DE10260862A1 - Correction of angle or distance measurements of a sensor system by derivation of one or more correction constants for angle, amplitude and or phase errors of sinusoidal and cosinusoidal measurement signals - Google Patents

Abstract

Method for correcting an angle and or distance measuring sensor arrangement in which the sine and cosine output signals (S1,i, S2,i) of sensors, which are obtained by sampling a measurement object, are evaluated. The angle, amplitude or phase errors of the measurement signal are corrected in that correction constants are estimated from a multiplicity of measurements. The invention also relates to a corresponding circuit arrangement for method implementation.

Description

State of technology

The invention relates to a method and a circuit arrangement for correcting an angle and / or distance measuring sensor system according to the preamble of the main claim.

They are sensor systems in themselves for one angle to be measured with a rotating measuring object or a distance to be measured for a linearly moving measurement object known in which the information to be obtained by a couple of sine and cosine Measurement signals is represented. The information is usually in the amplitude and / or in the phase of these measurement signals. Here occur in the measurement signals often angular or phase errors due to manufacturing tolerances or other circuit-specific features in the sensor arrangement are conditional.

For example, from the DE 41 02 655 A1 an inductive distance sensor is known in which a reference sensor is arranged for the correction of measurement errors caused by circuit technology, the reference signal of which is used for error compensation, in particular due to an insufficiently constant voltage supply to the measurement circuit arrangement.

Advantages of invention

With the generic method mentioned at the beginning for correcting an angle and / or distance measuring sensor system, with the sine and cosine Measurement signals are evaluated by scanning a moving Measurement object and thereby angle or phase errors the measurement signals are corrected in an advantageous manner further developed that from a plurality of measurement signals at least a constant for estimation the angle or phase error and / or the amplitude of the measurement signals is derived.

Conventional non-contact angle measuring sensors are for example on a so-called. AMR or GMR basis (AMR = Anisotrope Magneto Resistance or GMR = Giant Magneto Resistance constructed; distance measuring systems are preferably constructed on a lidar or radar basis. With the AMR and GMR sensors the angle of a magnetic field is measured relative to the sensor surface. This makes it easy to implement non-contact angle sensors If the magnetic field angle is denoted by Φ, ideally the following signals are obtained at the sensor output:
for the AMR sensor S ₁ = A * cos (2 * Φ) and S ₂ = A * sin (2 * Φ) or
for the GMR sensor S ₁ = A * cos (Φ) and S ₂ = A * sin (Φ), where A represents the signal amplitude.

With radar or lidar sensors (laser radar), a signal with a frequency f _{o is} emitted and reflected at a target at a distance d. The transmission signal delayed by the time dt = 2 * d / c (c speed of light) can then be observed at the receiver. This results in a phase shift Φ = -2π * f _o * 2 * d / c. If the transmit signal is multiplied by the receive signal and then high-frequency components are removed, a first signal S _{1 is obtained} (S ₁ = A * cos (Φ)). If the transmission signal is rotated through 90 ° by a phase shifter and the received signal is multiplied again and the high-frequency components are removed from the result, a second signal S ₂ (S ₂ = A * sin (Φ)) is obtained analogous to the signals in the previously described GMR sensor.

Here, however, the two signals S ₁ and S ₂ are often not exactly 90 ° out of phase with one another, namely S ₂ = A * sin (Φ + δΦ). The angular error is thus δΦ and can be caused by manufacturing tolerances in the GMR and AMR sensors. In the case of radar and lidar systems, the aforementioned 90 ° phase shifter is often subject to tolerance. If the phase error is not corrected, measurements of the angle or the distance are inaccurate. According to the invention, the phase error is now advantageously converted from a number N of measurement signals S _{1, i} and S _{2, i} , with i = 1 ... N, into one of the measurement systems mentioned above (for example AMR, GMR, radar, lidar) δΦ and optionally the amplitude A are approximately determined using the calculation steps specified in claims 2 to 5. If the phase error δΦ is now known, its effects on the measured variable can be eliminated by calculation.

The background of the method according to the invention is that for ideal measurement signals S _{1, i} = A * cos (Φ _i ) and S _{2, i} = A * sin (Φ _i + δΦ), with the definitions for the sizes specified in claim 2 M _i and T _i , the following relationship holds for an error e: e i = M i + r * T i - k = 0

If the measured values S _{1, i} and S _{2, i are} overlaid by noise and other disturbances, the right-hand equals sign applies only approximately. The constants r and k depend on A and δΦ as follows: r = -2 * sin (δΦ) k = A 2 * cos (δΦ) 2

und berechnet das Minimum von E bezüglich der Konstanten r und k, so ergeben sich die in den zuvor erwähnten Patentansprüchen angegebenen Lösungen für die Konstanten r und k.If one then forms the sum E of the error squares e _i

and calculating the minimum of E with respect to the constants r and k, the solutions for the constants r and k given in the aforementioned patent claims result.

A circuit arrangement according to the invention to carry out of a method described above advantageously assigns to Example at the output of the sensor system to which the analog measurement signals are applied each analog / digital converter for generating digital measurement signals on. Furthermore, a correction device for correcting the digital Measurement signals and a calculation block for calculating the constants r and k from the measurement signals are present, the first constant r to an input of the correction device for calculating the corrected measurement signal is.

