Method for determining rotational speed and / or motion variable derived from rotational speed, computer program product, computer device

Abstract

A method, a computer program product, and a computer device for determining a rotational speed and / or a motion variable derived from the rotational speed. The invention relates to a method for determining a rotational speed and / or a motion variable derived from the rotational speed, in particular a wheel circumferential speed, by means of a rotational speed sensor (1) having an incremental encoder wheel (2) and a plurality of sensor elements (4), in which a time interval between temporally adjacent measurement events of the rotational speed sensor (1) is determined in particular continuously, wherein the time interval is determined, determining at least one estimated value # imgabs0 # of at least one duty cycle of at least one of the periods of at least two of the measurement events, in particular of one or a corresponding signal curve of the sensor element (4), wherein at least two of the time intervals are corrected as a function of the estimated value # imgabs1 #, and wherein the rotational speed and / or the motion variable are / is determined as a function of the corrected time interval.

Description

Field of the Invention

[0001] The present invention relates to a method for determining the rotational speed and / or a motion variable derived from the rotational speed, in particular the circumferential speed, by means of a rotational speed sensor having an incremental encoder wheel and a plurality of sensor elements. Furthermore, the present invention also relates to a computer program product which, when implemented on a computer device, executes the above method. Furthermore, the present invention also relates to a computer device which is specifically arranged for implementing the computer program product or the method mentioned above. Background Art

[0002] Methods of the type mentioned at the beginning are known from the prior art. For example, such methods are applied in a vehicle environment. Thus, many functions presented in an ABS / ESP control device require the circumferential speed (ω·r) of the wheels of a motor vehicle as an input variable. The circumferential speed is typically determined by rotational speed measurement using a rotational speed sensor. For example, from the published document DE10 2015 213 572 A1, a method for operating such a rotational speed sensor is known, which rotational speed sensor has a signal generator ring having a plurality of magnet elements arranged in a uniformly alternating magnetic orientation on its circumference, and the rotational speed sensor has a signal receiver having a plurality of sensor elements for detecting the magnetic field of the magnets, wherein information bits are generated based on the magnetic field strength detected by the sensor elements and provided as a rotational speed information signal. In order to increase the resolution of the rotational speed sensor, in particular the output signals of three sensor elements of the rotational speed sensor are correlated with each other. Summary of the Invention

[0003] The method according to the invention having the features of claim 1 is characterized in that, in particular, the time intervals between temporally adjacent measurement events of a rotational speed sensor are continuously determined, and on the basis of the determined time intervals, at least one estimated value of at least one duty cycle of at least one period of, in particular, a signal curve of a sensor element is determined, the period including at least two of the measurement events, at least two of the time intervals are corrected on the basis of the estimated value, and the rotational speed and / or the movement variable is determined on the basis of the corrected time intervals. Thereby, the measurement accuracy of the rotational speed sensor is advantageously improved, in particular as mentioned at the beginning, in order to increase the accuracy in detecting the circumferential speed of a wheel of a motor vehicle. The method can advantageously be used in the rotational speed sensor based on magnetic field strength detection mentioned at the beginning, but is not limited thereto. More precisely, the method can be used with any type of rotational speed sensor having an incremental encoder. Preferably, the duty cycle is a value between 0 and 1 relative to the length of the period and describes the relative temporal position of the measurement event or measurement pulse within the respective period. The invention is based in this regard on the recognition that the duty cycle as the relative position of the measurement event deviates from its ideal value in practice, in particular due to non-uniform sampling of the incremental encoder wheel, and the measurement accuracy is improved by taking into account its actual value according to the estimation of the invention. By means of the correction, the corrected time intervals correspond to the time intervals of an ideally uniform sampling of the incremental encoder wheel, in particular of a magnetic pole pair, such that, for example, systematic measurement errors in the circumferential speed signal derived therefrom are minimized. In particular, the estimated value is continuously determined after a pre-given number of measurement events, respectively. In particular, the measurement events are generated on the basis of at least one signal curve of at least one of the sensor elements. In a specific example of a sine signal, where the zero crossings of the signal curve are detected as measurement events, the estimated value has a nominal value of 1 / 2, where the "average" zero crossing of the sine signal period is located, and the position of this zero crossing is thus described. Preferably, alternatively or additionally, the reciprocal value of the difference is corrected, and the rotational speed and / or the movement variable is determined on the basis of the corrected reciprocal value. This reciprocal value is then the frequency.

