Measurement data processor, position measurement device and computer implemented method

Abstract

The invention relates to a measurement data processor for correcting a position measurement error of a position measuring device having a magnetically measuring entity and a magnetic field sensor, and to a computer-implemented method. According to the invention, at least two periodic processes of a track signal of the magnetic field sensor are compared with each other in order to estimate a cross-periodic difference in a change process of the track signal. On the basis of the estimation, a correction signal in a periodic function is parameterized. By means of the correction signal, a corrected position signal of the position measuring device is generated. Advantageously, disturbing influences of external magnetic fields and / or of manufacturing-induced pole length differences of the measuring entity can be compensated in the described manner. Furthermore, the invention relates to a position measuring device having such a measurement data processor.

Description

Technical Field

[0001] This invention relates to a measurement data processor, particularly, but not limited to, for position measurement devices having a magnetic measuring entity and a magnetic field sensor. Furthermore, this invention relates to a position measurement device having such a measurement data processor and a computer-implemented method for correcting position measurement errors. Background Technology

[0002] Position measuring devices, such as rotary or linear encoders, are used in industry and science to obtain angle or position information. In many applications, magnetic position measuring devices are preferred due to their wear-free and maintenance-free nature. Such position measuring devices comprise a magnetic measuring entity and a non-contact magnetic field sensor, which provides raw measurement data for calculating angle or position information. However, position measurements calculated directly from this raw measurement data are often inaccurate and do not always correspond to the actual angle or position.

[0003] Traditional correction measures that consider measurement deviations through factory comparisons, such as correction values ​​statically stored in the corresponding position measuring equipment, are often insufficient because they do not provide the possibility of compensation for variable and / or external influences that are not present under factory conditions. Correction measures, such as factory comparisons, are also associated with specific components, such as the specific measuring entity. This association must then be considered over the entire product lifespan. Summary of the Invention

[0004] Therefore, the purpose of this invention is to realize the feasibility of improving the accuracy of magnetic position measuring devices through improved or at least additional correction measures.

[0005] According to the present invention, the objective is achieved by a measurement data processor according to the present invention.

[0006] The measurement data processor is particularly suitable for position measuring devices having a magnetic measuring entity consisting of alternating poles arranged in rows and a magnetic field sensor for generating a trajectory signal with a periodic process upon passing the poles. The measurement data processor is designed to retrieve measurement data representing at least two periodic processes of a first trajectory signal and a second trajectory signal of a magnetic field sensor moving relative to the poles at a pole frequency. Furthermore, the measurement data processor is designed to calculate, from the measurement data, at least one phase angle of the first trajectory signal and the second trajectory signal for each periodic process within the at least two periodic processes, and at least one characteristic value characterizing the change of the first trajectory signal and the second trajectory signal at said phase angle.

[0007] The measurement data processor is designed to determine at least one period comparison value by comparing at least a portion of the calculated feature values, the period comparison value representing the difference across periods of the change processes of the first trajectory signal and the second trajectory signal.

[0008] Furthermore, the measurement data processor is designed to: determine a correction signal from at least one periodic comparison value, wherein the correction signal is a periodic function or a superposition of periodic functions, each periodic function having an amplitude factor, a frequency factor, and a phase shift constant, wherein the amplitude factor and the phase shift constant are respectively related to at least one periodic comparison value, and wherein the frequency factor is a constant value independent of at least one periodic comparison value; calculate an uncorrected position signal from the measurement data; and calculate a corrected position signal from the uncorrected position signal and the correction signal.

[0009] The pole frequency reflects the number of poles passing through the magnetic field sensor per unit time and can be constant or variable over time, depending on the application. For example, the superposition of periodic functions can involve the summation of periodic functions. The periodic function can be a function related to the phase angle of the first and second trajectory signals. The corresponding amplitude factor, the corresponding phase shift constant, and the corresponding frequency factor are parameters of the correction signal, the determination of which is described in more detail below.

[0010] The advantages of this invention are described below:

[0011] The cross-cycle difference in the trajectory signal variation process mentioned above is mainly caused by the following: adjacent poles of the measured entity are not precisely complementary to each other, for example, due to differences in pole length caused by external magnetic fields and / or manufacturing. This situation is referred to as pole difference below.

[0012] Because the trajectory signal of a magnetic field sensor is related to the composition of the magnetic field at each pole, the pole differences cause different variations in the signal period during which the trajectory signals follow each other, thus distorting the accuracy of the uncorrected position signal. As discussed earlier, this effect of pole differences on the position signal cannot be adequately described by a constant correction value alone.

[0013] Because the measurement data processor according to the invention uses at least two cycle processes or measured data, it is advantageous to compare the signal cycles with each other. Therefore, by means of the measurement data processor according to the invention, particularly via at least one cycle comparison value, the degree or extent of polarity can be estimated. Based on this, the interference effect of polarity can be approximated in the form of a corrected signal, and subsequently minimized in the corrected position signal.

[0014] Therefore, the measurement data processor according to the present invention can be used to improve the accuracy of magnetic position measuring devices, and thus solve the initial objective.

[0015] The solution according to the invention can be further improved by different design schemes, each of which is advantageous in itself and can be arbitrarily combined with each other. The design forms and their associated advantages are discussed below.

[0016] According to the first feasible design, the measurement data processor can be designed to calculate the corrected position signal by subtracting the correction signal from the uncorrected position signal. This represents a type of computation for the corrected position signal that can be easily implemented.

