Method for operating a machine with a moving part, in particular a wheel

Abstract

Method for operating a machine (10) with a moving part (13), in particular a wheel, with a sensor device comprising several parts, wherein one part comprises at least two encoder elements (161, 162, 163, 164), and a part is a sensor element (22), and wherein the at least two encoder elements (161, 162, 163, 164) are connected to the moving part (13), wherein the at least two encoder elements (161, 162, 163, 164) act on the sensor element (22) during an operating period of the machine (10), and initial data are determined from an action of the at least two encoder elements (161, 162, 163, 164) on the sensor element (22) during the operation of the machine (10), and at least two encoder elements (161, 162, 163, 164) during a further operating period of the machine (10) act on the sensor element (22), and a part of the sensor device is changed and results from an action of the at least two sensor elements (161, 162, 163,164) during the second operation of the machine (10), second data are determined on the sensor element (161, 162, 163, 164), wherein the modified part of the sensor device causes an effect of the encoder element (161, 162, 163, 164) on the sensor element (22) that differs from the effect of the encoder element (161, 162, 163, 164) on the sensor element (22) before the modification, characterized in that second data are determined from the other effect of the encoder element (161, 162, 163, 164) on the sensor element (22) and at least part of the second data are compared at least indirectly with at least part of the first data.

Description

State of the art It is known from the operation of internal combustion engines that the position and speed of a crankshaft or camshaft are determined via a number of index flanks of a sensor wheel and / or their time intervals. Many speed-based functions, such as misfire detection (detection of ignition misfires), cylinder balancing, injection, and ignition, rely on precise position and speed information. Manufacturing-related positional errors, as well as differences or deviations of the sensor wheel's index flanks from the design ideal ("mechanical angular error"), are determined by adaptation algorithms. Correction values ​​are calculated, stored in non-volatile memory, and made available for future driving cycles.If such a sensor wheel is replaced due to a defect or by replacing a related component, the stored correction values ​​for the sensor wheel tooth errors must be reset by the workshop staff. However, this is often not guaranteed, so the old, invalid values ​​continue to be used in subsequent journeys. This can lead to incorrect diagnostic results or malfunctions. Embodiments of the invention According to a first aspect of the invention, a method for operating a machine with a moving part, in particular a wheel, is provided. A sensor device comprising several parts is provided, wherein one part has at least two encoder elements and another part is a sensor element, and wherein the at least two encoder elements are connected to the moving part. The at least two encoder elements act on the sensor element during one, in particular a first, operating cycle of the machine. Initial data are determined from the effect of the encoder element on the sensor element during the operation—in particular during one, in particular a first, operating cycle—of the machine. During a further, in particular a second, operating cycle of the machine, the at least two encoder elements act on the sensor element.From the further, especially second, effect of the at least two encoder elements on the sensor element during the second operation of the machine, second data are determined and in a comparison at least a part of the second data is compared at least indirectly with at least a part of the first data. Starting with an exemplary encoder wheel that has at least two encoder elements, the first data points to be determined are, for example, the time intervals between two encoder elements—especially those that are directly adjacent. If these two encoder elements are so-called teeth, they typically have two flanks: an ascending flank and a falling flank. An ascending flank is the first flank of a tooth approaching a sensor element, while a falling flank of the same tooth is the second flank approaching a sensor element. Based on such an arrangement, the times elapsed between two passages of ascending flanks past the sensor element can then be determined from the effects of the encoder elements on the sensor element. Alternatively, the times elapsed between two passages of falling flanks past the sensor element can also be determined. Alternatively, the time interval between the rising and falling flanks of a tooth as it passes the sensor element can be determined. Since the time elapsed in this way can be significantly shorter, this is a less preferred method. If such insights are gained during the initial operating period, initial data can be derived from them. In a first