Device and method for correcting a sensor signal

Abstract

The present invention relates to a device (10) for correcting a sensor signal, comprising a first input (11) for the sensor signal (12), a second input (13) for a reference sensor signal (14), a processor (15) designed to determine a corrected sensor signal (16) on the basis of a spatial transformation of the sensor signal (12) using the reference sensor signal (14), and an output (17) for the corrected sensor signal (16).

Description

[0001] Device and method for correcting a sensor signal

[0002] The present invention relates to devices and methods for improving the accuracy of sensor signals, in particular of redundantly designed torque and / or angle sensors.

[0003] Today's vehicle steering gears, especially those with electric power steering (EPS), can determine torque and angle signals from the driver's steering inputs. The torque and angle signals thus correspond to the steering inputs. Torque and angle sensors can be used for related measurements, which transmit the driver's steering input to an electronic control unit (ECU) to determine the assistance level.

[0004] Such sensors are inherently limited to a certain accuracy such as linearity and hysteresis, and the calibration accuracy of the sensors is also limited.

[0005] Therefore, the sensors remain with reduced accuracy for the application, necessitating the use of better and more costly sensors. These EPS systems typically include more than one sensor signal for safety and reliability / robustness reasons. In the context of ASIL (Automotive Safety Integrity Level), redundant sensors refer to the presence of multiple sensors monitoring or measuring the same function (e.g., steering input).

[0006] Against the background of systems with multiple redundant sensors, it can be seen as an object of the present invention to provide an improvement in the accuracy of the sensor signals.

[0007] This object is achieved by devices and methods according to the independent patent claims. Advantageous further developments are the subject of the dependent claims. According to a first aspect, the present invention provides a device for correcting a sensor signal. The device comprises a first input for a (first) sensor signal, a second input for a (second) reference sensor signal, a processor which is designed to determine a corrected sensor signal based on a spatial transformation of the sensor signal using the reference sensor signal and an output for the corrected sensor signal. By spatially transforming the sensor signal using the reference sensor signal, the corrected sensor signal can be obtained with improved accuracy. The spatial transformation can be selected depending on a signal shape of the sensor signal, for example with regard to hysteresis and / or.

[0008] A spatial transformation can include one or more translations (translations). A translation moves the position of an object along the x-, y-, or z-axis of the coordinate system. A spatial transformation can include one or more rotations (rotations). A rotation rotates the object by a specific angle around one or more axes. A spatial transformation can include one or more scalings. A scaling changes the size of an object, either uniformly in all directions or along specific axes. A spatial transformation can include one or more shears. A shear deforms an object by stretching or compressing it in one direction.

[0009] The sensor signal and the reference sensor signal can, for example, be sensor signals from redundant sensors that measure the same physical quantity (e.g., steering input, rotation, etc.). The redundant sensors can have the same or different measurement accuracies. One of the sensor signals is used as the reference sensor signal for the other. It is understood that embodiments of the present invention are not limited to two sensor signals; rather, the presented concept can be applied to any number of (redundant) sensor signals.

[0010] According to some embodiments, the processor of the device is configured to determine the corrected sensor signal based on averaging the sensor signal and a sensor signal transformed by means of the spatial transformation. Averaging can reduce an original measurement error of the sensor signal.

[0011] According to some embodiments, the spatial transformation comprises a rotation of a vector that has the reference sensor signal as a first vector component (e.g., x-component) and the sensor signal or an inverted version thereof as a second vector component (e.g., y-component). Thus, the spatial transformation allows the sensor signal to be mapped to a different point in 2- or 3-dimensional space.

[0012] According to some embodiments, the spatial transformation comprises an inversion of the sensor signal. If the sensor signal is considered as a y-value, the inversion of the sensor signal can correspond to a reflection of the sensor signal on the x-axis (reference sensor signal).

[0013] According to some embodiments, the spatial transformation comprises a rotation of the inverted sensor signal by a predetermined rotation angle or twist angle. The spatial transformation of the signal can comprise a transformation of the x-component (reference sensor signal) and the y-component (sensor signal). Each individual sensor measurement value (sensor signal) can be mapped to a new measurement point using the spatial transformation.

[0014] According to some embodiments, the predetermined angle of rotation or twist is 90°.

[0015] According to some embodiments, the spatial transformation provides a transformed sensor signal (or a coordinate of a transformed sensor signal) according to Sltrans = S2 * cosa + f CO rr * S1 mv * sina, where Sltrans is the transformed sensor signal, S2 is the reference sensor signal, S1inv is the inverted sensor signal, f CO rr represents a correction factor and a represents a predetermined rotation angle (relative to a transformed x-component). In some embodiments, the correction factor fcorr may be 1. During the transformation, the signal may be composed of an x-component (reference sensor signal) and a y-component (sensor signal).

