DE102020205943B3 - Method for processing the sensor signals of a rotation rate sensor arrangement, device and method for operating the device - Google Patents

Abstract

The invention relates to a method for processing the sensor signals of a rotation rate sensor arrangement with at least three rotation rate sensor components which supply sensor signals gx, gy, gz for three linearly independent spatial directions in a sensor reference system, in which an output signal with two signal components gy ', gz' for two linearly independent spatial directions of one Absolute reference system is generated by at least the sensor signals gy, gz for two of the three linearly independent spatial directions of the sensor reference system are transformed into the absolute reference system, whereby a rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is at least partially compensated. The compensation of the rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is based on the transformation of the at least two sensor signals gy, at least the third sensor signal gx for the third linearly independent spatial direction of the sensor reference system.

Description

The present invention relates to a method for processing the sensor signals of a rotation rate sensor arrangement. The invention also relates to a corresponding device and a method for operating the device. In particular, the invention relates to an electronic pointing device, an output signal generated by the device being a control signal for an electronic device.

State of the art

The change in orientation of the rotation rate sensor can be determined on the basis of the output signals of a rotation rate sensor. As a result, the rotation rate sensor can be used, for example, for angle measurement by attaching the rotation rate sensor to the object to be examined and determining the angular position or angle change of the object on the basis of the output signal.

Another application is to use the output signals of the rotation rate sensor to control an electronic device, such as a drone or a display. The rotation rate sensor can be built into a digital pointing device. With certain rotational movements of the pointing device, a user can move an object on the display in a horizontal or vertical direction.

The yaw rate sensor measures the pitch angle, roll angle and yaw angle in the sensor reference system. However, the user orients himself to the fixed absolute reference system. However, the sensor reference system does not match the absolute reference system either initially or at least because of the rotary movement. This can lead to incorrect angle specifications or unwanted control inputs.

From the US 7 158 118 B2 describes a correction of the deviation between the described coordinate systems using the direction of the acceleration due to gravity. However, this requires continuous measurement of the acceleration due to gravity using the output signals of an acceleration sensor, which is energy-intensive.

Disclosure of the invention

The invention provides a method for processing the sensor signals of a rotation rate sensor arrangement, a device and a method for operating the device with the features of the independent claims.

Preferred embodiments are the subject of the respective subclaims.

According to a first aspect, the invention thus relates to a method for processing the sensor signals of a rotation rate sensor arrangement with at least three rotation rate _{sensor components that supply sensor signals g x} , g _y , g _z for three linearly independent spatial directions in a sensor reference system. Here, an output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system is generated. At least the sensor signals g _y , g _z for two of the three linearly independent spatial directions of the sensor reference system are transformed into the absolute reference system, whereby a rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is at least partially compensated for. The compensation of the rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is based on at least the third sensor signal g _x for the third linearly independent spatial _{direction of the sensor reference system when transforming the at least two sensor signals g y} , g _z.

According to a second aspect, the invention accordingly relates to a device with a rotation rate sensor arrangement with at least three rotation rate _{sensor components which supply sensor signals g x} , g _y , g _z for three linearly independent spatial directions in a sensor reference system. The device further comprises a signal processing device for the sensor signals g _x , g _y , g _z , which generates an output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system using the method according to the invention.

According to a third aspect, the invention relates to a method for operating a device according to the invention, in which the rotation ϕ of the rotation rate sensor arrangement about at least one of the three linearly independent spatial directions of the sensor reference system at regular time intervals and / or event initiated and / or user-initiated is determined and the subsequent transformation of the sensor signals g _y , g _{z is} based.

Advantages of the invention

According to the method, the rotational movement of the yaw rate sensor arrangement can be calculated in the absolute reference system. For this purpose, the activation of an acceleration sensor is not necessary or only necessary for correction purposes. An acceleration sensor that may be present can be deactivated at least temporarily so that electricity can be saved. Depending on the application, an acceleration sensor can also be completely dispensed with.

wobei ϕ₀ eine zu bestimmende initiale Verdrehung bezeichnet, und wobei die Signalkomponenten g_y', g_z' des Ausgangssignals bestimmt werden als $$\left[\begin{array}{c}{g}_y{}^{\text{'}}\\ g_z{}^{\text{'}}\end{array}\right]=\left[\begin{array}{c}cos\varphi \cdot sin\varphi \\ -sin\varphi \cdot cos\varphi \end{array}\right]\cdot \left[\begin{array}{c}{g}_y\\ g_z\end{array}\right]$$

According to a development of the method, the rotation ϕ of the rotation rate sensor arrangement is determined on the basis of the third sensor signal g _x as $$\varphi ={\mathrm{<figure-callout\; id="0"\; label="\phi "\; filenames="DE102020205943B3\_0029.png"\; state="\{\{state\}\}">\varphi </figure-callout>}}_0{\displaystyle \sum G_x}\cdot \Delta t$$

where ϕ ₀ denotes an initial rotation to be determined, and where the signal components g _y ', g _z ' of the output signal are determined as $$\left[\begin{array}{c}{G}_y{}^{\text{'}}\\ G_z{}^{\text{'}}\end{array}\right]=\left[\begin{array}{c}cOs\varphi \cdot sin\varphi \\ -sin\varphi \cdot cOs\varphi \end{array}\right]\cdot \left[\begin{array}{c}{G}_y\\ G_z\end{array}\right]$$

Here, Δt denotes the duration of a predetermined time segment.