[0017] To save computational resources, a measurement data processor can be designed to calculate the uncorrected position signal from measurement data and already calculated possible phase angles. For example, the measurement data processor can be designed to determine the number of complete cycle processes contained in the measurement data and from which a coarse position measurement is determined. Furthermore, the measurement data processor can be designed to determine a fine position measurement based on the last calculated phase angle. The uncorrected position measurement is then derived from a linear combination, particularly a sum, of the coarse and fine position measurements.

[0018] A general design form is derived when the measurement data processor is designed to receive measurement data directly from the magnetic field sensor and / or read measurement data from the measurement data memory.

[0019] According to another feasible design approach, the first trajectory signal and the second trajectory signal can exist in analog and / or digital form. Therefore, the compatibility of the measurement data processor according to the invention extends to different interface types of magnetic field sensors. In other words, the measurement data processor is compatible not only with magnetic field sensors that output analog trajectory signals, but also with magnetic field sensors that output digital trajectory signals.

[0020] In designs with analog trajectory signals, the measurement data processor can be designed to calculate at least the uncorrected position signal in an analog manner. In designs with digital trajectory signals, the measurement data processor can be designed to calculate multiple uncorrected position measurements from the measurement data and form an uncorrected position signal from the uncorrected position measurements.

[0021] According to another feasible design, the first trajectory signal can be substantially sinusoidal, and the second trajectory signal can be substantially cosine-shaped. Preferably, the first and second trajectory signals have substantially the same trajectory signal frequency. In other words, the period lengths of the first and second trajectory signals are consistent. At least one phase angle can then be advantageously and simply calculated from the first and second trajectory signals via inverse trigonometric functions, especially via the atan2 function.

[0022] Furthermore, the trajectory signal frequencies of the first and second trajectory signals can correspond to the pole frequencies. Therefore, the measurement data processor according to the invention can be advantageously used with a high-resolution position measurement device, such as an AMR sensor in which the period length of the trajectory signal coincides with the pole length of the measured entity.

[0023] Reference measurements show that the effects of extreme differences in interference can be well approximated by the corrected signal if the corresponding amplitude factor in the corrected signal is proportional to at least one period comparison value. This is explained in more detail below.

[0024] Furthermore, if the corresponding phase shift constant in the correction signal is an inverse trigonometric function of at least one period comparison value, then the correction signal well approximates the effects of polarity interference. This is explained in more detail below.

[0025] Based on experience, if the appropriate frequency factor is selected according to the type of magnetic field sensor, the interference effects of polarity differences can be well approximated by a correction signal. This is explained in more detail below.

[0026] To improve signal quality, the measurement data processor can be designed to perform phase, offset, and / or amplitude adjustments on the first and second trajectory signals. This can be done, for example, via phase, offset, and amplitude control (POAC), as is known in the prior art for position measurement devices and will not be described further.

[0027] The measurement data processor is preferably designed to calculate the Euclidean norm values ​​of the first and second trajectory signals from the measurement data based on the phase angle. Here, the Euclidean norm is calculated from the square root of the sum of the squares of the signal values ​​of the first and second trajectory signals at their respective phase angles.

[0028] Furthermore, the measurement data processor can be designed to at least partially use the Euclidean norm values ​​of the first and second trajectory signals for calculating eigenvalues. Additionally, the Euclidean norm values ​​of the first and second trajectory signals can at least partially be used for phase, offset, and / or amplitude adjustments. It is also feasible to use the Euclidean norm values ​​of the first and second trajectory signals to perform an evaluation of the trajectory signals based on a threshold, thereby checking whether the trajectory signals are within a reasonable operating range. This results in a synergistic effect of saving computational costs, because the calculated Euclidean norm values ​​can be included not only in phase, offset, and / or amplitude adjustments, but also in the evaluation of the trajectory signals and the calculation of eigenvalues.

[0029] Here, the corresponding value of the Euclidean norm can be directly derived as an eigenvalue. In other words, the corresponding eigenvalue characterizing the change of the first and second trajectory signals in the case of the phase angle is equal to the value of the Euclidean norm of the first and second trajectory signals calculated in the case of the phase angle. This design is advantageous because computational modules pre-built for calculating the Euclidean norm are widely used and easy to use. Alternatively, the corresponding eigenvalues ​​can be calculated differently, for example, as the sum of the squares of the signal values ​​of the first and second trajectory signals or as the sum of the magnitude values ​​of the signal values ​​of the first and second trajectory signals. According to another feasible design, the measurement data processor can be designed to divide the measurement data into data groups, wherein each data group includes multiple data pairs, which respectively represent pairs of signal values ​​of the first and second trajectory signals, and wherein the data pairs of the data group belong to the same periodic process. Furthermore, the measurement data processor can be designed to associate the phase angle within the periodic process to which the corresponding data pair belongs with each data pair. In addition, the measurement data processor can be designed to associate the data groups with the respective poles of the measured entity. Therefore, measurement data, but also all other variables based on the measurement data (such as the value of the Euclidean norm, periodic comparison value, and correction signal), can be associated with the individual poles of the measured entity.

[0030] Therefore, on the one hand, the measurement data for the two polarities (i.e., the North Pole and the South Pole) can be averaged separately, and the correction signal can be calculated based on the average value. Alternatively, correction signals representing only one North Pole and one South Pole can be calculated separately, and then applied accordingly to all other North and South Poles. Of course, correction signals can also be calculated and applied individually for each North and South Pole.

[0031] Furthermore, the measurement data processor can be designed so that the POAC uses associated data sets. Therefore, on the one hand, measurement data for the two polarities (i.e., the North and South Poles) can be averaged separately, and phase, offset, and / or amplitude adjustments for the North and South Poles can be performed separately based on the average value. Alternatively, phase, offset, and / or amplitude adjustments can be calculated individually for each pole first, and then averaged separately for both polarities. Alternatively, phase, offset, and / or amplitude adjustments representing only one North and one South Pole can be calculated, and then applied accordingly to all other North and South Poles separately. Of course, phase, offset, and / or amplitude adjustments can also be calculated and applied individually for all North and South Poles.

