss difference component of a power loss difference phasor as a manipulated variable; ◯ Converting the power loss difference components of the power loss difference pointer (pwrPtrα, pwrPtrβ) into power loss differences to be set for each power driver module (21); ◯ Using the power loss differences to be provided for each power driver module (21) in a method for asymmetric power loss compensation.

Description

Technical field

[0001] The invention relates to electronically commutated electrical machines, which are controlled via power driver modules, each with one or more power transistors. The invention further relates to a method for compensating for asymmetrical power losses via the power transistors, e.g., for compensating for temperature differences and / or differing thermal heat dissipation. Technical background

[0002] Multiphase electric machines are typically controlled by a power driver module, for example, in the form of an array of half-bridge units, particularly in a B6 topology, with each power driver module supplying one phase of the electric machine. The power driver modules can each comprise series-connected power transistors, the heat generated by their power losses being dissipated via a heat sink and a coolant flow. Because the series arrangement of the power driver modules along the coolant flow results in uneven heat dissipation via the heat sink, a different operating temperature is established in each power transistor and in each power driver module, depending on the power consumption of the load.

[0003] For example, the power transistors are arranged on a heat sink, which is cooled by a flow of coolant. The coolant heats up due to the power dissipation of an upstream power transistor and thus heats downstream power transistors, resulting in different coolant temperatures for the power transistors. Since the temperature difference between the power transistor and the heat sink is generally the same for the same power dissipation and thermal resistance, the operating temperature of the power transistors will differ.

[0004] The power driver modules of the power driver are generally operated in such a way that the losses within them are distributed evenly, i.e., symmetrically. This means that the applied phase currents have the same RMS value, the same amplitude, and the same switching pattern, and, assuming the active areas of the power transistors are at the same or similar temperatures, generate symmetrical power losses in all phases. Due to the symmetrical phase currents and identical switching pattern, the power dissipation in the power driver modules is then the same, heating the coolant. This, however, leads to an increasing temperature rise in the coolant from one power transistor / power driver module to a power transistor / power driver module downstream, in the direction of coolant flow.As a result, with increasing coolant temperature and similar temperature fluctuations via the cooling system, the power transistors for the individual phases will reach different temperatures. These temperature differences can be compensated for, for example, by providing asymmetrical power dissipation in the power driver modules.

[0005] One possible method for asymmetric power loss compensation, as known, for example, from German patent application DE 10 2017 207 301 A1, involves determining a temperature distribution within a multiphase converter and setting predetermined phase currents in the converter based on this temperature distribution, resulting in a different effective phase current for each phase. Setting predetermined phase currents in the converter involves superimposing a fundamental frequency with at least one further harmonic of a predetermined phase current. By injecting second harmonics into the phase currents, asymmetric power loss is generated by influencing conduction losses.

[0006] However, this method requires a voltage component of the available voltage to impress the second harmonics. A voltage reserve is therefore necessary, meaning this method cannot be operated at maximum voltage on the electric machine or in field-weakened operation. Furthermore, if current control is required to set a predetermined motor torque, a fundamental difficulty arises in regulating the phase currents as the electrical frequency increases, since the switching count—that is, the number of possible control loops per electrical period—becomes too low. Thus, the voltage reserve and the switching count limit the availability of the second harmonic impression method for regulating asymmetrical power loss at high speeds.

[0007] Another method is known, for example, from DE 10 2016 220 893 A1 and describes a power converter for providing a multiphase alternating voltage for an electrical machine that can be coupled to the power converter, wherein control signals for controlling switching elements of half-bridges are generated in such a way that an angle-dependent offset voltage is set for the provided alternating voltage, which is set depending on the temperature of the switching elements.