iteration or stage, this initial data might, for example, only describe a time interval. In a second iteration or stage of the process, the initial data can be converted into speed-independent initial data. For example, with a encoder wheel featuring two encoder elements and an evaluation of the falling edges, it could be determined that the elapsed times—which are part of the initial data—between the falling edges during one revolution of the encoder wheel are in a ratio of 1:1.002. This ratio can be a characteristic pattern, or part of a characteristic pattern, of this encoder wheel with two encoder elements, which is determined during an initial operating period. This initial operating period could be a single journey of a motor vehicle with an internal combustion engine between the start-up process and the end of the engine's operation. During a second operating period, which can be another single journey of a motor vehicle with an internal combustion engine between a starting process and the end of operation of the internal combustion engine, it can be determined, for example, with the same encoder wheel with two encoder elements and an evaluation of the falling edges, that the elapsed times - which are parts of the second data - between the falling edges during one revolution of the encoder wheel are, for example, in a ratio of 1:1.002. Within the framework of such a procedure, it will thus be advantageously possible to determine by a subsequent comparison that the encoder wheel used during the first operating period and during the second operating period has the same shape, since the data difference is zero and therefore smaller than a data difference threshold. In this case, the comparison performed confirms that speed-based functions such as misfire detection, cylinder balancing, injection, and ignition, which rely on precise position and speed information, can be used without modification. The similarity of the data obtained during the first and second operating periods indicates that any existing mechanical angular errors of the sensor wheel are the same and do not need to be recalculated using adaptation algorithms; therefore, the original values ​​can be used. Previously stored correction values ​​for the sensor wheel tooth errors are used again in the subsequent drive without resulting in faulty diagnostic results or malfunctions. According to a further aspect of the invention, it is provided that a data difference is determined by comparing other second new data with stored first data. If the data difference reaches at least a data difference threshold, i.e., if the data difference is greater than or equal to the data difference threshold, then it can be—or will be—decided that the encoder wheel used does not have the same shape. If the evaluated system is modified, for example by replacing the encoder wheel with two encoder elements, the manufactured contour of the second encoder wheel – which occupies the original position of the first encoder wheel – can and usually will deviate from the contour of the first encoder wheel, for example, due to manufacturing processes. If this encoder wheel is then used during a second operating period, which could be a single additional journey of a motor vehicle with an internal combustion engine between a starting process and the end of the engine's operation, it can be determined, for example, using this encoder wheel with two encoder elements and an evaluation of the falling edges, that the elapsed times – which are part of the second data – between the falling edges during one revolution of the encoder wheel are, for example, in a ratio of 1:1.0023. In this case, the comparison reveals that speed-based functions such as misfire detection, cylinder balancing, injection, and ignition, which rely on precise position and speed information, can only be used after a modification. The similarity of the data obtained during the first and second operating periods indicates that any mechanical angular errors of the sensor wheel are different, meaning that a system change has occurred. Therefore, for speed-based functions like misfire detection, cylinder balancing, injection, and ignition, which depend on precise position and speed information, correction values ​​must be recalculated—for example, using adaptation algorithms. The old values ​​cannot be used if speed-based functions are to operate correctly.Previously known stored correction values ​​for sensor wheel tooth errors are changed and reused in the subsequent journey, without resulting in faulty diagnostic results or malfunctions despite the system change. According to a further aspect of the invention, it is provided that a data difference is determined by comparing the second set of new data with the stored first set of data. If the data difference reaches at least a certain threshold, i.e., if the data difference is greater than or equal to the threshold, it can be determined that the encoder wheel used does not have the same shape. Within the framework of such a procedure, it will