WO2022058100A1

WO2022058100A1 - Acceleration of electrical synchronous machines with an optimized thermal budget

Info

Publication number: WO2022058100A1
Application number: PCT/EP2021/072658
Authority: WO
Inventors: Andreas Speinle; Armin Ruf
Original assignee: Robert Bosch Gmbh
Priority date: 2020-09-21
Filing date: 2021-08-15
Publication date: 2022-03-24
Also published as: EP4214831A1; DE102020211759A1; CN116195180A

Abstract

Method (100) for operating an electrical synchronous machine (2), having the following steps: - a source DC voltage U_=,Q or source AC voltage U_~,Q with a source frequency f_Q is provided (110); - the source DC voltage U_=,Q or source AC voltage U_~,Q is converted (120) into a target AC voltage U_~,Z with a target frequency f_Z in an inverter or frequency converter (1) by means of temporally modulated switching of an arrangement (1a) of semiconductor switches, wherein, during at least one acceleration process of the synchronous machine (2), the arrangement (1a) of semiconductor switches is switched at an average modulation frequency f_M which is lower than the nominal frequency f_N used during steady-state operation of the synchronous machine (2) at a constant speed. - the synchronous machine (2) is fed (140) with the target AC voltage U_~,Z.

Description

description

Title:

Accelerating synchronous electric machines with optimized heat budget

The present invention relates to the operation of electrical synchronous machines that are supplied with an AC voltage of variable frequency via an inverter or frequency converter.

State of the art

Electrical synchronous machines run synchronously with a rotating field generated by the stator, in which the rotor moves. In order to be able to operate the synchronous machine with a variable speed, an AC voltage with a variable frequency is therefore required. Such an AC voltage can be generated, for example, in an inverter or frequency converter from a source DC voltage or from a source AC voltage. Such a power converter connects the three or more phases of its output according to a predetermined time program in alternation with the source DC voltage or with the source AC voltage, so that an at least approximately sinusoidal target AC voltage with the desired target frequency is present at its output .

It is important here that the rotating field approximates the sinusoidal shape as closely as possible, since deviations from the sinusoidal shape can lead to eddy current losses in the rotor. In order to approximate the sinusoidal shape as well as possible, it is again important to switch with the highest possible modulation frequency in the specified time program. The higher the modulation frequency, the less discretization effect that one phase of the output may or may not be connected to either the DC source voltage or the AC source voltage at any given time. From the DE 10 2018 217 051 A1 discloses a method with which a particularly high modulation frequency can be provided by operating several inverters or frequency converters in parallel.

Disclosure of Invention

A method for operating an electrical synchronous machine was developed as part of the invention.

In this method, a direct source voltage U=,Q or an alternating source voltage U˜,Q with a source frequency fo is provided. This source DC voltage U=,Q or source AC voltage U~,Q is converted into a target AC voltage U~,z with a target frequency fz in an inverter or frequency converter by time-modulated switching of an arrangement of semiconductor switches. The target AC voltage U˜,z can, in particular, be three-phase or multi-phase. By switching, each of the phases of the target AC voltage U~,z can be connected alternately to the poles of the source DC voltage U=,Q or source AC voltage U~,Q according to a predetermined time program. The synchronous machine is supplied with the target AC voltage U~,z.

During at least one acceleration process of the synchronous machine, the arrangement of semiconductor switches is switched with a mean modulation frequency fM, which is lower than the nominal frequency fu used during steady-state operation of the synchronous machine with a constant speed.

It was recognized that such acceleration processes were hitherto decisive for the dimensioning of the semiconductor switches used.

Switching with a high modulation frequency fu means that the target AC voltage is closely approximated to the sinusoidal shape and accordingly fewer eddy current losses occur in the rotor. If the rotating field is sinusoidal in each phase, the rotor is in a fixed with the rotor connected reference system in a constant magnetic field, so that no eddy currents are induced. On the other hand, deviations from the sinusoidal shape, such as harmonics, affect the rotor-fixed reference system and lead to eddy currents.