[0004] According to a preferred refinement of the invention, a pre-given signal value of the signal curve, in particular the zero crossing of the signal curve, is considered as the measurement event. By taking into account this signal value, the measurement event can be characterized particularly advantageously simply. Preferably, at least one of the measurement events, in particular all of the measurement events, is generated when a pre-given signal value of the respective signal curve, for example a zero crossing event, is reached.

[0005] It is particularly preferably provided that the measurement events are determined on the basis of the signal curves of at least two of the sensor elements, in particular on the basis of the differential signal curves of at least two of the sensor elements. Thereby, the measurement events can be determined particularly advantageously robustly. For example, the zero crossings of the respective signal curves are considered as the measurement events.

[0006] According to a preferred improvement of the present invention, it is provided that at least one deviation value of the duty cycle from a pre-given ideal value and / or at least one reference value characterizing the duty cycle are determined based on at least three consecutive time intervals, and the estimated value is determined based on the deviation value and / or the reference value. By this determination and consideration of the deviation value and / or the reference value, a particularly advantageously simple determination of the estimated value is ensured.

[0007] Particularly preferably, it is provided that before determining the estimated value, the determined deviation value or reference value is filtered, in particular by means of a discrete PT1 filter with a pre-given initialization value and a pre-given gain factor, wherein for filtering, in particular at least one difference between two consecutive deviation values or reference values assigned to the time intervals is determined. By this filtering of the deviation value or the reference value, the following advantage is obtained: the accuracy in determining the estimated value is further improved.

[0008] According to a preferred improvement of the present invention, it is provided that at least one absolute value, in particular a plurality of absolute values, of the difference between two consecutive estimated values, deviation values and / or reference values assigned to the time intervals is determined, and based on this absolute value or these absolute values, it is determined whether the consecutive estimated values converge. Only when convergence is recognized is the time interval corrected, and / or the convergence rate is determined and the time interval is corrected in a manner weighted by the convergence rate. By the determination and consideration of convergence, a robust determination of the estimated value is advantageously ensured. In the case where the estimated value does not converge, the method is particularly aborted in this regard, or the time interval is not corrected. Alternatively or additionally, in particular, a correction weighted by the convergence rate is performed.

[0009] Particularly preferably, it is provided that the absolute value, the plurality of absolute values and / or the average value of the plurality of absolute values are compared with a pre-given first threshold value to recognize convergence. When in particular a pre-given number of absolute values or the average value is below this threshold value, convergence is recognized, and / or when at least one of the absolute values and / or the average value at least reaches, in particular exceeds, a pre-given second threshold value higher than the first threshold value and / or at least one of the time intervals in the time interval at least reaches, in particular exceeds, a pre-given third threshold value, the recognized convergence is discarded. From this, the following advantage is obtained: the actual convergence is correctly recognized.

[0010] According to a preferred improvement of the present invention, it is provided that at least two signal curves are generated, in particular based on sensor signals of at least two sensor elements of a rotational speed sensor that are offset from each other differently, and consecutive measurement events are respectively assigned to different signal curves, and an estimated value of each duty cycle is determined for each signal curve. By this, it is advantageously ensured that a reliable distinction is made between the signal curves, and only the estimated values assignable to the corresponding signal curves are determined.

[0011] It is particularly preferably provided that at least one estimated value of at least one relative offset between two consecutive measurement events assigned to different signal curves is determined, and the time interval is corrected based on this estimated value. By taking into account the offset estimate, the following advantage is obtained: the accuracy in correcting the time interval is further improved. In this regard, the offset estimate is determined in addition to the duty cycle estimate and taken into account during correction.

[0012] According to a preferred refinement of the invention, for the estimated values, the convergence rate of the respective duty cycle and / or the deviation value of the respective duty cycle from a pre-given ideal value are determined respectively, and based on the determined convergence rate and / or deviation value, the duty cycle is assigned to one of at least two different signal curves respectively. This advantageously ensures that the respective estimated values and thus the underlying measurement events are correctly classified, that is, assigned to the correct signal curve.

[0013] It is particularly preferably provided that the duty cycle assigned to the estimated value with the higher convergence rate and / or the smaller deviation value is assigned to the first signal curve, and the respective other duty cycle is assigned to the second signal curve, in particular by exchanging the estimated values. This ensures a particularly simple classification of the estimated values and the underlying measurement events.

[0014] According to a preferred refinement of the invention, if at least one of the time intervals is below a pre-given fourth threshold, the measurement events assigned to one of the signal curves are discarded. This advantageously ensures that from the determined threshold, the computational effort of the method is reduced, for example, by ignoring the corresponding measurement events from a determined vehicle speed, while at the same time ensuring the required measurement accuracy.