thus be advantageously possible to determine by a subsequent comparison that the encoder wheel used during the first operating period and during the second operating period does not have the same shape. It can also be concluded that the encoder wheel used during the second operating period is not the same. In any case, the comparison makes it possible to determine that speed-based functions such as misfire detection, cylinder balancing, injection, and ignition can or must be used with modified data to ensure proper machine operation. The difference in the data obtained for the first and second operating periods indicates that any existing mechanical angular errors differ and must be recalculated using adaptation algorithms; therefore, the old values ​​cannot be reused. Previously stored correction values ​​for the encoder wheel tooth errors are recalculated to prevent subsequent erroneous diagnostic results or malfunctions.All previously relevant adaptation values, the initial data, in the control unit are deleted. The advantage of this procedure is that it becomes impossible to revert to incompatible adaptation values ​​after replacing the sensor wheel, thus preventing misdiagnoses and malfunctions. As part of the procedure, the second set of data, the modified data, the adaptation values, are stored in the control unit. It is then planned that operation will continue with this second set of data. It is provided that a encoder wheel has individual encoder elements, in particular tooth flanks, and that the initial data includes at least one of the following data types: - tooth period durations, each indicating a time between passages of two adjacent individual encoder elements, such as tooth flanks, past a sensor element; - a speed-independent, characteristic pattern, in particular a data pattern, which is determined from the tooth period durations, in particular tooth durations. This pattern is stored in non-volatile memory. Such a pattern is preferably generated during the start-up of the machine or internal combustion engine and then compared with a previously known pattern, as mentioned above. Furthermore, a computer program is provided which is configured to execute all steps of one of the methods, or which is programmed to execute one of the methods when run on a computer. Additionally, a machine-readable storage medium is disclosed on which the computer program is stored, or on which the computer program is stored for use in a method. A control unit is configured to execute all steps of one of the methods, or which is programmed for use in a method. Embodiments of the invention The invention is explained in more detail with reference to the following figures: Fig. 1 shows a schematically represented machine, exemplified as an internal combustion engine, Fig. 2 shows a schematically represented method, Fig. 3 shows measured tooth period durations (polygonal path) and calculated, modeled ideal tooth period durations over two working cycles of an internal combustion engine, Fig. 4 shows normalized tooth periods over modeled tooth period durations over two working cycles of an internal combustion engine. Fig. 1 shows a schematic representation of a machine 10, exemplified as an internal combustion engine. This machine 10 has a moving part 13, which is exemplified as a wheel (encoder wheel). The moving part 13, designed as a wheel or encoder wheel, rotates with a shaft (drive shaft or camshaft) of the machine 10. The moving part 13 has – exemplified as four – encoder elements 161, 162, 163, 164, which are exemplified here as teeth. The four encoder elements 161, 162, 163, 164 are arranged distributed around the moving part 13. In this example, the encoder elements 161, 162, 163, 164, designed as teeth, are arranged on a contour 19 (edge ​​contour, outer contour) of the moving part 13. The encoder elements 161, 162, 163, and 164 are arranged essentially uniformly, with their respective positions deviating from a theoretically ideal position as intended by the design. This is within the context of dhWithin manufacturing tolerances, this is generally typical. The design proposes, by way of example, an ideal position at exactly "12 o'clock" (α1=0°), exactly "3 o'clock" (α2=90°), exactly "6 o'clock" (α3=180°), and "9 o'clock" (α4=270°). The actual position will deviate from this ideal position in each case, within the tolerances specified in the design. The encoder element 161 at the 12 o'clock position leads slightly and therefore assumes an actual position α11=3°, the encoder element 162 at the 3 o'clock position lags slightly and therefore assumes an actual position α21=88°, the encoder element 163 at the 6 o'clock position leads slightly and therefore assumes an actual position α21=183°, the encoder element 164 at the 9 o'clock position lags slightly and therefore assumes an actual position α31=268°. A sensor element 22 is arranged such that it generates at least one signal 25 each time the encoder elements 161, 162, 163, 164 pass through.Each encoder element 161, 162, 163, 164 has both a