On the other hand, high modulation frequencies fu in connection with high currents lead to switching losses in the semiconductor switches, which in turn lead to these semiconductor switches heating up. Particularly high currents are required during an acceleration process. For example, the power requirement of a synchronous machine that drives an electric turbo compressor and requires around 15 kW of electrical power in steady-state operation at a constant speed can increase by around 5-6 kW when the speed is increased from 20,000 rpm to 100,000 rpm . The semiconductor switches must be designed so that their junction temperature never exceeds the maximum permissible value, even with the high currents required for this, otherwise there is a risk of irreversible damage to the semiconductor switches. The design for higher currents, on the other hand, cannot usually be achieved with improved cooling of the semiconductor switches alone, but also requires the semiconductors to be built larger.

However, semiconductor switches with larger semiconductors are disproportionately more expensive to manufacture in relation to the increase in current carrying capacity. This is due, among other things, to the fact that monocrystalline semiconductor material is usually only available with a specific defect density per unit area or volume. The larger the piece of semiconductor required to manufacture a semiconductor switch, the greater the likelihood that there will be a defect and the lower the yield in mass production. This is reflected in the final price of the semiconductor switch, especially when the starting material is an expensive semiconductor with a particularly wide band gap, such as silicon carbide.

It has now been recognized that, especially in the case of synchronous machines for driving turbocompressors, the high currents for increasing the speed are only required for a very short time, since these compressors have a very compact design. Typically, for said acceleration of 20,000 rpm 100,000 rpm only takes about one second. If the modulation frequency of the semiconductor switches is reduced for this short time, it is accepted that the target AC voltage U~,z contains harmonics and other deviations from the sinusoidal shape during this time and that the rotor is correspondingly heated by eddy current losses. However, the rotor has some heat capacity, so it can withstand heating for a short time before reaching its maximum allowable temperature. After the end of the acceleration process, the heat can be dissipated again by cooling the rotor. At the same time, the reduced modulation frequency fM avoids excessive heating of the semiconductor switches, despite the high currents required for acceleration, so that it is no longer necessary to oversize the semiconductor switches as was previously the case.

If the acceleration process lasts longer than the rotor with its heat capacity can buffer the heating caused by eddy currents, the increased stress caused by the eddy currents should be stopped. In a particularly advantageous embodiment, the mean modulation frequency fM is therefore initially reduced in response to the fact that acceleration of the synchronous machine is requested. Before the temperature TR of the rotor of the synchronous machine exceeds a predetermined threshold value, the modulation frequency fM is increased again to the nominal frequency fu and the current I supplied to the synchronous machine is reduced. This means that in case of doubt the acceleration process is weakened or even stopped if the reduced modulation frequency fM can no longer be maintained.

This is particularly advantageous when the semiconductor switches are not designed to carry the currents that are increased during the acceleration process even at a modulation frequency that corresponds to the nominal frequency. If use is made of the possibility created by the method of dimensioning the semiconductor switches smaller than hitherto, then this should not result in these semiconductor switches overheating in certain situations. In particular, rapid heating to a junction temperature that is too high causes the material to be subject to high mechanical stress due to thermal expansion. The by such Consequences caused by stress are not completely reversible, but add up over time, analogously to a summation poison, until the semiconductor switch eventually fails.

How long the rotor can buffer increased heating caused by eddy currents depends on the current operating situation of the synchronous machine. Therefore, in a further particularly advantageous embodiment, a maximum period of time for which the arrangement of semiconductor switches is switched with the lower mean modulation frequency fM is determined using a model that consists of at least

• the current-time profile of the current I supplied to the synchronous machine,

• the speed n of the synchronous machine and

• the ambient temperature predicts the development of the rotor temperature TR. These variables are typically monitored anyway during operation of the synchronous machine and are therefore available without additional effort.

Alternatively or in combination with this, the temperature T of the rotor can be monitored by measurement. A non-contact infrared thermometer, for example, can be used for this purpose.