[0015] It is particularly preferably provided that the incremental encoder wheel is magnetically configured, and at least three sensor elements, in particular Hall sensors or magnetoresistive sensors, are arranged equidistantly in a row, wherein the difference between the sensor signals of two, in particular two outer, sensor elements is generated as the first signal curve, and the difference between the sensor signal of one of the sensor elements, in particular the middle sensor element, and the average value of the sensor signals of two other, in particular two outer, sensor elements of the sensor elements is generated as the second signal curve. The following advantage is obtained: the accuracy in determining the motion variable is further improved. In this regard, a virtual additional signal is generated between two measurement events. In particular, as the respective signal values, a periodic signal curve, in particular the zero-crossing event of the magnetic flux density, corresponding to the arc length of one magnetic pole pair is pre-given respectively. For example, Hall sensor elements are used, where as the measurement event, the zero-crossing of the curve of the magnetic flux density of the output sensor element is output, that is, two zero-crossings per magnetic pole pair. In order to improve the accuracy, the mentioned quasi-virtual additional measurement events based on the differential signal of the Hall sensor element are then generated.

[0016] The computer program product according to the invention for implementation on a computer device, having the features of claim 14, is characterized in that the computer program product, when used as prescribed, implements the method according to the invention. The advantages already mentioned thereby result. Preferably, a data carrier is provided which has stored thereon the computer program product according to the invention.

[0017] The computer device having the features of claim 15 is characterized in that the computer device is specifically arranged for executing the method according to the invention or implementing the computer program product according to the invention. The advantages already mentioned above thereby also result. Preferably, the computer device is a control device assigned to a motor vehicle and in particular arranged in the motor vehicle. Description of the Drawings

[0018] Other preferred features and combinations of features result from what has been described previously and from the claims. The invention will be explained in more detail below with the aid of the drawings. For this purpose:

[0019] Figure 1 A rotational speed sensor is shown,

[0020] Figure 2 A first signal curve is shown,

[0021] Figure 3 A method for determining the rotational speed and / or a movement variable derived from the rotational speed is shown, and

[0022] Figure 4 A second signal curve is shown. Detailed Description of the Invention

[0023] Figure 1 A previously known rotational speed sensor 1 for determining, for example, the circumferential speed ω·r is shown, which circumferential speed can then be used as an input variable for ABS / ESP regulation as mentioned at the beginning. The rotational speed sensor 1 is configured to detect (measurement) events, such as a change in magnetic flux, of an encoder or incremental encoder wheel 2 rotating with the wheel and to transmit them to a computer device 3 (currently a control device). The computer device 3 is in turn configured to determine, in particular with the aid of a high-frequency counter (typically 10 MHz), the time interval Δt(i) between the i-th event and the preceding (i - 1)-th event, and to estimate the instantaneous wheel rotational speed based on this time interval and to estimate the circumferential speed in a manner derived from this instantaneous wheel rotational speed.

[0024] Magnetic multi-pole wheels are generally used as the encoder or incremental encoder wheel 2. In the case of an incremental encoder, a plurality of, currently three, sensor elements 4 are used, which are arranged equidistantly in a row (in Figure 1are presented one above the other vertically. The sensor element 4 is in particular configured as a Hall sensor, an AMR sensor, a GMR sensor, and / or a TMR sensor and is configured to detect, for example, in the tangential direction, a magnetic field emitted from the incremental encoder wheel 2 and output a measurement variable depending on the magnetic field, such as a Hall voltage.

[0025] Measurement events transmitted to the computer device 3 are derived accordingly from the measurement variables of the sensor element 4. For example, the signal zero crossing of the magnetic flux density is used as a measurement event. During wheel movement, therefore, exactly two events occur for each magnetic pole pair of the incremental encoder wheel 2 over the arc length D. In order to reduce the offset error of the zero position due to, for example, temperature effects, in this case, the differential signal of the signals of two external sensor elements, currently referred to as A and B, in the sensor element 4 (corresponding to the lower and upper sensor elements 4 in Figure 1 is regarded as the first signal curve. The signal of the intermediate sensor element, currently referred to as M, in the sensor element 4 (corresponding to the intermediate sensor element 4 in Figure 1 is used, for example, for rotation direction determination.