[0008] By selectively shifting an offset voltage, switching operations can be reduced, thus relieving half-bridge units / phases by saving switching losses. However, reducing switching losses in half-bridge units / phases with excessive power dissipation results in asymmetrical power losses compared to the other phases. Shifting the offset voltage to reduce switching operations is also known as DPWM (Discontinuous PWM) or Flat Top. However, such methods, which rely on shifting the offset voltage, cause very high DC link ripple at medium output levels. This ripple decreases with increasing operating voltage.

[0009] The object of the present invention is to provide an improved method for operating a power driver for supplying phase currents for an electric machine that can be operated with asymmetrical power losses over the entire speed ranges. Disclosure of the invention

[0010] This problem is solved by the method for operating an electric machine using an asymmetric power loss compensation according to claim 1 and by the device and motor system according to the dependent claims.

[0011] Further details are specified in the dependent claims.

[0012] According to a first aspect, a method for operating an electronically commutated, multiphase electrical machine using a power driver with several power driver modules is provided, comprising the following steps: - Operating the electric machine by setting phase voltages, which are adjusted using pulse-width modulated control; - Performing a procedure for asymmetric power loss compensation with the following further steps: ◯ Measuring phase temperatures for each of the power driver modules; ◯ Converting the phase temperatures into a temperature space vector with two vector components in a Cartesian coordinate system; ◯ Adjust the pointer components to a predetermined setpoint in order to obtain a power loss differential component of a power loss differential pointer as a manipulated variable for each pointer component; ◯ Converting the power loss difference components of the power loss difference pointer into power loss differences to be set for each power driver module; ◯ Using the power loss differences to be provided for each power driver module in a method for asymmetric power loss compensation.

[0013] Furthermore, it may be provided that the power loss differences for each power driver module are taken into account in such a way that the power loss difference indicates a measure of a reduction or an increase in the power loss for the power driver module in question.

[0014] It may be provided that the temperature space vector is determined from the phase temperatures by assuming for each of the phase temperatures an angle with respect to a stator-fixed coordinate system, which corresponds to the angle of the phase determined by the power driver module associated with the phase temperature.

[0015] Multiphase power drivers are constructed with a number of power driver modules corresponding to the number of phases. These power driver modules can be configured, for example, as half-bridge units, each comprising at least one high-side semiconductor switch and one low-side semiconductor switch. The semiconductor switches can be implemented using power electronic semiconductor devices such as MOSFETs, IGBTs, IGCTs, and the like.

[0016] The power driver modules are typically connected to a cooling system through which a coolant flows. However, conventional cooling system topologies lead to inhomogeneous heat dissipation for the semiconductor switches in the individual power driver modules due to the successive heating of the coolant. This results in different temperatures being reached in the semiconductor switches during operation. These temperatures are determined by the heat loss generated in the respective semiconductor switches and the temperature difference between the semiconductor switch and the coolant at the point of thermal coupling between the semiconductor switch and the cooling system.

[0017] Conventional methods involve adjusting the power dissipation for individual power driver modules, or their heat dissipation (e.g., via a cooling device), to compensate for temperature differences between phases as much as possible. However, compensating for temperature differences becomes complex with more than two phases and an increasing number of phases, and is typically only designed for a single operating point. Therefore, a simplified and more flexible method for adjusting the power dissipation of each of the multiple phases is planned.

[0018] For this purpose, it is planned to convert the phase temperatures, i.e. the temperatures of the semiconductor switches of the power driver modules assigned to the phases, into a temperature vector which can be specified as a space vector with respect to a fixed, in particular stator-fixed, reference system in a Cartesian coordinate system or in a polar coordinate system.

[0019] This temperature range indicator can be uniquely determined by two quantities, regardless of the number of phases of the power driver. For this purpose, the phases, i.e., the individual power driver modules, can each be assigned a pointer angle, where the pointer angle assigned to the phases is uniformly distributed over a unit circle according to the number of phases (for example, 0°, 120°, and 240° for three phases).

[0020] The phase angle of the temperature space vector thus provides an assignment to the affected adjacent phases or power driver modules, i.e., to the phase that is warmest.

