Method for processing an electrical signal and control unit for a camshaft adjuster or internal combustion engine for carrying out the method

Abstract

The invention relates to a method for processing an electrical signal (S) that is generated continuously over time based on measured values ​​from a sensor (23, 24, 25) and transmitted via a line (20, 22) exposed to disturbances, comprising the following features: setting a sampling rate interval (T), setting quantization levels (Q1, Q2), assigning the electrical signal to a quantization level (Q1, Q2), determining the number of changes of the quantization level (Q1, Q2) within the sampling rate interval (T), and discarding all even-numbered changes of the quantization level (Q1, Q2) within the sampling rate interval (T). It also relates to a control unit (17, 21) for a camshaft adjuster (7, 8) or for an internal combustion engine (1) that performs such a method.

Description

[0001] The invention relates to a method for processing an electrical signal that is generated continuously over time and based on measured values ​​from a sensor and transmitted through a conductor exposed to disturbances.

[0002] Furthermore, the invention relates to a control unit for an internal combustion engine or a camshaft adjuster.

[0003] Various operating methods for camshaft adjusters are described in documents DE 102 59 133 A1 and DE 102 42 659 A1. These documents specifically address the interaction between the engine control unit and the camshaft adjustment. Swashplate drives, functioning as three-shaft drives, are used as the actuating mechanisms for the camshaft adjusters.

[0004] German patent DE 10 2012 219 297 A1 describes a method for operating a motor vehicle that includes an engine control unit and an additional camshaft adjustment control unit. A CAN bus is provided for data transmission. The camshaft adjustment control unit can be started before the engine control unit has finished booting up.

[0005] WO 2006 / 122 665 A1 describes a topology for generating a control signal for an electrically actuated camshaft adjuster. A control unit is integrated into a control unit. Hall sensors are used to detect the states of the camshaft adjuster.

[0006] DE 10 2004 041 232 B4 describes a method for operating a camshaft adjuster, which may include an electric or a hydraulic actuating device. In each case, reference and / or base values ​​relating to the state of the camshaft adjuster are obtained within the operating procedure by averaging values ​​taken at different times.

[0007] In electric camshaft adjusters, as in engine control units, the electrical signals from sensors on the crankshaft and camshaft trigger wheels are evaluated. Unlike engine control units, however, the cables between the control units and the sensors are comparatively long. These cables are exposed to a multitude of interferences in the engine compartment.

[0008] Ground shifts due to high current loads and electromagnetic interference cause voltage spikes and signal noise in the useful signal. With increasing electrification in vehicles, the number of interference sources also grows, making reliable signal transmission over long distances no longer guaranteed.

[0009] Signal evaluation is therefore usually difficult and more complex than with engine control units. Object of the invention

[0010] The invention is based on the objective of avoiding the disadvantages of the prior art and, in particular, of processing an electrical signal that has been exposed to electrical interference.

[0011] This problem is solved according to the invention by a method for processing an electrical signal according to claim 1. Likewise, the problem is solved by a control unit with the features of claim 10. The embodiments and advantages of the invention explained below in connection with the control unit also apply mutatis mutandis to the method and vice versa.

[0012] The method according to the invention is based on the understanding that the voltage peaks and signal-noise components radiated into many useful signals have a different duration than their average change. If the temporal and / or amplitude change of the useful signal is approximately known beforehand, the proposed method offers a simple way to eliminate interference without having to resort to a computationally intensive Fourier transform unit.

[0013] The method according to the invention is suitable, for example, for signals whose value changes continuously. Continuous change is to be understood in the mathematical sense that there are no abrupt changes in the signal. This applies, for example, to signals that are measured values ​​of continuously changing quantities.

[0014] The method is also well suited for signals whose value changes abruptly, if the time of the change is predictable within certain limits, especially if the changes occur periodically.

