CN105897101B

CN105897101B - Method for generating an electric rotating field and device for carrying out said method

Info

Publication number: CN105897101B
Application number: CN201610085586.5A
Authority: CN
Inventors: J.勒斯纳
Original assignee: Sanger Automotive Germany GmbH; Robert Bosch GmbH
Current assignee: Sanger Automotive Germany GmbH
Priority date: 2015-02-16
Filing date: 2016-02-15
Publication date: 2020-06-05
Anticipated expiration: 2036-02-15
Also published as: CN105897101A; DE102015202767A1

Abstract

A method for generating an electric rotating field rotating with a rotation period in a stator (10) of an electric machine, the stator having a plurality of stator connections (U-Y), each of which in an inverter (20) that can be triggered can be connected to a first potential (B +) of a DC voltage by triggering a first current valve (UH-YH) connected to the stator connection (U-Y) and to a second potential (B-) of the DC voltage by triggering a second current valve (UL-YL) connected to the stator connection (U-Y), respectively. It is provided that during at least one partial time interval of a rotation cycle either the first current valve (UH-YH) connected to at least two of the stator connections (U-Y) or the second current valve (UL-YL) connected to at least two of the stator connections (U-Y) is permanently activated, respectively, and that during at least one partial time interval the first and second current valves (UH-YH, UL-YL) connected to the remaining stator connections (U-Y) are alternately activated a plurality of times, respectively.

Description

Method for generating an electric rotating field and device for carrying out said method

Technical Field

The invention relates to a method for generating an electric rotating field rotating with a rotation period in a stator of an electric machine and to a device for carrying out the method.

Background

Electric rotating field drivers are well known. In order to generate their electrical rotating field or corresponding current, different modulation types can be used here, as are also known. Pulse Width Modulation (PWM) is popular for use in rotating field drivers. The adjustment in such a rotating field drive takes place in different ways depending on the respective current requirements. The main known methods are as follows: the use of sinusoidal current regulation (sinusoidal rectification), the use of block current regulation (block energization), the use of block voltage (control with block voltage), and the use of sinusoidal voltage regulation with superimposed zero voltage.

The previously described techniques can in principle be used in motors with any number of phases. In practice, motors with three phases are most commonly used. However, motors with other numbers of phases are also available, for example motors with one, two, four, five, six, seven or nine phases.

In order to generate the trigger signals required for the PWM operation of the inverter (pulse inverter) used, various techniques are known from the prior art, in addition to the so-called vector-based method and the so-called flat-top triggering (flat-panel triggering), as will be explained below. The known technology has disadvantages, however, and the technical problem to be solved by the invention is therefore to specify a better method for generating an electric rotating field in the stator of an electric machine and a device for carrying out the method.

Disclosure of Invention

Against this background, the invention proposes a method for generating an electrical rotating field in a stator of an electric machine, which rotates with a rotation cycle, and a device for carrying out the method. In a method for generating an electric rotating field rotating with a rotation period within a stator of an electric machine, the stator having a plurality of stator connections, each of which can be connected in a triggerable inverter to a first potential of a direct voltage by triggering a first current valve connected to the stator connection and to a second potential of the direct voltage by triggering a second current valve connected to the stator connection, respectively, the method is characterized in that during at least one local time interval of the rotation period either the first current valve connected to at least two of the stator connections or the second current valve connected to at least two of the stator connections is permanently triggered and during at least one local time interval the first current valve and the second current valve connected to the remaining stator connections are triggered alternately a plurality of times, respectively. The embodiments of the invention are described below.

Advantages of the invention

The invention relates to a method for generating an electric rotating field in a stator of an electric machine, said field rotating with a period of rotation. As is usual in this respect, the stator has a plurality of stator connections, each of which in a triggerable inverter (pulse inverter) can be connected to a first potential of the direct voltage by triggering a first current valve coupled to the stator connection and to a second potential of the direct voltage by triggering a second current valve coupled to the stator connection.

The term "rotation period" here refers to the time interval during which the electric rotating field completes a complete rotation. The rotation period thus comprises an electrical angle of 360 deg.. The current valve used can be designed in a manner known per se, for example as a metal oxide field effect transistor (MOSFET).

According to the invention, it is provided that during at least one partial time interval of the rotation cycle, either the first current valve connected to at least two of the stator connections or the second current valve connected to at least two of the stator connections is permanently activated, and that during the at least one partial time interval, the first and second current valves connected to the remaining stator connections are activated in each case a plurality of times in an alternating manner.