[0032] To determine at least one period comparison value in a simple manner, a measurement data processor can be designed to calculate a first value of the Euclidean norm of the trajectory signal from a first data pair of a first data group, and a second value of the Euclidean norm of the trajectory signal from a second data pair of a second data group, wherein the phase angles of the first and second data pairs are consistent or at least approximately consistent, and wherein the first and second data groups are associated with adjacent poles. Thus, the measurement data processor is preferably designed to determine at least one period comparison value based on the difference between the first and second values ​​of the Euclidean norm.

[0033] Optionally, the measurement data processor can be designed to calculate the difference, consisting of the values ​​of their Euclidean norms, from data sets of multiple data pairs that are consistent in phase angle among adjacent poles, and determine at least one periodic comparison value based on the difference. Therefore, sensitivity to isolated outliers in the measurement data can be reduced.

[0034] Here, each difference yields a period comparison value. Therefore, the measurement data processor can be designed to calculate multiple period comparison values. In particular, the measurement data processor can be designed to calculate at least two period comparison values ​​from data pairs with a spacing of half a period (π) between their corresponding phase angles. The measurement data processor is preferably designed to calculate a first period comparison value for a phase angle of -0.25π and a second period comparison value for a phase angle of 0.75π. This is advantageous because the values ​​of the Euclidean norm required for this purpose in the cases of phase angles of -0.25π and 0.75π can additionally be used in the phase adjustment of the POAC.

[0035] Alternatively or additionally, the measurement data processor can be designed to calculate the average, effective value, or rectified value (Gleichrichtwert) of the differences mentioned above as periodic comparison values. The average value is derived as the quotient of the sum of the differences divided by the number of differences. The effective value is derived as the square root of the quotient of the sum of the squares of the differences divided by the number of differences. The rectified value is derived as the quotient of the sum of the magnitudes of the differences divided by the number of differences.

[0036] To determine the parameters of the correction signal, the measurement data processor according to the present invention can be designed to calculate the corresponding frequency factor, the corresponding amplitude factor, and the corresponding phase shift constant by means of empirical equations or correlations. The empirical equations or correlations can be obtained in advance, for example, experimentally, by comparing the uncorrected position signal with an ideal reference signal during reference measurement, forming an error signal from the difference between the uncorrected position signal and the ideal reference signal, and examining the spectrum of the error signal. The spectrum can be obtained, for example, by means of the Discrete Fourier Transform, especially by means of the Fast Fourier Transform. Advantageously, the error signal and the spectrum obtained therefrom can be obtained piecewise, for example, at two adjacent poles or over two signal periods, in order to minimize the average effect of the Fourier Transform.

[0037] In order to apply the measurement data processor according to the present invention in a position measurement device with an AMR sensor, empirical equations or correlations described below are proposed:

[0038] Advantageously, the corresponding frequency factor in the correction signal is chosen to be an odd multiple of 0.5. For example, the correction signal can be a single-period function with a frequency factor of 0.5. Therefore, the second harmonic oscillation component of the error signal can be approximated by the correction signal and calculated from the corrected position signal.

[0039] Alternatively, the correction signal can also be the sum of a first periodic function with a frequency factor of 0.5 and a second periodic function with a frequency factor of 1.5. Therefore, in addition to the second harmonic oscillation component of the error signal, at least other oscillation components of the error signal can still be detected.

[0040] If at least one periodic comparison value is calculated from the effective values ​​of the differences already introduced above, the amplitude factor of the first periodic function (i.e., the periodic function with a frequency factor of 0.5) can be calculated as the product of the effective value multiplied by the square root of 3 + / - 0.17. The amplitude factor of the second periodic function (i.e., the periodic function with a frequency factor of 1.5) can be calculated as the product of the effective value multiplied by the square root of 0.125 + / - 0.08.

[0041] If at least one periodic comparison value is alternatively calculated from the rectified difference values ​​already introduced above, the magnitude factor of the first periodic function can be calculated as the product of the rectified value and 2.5 + / - 0.25. The magnitude factor of the second periodic function can be calculated as the product of the rectified value and 0.53 + / - 0.1.

[0042] Alternatively, the magnitude factor of the second periodic function can also be estimated as the magnitude factor of the first periodic function multiplied by 0.25.

[0043] As described above, if there exists a first period comparison value for a phase angle of -0.25π and a second period comparison value for a phase angle of 0.75π, then the phase shift constant for the first periodic function can be calculated as the sum of the function value of the atan2 function with the first and second period comparison values ​​in the form of independent variables plus 2 + / - 0.4, and the phase shift constant for the second periodic function can be calculated as the function value of the atan2 function with the first and second period comparison values ​​in the form of independent variables minus 2.7 + / - 0.4.

[0044] The initial objective can also be achieved by a position measuring device having: a magnetic measuring entity consisting of alternating polarity poles arranged in rows; a magnetic field sensor for generating a trajectory signal with a periodic process when passing the poles; and a measurement data processor according to one of the above design schemes.

[0045] The position measuring device according to the invention can be designed in particular as a magnetic rotary encoder. In this case, the measuring entity may, for example, involve a magnetic pole wheel. Alternatively, the position measuring device according to the invention can be designed as a magnetic linear encoder and have a magnetic tape or magnetic track as the measuring entity. The magnetic field sensor can be designed to send the trajectory signal directly to a measurement data processor. An analog-to-digital converter may optionally be provided between the magnetic field sensor and the measurement data processor.

