DE102014226604A1 - Method and device for compensating an arrangement tolerance between two sensor elements of a position sensor arrangement - Google Patents

Abstract

The invention relates to a method for determining a corrected positional angle (φ _korr ) of a shaft (3) by correcting an arrangement tolerance between sensor elements (22, 23) of a sensor arrangement (2), wherein the sensor elements (22, 23) provide a first and a second phase-shifted by 90 ° sensor signal (S1, S2) are arranged, comprising the following steps:
Detecting the sinusoidal sensor signals (S1, S2) of the sensor elements (22, 23) during a rotation of the shaft (3),
- determining a correction quantity (Δφ) dependent on the presence of a deviation of the phase shift of the phase-shifted sensor signals (S1, S2) from a phase shift of 90 °;
- Correcting at least the first sensor signal (S1) as a function of the correction variable (Δφ), so that at least one corrected first sensor signal (S1 _corr ) is formed, with which the corrected attitude angle (φ _corr ) can be determined.

Description

Technical area

The invention relates to position sensors for detecting a positional angle of a shaft, in particular for using a detection of a positional angle of a rotor in an electrical machine. Furthermore, the present invention relates to measures for correcting assembly-related measurement tolerances in sensor arrangements.

Technical background

For detecting a position angle of a shaft, in particular a rotor shaft of an electric machine, position sensor arrangements are provided. Frequently, a position sensor arrangement with magnetic field sensors is provided as sensor elements which detect a wave-proof magnetic field. Normally, the sensor elements are arranged on the shaft in such a way that mutually phase-shifted sinusoidal sensor signals are provided when the shaft rotates. It is often provided to arrange the sensor elements such that they provide sensor signals with a phase offset of exactly 90 ° when the rotor shaft rotates.

In reality, the phase offset between the sensor signals provided by the sensor elements is not exactly a predetermined angle of z. B. 90 °, but it comes due to arrangement tolerances to corresponding deviations. If the position angle of the shaft is derived from the sinusoidal sensor signals of the sensor elements, an error in the position detection can occur due to these arrangement tolerances.

In order to compensate for location detection errors caused by arrangement tolerances, a calibration usually takes place at the end of production. During the calibration, a suitable correction variable is determined in order to determine the arrangement tolerances by presetting for the application of a positional angle detected by the position sensor and to store it in a suitable manner in a unit which processes an indication of the positional angle. This represents a further manufacturing step, whereby the entire manufacturing process becomes more complex.

It is an object of the present invention to provide a simplified possibility, whereby a measurement of a position angle by a position sensor, which is inaccurate due to arrangement tolerances, can be corrected.

Disclosure of the invention

This object is achieved by the method for determining a corrected attitude angle of a wave by correcting an arrangement tolerance between sensor elements of a sensor arrangement according to claim 1 and by the apparatus and the sensor system according to the independent claims.

Further embodiments are specified in the dependent claims.

• – Erfassen von sinusförmigen Sensorsignalen der Sensorelemente bei einer Drehung der Welle,
• – Bestimmen einer Korrekturgröße abhängig von dem Vorliegen einer Abweichung der Phasenverschiebung der phasenverschobenen Sensorsignale von einer Phasenverschiebung von 90°;
• – Korrigieren mindestens des ersten Sensorsignals abhängig von der Korrekturgröße, so dass ein korrigiertes erstes bzw. zweites Sensorsignal gebildet wird, mit dem der korrigierte Lagewinkel bestimmbar ist bzw. bestimmt wird.

According to a first aspect, a method is provided for determining a corrected position angle of a shaft by correcting an arrangement tolerance between sensor elements of a sensor arrangement, wherein the sensor elements are arranged to provide a first and a second sensor signal phase-shifted by 90 °. The method comprises the following steps:

• Detecting sinusoidal sensor signals of the sensor elements during rotation of the shaft,
• Determining a correction quantity dependent on the presence of a deviation of the phase shift of the phase-shifted sensor signals from a phase shift of 90 °;
• Correction of at least the first sensor signal as a function of the correction variable, so that a corrected first or second sensor signal is formed with which the corrected position angle can be determined or determined.

One idea of the above method is to check the resulting curves of the sensor signals of the sensor elements for their phase offset during rotation of the shaft. The sensor signals should ideally have a phase offset of 90 °.