The invention is in principle applicable to motors having five phases, but can also be used in motors having more than five phases. The invention will be explained with the focus on a five-phase motor. In this case, during at least one partial time interval of the rotation cycle, either the first current valves connected to exactly two of the stator connections or the second control valves connected to these stator connections are permanently controlled.

The alternating activation of the remaining current valves is carried out in each case with a clocked cycle duration which corresponds in time to a fraction of the rotation cycle. The skilled person selects the respective beat timing in a suitable manner, as explained below. The duration of the partial time intervals of the rotation cycle, during which either the first current valve connected to at least two of the stator connections or the second current valve connected to at least two of the stator connections is permanently triggered, corresponds to a multiple of the cycle duration of the illustrated clock timing, so that the permanently triggered current valve remains in its triggered state at a plurality of clock timings.

The invention significantly reduces the switching losses that occur in conventional vector-based methods due to the high-frequency triggering of all current valves by means of the described measures. Within the framework of the invention, the great advantage over the flat top (Flattop) method described at the outset is also achieved, since instead of the current valves of only one phase connection, the current valves of at least two phase connections can remain in their respective activated state for a longer time interval (i.e. the above-mentioned partial time intervals of the rotation cycle). The rotation period can be divided into a plurality of corresponding local time intervals in order to distribute the load evenly to the involved current valves, for example. This is explained below.

In conventional vector-based triggering, one first observes possible switching vectors with which the inverter used can be triggered. If one looks at a five-phase inverter, this inverter has ten current valves, for example MOSFETs, which are arranged in the illustrated manner. The current valves provided for connecting the stator connections to a first potential of the dc voltage and for connecting the stator connections to a second potential (in the generic term of the present application "first" and "second" current valves or so-called "high-side" and "low-side" current valves) must always be actuated counter to one another in order to avoid a lateral short circuit, i.e. the first and second current valves connected to the stator connections are never actuated simultaneously. The switching states of the five current valves can thus be freely selected, so that 2 results⁵= 32 possible switch states. Each switching state produces a voltage vector, which is produced by the geometric addition of the voltages on the individual phases.

Two of these switching states (also called 00000 and 11111, that is to say all the stator connections are connected either to the first potential or to the second potential of the direct voltage) lead to a short circuit of the stator connections and therefore to a voltage vector of zero length.

Those switching states having ten numerically largest voltage vectors are typically observed from the remaining switching states. These voltage vectors are spread out decagonally. The details and the acronyms of the switch states can be referred to the relevant technical literature. Each arbitrary voltage vector within this decagon may be generated by a combination of two adjacent voltage vectors and a zero vector. The provision of a specific voltage vector can be achieved, for example, by providing a switching state 11001 for a triggering duration ta and a switching state 11000 for a triggering duration tb within a switching cycle or switching cycle of length Ts. One of the two zero vectors, that is, one of the switch states 00000 or 11111, is set for the remaining time Ts-ta-tb. The switching states of the switches during a switching period of length Ts result in a total of one pulse pattern, which in sequence comprises the respective switching states. The pulse pattern may include switching states in any order (e.g., center aligned or edge aligned PWM).

The illustrated triggering method is also known as Space Vector Pulse Width Modulation (SVPWM), a more precise description being found, for example, in the publication "Space Vector Modulation Schemes for a Five-phase voltage Source Inverter" (EPE 2005, Dresden, ISBN 90-75815-08-5) by A.Iqbal and E.Levi.

In the pulse-width-modulated triggering, the current valves of the inverter used, for example MOSFETs, are clocked at a constant frequency, which is significantly higher than the frequency of the fundamental wave of the electricity, i.e. the frequency of the current profile modulated for generating the rotating field. In this case, losses occur in the current valve during each switching operation. In order to keep this loss small, the beat frequency is selected to be as small as possible. However, the low clock frequency leads to an increased ripple of the phase current generated and thus of the torque generated, which also increases the noise emission of the electric drive.

As is also explained in the framework of the drawing description below, a voltage profile is generated by SVPWM, which differs significantly from the sinusoidal voltage profile of a conventional sinusoidal PWM control. But overall the same voltage space vector is generated. The important difference from sinusoidal PWM control is that at each point in time the two phase voltages have the same amplitude. This situation plays an advantageous role when SVPWM triggering is combined with a flat-top method, which is also known from the literature.

