WO2006032241A1

WO2006032241A1 - Method and arrangement for compensating errors in displacement or angle sensors

Info

Publication number: WO2006032241A1
Application number: PCT/DE2005/001616
Authority: WO
Inventors: Matthias Moerbe; Axel Wenzler; Ralf Noltemeyer; Cord Hoersten; Oliver Glitza; Klaus Sekot
Original assignee: Robert Bosch Gmbh; Bishop Innovation Limited
Priority date: 2004-09-22
Filing date: 2005-09-15
Publication date: 2006-03-30
Also published as: AU2005287781B2; DE102004045811A1; AU2005287781A1; EP1794548A1

Abstract

The invention relates to a message and arrangement for detecting the movement or angle of rotation of moving mechanical components, wherein code signals are evaluated, said signals being produced by scanning several code tracks (2,3,4,5;2',5') which are disposed in the vicinity of each other perpendicular to the direction of movement on the moving component by means of respectively associated detector lines (6,7,8,9,10) of a stationary sensor. Additional lines of detectors (10) are arranged in such a way that at least one of the detector lines (6,7,8,9,10) can be allocated, with maximum tolerance (AT) of the moving component, perpendicular to the direction of movement of each code track (2, 3, 4, 5; 2', 5' ).

Description

Method and arrangement for compensation of errors in

Displacement or angle sensors

State of the art

The invention relates to a method and an arrangement for compensating errors in displacement or angle sensors, in particular when detecting a rotation angle on axes or shafts, according to the preamble of the main claim.

For example, to detect the torque acting on a steering wheel axis of a motor vehicle during the rotation of the steering wheel, very small angle changes must be measured in both directions of rotation of the steering wheel. Incremental angle sensors are applicable which evaluate an angular position on the basis of optical, magnetic or otherwise generated by the rotation and detected by suitable means pulses. In order to increase the measurement accuracy and to increase the uniqueness range of the angle measurement, usually a multiple

number of such incremental, usually periodically occurring, measured values, so that here several phase measured values occur, from which the quantity to be measured, such as, e.g. the angle of rotation, an angular difference or the distance to a target, is to be determined.

For example, in WO 00/28285 an optical sensor for detecting the position of movable surfaces is described, which detects an optical pattern with high- and low-rigreflektierenden sections. In WO 00/36377, a signal processing method is based on this, which is based on assigning specific detector rows to the optically detectable code or pattern in such a way that for each high and low-reflectance sections of the code track is exactly one detector row is provided.

In order to evaluate such phase or angle measured values, in the case of more than two phase signals, for example, a method described in DE 101 42 449 A1 is proposed. In this case, signals for determining distances with a displacement sensor or rotational angles with an angle sensor can be obtained from one or more detector rows. The known method has the effect that a unique phase or angle measurement value for position determination of the code-carrying object is generated from N ambiguous phase signals. Here the measured phase values are computationally converted by means of a linear transformation method and evaluated with a predetermined weighting.

Thus, in this prior art, a method is described which generates from N ambiguous, possibly also disturbed phase measured values, a highly accurate, robust and definite phase measured value. The procedure is applicable For example, in an optical angle sensor, in which N pa¬ rallele tracks are applied to a cylinder. On each of the N tracks (i = l ... N) are ni periods of phase information, which are represented, for example, in the optical case by ni periods of light-dark transitions, the. Other sensor principles, eg magnetic or capacitive, are also possible here. The traces of the sensor can also be applied on a plane instead of on a cylinder, for example in the case of a displacement sensor.

The known devices or methods depend on the fact that the code can be detected and evaluated at any time by the sensor via the information-carrying sections of the surface of the object whose position (path or angle) is to be detected. However, this is no longer the case if e.g. Due to mechanical tolerances of the underlying system, a relative deposition of sensor and code occurs. For example, in the case of an optoelectronic angle or torque sensor, this is the case when the cylindrical code disks mounted on a rotary shaft move so far in the axial direction that they do not move with the aid of the lens of the optoelectronic angular or rotational sensor can be imaged more on the associated Detektorzei¬ len of the detector.

Advantages of the invention

With the aforementioned generic method for detecting the movement or the angle of rotation of moving mechanical components code signals can be evaluated by scanning several perpendicular to Be¬ wegungsrichtung adjacent code tracks on -A-

the moving component by means of respectively associated Detek¬ torzeilen a stationary sensor are generated. According to the invention, the known method is advantageously developed in such a way that additional detector lines are arranged such that at maximum tolerance of the moving component perpendicular to the direction of movement of each code track at least one detector line of a detector can be switched on.

For example, if the code consists of several tracks on the circumference of a code disc of an angle sensor, e.g. four tracks, the tracks are normally assigned to each one clearly a detector line on the stationary detector. This allows a shift of the L / 2 code imaged onto the detector in the axial direction, where L is the width of a code track in the image area when L is large compared to the extension of the detector row in the axial direction. A shift beyond this level leads in the prior art to a failure of the sensor principle. An unambiguous assignment of the code track to a detector line here means that a code track of only exactly one detector line in each case is read correctly, i. can be evaluated.

