WO2011144197A2

WO2011144197A2 - Method for adjusting the rotor position of an electrically commutated motor

Info

Publication number: WO2011144197A2
Application number: PCT/DE2011/000420
Authority: WO
Inventors: Martin Rapp; Carsten Angrick; Martin Zimmermann
Original assignee: Schaeffler Technologies Gmbh & Co. Kg
Priority date: 2010-05-20
Filing date: 2011-04-18
Publication date: 2011-11-24
Also published as: WO2011144197A3; CN103081343A; DE112011101703A5; DE102011017395A1

Abstract

The invention relates to a method for adjusting a rotor position of an electrically commutated motor, in particular a brushless DC motor, in which the phases of the motor are actuated after determination of the rotor position via a block commutation. The invention is characterized in that the actuation of the motor phases is switched from block commutation to time-controlled commutation in order to adjust the rotor position with high precision.

Description

Method for controlling the rotor position of an electrically commutated motor

The invention relates to a method for controlling the Rotoriage an electrically commutated motor, in particular a brushless DC motor, in which the phases of the motor are driven by determining the position of the rotor via an event-controlled commutation.

"Event-controlled" in the context of the present invention means that an event (eg Hall interrupt or a certain angle) is detected and, due to the detected event, a commutation (preferably via the control software) is performed. An example of event-driven commutation is "block commutation".

In electrically commutated electric motors, which are used for example in clutch actuators, it is necessary to regulate the rotor position of the engine high resolution. For this purpose, a vector control can be performed. This vector control or field-oriented control can consist of a speed controller based on a subordinate current controller. In this case, the phase current of at least two phases of the motor can be measured in a high-frequency time grid, from which the further energization of the rotor is determined. Such a vector control leads to a very high computational effort in the drive unit.

In the application of the block commutation (as an example of the event-driven commutation), in which always two of the three three-phase windings of the motor are energized, a resolution of the Rotoriage can only be adjusted depending on the number of pole pairs of the motor. Furthermore, the block commutation has the disadvantage that it leads to a discontinuous torque curve with strong amplitudes.

The invention is therefore based on the object of specifying a method for controlling a rotor position of an electrically commutated motor, in which the rotor of the motor is precisely adjusted to at least 1 °, although the utilization of the evaluation unit with the motor control for vector control is substantially reduced , According to the invention the object is achieved in that the control of the phases of the motor is switched from the event-controlled commutation to a timed commutation for highly accurate adjustment of the position of the rotor. This is always at a certain time (time interrupt), the z. B. depending on a particular topology of the engine, a Umkommutierung made. This has the advantage that a process interrupt to initiate the commutation during the time-controlled commutation takes place in a fixed or customizable time grid. "Time frame" in connection with the present invention means a cyclically recurring software interrupt with the same time step size. "Fixed" here means that the software interrupt remains the same for a longer period of time or even the entire service life. "Adaptable" means that after a certain time and / or depending on preset conditions (eg speed limits) a change is made. In addition, the time-controlled commutation makes it possible to adjust the position of the rotor very precisely and in phase, which leads to a very precise regulation of the position of the rotor. An example of the time-steered commutation is a sine commutation.

Advantageously, the time-controlled commutation takes place at fixed predetermined time intervals, wherein in dependence on the determined position of the rotor of the motor drive values for the phases of the motor are read from a table. By the tabular assignment of the control values for the further energization of the three phases of the motor, the control of the position of the rotor is designed very accurately, since any small change in the rotor position can be occupied in the table with an exact control value. Furthermore, the computing power of the evaluation unit is reduced, whereby the utilization of the evaluation unit with tasks of the motor control is severely limited.