Furthermore, another Output of the calculation block, the constant k for the calculation the amplitude of the measurement signals.

drawing

An embodiment of a circuit arrangement to carry out of the method according to the invention is explained using the drawing. Show it:

1 2 shows a block diagram of such a circuit arrangement for correcting the output signals of a sensor system according to the invention,

2 a diagram of the uncorrected sine and cosine output signals of the sensor system according to the 1 and

3 a diagram of the corrected sine and cosine output signals of the sensor system according to the 1 ,

description of the embodiment

In 1 is shown in a block diagram of a circuit arrangement with which the sensor system 1 Sine and cosine signals supplied for angle or distance measurement, for example with an AMR or GMR sensor mentioned in the introduction to the description or with a radar or lidar system, for error correction in analog-digital converters 2 and 3 digitized and further processed. If the measurement signals contain an offset, this should be removed beforehand in a manner known per se.

In 2 are the digitized measured values S ₁ and S ₂ , for example in the case of an AMR sensor as S ₁ = A * cos (2 * Φ) and S ₂ = A * sin (2 * Φ + δΦ) for a number of N measurements, for example N = 400, with the respective measurement signals S _{1, i} and S _{2, i} (i = 1 .... N) in an S ₁ / S ₂ plane, from which it can be seen that an oval contour occurs due to an angular error which sets measurement signals.

The measurement signals S ₁ and S ₂ are now based on the in 1 Circuit arrangement shown a correction device 4 and a calculation block 5 fed. At the output of the calculation block 5 is a constant r as a correction quantity for the correction device 4 , for calculating corrected measurement signals S _{1, i} and Sk _{2, i} , and a constant k for correcting the amplitude A of the measurement signals. The value for r can also be obtained, for example, once as a constant offline from a calibration data set. However, the following describes how the value for r is continuously recalculated online from the last N current measurement signals as a data record.

The constants r and k are determined in accordance with the exemplary embodiment in such a way that first the square sum M _{i of} the measurement signals M i = S 2 1, i + S 2 2, i (1) and the product T i = S 1, i * S 2, i (2) is determined.

D=N*Stq–St2 (7). Sum values Sm, St, Smt and Stq and a determinant D are now determined from these variables M _i and T _i as follows:

D = N * Stq-St 2 (7).

These variables are then used in the calculation block 5 the constants are determined as follows:

In the correction facility 4 then in the exemplary embodiment described here, the respective measurement signals S ₁ and S ₂ , including the constants r and k, each result in a corrected measurement signal according to the relationship

The amplitudes A of the measurement signals S ₁ and S ₂ can then still according to the relationship

In 3 are now analogous to the representation according to the 2 the corrected measurement signals are shown in an S ₁ / Sk ₂ plane, it being evident that the measurement signals corrected in this way form a circle.

Claims (8)

Method for correcting an angle and / or distance measuring sensor arrangement ( 1 ), in which - sinusoidal and cosine-shaped measurement signals (S _{1, i} , S _{2, i} ) are evaluated, which have been obtained by scanning a moving measurement object and in which - angle or phase errors of the measurement signals are corrected, characterized in that - From a plurality (N) of measurement signals (S _{1, i} , S _{2, i} ) at least one constant (r, k) for the approximate determination of the angle or phase error (6Φ) and / or the amplitude (A) of the measurement signals (S _{1, i} , S _{2, i} ) is derived. A method according to claim 1, characterized in that - The value for the constants (r, k) is obtained once offline from a calibration data record. A method according to claim 1, characterized in that - for a plurality (N, i = 1 ... N) measurements from the respective measurement signals (S _{1, i} , S _{2, i} ) the sum of squares M i = S 2 1, i + S 2 2, i (1) and - the product T i = S 1, i * P 2, i (2) - are determined and - sum values (Sm, St, Smt, Stq) and a determinant (D) for calculating the constants (r, k) are determined from these quantities (M _i , T _i ). A method according to claim 3, characterized in that - the sum values and the determinant (Sm, St, Smt, Stq, D) are calculated as follows:

- D = N * Stq-St 2 (7) - and that the constants (r, k) are determined from these quantities as follows:

Method according to one of the preceding claims, characterized in that - from the respective measurement signals (S _{1, i} , S _{2, i} ) and the constants (r, k) in each case a corrected measurement signal according to the relationship

Method according to one of the preceding claims, characterized in that - from the constants (r, k) the amplitudes of the measurement signals according to the relationship

Circuit arrangement for carrying out a method according to one of the preceding claims, characterized in that - at the output of the sensor system ( 1 ) to which the analog measurement signals (A * cos (Φ), A * sin (Φ + δΦ)) are present, each analog / digital converter ( 2 . 3 ) for generating digital measurement signals (S _{1, i} , S _{2, i} ) are switched on that - a correction device ( 4 ) for the correction of the digital measurement signals (S _{1, i} , S _{2, i} ) is available and that - a calculation module ( 5 ) for calculating the constant (r, k) from the measurement signals (S _{1, i} , S _{2, i} ) is present, the first constant (r) being applied to a correction input of the correction device ( 4 ) is used to calculate the corrected measurement signal (Sk _{2, i} ). Circuit arrangement according to claim 7, characterized in that - at a further output of the calculation module ( 5 ) the constant (k) for calculating the amplitude (A) of the measurement signals (A * cos (Φ), A * sin (Φ + δΦ)) is present.