As is also explained in detail in the description of the figures, flat-top triggering is based on the simplification of the corresponding pulse pattern. As already mentioned and explained in detail below, the respective pulse patterns comprise several switching states in which all first (high-side) current valves are closed and several switching states in which all second (low-side) current valves are closed (00000 or 11111), respectively.

But these switch states are equally significant for providing the respectively desired voltage vectors. The switch state 00000 can be replaced by the switch state 11111 in each pulse mode. In this way, the current valves of the phase connections can be permanently held in a switched state, so that no switching losses occur for the phase connections. The switching losses can thus be reduced by a factor of three by flat-top triggering in a three-phase motor, since exactly one phase benefits from the explained advantages.

On the other hand, in sinusoidal PWM control of a five-phase motor, the current valves of the phase connections are permanently held in a switching state, as is already known, so that the switching losses are reduced by a factor of five in this case.

The invention provides for combining the SVPWM method and the top trigger in the framework of the measures explained. Since, during the SVPWM activation of the five phases, two adjacent phases always have the same voltage amplitude, this results, after the flat-top method has been used, in the two current valves of the five phases remaining in their respective switching state for a longer time. Switching losses are reduced by two fifths in this way, the reduction being greater in magnitude than by using the sinusoidal PWM method and only conventional flat-top triggering.

The method according to the invention generally offers advantages when the first and second current valves connected to the remaining stator connections are activated alternately within the framework of the described clock cycle with a period of at most one tenth of the rotation period. Values are typically up to about 10 hertz. Smaller frequencies are generally not used, since the current ripple increases and the noise formation increases here. In addition, the triggering takes place here in each case for a predetermined triggering duration. The remaining current valves are therefore triggered at a frequency, which is preferably constant, greater than the frequency of the electrical fundamental wave. The trigger time is corrected according to the respective current vector to be set.

As already mentioned, it may be advantageous to trigger only the first or the second current valve discontinuously in order to distribute its load. In an advantageous embodiment of the method according to the invention, it is therefore provided that, in a first partial time interval of the rotation cycle, the first current valves connected to at least two stator connections and the second current valves connected to at least two stator connections during a second partial time interval of the rotation cycle are continuously activated, and the first and second current valves connected to the remaining stator connections during the first and second partial time intervals are activated in each case in a plurality of alternating fashion. Although the time interval for keeping the triggering of the current valve for the phase connection constant is somewhat shortened in this way; however, the switching losses are significantly reduced without overloading the current valve.

The method according to the invention advantageously comprises: the rotation cycle is divided into a plurality of pulses and successive triggering patterns for the current valves are determined for each pulse by means of a modulation method, wherein for at least one partial time interval of the rotation cycle, the triggering pattern in which all the first current valves are triggered is replaced by the triggering pattern in which all the second current valves are triggered, or vice versa. Thus, the modulation method used, e.g., SVPWM triggering, is supplemented with flat-top triggering.

The computing unit according to the invention, for example a controller of an electric machine, is provided as a means for carrying out the method, in particular in terms of program technology, for carrying out the method.

The implementation of the method in the form of software is also advantageous, since this results in particularly low costs, in particular if the controller that is executed is also used for other tasks and is therefore always available. Suitable data carriers for supplying the computer program are, in particular, floppy disks, hard disks, flash memories, EEPROMs, CD-ROMs, DVDs etc. Programs can likewise be downloaded via computer networks (internet, intranet, etc.).

Further advantages and embodiments of the invention emerge from the description and the enclosed drawing.

The features mentioned above and those yet to be explained below can of course be used not only in the respectively specified combination but also in other combinations or alone without departing from the framework of the invention.

The invention is illustrated schematically in the drawings by means of embodiments and is described hereinafter with reference to the drawings.

Drawings

Fig. 1 shows a simplified partial schematic illustration of an arrangement with an electric machine and an inverter.

Fig. 2 shows a signal profile for explaining the basis of the method according to an embodiment of the invention.

Fig. 3A to 3D show signal profiles for explaining the basis of a method according to an embodiment of the invention.

Fig. 4 shows a signal profile for explaining the basis of the method according to an embodiment of the invention.

Fig. 5 shows a signal profile for explaining the basis of the method according to an embodiment of the invention.

Fig. 6 shows a signal profile for explaining the basis of the method according to an embodiment of the invention.

Detailed Description

Fig. 1 shows a schematic illustration of a device with an electric machine and a controllable inverter, which may be the basis of an embodiment of the invention, for example.

