WO2022057963A1

WO2022057963A1 - Method for detecting an angular position, and detection system

Info

Publication number: WO2022057963A1
Application number: PCT/DE2021/100519
Authority: WO
Inventors: Jie Zhou
Original assignee: Schaeffler Technologies AG & Co. KG
Priority date: 2020-09-18
Filing date: 2021-06-17
Publication date: 2022-03-24
Also published as: DE102020124419B4; CN116490750A; DE102020124419A1

Abstract

The invention relates to a method (100) for detecting an angular position (a) of a rotary component (22) rotatable about an axis of rotation (24) via a sensor unit (26), which has a fixed sensor element (30) and a rotary element (28) rotatable with respect to the fixed sensor element and together with the rotary component (22), wherein: the sensor element (30) outputs to an evaluation unit (102) a first and second sensor signal (S₁, S₂) each of which is dependent on the angular position, and the evaluation unit calculates therefrom an angle signal indicating the angular position and an angle error (e); an orthogonal error (φ) describing the orthogonal deviation between the first and second sensor signals (S₁, S₂) and assigned to the angle error (ϵ) is calculated by detecting a first angle gradient (G₁) of the angle signal when the rotary component (22) is in a first rotary position and by detecting a second angle gradient (G2) of the angle signal when the rotary component (22) is in a second rotary position different from the first rotary position, then calculating the orthogonal error (φ) on the basis of the first and second angle gradients (G₁, G₂) and adjusting the angular position (α_f) by this orthogonal error (φ). The invention also relates to a detection system.

Description

Angular position detection method and detection system

description introduction

The invention relates to a method for detecting an angular position according to the preamble of claim 1. The invention also relates to a detection system for detecting an angular position.

A method for detecting an angular position of a rotary component is known, for example, from WO 2018/219388 A1. A method for detecting an angular position of a rotary component that can be rotated about an axis of rotation is described therein, in which the angular position of the rotary component is recorded by a sensor system arranged radially at a distance from the axis of rotation. A magnetic ring arranged firmly and concentrically on the rotating component causes a magnetic field that changes relative to the sensor system and is detected by the sensor system, with a signal picked up by the sensor system being evaluated with regard to the angular position. The signal picked up by the sensors is evaluated with regard to amplitude information of the magnetic field and a correction parameter is determined from the amplitude information, by means of which an angular error of the angular position picked up from the signal of the sensors is determined. The angular error is then used to correct the angular position determined from the signal emitted by the sensors.

The object of the present invention is to detect an angular position of a rotary component more accurately and quickly. The angular position of the rotary component should be able to be determined with as little calculation effort as possible. Furthermore, the angular position should be detected more cost-effectively.

At least one of these objects is achieved by a method for detecting an angular position of a rotary component that can be rotated about an axis of rotation via a sensor unit that has a fixed sensor element and a rotary element that can be rotated relative to it and together with the rotary component, with the sensor element in each case having a first and second outputs a sensor signal to an evaluation unit, which uses this to calculate an angle signal indicating the angular position and an angle error, wherein an orthogonal error describing the orthogonal deviation between the first and second sensor signals and associated with the angle error is calculated by a first angle gradient at a first rotary position of the rotary component of the angle signal and a second angular gradient of the angle signal is detected when the second rotational position of the rotating component differs from the first rotational position, the orthogonal error is then calculated as a function of the first and second angular gradient and the angular position is also corrected for this orthogonal error.

This allows the angular position of the rotary member to be detected more accurately. The measurement error can be reduced with little calculation effort.

The orthogonal error can be triggered by a tilting of the rotary component about the axis of rotation.

The sensor unit and the rotary component can be arranged in a vehicle. The rotary component can be assigned to a parking lock device. The parking lock device may include a parking lock actuator having the rotary member. The parking lock actuator can have an electric motor with a stator and a rotor that can rotate with respect to this. The rotary component can be connected to the rotor in a rotationally fixed manner. The electric motor can be drivingly connected to a hydraulic pump. The sensor unit can detect the angular position of the rotor relative to the stator. As a result, the commutation of the electric motor can be improved.

The rotary component can be arranged on an actuator for actuating a clutch of the vehicle.

The rotating member and the rotating member may be concentrically rotatable.

The sensor unit can be designed as an angle sensor. The sensor element can be a Hall sensor. The sensor unit can have at least two sensor elements. The sensor unit can have four sensor elements.

The rotating element can be a magnetic ring. The rotating element can be a permanent magnet. The rotating element can be diametrically magnetized.