[0016] According to some embodiments, the processor is configured to adjust the correction factor and / or the predetermined rotation angle such that a linearity and / or hysteresis error of the corrected sensor signal is reduced compared to the sensor signal. This can be done iteratively, for example.

[0017] According to some embodiments, the sensor signal indicates a first measured variable of a measurement object (e.g. handlebar, torsion bar, etc.) measured by a first sensor and the reference sensor signal indicates a second measured variable of the measurement object measured by a second sensor. The first measured variable can be, for example, a torque applied to a torsion bar. The second measured variable can be, for example, a twist angle of the torsion bar, from which the torque is determined. Accordingly, the sensor signal can indicate a torque of the measurement object (e.g. torsion bar) measured by a torque sensor and the reference sensor signal can indicate a twist angle of the measurement object (e.g. torsion bar) measured by an angle sensor.

[0018] According to a further aspect, the present invention provides a method for correcting a sensor signal. The method comprises receiving a sensor signal and a reference sensor signal, determining a corrected sensor signal based on a spatial transformation of the sensor signal using the reference sensor signal, and outputting the corrected sensor signal.

[0019] According to some embodiments, the sensor signal indicates a first measured variable (e.g., torque) of a measurement object (e.g., shaft, torsion bar) measured by a first sensor (e.g., torque sensor), and the reference sensor signal indicates a second measured variable (e.g., angle of rotation) of the measurement object (e.g., shaft, torsion bar) measured by a second sensor (e.g., angle sensor). According to some embodiments, the sensor signal and the reference sensor signal are received substantially simultaneously. That is, the first and second measured variables are measured substantially simultaneously.

[0020] According to some embodiments, the corrected sensor signal is determined based on an averaging of the sensor signal and the sensor signal transformed by means of the spatial transformation.

[0021] According to some embodiments, the spatial transformation comprises an inversion of the sensor signal and a rotation of the inverted sensor signal by a predetermined rotation angle (e.g., 90°).

[0022] They show:

[0023] Fig. 1 is a schematic representation of a device for correcting a sensor signal according to an embodiment;

[0024] Fig. 2 shows a more detailed representation of sensor signal processing according to an embodiment;

[0025] Fig. 3 is a schematic representation of a sensor signal relative to a reference sensor signal;

[0026] Fig. 4 is a schematic representation of an inverted sensor signal relative to the reference sensor signal;

[0027] Fig. 5 is a schematic representation of a sensor signal transformed by means of spatial transformation;

[0028] Fig. 6 shows an example of a transformed and additionally scaled sensor signal; Fig. 7 shows a schematic representation of a corrected sensor signal based on averaging the sensor signal and the transformed sensor signal;

[0029] Fig. 8 is an enlarged view of an improved linearity of the corrected sensor signal; and

[0030] Fig. 9 is an enlarged view of a reduced hysteresis of the corrected sensor signal.

[0031] Fig. 1 shows a schematic representation of a device 10 for correcting a sensor signal 12.

[0032] The device 10 comprises a first input 11 for the sensor signal 12 and a second input 13 for a reference sensor signal 14. The device 10 also comprises a processor 15 configured to determine a corrected sensor signal 16 based on a spatial transformation of the sensor signal 12 using the reference sensor signal 14. The device 10 further comprises an output 17 for the corrected sensor signal 16.

[0033] The device 10 or the processor 15 can, for example, be a computing unit, such as a control unit in a motor vehicle. A control unit in a vehicle, also referred to as a control unit or electronic control unit (ECU), is an electronic device used in modern vehicles to control and regulate various functions. These control units are part of the so-called vehicle electronics system and work together to ensure the smooth and efficient functioning of the vehicle. A control unit can work together with a variety of sensors and actuators in the vehicle to continuously acquire data, calculate appropriate control commands, and regulate and optimize the vehicle's functions.Communication between the control units and other components in the vehicle can take place via bus systems such as the Controller Area Network (CAN bus) to ensure efficient and coordinated cooperation. The first input 11 can also be understood as the first interface to a first sensor. The first sensor can, for example, be a torque sensor installed in the vehicle, which serves to measure the torque or torsional moment acting on a shaft or a rotating component (e.g. torsion bar) in the vehicle. Torque is a physical quantity that indicates how much rotational force is exerted on an axle or a rotating body. The sensor signal 12 can therefore, for example, comprise a first torque sensor signal.