[0046] The position measuring device according to the invention benefits from the advantages already described in the measurement data processor and is thus characterized by improved measurement accuracy. Furthermore, the position measuring device according to the invention is suitable for use in regulating loops, such as speed or rotation speed regulators, because the measurement data processor allows the calculation of the variables needed to calculate the corrected position signal solely from the measurement data representing the trajectory signal, independent of the system's motion process. Moreover, the position measuring device according to the invention contributes to improved system stability with improved regulating characteristics because it reduces interference without adding additional phase rotation to the signal.

[0047] According to a feasible design approach, the magnetic field sensor of the position measuring device can be an AMR sensor (i.e., a sensor that operates based on the principle of anisotropic magnetoresistive effect). Due to the AMR sensor, the position measuring device according to the present invention achieves relatively high resolution.

[0048] Alternatively or additionally, the magnetic field sensor may have a sensor length less than or equal to the pole length of the measured entity. In particular, the magnetic field sensor may be a so-called fixed-pitch AMR sensor, where the sensor length is substantially equal to the pole length. Alternatively, the magnetic field sensor may be a so-called free-pitch AMR sensor, where the sensor length is shorter than the pole length. In rotary encoders, the sensor length is measured tangentially to the direction of relative motion between the magnetic field sensor and the measured entity. In linear encoders, the sensor length is measured parallel to the direction of relative motion between the magnetic field sensor and the measured entity.

[0049] By choosing this sensor length, a more cost-effective magnetic field sensor with a smaller sensor surface can be used in the position measuring device according to the invention. Measurement errors resulting from the small sensor surface are typically corrected advantageously by means of a measurement data processor.

[0050] The initial objective can also be achieved through a computer-implemented method according to the present invention.

[0051] The method is suitable for correcting measurement errors in a position measuring device having a magnetic measuring entity composed of alternating polarities arranged in rows and a magnetic field sensor for generating a trajectory signal with a periodic process when passing over the poles. The method includes the following steps:

[0052] - Receive measurement data from at least two cycles of the first and second trajectory signals of a magnetic field sensor representing the polar motion relative to the measured entity.

[0053] - Calculate at least one phase angle of the first trajectory signal and the second trajectory signal for each of at least two cycle processes from the measurement data, and calculate at least one characteristic value characterizing the change process of the first trajectory signal and the second trajectory signal under the condition of said phase angle.

[0054] - At least one period comparison value is determined by comparing at least a portion of the calculated feature values, said period comparison value representing the difference across periods in the variation process of the first trajectory signal and the second trajectory signal.

[0055] - A correction signal is determined from at least one periodic comparison value, wherein the correction signal is a periodic function or a superposition of periodic functions, each periodic function having an amplitude factor, a frequency factor, and a phase shift constant, wherein the amplitude factor and the phase shift constant are respectively related to at least one periodic comparison value, and wherein the frequency factor is a constant value independent of at least one periodic comparison value.

[0056] - Calculate the uncorrected position signal from the measurement data, and calculate the corrected position signal from the uncorrected position signal and the corrected position signal.

[0057] As in the measurement data processor according to the invention, the method implemented by means of the computer can improve the measurement accuracy of the position measuring device because the interference effect of polarity difference is subtracted when calculating the corrected position signal.

[0058] Computer programs including instructions also provide the advantages mentioned above and thus solve the initial objectives, wherein the instructions, when executed by a computer, cause the computer to perform the method steps of a computer-implemented method. The computer program according to the invention particularly allows the computer-implemented method to be run on a general-purpose computer, such as a commercially available PC. This expands the applicability of the invention. The same applies to computer-readable storage media comprising instructions, which, when executed by a computer, cause the computer to perform the method steps of a computer-implemented method. The computer program according to the invention can particularly be stored on a computer-readable storage medium. Such a storage medium also contributes to the improved portability of the invention.

[0059] The advantages described regarding the measurement data processor and position measurement device also apply to the computer-implemented method according to the invention, and vice versa. Attached Figure Description

[0060] The invention is described in detail below with reference to the accompanying drawings. The combinations of features exemplarily shown in the illustrated embodiments can be supplemented by other features corresponding to the characteristics necessary for a particular application of the measurement data processor and / or the position measuring device according to the invention, based on the above embodiments. Similarly, according to the above embodiments, individual features in the described embodiments may be omitted if their function is not important in a particular application. In the drawings, the same reference numerals are always used for elements with the same function and / or the same construction.

[0061] The attached diagram shows:

[0062] Figure 1 A schematic diagram of a measurement data processor according to an exemplary embodiment of the present invention is shown;

[0063] Figure 2 A schematic diagram of a position measuring device according to an exemplary embodiment of the present invention is shown; and

[0064] Figure 3 Show Figure 2 Another schematic diagram of the position measuring device according to the present invention. Detailed Implementation

[0065] The following is for reference. Figure 1 The measurement data processor 1 according to the present invention is described. Furthermore, according to... Figure 2 and Figure 3 The position measuring device 2 according to the present invention is described.

[0066] While some aspects of the invention are described only within the scope of the device, it is of course possible that said aspects also represent descriptions of corresponding methods, wherein, for example, blocks, modules, units, or devices correspond to method steps or the functions of method steps. Similarly, aspects described within the scope of method steps also correspond to descriptions of blocks, modules, units, or characteristics of the device.