According to the invention, by adding further detector rows and a suitable interconnection of the detector rows or by adapting the signal evaluation of the detector rows, it is now achieved that a code track can also be evaluated by one or more adjacent detector rows. It is thus advantageously possible with an appropriate interconnection of the detector signals to achieve an ambiguous assignment and a displacement of the code and thus of the measurement object beyond the value L / 2. For carrying out the method according to the invention, as described above, it must be ensured that the code tracks are readable at all times by detector lines, whereby according to the invention it is avoided that mechanical measures can be dispensed with, eg that a displacement of a rotating measuring object in the axial direction Direction is limited by elaborate and therefore expensive mechanical bearings. Another complex solution, which consists in increasing the code width and the resulting distance between the detector rows, can be avoided with the invention. Thus, both an enlargement of the code region of the measurement object and a magnification of the detector surface can be dispensed with, which is undesirable or would entail higher costs for the sensor.

The solution according to the invention is essentially based on making the detector more tolerant of displacements in the axial direction by virtue of the fact that an electronic solution becomes possible with a suitable arrangement of detector rows and an electronic evaluation concept adapted thereto. This is usually cheaper than a mechanical solution or a code track and / or detector magnification.

According to an advantageous embodiment, it is proposed that the detector rows lie next to one another at a distance smaller than the code track width and that the inner detector lines can detect one of two adjacent code tracks, depending on the position within the tolerance, and respectively in the Extreme positions of tolerance an outer detector line is adjacent to a code track.

In a first variant of the proposed embodiment, the distance of the detector lines is smaller by a factor than the code track width L, which results from the number of code tracks divided by the number of Detek¬ torzeilen is formed. According to a second embodiment, the distance between the detector lines is smaller than the code track width L by the width of the detector line perpendicular to the direction of movement, that is to say in the case of an angle sensor in the axial direction.

According to a further embodiment, it is also possible that an auxiliary code track is attached to at least one of the outer code tracks, which in the extreme positions of the tolerance of the measurement object can also be detected by the outer detector line of the detector.

In a sensor arrangement according to the invention for carrying out the method described above, the code tracks are arranged on code disks of a rotating shaft in an advantageous manner for detecting the angle of rotation and / or the torque of the rotating shaft. In this case, the code tracks are preferably formed by light-dark markings on the code disks and the detector lines consist of optoelectronic sensor elements.

drawing

Embodiments of a sensor arrangement for carrying out the method according to the invention will be explained with reference to the drawing. Show it:

FIG. 1 shows a basic arrangement of four code traces and four associated detector rows of an angle sensor according to the prior art, FIG. 2 shows the geometric relationships of the arrangement according to FIG. 1 in extreme positions, with a shift between the code tracks and the detector rows within the tolerance,

FIG. 3 shows an arrangement according to the invention with five detector lines at four code tracks,

FIG. 4 shows the geometric relationships of the assignment of detector lines and code tracks according to FIG. 3,

FIG. 5 shows a further exemplary embodiment with five detector rows,

FIG. 6 shows an exemplary embodiment with additional auxiliary code tracks and

FIG. 7 shows an exemplary embodiment with a doubled doubling of the detector lines in the middle region of the code tracks.

Description of the embodiments

FIG. 1 shows, in a schematic view, a developed code disk 1 with four code tracks 2, 3, 4, 5 of the width L. The code tracks 2 to 5 are hereby assigned to detector rows 6, 7, 8, 9 with a spacing P in the usual manner. With this arrangement known per se, it is possible, for example, to determine the angle of rotation from a movement of the code tracks 2 to 5 along the detector lines 6 to 9. If, due to the mechanical axial tolerance of the arrangement, ie a displacement in the y direction perpendicular to the direction of movement of the code disk 1, both the code tracks 2 to 5 and the detector lines 6 to 9 are displaced relative to one another, possibly also in opposite directions , it must be ensured that the code is currently legible, ie the code tracks 2 to 5 can be scanned, for example, optically from the associated detector line 6 to 9.

The sizes P used herein was the Detektorzeilenab-, the code width L and the width of the detector line PD in the y direction are also shows in Figure 2 can be seen, DJ_ _e the geometric conditions of the assembly according to the Fi¬ gur 1 in two extreme positions, wherein the right position shows a shift between the code tracks 2 to 5 and the detector lines 6 to 9 by an axial tolerance AT. Here, a conventional detector layout is thus exemplified for four code tracks with shifted in the y direction code track images in the "lower" (left) and "upper" (right) extreme positions. The digits on the right edge of the detector lines 6 to 10 indicate which code tracks 2 to 5 of these detector lines can be evaluated.

The quantities L, P and AT are linked as follows: P = f (L) and AT = g (P), where f and g symbolize functions; The quantities L and P in each case represent the area requirement of the code area on the measuring object, e.g. here a rotating shaft, and S = h (P), with h as a function, the area required by the detector. In particular, according to the invention, the quantities L and S can now be minimized for a given value of AT.