According to a development of the method, the signal components g _y ', g _z ' of the output signal are calculated with the aid of sine and cosine functions and / or approximated using look-up tables.

According to a development of the method, the initial rotation ϕ _{0 is determined} on the basis of sensor signals of at least one acceleration sensor and / or at least one magnetic field sensor. For example, the acceleration sensor signal can indicate the current acceleration of the rotation rate sensor arrangement. A constant component, which corresponds to the acceleration due to gravity, can be determined over several measurements. The rotation (about a roll angle in the absolute reference system) can be compensated for by rotating the sensor reference system. Furthermore, the rotation can also be determined taking into account a weighting of the acceleration sensor signals and / or the magnetic field sensor signals.

According to a development of the method, the transformation of the at least two sensor signals g _y , g _{z is based on} all three sensor signals g _x , g _y , g _z , in that the transformation takes place on the basis of quaternions.

wobei mittels der so berechneten halben Radianten die Aktualisierung der Quaternionen berechnet wird als $$\begin{array}{c}\overline{q_w}=q_{w_{t-1}}+\alpha \cdot q_{x_{t-1}}+\beta \cdot q_{y_{t-1}}+\gamma \cdot q_{z_{t-1}}\\ \overline{q_x}=q_{x_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{y_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_y}=q_{y_{t-1}}-\alpha \cdot q_{w_{t-1}}-\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_z}=q_{z_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{y_{t-1}}\end{array}$$

According to a further development of the method, the three sensor signals g _x , g _y , g _{z are converted} from degrees per second into frequency-dependent half radians, according to FIG $$\left(\begin{array}{c}\alpha \\ \beta \\ \gamma \end{array}\right)=\left(\begin{array}{c}{G}_x\\ G_y\\ G_z\end{array}\right)\cdot \frac{\pi }{360}\cdot \Delta t$$

using the half radians thus calculated, the update of the quaternions is calculated as $$\begin{array}{c}\overline{q_w}=q_{w_{t-1}}+\alpha \cdot q_{x_{t-1}}+\beta \cdot q_{y_{t-1}}+\gamma \cdot q_{z_{t-1}}\\ \overline{q_x}=q_{x_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{y_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_y}=q_{y_{t-1}}-\alpha \cdot q_{w_{t-1}}-\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_z}=q_{z_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{y_{t-1}}\end{array}$$

A normalization factor is also calculated as: $$factOr=\frac{1}{\sqrt{{\overline{q_w}}^2+{\overline{q_x}}^2+{\overline{q_y}}^2+{\overline{q_z}}^2}}$$

The quaternions are normalized according to: $$\begin{array}{c}{q}_{w_t}=\overline{q_w}\cdot factOr\\ q_{x_t}=\overline{q_z}\cdot factOr\\ q_{y_t}=\overline{q_y}\cdot factOr\\ q_{z_t}=\overline{q_z}\cdot factOr\end{array}$$

The signal components g _y ', g _z ' of the output signal are calculated as: $$\begin{array}{l}{G}_y{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_x\cdot q_{z_t}+q_{x_t}\cdot G_x\cdot q_{y_t}-q_{w_t}\cdot G_z\cdot q_{x_t}+q_{y_t}\cdot G_z\cdot q_{z_t}\right)+G_y\\ \cdot \left(q_{w_t}^2-q_{x_{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="DE102020205943B3\_0025.png,DE102020205943B3\_0026.png"\; state="\{\{state\}\}">t</figure-callout>}}}^2+q_{y_t}^2-q_{z_t}^2\right)\\ G_z{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_y\cdot q_{x_t}+q_{w_t}\cdot G_x\cdot q_{y_t}+q_{x_t}\cdot G_x\cdot q_{z_t}+q_{y_t}\cdot G_y\cdot q_{z_t}\right)+G_z\\ \cdot \left(q_{w_t}^2-q_{x_t}^2-{\mathrm{<figure-callout\; id="2"\; label="q\; y\; t"\; filenames="DE102020205943B3\_0025.png,DE102020205943B3\_0026.png"\; state="\{\{state\}\}">q</figure-callout>}}_{y_t}^{{}_{}}+q_{z_t}^2\right)\end{array}$$

According to a development of the device, the output signal is used as a control signal for a man-machine interface. For example, a mouse pointer on a display can be controlled using the output signal. The device can be, for example, a pointing device, or a component of such a pointing device.

