JPH10253360A - Angle-of-rotation detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an angle-of-rotation detector which enables highly accurate detection of the angle of rotation with limited volume of calculating processing in the calculation of the angle of rotation of a rotor based on an angular velocity signal. SOLUTION: A signal processing section which obtains an angular velocity signal corresponding to an angular velocity about a specified axis of a rotor to output an angle signal is contained to form an angle of rotation detector. The signal processing section extracts a high frequency component of the angular velocity signal offset corrected with a bypass filter 23 and the extracted high frequency component is subtracted from the angular velocity signal to determine a low frequency component contained in the angular velocity signal. The low frequency component is inputted into a dead zone processing section 25 to remove a low frequency noise component. The results are added and synthesized to the high frequency component and are sent to an integration circuit 26 to obtain an angle signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a rotation angle detecting device, and more particularly, to detecting a one-dimensional or multidimensional rotation angle (posture angle) of a rotating body (inertia body) with high accuracy and stability. Device.

[0002]

2. Description of the Related Art In recent years, for example, virtual reality (Ve)
For example, a sensor for realizing (real reality) or a pointing device for a computer such as a mouse or a digitizer requires a rotation angle detecting device for detecting a posture angle of a rotating body that rotates about a certain reference line. I have.

[0003] That is, in virtual reality,
For example, Head Mount Displ
ay; HMD; head-mounted display) equipped with this kind of rotation angle detection device, and provides the wearer of the HMD with an image corresponding to the posture angle of the head, so that the user can experience the world of virtual reality. Technology is being developed.

Further, in a pointing device for a computer, by incorporating this kind of rotation angle detecting device, the viewpoint and the position of a cursor on a computer screen are changed according to the attitude angle of the device. .

As such a rotation angle detecting device, a device using a magnetic sensor or a geomagnetic sensor to measure an angle, or a device using an angular velocity sensor such as a gyro sensor is known. However, a device using a magnetic sensor or a geomagnetic sensor uses magnetism (geomagnetism), so the range of use is limited, and the device is easily affected by the external environment. Since the circuit configuration for measuring the angle is complicated, there is a disadvantage that the cost is increased. For this reason,
Devices using an angular velocity sensor such as a gyro sensor that does not have such a defect are more widely used.

When the rotation angle is detected using a gyro sensor, generally, the angular velocity signal of the rotating body output from the gyro sensor is A / D converted (analog / digital conversion, the same applies hereinafter) and converted. A configuration is adopted in which after the subsequent angular velocity signal is subjected to noise removal by digital signal processing, an angle signal proportional to the rotation angle of the rotating body is obtained by integrating the obtained signal.

However, in this case, the offset component (0 output value) from the gyro sensor drifts from the reference voltage _Vref as shown by the hatched line in FIG. (DC component) is included as an error. For this reason, as shown by the shaded line in FIG. 11B, the errors accumulate in the angle signal obtained by integrating the output of the gyro sensor, and increase with time. This is an obstacle in accurately obtaining the rotation angle of the rotating body.

As one method for eliminating the error due to the drift, it is conceivable to use a high-precision gyro sensor such as a fiber gyro sensor instead of an inexpensive and small gyro sensor such as a vibration gyro sensor. However, such a method has disadvantages in that the size and weight of the rotation angle detection device are increased and the cost is increased. Therefore, it is difficult to incorporate the rotation angle detection device into a pointing device and spread it.

Further, as disclosed in Japanese Patent Publication No. 4-76611, there has been proposed a method of eliminating this kind of error by signal processing called dead zone processing. This dead zone processing stops the integration process when the output value of the gyro sensor signal is within a certain range from the reference value of the integration, that is, when the output value of the gyro sensor signal deviates from the range of the dead zone. This is a method of executing the integration process only during the deviated period. By setting the above range as a range where the offset output drifts as shown in FIG. 12A, the drift component can be effectively removed, and as shown in FIG. The rotation angle detection is accurate and stable.

[0010]

However, in the case of the rotation angle detecting apparatus using the above-mentioned dead zone processing, the signal processing itself is simple, but when performing a stable rotation angle detection, the range of the dead zone is increased. Must be set. However, when the rotation speed of the rotating body is low, the angular velocity signal from the gyro sensor has a small value. Therefore, if the range of the dead zone is too large, the movement of the rotating body rotating at a low speed cannot be detected.