In one embodiment, by addition or subtraction of an offset to the current

certain position of the rotor caused by the read out from the table associated control values for the individual phases of the motor field weakening. A field weakening occurs whenever the stator field is not at an angle of 90 ° to the rotating field of the rotor. By changing the input values of the table changed control values are read out by the evaluation unit, with which the motor is controlled and resulting in the field weakening. As a result, an improved speed efficiency of the engine is achieved. This field weakening is not only in the time-controlled commutation, but also in event-driven commutation, z. B. Blockommutierung, feasible. In a further development, the event-controlled commutation, z. B. Block Kommutierung, triggered by a process interruption, which is triggered in response to the position of the rotor, in particular a predetermined angle change. Thus, the timing of the block commutation is determined by the position of the rotor, while the time-controlled commutation always takes place after the expiration of a fixed time window.

In one variant, a sensor determining the Lagades rotor detects an absolute angle of the rotor which is forwarded via an incremental interface generating at least two square-wave signals to an evaluation unit which counts the edges of the at least two square-wave signals generated by the incremental interface and from this the position of the rotor determined by the motor, wherein after a certain number of counted edges of the process interrupt is triggered. By using a high-resolution sensor, as is the case with the use of an absolute encoder, the rotor position of the motor can be determined very accurately, which is the starting point for a degree-accurate control of the rotor.

Advantageously, after a predetermined number of process interrupts occurs

Correction interrupt which is triggered after a smaller number of edges than the process interrupt. This correction is necessary because the process interrupt must be generated after a certain electrical angle of the drive signal. When converting to mechanical angular distances, values occur that are not integer, which is why a correction interrupt must be performed. In one variant, switching between event-controlled commutation (e.g.

Block commutation) and time-controlled commutation (eg sinusoidal commutation) as a function of a speed of the motor. Since the control of the motor is sinusoidal at low speeds, the impressed motor current corresponds approximately to the shape of the generator voltage of the motor. In this case, the phase offset is very small, so that it can be neglected for the highly accurate adjustment of the rotor position. A constant torque curve supports the exact adjustment of the position of the rotor. At higher speeds is switched to block commutation because the motor current of the motor voltage due to the motor inductance lags increasingly. As a result, the motor current is no longer impressed at the ideal time depending on the rotor position, so that no longer the maximum possible engine torque is generated. Therefore, the high-precision adjustment The phasing used the advantages that are present at low speeds by the sinusoidal control of the motor.

00013] In one embodiment, when the first predetermined speed threshold is undershot, the rotational speed of the engine is switched from the block commutation to the time-controlled commutation. Since there is almost no phase offset between motor current and motor voltage, a reliable 1 ° -accurate adjustment of the rotor position of the motor is possible.

00014] In a further embodiment, when a second predetermined

Speed threshold by the speed of the motor, which is greater than the first speed threshold, switched from the timed commutation to the block commutation. This ensures that the control of the rotor position can be carried out to the degree possible as long as possible.

00015] In a further development, the switchover from the time-controlled commutation to the

Block commutation when one phase of the motor is de-energized. This takes into account that the ratio of motor current to motor torque behaves differently in block commutation than in the case of timed commutation.

00016] The invention allows numerous embodiments. One of them should be based on in the

Drawing illustrated figures are explained in detail.

00017] It shows:

00018] Figure 1: an equivalent circuit diagram of a phase of an electrically commutated motor

00019] FIG. 2: a block diagram of the hardware connection of a rotor position sensor

00020] FIG. 3: a representation of the realization of a time-dependent commutation and a block commutation

00021] Figure 4: a representation of the generation of a process interrupt in a block commutation Figure 5: a representation of the terminal voltages in sinusoidal commutation

(Solid lines L1 to L3 and the terminal voltages in block commutation (colored lines 1 to 3), where the printed in bold, dashed vertical lines represent favorable switching times (since there is a zero crossing of one of the phases).

Figure 1 shows an equivalent circuit diagram for a phase of an electrically commutated motor 1, which is designed as a brushless DC motor (BLDC motor). Such an engine is used, for example, for driving clutches in motor vehicles. It is a three-phase motor, which is operated with a three-phase current, which is caused by three alternating voltages 120 ° out of phase and produces a rotating field. The motor has three three-phase windings, with one of the three alternating voltages applied to each winding. The connection of the winding is referred to as phase. Each winding represents an impedance that is characterized by the L-phase (inductive reactance) and an R-phase (real resistance).