[0015] It is particularly well-suited for digital signals or analog signals with pronounced edges. For example, a Hall sensor can detect the rotational speed of a shaft using a trigger wheel. Trigger wheels have teeth that are easily detected using magnetic sensing because the proportionality of the induced voltage to the first derivative of the field change generates an edge peak. This peak does not need to be measured in its absolute value, but can simply be registered as present / absent and thus only needs to be counted.

[0016] One prerequisite for the method is that a suitable sampling rate interval can be selected. This requirement is easily met for many applications, since a measurement can, for example, be based on previously measured values, and many physical quantities do not change abruptly. If no past measurements are available that can be meaningfully used, assumptions can be made plausible. For example, the rotational speed of a stationary motor can be assumed to be zero.

[0017] The sampling rate interval can be fixed permanently. For example, it can be hardware-coded or set once at the start of the process, without subsequent changes. Alternatively, the sampling rate interval can be controlled or regulated. Resetting can occur after each measurement or be linked to predetermined time intervals, measurement results, and / or external conditions.

[0018] The process also defines quantization levels. Their number and magnitude depend on the application; two quantization levels are sufficient for digital signals.

[0019] According to the inventive method, the number of changes in the quantization level is determined. A change in the measured value only occurs when a certain threshold is exceeded, so that the measured value is assigned to a different quantization level. This reliably eliminates noise components and random fluctuations.

[0020] Interference affecting the electrical signal is often due to long cables in many applications. This interference typically manifests in the signal waveform as the first derivative of the interference, resulting in voltage spikes. The invention takes advantage of the fact that these voltage spikes are relatively short and cause no change or only a double change in the quantization interval. If the sampling interval is significantly larger, for example, an order of magnitude larger, than the duration of the average voltage spikes caused by interference radiation, a double change in the quantization stage can be reliably attributed to interference. Therefore, the inventive method discards all even-numbered changes in a quantization stage within the sampling interval.

[0021] In a further development process, if there is an odd number of quantization level changes within a sampling rate interval, all subsequent quantization level changes within that interval are discarded. Thus, the process only determines whether there is an odd or even number of quantization level changes within the sampling rate interval. Consequently, it is irrelevant whether a voltage peak due to interference occurs before or after the change in the desired signal.

[0022] The proposed method is particularly suitable for digital signals. However, it is also suitable for analog signals that are quantized into two or more quantization levels after transmission through the conductor.

[0023] Due to its high robustness, the method according to the invention allows the use of comparatively long cables without requiring additional shielding. This also opens up possibilities for placing the signal-evaluating control unit further away, allowing for a more flexible installation location in an internal combustion engine, thus reducing the control unit's exposure to heat or enabling better packaging.

[0024] In addition to voltage spikes caused by EMC interference, noise further complicates the evaluation of edge peaks. The signal quality can be improved by filtering it beforehand. A low-pass filter, for example, can be used for this purpose. Furthermore, hysteresis can be addressed using a logic filter that evaluates the individual edge patterns.

[0025] In a training course, if two consecutive sampling intervals with odd number of quantization level changes occur, the number of quantization level changes in the current and preceding sampling intervals is set to zero. This is based on the assumption that the measured value either fluctuates around the threshold separating the quantization levels or that the two edges of a disturbance variable are randomly distributed across two different sampling intervals.

[0026] The proposed method is particularly suitable for a control unit of an electric or hydraulic camshaft adjuster of an internal combustion engine, which executes the process steps. The internal combustion engine is designed as a reciprocating piston engine and, in a basic structure known per se, comprises a crankshaft and at least one camshaft. An electromechanically adjustable actuator is provided for adjusting the camshaft; this actuator is, for example, a three-shaft transmission, in particular a wave gear transmission. The camshaft adjuster can be operated as follows: - The angular position of the crankshaft is continuously determined, with incremental detection of angular changes starting from a detected reference angular position; - a reference position of the camshaft is detected, in particular with the help of a trigger disc; - The current angular position of the camshaft and the crankshaft are calculated from the recorded angular changes of the rotor; - The difference between the two mentioned angular positions, that is, the phase difference between the crankshaft and the camshaft, is calculated and used to control the electric motor that drives the adjusting shaft.