[0021] The input transformation for determining the quantities used to represent the temperature space vector can be performed using a well-known Clarke transformation. For a three-phase system with phases at 0°, 120°, and 240° phase angles, the temperature vector tTPtrα, tTPtrβ can be derived from the typically measured phase temperatures t PhaU , t PhaV , t PhaW as follows: (tTPtrαtTPtrβ)=GTPtr⋅(1−12−12032−32)⋅(tPhaUtPhaVtPhaW)

[0022] G is involved TPtr Freely selectable, depending on the desired interpretation of the pointer length, for example with: GTPtr=13 To convert to polar coordinates, the length tTPtrAmp of the temperature space vector can be calculated using known methods for vectors: tTPtrAmp=|tTPtrαtTPtrβ|

[0023] The angle agTPtrAg of the temperature space pointer corresponds to: agTPtrAg=atan2(tTPtrβ, tTPtrα)

[0024] Using a control system, the phasor values ​​of the temperature space phasor can be assigned to power loss differential values ​​of a power loss differential phasor. The power loss differential phasor can then be transformed into the required power loss differentials for the individual phases, while maintaining the assignment of the phases to a phase angle corresponding to the phase angles used for determining the temperature phasor.

[0025] The power loss differential vector can be controlled using, for example, a PI controller with a proportional and an integrator component, or other controller structures. For this purpose, both temperature vector values, tTPtrα and tTPtrβ, can be controlled separately to zero. Since the setpoint of the control is usually zero, a control mechanism, e.g., using a lookup table or similar, can be implemented instead of a control mechanism.

[0026] To limit the adjustment range of the control, an anti-windup can be taken into account for the integrator component of each individual control.

[0027] Instead of PI controls, alternative control approaches, such as multivariable control and the like, can also be implemented.

[0028] The regulation can also include a feedforward control.

[0029] The setpoint for the controls can be set to 0, as the goal is to balance temperature differences as much as possible. However, other setpoints are also conceivable to positively influence the system's behavior.

[0030] Furthermore, the control for balancing the phase temperatures can be carried out continuously during the operation of the electric machine.

[0031] Alternatively, release conditions can be checked, whereby temperature differences in the phases are only equalized if a release condition is met. Release criteria can include, for example, exceeding a temperature threshold for a maximum phase temperature, exceeding a temperature threshold for the coolant temperature, exceeding a power loss threshold for the total power loss of the power driver when operating an electric machine, and / or exceeding a torque / phase current threshold of the electric machine.

[0032] The manipulated variable (control output) of both the control and the regulation results in a power loss differential vector with two power loss differential vector quantities pwrPtrα, pwrPtrβ, which, according to a back transformation in the phases, represent the power loss differences pwr DeltaU , pwr DeltaV , pwrDel-taW They can be converted. This can be done, for example, using the following transformation formula: (pwrDeltaUpwrDeltaVpwrDeltaW)=1GPwrPtr(10−1232−12−32)⋅(pwrPtrαpwrPtrβ)

[0033] The method for asymmetric power loss compensation may include an asymmetric control method or an asymmetric phase current method to achieve power loss equalization. For example, the aforementioned asymmetric control method can adjust the power losses of the individual phases according to the power loss differences. This method uses predefined functions to determine a flat-top pointer length. agFlatTopPtrAmp=f1(pwrPtrAmp,...) to determine, whereby pwr PtrAmp corresponds to the amplitude of the power loss difference indicator. and a flat-top pointer angle agFlatTopPtrAg=f2(agpwrPtrAg,...) where ag pwrPtrAg , corresponds to the position angle of the power loss difference pointer. to determine. The function f1 describes the relationship between the target power loss pointer difference length and the flat-top pointer length, which describes how long the pulse width modulation is suspended within a cycle period.