[0034] There are different types of torque sensors that work differently depending on the application and measuring principle, such as:

[0035] • Strain-gauge torque sensors: These sensors use strain gauges mounted on a shaft or element to measure torque. When torque is applied, the length, or elongation, of the strain gauges changes, resulting in a change in electrical resistance proportional to the torque. The change in resistance is then converted into an electrical signal and measured.

[0036] • Torque transducers: Torque transducers operate on the principle of dynamic torque measurement. They consist of a rotating component on which the torque to be measured acts. This moving element is connected to a measuring system that directly measures the torque and determines the torque.

[0037] • Capacitive torque sensors: These sensors rely on the change in capacitance between two capacitor plates, which is influenced by the torque. The torque causes a displacement of the plates, resulting in a change in capacitance and thus in the measured torque.

[0038] • Magnetostrictive torque sensors: These sensors use a magnetostrictive material that changes its magnetic properties when subjected to a mechanical load such as torque. The magnetic properties are monitored and serve as a measure of the torque. Dynamic torque sensors can measure both the torque and the angle at which the shaft or rotating component moves. They can therefore combine torque and angle sensors. Such sensors can be used in applications where both the torque and the change in angle during rotation are of interest. For example, magnetostrictive or capacitive torque sensors can be used to simultaneously measure the angle of the rotating part.

[0039] The second input 13 can also be understood as a second interface to a second sensor. The second sensor can, for example, be a torque sensor redundant to the first sensor, which serves to redundantly measure the torque or torsional moment acting on the shaft or the rotating component. The second sensor can, for example, also be an angle sensor, which serves to measure the angle of rotation of the shaft or the rotating component. The angle of rotation depends on the torque. The angle sensor can, for example, be installed in a dynamic torque sensor.

[0040] There are different types of angle sensors that use different measuring principles:

[0041] • Potentiometric angle sensors: Potentiometric angle sensors are based on voltage division across a resistor. The angle is determined by the position of a movable wiper or sliding track on the resistor, which changes the output voltage accordingly.

[0042] • Optical angle sensors: Optical angle sensors use light or infrared rays and receive the reflected signals to determine the angle. There are various optical measurement methods, such as rotary encoders, which generate a discrete number of pulses per revolution, or optical absolute angle sensors, which provide the absolute angle in a continuous range of values.

[0043] • Magnetic angle sensors: Magnetic angle sensors use magnetic fields and Hall-effect sensors or magnetoresistive elements to measure the angle. These sensors are often used in combination with magnetic markers on the rotating shaft. • Inductive angle sensors: Inductive angle sensors measure the angle based on the change in an induced voltage in a coil when a magnetic field changes. These sensors are often used for applications with very high demands on robustness and reliability.

[0044] Sensor signal 12 and reference sensor signal 14 thus relate to different sensor signals of physical measurement quantities that are, however, related to one another, such as torque - torque, torque - angle of rotation, or angle of rotation - angle of rotation, to name just a few examples. Sensor signal 12 and reference sensor signal 14 both relate to the same measurement object, such as the shaft, torsion bar, or rotating component. While sensor signal 12 and reference sensor signal 14 can have approximately the same accuracies, embodiments are also conceivable in which sensor signal 12 and reference sensor signal 14 have different accuracies. For example, reference sensor signal 14 could have a higher accuracy than sensor signal 12.

[0045] The spatial transformation performed by processor 15 may include one or more translations (translations). A translation moves the position of an object along the x-, y-, or z-axis of the coordinate system. The spatial transformation may additionally or alternatively include one or more rotations (rotations). During a rotation, the object is rotated by a specific angle about one or more axes. The spatial transformation may additionally or alternatively include one or more scalings. Scaling changes the size of an object, either uniformly in all directions or along specific axes. The spatial transformation may include one or more shears. Shearing deforms an object by stretching or compressing it in one direction.

[0046] An embodiment of the device 10 or the processor 15 will now be described in more detail with reference to Fig. 2.

[0047] First, the processor 15 receives the sensor signal (S1) 12 from a first sensor and the reference sensor signal (S2) 14 measured independently by a second sensor. In the illustrated embodiment, the sensor signal S1 is a first torque sensor signal from a first torque sensor, and the reference sensor signal S2 is a second torque sensor signal from a second (redundant) torque sensor. The reference sensor signal S2 can (but does not have to) have a higher accuracy than the sensor signal S1. The reference sensor signal S2 preferably does not have the same error characteristics as the sensor signal S1. For example, the reference sensor signal S2 could also be an angle sensor signal related to the first torque sensor signal, which indicates a twist angle caused by the applied torque.