[0027] The reference position of the camshaft is detected, with the corresponding signal typically already being provided to the engine control unit of the internal combustion engine. This signal contains edge peaks that can be counted.

[0028] In an advantageous embodiment, the angular changes of the crankshaft are detected with a finer resolution than the angular changes of the electric motor's rotor, which is fixedly coupled to the adjusting shaft of the actuator. Due to the given positive or negative reduction ratio of the actuator, a very fine resolution of the camshaft's angular position is possible.

[0029] In a preferred method, the angular positions of both the crankshaft and the rotor of the electric motor, which lie between two positions discretely distinguishable from one another by means of sensor signals, are approximately determined computationally by temporal extrapolation. This assumes that during the period to which the interpolation refers, the shaft in question, i.e., the motor shaft of the electric motor or the crankshaft, rotates at a practically constant speed.

[0030] The internal combustion engine comprises a crankshaft, at least one camshaft adjustable electromechanically via an actuator, in particular a wave gear, an engine control unit and a camshaft control unit provided for controlling an actuator motor, namely an electric motor, which actuates the actuator, wherein the engine control unit is connected to a device for detecting the angular position of the crankshaft and the camshaft control unit is connected to the engine control unit.and wherein the only means for detecting the angular position of the camshaft are a device for detecting a reference position of the camshaft to be adjusted and a device for detecting the angular position of the shaft of the actuator motor, and the camshaft control unit is designed to determine the phase position of the camshaft in relation to the crankshaft on the basis of the information provided by these devices in combination with the detected angular position of the crankshaft and the transmission ratio of the actuator.

[0031] The electric actuator of the camshaft adjuster is, for example, a permanent magnet synchronous motor. This electric motor has, for instance, four or six pole pairs. Changes in the angular position of the electric motor's rotor can be detected, for example, using Hall sensors.

[0032] According to one possible configuration, the engine control unit includes a memory that stores the flanks of a crankshaft trigger wheel detected during crankshaft rotation. By detecting both rising and falling flanks, a higher resolution can be achieved compared to detecting only similar flanks. Furthermore, a verification mechanism can be implemented to ensure the recorded data is free of logical inconsistencies. This type of signal processing is also feasible for the camshaft control unit. Similarly, the camshaft control unit includes a memory.

[0033] Within the sampling rate interval, a maximum of one pulse is evaluated; all other pulses are discarded. The following applies: If an even number of pulses occur (e.g., rising-falling, falling-rising, falling-rising-falling-rising...), then this is considered a disturbance, and these pulses are completely ignored. If an odd number of pulses occur (e.g., rising, rising-falling-rising, rising-falling-rising-falling-rising, falling-rising-falling-rising-falling...), then only the first edge in each direction is evaluated; subsequent edges are discarded.

[0034] The evaluation of edge patterns from the useful signal and the suppression of edge peaks caused by disturbances works particularly reliably when the sampling rate interval is not chosen to be constant, but rather tailored to the expected signal sequence. Therefore, a further development proposes that the sampling rate interval be dependent on the rotational speed of the shaft being measured. The sampling rate interval must be chosen such that it is smaller than the expected minimum duration between two useful pulses and larger than the typical duration of a disturbance.

[0035] This simple method of evaluation in the control unit, involving the counting of pulses, frees up computing resources that can be used for other tasks. For example, a multiple edge angle calculation of the camshaft and crankshaft can be performed to enable a fallback calculation in case of loss of the crankshaft sensor signal or one of the Hall signals.