According to a further development of the method, the compensation for the rotation includes a compensation for a rotation generated by a rolling movement of the rotation rate sensor arrangement in the absolute reference system. In particular, rotations around a single axis in the absolute reference system can be compensated, for example only a rolling movement. Such a twist should be compensated for when using a pointing pen, for example. Only the rotary movements around a pitch axis and a yaw axis in the absolute reference system should contribute to the generation of an output signal. The invention is therefore suitable for using the rotation rate sensor arrangement for controlling a two-dimensional measurement or movement, for example a mouse pointer.

According to a further development of the method, rotations about two orthogonal axes can be compensated for in the absolute reference system. Only the rotary movement around the remaining third orthogonal axis should contribute to the generation of an output signal. The invention is therefore suitable for using the rotation rate sensor arrangement for controlling a one-dimensional measurement or movement.

According to a development of the method, the determination of initial values for angles of rotation of the rotation rate sensor arrangement is repeated at predetermined time intervals. Instantaneous angles of rotation of the rotation rate sensor arrangement are calculated in the absolute reference system on the basis of the repeatedly determined initial value for the angle of rotation and on the basis of the sensor signals generated. Such a recalibration can prevent errors in the compensation of the rotation from becoming too large, which errors may arise over time due to measurement and calculation inaccuracies.

According to a development of the method, the predefined time intervals are selected periodically or are selected as a function of a time dependency of the calculated instantaneous angle of rotation. In particular, the time intervals can be selected to be smaller, the more the calculated instantaneous angle of rotation changes, since in this case the probability of increasing errors in the compensation of the rotation is greater.

According to a further development, the device comprises at least one optionally activatable acceleration sensor and / or at least one optionally activatable magnetic field sensor, the signal processing device being designed to rotate eine the rotation rate sensor arrangement about at least one of the three linearly independent spatial directions on the basis of the acceleration sensor signals and / or the magnetic field sensor signals to determine the sensor reference system.

Figure list

• 1 ein schematisches Blockdiagramm einer Vorrichtung gemäß einer Ausführungsform der Erfindung;
• 2 schematische Darstellungen zur Erläuterung einer Verwendung der Drehratensensoranordnung zur Steuerung eines Mauszeigers auf einer Anzeige;
• 3 eine schematische Darstellung einer Drehratensensoranordnung zur Erläuterung eines Rollwinkels, Gierwinkels und Nickwinkels;
• 4 schematische Darstellungen zur Erläuterung einer Kompensation einer Verdrehung der Drehratensensoranordnung;
• 5 eine schematische Illustration zur Erläuterung einer Aktivität von einer Drehratensensoranordnung und einem Beschleunigungssensor gemäß einer Ausführungsform der Erfindung;
• 6 eine schematische Illustration zur Erläuterung einer Aktivität von einer Drehratensensoranordnung und einem Beschleunigungssensor gemäß einer weiteren Ausführungsform der Erfindung;
• 7 eine schematische Illustration zur Erläuterung eines Orientierungsfehlers nach kombinierten Drehungen der Drehratensensoranordnung;
• 8 eine schematische Illustration zur Erläuterung einer Kompensation einer Verdrehung unter Berücksichtigung einer Drehung um eine einzige Koordinatenachse;
• 9 eine schematische Illustration zur Erläuterung einer Kompensation einer Verdrehung unter Berücksichtigung einer Drehung um alle drei Koordinatenachsen; und
• 10 ein schematisches Flussdiagramm eines Verfahrens zum Verarbeiten von Sensorsignalen gemäß einer Ausführungsform der Erfindung.

Show it:

• 1 a schematic block diagram of an apparatus according to an embodiment of the invention;
• 2 schematic representations to explain a use of the rotation rate sensor arrangement for controlling a mouse pointer on a display;
• 3 a schematic representation of a rotation rate sensor arrangement for explaining a roll angle, yaw angle and pitch angle;
• 4th schematic representations to explain a compensation for a rotation of the rotation rate sensor arrangement;
• 5 a schematic illustration to explain an activity of a rotation rate sensor arrangement and an acceleration sensor according to an embodiment of the invention;
• 6th a schematic illustration to explain an activity of a rotation rate sensor arrangement and an acceleration sensor according to a further embodiment of the invention;
• 7th a schematic illustration to explain an orientation error after combined rotations of the rotation rate sensor arrangement;
• 8th a schematic illustration to explain a compensation of a twist taking into account a rotation about a single coordinate axis;
• 9 a schematic illustration to explain a compensation of a twist taking into account a rotation about all three coordinate axes; and
• 10 a schematic flow diagram of a method for processing sensor signals according to an embodiment of the invention.