The correction quantity is determined as a function of the amount of deviation of the phase offset from a phase offset of 90 ° and at least one of the sensor signals is corrected as a function of the determined correction quantity. This allows a permanent compensation of a position angle erroneously determined from the sensor signals due to arrangement tolerances, so that also changing arrangement tolerances between the sensor elements, which may possibly occur during operation of the sensor system due to vibration or the like, can be compensated.

By the above method, it is possible to adaptively adjust the correction quantity online so that sensor signals having an exact phase shift of 90 ° are always provided. In particular, eliminates the need to end of manufacturing, the assembly tolerances between determine the sensor elements and make a corresponding correction of the resulting phase shift to 90 °.

The dynamic detection and subsequent correction of the error in an online process yields significantly better results than a static adjustment at the end of production. The production time can be reduced. Furthermore, it is possible in dynamic recognition to compensate for effects that can lead to a change in the arrangement tolerances over the life of the sensor system.

Furthermore, the correction can be carried out by means of a control depending on an error value which is determined from the current signal values of the corrected first sensor signal.

It can be provided that the correction is carried out with the aid of a control as a function of an error value which is a deviation of a sum of the squared signal values of the corrected first sensor signal and the second sensor signal from a constant value, in particular a sum of the squared amplitudes of the sensor signals, depends.

The checking can thus be carried out by determining the squared sum of the signal values of the sensor signals. The sum of the squared sensor values of the sensor signals is zero with an exact phase shift of 90 ° between the two sensor signals. However, with a departure from the phase shift of 90 °, the plot of the sum of the squared sensor values of the sensor signals will oscillate at twice the frequency of the sensor signals and at an amplitude that is the magnitude of the difference between the phase offset of the sensor signals and a phase offset of 90 ° depends. As soon as a corresponding oscillation of the sum of the squared sensor values is detected, a correction of one of the two sensor signals with a correction variable can be made.

According to an embodiment, the correction may be performed with the aid of the regulation as a function of the error value, the error value being an amount of deviation of the sum of the squared signal values of the corrected first sensor signal and the second sensor signal from a constant value, in particular from the sum of the squared amplitudes Sensor signals corresponds.

Furthermore, the correction variable can be determined by integrating the error value applied with a sign value.

The sign value can be determined iteratively by inverting the previously valid sign value if the magnitude of the correction quantity exceeds a predetermined threshold.

It can be provided that the sign value is determined as a function of the at least one corrected first sensor signal and the second sensor signal by determining a direction of rotation of the shaft from the sensor signals and a sign of the deviation of the phase shift from the phase shift of 90 ° between the at least one corrected first sensor signal and the second sensor signal is determined, wherein the sign value from the direction of rotation and the sign of the deviation is determined.

According to an embodiment, the correction quantity may be provided as a correction angle Δφ indicating the deviation of the phase shift from a phase shift of 90 °, the correction angle Δφ for correcting the first sensor signal having the following formula S1 _corr = (S1-sin (Δφ) * S2) * 1 / cos (Δφ) is used to obtain the corrected first sensor signal S1 _corr .

• – um sinusförmige Sensorsignale der Sensorelemente bei einer Drehung der Welle zu erfassen,
• – um eine Korrekturgröße abhängig von dem Vorliegen einer Abweichung der Phasenverschiebung der phasenverschobenen Sensorsignale von einer Phasenverschiebung von 90° zu bestimmen;
• – um mindestens eines des ersten und zweiten Sensorsignals abhängig von der Korrekturgröße zu korrigieren, so dass mindestens das korrigierte erste Sensorsignal gebildet wird, mit dem der korrigierte Lagewinkel bestimmbar ist.

According to a further aspect, an evaluation unit is provided for determining a corrected positional angle of a shaft by correcting an arrangement tolerance between sensor elements of a sensor arrangement, wherein the sensor elements are arranged to provide a first and a second sensor signal phase-shifted by 90 °. The evaluation unit is designed:

• To detect sinusoidal sensor signals of the sensor elements upon rotation of the shaft,
• To determine a correction quantity dependent on the presence of a deviation of the phase shift of the phase-shifted sensor signals from a phase shift of 90 °;
• To correct at least one of the first and second sensor signals as a function of the correction variable, so that at least the corrected first sensor signal is formed with which the corrected position angle can be determined.

• – eine Sensoranordnung mit Sensorelementen zum Bereitstellen eines ersten und eines zweiten Sensorsignals und einem Geberrad, an dem die Sensorelemente angeordnet sind;
• – die obige Auswerteeinheit.

According to a further aspect, a sensor system is provided, comprising:

• A sensor arrangement with sensor elements for providing a first and a second sensor signal and a transmitter wheel on which the sensor elements are arranged;
• - the above evaluation unit.

