Method for Determining Phase Currents of a Rotating Multiphase Electrical Machine Fed by Means of a PWM-Controlled Inverter

Abstract

The disclosure relates to a method for determining phase currents of a rotating multiphase electrical machine fed by means of a PWM-controlled inverter. In this case, injection voltages applied in at least one stipulated PWM period are determined. An evaluation direction for a phase current vector is also determined and a division of current measurements for the individual phase currents is determined on the basis of the evaluation direction. The phase currents are then determined on the basis of the previously determined division of the current measurements.

Claims

1. A method for determining phase currents of a rotating multiphase electrical machine fed by a PWM-controlled inverter, the method comprising: determining injection voltages at a time point during one controller sampling period that are applied to drive the rotating multiphase electrical machine in at least one PWM period following the time point; determining an evaluation direction of a phase current vector of the rotating multiphase electrical machine based at least one of (i) on the injection voltages, (ii) an operating point of the rotating multiphase electrical machine, and (iii) a temperature-dependency of the phase current vector; determining, for each of the phase currents, a number of current measurements to be carried out in each case in at least one passive switching state of the PWM-controlled inverter within the at least one PWM period, based on the evaluation direction; and determining the phase currents that flow during the at least one PWM period of the rotating multiphase electrical machine, based on the respective number of current measurements.

2. The method as claimed in claim 1, wherein the number of current measurements for each phase current corresponds, for all the phase currents in total, to a maximum possible number of current measurements within the at least one passive switching state of the PWM-controlled inverter.

3. The method as claimed in claim 1, the determining the phase currents further comprising: determining at least one of the phase currents using a plurality of A/D converters.

4. The method as claimed in claim 3, the determining the phase currents further comprising: sampling, with the plurality of A/D converters, the at least one phase current with a time offset with respect to one another.

5. The method as claimed in claim 1, the determining the phase currents further comprising: determining the phase currents based on current measurements of the respective number of current measurements that take place midway in the at least one passive switching state.

6. The method as claimed in claim 1, the determining the phase currents further comprising: determining the phase currents based on current measurements of the respective number of current measurements that take place with mirror symmetry with respect to one another.

7. The method as claimed in claim 1, wherein: the determining, for each of the phase currents, the number of current measurements includes determining as zero the number of current measurements for a first phase current of the phase currents that deviates most in terms of direction from the evaluation direction; and the determining the phase currents includes determining the first phase current based on others of the phase currents using Kirchoff's first law.

8. The method as claimed in claim 1, further comprising: determining a Clarke transformation based on at least one of (i) the number of current measurements for each of the phase currents and (ii) the evaluation direction; and determining the phase current vector in the evaluation direction based on the Clarke transformation and those of the phase currents whose number of current measurements is greater than zero.

9. An electrical machine, wherein the electrical machine is of rotating, multiphase design and is fed by means of a PWM-controlled inverter, the electrical machine being, in order to determine phase currents thereof, configured to, determine injection voltages at a time point during one controller sampling period that are applied to drive the rotating multiphase electrical machine in at least one PWM period following the time point; determine an evaluation direction of a phase current vector of the rotating multiphase electrical machine based at least one of (i) on the injection voltages, (ii) an operating point of the rotating multiphase electrical machine, and (iii) a temperature-dependency of the phase current vector; determine, for each of the phase currents, a number of current measurements to be carried out in each case in at least one passive switching state of the PWM-controlled inverter within the at least one PWM period, based on the evaluation direction; and determine the phase currents that flow during the at least one PWM period of the rotating multiphase electrical machine, based on the respective number of current measurements.

Description

DRAWINGS

[0040] FIG. 1 shows an exemplary embodiment of a method according to the invention for determining phase currents of a rotating multiphase electrical machine fed by means of a PWM-controlled inverter.

[0041] FIG. 2 shows an exemplary phase current-time diagram for the method illustrated in FIG. 1 and the time points of the current measurements of the respective phase currents carried out by means of the A/D converter.

[0042] FIG. 3 shows an exemplary embodiment of a multiphase electrical machine that is designed to carry out a method according to the invention.

DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

[0043] FIG. 1 shows a first exemplary embodiment of a method according to the invention.

[0044] The phase currents i.sub.a, i.sub.b, i.sub.c of a three-phase rotating electrical machine fed by means of a PWM-controlled inverter are determined by means of the method. The rotor position of the electrical machine can, for example, be estimated on the basis of the determined phase currents i.sub.a, i.sub.b, i.sub.c.

[0045] At the beginning, in a method step a, at a time point during a controller sampling period, injection voltages that will be applied in at least one PWM period following this time point for drive of the electrical machine are determined. The injection voltages are determined by means of a controller. The injection voltages are here composed of voltages specified by the controller for drive of the electrical machine, which generate a corresponding fundamental wave current together with additional, high-frequency voltages. The fundamental wave current here provides the desired operation of the machine with, for example, a predetermined rotation speed or also with a predetermined torque. Through the high-frequency voltages, a position-dependent current change can for example, furthermore, be achieved, while conclusions can, for example, be drawn from this current change as to the rotor position of the electrical machine.