Identical or functionally identical elements and devices are provided with the same reference symbols in all figures.

Description of the exemplary embodiments

1 shows a schematic block diagram of an apparatus 1 with a rotation rate sensor arrangement 2 and a signal processing device 3 . The rotation rate sensor arrangement 2 includes three rotation rate sensor components 21 , 22nd , 23 , which provide sensor signals g _x , g _y , g _z for three linearly independent spatial directions in a sensor reference system. For example, the spatial directions can be orthogonal directions x , y , z correspond. The sensor signals can include rotation rates, for example about a roll axis, yaw axis and / or pitch axis in the sensor reference system.

The signal processing device 3 for the sensor signals g _x , g _y , g _z generates an output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system. The signal processing device 3 can include one or more microprocessors, application processors, sensor hubs, FPGAs, ASICs or the like, as well as a memory device.

The signal processing device 3 generates the output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system by the signal processing device 3 at least the sensor signals g _y , g _z for two of the three linearly independent spatial directions of the sensor reference system are transformed into the absolute reference system. As a result, a rotation ϕ of the yaw rate sensor arrangement about the third linearly independent spatial direction of the sensor reference system is compensated. The signal processing device 3 uses at least the third sensor signal g _x for the third linearly independent spatial direction of the sensor reference system as a basis for the compensation of the rotation Dre of the rotation rate sensor arrangement about the third linearly independent spatial direction of the sensor reference system when transforming the at least two sensor signals g _y , g _z.

wobei ϕ₀ eine zu bestimmende initiale Verdrehung bezeichnet und Δt ein vorgegebenes Zeitintervall bezeichnet. Die Signalkomponenten g_y', g_z' des Ausgangssignals werden bestimmt als: $$\left[\begin{array}{c}{g}_y{}^{\text{'}}\\ g_z{}^{\text{'}}\end{array}\right]=\left[\begin{array}{c}cos\varphi \cdot sin\varphi \\ -sin\varphi \cdot cos\varphi \end{array}\right]\cdot \left[\begin{array}{c}{g}_y\\ g_z\end{array}\right]$$

The signal processing device 3 can determine the rotation ϕ of the yaw rate sensor arrangement on the basis of the third sensor signal g _x as: $$\varphi ={\mathrm{<figure-callout\; id="0"\; label="\phi "\; filenames="DE102020205943B3\_0029.png"\; state="\{\{state\}\}">\varphi </figure-callout>}}_0{\displaystyle \sum G_x}\cdot \Delta t$$

where ϕ ₀ denotes an initial rotation to be determined and Δt denotes a predetermined time interval. The signal components g _y ', g _z ' of the output signal are determined as: $$\left[\begin{array}{c}{G}_y{}^{\text{'}}\\ G_z{}^{\text{'}}\end{array}\right]=\left[\begin{array}{c}cOs\varphi \cdot sin\varphi \\ -sin\varphi \cdot cOs\varphi \end{array}\right]\cdot \left[\begin{array}{c}{G}_y\\ G_z\end{array}\right]$$

The signal processing device 3 the signal components g _y ', g _z ' of the output signal can also approximate using look-up tables.

The signal processing device 3 the initial rotation ϕ ₀ on the basis of sensor signals from an acceleration sensor 4th and / or a magnetic field sensor 5 determine.

The signal processing device 3 The transformation of the at least two sensor signals g _y , g _{z can be based on} all three sensor signals g _x , g _y , g _z , in that the transformation takes place on the basis of quaternions.

wobei mittels der so berechneten halben Radianten die Aktualisierung der Quaternionen berechnet wird als $$\begin{array}{c}\overline{q_w}=q_{w_{t-1}}+\alpha \cdot q_{x_{t-1}}+\beta \cdot q_{y_{t-1}}+\gamma \cdot q_{z_{t-1}}\\ \overline{q_x}=q_{x_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{y_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_y}=q_{y_{t-1}}-\alpha \cdot q_{w_{t-1}}-\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_z}=q_{z_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{y_{t-1}}\end{array}$$

The signal processing device 3 For this purpose, the three sensor signals g _x , g _y , g _{z can be converted} from degrees per second into frequency-dependent half radians, according to $$\left(\begin{array}{c}\alpha \\ \beta \\ \gamma \end{array}\right)=\left(\begin{array}{c}{G}_x\\ G_y\\ G_z\end{array}\right)\cdot \frac{\pi }{360}\cdot \Delta t$$

using the half radians thus calculated, the update of the quaternions is calculated as $$\begin{array}{c}\overline{q_w}=q_{w_{t-1}}+\alpha \cdot q_{x_{t-1}}+\beta \cdot q_{y_{t-1}}+\gamma \cdot q_{z_{t-1}}\\ \overline{q_x}=q_{x_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{y_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_y}=q_{y_{t-1}}-\alpha \cdot q_{w_{t-1}}-\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_z}=q_{z_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{y_{t-1}}\end{array}$$