Brief description of the drawings

Embodiments are explained below with reference to the accompanying drawings. Show it:

1 a sensor system having a sensor arrangement with two staggered sensor elements for detecting a dependent of a positional angle of a wave magnetic field;

2 a functional diagram for illustrating a function for correcting a determined by means of the sensor elements attitude angle;

3 a functional diagram for illustrating a further function for correcting a determined by means of the sensor elements attitude angle; and

4a and 4b Traces of the sensor signals and an error value and a correction angle for a positive and negative arrangement tolerance of the sensor elements.

Description of embodiments

1 shows a sensor system 1 with an exemplary sensor arrangement 2 , The sensor arrangement 2 is on a wave 3 , In particular, a rotor shaft of an electric machine arranged, which rotates in operation.

The wave 3 is with a donor wheel 21 the sensor arrangement 2 firmly connected. They are a first and a second sensor element 22 . 23 each provided by means of a magnetic field detection unit 26 can detect a strength of a magnetic field. The sensor elements 22 . 23 are preferably formed similar. When turning the sender wheel 21 is on the sensor elements 22 . 23 depending on the position of the sender wheel 21 dependent magnetic field on the sensor elements 22 . 23 exercised.

To provide the magnetic field, the sender wheel 21 be magnetized, so this is an even number of poles 24 having circumferentially adjacent poles 24 polarized opposite to each other. In the illustrated embodiment, the sender wheel 21 two poles 24 on. Alternatively, the sender wheel 21 also be constructed of a structured soft magnetic material and the sensor elements 22 . 23 between a magnetic field detection unit 26 and the donor wheel 21 be provided with corresponding (not shown) permanent magnet, so that the magnetic flux through the magnetic detection unit 26 depending on the position of the sender wheel 21 varied.

The first and the second sensor element 22 . 23 are designed to provide an electrical first or second sensor signal S1, S2 dependent on the strength of the applied magnetic field. Furthermore, the sensor elements 22 . 23 and the donor wheel 21 so formed and arranged to each other that upon rotation of the shaft 3 the donor wheel 21 rotates while sinusoidal sensor signals S1, S2 through the sensor elements 22 . 23 generated and deployed.

Furthermore, the sensor units 22 . 23 so on the donor wheel 21 arranged that the sinusoidal sensor signals have a phase offset of about 90 °. This can be achieved in the present case by the sensor elements 22 . 23 an arrangement angle of 90 ° (optimum arrangement angle) to each other, that is, the direction of the maximum sensitivity of the magnetic field detection units 26 is radial to the encoder wheel 21 aligned and includes an angle of 90 °. Depending on the number of poles 24 the donor wheel 21 can the optimal arrangement angle of the sensor elements 22 . 23 also differ and may be general 360 ° / 2 · N where N is the number of poles 24 the donor wheel 21 equivalent. For example, the arrangement angle of the sensor elements 22 . 23 45 ° correspond when the sender wheel 21 four poles 24 having.

The sensor signals S1, S2 become an evaluation unit 5 supplied, which provides a corrected attitude angle φ _corr . The evaluation unit 5 may be part of a control unit that performs other tasks based on the corrected attitude angle φ _corr , such as driving an electric machine whose rotor is connected to the shaft 3 connected is. The evaluation unit 5 samples the sensor signals S1, S2 in successive measuring cycles and respectively sets current signal values for processing and in particular for determining the corrected positional angle of the shaft 3 to disposal.

2 illustrates the function of the evaluation unit 5 based on a functional diagram. The function of the evaluation unit 5 includes an adjustment function 51 , The one of the sensor signals, in this embodiment, the first sensor signal S1, depending on a determined deviation from the optimum arrangement angle of the sensor elements 22 . 23 corrected. The determined arrangement deviation is indicated as a correction angle Δφ (correction variable) of the phase shift between the sensor signals S1, S1 to the phase shift of 90 °

The corrected first sensor signal S1 _korr and the second sensor signal S2 become a position angle detection function 52 supplied based on the corrected first sensor signal S1 _corr and the second sensor signal S2, whose curves now have a phase shift of exactly 90 ° due to the correction, the corrected position angle φ _corr determined. The attitude angle determining function calculates particular φ _corr following the corrected position angle: φ _corr = arctane (S1 / S2)

The correction function 51 determined based on the arrangement deviation, with which the sensor elements from each other from the optimal arrangement on the sender wheel 21 deviate, the corrected first sensor signal S1 _korr as follows: S1 _corr = (S1 - sin (Δφ) · S2) · 1 / cos (Δφ)

By the arrangement of the correction function 51 and the attitude angle determination function 52 In the manner described above, two sensor signals S1, S2 with a phase shift deviating from a phase shift of 90 ° can be corrected. The correction can be made with the aid of a correction angle Δφ, which indicates the deviation from the optimum phase shift of 90 °. It is (the first sensor signal S1, for. Example) corrects one of the sensor signals corresponding to the correction angle Δφ _korr and from the corrected first sensor signal S1 and the second sensor signal S1, the corrected positional angle φ _corr determined.