[0067] exist Figure 1 The diagram shows a simplified schematic of an exemplary embodiment of the measurement data processor 1. The measurement data processor 1 may have a separate processor circuit board 4 and / or be integrated on a circuit board (not shown) of the position measuring device 2. The blocks, modules, and units of the measurement data processor 1 described below may be implemented in hardware, software, or a combination of both.

[0068] Measurement data processor 1 is configured for position measuring device 2. Therefore, measurement data processor 1 can be used, for example, but not only, in magnetic position measuring device 6, which has a magnetic measuring entity 8 consisting of poles 10 of alternating polarities 12a, 12b arranged in rows, and a magnetic field sensor 14 for generating trajectory signals 16a, 16b with periodic processes 18a, 18b when passing over poles 10. The magnetic field sensor 14 and the measuring entity 8 are correspondingly movably designed relative to each other (see [reference needed]). Figure 2 and Figure 3 ).

[0069] exist Figure 2 and Figure 3A simplified schematic diagram of an exemplary embodiment of a position measuring device 2 is shown. The position measuring device 2 according to the invention includes a magnetic measuring entity 8, a magnetic field sensor 14, and a measurement data processor 1. In the illustrated embodiment, the position measuring device 2 is designed, for example, as a magnetic linear encoder 20. In this case, the measuring entity 8 may involve a magnetic tape 22 or a magnetic track 24. According to an alternative embodiment, the position measuring device 2 may also be designed as a magnetic rotary encoder 21 and have magnetic pole wheels 25 as the measuring entity 8 (see [link to documentation]). Figure 1 ).

[0070] The magnetic field sensor 14 can be an AMR sensor and can include two sampling heads 26a and 26b that operate based on the principle of anisotropic magnetoresistive effect. Figure 2 and Figure 3 In the illustrated embodiment, the magnetic field sensor 14 has a sensor length 30 measured parallel to the direction 28 of relative motion, said sensor length being equal to the pole length 32 of the pole 10 of the measuring entity 8. Alternatively, the sensor length 30 may not be equal to the pole length, and in particular, the sensor length 30 may be less than the pole length 32.

[0071] The measurement data processor 1 is designed to recall measurement data 34 representing at least two cycle processes 18a, 18b of the first trajectory signal 16a and at least two cycle processes 18a, 18b of the second trajectory signal 16b. The measurement data processor 1 is particularly designed to receive measurement data 34 directly from the magnetic field sensor 14 and / or read measurement data 34 from the measurement data memory 36. This is in… Figure 1 The values ​​are indicated either centered on the left or at the top left, respectively. To receive measurement data 34, the measurement data processor 1 may have a corresponding receiving unit 38 and a receiving interface 40. Correspondingly, to read measurement data 34, the measurement data processor 1 may have a reading unit 42 and a reading interface 44.

[0072] Depending on whether the magnetic field sensor 14 outputs analog and / or digital trajectory signals, the trajectory signals 16a and 16b can exist in analog and / or digital form. The magnetic field sensor 14 can also be designed to directly send the trajectory signals 16a and 16b to the measurement data processor 1. Optionally, an analog-to-digital converter 46 (see [reference]) can be provided between the magnetic field sensor 14 and the measurement data processor 1. Figure 2 and Figure 3 ).

[0073] Trajectory signals 16a and 16b are generated by the magnetic field sensor 14, and in particular the two sampling heads 26a and 26b, moving relative to the poles 10 of the measuring entity 8 at a time-constant or variable pole frequency, or by the poles 10 moving past the magnetic field sensor. The first trajectory signal 16a is the change in the local magnetic field at the magnetic field sensor 14 recorded by the sampling head 26a. The second trajectory signal 16b is correspondingly the change in the local magnetic field at the magnetic field sensor 14 recorded by the sampling head 26b. If adjacent poles 10a and 10b of the measuring entity 8 are not precisely complementary to each other, for example, due to differences in pole lengths caused by external magnetic fields and / or manufacturing processes, a pole difference exists. This pole difference impairs the measurement accuracy of the position measuring device 2 and is compensated for according to the invention by means of a measurement data processor 1 as described below.

[0074] exist Figure 1 As can be seen from the example, the first trajectory signal 16a is substantially sinusoidal and the second trajectory signal 16b is substantially cosine-shaped, wherein trajectory signals 16a and 16b have substantially the same trajectory signal frequency. In other words, the period length 48a of the first trajectory signal 16a coincides with the period length 48b of the second trajectory signal 16b. Furthermore, in this embodiment, the trajectory signal frequency coincides with the pole frequency. This means that each pole 10 of the magnetic field sensor 14 of the measuring entity 8 generates a periodic process 18 in the trajectory signals 16a and 16b, respectively.

[0075] To ensure that the trajectory signals 16a and 16b are matched to each other as well as possible, the measurement data processor 1 is preferably designed to perform phase, offset, and / or amplitude adjustments on the trajectory signals 16a and 16b. This can be done, for example, via phase, offset, and amplitude adjustment (POAC for short). The measurement data processor 1 may correspondingly have a POAC block 52.

[0076] To edit the measurement data 34, the measurement data processor 1 can also be designed to divide the measurement data 34 into data groups 54, each data group 54 comprising multiple data pairs 56, which represent signal value pairs 58 of trajectory signals 16a and 16b, respectively, and wherein the data pairs 56 of the data group 54 belong to the same periodic process 18. For this purpose, the measurement data processor 1 can have a data editing block 60. Furthermore, the measurement data processor 1 can be designed to associate the data groups 54 with the respective poles 10 of the measurement entity 8 in the data editing block 60. This association is feasible, in particular, due to the already mentioned consistency between the trajectory signal frequency and the pole frequency.