In a further advantageous embodiment, the mean modulation frequency fM is only reduced in response to an acceleration of the synchronous machine being requested if the absolute difference between the requested setpoint speed ns and the current actual speed ni exceeds a predefined threshold value. Such a large deviation is associated with large requested currents IB. On the other hand, smaller deviations can also be corrected with smaller currents IB, for which it is not necessary to use the rotor as a heat buffer. In particular, the threshold value prevents the small deviations that occur again and again in normal operation between the target speed ns and the actual speed ni from leading to a permanent heat input into the rotor, so that the rotor can no longer keep up with its speed when greater acceleration is requested full heat capacity is available as a heat buffer. If the arrangement of semiconductor switches is operated in pulse width modulation, the mean modulation frequency can be reduced, for example, in particular by reducing the frequency of the pulse width modulation. The time program for switching is then only scaled in terms of time and otherwise remains unchanged. Alternatively or in combination with this, for example, the switching state of one phase of the arrangement can be kept constant over one period of the pulse width modulation, with the phase affected by this being changed in turn. This modification of the time program causes only a slight deviation of the target AC voltage U~,z from the sinusoidal shape, but reduces the average modulation frequency fM and thus the heat load on the semiconductor switches by a third.

If the semiconductor switches are controlled by a current controller in order to regulate the current supplied to the synchronous machine to a desired value, the mean modulation frequency fM can be reduced, for example, by increasing a hysteresis and/or a time constant of the current controller. The "sacrifice" to be made in the interest of not heating up the semiconductor switches too much is then not made in the form of a deviation of the target AC voltage U~,z from the sinusoidal shape, but in the form of a somewhat reduced control quality of the current controller. This means that no additional eddy currents are generated in the rotor.

As explained above, the mean modulation frequency fM is advantageously reduced to such an extent that the maximum junction temperature Tj in the arrangement of semiconductor switches remains below a predetermined threshold value during the acceleration process. This temperature Tj can be measured directly. The change in the mean modulation frequency fM can therefore be adapted dynamically. If the heat buffer provided by the rotor is not quantitatively sufficient and the junction temperature Tj approaches the critical limit, the current I supplied to the synchronous machine can be reduced in order to weaken the acceleration process or stop it altogether. In particular, the method can be fully or partially computer-implemented. The invention therefore also relates to a computer program with machine-readable instructions which, when executed on one or more computers, cause the computer or computers to carry out one of the methods described. In this sense, embedded systems for technical devices and control units for vehicles that are also able to execute machine-readable instructions are also to be regarded as computers.

The invention also relates to a machine-readable data carrier and/or a download product with the computer program. A downloadable product is a digital product that can be transmitted over a data network, i.e. can be downloaded by a user of the data network and that can be offered for sale in an online shop for immediate download, for example.

Furthermore, a computer can be equipped with the computer program, with the machine-readable data carrier or with the downloadable product.

As explained above, the method described above makes it possible to dispense with large oversizing of these semiconductor switches when generating a target AC voltage U˜,z with a variable frequency fz for an electrical synchronous machine with an arrangement of semiconductor switches. The invention therefore also relates to a system with an electrical synchronous machine and an arrangement of semiconductor switches, in which this waiver is manifested.

The arrangement of semiconductor switches is designed to convert a source DC voltage U=,Q or a source AC voltage U~,Q with a source frequency fo by time-modulated switching into a target AC voltage U~,z with a target frequency fz to convert, which is performed as a supply voltage to the electrical synchronous machine. The arrangement of semiconductor switches is designed in such a way that it is used for acceleration processes the current IB provided for the synchronous machine can only carry at a lower mean modulation frequency fs than the nominal frequency fu provided for stationary operation of the synchronous machine with a constant speed.

Further measures improving the invention are presented in more detail below together with the description of the preferred exemplary embodiments of the invention with the aid of figures.

exemplary embodiments

It shows:

FIG. 1 embodiment of the method 100;

FIG. 2 Effects of an exemplary use of the method 100;

Figure 3 embodiment of the system 10.