rising edge Fs1, Fs2, Fs3, Fs4 and a falling edge Ff1, Ff2, Ff3, Ff4. For example, only the signals 25 generated by the falling edges Ff1, Ff2, Ff3, Ff4 are evaluated. A signal 25 generated by the sensor element 22 is sent to a computer 28 and evaluated in this computer 28. The proposed method for determining the relative position of at least two encoder elements 161, 162, 163, 164 to each other proceeds as follows (see also Fig. 2): During the operation of the machine 10 and its moving part 13, the encoder elements 161, 162, 163, 164 pass the sensor element 22. Several – preferably all – falling edges Ff1, Ff2, Ff3, Ff4 act on the sensor element 22 in one step S10 and successively generate signals 25, which are supplied to the computer 28 and evaluated in this computer 28. In this process – preferably while the machine 10 is rotating – timestamps ti are accessed in step S20 for a defined number N+1 of signals 25. These timestamps are stored in a memory 31 assigned to the computer 28 in step S30 for each signal 25 (e.g., voltage edge) of the sensor element 22. That is, a point in time (time, "timestamp ti") is determined and stored for each signal 25. From the determined orIn step S40, N tooth period durations Ti = ti - ti-1 are determined between two signals 25 (e.g., voltage edges) at times ti and ti-1 from the measured N+1 timestamps ti (time points). Each such tooth period corresponds to a measurement tooth period and is equal to a measured time interval. In step S50, a sequence is generated from these tooth periods Ti, which is also referred to here as a measurement sequence. Furthermore, ideal tooth period durations T* are calculated or determined for the falling flanks Ff1, Ff2, Ff3, Ff4 (step S15) and a sequence (model sequence) is formed from these ideal tooth period durations T* (step S25). In a further step S60, the tooth period durations T (measured tooth period durations) determined from specific or measured times or timestamps are compared with the ideal tooth period durations T*i (determination of deviations, calculation of differences). The deviations from T*i determined in this process are used as the basis for determining the pattern. In particular, it is intended to average both the used tooth period durations Ti (measured tooth period durations) and the modeled tooth period durations T*i (model tooth period durations) over several revolutions of the moving part 13. Two possible examples for the formation of a model are explained in more detail below. According to a first example, the model sequence of tooth period durations is examined using Discrete Fourier Transform (DFT) for the frequency components of the single, double and triple firing frequency (amplitude and phase in each case), a linear gradient is determined and a corresponding model tooth period sequence is generated from it. In a second example, a simple low-pass filter is applied to the sequence of tooth period durations used. A suitable filter type and cutoff frequency are selected. Any resulting phase shift must be corrected accordingly. For example, a 30th-order FIR low-pass filter with a Hamming window can be used. N = number of teeth per revolution + 2*(filter order / 2)), relative cutoff frequency = 0.1. To correct the phase shift, the filtered tooth period durations i = (filter order / 2 + 1) ... (N - filter order / 2) are then used. Several implementations are conceivable for defining the features: In a first embodiment, the deviations of the measured tooth period from a model value are represented as ratios according to the formula ri = 10⁶ * ((Ti / T*i) - 1) and the resulting values ​​are assigned to the following categories 0 to 9: Category 0: ri = <-4000. Category 1: ri = -4000 ... <-3000. Category 2: ri = -3000 ... <-2000. Category 3: ri = -2000 ... <-1000. Category 4: ri = -1000 ... <-0. Category 5: ri = 0 ... <1000. Category 6: ri = 1000 ... <2000. Category 7: ri = 2000 ... <-3000. Category 8: ri = 3000 ... <4000. Category 9: ri= >4000. Applied to a motion element 13, designed as a so-called 60-2 encoder wheel – an encoder wheel with 58 regularly arranged teeth and two missing teeth, i.e., a large gap with a width equal to three regular tooth gaps plus the two correspondingly missing teeth – this results in a 57-digit decimal number or sequence of digits consisting of 57 tooth period durations (the large gap with the two missing teeth is not taken into account), which can either be stored or further compressed. This 57-digit or multi-digit decimal number or sequence of digits consisting of 57 (many) tooth period durations is the feature number M. For example, for a trip 1, the following numerical values ​​or determined tooth period durations are given for the first nine considered tooth gaps i = 1 to 9: T1 = 673.7 µs; T2 = 675.35 µs; T3 = 674.90 µs; T4 = 669.55 µs; T5 = 663.95 µs; T6 = 658.45 µs; T7 = 655.75 µs; T8 = 655.3 µs; T9 = 654.95 µs. The corresponding tooth gaps i = 1 are assigned ideal, modeled tooth period durations: T*1 = 673.8 µs; T*2 = 673.19 µs; T*3 = 671.35 µs; T*4 = 668.55 