In the arrangement according to FIG. 2, the following equations apply: P = L,

AT = P PD,

S = 3P + PD.

From FIG. 3 and FIG. 4, it can be seen that by adding another detector row 10 with as many as four code tracks, the axial tolerance ΔT can be increased if the rows 6 to 10 of the detector are modified such that they now have several, e.g. read two code tracks, i. can evaluate. The detector row 7 in FIG. 4 can evaluate the code tracks 2 + 3, the detector row 8 can evaluate the code tracks 3 + 4 and the detector row 9 can evaluate the code tracks 4 + 5. From a so-called vernier arrangement according to FIG. 3 with e.g. five locally uniformly distributed detector lines 6 to 10 results to scan four code tracks 2 to 5 then a layout according to the figure 4. When shifting the code tracks it comes with use of this arrangement to equally distributed switching operations. This is quite desirable since this minimizes the influence of the switching on the resulting measured value.

The equations apply here:

P = 4/5 L,

AT = 3/2 P PD, S = 4P + PD.

In an arrangement with seven detector rows, not shown here, the following equations would result:

P = 4/7 L,

AT = 5/2 P PD, S = 6P + PD. Instead of the secondary condition P = 4/5 L according to FIG. 4, P + PD = L can also be required, so that according to FIG. 5, a maximum distance of the detector rows and the greatest possible space gain result. The switching of the detector lines 6 to 10, however, is no longer uniformly distributed when using this arrangement, but at the same time.

The equations apply here:

P = L -PD, AT = 2P -3PD. S = 4P + PD.

It can be seen from FIG. 4 that the maximum shift in the code tracks is always achieved when either the code track 2 or 5 moves out of the detector field and thus can no longer be evaluated. This leads to an exemplary embodiment according to FIG. 6, in which the code is continued by applying to the code track 2 an auxiliary code track 2 '(left-hand representation) and to the code track 5 an auxiliary code track 5' (right-hand representation). Then, the maximum shift is achieved when the code track 2 'touches the detector row 6 or when the code track 10 touches the detector row 5', e.g. It is required that so-called dual lines be used, i. a detector row can process a maximum of two adjacent code tracks. The attached code segments as auxiliary code tracks 2 'and 5' are therefore shorter than the main code tracks 2 to 5. A further feature here is that detector lines 6 and 10 now also have to be designed as dual lines.

The equations apply here:

P = 4/5 L, AT = 2P - PD, S = 4P + PD.

FIG. 7 also shows an exemplary embodiment in which the three inner detector rows 7 to 9 according to FIGS. 4 or 6 are shown as dual lines, ie. These detector lines are each divided into two separate, immediately adjacent detector rows.

Here then the equations apply:

P = L-PD, AT = 2 PD, S = 4P + 4PD.

Claims

claims

1) Method for detecting the movement or Drehwin¬ angle on moving mechanical components, in which

Code signals are evaluated by scanning a plurality of code tracks (2, 3, 4, 5, 2 ', 5') perpendicular to the direction of movement on the moved component by means of respectively associated detector lines (6, 7, 8, 9, 10) of a stationary sensor, characterized in that

- Additional detector rows (10) are arranged such that at maximum tolerance (AT) of the moving component perpendicular to the direction of each code ^¬ track (2, 3, 4, 5, 2 ', 5') len at least one of the detector (6 , 7, 8, 9, 10).

2) Method according to claim 1, characterized in that

- The detector lines (6,7,8,9,10) in a smaller distance Ab (P) than the code track width (L) side by side

Depending on the position within the tolerance (AT), the inner detector rows (6, 7, 8, 9, 10) can each detect one of two adjacent code tracks (2, 3, 4, 5) and respectively in the extreme positions of the tolerance lie an outer detector line (6, 10) next to a code track (2, 5).

3) Method according to claim 2, characterized in that

- The distance (P) of the detector lines (6,7,8,9,10) by ei¬ nen factor is smaller than the code track width (L), the number of code tracks (2,3,4,5) divided by the number of detector lines (2,3,4,5) is formed.

4) Method according to claim 2, characterized in that

- The distance (P) of the detector lines (6,7,8,9,10) by the width (PD) of the detector line (6,7,8,9,10) perpendicular to the direction of movement is smaller than the Codespur¬ width (L ).

5) Method according to one of the preceding claims, characterized in that

- An auxiliary code track (2 ', 5') is attached to at least one of the outer code tracks (2,5), which in each case in the extreme positions of the tolerance (AT) of the äuße¬ ren detector line (6,10) is detectable. 6) Sensor arrangement for carrying out a method according to one of the preceding claims, characterized gekennzeich¬ net that

- The code tracks (2, 3, 4, 5, 2 ', 5') are arranged on code discs of a rotating shaft for detecting the angle of rotation and / or the torque of the rotating shaft.

7) Sensor arrangement according to claim 6, characterized gekennzeich¬ net that

- The code tracks (2, 3, 4, 5, 2 ', 5') are formed by light-dark markings on the code discs and the detector lines consist of optoelectronic Sensorele¬ elements.