[0048] As indicated in Fig. 3, at least the sensor signal S1 applied to the first input 11 may be subject to a hysteresis and / or linearity error. Such errors are to be reduced by the proposed invention.

[0049] In a further step (see Fig. 4), the sensor signal S1 is inverted by the processor 15. If the sensor signal S1 is considered a y-value (and the reference sensor signal S2 as the x-value), the inversion of the sensor signal S1 can correspond to a reflection of the sensor signal on the x-axis. Mathematically, the signal comprises a pair of values ​​(x = the quantity to be measured, y = the measured quantity). A rotation or reflection of the signal is mathematically feasible even without a reference signal. If the quantity to be measured / reference is neglected, usually only the y-component is referred to as the "signal." For example, a torque sensor that outputs a torque signal is not questioned for its reference (this would be ideal); only upon closer inspection, particularly during analysis for signal enhancement, does the reference "x" come into focus. In the following description of the transformation, it is important to consider and label the assignment of the signal coordinates.

[0050] During inversion, the sensor signal pair (x = S2, y = S1) becomes (x = S2, y = S1inv), with S1inv = -S1. The S1 signal is mirrored and also exhibits an inverted behavior for linearity, hysteresis, and other accuracy contributions. To apply the graph in Fig. 4 to the original sensor signals S1 for optimization, the inverted S1 inv signals can be transformed to the known independent sensor reference S2 (e.g., torque sensor, angle sensor) using a known spatial transformation of the coordinate description.

[0051] During a rotation, a point P = (x, y) passes into the point P' = (x', y'), whose position vector (written as a column vector) one from by applying the formula:

[0052] If we consider the reference sensor signal S2 as the x-value and the inverted sensor signal S1 mv as the y-value of a pair of values, the transformed sensor signal Sltrans (or a transformed coordinate) results according to

[0053] S1 trans = S2 * cosa + S1 inv * sina, where a represents a predetermined rotation angle. Accordingly, the transformed reference sensor signal S2trans (or the corresponding transformed coordinate) can be calculated according to

[0054] S2trans = S2 * sina - S1 inv * COS(X). According to this embodiment, the inverted sensor signal S1 inv is rotated positively within the coordinate system (clockwise) with its signal shape. If the inverted sensor signal S1i nv For example, rotated by a rotation angle of a = 90°, the graph, based on the transformed reference sensor signal S2t rans, as shown in Fig. 5 (Sltrans is also referred to as Tn'inv in Fig. 5). It should be noted that, in principle, other rotation angles other than 90° are also conceivable.

[0055] According to a further embodiment, the spatial transformation into a scaled sensor signal S1 trans, seal can be carried out according to

[0056] S1 trans, seal — S2 * cosa + fcorr S1 inv sina, where S2 is the reference sensor signal (x-value), S1mv (y-value) is the inverted sensor signal, f CO rr t 1 represents a correction factor and the predetermined rotation angle. An example of a transformed and scaled sensor signal Sl trans, scai is shown in Fig. 6 (Sltrans, scai is also referred to as Tn'mv.sciaed in Fig. 6). The scaling occurs "automatically" when the transformed y-coordinates are projected onto the same reference system (x-axis) as the sensor signals.

[0057] In a further step, the processor 15 averages the sensor signal S1 with the transformed sensor signal Sl trans according to (S1 + S1 trans) / 2 or (S1 + S1 trans, scai) / 2. This averaging process produces the corrected sensor signal (Slkorr) 16 with improved linearity and hysteresis at the output of the processor 15.

[0058] The improvements in the sensor signal Slkorr can be seen in the enlarged views of Figures 8 and 9. The error curves of the signals demonstrate the improved accuracy across the entire sensor range.

[0059] The present invention describes a concept that improves the accuracy of the sensor signal S1 with an inherently existing, independent reference sensor signal S2 (e.g. angle sensor within the torque and angle sensor system) by specifically combining and calculating the signals. The reference sensor should be an "external sensor" that does not have the same error characteristics as the sensor to be improved. Higher accuracy is always desirable. For example, if torque sensors are always equally "poor" with similar linearity curves, a referenced angle sensor with a different linearity curve can reveal the error of the torque sensors. The error that is added to the sensor signal in step 1 is the same error that is also added to the transformation in step 3 and could be calculated out to a certain extent. The basis of this method is the use of the shape of the sensor signal S1 to improve the accuracy.The signal shape is transformed in such a way that a generated correction signal Sltrans or S1 trans, seal is added to improve the sensor signal S1.