The signal processing device can also 3 calculate a normalization factor as: $$factOr=\frac{1}{\sqrt{{\overline{q_w}}^2+{\overline{q_x}}^2+{\overline{q_y}}^2+{\overline{q_z}}^2}}$$

$$\begin{array}{l}{q}_{w_t}=\overline{q_w}\cdot factor\\ q_{x_t}=\overline{q_z}\cdot factor\\ q_{y_t}=\overline{q_y}\cdot factor\\ q_{z_t}=\overline{q_z}\cdot factor\end{array}$$

The quaternions are normalized according to: $$$$

$$\begin{array}{l}{q}_{w_t}=\overline{q_w}\cdot factOr\\ q_{x_t}=\overline{q_z}\cdot factOr\\ q_{y_t}=\overline{q_y}\cdot factOr\\ q_{z_t}=\overline{q_z}\cdot factOr\end{array}$$

The signal processing device 3 calculates the signal components g _y ', g _z ' of the output signal as: $$\begin{array}{l}{G}_y{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_x\cdot q_{z_t}+q_{x_t}\cdot G_x\cdot q_{y_t}-q_{w_t}\cdot G_z\cdot q_{x_t}+q_{y_t}\cdot G_z\cdot q_{z_t}\right)+G_y\\ \cdot \left(q_{w_t}^2-{\mathrm{<figure-callout\; id="2"\; label="q\; x\; t"\; filenames="DE102020205943B3\_0025.png,DE102020205943B3\_0026.png"\; state="\{\{state\}\}">q</figure-callout>}}_{x_t}^{{}_{}}+q_{y_t}^2-q_{z_t}^2\right)\\ G_z{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_y\cdot q_{x_t}+q_{w_t}\cdot G_x\cdot q_{y_t}+q_{x_t}\cdot G_x\cdot q_{z_t}+q_{y_t}\cdot G_y\cdot q_{z_t}\right)+G_z\\ \cdot \left(q_{w_t}^2-q_{x_t}^2-{\mathrm{<figure-callout\; id="2"\; label="q\; y\; t"\; filenames="DE102020205943B3\_0025.png,DE102020205943B3\_0026.png"\; state="\{\{state\}\}">q</figure-callout>}}_{y_t}^{{}_{}}+q_{z_t}^2\right)\end{array}$$

The output signal can be used as a control signal for a man-machine interface 6th be used. In particular, the human-machine interface 6th an external electronic device 6th be. For example, a mouse pointer can be controlled on a display.

According to further embodiments, the electronic device 6th also a component of the device 1 be.

Optionally, the device includes 1 further an acceleration sensor 4th and / or magnetic field sensor 5 which output acceleration sensor signals or magnetic field sensor signals.

The signal processing device 3 can, on the basis of the acceleration sensor signals and / or the magnetic field sensor signals, a rotation ϕ of the rotation rate sensor arrangement 2 to determine at least one of the three linearly independent spatial directions of the sensor reference system.

The components of the device 1 are preferably manufactured using MEMS technology.

2 shows schematic representations to explain a use of the rotation rate sensor arrangement 2 to control a mouse pointer 8th on an advertisement 7th . When the rotation rate sensor arrangement is rotated 2 around a yaw axis z the rotation rate sensor arrangement 2 becomes the mouse pointer 8th on the display 7th shifted in the horizontal direction ( 2 Left). When the rotation rate sensor arrangement is rotated 2 around a pitch axis y the yaw rate sensor arrangement becomes the mouse pointer 8th on the display 7th shifted in the vertical direction ( 2 right).

3 shows a schematic representation of the rotation rate sensor arrangement 2 to explain a roll angle, yaw angle and pitch angle in the sensor reference system. For this purpose there are three mutually orthogonal coordinate axes x , y and z drawn in the sensor reference system, with the first coordinate axis x corresponds to a roll axis, the second coordinate axis y corresponds to a pitch axis and the third coordinate axis z corresponds to a yaw axis. In the case of a pen-shaped pointing device, the first coordinate axis x for example, correspond to the axis of symmetry of the pointing device.

The angular position of the yaw rate sensor arrangement 2 with reference to the first coordinate axis x is described by the roll angle in the sensor reference system. The angular position of the yaw rate sensor arrangement is also shown 2 with reference to the second coordinate axis y described by the pitch angle in the sensor reference system. Finally, the angular position of the yaw rate sensor arrangement 2 with reference to the third coordinate axis z described by the yaw angle in the sensor reference system.