[0034] The function f2 describes the relationship between the target power loss pointer difference angle and the flat-top pointer angle, which describes the position (phase) at which the pulse width modulation is applied. The flat-top pointer length and angle can then be converted into flat-top angle widths using the following transformation. (agFlatTopPtrαagFlatTopPtrβ)=(cos(agFlatTopPtrAg)sin(agFlatTopPtrAg))⋅agFlatTopPtrAmp (agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW)=23(10−1232−1232)⋅(agFlatTopPtrαagFlatTopPtrβ) (agFlatTopAlphaDesUagFlatTopAlphaDesVagFlatTopAlphaDesW)=(agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW)−min(agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW) where (agFlatTopPtrαagFlatTopPtrβ) em flat-top pointer in in static reference system, (agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW) the flat-top angle widths for each phase as an intermediate size and (agFlatTopAlphaDesUagFlatTopAlphaDesVagFlatTopAlphaDesW) The target flat-top angle widths for each phase must correspond.

[0035] This transforms the flat-top widths such that one phase always has a flat-top width of 0° and therefore does not perform "flat-top" operation in this phase, and the others have an angle greater than or equal to 0° as their flat-top width.

[0036] Alternatively, to convert the control pointer of the power loss differences, the method with asymmetric phase currents can be achieved by imprinting a second harmonic as a current harmonic onto the DQ motor currents. This current harmonic is realized by two current harmonics with complementary, opposing oscillation patterns and can be described by a total of four parameters (IdPos, IdNeg, IqPos, IqNeg).

[0037] According to another aspect, a device, in particular a control unit, is provided for operating an electronically commutated, multi-phase electrical machine, wherein the device is designed to: - Operating the electric machine by setting phase voltages, which are adjusted using pulse-width modulated control; - Performing a procedure for asymmetric power loss compensation with the following steps: ◯ Measuring phase temperatures for each of the power driver modules; ◯ Converting the phase temperatures into a temperature space vector with two vector components in a Cartesian coordinate system; ◯ Adjust the pointer components to a predetermined setpoint in order to obtain a power loss differential component of a power loss differential pointer as a manipulated variable for each pointer component; ◯ Converting the power loss difference components of the power loss difference pointer into power loss differences to be set for each power driver module; ◯ Using the power loss differences to be provided for each power driver module in a method for asymmetric power loss compensation. Brief description of the drawings

[0038] The embodiments are explained in more detail below with reference to the accompanying drawings. These show: Fig. 1 a schematic representation of a motor system with an electric machine controlled by a power driver; and Fig. 2 A functional block diagram to illustrate the control of the electric machine using a temperature space vector to achieve asymmetric power loss compensation. Description of embodiments

[0039] Fig. Figure 1 schematically shows a motor system 1 with a power driver 2 for providing phase voltages for controlling phases of an electronically commutated electric machine 3. The power driver 2 can in particular be provided as a B6 bridge circuit with three power driver modules in the form of half-bridge units 21, each comprising series connections of two semiconductor switches 22, e.g. MOSFETs, IGBTs or the like.

[0040] The half-bridge units 21 are supplied by a supply voltage with a high and a low supply voltage potential VH, VL.

[0041] The semiconductor switches 22 are provided with heat dissipation via a cooling system 5. The cooling system can be designed to achieve heat removal from the semiconductor switches 22 via a coolant flow channel 51, which can be, for example, air or a liquid. Heat removal typically occurs serially along the semiconductor switches 22, so that the coolant heats up as it flows through the coolant flow channel 51, resulting in different temperatures for the semiconductor switches 22.

[0042] The phase temperatures of the power driver modules 21 can be measured by suitably attached temperature sensors 24.

[0043] The electrical behavior of semiconductor switches 22 depends significantly on the temperature, even with otherwise identical control. To compensate for the effect of operation at different temperatures of the active areas of the semiconductor switches 22, an asymmetric power dissipation compensation (also called asymmetric power dissipation generation) is therefore provided, which results in different power dissipations for the power driver modules.