The sensor unit can output a first sensor signal assigned to a first detection position and a second sensor signal assigned to a second detection position, which is perpendicular to the first detection position about the axis of rotation. The sensor element can be arranged radially offset with respect to the axis of rotation. As a result, the first and/or second sensor signal can be a periodic signal, caused in particular by the rotation of the rotary element. The first sensor signal can be a cosine signal and the second sensor signal can be a sine signal.

The angle signal can be calculated from the first and second sensor signals by applying an atan2 function. The proposed procedure can then be carried out.

The angle error can have an amplitude error describing the amplitude difference between the first and second sensor signal and/or an offset error. These errors can already be reduced or eliminated before the orthogonal error is calculated. Before the proposed method is used, the first and second sensor signals can already be corrected for possible offset errors and/or possible amplitude errors, for example by using a max-min method.

The proposed method can be carried out during an application-side operation of the rotary component, for example in the vehicle. Before detecting the first and second angle gradients, the angle signal can be smoothed using a moving average method.

In a preferred embodiment of the invention, it is advantageous if the first angular gradient and the second angular gradient are detected by rotating the rotary component at a constant rotational speed. The rotating member may be free from driving torque. During the capture of the first and second Angular gradients, the moment of inertia of the components involved in the rotational movement, including the rotary component, can cause a constant, this includes a virtually constant rotational speed with deviations of less than 10%, in particular less than 5%, particularly preferably less than 1%, rotational speed.

The angular gradient G* can be calculated as follows

If the first and second angular gradients are recorded while the rotational speed remains the same, the angular gradient G can be calculated in a simplified manner according to

G = Δα

If the angular gradient is calculated according to this relationship, the calculation effort and the numerical noise caused by the division can be reduced.

In a preferred embodiment of the invention, it is advantageous if the first angular gradient and the second angular gradient are detected by rotating the rotating component through at least 90°, in particular around the axis of rotation. As a result, the two measured values can be recorded and at the same time the method can be carried out more quickly. The rotary member can be rotated at least one full revolution to capture the first and second angular gradients.

In a preferred embodiment of the invention, it is advantageous if the first rotational position is 45° and the second rotational position is 135°. The first angular gradient G ₁ can be detected at 135° and the second angular gradient G ₂ can be detected at 45°.

A preferred embodiment of the invention is advantageous in which an error parameter is calculated as a function of a quotient from the first and second angle gradient. The error parameter γ can be calculated as follows

In a preferred embodiment of the invention, it is advantageous if the orthogonal error φ is calculated as a function of the error parameter γ. The orthogonal error φ can be calculated as follows

In a special embodiment of the invention, it is advantageous if the angular error ϵ(α, φ) is calculated as a function of the orthogonal error φ. The angular error ϵ(α, φ) can be calculated as follows

With a small orthogonal error φ, the relationship (1) can be

be simplified.

In a preferred embodiment of the invention, it is provided that the angular error ϵ(α, φ) is subtracted from the measured angular position.

At least one of the objects specified above is also achieved by a detection system for detecting an angular position of a rotating component using the method specified above, having an evaluation unit and a sensor unit which has a fixed sensor element and a rotating element which can be rotated relative to this and together with the rotating component.

In an advantageous embodiment of the invention, it is provided that the detection system is assigned to a parking lock device.

Further advantages and advantageous configurations of the invention result from the description of the figures and the illustrations. character description

The invention is described in detail below with reference to the figures. They show in detail:

Figure 1: A cross section of a parking lock actuator with a detection system in a special embodiment of the invention.

FIG. 2: A flow chart of a method in a special embodiment of the invention.

FIG. 3: A profile of a measured and an ideal angle signal.

Figure 4: An angular gradient over the rotational position.

Figure 5: A course of the angular error over the rotational position.

Figure 6: A comparison of the compensation performance when calculating the angular position.

Figure 1 shows a cross section of a parking lock actuator 10 with a detection system in a specific embodiment of the invention. The parking lock actuator 10 is preferably an electronic parking lock actuator 10 and is arranged in a vehicle to actuate a parking lock device. The parking lock device can be switched between a locked position, in which a parking position of the vehicle is fixed, and a release position, in which the vehicle can be moved. Switching takes place via the parking lock actuator 10, which for this purpose has an electric motor 12 with a stator 14 and a rotor 16 that can be moved relative to this.