The sensor reference system can be transformed into an absolute reference system. In the absolute reference system, in turn, three orthogonal axes x ', y', z 'can be defined, as exemplified in 2 drawn. Analogously, these can be referred to as roll axis x ', pitch axis y' and yaw axis z 'in the absolute reference system, with roll angle, yaw angle and pitch angle being measured in the absolute reference system relative to these axes. For the in 2 Starting position of the yaw rate sensor arrangement shown on the left 2 do the roll axes, pitch axes and yaw axes in the sensor reference system and in the absolute reference system match.

In the in 2 A rolling movement in the absolute reference system should not affect the control of the mouse pointer 8th on the display 7th to have. This movement should therefore be compensated.

4th shows schematic representations to explain a compensation for a rotation of the rotation rate sensor arrangement 2 . As shown on the left, the rotation rate sensor arrangement 2 be rotated by a roll angle around the roll axis x 'in the absolute reference system. The rotation rate sensor arrangement 2 is thus in an inclined position so that the sensor reference system and the absolute reference system are rotated relative to one another. When the rotation rate sensor arrangement is rotated 2 The rotation rate sensor arrangement rotates around the yaw axis z 'in the absolute reference system 2 in the sensor reference system both around the yaw axis z as well as about the pitch axis y . If the rotation were not compensated for to calculate the output signal, this would result in a movement of the mouse pointer that deviates from the horizontal direction, as shown on the left in 4th shown. However, since the user orients himself to the absolute reference system, this is undesirable. By compensating for the rotation, the mouse pointer can only move horizontally, as on the right in 4th shown. The compensation will be described in more detail below.

ermittelt werden kann. Mit Hilfe folgender Formel kann die anfängliche Verdrehung ϕ₀ um die Rollachse im absoluten Bezugssystem berechnet werden: $${\varphi }_0=ta{n}^{-1}\left(\frac{{\alpha }_y}{{\alpha }_z}\right).$$

At the beginning, the starting position of the yaw rate sensor arrangement 2 be determined. This can either be known because the rotation rate sensor arrangement 2 is in a known position at the beginning, or can be measured. The measurement can be carried out using a measurement signal from the acceleration sensor 4th and / or magnetic field sensor 5 respectively. This is intended below for an acceleration signal from the acceleration sensor 4th will be explained in more detail. The accelerometer 4th determines an acceleration from which, with the help of a low-pass filter, the gravitational acceleration vector: $$\left(\begin{array}{c}{\alpha }_x\\ {\alpha }_y\\ {\alpha }_z\end{array}\right)$$

can be determined. The following formula can be used to calculate the initial rotation ϕ ₀ around the roll axis in the absolute reference system: $${\varphi }_0=ta{n}^{-1}\left(\frac{{\alpha }_y}{{\alpha }_z}\right).$$

Furthermore, from the yaw rate sensor arrangement 2 measured yaw data g _y (ie the rate of rotation relative to the yaw axis in the sensor reference system) and pitch data g _z (ie the rate of rotation relative to the pitch axis in the sensor reference system) are used to calculate the rotation ϕ as a function of time.

Since the sensor signals of the yaw rate sensor arrangement 2 for example for controlling a mouse pointer, the roll data g _x (ie the rate of rotation relative to the roll axis in the sensor reference system) can therefore be used to compensate for the rotation of the rotation rate sensor arrangement without increasing the energy consumption 2 , ie the inclination, can be used.

By integrating the measured rate of rotation g _{x of} the roll axis, the rotation ϕ of the rate of rotation sensor arrangement becomes 2 determined and can be corrected using formulas (1) and (2).

The correction of the twist can be calculated with the help of sine and cosine functions. Alternatively, the use of a look-up table and, for example, linear interpolation between the table entries allows these functions to be calculated approximately and thus the angle correction to be calculated even on computing units with particularly limited resources.

Compared to the compensation of the twist with the help of the gravitational acceleration vector, the correction by the measured rate of rotation is much more precise and almost without delay, while the low-pass filter of the gravitational acceleration calculation can result in a delayed compensation.

If the rotation rate sensor arrangement 2 is aligned with the absolute reference system at the start of the control process and the length of the control process is limited to a few minutes, ϕ ₀ can be assumed to be 0, so it does not have to be measured initially. As a result, a correction with the aid of the acceleration due to gravity and thus the energy consumption required for this can be dispensed with.