[0036] In a training course, it is suggested to perform a calculation for multiple cam flanks (rising and falling) and / or a partial selection of signals (e.g., using only every second or third value). For example, this increases the chance of obtaining a plausible measurement in the case of short interference events. Example of implementation

[0037] The invention is explained in more detail below with reference to an exemplary embodiment. This embodiment relates to a control unit for an electric camshaft adjuster and is shown schematically in the figures. These figures show: Fig. 1 Components of an internal combustion engine with electromechanical camshaft adjustment in an overview diagram, Fig. 2 the interaction between an engine control unit and a camshaft control unit of the internal combustion engine, Fig. 3 the relationship between measurements on the crankshaft and a camshaft of the internal combustion engine, Fig. 4a a schematic, analog useful signal with interference signals, Fig. 4b a schematic, digital signal with interference signals, Fig. 5a-h different possible signal patterns that can occur within a sampling rate interval.

[0038] A in Fig. 1. An internal combustion engine, designated as an inline engine and marked with reference numeral 1, comprises a crankshaft 2 and two camshafts 3, 4, namely an intake camshaft 3 and an exhaust camshaft 4. In contrast to the illustrated embodiment, the internal combustion engine could also be a reciprocating piston engine of another design, for example a V-engine, which has two intake and two exhaust camshafts.

[0039] The camshafts 3, 4 are driven by the crankshaft 2 via chain drives 5, 6. Each camshaft 3, 4 is adjustable by means of an electromechanical camshaft adjuster 7, 8. The camshaft adjuster 7, 8 each has a three-shaft gear unit 9, 10 constructed as a wave gear unit. An input shaft of the adjustment unit 9, 10 is driven by the chain drive 5, 6. The output shaft of the adjustment unit 9, 10 is non-rotatably connected to the camshaft 3, 4 to be adjusted. A third shaft of each adjustment unit 9, 10 can be driven by an electric motor 11, 12 belonging to the respective camshaft adjuster 7, 8. The shaft marked 29 ( Fig.2) The motor shaft of the electric motor 11, 12, on which a rotor 28 is mounted, is coupled to the third shaft of the actuator 9, 10 in a rotationally fixed manner, optionally via a compensating coupling. In the exemplary embodiment, the so-called third shaft is an inner ring of a wave generator of the actuator 9, 10, which is designed as a wave gear.

[0040] The electric motors 11 and 12 are connected to a camshaft control unit 17 via connecting lines 13 and signal lines 14. The connectors of the electric motors 11 and 12 for the connecting lines 13 are designated 15, and the connectors for the signal lines 14 are designated 16. These lines 13 and 14 are connected to a connector 18 of the camshaft control unit 17. Hall signals, acquired by Hall sensors (not shown), are transmitted via the signal lines 14 and provide information about changes in the angular position of the rotor 28. The Hall sensors are part of a rotor position detection device designated 44.

[0041] The camshaft control unit 17 is connected to the engine control unit 21 of the internal combustion engine 1 via a data bus 19, namely a CAN bus, and a signal line 20. A crankshaft sensor 23 is connected to the engine control unit 21 via a crankshaft line 22. The crankshaft sensor 23 senses a crankshaft trigger wheel 27, which is fixedly connected to the crankshaft 2. Furthermore, sensors 24 and 25 are connected to the engine control unit 21, each of which interacts with a trigger disc 26 connected to a camshaft 3 or 4.

[0042] The Fig.Figure 2 illustrates data processing operations in the engine control unit 21 (left) and in the camshaft control unit 17 (right). As shown in the illustration, the signal generated by the trigger disk 26 is processed within the engine control unit 21. In the exemplary embodiment, the trigger disk 26 schematically has a single projection 32. One flank of the projection 32 is labelled 33. The flank 33 of the trigger disk 26 provides a camshaft trigger in a manner known per se. A logical connection is established between the camshaft trigger and the scanning of the crankshaft trigger wheel 27.

[0043] The crankshaft trigger wheel 27 has teeth 35, which, together with an adjacent gap located between two teeth 35, each cover an angle of 6°. By omitting two teeth, a recess 36 is formed, with the first tooth 35 adjacent to the recess 36 representing a reference mark 34. The signal detected using the reference mark 34 is also referred to as the TD signal. A copy of this TD signal, to which a further mark may be added, is transmitted from the engine control unit 21 to the camshaft control unit 17 via the signal line 20. Within the camshaft control unit 17, the TD signal, which indicates a reference angle position of the crankshaft, is logically linked with characteristics of the electric motor 11, 12.