On the other hand, in addition to the output from the gyro sensor, the output from another sensor such as a geomagnetic sensor is used, and these outputs are input-processed through a non-linear filter to obtain a more precise angle signal. A system proposal has also been made. This is, for example, GPS (Global Posit
It is used for interpolation of an output from a gyro sensor, an output from a magnetic sensor, a vehicle speed pulse output, and the like to determine a current position, an azimuth angle, and the like, for interpolation in an ioning system. However, in the method using the nonlinear filter, the calculation of the nonlinear filter itself is complicated, so that a heavy load is imposed on the signal processing unit, and the cost increases due to the use of a plurality of sensors.

SUMMARY OF THE INVENTION It is an object of the present invention to improve a rotation angle of a rotating body by processing an angular velocity signal so that a highly accurate rotation angle detection can be stably performed with a small amount of calculation processing. To provide a rotation angle detecting device.

[0013]

According to the present invention, there is provided a rotational angle detecting device for detecting an angular velocity of a rotating body about a predetermined axis, and an angular velocity signal obtained from the angular velocity detecting section for processing the rotational speed. A signal processing unit for determining a rotation angle of the body about the axis. The signal processing unit includes, for example, a dead band processing unit that removes a frequency variation component within a predetermined range included in a low frequency component of the angular velocity signal, and a composite value of a high frequency component of the angular velocity signal and an output of the dead band processing unit. And an integration processing unit that calculates the rotation angle by integrating the low frequency component or the high frequency component, and one of the low frequency component and the high frequency component is extracted from the angular velocity signal by a predetermined filter, and the other of the low frequency component or the high frequency component Is obtained by subtracting the output signal of the filter from the angular velocity signal.

With this configuration, the amount of calculation processing is reduced as compared with the case where the high-pass and low-pass filters are always used to extract the high-frequency component and the low-frequency component. Before (before integration)
, The offset drift component is removed.

When the dead zone processing section includes a means for outputting a signal of a different level in accordance with a result of comparison between the frequency fluctuation component and a predetermined threshold value, a plurality of threshold values different in steps are provided. Is preferably set in advance so as to enable multi-stage dead zone processing.

As a more preferred configuration of the present invention, the signal processing section is provided with an offset correction section for offset-correcting the angular velocity signal obtained from the angular velocity detection section. An offset value used for offset correction is, for example, an average value of offset drift components included in the angular velocity signal, or a value based on an offset drift component obtained by subtracting the output of the dead zone processing unit from the low frequency component,
It is dynamically updated sequentially. In this case, the integration processing unit is configured to calculate the rotation angle by integrating a composite component of a high-frequency component of the angular velocity signal subjected to the offset correction and an output of the dead zone processing unit.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the rotation angle detecting device according to the present invention having the above-mentioned configuration will be described in detail with reference to the drawings. As shown in FIG. 1, the rotation angle detection device includes an angular velocity detection unit 1 and a signal processing unit 2.
The angular velocity detecting unit 1 includes, for example, a gyro sensor and an amplifier for the output of the gyro sensor, detects an angular velocity of a rotating body (not shown) around a predetermined axis, and outputs an angular velocity signal 11 corresponding to the detected angular velocity. is there. Note that the output of the angular velocity detecting unit 1 is an analog signal in which an offset drift component is superimposed on the signal of the gyro sensor. Further, specifically, for example, a piezoelectric vibrating gyro can be used as the gyro sensor.

The signal processing unit 2 performs required signal processing on the angular velocity signal 11 from the angular velocity detecting unit 1 to obtain an angle signal. For example, a one-chip microcomputer (microcomputer) including a CPU and a memory element And DSP
(Digital signal processor) reads and executes a program stored in the memory element, thereby, for example, an A / D conversion unit 21, an offset correction unit 22, a high-pass filter 23, and a low-pass filter as shown in FIGS. The function of the filter 24, the dead zone processing unit 25, and the integration circuit 26 is realized.

The A / D converter 21 converts the analog angular velocity signal 11 into a digital signal. The offset corrector 22 removes (offset corrects) an offset drift component included in the digitally converted angular velocity signal. Specifically, this offset correction is performed by the angular velocity signal (ω _in ) converted by the A / D converter 21.
, By subtracting the offset value (ω _off ) output from the offset correction unit 22 from. As the offset value, for example, immediately after the power is turned on or after a certain time has elapsed, the angular velocity signal 11 from the angular velocity detection unit 1 is output.
Is read one or more times and the averaged value is used so that it is dynamically updated sequentially. Further, as will be described later, the difference between the input and output of the dead zone processing unit 25 can be used as a change in the offset value. The offset-corrected angular velocity signal is converted into a positive / negative corrected angular velocity signal (ω _n = ω _in −ω _off ) proportional to the angular velocity including the rotation direction.