[0077] The measurement data processor 1 is also designed to calculate at least one phase angle of the trajectory signals 16a, 16b from the measurement data for each of at least two cycle processes 18a, 18b. and characterizing trajectory signals 16a, 16b at the phase angle The characteristic value r of the change process under the condition. For this purpose, the measurement data processor 1 has a phase angle calculation unit 62 and an characteristic value calculation unit 64.

[0078] Preferably, the corresponding phase angle during the periodic process The value can be between 0 and 2π or between -π and π, expressed in radians. In the illustrated embodiment, for example, the corresponding phase angle can be calculated from one of the data pairs 56, namely from the signal value S1 of the first trajectory signal 16a and the signal value S2 of the second trajectory signal 16b, via the atan2 function.

[0079]

[0080] Therefore, the measurement data processor 1 is also designed to convert each data pair 56 or each signal value S1 and S2 with its phase angle. Related.

[0081]

[0082]

[0083] The eigenvalue r can be calculated from the signal values ​​S1 and S2, for example, using the Euclidean norm, and then compared with the phase angle calculated from the same signal values ​​S1 and S2. Related:

[0084]

[0085] In this manner, for example, at least one characteristic value r1 for periodic process 18a and at least one characteristic value r2 for periodic process 18b can be calculated. Preferably, the measurement data processor 1 is designed to calculate characteristic value r1 from a first data pair 56a of a first data group 54a and characteristic value r2 from a second data pair 56b of a second data group 54b, wherein the phase angles of the first data pair and the second data pairs 56a, 56b are consistent or at least approximately consistent, and wherein the first data group and the second data group 54a, 54b are associated with adjacent poles 10a, 10b. In other words, the measurement data processor 1 is designed to calculate the same or approximately the same phase angle in the data editing block 60 from the data groups 54a, 54b of adjacent poles 10a, 10b. Two feature values ​​r1 and r2 in two related data pairs:

[0086]

[0087]

[0088] Independent variable It should be noted here that they involve the same phase angle. However, the corresponding signal values ​​S1 and S2 belong to the data groups of adjacent poles.

[0089] Optionally, the measurement data processor 1 is designed to adjust the phase, offset, and / or amplitude in the POAC block 52 according to the phase angle. The Euclidean norm values ​​of the trajectory signals 16a and 16b are fully or at least partially retrieved from the eigenvalue calculation unit 64 and used for phase, offset, and / or amplitude adjustments. This is in Figure 1 Arrow 66 is used to indicate this.

[0090] Furthermore, the measurement data processor 1 is designed to determine at least one period comparison value u by comparing at least a portion of the calculated feature values ​​r, said period comparison value representing the difference across periods in the variation process of the trajectory signals 16a, 16b. The degree or manifestation of the extreme differences mentioned above can be estimated via at least one period comparison value u.

[0091] For this purpose, the measurement data processor 1 has a period comparison block 68. The measurement data processor 1 is designed, for example, to calculate at least one period comparison value u in the subtraction module 70 based on the difference between the characteristic values ​​r1 and r2.

[0092]

[0093] Optionally, the measurement data processor 1 can be designed to calculate different phase angles from data sets 54a, 54b of adjacent poles 10a, 10b. Multiple such differences. Then, periodic comparison block 68 can output one periodic comparison value each: u1, u2, u3...

[0094] Where i = 1, 2, 3, ...

[0095] In particular, the measurement data processor 1 can be additionally designed to calculate the phase angle at intervals having half a cycle (π). In the case of at least two periodic comparison values ​​u1 and u2.

[0096]

[0097] in

[0098] Preferably, the measurement data processor is designed to calculate a first period comparison value u1 for a phase angle of -0.25π and a second period comparison value u2 for a phase angle of 0.75π.

[0099]

[0100]

[0101] It is particularly advantageous if the eigenvalues ​​r1 and r2, with phase angles of -0.25π and 0.75π, are also used for phase adjustment in POAC block 52.

[0102] Alternatively or additionally, the measurement data processor 1 may be designed to average the period comparison values ​​u over their quantity l. For this purpose, the period comparison block 68 may have an averaging module 72 designed to calculate the average value M, the RMS value E, or the rectified value G.

[0103]

[0104]

[0105]

[0106] Provided that the estimation of the degree of the extreme difference mentioned above exists in the form of at least one periodic comparison value u, the interference effect of the extreme difference can be approximated by means of a correction signal K. For this purpose, the measurement data processor 1 may have a correction block 74 designed to determine the correction signal K from at least one periodic comparison value u, wherein the correction signal K is a relationship between the amplitude factor a, the frequency factor f, and the phase shift constant p and the phase angle. The relevant periodic function F. For example:

[0107]

[0108] As can be seen from the equations above, the amplitude factor *a* and the phase shift constant *p* are parameters related to at least one periodic comparison value *u*. Conversely, the frequency factor *f* is a constant parameter. The specific determination of these parameters will be discussed below.

[0109] Alternatively, the correction signal K can also be formed by superimposing periodic functions F:

[0110]

[0111] The measurement data processor 1 is also designed to calculate the uncorrected position signal U from the measurement data 34 in the first position calculation unit 76a. The measurement data processor 1 may preferably be designed to consider possible already calculated phase angles when calculating the uncorrected position signal U. This is Figure 1 The arrow 78 indicates this.

[0112] In the digital trajectory signal, the measurement data processor 1 can be designed to calculate multiple uncorrected position measurements and form an uncorrected position signal U from the uncorrected position measurements. Specifically, the measurement data processor 1 can be designed to determine the number of complete cycle processes 18 and identify coarse position measurements from them. Furthermore, the measurement data processor 1 can be designed to determine the position based on the last calculated phase angle. Determine the fine position measurement values. Then, derive the uncorrected position measurement values ​​from the linear combination, and especially the sum, of the coarse and fine position measurement values.