The rotor 16 is connected in a torque-proof manner to a hydraulic pump 20 via a drive shaft 18 for the transmission of a drive torque. Rotational movement of the rotor 16 is transmitted to the hydraulic pump 20, which thereby causes a fluid pressure change to control the parking lock device. As a rotary component 22 , the drive shaft 18 can be rotated about an axis of rotation 24 and an angular position of the rotary component 22 can be detected by a sensor unit 26 . The sensor unit 26 has a rotary element 28 which is firmly connected to the rotary component 22 . The rotary element 28 is firmly connected to the drive shaft 18 and is preferably designed as a diametrically magnetized permanent magnet. The sensor unit 26 further comprises a sensor element 30 which is connected to a housing 32 axially opposite the rotary element 28 . The rotary element 28 can be rotated in relation to the sensor element 30 . The sensor element 30 preferably comprises a Hall sensor which is arranged radially offset with respect to the axis of rotation and which detects an angular position of the rotary component 22 via the magnetic field provided by the rotary element.

The sensor element 30 preferably has two sensor components lying in a plane whose normal is parallel to the axis of rotation 24, perpendicular to one another and each radially offset from the axis of rotation, each of which emits a sensor signal. In the ideal case, the sensor signals are periodic signals that are offset by 90° from one another when the rotary component 22 rotates.

FIG. 2 shows a flow chart of a method 100 in a special embodiment of the invention. The following description refers to FIG. 2, but expressly refers to other figures in places where they are indicated.

The sensor unit 26 outputs a first sensor signal S ₁ and a second sensor signal S ₂ which is ideally offset by 90° thereto, ie ideally orthogonal. For example, the first sensor signal S ₁ is a cosine signal and the second sensor signal S ₂ is a sine signal. The first and second sensor signal S _1, S ₂ is received by an evaluation unit 102 and processed therein to form an angular position α _f of the rotary component. The angular position α _f is preferably calculated from the first and second sensor signals S _{1 ,} S ₂ using an atan2 function. The atan2 function is an extension of the inverse trigonometric function arctangent and, like it, an inverse function of the trigonometric function tangent. It takes two real numbers as arguments, unlike the normal arctangent, which takes only one real number as an argument. It therefore has enough information to be able to output the function value in a value range of 360° (i.e. all four quadrants) and does not have to be limited to two quadrants (like the normal arctangent). The angular position α _f can have an angular error. For example, the first and second sensor signals S _{1 ,} S ₂ can be offset from one another by a phase value that is not equal to 90°. This deviation from an orthogonal position of the first and second sensor signals S _{1 ,} S ₂ causes an orthogonal error as an error component in the angular error of the angular position α _f .

The method 100, described in more detail below, is preferably designed to reduce or compensate for this orthogonal error. First, the rotating component is allowed to rotate through at least 90°, with the rotating speed remaining as constant as possible. The rotational speed can be adjusted, for example, by decoupling the rotational component from a drive torque and allowing it to rotate freely. Due to the moment of inertia of the rotary component and the components connected to it in a rotationally effective manner, the rotational speed can temporarily be maintained approximately constant.

During the rotary movement, the angular position α _f detected by the sensor unit 26 and output by the evaluation unit 102 is recorded as an angle signal. The angle signal can be corrected for a possible amplitude error and/or a possible offset error, for example by using a max-min method, before the atan2 function is applied.

FIG. 3 shows a progression of the angular position a over the rotary position D of the rotary component in comparison between the measured angle signal α _f and an ideal angle signal α _t that is free from an orthogonal error. It can be seen that the orthogonal error triggers a periodic change in the ideal angle signal α _t . For example, the orthogonal error can result from a tilting of the rotary component with respect to the axis of rotation.

Returning to Figure 2, the rotary member is allowed to rotate at least 90°, preferably a full revolution. Then, in a calculation step 104, the gradient of the angle signal is recorded as the respective angle gradient at two measurement positions. A first angular gradient G ₁ is preferably found at 135° and a second angular gradient G ₂ is preferably found at 45°. In FIG. 4, the angular gradient G is shown as a function of the rotational position D of the rotating component. The measured angular gradient G _m over a complete revolution is preferably averaged using a moving average method in order to reduce the influence of noise. This averaged angular gradient G _p is preferably used further. It can be established that there is a minimum angular gradient at 45° and a maximum angular gradient at 135°.

Coming back to Figure 2, the first and second angular gradients G ₁ , G ₂ , i.e. the angular gradient at 45° and the angular gradient at 135°, are then transferred to a processing step 106 in which, depending on a quotient of the first and second angular gradient G ₁ , G ₂ an error parameter γ is calculated.

The error parameter y is transferred to a subsequent processing step 108, which uses it to calculate an orthogonal error φ of the angle signal. In a subsequent conversion step 110, the orthogonal error φ is converted into a calculated angular error ϵ _b , which is transferred to a subsequent output step in which the measured angular position α _f is corrected for the calculated angular error ϵ _b and output as the calculated angular position a.