[0026] Ideally, the magnetic poles within a pole pair have the same length and there is no offset error in the zero position of the differential signal, so that the events occur at a spatial interval of D / 2. In practice, the relative position of the events, called the duty cycle α, deviates from the ideal value of 1 / 2, typically by ±5% (the duty cycle is thus a value between 0 and 1).

[0027] In the case of an ideal incremental encoder wheel 2, each pole pair has the same arc length D, that is, all pole pairs are evenly distributed on the encoder circumference. Due to manufacturing, in practice, there are indexing errors, typically up to ±5% here as well. The resulting, wheel-periodically repeating pattern is learned and compensated for in particular by evaluating the event sequence in the computer device 3. The corresponding method is described, for example, in “Increasing signal accuracy of automotive wheel-speed sensors by on-line learning” (R. Schwarz, O. Nelles, P. Scheerer, and R. Isermann in American Control Conference, Albuquerque, New Mexico, 1997). For the method to be described later, this possible non-uniformity (indexing error) of the pole pairs of the encoder wheel relative to each other is preferably regarded as non-existent or is adequately compensated.

[0028] In particular, the more recent developments in the field of automated parking require a higher path resolution in the sense of a higher number of events per revolution of the wheel. Here, due to technical and economic reasons, encoder wheels with significantly larger numbers of pole pairs are not effective. However, it is possible to increase the number of measurement events per pole pair, for example, by considering additional signal curves of individual signals offset from each other. In particular, an additional second signal curve is derived from the zero crossings of the differential signal of the middle sensor element M and the signal average value of the two outer sensor elements A and B.

[0029] The relative positions of the now four events per pole pair can be described with the aid of three independent variables. Subsequently, for the description, the existing duty cycle α (ideal value 1 / 2) of the measurement events of the first signal curve, the duty cycle β (ideal value 1 / 2) of the measurement events of the second signal curve, and the relative displacement γ (ideal value 1 / 4) of the measurement events as the offset of the signal curves relative to each other are used. This is not a limitation of the invention, and independently selected variables can always be transformed into the three selected variables.

[0030] Generally speaking, the statistical properties of the measurement events of the second signal curve as intermediate events are different from the statistical properties of the measurement events of the first signal curve as standard events. For example, different correlations of the measurement variables result in a typically smaller amplitude of the differential signal M - (A + B) / 2, which depends on Figure 2 the ratio of the indicated sensor element spacing d to the arc length D, and thus results in a higher sensitivity in the determination of the time position of the offset error, for example, relative to the zero position of the corresponding measurement events and / or in the case of an incremental encoder with sensor element 4 being tilted in the magnetic field of the incremental encoder wheel 2 acting as an encoder.

[0031] This manifests itself in an increased spread (Streuung) of β compared to α in the measurement. Due to the spatial and temporal sampling of the magnetic field, where the spatial component now becomes relevant, the accuracy of the geometric mid - position of the sensor element M also has an impact on the duty cycle value β of the intermediate events.

[0032] In Figure 2 these basic relationships are presented. In Figure 2 for this purpose, a part of the incremental encoder wheel 2 acting as an encoder is projected, which has alternating magnetic north poles N and magnetic south poles S. Each of these pole pairs has an arc length D. In the following, in the corresponding figures, with respect to the encoder rotation angle exemplary corresponding sine curves of the magnetic flux density B in the far - field of the encoder at the three sensor elements A, M, and B are plotted when the encoder moves to the right. In another corresponding figure, with respect to this rotation angle The differential signals A-B and M-(A+B) / 2 used as described are plotted. Each arc length D accordingly includes one period of the respective signal and thus includes two zero-crossing events for each signal curve in the signal curves.

[0033] The following reference Figure 3 describes an advantageous method for determining the rotational speed and / or a motion variable derived from the rotational speed, in particular by means of a rotational speed sensor, in particular by means of the rotational speed sensor 1 described above. For this purpose, Figure 3 the method is illustrated by means of a flow chart. In particular, by means of this method it is ensured that the accuracy in determining the respective variable is improved by taking into account the deviation of the duty cycle from its respective ideal value as described above. The method is in particular carried out by means of a computer device 3.

[0034] In step S1, the method in particular continuously determines the start of the respective time interval between temporally adjacent measurement events of the rotational speed sensor. As measurement events, preferably predefined signal values of the signal curve, in particular the zero-crossing of the signal curve, are considered. Particularly preferably, the measurement events are determined according to the signal curves of at least two sensor elements in the sensor element, in particular according to the differential signal curve of at least two sensor elements in the sensor element. For example, reference is made to Figure 1 and Figure 2 the zero-crossings of two signal curves derived from the respective associated differential signals of the sensor element 4 as described. However, for the method according to the invention, in general it is sufficient if only one signal curve is correspondingly considered, for example the first signal curve as the differential signal of the sensor elements A and B.