Figure 1 is a schematic flowchart of an embodiment of the method 100 for operating a. In step 110, a direct source voltage U=,Q or an alternating source voltage U˜,Q with a source frequency fo is provided. In step 120, the source DC voltage U=,Q or source AC voltage U~,Q in an inverter or frequency converter 1 is converted into a target AC voltage U~,z with a target frequency fz by time-modulated switching of an arrangement la of semiconductor switches transformed. In step 130, during at least one acceleration process of the synchronous machine 2, the arrangement la of semiconductor switches is switched with a mean modulation frequency fM, which is lower than the nominal frequency fu used during steady-state operation of the synchronous machine 2 at a constant speed. In step 140, the electrical synchronous machine 2 is supplied with the target AC voltage U.about.z, which has the frequency fz. According to block 131, the mean modulation frequency fM can be reduced in response to the fact that an acceleration of the synchronous machine 2 is requested. According to block 132, it can then be checked whether the temperature TR of the rotor of the synchronous machine 2 exceeds a predetermined threshold value. If this is the case (truth value 1), the modulation frequency fM is increased again to the nominal frequency fu according to block 133, and according to block 134 the current I supplied to the synchronous machine 2 is reduced.

According to block 132a, a maximum time At, for which the arrangement 1a of semiconductor switches is switched with the lower mean modulation frequency fM, can be determined using a model 3 that predicts the development of the temperature T of the rotor. This temperature TR can also be measured directly according to block 132b.

According to block 135, it can be checked whether the absolute difference between the requested setpoint speed ns and the current actual speed ni exceeds a predetermined threshold value. The reduction of the mean modulation frequency fM according to block 136 can then be restricted to the case where the threshold value is exceeded (truth value 1). Otherwise (truth value 0) this reduction does not take place.

According to block 121, the arrangement 1a of semiconductor switches can be operated, for example, in pulse width modulation. According to block 137a, the mean modulation frequency fM can then be reduced by reducing the frequency of the pulse width modulation. Alternatively or in combination with this, according to block 137b, the switching state of one phase of the arrangement 1a can be kept constant over a period of the pulse width modulation, with the phase affected by this being changed in turn.

The semiconductor switches can be controlled by a current regulator, for example, in order to regulate the current I supplied to the synchronous machine to a desired value. According to block 138, the mean modulation frequency fM can then be reduced by increasing a hysteresis and/or a time constant of the current controller. In general, according to block 139, the mean modulation frequency fM can be reduced to such an extent that the maximum junction temperature Tj in the arrangement la of semiconductor switches remains below a predetermined threshold value during the acceleration process.

Figure 2 shows the effects of an exemplary use of the method 100 in an acceleration process of the synchronous machine 2. Above the time t measured in seconds are the speeds n, the current I supplied to the synchronous machine 2, the modulation frequency fM of the arrangement la of semiconductor switches in the inverter or frequency converter 1 , the junction temperature Tj in the arrangement la of semiconductor switches and the temperature TR of the rotor of the synchronous machine 2 are plotted.

Method 100 is based on the situation in which setpoint speed ns of synchronous machine 2 jumps from 20,000 rpm to 120,000 rpm at t=1 second. The actual speed ni reflects this request with a flatter edge.

At the moment when the setpoint speed ns jumps up, the current I supplied to the synchronous machine also jumps up to an acceleration level IB. Shortly before the actual speed ni reaches the target speed ns, the current I begins to fall until it reaches the steady-state level Is and remains there.

So that the arrangement la of semiconductor switches can carry the high current IB, it is operated with a modulation frequency fM below the nominal frequency fu for the steady-state current Is during the period in which the current I is above the steady-state level Is. Therefore, during the acceleration process, the junction temperature Tj overshoots the eventual steady state far less than if the device la were operated constantly at the nominal modulation frequency fu.

In return, the temperature TR of the rotor of the synchronous machine 2 increases more during the acceleration process than if the arrangement la were operated constantly at the nominal modulation frequency fu. This However, excess temperature TR is gradually reduced after the end of the acceleration process.