[0113] In the analog trajectory signal, the measurement data processor 1 can be designed to calculate the uncorrected position signal U, at least in an analog manner.

[0114] Furthermore, the measurement data processor 1 is designed to calculate the corrected position signal P from the uncorrected position signal U and the correction signal K in the second position calculation unit 76b. Specifically, the measurement data processor 1 can be designed to calculate the corrected position signal P by subtracting the correction signal K from the uncorrected position signal U.

[0115] P=UK

[0116] In the corrected position signal P, the interference from the polarity difference mentioned above is minimized by calculating the correction signal K. Therefore, the position measuring device 2 provides a more accurate result in the form of the corrected position signal P.

[0117] Due to the correlation between data set 54 and the poles 10 of measurement entity 8, the measurement data 34 for the two polarities 12a, 12b (i.e., the North Pole and the South Pole) can be averaged separately, and a correction signal K can be calculated based on the average value. Alternatively, a correction signal K representing only one North Pole and one South Pole can be calculated, and then applied separately to all other North and South Poles accordingly. Of course, the correction signal K can also be calculated and applied separately for each North and South Pole.

[0118] To determine the specific parameters in the correction signal K, the correction block 74 may have a parameterization module 80a, 80b, 80c, each based on empirical equations or correlations. The empirical equations or correlations used can be determined in advance experimentally, for example, by comparing the uncorrected position signal U with the ideal reference signal during reference measurement, generating an error signal from the difference between the uncorrected position signal U and the ideal reference signal, and examining the spectrum of the error signal. The spectrum can be obtained, for example, using a discrete Fourier transform, particularly a fast Fourier transform, and further, segment by segment.

[0119] Based on experience, if the appropriate frequency factor f is selected according to the type of magnetic field sensor 14, the interference effect of polarity can be well approximated by the correction signal K. To input or read the type of the magnetic field sensor 14 used, the measurement data processor 1 may have an interface 82, through which the corresponding parameterization module 80a obtains type information 84.

[0120] For AMR sensors, it is proposed here that the frequency factor f be chosen as an odd multiple of 0.5:

[0121]

[0122] Then, the equation for the correction signal K, which has already been listed above, is applied, for example:

[0123]

[0124]

[0125] Generally, it is sufficient to superimpose the periodic functions F of the first two odd numbers:

[0126]

[0127] The reference measurements mentioned above indicate that if the corresponding amplitude factor a is calculated in the parameterization module 80b in proportion to at least one period comparison value u, the interference effect of the extreme difference can be well approximated by the correction signal K. If an effective value E, as introduced above, exists, the following correlation can be used in the equation of the correction signal K in the case of an AMR sensor:

[0128]

[0129]

[0130] If, alternatively, the rectified value G already introduced above exists, then in the case of an AMR sensor, the following correlation is proposed for the equation of the correction signal K:

[0131] a = a1 = 2, 5·G(u i )

[0132] a2 = 0, 53·G(u) i ).

[0133] Alternatively, the following simplified association can be used:

[0134]

[0135] If the corresponding phase shift constant p is calculated from the inverse trigonometric function of at least one period comparison value u in the parameterization module 80c, the correction signal K approximates the interference effects of the extreme difference well.

[0136] If there exist a first period comparison value u1 for a phase angle of -0.25π and a second period comparison value u2 for a phase angle of 0.75π, as already described, then, for example, the following correlation can be used in the equation of the correction signal K, respectively, in the case of an AMR sensor:

[0137]

[0138]

Claims

1. A measurement data processor (1) for a position measuring device (2) having a magnetic measuring entity (8) composed of rows of poles (10, 10a, 10b) of alternating polarity (12a, 12b) and a magnetic field sensor (14) for generating a track signal (16a, 16b) having a periodic course (18, 18a, 18b) when passing the poles (10, 10a, 10b), wherein the measurement data processor (1) is designed for: - calling measurement data (34) representing at least two periodic courses (18a, 18b) of a first and a second track signal (16a, 16b) of the magnetic field sensor (14) moving with a pole frequency relative to the poles (10, 10a, 10b), wherein the pole frequency reflects the number of poles passing the magnetic field sensor per unit of time and can be constant in time or variable as a function of the application, - calculating at least one phase angle of the first trajectory signal and the second trajectory signal (16a, 16b) from the measurement data (34) for each of the at least two periodic processes (18a, 18b) and at least one characteristic value r of a variation process of the first trajectory signal and the second trajectory signal (16a, 16b) at the phase angle ​ - determining at least one periodic comparison value u from a comparison of at least a portion of the calculated characteristic values r, which periodic comparison value u represents a cross-periodic difference of the varying courses of the first and the second track signal (16a, 16b), - determining a correction signal K from the at least one periodic comparison value u, wherein the correction signal K is a periodic function F or a superposition of periodic functions F, each having an amplitude factor a, a frequency factor f and a phase shift constant p, wherein the amplitude factor a and the phase shift constant p are each related to the at least one periodic comparison value u and wherein the frequency factor f is a constant value independent of the at least one periodic comparison value u, - calculating an uncorrected position signal U from the measurement data (34), and - calculating a corrected position signal P from the uncorrected position signal U and the correction signal K.

2. The measurement data processor according to claim 1, wherein the measurement data processor (1) is designed for calculating the corrected position signal P by subtracting the correction signal K from the uncorrected position signal U.