[0035] For example, at least three sensor elements, in particular Hall sensors or magnetoresistive sensors, are arranged in a row equidistantly as described in Figure 1 and Figure 2 such that the difference between the sensor signals of two sensor elements, in particular two outer sensor elements, is generated as the first signal curve, and the difference between one sensor signal of one of the sensor elements, in particular the middle sensor element, and the average value of the sensor signals of two other sensor elements, in particular two outer sensor elements, of the sensor element is generated as the second signal curve. Herein, in particular, but not mandatorily, the incremental encoder wheel is magnetically formed. Thus, the method can also be advantageously applied in the case of optical sensing or other types of event transmission or measurement.

[0036] In the subsequent step S2, at least one estimated value of at least one duty cycle of one signal curve or the respective signal curve of the sensor element is determined according to the determined time interval, the period of which includes at least two of the measurement events.

[0037] With reference to Figure 1 and Figure 2 , for example, the duty cycles α and β and the relative position in the form of a relative displacement γ of the measurement events within the pole pair are estimated based on the measured time intervals of all the measurement events identified by the rotational speed sensor, and based on this, the corresponding measurement events are classified according to the estimated duty cycle values and their convergence speeds, that is, the corresponding measurement events are assigned to their corresponding signal curves, as will be described in detail later.

[0038] In Figure 4 a corresponding example of the rotational speed sensor 1 having three sensor elements 4 described above is presented. Figure 4 For this purpose, the first and second signal curves plotted over a large range of the rotational angle are shown, such that two similar measurement events (two consecutive arc lengths D or periods) out of four different events for each arc length D are presented respectively, and the measurement events are spaced apart from each other by D 1 , D 2 , D 3 , D 4 . In addition, the duty cycle α of the measurement events of the first signal curve, the duty cycle β of the measurement events of the second signal curve, and the relative offset γ are presented respectively, as described above.

[0039] In the specific embodiment described with reference to Figure 1 , Figure 2 and Figure 4 , the relative position of the events within the arc length D of the pole pair is estimated based on the continuous sequence Δt(i) of the measured time intervals between the measurement events. In the case of doubling the resolution, that is, when considering the second signal curve, it is assumed in this case that: each pole pair is always characterized by four adjacent time intervals, that is, in particular, the measurement events are not detected too many or too few.

[0040] Preferably, for this purpose, at least one deviation value of the duty cycle from a pre-given ideal value and / or at least one reference value characterizing the duty cycle is determined based on at least three consecutive time intervals, and the estimated value is determined based on the deviation value and / or the reference value.

[0041] In the simplest case, only one signal curve is considered, for example Figure 2 the first signal curve (A - B) in . Then the estimation of the duty cycle value is carried out especially (first manifestation) after every second event (i = 2n, where

[0042]

[0043] Alternatively, the deviation from the ideal value is estimated first

[0044]

[0045] Subsequently, the estimated value is determined based on this deviation

[0046]

[0047] Preferably, at least two signal curves are generated, in particular based on the sensor signals of at least two sensor elements of a rotational speed sensor that are offset from each other differently, as described above. Successive measurement events are each assigned to a different signal curve, and an estimated value of the duty cycle is determined for each signal curve. In this case, preferably, an estimated value of at least one relative offset of two successive measurement events assigned to different signal curves is additionally determined.

[0048] In a specific embodiment, in particular (first manifestation), after each fourth event (i = 4n, where ) the duty cycle and and the relative offset are estimated:

[0049]

[0050] For further resolution improvement, the corresponding processing method can be generalized accordingly. If, for example, 8 events are provided for each arc length, in particular for one pole pair, then after each eighth event (i = 8n, where ) the then four duty cycle values and and the relative offset values and

[0051]

[0052]

[0053] Returning to this specific embodiment, in an alternative representation, the deviation from each ideal value is estimated first (similarly, for example, a generalization to the 8 events mentioned above is possible):

[0054]

[0055] Subsequently, the estimated value is determined based on this

[0056]

[0057] In particular, before determining the estimated value, the determined deviation value or reference value is filtered, in particular by means of a discrete PT1 filter having a pre-given initialization value and a pre-given gain factor, wherein for the filtering, at least the difference between two consecutive deviation values or reference values assigned to a time interval is determined in particular.