Figure 3 shows an embodiment of the system 10 from the synchronous machine 2 and the inverter 1, which consists of a source

DC voltage U=,Q generates a target AC voltage U~,z with target frequency fz. In terms of circuitry, the arrangement la of semiconductor switches corresponds to an arrangement that is customary for an inverter. However, it differs from this in the dimensioning of the semiconductors used. The semiconductors are dimensioned in such a way that the modulation frequency fM must be lowered below the nominal frequency fu for the stationary state when the current I supplied to the synchronous machine 2 becomes greater than the current Is provided for the stationary state.

Claims

Expectations

1. Method (100) for operating an electrical synchronous machine (2) with the steps:

• a source DC voltage U=,Q or source AC voltage U˜,Q with a source frequency fo is provided (110);

• the source DC voltage U=,Q or source AC voltage U~,Q is converted into a target AC voltage U~,z with a target Frequency fz converted (120), during at least one acceleration process of the synchronous machine (2), the arrangement (la) of semiconductor switches is switched (130) with a mean modulation frequency fM, which is lower than that during stationary operation of the synchronous machine (2). nominal frequency fu used at constant speed.

• the synchronous machine (2) is supplied with the target AC voltage U~,z (140).

2. The method (100) of claim 1, wherein

• the mean modulation frequency fM is reduced (131) in response to an acceleration of the synchronous machine (2) being requested and,

• Before the temperature TR of the rotor of the synchronous machine (2) exceeds a predetermined threshold value (132), the modulation frequency fM is increased again to the nominal frequency fu (133) and the current I supplied to the synchronous machine (2) is reduced (134).

3. The method (100) according to claim 2, wherein a maximum period of time At, for which the arrangement (la) of semiconductor switches with the lower average Modulation frequency fM is switched, based on a model (3) is determined (132a), which consists of at least

• the current-time profile of the current I supplied to the synchronous machine (2),

• the speed n of the synchronous machine (2) and

• the ambient temperature predicts the development of the rotor temperature TR.

4. The method (100) according to any one of claims 2 to 3, wherein the temperature T of the rotor is monitored by measurement (132b).

5. The method (100) according to any one of claims 2 to 4, wherein the mean modulation frequency fM in response to the fact that an acceleration of the synchronous machine (2) is requested, is only reduced (136) if the absolute difference between the requested target - Speed ns and the current actual speed ni exceeds a predetermined threshold value (135).

6. The method (100) according to any one of claims 1 to 5, wherein the arrangement (la) is operated from semiconductor switches in pulse width modulation (121) and wherein the average modulation frequency fM is reduced by

• the frequency of the pulse width modulation is reduced (137a) and/or

• the switching state of one phase of the arrangement (1a) is kept constant (137b) over a period of pulse width modulation, with the phase affected by this being changed in turn.

7. The method (100) according to any one of claims 1 to 6, wherein the semiconductor switches are controlled by a current regulator (122) in order to regulate the current I supplied to the synchronous machine (2) to a desired value, and wherein the mean modulation frequency fM is reduced by increasing a hysteresis and/or a time constant of the current controller (138).

8. The method (100) according to any one of claims 1 to 7, wherein the average modulation frequency fM is reduced so far (139) that during the acceleration process, the maximum junction temperature Tj in the - 14 -

Arrangement (la) remains from semiconductor switches below a predetermined threshold.

9. Computer program containing machine-readable instructions which, when executed on one or more computers, cause the computer or computers to carry out the method (100) according to any one of claims 1 to 8.

10. Machine-readable data carrier and/or download product with the computer program according to claim 9.

11. Computer with the computer program according to claim 9 and/or with the machine-readable data carrier and/or downloadable product according to claim 10.

12. System (10), comprising an electrical synchronous machine (2) and an inverter or frequency converter (1) with an arrangement (la) of semiconductor switches, which is designed to have a source DC voltage U=, Q or source AC voltage U~ To convert Q with a source frequency fo by time-modulated switching into a target AC voltage U~,z with a target frequency fz, which is routed as supply voltage to the electrical synchronous machine (2), the arrangement (la) of semiconductor switches is designed in such a way that it can carry a current IB provided for acceleration processes of the synchronous machine (2) only at a lower mean modulation frequency fM than the nominal frequency fN provided for stationary operation of the synchronous machine (2) with a constant speed.