[0044] In Fig.Figure 2 shows permanent magnets 30 and windings 31 of the electric motor 11, 12. A possible path of Hall signals HSA, HSB, HSC, which provide information about changes in the angular position of the rotor 28, is shown schematically. Each combination of the Hall signals HSA, HSB, HSC corresponds to a bit pattern, in this embodiment the bit patterns 010, 011, 001, 101, 100 and 110.

[0045] How Fig.As shown in Figure 3, each tooth 35 has a rising flank Fs and a falling flank Ff. It can be assumed with good approximation that the crankshaft speed does not change when rotating by one tooth 35. Therefore, the time interval, which represents a partial period during the rotation of the crankshaft 2 from one tooth 35 to the next, can be used to calculate any angular position of the crankshaft 2 lying between two teeth 35. In this way, the camshaft reference position, that is, the angular position of the camshaft 3, 4 at which flank 33 is detected, can also be assigned to a precise angular position of the crankshaft 2.

[0046] Similarly, angular positions of the camshaft 3, 4 are extrapolated using bit patterns generated during the operation of the electric motor 11, 12. This calculation also assumes that the motor shaft 29 rotates within the relevant angular range at an approximately constant angular velocity.

[0047] From the Fig. A signal S is generated at 4a, which is received and processed by one of the control units 17, 21. The applied voltage is plotted over time. The signal S consists of a useful signal 45, which is superimposed with interference pulses 46a to 46h. Due to the short duration of the interference signals, they induce voltages that, because of their short duration, are represented as peaks.

[0048] The electrical signal S is analog and is first transformed into a digital signal S'. Two quantization levels, Q1 and Q2, are provided for the digital signal; these are also designated as low and high. The assignment to one of the quantization levels is as follows:

[0049] If the signal voltage is above an upper threshold U2, it is assigned to quantization stage Q2 regardless of its actual amplitude. Similarly, if the signal voltage is below a lower threshold U1, it is assigned to quantization stage Q1. Therefore, for example, interference pulses 46b and 46h have no influence on the assignment to quantization stage Q1.

[0050] The region between thresholds U1 and U2 defines a transition region. If the voltage exceeds or falls below one of the thresholds over time, thus reaching the transition region, without exceeding or falling below the other threshold, no reassignment of the quantization level Q1 or Q2 occurs. The disturbance pulse 46e therefore does not cause a reassignment because the voltage does exceed the lower threshold U1, but then falls below it again without having first reached the level of the upper threshold U2.

[0051] All other interference pulses 46a, 46c, 46d, 46f and 46g are incorporated into the digital signal S'.

[0052] In the next step, the number of changes to the quantization stage Q1, Q2 within the respective sampling rate interval T is determined. All even-numbered changes to the quantization stage Q1, Q2 within the sampling rate interval T are discarded, so that only the first change is considered. As described below, this eliminates the noise pulses.

[0053] The interference pulse 46a, which occurs in the first sampling rate interval T, is in Fig. 5d is shown in more detail. It forms a double edge and thus a double change in the quantization level Q1, Q2, and is therefore not counted. The same applies to the perturbation pulses 46d and 46f, which are shown in Fig. Figure 5c is shown enlarged. They also form a double flank and are therefore not counted.

[0054] The noise pulse 46c lies in the region where the useful signal 45 changes from quantization stage Q2 to quantization stage Q1. The relevant sampling rate interval T is in Fig. Figure 5a illustrates this. There is an odd number of edges, so the first edge is recorded as a change in the quantization level from Q2 to Q1. The same applies to the reverse transition, which is shown in the Fig. 4a is provided with a flatter signal rise and the two interference pulses 46f, 46g. The corresponding sampling rate interval is in Fig. 5g is shown. The five flanks again form an odd number, so they count as a change in the quantization level Q1, Q2.