µs; T*5 = 665.18 µs; T*6 = 661.73 µs; T*7 = 658.62 µs; T*8 = 656.2 µs; T*9 = 654.61 µs. This results in the following ratios: r1 to r9: r1 = -181.21; r2 = 3197.77; r3 = 5281.02; r4 = 1500.00; r5 = -1851.04; r6 = -4954.95; r7 = -4356.17; r8 = -1346.27; r9 = 523.56. In case a careful person should notice differences between their own result and the figures given here, it should be noted that when determining the figures given here, further decimal places not shown here were taken into account. Following the above instructions for determining the characteristic numbers based on the categories, the following corresponding numbers (category values ​​Ki) result: K1= 4, K2= 8, K3= 9, K4= 6, K5= 3, K6= 0, K7= 0, K8= 3, K9= 5, ... . This results in the following for the feature number Mold = M2 = 489630035 ... . The exact number of features is: If another trip is undertaken with the same vehicle (trip 2), the following values ​​should be obtained as examples. For trip 2, the following numerical values ​​or determined tooth period durations are given below for the first nine considered tooth gaps i = 1 to 9: T1 = 657.10 µs; T2 = 655.95 µs; T3 = 660.25 µs; T4 = 661.90 µs; T5 = 666.75 µs; T6 = 667.70 µs; T7 = 670.90 µs; T8 = 670.20 µs; T9 = 673.20 µs. The corresponding tooth gaps i = 1 are assigned ideal, modeled tooth period durations: T*1 = 656.35 µs; T*2 = 658.25 µs; T*3 = 660.54 µs; T*4 = 663.10 µs; T*5 = 665.77 µs; T*6 = 668.35 µs; T*7 = 670.60 µs; T*8 = 672.20 µs; T*9 = 672.91 µs. This results in the following ratios: r1 to r9: r1 = 1138.05; r2 = -3501.67; r3 = -436.10; r4 = -1798.44; r5 = 1475.62; r6 = -977.74; r7 = 469.10; r8 = -2948.25; r9 = 431.32.In the event that an attentive person should find differences between their own result and the figures given here, it is again pointed out that when determining the figures given here, further decimal places not shown here were taken into account. Following the above instructions for determining the characteristic numbers based on the categories, the following corresponding numbers (category values ​​Ki) result: K1= 6, K2= 1, K3= 4, K4= 3, K5= 6, K6= 4, K7= 5, K8= 2, K9= 5, ... . This results in the characteristic number Mnew = M1 = 614364525 ... . In another implementation (second embodiment), ratios Ti / T*i, Ti+1 / T*i+1, Ti+2 / T*i+2, ... for various adjacent tooth periods are calculated from determined tooth period durations Ti, Ti+1, Ti+2, ... (measured tooth period durations) and modeled tooth period durations T*i, T*i+1, T*i+2, ... (model tooth period durations). Deviations of the standardized tooth periods from their respective subsequent standardized tooth periods are then represented or determined as differences di = ((Ti+1 / T*i+1) - (Ti / T*i)) * 10⁶. The resulting values ​​can be categorized as described for the first embodiment. In both embodiments, a decimal number or sequence of digits is determined, which is identified as a resulting feature number M2 or Mold – which is an “old” determined feature number M that is present during a, in particular the first, operating period of the machine 10. A feature number M is preferably determined at the beginning of the operating period of the machine 10. Adaptation values ​​are reset according to this proposal when the following condition is met: To determine whether a part of the sensor device has been modified, the following procedure is performed: For a second operating period of machine 10, a further decimal number or sequence of digits is determined according to the procedure described above. This is determined as a resulting characteristic number M1 or Mnew = 614364525... – which is a "new" or most recently determined characteristic number M during a subsequent, particularly a second, operating period of machine 10. This characteristic number M is also preferably determined at the beginning of this operating period of machine 10. For example, the threshold value S for deciding whether the moving part (encoder wheel) has been changed is defined as follows: the threshold value S should be the product of a chosen factor (here F = 1.5) and the tooth periods taken into account (here N = 57), i.e., S = F * N = 1.5 * 57 = 85 (rounded down). The following steps are taken with the two feature numbers M1 and M2 (or Mnew and Mold). Differences are calculated between the digits of the new characteristic value M1 and the old characteristic value M2. This difference is then squared. It follows that: If the sum of the squared errors E of two feature numbers M1 and M2 (or Mnew and Mold) exceeds an applicable threshold, the state "movement part 13 (encoder wheel) replaced" is recognized and the "old" adaptation values ​​dependent on movement part 13 (encoder wheel) are deleted. Since E > S, it is concluded that the moving part (sender wheel) has been replaced. Fig. 3 shows measured tooth period durations (polygonal path) and calculated, modeled ideal tooth period durations over two working cycles of an internal combustion engine. Fig. 4 shows normalized tooth periods over modeled tooth period durations.