The correction angle Δφ becomes in a correction angle detection function 53 certainly. The correction angle detection function 53 includes an error value function 54 of which the corrected first sensor signal S1 _corr and the second sensor signal S2 are supplied.

In the error value function 54 For example, the signal values of the corrected first sensor signal S1 _corr and the second sensor signal S2 are squared and summed and then normalized to an error value FW as follows FW = | S1 2 / corr + S2 ² - (amp (S1 _corr ) ² + amp (S2) ² ) | where amp () corresponds to an amplitude function of the sensor signals indicative of the amplitude of the sinusoidal sensor signals S1, S2. The normalization can also be done by division by the term amp (S1 _corr ) ² + amp (S2) ² . If the sensor signals S1, S2 are already normalized beforehand, their amplitudes are each 1, and the term amp (S1 _corr ) ² + amp (S2) ² can be replaced by 1.

A positive error value FW is obtained, which deviates from zero as soon as the phase shift between the corrected first sensor signal S1 _korr and the second sensor signal S2 deviates from 90 °. The error value FW can be a multiplication element 56 are fed, in which the error value FW is provided with a sign, that is multiplied by a sign value.

The product of the sign value V and the error value FW becomes an integrator 55 fed. The output of the integrator 55 corresponds to the correction angle Δφ of the adjustment function 51 provided.

The sign value V can be either +1 or -1 and is a sign function 57 provided. The sign function 57 analyzes the determined correction angle Δφ and detects when it permanently changes in one direction due to the provided error value FW and its magnitude | Δφ | exceeds a predetermined threshold. If the predetermined threshold value, which may correspond, for example, 10 °, exceeded, it is assumed that the correction angle Δφ by the integrator 55 was changed with a wrong sign and the sign of the sign value V is reversed, ie the previous sign value multiplied by -1.

In this way, by probing the sign, the direction of the arrangement deviation between the sensor elements 22 . 23 can be detected and in another direction by changing the Δφ first in one direction, and if this does not lead to a suitable correction of the arrangement tolerance. That the direction of change of the correction angle Δφ does not converge can be detected by an increasing error value FW, which leads to an accelerated increase in the amount of the correction angle as the integrator value.

In 3 a further embodiment is shown in which the correction angle Δφ by a functional determination of the sign value V in a modified sign function 57 ' is determined. Ultimately, the direction of the arrangement deviation from the optimal arrangement of the sensor elements 22 . 23 That is, the direction of the resulting phase shift deviating from the phase shift of 90 ° are detected on the basis of the profiles of the sensor signals S1, S2.

In particular, by the signs of the gradients of the corrected first sensor signal S1 _corr and the second sensor signal S2 at successive zero crossings, the direction of rotation D of the shaft 3 be derived. In particular, given a first sign of the gradient of the corrected first sensor signal S1 _corr in a zero crossing, the sign of the gradient of the second sensor signal S2 indicates the direction of rotation in the case of an immediately following zero crossing.

Furthermore, the sign of the second sensor signal S2 at the time of the zero crossing of the first corrected sensor signal S1 _korr can indicate the direction of the phase shift. Thus, the sign of the second sensor signal S2 at the time of zero crossing of the first corrected sensor signal S1 _korr in conjunction with the rotational direction may indicate the sign value V applied to the error value FW to achieve convergence of the integrator value.

In 4a are first and second sensor signals S1, S2, which have a phase shift of 95 ° shown. This results in a curve of the error values FW, which has a double frequency oscillation of the sensor signals S1, S2. Furthermore, the correction angle Δφ is shown, which converges very fast.

In 4b are the sensor signals S1, S2 and the resulting course of the error values at a phase shift of 85 ° between the sensor signals S1, S2 shown. One recognizes the resulting correction angle Δφ, which is now correspondingly positive in order to compensate for the missing phase shift.