[0046] In a method step b an evaluation direction can then be determined depending on the injection voltages determined in method step a. The evaluation direction is here, for example, oriented to the direction of the high-frequency voltage injection which is fed in in addition to the voltages required to adjust the manipulated variables. The reason for this is that the high-frequency current changes generated by the high-frequency voltage contain corresponding rotor position information, wherein the direction of the current change is correlated to the direction of the high-frequency voltage injection. If it is now, for example, assumed that the total rotor position information is contained in phase current i.sub.a, the direction of the phase current i.sub.a is determined as the evaluation direction for the phase current vector.

[0047] After method step b, the number of current measurements that should be carried out for each of the three phase currents i.sub.a, i.sub.b, i.sub.c in at least one passive switching state of the inverter within the at least one PWM period is determined in a method step c. This is determined here depending on the evaluation direction determined in method step b. Since, in method step b, the evaluation direction is for example determined in the direction of the phase current i.sub.a, the number of current measurements for the phase current i.sub.a is maximized in order to optimize the signal-to-noise ratio for the phase current i.sub.a. The number of current measurements assigned to the phase current i.sub.b can, moreover, be merely one, and the number of current measurements can be simply zero for the phase current i.sub.c. The maximum number of current measurements depends, for example, on the sampling time of the A/D converters and their dead time. Only a limited total number of current measurements can thus be carried out in each passive switching state, and these can then be appropriately distributed over the individual phase currents i.sub.a, i.sub.b, i.sub.c.

[0048] The phase currents i.sub.a, i.sub.b, i.sub.c are then determined in a method step d, wherein this determination takes place depending on the number of current measurements respectively determined in method step c. The single current measurements of the phase current i.sub.b is necessary in order to be able to uniquely determine the phase current vector. As already explained, in this case no current measurement is necessary at all for the phase current i.sub.c, since this can be determined, for example, from the phase current i.sub.a and the phase current i.sub.b by means of Kirchoff's first law as follows:

i.sub.c=−i.sub.a−i.sub.b

[0049] Optionally, after method step d, a further method step e and a method step f can take place. In method step e a Clarke transformation is here determined for each of the phase currents depending on the number of current measurements determined in method step c or also depending on the evaluation direction determined in method step b.

[0050] In method step f, the phase current vector in the evaluation direction is subsequently determined depending on the Clarke transformation determined in method step f and on the phase currents whose number of current measurements is greater than zero determined in method step d.

[0051] In this example, the Clarke transformation used for evaluation is therefore only dependent on the phase currents i.sub.a and i.sub.b.

[0052] FIG. 2 shows an exemplary phase current-time diagram for the method illustrated in FIG. 1 and the time points of the current measurements of the respective phase currents carried out by means of the A/D converters.

[0053] The curves of the phase currents i.sub.a and i.sub.b are plotted against time t. The duration of a PWM period 10 is also illustrated, wherein the PWM period 10 has a PWM start 11 and a PWM midway point 12.

[0054] The sampling behavior of a first A/D converter AD1 and a second A/D converter AD2 is also illustrated, wherein the first and second A/D converters AD1 and AD2 can sample respectively all the phase currents i.sub.a, i.sub.b and i.sub.c by means of current sensors as required. The current measurements are carried out here both around the PWM start 11 and the PWM midway point 12.

[0055] As described in FIG. 1, the phase current i.sub.a should be measured as often as possible, since this has the greatest information content. For this reason, the phase current i.sub.a is measured twice and the phase current i.sub.b only once by means of the first A/D converter AD1 around the PWM start 11, so leading to the current values i.sub.a(0), i.sub.b(0) and i.sub.a(1). Only the phase current i.sub.a is, on the other hand, measured by the A/D converter AD2 around the PWM start 11, which accordingly leads to the current values i.sub.a(2), i.sub.a(3) and i.sub.a(4).

[0056] It is to be noted that the phase currents i.sub.a and i.sub.b are sampled by the first A/D converter AD1 with mirror symmetry with respect to one another. The two measurements of the phase current i.sub.a thus surround the measurement of the phase current i.sub.b. In addition, the A/D converters AD1 and AD2 sample with a small mutual offset.

[0057] The distribution and performance of the current measurements by the A/D converters AD1 and AD2 are correspondingly repeated around the PWM midway point 12 and around the PWM start of the following PWM period.

[0058] In order then finally to determine the phase currents i.sub.a, i.sub.b and i.sub.c, the current values i.sub.a(0), i.sub.a(1), i.sub.a(2), i.sub.a(3) and i.sub.a(4) recorded for the phase current i.sub.a are averaged. The current value i.sub.b(0) recorded for the phase current i.sub.b is also used, and the phase current i.sub.c calculated by means of Kirchoff's first law from the averaged phase current i.sub.a and the phase current i.sub.b.

[0059] FIG. 3 shows an exemplary embodiment of a multiphase electrical machine (100) which is designed to carry out a method according to the invention.

[0060] The electrical machine 100 here has a stator-rotor unit 110. This stator-rotor unit 110 is fed by an inverter 120 which can, for example, be designed as a B6 bridge. The inverter 120 is in turn driven by a PWM generation unit 130 which converts the specifications received from a current controller 140 into corresponding PWM duty ratios. The current controller 140 is designed to sample the phase currents i.sub.a, i.sub.b and i.sub.c of the stator-rotor unit 110 which is not, however, illustrated here. The electrical machine 100 can furthermore comprise a processing unit 150 that is designed to control the current control in such a way that the phase currents i.sub.a, i.sub.b and i.sub.c are determined by means of the method according to the invention, for example according to FIG. 1.