[0058] In the simplest case, only one signal curve is considered here, for example Figure 2 the first signal curve (A - B) in and a gain factor k (for example k = 3%) in the form of a discrete PT1 filter:

[0059]

[0060] In this specific embodiment, in particular (second embodiment), the deviation from the ideal value is filtered for two signal curves (similarly, for example, a generalization to the 8 events mentioned above is possible), in particular in the form of a discrete PT1 filter having an initialization value and a gain factor k (for example k = 3%):

[0061]

[0062] In particular, when the measured time interval changes due to a change in speed, another preferred calculation rule (third embodiment) provides an advantageously improved estimation quality, in which the reference value is distributed equally in the denominator to two pole pairs obtained from the measured duration of the pole and its predecessor or successor, and the corresponding filtering is performed based on this.

[0063] In the simplest case, only one signal curve is considered here, for example Figure 2 the first signal curve (A - B) in

[0064]

[0065] For the specific embodiment considered with two signal curves, it follows (similarly, for example, a generalization to the 8 events mentioned above is possible):

[0066]

[0067] Particularly preferably, the convergence rate of the corresponding duty cycle and / or the deviation value of the corresponding duty cycle from a pre-given ideal value are determined for the estimated values respectively, and the duty cycles are respectively assigned to one of at least two different signal curves according to the determined convergence rate and / or deviation value. In particular, the duty cycle assigned to the estimated value with a higher convergence rate and / or a smaller deviation value is assigned to the first signal curve, and the other duty cycle is respectively assigned to the second signal curve, in particular by exchanging the estimated values.

[0068] If, for example, the type of event is unknown in this specific embodiment, the assignment of the estimated duty cycle to the curve is initially arbitrary. If, for example, the number of times below the threshold during convergence identification exceeds twice, one of the two duty cycle estimates or converges significantly faster than the other duty cycle estimate, then this duty cycle estimate is classified as the duty cycle of the first signal curve, and this other duty cycle estimate is classified as the duty cycle of the second signal curve. The background is the assumption of a higher spread when determining the time position of the measurement event of the second signal curve.

[0069] Alternatively or additionally, the duty cycle estimate that is closer to the ideal value of 1 / 2 or is classified as the duty cycle of the first signal curve, and the other duty cycle estimate is classified as the duty cycle of the second signal curve. The background is the assumption of a poorer offset adjustment for the measurement event of the second signal curve. These two criteria can be related in different ways, preferably classified first according to the convergence speed, and secondly according to the convergence value in case of lack of clarity.

[0070] If the initially arbitrary assignment of the estimated duty cycle to the event type during classification proves to be incorrect ( the duty cycle classified as the first signal curve), then preferably the estimated values are exchanged:

[0071]

[0072] Even if the type of event is known, the corresponding classification based on the statistical signal characteristics is still beneficial for credibility checking and / or monitoring. The classification can be similarly applied to and generalized to the 8 events mentioned above.

[0073] In particular, when at least one of the time intervals is below a pre-given fourth threshold, the measurement event assigned to one of the signal curves is discarded. Thus, the calculation of the motion variable can in particular be based only on the measurement events classified as such of the first signal curve and explicitly without using the measurement events of the second signal curve. Thereby, the computational cost and / or the signal noise in the calculated output signal can be advantageously reduced.

[0074] Preferably, the estimation is aborted in the case of a very large time interval in order to avoid special cases of reverse rotation direction, the time interval being close to the wheel rest state when using a rotational speed sensor at the wheel of a motor vehicle, and the assumption of four measurement events per pole pair may not be applicable in the special case. In particular, for example, in order to temporarily reduce the computational load or in the case of predictably poor convergence conditions, such as when strong speed changes are applied to the motor vehicle or when driving on an unpaved road, a temporary suspension of the estimation is carried out.

[0075] After the estimation is interrupted as described above (especially close to the wheel rest state), re-initialization is preferably carried out in order to achieve faster convergence of the estimated value in the case of reverse rotation direction.

[0076] In step S3, now at least two time intervals in the time interval are corrected according to one or more estimated values, such that the corrected time intervals correspond to the time intervals of ideally uniform sampling of the incremental encoder wheel, especially with respect to Figure 1 and Figure 2 one magnetic pole pair, and the systematic measurement error in the wheel circumference speed signal derived therefrom is minimized.