3. The measurement data processor according to claim 1 or 2, wherein the first track signal (16a) is essentially sinusoidal, wherein the second track signal (16b) is essentially cosinusoidal and / or wherein the track signal frequencies of the first and the second track signal (16a, 16b) correspond to the pole frequency.

4. The measurement data processor according to claim 1 or 2, wherein in the correction signal K the respective amplitude factor a is proportional to the at least one periodic comparison value u and / or the respective phase shift constant p is an inverse trigonometric function of the at least one periodic comparison value u and / or the respective frequency factor f is an odd multiple of 0.

5.

5. The measurement data processor of claim 1 or 2, wherein the measurement data processor (1) is designed to calculate the value of the Euclidean norm of the first trajectory signal and the second trajectory signal (16a, 16b) and the eigenvalue r from the measurement data (34) according to the phase angle the Euclidean norm of the first trajectory signal and the second trajectory signal (16a, 16b) and the eigenvalue r from the measurement data (34).

6. The measurement data processor according to claim 5, wherein the measurement data processor (1) is designed for: - dividing the measurement data (34) into data sets (54, 54a, 54b), wherein each data set (54, 54a, 54b) comprises a plurality of data pairs (56, 56a, 56b) representing signal value pairs (58) of the first and second track signals (16a, 16b), respectively, and wherein the data pairs (56, 56a, 56b) of a data set (54, 54a, 54b) belong to the same periodic process (18, 18a, 18b), respectively, - the phase angle within the periodic process (18, 18a, 18b) is determined associated with each data pair (56, 56a, 56b), - associating the data sets (54, 54a, 54b) with the poles (10, 10a, 10b) of the measurement entity (8), respectively, - calculating a first value of the Euclidean norm of the first and second track signals (16a, 16b) from a first data pair (56a) of a first data set (54a), - calculating a second value of the Euclidean norm from a second data pair (56b) of a second data set (54b), wherein the phase angles of the first data pair and the second data pair (56b, 56b) are correspond, and wherein the first data set and the second data set (54a, 54b) are associated with adjacent poles (10a, 10b), and - determining the at least one periodic comparison value u from a difference of the first and second values of the Euclidean norm.

7. The measurement data processor of claim 6, wherein the measurement data processor (1) is designed to calculate a difference of values of Euclidean norms of the first and second trajectory signals (16a, 16b), respectively, from data sets (54a, 54b) of adjacent poles (10a, 10b) of a plurality of data pairs (56a, 56b) that coincide in the phase angle and to determine the at least one cycle comparison value u from the difference.

8. The measurement data processor of claim 5, wherein the measurement data processor (1) is designed to perform a phase, offset and / or amplitude adjustment of the first and second track signals (16a, 16b) and for this purpose to use the values of the Euclidean norm of the first and second track signals (16a, 16b).

9. A position measuring device (2) having a magnetic measurement entity (8) composed of poles (10, 10a, 10b) of alternating polarity (12a, 12b) in rows with each other, a magnetic field sensor (14) for generating track signals (16a, 16b) having periodic processes (18, 18a, 18b) upon passing the poles (10, 10a, 10b), and a measurement data processor (1) according to any one of claims 1 to 8.

10. The position measuring device of claim 9, wherein the magnetic field sensor (14) is an AMR sensor, and / or the magnetic field sensor (14) has a sensor length (30) which is less than or equal to a pole length (32) of the poles (10, 10a, 10b) of the measurement entity (8).

11. A computer-implemented method for correcting measurement errors of a position measuring device (2) having a magnetic measurement entity (8) composed of poles (10, 10a, 10b) of alternating polarity (12a, 12b) in rows with each other and a magnetic field sensor (14) for generating track signals (16a, 16b) having periodic processes (18, 18a, 18b) upon passing the poles (10, 10a, 10b), wherein the method comprises the following steps: - calling measurement data (34) representing at least two periodic processes (18a, 18b) of first and second track signals (16a, 16b) of the magnetic field sensor (14) moving relative to the poles (10, 10a, 10b) of the measurement entity (8), - dividing the measurement data (34) into data sets (54, 54a, 54b), wherein each data set (54, 54a, 54b) comprises a plurality of data pairs (56, 56a, 56b) representing signal value pairs (58) of the first and second track signals (16a, 16b), respectively, and wherein the data pairs (56, 56a, 56b) of a data set (54, 54a, 54b) belong to the same periodic process (18, 18a, 18b), respectively, - associating the data sets (54, 54a, 54b) with the poles (10, 10a, 10b) of the measurement entity (8), respectively, - calculating a first value of the Euclidean norm of the first and second track signals (16a, 16b) from a first data pair (56a) of a first data set (54a), - determining the at least one periodic comparison value u from a difference of the first and second values of the Euclidean norm. - calculating at least one phase angle of the first trajectory signal and the second trajectory signal (16a, 16b) from the measurement data (34) for each of the at least two periodic processes (18a, 18b) and at least one characteristic value r of a variation process of the first trajectory signal and the second trajectory signal (16a, 16b) at the phase angle r, - determining at least one period comparison value u from a comparison of at least a part of the calculated characteristic values r, which period comparison value represents a period-crossing difference of the change processes of the first and second trajectory signals (16a, 16b), - determining a correction signal K from the at least one period comparison value u, wherein the correction signal K is a period function F or a superposition of period functions F, which each have an amplitude factor a, a frequency factor f and a phase shift constant p, wherein the amplitude factor a and the phase shift constant p are each related to the at least one period comparison value u, and wherein the frequency factor f is a constant value independent of the at least one period comparison value u, - calculating an uncorrected position signal U from the measurement data (34), and - calculating a corrected position signal P from the uncorrected position signal U and the correction signal K.

Images