[0055] By reducing interference pulses 46a to 46h in this way, computing power is freed up in the control unit 17, 21, which is available for other calculations. Fig. Figure 3 shows a trigger disc 26, which is located in Fig.2 differs in that calculations are performed on three flanks. Using the values ​​shown there, the phase angle is calculated in the case of Arrow 1 2 × 0° - 0° = 0°, Arrow 2 2 × 15° - 30° = 0°, Arrow 3 2 × 90° - 160° = -20°.

[0056] This allows for redundancy, and also makes a control loop more stable. Reference symbol list 1 Internal combustion engine 2 Crankshaft 3 camshaft 4 camshaft 5 chain drives 6 chain drives 7 camshaft adjusters 8 camshaft adjusters 9 actuators 10 actuators 11 Electric motor 12 Electric motor 13 Connection cable 14 Signal line 15 Plug connectors for connecting cables 16 Connectors for signal lines 17 Camshaft control unit 18. Camshaft control unit connector 19 Data bus 20 Signal line 21 Engine control unit 22 Crankshaft line 23 Crankshaft sensor 24 Sensor 25 Sensor 26 trigger disc 27 Crankshaft trigger wheel 28 Rotor 29 Motor shaft 30 permanent magnets 31 windings 32 Survey 33 Flank 34 Reference marking 35 teeth 36 recess 37 Evaluation unit 38 ring storage units of the camshaft control unit 39 Storage area 40 storage area 41 Ring memory of the engine control unit 42 Memory area 43 Storage area 44 Rotor position detection device 45 Useful signal 46a - 46h Interference impulses Ff falling flank Fs rising flank HSA Hall signal HSB Hall signal HSC Hall signal S Signal S' digital signal T sampling rate interval TD TD signal P1 Arrow 1 P2 Arrow 2 P3 Arrow 3

Claims

[1] Method for processing an electrical signal (S) that is generated continuously over time and based on measured values ​​of a sensor (23, 24, 25) and transmitted via a line (20, 22) exposed to disturbances, having the following features: a) setting a sampling rate interval (T), b) setting quantization levels (Q1, Q2), c) assigning the electrical signal to a quantization level (Q1, Q2), d) determine the number of changes in the quantization level (Q1, Q2) in the sampling rate interval (T), e) discard all even-numbered changes of the quantization level (Q1, Q2) within the sampling rate interval (T). [2] Method according to claim 1, characterized by, that if there is an odd number of changes to the quantization stage (Q1, Q2) within a sampling rate interval (T), all subsequent changes to the quantization stage (Q1, Q2) within the sampling rate interval (T) are discarded. [3] Method according to claim 1 or 2, characterized by , that the electrical signal (S) is a digital signal or an analog signal which, after transmission through the line (20, 22), is quantized into exactly two quantization levels (Q1, Q2). [4] Method according to any one of the preceding claims, characterized by , that the electrical signal (S) is filtered. [5] Method according to any one of the preceding claims, characterized by, that in the case of two consecutive sampling rate intervals (T) with odd number of changes in the quantization level (Q1, Q2), the number of changes in the quantization level (Q1, Q2) in the current and preceding sampling rate interval (T) is set to zero. [6] Method according to any one of the preceding claims, characterized by , that the sensor (23, 24, 25) measures a constantly changing quantity. [7] Method according to claim 6, characterized by , that the sensor (23, 24, 25) scans a trigger wheel of a camshaft (3, 4) or a crankshaft (2). [8] Method according to any one of the preceding claims, characterized by , that the sampling rate interval (T) is controlled or regulated by the measured values. [9] Method according to any one of the preceding claims, characterized by , that the sampling rate interval (T) is less than 50µs. [10] Control unit (17, 21) for a camshaft adjuster (7, 8) or for an internal combustion engine (1) performing a method according to one of the preceding claims.

Images