[0077] In the simplest case, only one signal curve is considered again, for example Figure 2 the first signal curve (A - B) in Then, in particular, the estimated value is continuously used to correct the measured time difference Δt(i), such that the corrected time difference

[0078]

[0079] corresponds to the time interval of ideally uniform sampling of this magnetic pole pair: and In this specific embodiment, similarly, the estimated values

[0080]

[0081] are preferably continuously used to correct the measured time difference Δt(i) (the following considerations can be similarly applied to and generalized to the 8 events mentioned above, for example):

[0082]

[0083] Similarly, in this specific embodiment, when considering two signal curves, the following corrected pulse frequency sequence is used:

[0084]

[0085] Regarding the determination of the motion variable, especially the wheel speed or the circumferential speed, the advantages of a higher-resolution incremental encoder are mainly limited to low speeds. Determined by the incremental encoder principle, the information aging of the speed signal (the time interval between the effective time points of two events and the available time points after the second event) increases reciprocally as the speed decreases.

[0086] Doubling the resolution set in this specific embodiment halves the information aging of the speed signal. This is relevant in the speed range where there is less than one event every 5 ms, for example. Relative errors, such as the still unconverged and thus uncorrected duty cycle of the second signal curve (in the case of low speeds), also only have a small impact on the absolute accuracy of the speed estimation.

[0087] In the case of speeds above this range, the halving of the information aging is negligible, and here the possible accuracy drawbacks of the intermediate pulses as measurement events of the second signal curve and the drawbacks of the computational effort increasing linearly with the number of events are dominant.

[0088] Therefore, preferably, for higher speeds, as described above, the motion variable is estimated using the following alternative pulse frequency sequence explicitly without using the intermediate pulses classified as such, that is, only for every second event and by accumulating the time differences measured by skipping over the intermediate pulses:

[0089]

[0090] Particularly preferably, at least one absolute value, especially a plurality of absolute values, of the difference between two consecutive estimated values, deviation values, and / or reference values assigned to a time interval is also determined in advance. In particular, then, based on this absolute value or plurality of absolute values, it is determined whether the consecutive estimated values converge, and the time interval is corrected only when convergence is recognized. Alternatively or additionally, the convergence rate is determined and the time interval is corrected weighted using the convergence rate.

[0091] Here, preferably, the absolute value, the plurality of absolute values, and / or the average value of the plurality of absolute values are compared with a pre-given first threshold value to identify convergence, and when, in particular, a pre-given number of absolute values or the average value is below the threshold value, convergence is recognized.

[0092] Alternatively or additionally, when at least one of the absolute value and / or the average value reaches, in particular exceeds, a predefined second threshold that is higher than a first threshold and / or when at least one of the time intervals reaches, in particular exceeds, a predefined third threshold, the identified convergence is discarded.

[0093] In the simplest case, again only one signal curve is considered, for example Figure 2 the first signal curve (A - B) in or alternatively, in the second and third embodiments, according to the absolute value of the change difference the estimated convergence is inferred.

[0094] In this specific embodiment, when considering two signal curves, in particular according to the absolute value of the change of the estimated value or or alternatively, in the second and third embodiments, according to the absolute value of the change difference or the corresponding estimated convergence is inferred.

[0095] As a particularly low - computationally - expensive implementation, preferably each absolute value is compared with a threshold of, for example, 0.1%, and convergence is identified when it is continuously below this threshold for, for example, 100 multi - poles n. Alternatively, for example, an average / filtering of the absolute value and a corresponding threshold comparison of the average value are set.

[0096] Preferably, as described, only the estimated value of the convergence is used to correct the measured time interval. Alternatively, a correction weighted by the convergence rate is set. In particular, when, for example, there is a large time difference (close to the stationary state), or if the absolute value exceeds, for example, a higher second threshold of 0.5% due to increased signal noise when driving on an unpaved road, the identified convergence is discarded again.

[0097] Finally, in step S4, the rotational speed and / or the motion variable are determined according to the corrected time interval. For example, as described above, the circumferential speed of the wheel. Preferably, this method is performed continuously.

[0098] In particular, in order to determine, in addition to the correction of the duty cycle, other signal - processing steps are performed, such as forming the reciprocal value, correcting the graduation error, anti - aliasing filtering, sampling rate conversion, considering the number of multi - poles of each encoder, and the wheel circumference. Preferably, the duty - cycle correction according to the present invention is first performed in combination with the formation of the reciprocal value, and for all subsequent steps, preferably one of the described pulse - frequency sequences is used.