High-pass filter 23 and low-pass signal 2
3 receives the angular velocity signal after the above-described offset correction, and outputs a high frequency component (ωH _n = HPF (ω _n )) and a low frequency component (ωL _n = LPF (ω _n )). here,
The high-pass filter 23 and the low-pass filter 24 have the same cutoff frequency. FIG. 5 shows an example of such a frequency characteristic.

FIG. 5A shows the frequency characteristics of the low-pass filter 24, and FIG.
3 shows the frequency characteristics of No. 3 respectively. This figure 5
In the cases (a) and (b), the low-pass filter 23 and the high-pass filter 24 each have a cutoff frequency of about 1.5 [Hz].

The cutoff frequency varies depending on the frequency component of the movement (rotation) of the rotating body to be detected and the frequency of the angular velocity signal from the angular velocity detector 1.
For example, when detecting the motion of a person, it is set to about 1 [Hz]. As the filter, a known digital filter such as FIR or IIR is used.

The dead zone processing section 25 includes a low-pass filter 2
4 is a low frequency component (ωL _n = L) of the angular velocity signal output from
PF (ω _n )).
The dead band process is to set the input signal to “0” when the low frequency component (ωL _n = LPF (ω _n )) is “0”, that is, within a fixed frequency fluctuation range from the time of the stationary state, that is, within the dead band. Assuming the output, otherwise output the input signal as it is.

FIG. 6 shows an example of the input / output characteristics in the dead zone processing section 25. Here, the threshold value TH _{1 is used} as a dead zone for the input signal ωL _n from the low-pass filter 24.
Range between the threshold TH ₂ is set to. When the input signal .omega.L _n is in the range between the thresholds TH _1, TH _2, the dead zone processing result in the dead zone processing unit 25, an output signal _{(ωL' n = f (ωL n} )) to " 0 ". On the other hand, if it is outside the above range, as the output signal OmegaL' _n, as to output the value of the input signal .omega.L _n. In this way, that all input signals .omega.L _n in the dead zone is "0", and remove the offset drift component which is a low-frequency noise contained in the dead band.

FIG. 7 shows an example in which the range in which the dead zone processing is performed on the input signal ωL _n is further subdivided, that is, an example in which the multi-stage dead zone processing is performed. In this example, four thresholds T at different levels are used.
H ₁ , TH ₂ , TH ₃ , and TH ₄ are set, and in sections I to V set corresponding to these thresholds TH ₁ , TH ₂ , TH ₃ , and TH ₄ , for example, output signal OmegaL' _n as is to be defined.

Section I ωL _n <TH ₁ : ωL ′ _n = ωL _n Section II TH ₁ ≦ ωL _n <TH ₂ : ωL ′ _n = ωL _n −TH ₂ Section III TH ₂ ≦ ωL _n <TH ₃ : ωL ′ _n = 0 section _{IV TH 3 ≦ ωL n <TH} 4: ωL' n = ωL n -TH 4 section _{V TH 4 ≦ ωL n: ωL'} n = ωL n

By performing such multi-stage dead zone processing, the area of the complete dead zone (section III), that is, the area where "0" is output can be made smaller than that in the case of FIG. Further, the dead zone processing can be performed at a finer level than in the case of FIG. For this reason, the range of the detectable angular velocity is widened, and it is possible to prevent the motion of the rotating body rotating at a low speed from being undetectable.

The output signal OmegaL' _n of the dead zone processing unit 25,
The output (HPF (ω _n )) of the high-pass filter 23 is added and synthesized. Also, the composite value (ω ′ _n = ωL ′ _n + HP
F (ω _n )) is input to the integration circuit 16. The integration circuit 16 performs an integration process on the input composite value based on an algorithm such as the trapezoidal method, and outputs an angle signal representing a value proportional to the rotation angle.

[0029]

【Example】

(First Embodiment) Next, a specific operation in the signal processing section 2 will be described. For example, the signal processing unit 2 having the configuration shown in FIG. 2 performs angle detection according to the procedure shown in FIG. That is, first, in advance to determine the offset value omega _off in the manner described above (step S11), and performs A / D conversion of the angular velocity signal 11 in the A / D converter 21 (step S1
2). Thereafter, the offset correction unit 22 performs an offset correction process (ω _n = ω _in −ω _off ) (step S13),
Next, the high-pass filter 23 calculates a high-pass filter for passing high-frequency components,
, Low-pass filter calculations for passing low-frequency components are performed (steps S14 and S15).
Note that the order of each filter calculation may be reversed or may be simultaneous. With respect to the low-pass filter 24, the dead band processing unit 25 further performs dead band processing (step S16), and the signal ωL ′ _n after the dead band processing is performed.
And adding combines the output signal .omega.H _n of the high-pass filter 23 (step S17). Thereafter, the integration circuit 26 performs integration processing (θ _n = ∫ω ′ _n dt) to obtain an angle signal (step S18). If the next angular velocity signal 11 has been input, the processing after step S12 is repeated, and if there is no more angular velocity signal 11 to be processed, the processing ends (step S19).