The machine comprises a stator 10 configured in a five-phase earth and in a five-pointed star circuit. The rotor of the electric machine, which is also present, and other details are not shown. The windings 11 of the stator 10 are each supplied with current via five phase connections U to Y via controllable current valves, which are likewise shown in a greatly simplified manner and can be switched on and off, here denoted by UL to YL and UH to YH, of the inverter 20.

The phase connections U to Y can be selectively connected in the inverter 20 via the current valves UL to YL and UH to YH to a first potential B + of the dc voltage or to a second potential B-of the dc voltage, for example a battery pole of an on-board electrical system of the motor vehicle. Current valves UH to YH, referred to herein as "first" current valves, are arranged in the upper rectifier branch H ("high-pressure side"). The current valves UL to YL, which are referred to herein as "second" current valves, are arranged in the second rectifier branch L ("low-side"). The current valves UL to YL and UH to YH are illustrated in the drawing in a simplified manner as transistors with parallel diodes. The diode here symbolizes the reverse diode which is typically present in a MOSFET. The current valves UH to YH and UL to YL can be triggered by a centralized or decentralized control device (not shown), for example by means of the described SVPWM method.

Fig. 2 shows the profiles of the nominal voltages Uu to Uy, as they occur in the individual phases or phase connections U to Y during SVPWM control. In fig. 2, the angle of the rotating field to be generated is plotted in radians on the abscissa and the voltage amplitude in percent is shown on the ordinate. Although the voltage profiles of the nominal voltages Uu to Uy differ significantly from the sinusoidal voltage profiles of conventional sinusoidal PWM control systems, they result overall in the same voltage space vector.

From the illustrated voltage profile of the nominal voltages Uu to Uy, a trigger signal for the current valve is generated, for example, by means of sinusoidal triangular modulation. In this case, a voltage amplitude of, for example, 70% means that at the time point for generating the nominal voltages Uu to Uy or in the instantaneous switching cycle, the respective first (high-side) current valve of the phase under consideration is switched on for a period duration of 70% and the respective second (low-side) current valve is switched on for a period duration of 30%.

As described, the important difference from sinusoidal PWM control is that at each point in time the two phase voltages have the same amplitude. This situation plays a particularly advantageous role when SVPWM triggering is combined with the flat top (Flattop) method also mentioned.

A known flat-top method for sinusoidal PWM triggering is shown in fig. 3A and 3B, and a flat-top method for SVPWM triggering, which is proposed according to one embodiment of the present invention, is shown in fig. 3C and 3D. The switching state of the current valves of each stator terminal U to Y is shown in each case during time t in one switching cycle. A very high signal level means that the first (high-pressure-side) current valve is switched on and the second (low-pressure-side) current valve is switched off, whereas at a low signal level the first (high-pressure-side) current valve is switched off and the second (low-pressure-side) current valve is switched on.

As already explained, the pulse pattern for the sinusoidal PWM triggering (fig. 3A) contains, within the shown switching cycle, time intervals in which all first (high-side) current valves are switched on simultaneously (time interval 301) and also time intervals in which all second (low-side) current valves are switched on simultaneously (time intervals 302 and 303).

The resulting short circuits (high-side and low-side) are synonymous for providing the desired voltage vector. Thus in the flat top trigger shown in fig. 3B, the high side short is replaced by a low side short accordingly, in other words,

time interval

302 or 303 in fig. 3A is extended by half of time interval 301 of fig. 3A to time

intervals

304 and 305, respectively. All high side firing intervals are shortened by the length of the shortest firing interval in phase W, and the low side firing intervals are lengthened by the same amount. The result of this is that in phase W the low-side switch is permanently switched on, and no switching losses occur anymore.

In a three-phase motor, the switching losses can be reduced by one third as described, since one phase is always switched on. In a five-phase motor with sinusoidal PWM triggering, one phase is also always switched on, so that the switching losses are reduced by a factor of five.

The invention now provides that the SVPWM method and the top trigger are combined as described, as shown in the embodiments of fig. 3C and 3D. Reference is made herein to fig. 2.

For setting the values, at the beginning of the voltage profile shown in fig. 2, phases U and V (see voltages Uu and Uv) are triggered to 85%, phases W and X are triggered to 15% and phase Y is triggered to 15%, respectively. The "triggering" of a phase means here that the first (high-pressure-side) current valve is switched on and the second (low-pressure-side) current valve is switched off. The pulse pattern shown in fig. 3C corresponds to this. This pulse pattern also includes, in the illustrated switching cycle, the time intervals in which all first (high-side) current valves are simultaneously switched on (time interval 306); and also includes time intervals during which all of the second (low-side) current valves are turned on simultaneously (time intervals 307 and 308).