Claims

1. A method for determining a rotational speed and / or a motion variable derived therefrom, in particular a wheel peripheral speed, by means of a rotational speed sensor (1) having an incremental encoder wheel (2) and a plurality of sensor elements (4), - in which the time intervals between temporally adjacent measuring events of the rotational speed sensor (1) are determined in particular continuously, - wherein at least one estimated value of at least one duty cycle of a period of at least two of the measuring events, in particular of one or a corresponding signal curve of the sensor element (4), is determined as a function of the determined time interval - where according to the estimated value correcting at least two of the time intervals, and wherein the rotational speed and / or the movement variable is determined as a function of a corrected time interval.

2. The method according to claim 1, characterized in that A predefined signal value of the signal curve, in particular a zero crossing of the signal curve, is considered as a measuring event.

3. The method according to any one of the preceding claims, characterized in that A measuring event is determined based on the signal curves of at least two of the sensor elements (4), in particular based on the differential signal curves of at least two of the sensor elements (4).

4. The method according to any one of the preceding claims, characterized in that at least one deviation value of the pulse duty cycle from a predetermined ideal value and / or at least one reference value characterizing the duty cycle is determined based on at least three consecutive time intervals, and the estimated value According to the deviation value and / or reference values.

5. The method according to claim 4, characterized in that In determining the estimated value Previously, the determined deviation value was obtained, in particular by means of a discrete PT1 filter with a predetermined initialization value and a predetermined gain factor. or reference values, wherein for filtering, in particular two consecutive deviation values ​​assigned to a time interval are determined Or at least a difference of a reference value.

6. The method according to any one of the preceding claims, characterized in that Determine the values ​​of two consecutive estimates assigned to the time interval Deviation value and / or at least one absolute value, in particular a plurality of absolute values, of a difference between a reference value and / or a reference value, from which the continuous estimated value is determined Convergence or not, the time interval is corrected only when convergence is identified, and / or the convergence rate is determined and the time interval is corrected in a manner weighted by the convergence rate.

7. The method according to claim 6, characterized in that The absolute value, the multiple absolute values ​​and / or the average value of the multiple absolute values ​​are compared with a predetermined first threshold value to identify convergence, and convergence is identified when, in particular, a predetermined number of absolute values ​​or average values ​​are below the threshold value, and / or the identified convergence is discarded when at least one of the absolute values ​​and / or average values ​​at least reaches, in particular exceeds, a predetermined second threshold value higher than the first threshold value and / or at least one of the time intervals at least reaches, in particular exceeds, a predetermined third threshold value.

8. The method according to any one of the preceding claims, characterized in that In particular, at least two signal curves are generated based on sensor signals of at least two sensor elements (4) of the rotational speed sensor (1) that are offset differently from one another, consecutive measurement events are respectively assigned to different signal curves, and an estimated value for each duty cycle is determined for each of the signal curves.

9. The method according to any one of the preceding claims, characterized in that Determining at least one estimated value of at least one relative shift of two consecutive measurement events assigned to different signal curves And based on this estimate The time interval is corrected.

10. The method according to any one of the preceding claims, characterized in that For the estimated value Determine the convergence rate of the corresponding duty cycle and / or the deviation value of the corresponding duty cycle from a predetermined ideal value respectively And according to the determined convergence rate and / or deviation value A pulse duty factor is respectively assigned to one of the at least two different signal curves.

11. The method according to claim 10, characterized in that will be assigned to the cluster with higher convergence rate and / or smaller deviation value Estimated value of A duty factor of one embodiment is assigned to the first signal curve and another duty factor is assigned to the second signal curve, in particular by exchanging the estimated values 12. The method according to any one of the preceding claims, characterized in that A measuring event assigned to one of the signal curves is discarded if at least one of the time intervals falls below a predefined fourth threshold value.

13. The method according to any one of the preceding claims, characterized in that The incremental encoder wheel is designed magnetically and at least three sensor elements (4), in particular Hall sensors or magnetoresistive sensors, are arranged equidistantly in a row, wherein as a first signal curve, a difference between sensor signals of two, in particular two outer sensor elements (4) is generated, and as a second signal curve, a difference between one of the sensor signals of one of the sensor elements (4), in particular the middle sensor element, and an average value of the sensor signals of two other, in particular two outer sensor elements of the sensor elements (4) is generated.

14. A computer program product for implementation on a computer device (3), characterized in that The computer program product, when used as intended, carries out the method according to one of the preceding claims.

15. A computer device (3), in particular for a control device of a motor vehicle, characterized in that The computer device (3) is specially designed to implement the computer program product according to claim 14.

Images