At high speeds, the motor current I Mot of the motor voltage U_Gen rushes due to the

Motor inductance L_Phase after. To determine the position of the rotor of the motor to at least a degree exactly, the motor current I Mot must match exactly to the rotor position, that is, the motor current I Mot must be in phase with the motor voltage U Gen. To realize this, the angle at which the voltage U Vor must precede the motor voltage U Gen would have to be determined via a vector operation. Since such a procedure is very time-consuming and requires a high computing power of an evaluation unit, a method is proposed which dispenses with the tedious calculation processes and nevertheless has a setting resolution of the rotor position of the motor 1 of at least 1 °.

In Figure 2 is a block diagram for the connection of a rotor position sensor 2 to a

Evaluation unit 3 shown, which allows such a highly accurate setting resolution. With the rotor position sensor 2, the absolute angle of the rotor is measured. The rotor position sensor 2 is connected via a serial interface 4 and an incremental interface 5 to the evaluation unit 3, which is realized in the present case by a microcontroller.

The combination of the microcontroller 3 with the rotor position sensor 2 allows on the

incremental interface 5 to generate the needed for a commutation 8. 9 interrupts. The interface 5 consists of at least two lines, each line provides a square wave signal and the two square-wave signals are phase-shifted by 90 °. Around To determine the position of the rotor, attention must be paid to the rising or falling edge of the signal, also referred to as increments. Typical are counts of 2048 increments per one revolution of the rotor. This means 1024 rising and 1024 falling edges per revolution.

The decision as to whether the engine 1 should be controlled via a block commutation 8 or a timed commutation 9 for the precise setting resolution of 1 degree is made as a function of the engine speed. Thus, for example, a block commutation 8 is carried out at a speed of greater than 100 rpm, while at a speed of less than or equal to 100 rpm, the time-controlled commutation 9 is used. The reason for switching between timed commutation 9 and block commutation 8 is the speed-dependent impedance of motor 1. Due to the impedance, the inductive reactance increases in proportion to the speed, resulting in the phase offset α between motor voltage and motor current. (N) = arctanj

R

00028]

00029] where N: speed in min -1,

00030] p: pole pair of the motor

00031] Inductance of the phase

00032] R: Phase resistance

00033] FIG. 3 shows that both types of commutation 8, 9 are controlled by the signal

which outputs the incremental interface 5. The time-controlled commutation 9, in which all three phases of the electric motor 1 are controlled by means of a PWM signal, is called in a fixed time grid t. Ideally, this takes place as a function of the period TPWM of the PWM frequency. At each of these times t, t + TPWM. T + 2TPWM, a process interrupt 6 is carried out, in which for the high-precision position control of the rotor, the increments counted in the time frame are read, with their Help the exact position determination of the rotor takes place. The thus determined position of the rotor is visited in a table and read out a corresponding control pattern for energizing all three phases of the motor 1 of the microcontroller 3.

Ό0034] For block commutation 8, the process interrupt 6 depends on the position of the

Rotor generated. In the microprocessor 3, a process interrupt 6 is triggered by means of a position counter which counts the edges of the square-wave signals of the incremental interface 5. The particular determined by evaluation of the edges of the incremental interface 5 process interrupt 6 is always triggered at the position of the rotor, which is electrically characterized by 60 °. For a 7-pole motor 1, 60 ° electrical, which characterizes the rotating field position, corresponds to 360 (7 ^* 6) = 8, 57 ° mechanical, which corresponds to the rotor position. The 6 in the denominator of the equation corresponds to the 6 block steps per 360 ° electrical. The conversion between electrical and mechanical position is necessary to assign the position of the rotating field of the rotor.