As described above, in the first embodiment, the offset drift component is removed by performing the offset correction on the digitally converted angular velocity signal, and the low frequency component included in the angular velocity signal after the offset correction is converted into the dead band processing section. 25, the error component is removed, and the low-frequency signal 0 component subjected to the dead zone processing is added and synthesized with the high-frequency signal component included in the angular velocity signal, so that the integration process is performed. A rotation angle with small fluctuation and a rotation angle with large frequency fluctuation than the offset drift component can be effectively detected, and an angle signal with high accuracy can be obtained. At this time, the dead zone processing unit 2
In 5, by performing the multi-step dead zone processing as shown in FIG. 7, it is possible to detect the rotation angle corresponding to the fine movement of the rotating body.

(Second Embodiment) As a modification of the first embodiment, even if the low-pass filter 24 is omitted, the first embodiment
A processing result similar to that of the embodiment can be obtained. FIG. 3 shows an example of such a configuration. That is, in this embodiment, the output signal HPF (ω _n ) of the high-pass filter 23 is subtracted from the angular velocity signal (ω _n ) subjected to the offset correction in the offset correction unit 22,
Extracting the low frequency component .omega.L _n of the angular velocity signal omega _n, are configured to enter the low-frequency component to the dead-zone processing unit 25. Other configurations and operations are the same as those in the first embodiment.

The procedure of angle detection by the signal processing unit 2 of the second embodiment configured as shown in FIG. 3 is as shown in FIG. 9, and the low-pass filter calculation is performed by "ωL _n = ω _n -ωH _n ".
The processing is substantially the same as the processing in FIG. 8 except that the processing is performed by the calculation processing (step S15 ′). According to the second embodiment, even if the low-pass filter 24 is omitted, the rotation angle can be detected in substantially the same procedure as in the first embodiment.
Calculation processing is reduced, and the load on the signal processing unit 2 is reduced.

(Third Embodiment) Further, in the second embodiment, the offset value used in the offset correction unit 22 is dynamically updated so that the offset value is always set to the optimum value for removing the offset drift component. Other configurations are also possible. FIG. 4 shows a configuration example of the signal processing unit 2 in this case.

[0034] In the third embodiment, in the configuration of FIG. 3, it subtracts the output signal OmegaL' _n of the dead zone processing unit 25 from the low-frequency component .omega.L _n of the angular velocity signal omega _n. That is, the difference between the input and output of the dead zone processing unit 25 is calculated. Input signal to the dead zone processing unit 25 is composed of an offset drift [Delta] [omega _OFFN a low frequency component .omega.L _n and low frequency noise of the angular velocity signal omega _n, also components of the offset drift [Delta] [omega _OFFN is removed by the dead zone processing unit 25 Therefore, the difference between the input and output of the dead zone processing unit 25 is the offset drift Δω _offn . The offset drift component Δω _offn (= ωL _n −ωL ′ _n ) obtained in this manner is converted to the offset correction unit 2
Give feedback to 2.

The offset correction section 22 multiplies the offset drift component Δω _offn fed back by a constant correction coefficient, and then adds it to the latest offset value currently set. That is, assuming that the offset value at time n is ω _offn and the correction coefficient is α, the offset value at time n + 1 is updated as follows.

[0036]

[ _Equation 1] ω _{offn + 1} = α · (ω _offn + △ ω _offn )

The value of the correction coefficient α changes depending on the drift characteristics of the gyro sensor constituting the angular velocity detector 1 and the like. Normal is "1", also, the _ω offn
If the angle output is too large to stabilize the angle output, an optimum value can be determined according to the characteristics of the system, such as setting a value smaller than "1".

The procedure of angle detection by the signal processing unit 2 having the configuration shown in FIG. 4 is as shown in FIG. The steps here are
The procedure according to the second embodiment is different from the procedure according to the second embodiment in that an offset drift component extraction process (step S19) and an offset value update process (step S20) are further added to the procedure according to the second embodiment shown in FIG. It becomes the same.