The calculation step 104, the processing step 106 and the conversion step 110 can be carried out in a training process or a learning process and can output the orthogonal error φ in a retrievable manner. For example, the teaching process and/or the learning process can be carried out before an application-side operation of the rotary component and/or during an application-side operation of the rotary component, for example in the vehicle. The conversion step 110 may retrieve the learned orthogonal error φ during field operation of the rotating component and process it as described.

FIG. 5 shows a profile of the angle error ε over the rotational position D. The angle error ε _m of the measured angle signal runs periodically because of the orthogonal error involved. The calculated angle error ε _b can be detected and compensated for using the method described above. This reduces the angle error in the angle signal to the remaining angle error ϵ _r .

FIG. 6 shows a comparison of the compensation power A when calculating the angular position. In FIG. 6a), the angular error e is shown as a function of the orthogonal error φ. The angle error ϵ _m of the measured angle signal is greater than the remaining angle error ϵ _r when using the proposed method. The smaller the orthogonal error φ, the greater the relative difference.

FIG. 6 b) shows the compensation power A of an error compensation for calculating the angular position. With a smaller orthogonal error φ, the error compensation is maximally possible. Since the orthogonal errors φ to be expected in practice are usually below 20°, the compensation power A itself is over 90%.

Reference List

10 parking lock actuator

12 electric motor

14 stator

16 rotors

18 drive shaft

20 hydraulic pump

22 rotary component

24 axis of rotation

26 sensor unit

28 rotary element

30 sensor element

32 housing

100 procedures

102 evaluation unit

104 calculation step

106 preparation step

108 processing step

110 conversion step

S1 first sensor signal

S ₂ second sensor signal α calculated angular position αf angular position α _t angular position

D rotation position

G Angular gradient

G m measured angular gradient

G _p mean angle gradient

G ₁ first angle gradient

G ₂ second angle gradient γ error parameter φ orthogonal error ϵ angle error ϵb calculated angle error

Claims

patent claims

1. Method (100) for detecting an angular position (α) of a rotary component (22) which can be rotated about an axis of rotation (24) via a sensor unit (26) which has a fixed sensor element (30) and, opposite this and together with the rotary component (22 ) has a rotatable rotating element (28), wherein the sensor element (30) outputs a first and second sensor signal (S ₁ , S ₂ ) dependent on the angular position to an evaluation unit (102), which uses this to generate an angular signal indicating the angular position and an angular error ( ϵ) is calculated, characterized in that an orthogonal error (φ) describing the orthogonal deviation between the first and second sensor signal (S ₁ , S ₂ ) and assigned to the angular error (ϵ) is calculated in that at a first rotational position of the rotary component (22 ) a first angular gradient (G ₁ ) of the angle signal and, if the second rotational position of the rotary component (22) differs from the first rotational position, a second angular gradient (G ₂ ) of the angle signal is detected, the orthogonal error (φ) is then calculated as a function of the first and second angular gradient (G ₁ , G ₂ ) and the angular position (α _f ) is also corrected for this orthogonal error (φ).

2. Method (100) for detecting an angular position (α) according to claim 1, characterized in that the first angular gradient (G ₁ ) and second angular gradient (G ₂ ) is detected by rotating the rotary component (22) at a constant rotational speed.

3. The method (100) for detecting an angular position (α) according to claim 1 or 2, characterized in that the first angular gradient (G ₁ ) and second Angle gradient (G ₂ ) is detected by rotating the rotary member (22) through at least 90°.

4. Method (100) for detecting an angular position (α) according to claim 3, characterized in that the first rotational position is 45° and the second rotational position is 135°.

5. Method (100) for detecting an angular position (α) according to one of the preceding claims, characterized in that an error parameter (y) is calculated depending on a quotient of the first and second angular gradient (G ₁ , G ₂ ).

6. Method (100) for detecting an angular position (α) according to claim 5, characterized in that the orthogonal error (φ) is calculated as a function of the error parameter (y).

7. Method (100) for detecting an angular position (α) according to claim 6, characterized in that the angular error (ϵ _b ) is calculated as a function of the orthogonal error (φ).

8. Method (100) for detecting an angular position (α) according to claim 7, characterized in that the angular error (ϵ _b ) is subtracted from the measured angular position (α _f ).

9. Detection system for detecting an angular position (α) of a rotary component (22) by a method (100) according to any one of the preceding claims and having an evaluation unit (102) and a sensor unit (26) having a fixed sensor element (30) and an opposite this and together with the rotary member (22) rotatable rotary element (28).

10. Detection system for detecting an angular position (α) according to claim 9, characterized in that the detection system is associated with a parking lock device.