Ό0035] For the incremental interface 5, which generates 2048 edges or increments per revolution, every 2048 / (7 * 6) = 48.76 increments a process interrupt 6 must be triggered. That is, per increment, the position of the rotor changes by an average of 8, 57 °. Since no even numbers of increments are obtained in this case, a correction must be carried out, as shown in FIG. Thus, a process interrupt 6 is always triggered after 49 increments, which corresponds to a change in the rotor position per increment of 8.61 °. However, an error of approximately 0.042 ° is made per process interrupt 6, which is summed up. To correct this error, the correction interrupt 7 is triggered after 48 increments.

Lösung0036] By the solution according to the invention, both the high setting accuracy of

time-controlled commutation 9 and the high efficiency of block commutation 8 use. The time-controlled commutation 9 takes place via the readout of suitable control patterns on the basis of the determined rotor position from a stored table. Due to the increasing phase shift in motor current and motor voltage with increasing speed must be switched from a certain speed to block commutation 8, which is permissible because the high setting accuracy of the rotor position is needed only at low speeds. FIG. 5 illustrates a possible embodiment of the switching from event-controlled commutation (block commutation) to time-controlled commutation (sinusoidal commutation), preferably at those points at which one of the phases has a zero crossing.

00037] The commutation can not only depend on a signal value (e.g., speed) of

Block can be switched to sine or vice versa, but also depending on the phase position (x-axis). By choosing a favorable switching time, as shown for example in Figure 5, a gentler switching process (without torque jump, or current jump) can be realized. Due to the different control, a factor should be added to the setpoint voltage when switching the commutation.

Electromotor

Rotor position sensor

evaluation

serial interface

incremental interface

Process Interrupt

Korrekturinterrupt

Block commutation

timed commutation

Claims

claims

1. A method for controlling the rotor position of an electrically commutated motor, in particular a brushless DC motor, in which the phases of the motor (1) are driven by determining the position of the rotor via an event-controlled commutation (8), characterized in that the highly accurate setting of Position of the rotor, the control of the phases of the motor (1) from the event-controlled commutation (8) is switched to a timed commutation (9).

2. The method according to claim 1, characterized in that the time-controlled commutation (9) at fixed predetermined time intervals (t, t + TpwM, t + 2T _PW M) takes place, wherein, depending on the determined position of the rotor of the motor (1) Control values for the phases of the motor (1) are read from a table.

3. The method according to claim 2, characterized in that caused by addition or subtraction of an offset to the currently determined rotor position by the read out of the table drive values for the phases of the motor (1), a field weakening.

4. The method according to at least one of claims 1, 2 or 3, characterized in that the event-controlled commutation (8) by a process interrupt (6) is triggered, which is triggered in response to the position of the rotor, in particular a predetermined angle change.

5. The method according to claim 4, characterized in that a position of the rotor determining sensor (2) detects an absolute angle of the rotor, which is forwarded via an at least two square wave generating incremental interface (5) to an evaluation unit (3) which the flanks of the at least two square-wave signals generated by the incremental interface (5) are counted and the position of the rotor of the motor (1) is determined therefrom, the process interrupt (6) being triggered after a certain number of counted edges.

6. The method of claim 4 and 5, characterized in that after a predetermined number of process interrupts (6), a correction interrupt (7), which is triggered after a smaller number of edges than the process interrupt (6).

7. The method according to any one of the preceding claims, characterized in that the switchover between event-controlled commutation (8) and time-controlled commutation (9) in dependence on a rotational speed of the motor (1).

8. The method according to claim 7, characterized in that when falling below a first predetermined speed threshold by the rotational speed of the motor (1) of the event-controlled commutation (8) of the motor (1) is switched to the timed commutation (9).

9. The method according to claim 7 or 8, characterized in that when exceeding a second predetermined speed threshold by the speed of the motor (1), which is greater than the first speed threshold, from the time-controlled commutation (9) to the event-controlled commutation (8). is switched.

10. The method according to claim 9, characterized in that the switching from the time-controlled commutation (9) to the event-controlled commutation (8) takes place when a phase of the motor (1) is de-energized.