The time interval 306 represents 15% of the entire beat duration. If the total time interval during which the first (high-pressure-side) current valve is now switched on is shortened by 15% as shown in fig. 3D, a continuous triggering over the entire period of the cycle occurs for phases W and X. Time intervals 306 to 308 (fig. 3C) are in this case identical in their effect to time intervals 309 and 310 (fig. 3D). As a result, phases U and V were triggered to 60%, phases W and X were 0% and phase Y was 35%.

Since two adjacent phases always have the same voltage amplitude during SVPWM triggering of the five phases, this results in two of the five phases always being turned on in the plateau, and the switching losses are therefore reduced by two fifths.

Fig. 4 shows in the same way as fig. 2 the trigger patterns for SVPWM for the phases with flat-top triggering (voltage trend UuFT) and without flat-top triggering (Uu). In this connection, to ensure a symmetrical loss distribution between the first and second (high-pressure-side and low-pressure-side) switching valves, all 36 ° electrical grounds alternate between a high-pressure-side plateau and a low-pressure-side plateau.

Fig. 5 shows in the corresponding diagram the voltage profiles UuFT to UyFT of all five phases during a flat triggering. It can be seen that at each time point, both phases were triggered to either 0% (low pressure side plateau) or 100% (high pressure side plateau).

Fig. 6 shows, for example, a trigger voltage 601 with a low-side plateau and two phase currents (602 and 603) as a time-based voltage or current profile.

It can also be seen here that the voltage is clocked only for 60% of the total duration and is switched to the low-voltage side for 40% of the total duration.

Claims

1. Method for generating an electric rotating field rotating with a rotation period within a stator (10) of an electric machine, which stator has a plurality of stator connections (U-Y), each of which can be connected in an inverter (20) that can be triggered to a first potential (B +) of a direct voltage by triggering a first current valve (UH-YH) connected to the stator connection (U-Y) and to a second potential (B-) of the direct voltage by triggering a second current valve (UL-YL) connected to the stator connection (U-Y), respectively, characterized in that during at least one local time interval of the rotation period either the first current valves (UH-YH) connected to at least two of the stator connections (U-Y) or the second current valves (UL-YL) connected to at least two of the stator connections (U-Y) are permanently triggered, respectively, and during at least one partial time interval the first and second current valves (UH-YH, UL-YL) connected to the remaining stator connections (U-Y) are each activated a plurality of times alternately.

2. Method according to claim 1, in which method the alternating activation of the first and second current valves (UH-YH, UL-YL) connected to the remaining stator connections (U-Y) takes place with a predetermined activation frequency of at least 10 kHz in cycles and with a predetermined activation duration in each case.

3. Method according to any one of the preceding claims, in which method the first current valve (UH-YH) connected to at least two of the stator connections (U-Y) during a first partial time interval of a rotation cycle and the second current valve (UL-YL) connected to at least two of the stator connections (U-Y) during a second partial time interval of a rotation cycle are continuously activated, and in which method the first and second current valves (UH-YH, UL-YL) connected to the remaining stator connections (U-Y) during the first and second partial time intervals, respectively, are alternately activated a plurality of times.

4. Method according to claim 1 or 2, which is used for generating an electric rotating field rotating with a rotation period in a stator (10) of an electric machine having five phase connections (U-Y), and in which method either a first current valve (UH-YH) connected to exactly two of the stator connections (U-Y) or a second current valve (UL-YL) connected to exactly two of these stator connections (U-Y) is permanently triggered during at least one partial time interval of the rotation period.

5. A method according to claim 1 or 2, the method comprising: dividing the rotation period into a plurality of beats; and for each cycle, determining successive activation patterns for the current valves (UH-YH, UL-YL) by means of a modulation method, wherein the activation pattern in which all first current valves (UH-YH) are activated is replaced by the activation pattern in which all second current valves (UL-YL) are activated, or vice versa, for at least one partial time interval of the rotation cycle.

6. A computing unit arranged to perform the method according to any one of the preceding claims.

7. A storage medium readable by a machine, on which is stored a computer program that triggers a computing unit for: the computer program, when executed on the computing unit, performs the method according to any of claims 1 to 5.