5 shows a schematic illustration to explain an activity of a rotation rate sensor arrangement and an acceleration sensor according to an embodiment of the invention. Al shows the activity of the accelerometer 4th on while A2 the activity of the yaw rate sensor arrangement 2 indicates. Initially, that is, up to a point in time t1 is just the acceleration sensor 4th activated, for example to recognize gestures, activities or steps. At the time t1 becomes the rotation rate sensor arrangement 2 activated, for example to provide a pointer function. Using the sensor signals from the acceleration sensor 4th becomes, as described above, the initial rotation ϕ ₀ at the point in time t1 determined. The accelerometer 4th can now be deactivated to save power. The momentary rotation ϕ is for this purpose based on the sensor signals of the rotation rate sensor arrangement 2 calculated. At a time t2 becomes the rotation rate sensor arrangement 2 deactivated again and the accelerometer 4th activated again.

If there is therefore no initial alignment with the world reference system, a prior use of the acceleration _{sensor can be used to determine ϕ 0 at the start of the control process} 4th determined gravitational acceleration vector can be used to compensate for the rotation. During the control process, the accelerometer 4th remain switched off to save power.

6th shows a schematic illustration to explain an activity of a rotation rate sensor arrangement 2 and an accelerometer 4th according to a further embodiment of the invention. The acceleration sensor is used for this 4th activated at specified times in order to carry out a new calibration. The inclination of the yaw rate sensor arrangement is thus at these points in time 2 based on the sensor signals of the acceleration sensor 4th determined. The accelerometer 4th can be activated periodically for this purpose, for example. Instead of the accelerometer 4th to deactivate, provision can be made for the measurement frequency of the acceleration sensor 4th to reduce.

Instead of or in addition to the accelerometer 4th can also use the magnetic field sensor 5 can be used for orientation correction, preferably for twist compensation around the yaw axis.

7th shows a schematic illustration to explain an orientation error after combined rotations of the rotation rate sensor arrangement 2 . A concatenation of rotations of the rotation rate sensor arrangement is illustrated 2 about the pitch axis y 'and the yaw axis z in the absolute frame of reference. Used to compensate for the inclination of the yaw rate sensor arrangement 2 If only the roll angle is used, a noticeable error in the rotation can occur.

This shows 8th a schematic illustration to explain the compensation of a twist taking into account a rotation about a single coordinate axis. U denotes the dependence of the angle of rotation around the yaw axis as a function of time t, V the dependence of the angle of rotation around the pitch axis as a function of time t and W. the dependence of the compensated angle of rotation around the roll axis as a function of time t. The compensation is used to ensure that the angle of rotation about the roll axis should essentially remain equal to zero. If only the rotation around a single coordinate axis is taken into account to compensate for the inclination, this is no longer the case for longer times.

9 shows a schematic illustration to explain a compensation for a twist taking into account a rotation about all three coordinate axes. As can be seen, the roll angle is now close to zero degrees for longer periods as well.

10 shows a schematic flow diagram of a method for processing the sensor signals of a rotation rate sensor arrangement 2 with at least three rotation rate sensor components 21 , 22nd , 23 .

In a first process step S1 deliver the yaw rate sensor components 21 , 22nd , 23 Sensor signals g _x , g _y , g _z for three linearly independent spatial directions in a sensor reference system.

In a second process step S2 an output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system is generated by at least the sensor signals g _y , g _z for two of the three linearly independent spatial directions of the sensor reference system are transformed into the absolute reference system, whereby a Rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is at least partially compensated. The compensation of the rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system is based on at least the third sensor signal g _x for the third linearly independent spatial _{direction of the sensor reference system when transforming the at least two sensor signals g y} , g _z.

Claims (10)

Method for processing the sensor signals of a rotation rate sensor arrangement (2) with at least three rotation rate sensor components (21, 22, 23) which _{supply sensor signals g x} , g _y , g _z for three linearly independent spatial directions in a sensor reference system in which an output signal with two signal components g _y ', g _z ' is generated for two linearly independent spatial directions of an absolute reference system in that at least the sensor signals g _y , g _z for two of the three linearly independent spatial directions of the sensor reference system are transformed into the absolute reference system, whereby a rotation ϕ of the rotation rate sensor arrangement by the third linearly independent spatial direction of the sensor reference system is at least partially compensated, characterized in that the compensation of the rotation ϕ of the rotation rate sensor arrangement around the third linearly independent spatial direction of the sensor reference system in the transformation of the at least two sensor signals g _y , g _z at least the third sensor Signal g _{x is used as the basis} for the third linearly independent spatial direction of the sensor reference system. Procedure according to Claim 1 , characterized in that the rotation ϕ of the rotation rate sensor arrangement on the basis of the third sensor signal g _{x is} determined as $$\varphi ={\varphi }_0{\displaystyle \sum G_x}\cdot \Delta t$$

where ϕ ₀ denotes an initial rotation to be determined, and that the signal components g _y ', g _z ' of the output signal are determined as $$\left[\begin{array}{c}{G}_y{}^{\text{'}}\\ G_z{}^{\text{'}}\end{array}\right]=\left[\begin{array}{c}cOs\varphi \cdot sin\varphi \\ -sin\varphi \cdot cOs\varphi \end{array}\right]\cdot \left[\begin{array}{c}{G}_y\\ G_z\end{array}\right]$$