As described above, by updating the offset value with the offset drift component obtained from the difference between the input and output of the dead zone processing unit 25, the offset value changes with time, for example, the offset value due to factors such as temperature drift. Fluctuation can be prevented, and highly accurate angle detection can be stably performed.

In the second and third embodiments, the case where the low-pass filter 24 is omitted and the low-frequency component of the angular velocity signal ωn is calculated using the output of the high-pass filter 23 has been described. omitted, a low-pass filter 24 a high-frequency component .omega.H _n of the angular velocity signal omega _n
Can also be calculated using the output of In this case, the high frequency component .omega.H _n of the angular velocity signal ωn is calculated as "ω _n -ωL _n". Thus, by replacing the filter calculation processing in one filter with a simple subtraction processing,
Processing in the signal processing unit is greatly reduced.

[0041]

As is clear from the above description, according to the rotation angle detecting device of the present invention, when calculating the rotation angle of the rotating body by processing the angular velocity signal, it is possible to obtain a high accuracy with a small amount of calculation processing. The rotation angle can be stably detected. As a result, even when a sensor having a low accuracy of the angular velocity signal is used, the rotation angle can be detected with high accuracy, and the conventional problems can be solved.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of a rotation angle detection device of the present invention.

FIG. 2 is a configuration diagram of a signal processing unit according to the first embodiment.

FIG. 3 is a configuration diagram of a signal processing unit according to a second embodiment.

FIG. 4 is a configuration diagram of a signal processing unit according to a third embodiment.

5A is a frequency characteristic graph of a low-pass filter included in a signal processing unit, and FIG. 5B is a frequency characteristic graph of a high-pass filter.

FIG. 6 is an input / output characteristic graph of a dead zone processing unit constituting the signal processing unit.

FIG. 7 is an input / output characteristic graph showing the contents of multi-stage dead zone processing in the dead zone processing unit.

FIG. 8 is an explanatory diagram of a procedure of angle detection according to the first embodiment.

FIG. 9 is an explanatory view of a procedure of angle detection according to a second embodiment.

FIG. 10 is an explanatory diagram of a procedure of angle detection according to a third embodiment.

11A is an explanatory diagram of an angular velocity signal from a gyro sensor as an example of an angular velocity detecting unit, and FIG. 11B is an explanatory diagram of an angular signal obtained by integrating the angular velocity signal.

12A and 12B are explanatory diagrams of dead zone processing of an angular velocity signal.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Angular velocity detection part 2 Signal processing part 21 A / D conversion part 22 Offset correction part 23 High pass filter 24 Low pass filter 25 Dead zone processing part 26 Integration circuit

Continued on the front page (72) Inventor Satoshi Takahata 2-26-10 Tenjin, Nagaokakyo-shi, Kyoto Murata Manufacturing Co., Ltd.・ Tech Inside (72) Inventor Tsutomu Miyasaka 4-42-12 Kamata, Ota-ku, Tokyo Shinsei Building Data Tech Co., Ltd.

Claims (5)

[Claims] 1. An angular velocity detecting section for detecting an angular velocity of a rotating body about a predetermined axis, and a signal processing section for processing an angular velocity signal acquired from the angular velocity detecting section to obtain a rotation angle of the rotating body about the axis. A dead band processing unit that removes a frequency variation component within a predetermined range included in a low frequency component of the angular velocity signal, and a high frequency component of the angular velocity signal and an output of the dead band processing unit. An integration processing unit that integrates a composite value to calculate the rotation angle, wherein one of the low frequency component or the high frequency component is extracted from the angular velocity signal by a predetermined filter, and the low frequency component or the high frequency component is The other of the two is obtained by subtracting the output signal of the filter from the angular velocity signal. 2. The method according to claim 1, wherein the dead zone processing unit includes a unit that outputs a signal of a different level in accordance with a result of comparison between the frequency fluctuation component and a predetermined threshold. 2. The rotation angle detecting device according to claim 1, wherein the following value is set. 3. The signal processing unit further includes an offset correction unit configured to remove an offset drift component included in the acquired angular velocity signal, and the integration processing unit performs processing on the angular velocity signal from which the offset drift component has been removed. 3. The rotation angle detection device according to claim 1, wherein the rotation angle is calculated by integrating a composite value of a high frequency component and an output of the dead zone processing unit. 4. The offset correction unit according to claim 3, wherein the offset correction unit performs the offset correction based on an offset drift component obtained by subtracting an output of the dead band processing unit from a low frequency component of the angular velocity signal. The rotation angle detection device according to any one of the preceding claims. 5. The rotation angle detecting device according to claim 4, wherein an offset value used for said offset correction is dynamically updated.