[0044] In principle, the power driver 2 is controlled via a control unit 4 to control the semiconductor switches 22 via a suitable commutation method in order to operate the electric machine 3. In particular, phase voltages are set using pulse-width modulated control of the half-bridge units 21. For this purpose, the semiconductor switches 22 are controlled in a manner known per se with a control signal that produces a specific phase voltage according to a predetermined duty cycle.

[0045] Furthermore, methods for asymmetric power loss compensation are known. For example, DE 10 2017 207 301 A1 discloses a method by which a second harmonic is impressed into the phase currents in order to generate asymmetric power loss by influencing the conduction losses.

[0046] Another method for asymmetric power loss compensation is the asymmetric control method, as described, for example, in DE 10 2016 220 893 A1. This involves shifting the offset voltage by maintaining a phase voltage at either the high or low supply voltage potential over the control cycles of a PWM control signal, in order to reduce the number of switching operations. This allows the switching losses for individual half-bridge units / phases to be reduced and asymmetric power losses compared to the other phases to be achieved.

[0047] Asymmetric power loss compensation is applied to power loss differences for each of the phases resulting from temperature control of the phase controls.

[0048] In Fig.Figure 2 shows a functional block diagram illustrating a method for compensating phase temperatures. This diagram describes the control of the phase losses of the individual phase / power driver modules of the power driver, thus compensating for temperature differences between them.

[0049] In a temperature pointer block 31, the phase temperatures t detected by the temperature sensors 24 on the individual power driver modules 21 are displayed. PhaU , t PhaV , t PhaW according to a pointer transformation into a temperature space pointer with temperature pointer sizes t TPtrα , t TPtrβThe data is provided in either the Cartesian or polar coordinate system. In the following, it is assumed that the temperature pointer quantities represent the pointer quantities (α, β) in the Cartesian coordinate system. Here, phase U is assigned an angle of 0°, phase V an angle of 120°, and phase W an angle of 240°.

[0050] The following applies: (tTPtrαtTPtrβ)=GTPtr⋅(1−12−12032−32)⋅(tPhaUtPhaVtPhaW)

[0051] These are each fed to a difference element 32, 33, in which the difference is compared to a respective target value t. refα , t refß is determined. The target values ​​t refα , t refβ are preferably 0, which indicates that there should be no temperature differences between the phase / power driver modules 21.

[0052] In alternative embodiments, the target values ​​can also take on values ​​other than 0.

[0053] The resulting control deviations are each fed to a control element 34, 35, which can preferably be configured as a polar controller. The control elements 34, 35 provide power loss differential vector quantities pwrPtrα, pwrPtrβ with an α and β component of a power loss differential vector. To limit the intensity of the manipulated variables, an anti-windup for the integrator component of the control elements 34, 35 can be provided. Furthermore, the output variables of the control elements 34, 35 can be limited by limiting the length of the resulting power loss differential vector with respect to a limit value. This can be achieved, for example, using a polar limiter.

[0054] The power loss difference pointer is now converted into power loss difference quantities in a conversion block 36 using a corresponding method for asymmetric power loss compensation, which together with the average power loss of all phases / power driver modules can be used to control the power driver modules of the power driver.

[0055] The asymmetric power loss compensation is carried out in the power loss compensation block 37. (agFlatTopPtrαagFlatTopPtrβ)=(cos(agFlatTopPtrAg)sin(agFlatTopPtrAg))⋅agFlatTopPtrAmp (agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW)=23(10−1232−1232)⋅(agFlatTopPtrαagFlatTopPtrβ) (agFlatTopAlphaDesUagFlatTopAlphaDesVagFlatTopAlphaDesW)=(agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW)−min(agFlatTopTmpUagFlatTopTmpVagFlatTopTmpW)

[0056] The sizes ag FlatTopAlphaDesU , ag FlatTopAlphaDesV , ag FlatTopAlphaDesWThese are now used to control the electric machine 3. The position of the control cutoff is set symmetrically to the phase voltage maximum.