Procedure according to Claim 2 , characterized in that the signal components g _y ', g _z ' of the output signal are calculated with the aid of sine and cosine functions and / or are approximated using look-up tables. Method according to one of the Claims 2 or 3 , characterized in that the initial rotation ϕ _{0 is determined} on the basis of sensor signals of at least one acceleration sensor (4) and / or at least one magnetic field sensor (5). Procedure according to Claim 1 , characterized in that the transformation of the at least two sensor signals g _y , g _{z is based on} all three sensor signals g _x , g _y , g _z , in that the transformation takes place on the basis of quaternions. Procedure according to Claim 5 , characterized in that the three sensor signals g _x , g _y , g _{z are converted} from degrees per second into frequency-dependent angles of rotation in half radians, according to $$\left(\begin{array}{c}\alpha \\ \beta \\ \gamma \end{array}\right)=\left(\begin{array}{c}{G}_x\\ G_y\\ G_z\end{array}\right)\cdot \frac{\pi }{360}\cdot \Delta t$$

that using the half radians calculated in this way, the update of the quaternions is calculated as $$\begin{array}{c}\overline{q_w}=q_{w_{t-1}}+\alpha \cdot q_{x_{t-1}}+\beta \cdot q_{y_{t-1}}+\gamma \cdot q_{z_{t-1}}\\ \overline{q_x}=q_{x_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{y_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_y}=q_{y_{t-1}}-\alpha \cdot q_{w_{t-1}}-\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{z_{t-1}}\\ \overline{q_z}=q_{z_{t-1}}-\alpha \cdot q_{w_{t-1}}+\beta \cdot q_{x_{t-1}}-\gamma \cdot q_{y_{t-1}}\end{array}$$

that a normalization factor is calculated as $$factOr=\frac{1}{\sqrt{{\overline{q_w}}^2+{\overline{q_x}}^2+{\overline{q_y}}^2+{\overline{q_z}}^2}}$$

that the quaternions are normalized according to $$\begin{array}{c}{q}_{w_t}=\overline{q_w}\cdot factOr\\ q_{x_t}=\overline{q_z}\cdot factOr\\ q_{y_t}=\overline{q_y}\cdot factOr\\ q_{z_t}=\overline{q_z}\cdot factOr\end{array}$$

and that the signal components g _y ', g _z ' of the output signal are then calculated as $$\begin{array}{l}{G}_y{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_x\cdot q_{z_t}+q_{x_t}\cdot G_x\cdot q_{y_t}-q_{w_t}\cdot G_z\cdot q_{x_t}+q_{y_t}\cdot G_z\cdot q_{z_t}\right)+G_y\\ \cdot \left(q_{w_t}^2-q_{x_t}^2+q_{y_t}^2-q_{z_t}^2\right)\\ G_z{}^{\text{'}}=2\cdot \left(q_{w_t}\cdot G_y\cdot q_{x_t}+q_{w_t}\cdot G_x\cdot q_{y_t}+q_{x_t}\cdot G_x\cdot q_{z_t}+q_{y_t}\cdot G_y\cdot q_{z_t}\right)+G_z\\ \cdot \left(q_{w_t}^2-q_{x_t}^2-q_{y_t}^2+q_{z_t}^2\right)\end{array}$$

Device (1), at least comprising: a. A rotation rate sensor arrangement (2) with at least three rotation rate sensor components (21, 22, 23) which _{supply sensor signals g x} , g _y , g _z for three linearly independent spatial directions in a sensor reference system, and b. A signal processing device (3) for the sensor signals g _x , g _y , g _z , which generates an output signal with two signal components g _y ', g _z ' for two linearly independent spatial directions of an absolute reference system using a method according to one of the Claims 1 to 6th generated. Device according to Claim 7 , characterized in that the output signal is used as a control signal for a man-machine interface (6). Device according to one of the Claims 7 or 8th , with at least one optionally activatable acceleration sensor (4) and / or at least one optionally activatable magnetic field sensor (5), wherein the signal processing device is designed on the basis of the acceleration sensor signals and / or the magnetic field sensor signals a rotation ϕ of the rotation rate sensor arrangement by at least one of the three linear to determine independent spatial directions of the sensor reference system. Method for operating a device according to Claim 9 , in which the rotation ϕ of the rotation rate sensor arrangement by at least one of the three linearly independent spatial directions of the sensor reference system is determined at regular time intervals and / or event-initiated and / or user-initiated and the subsequent transformation of the sensor signals g _y , g _z according to one of the Claims 1 to 6th is taken as the basis.

Images