[0060] The invention comprises a non-contact position sensor principle with n (> 2) independent sensor signals to meet safety and robustness requirements. It can be assumed that the sensor signals are calibrated, connected to, and controlled by the system control unit.

[0061] The exemplary embodiments described and shown in the figures are selected only as examples. Different exemplary embodiments can be combined with each other in their entirety or with regard to individual features. Furthermore, one exemplary embodiment can be supplemented by features of another exemplary embodiment.

[0062] Furthermore, the following claims are hereby incorporated into the Detailed Description, each claim may stand on its own as a separate example. While each claim may stand on its own as a separate example, it should be noted that although a dependent claim may refer to a particular combination with one or more other claims in the claims, other examples may include a combination of the dependent claim with the subject matter of any other dependent or independent claim. Such combinations are explicitly suggested herein unless it is stated that a particular combination is not intended. Furthermore, features of a claim for any other independent claim are also intended to be included, even if that claim is not made directly dependent on the independent claim.

[0063] Reference numeral Device for correcting a sensor signal First input Sensor signal Second input Reference sensor signal Processor Corrected sensor signal Output

Claims

Patent claims 1. A device (10) for correcting a sensor signal, comprising a first input (11) for the sensor signal (12); a second input (13) for a reference sensor signal (14); a processor (15) which is designed to Determining a corrected sensor signal (16) based on a spatial transformation of the sensor signal (12) using the reference sensor signal (14); and an output (17) for the corrected sensor signal (16).

2. Device (10) according to claim 1, wherein the processor (15) is designed to determine the corrected sensor signal (16) based on an averaging of the sensor signal (12) and a sensor signal transformed by means of the spatial transformation.

3. Device (10) according to one of the preceding claims, wherein the spatial transformation comprises a rotation of a vector having the reference sensor signal (14) as a first vector component and the sensor signal (12) or an inverted version thereof as a second vector component.

4. Device (10) according to one of the preceding claims, wherein the spatial transformation comprises an inversion of the sensor signal (12).

5. The device (10) of claim 4, wherein the spatial transformation comprises a rotation of the inverted sensor signal by a predetermined rotation angle.

6. Device (10) according to claim 5, wherein the predetermined angle of rotation is 90°.

7. Device (10) according to one of the preceding claims, wherein the spatial transformation comprises a coordinate of a transformed sensor signal corresponding to S1 trans —S2 * COS(X + fcorr * S1 inv * sina, where S1 trans is the coordinate of the transformed sensor signal, S2 is the reference sensor signal, S1inv is the inverted sensor signal, fcorr is a correction factor and a is a predetermined rotation angle.

8. Device (10) according to claim 7, wherein the processor (15) is designed to adjust the correction factor and / or the predetermined rotation angle such that a linearity and / or hysteresis error of the corrected sensor signal (16) is reduced compared to the sensor signal (12).

9. Device (10) according to one of the preceding claims, wherein the sensor signal (12) indicates a first measured variable of a measurement object measured by a first sensor and the reference sensor signal (14) indicates a second measured variable of the measurement object measured by a second sensor.

10. Device (10) according to claim 9, wherein the sensor signal (12) comprises a torque of the measuring object measured by a torque sensor and the reference sensor signal (14) comprises a rotation angle of the measuring object measured by an angle sensor.

11. Device (10) according to one of the preceding claims, wherein the sensor signal (12) and the reference sensor signal (14) are sensor signals from redundant sensors.

12. Device (10) according to one of the preceding claims, wherein a measurement accuracy of the reference sensor signal (14) is greater than a measurement accuracy of the sensor signal (12).

13. A method for correcting a sensor signal, comprising receiving a sensor signal and a reference sensor signal; Determining a corrected sensor signal based on a spatial transformation of the sensor signal using the reference sensor signal; and Output the corrected sensor signal.

14. The method according to claim 13, wherein the sensor signal indicates a first measured variable of a measurement object measured by a first sensor and the reference sensor signal indicates a second measured variable of the measurement object measured by a second sensor.

15. The method according to claim 13 or 14, wherein the sensor signal and the reference sensor signal are received substantially simultaneously.

16. The method according to any one of claims 13 to 14, wherein the corrected sensor signal is determined based on an averaging of the sensor signal and the sensor signal transformed by means of the spatial transformation.

17. The method according to any one of claims 13 to 16, wherein the spatial transformation comprises an inversion of the sensor signal and a rotation of the inverted sensor signal by a predetermined rotation angle.

Images