[0057] Alternatively, the control can also be achieved using the method of asymmetric phase currents. Other methods for converting the phase-related power loss differences can also be applied. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2017 207 301 A1 [0005, 0045] DE 10 2016 220 893 A1 [0007, 0046]

Claims

[1] Method for operating an electronically commutated, multiphase electrical machine (3) using a power driver (2) with several power driver modules (21), comprising the following steps: - Operating the electrical machine (3) by setting phase voltages, in particular by means of pulse width modulated control; - Performing a procedure for asymmetric power loss compensation with the following further steps: ◯ Measuring phase temperatures (t PhaU , t PhaV , t PhaW ) for each of the power driver modules (21); ◯ Conversion of phase temperatures (t PhaU , t PhaV , t PhaW ) into a temperature space pointer with two pointer components (t TPtrα , t TPtrβ ) in a Cartesian coordinate system; ◯ Rules of the pointer components (t TPtrα , t TPtrβ) to a predetermined setpoint, in order to determine for each pointer component (t TPtrα , t TPtrβ ) to obtain a power loss difference component of a power loss difference phasor as a manipulated variable; ◯ Converting the power loss difference components of the power loss difference pointer (pwrPtrα, pwrPtrβ) into power loss differences to be set for each power driver module (21); ◯ Using the power loss differences to be provided for each power driver module (21) in a method for asymmetric power loss compensation. [2] Method according to claim 1, wherein the power loss differences for each power driver module (21) are taken into account such that the power loss difference indicates a measure of a reduction or an increase in the power loss for the power driver module (21) concerned. [3] Method according to claim 1 or 2, wherein the temperature space indicator is derived from the phase temperatures (tPhaU , t PhaV , t PhaW ) is determined by assuming for each of the phase temperatures an angle with respect to a stator-fixed coordinate system that corresponds to the angle of the phase determined by the power driver module associated with the phase temperature. [4] Method according to any one of claims 1 to 3, wherein the control of the pointer components is carried out with the setpoint of 0. [5] Method according to one of claims 1 to 4, wherein the method is carried out only when a release condition is present, which in particular includes exceeding a temperature threshold for a maximum phase temperature, exceeding a temperature threshold for the coolant temperature, exceeding a power loss threshold of the total power loss of the power driver (2) during operation of the electric machine (3) and / or exceeding a threshold value of a torque / phase current of the electric machine (3). [6] Method according to any one of claims 1 to 5, wherein the method for asymmetric power loss compensation comprises a method of asymmetric control or a method of asymmetric phase currents. [7] Device, in particular a control unit (4), for operating an electronically commutated, multi-phase electrical machine (3), wherein the device is configured to: - Operating the electric machine (3) by setting phase voltages which are adjusted using pulse width modulated control; - Performing a procedure for asymmetric power loss compensation with the following steps: ◯ Measuring phase temperatures (t PhaU , t PhaV , t PhaW ) for each of the power driver modules (21); ◯ Conversion of phase temperatures (t PhaU , t PhaV , t PhaW ) into a temperature space pointer with two pointer components (t TPtrα , t TPtrβ ) in a Cartesian coordinate system; ◯ Rules of the pointer components (t TPtrα , t TPtrβ ) to a predetermined setpoint, in order to determine for each pointer component (t TPtrα , t TPtrβ ) to obtain a power loss difference component of a power loss difference phasor as a manipulated variable; ◯ Converting the power loss difference components of the power loss difference pointer (pwrPtrα, pwrPtrβ) into power loss differences to be set for each power driver module (21); ◯ Using the power loss differences to be provided for each power driver module (21) in a method for asymmetric power loss compensation. [8] Motor system (1) with the device according to claim 7 and an electric machine (3).

Images