Claims

Method for operating a machine (10) with a moving part (13), in particular a wheel, with a sensor device comprising several parts, wherein one part comprises at least two encoder elements (161, 162, 163, 164), and a part is a sensor element (22), and wherein the at least two encoder elements (161, 162, 163, 164) are connected to the moving part (13), wherein the at least two encoder elements (161, 162, 163, 164) act on the sensor element (22) during an operating period of the machine (10), and initial data are determined from an effect of the at least two encoder elements (161, 162, 163, 164) on the sensor element (22) during the operation of the machine (10), and at least two encoder elements (161, 162, 163, 164) during a further operating period of the machine (10) act on the sensor element (22), and a part of the sensor device is changed and results from an action of at least two sensor elements (161, 162, 163,164) during the second operation of the machine (10), second data are determined on the sensor element (161, 162, 163, 164), wherein the modified part of the sensor device causes an effect of the encoder element (161, 162, 163, 164) on the sensor element (22) that differs from the effect of the encoder element (161, 162, 163, 164) on the sensor element (22) before the modification, characterized in that second data are determined from the other effect of the encoder element (161, 162, 163, 164) on the sensor element (22) and at least part of the second data are compared at least indirectly with at least part of the first data. Method according to claim 1, characterized in that a data difference is determined by comparing parts of the second data with parts of the first data. Method according to claim 2, characterized in that the data difference does not reach a data difference threshold. Method according to claim 2, characterized in that the data difference reaches at least a data difference threshold. Method according to claim 4, characterized in that the first data are deleted. Method according to one of claims 1 to 5, characterized in that the second data is stored. Method according to claim 6, characterized in that the operation is continued with the second data. Method according to claim 7, characterized in that the first data remain in the memory (31) and continue to be used for operation of the machine (10). Method according to one of claims 1 to 8, characterized in that the encoder elements (161, 162, 163, 164) - in particular of a moving part (13), preferably encoder wheel - have individual encoder elements - in particular tooth flanks (Fs1, Fs2, Fs3, Fs4, Ff1, Ff2, Ff3, Ff4) - and the first data have at least one of the following type of data: tooth period durations, each of which specifies a time between passages of two adjacent individual encoder elements past a sensor element. Method according to claim 9, characterized in that a sequence is formed from the tooth period durations. Method according to claim 10, characterized in that a model is provided by which - in particular ideal - tooth period durations are determined and a model sequence is determined from the tooth period durations, and then the measurement sequence of the measured tooth period durations is compared with the model sequence from the model tooth period durations in a comparison (S60). Method according to claim 11, characterized in that deviations between the measuring tooth period durations and the model tooth period durations are determined from the comparison (step S60). Method according to claim 11 or 12, characterized in that the model tooth period durations and the measuring tooth period durations are determined over several measuring cycles - in particular of a moving part (13), preferably encoder wheel. Method according to one of claims 10 to 13, characterized in that frequency components are determined by means of discrete Fourier transformation - in particular a single or a double or a triple ignition frequency - then a linear gradient is determined and the model sequence of model tooth period durations is generated from it. Method according to one of claims 10 to 13, characterized in that the sequence of time durations is low-pass filtered. Method according to claim 12, characterized in that a characteristic number (ri) is determined by means of a ratio of a measuring tooth period and a model tooth period - in particular for the multiple distances of two adjacent encoder elements (161, 162, 163, 164) present on the moving part (13), preferably encoder wheel. Method according to claim 16, characterized in that a multi-digit key figure is determined from at least several - in particular all - determined individual key figures (ri), wherein each determined individual key figure (ri) determines a position in the multi-digit key figure. Method according to claim 12, characterized in that a deviation of a normalized measuring tooth period to a subsequent normalized measuring tooth period is determined by means of a difference (di). Computer program (34) configured to perform all steps of one of the methods according to any one of claims 1 to 18 or programmed to perform a method according to any one of claims 1 to 18 when executed on a computer. Machine-readable storage medium on which the computer program according to claim 19 is stored or on which the computer program according to claim 19 is stored for use in a method of claims 1 to 18. Control unit that is configured to perform all steps of one of the methods according to any one of claims 1 to 18 or that is programmed for use in a method according to any one of claims 1 to 18.

Images