LIST OF REFERENCE NUMBERS

1

sensor system

2

sensor arrangement

3

wave

21

sensor wheel

22, 23

sensor elements

24

Pole

26

Magnetic field detection unit

5

evaluation

51

Customizer

52

Position angle detection function

53

Correction angle discovery feature

54

Error value function

55

integrator

56

multiplication element

57

sign function

57 '

modified sign function

S1, S2

sensor signals

φ _corr

corrected attitude angle

Δφ

correction angle

Claims (10)

Method for determining a corrected positional angle (φ _corr ) of a shaft ( 3 ) by correcting an arrangement tolerance between sensor elements ( 22 . 23 ) a sensor arrangement ( 2 ), wherein the sensor elements ( 22 . 23 ) are provided for providing a first and a second phase-shifted by 90 ° sensor signal (S1, S2), comprising the following steps: - detecting the sinusoidal sensor signals (S1, S2) of the sensor elements ( 22 . 23 ) upon rotation of the shaft ( 3 ), - determining a correction quantity (Δφ) dependent on the presence of a deviation of the phase shift of the phase-shifted sensor signals (S1, S2) from a phase shift of 90 °; - Correcting at least the first sensor signal (S1) as a function of the correction variable (Δφ), so that at least one corrected first sensor signal (S1 _corr ) is formed, with which the corrected attitude angle (φ _corr ) can be determined. The method of claim 1, wherein the correcting is performed by means of a control depending on an error value determined from the current signal values of the corrected first sensor signal. The method of claim 2, wherein the correction is performed by means of a control dependent on an error value, which is a deviation of a sum of the squared signal values of the at least one corrected first sensor signal (S1 _korr ) and the second sensor signal (S2) from a constant value, in particular a sum of the squared amplitudes of the sensor signals (S1, S2) depends. The method of claim 3, wherein the correction is performed by the control depending on the error value, the error value being an amount of deviation of the sum of the squared signal values of the corrected first sensor signal (S1 _corr ) and the second sensor signal (S2) from a constant value , in particular the sum of the squared amplitudes of the sensor signals corresponds. Method according to one of Claims 2 to 4, wherein the correction quantity (Δφ) is determined by integrating the error value (FW) applied with a sign value (V). The method of claim 5, wherein the sign value (V) is iteratively detected by inverting the sign value (V) when the amount of the correction quantity (Δφ) exceeds a predetermined threshold. The method according to claim 5, wherein the sign value is determined as a function of the at least one corrected first sensor signal (S _{1 corr} ) and the second sensor signal ( _{S 2} ), by a direction of rotation of the shaft ( 3 ) is determined from the sensor signals (S1, S2) and a sign of the deviation of the phase shift from the phase shift of 90 ° between the corrected first sensor signal (S1 _korr ) and the second sensor signal (S2) is determined, wherein the sign value (V) from of the Direction of rotation and the sign of the deviation is determined. Method according to one of claims 1 to 7, wherein the correction amount is provided as a correction angle (Δφ) indicating the deviation of the phase shift of a phase shift of 90 °, wherein the correction angle (Δφ) for correcting at least the first sensor signal (S1) with following formula S1 _corr = (S1-sin (Δφ) * S2) * 1 / cos (Δφ) is used to obtain the corrected first sensor signal (S1 _corr ). Evaluation unit ( 5 ) for determining a corrected attitude angle (φ _corr ) of a wave ( 3 ) by correcting an arrangement tolerance between sensor elements ( 22 . 23 ) a sensor arrangement ( 2 ), wherein the sensor elements ( 22 . 23 ) are provided for providing a first and a second phase-shifted by 90 ° sensor signal (S1, S2), wherein the evaluation unit ( 5 ) is formed: - by sinusoidal sensor signals of the sensor elements ( 22 . 23 ) upon rotation of the shaft ( 3 ) - to determine a correction quantity (Δφ) dependent on the presence of a deviation of the phase shift of the phase-shifted sensor signals from a phase shift of 90 °; To correct at least one of the first and second sensor signals (S1, S2) as a function of the correction variable (Δφ), so that the corrected first sensor signal (S1 _corr ) is formed with which the corrected position angle (φ _corr ) can be determined. Sensor system ( 1 ) comprising: a sensor arrangement ( 2 ) with sensor elements ( 22 . 23 ) for providing a first and a second sensor signal (S1, S2) and a transmitter wheel ( 21 ), on which the sensor elements ( 22 . 23 ) are arranged; - an evaluation unit ( 5 ) according to claim 9.

Images