Sensor apparatus, sensor module, integrated circuit, inclination angle estimation method, and azimuth detection method

Abstract

To provide a sensor apparatus, an azimuth detection device, a sensor module, and an inclination angle estimation method capable of estimating an inclination angle without any dependence on a scale factor.SOLUTION: A sensor apparatus comprises: a mechanism for rotating a rotation system while an accelerometer having an axis not parallel to any of a surface vertical to a rotation axis and the rotation axis is provided in the rotation system that is rotatable with the rotation axis as a center; and a processor that acquires acceleration measured by the accelerometer in accordance with rotational angles in three or more points of the rotation system to be rotated, separates the acceleration into a first component following the rotation of the rotation system and a second component not following the rotation, and estimates an inclination angle of the rotation system on the basis of the first component and the second component.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present embodiment relates to a sensor device, an orientation detection device, a sensor module, and an inclination angle estimation method. [Background technology]

[0002] Magnetism is commonly used to detect true north. However, the geomagnetic field is very weak and is strongly affected by nearby magnetized metals, so it often does not indicate the correct north. GPS (Global Positioning System) options include a function to display direction, but this cannot be used indoors where line of sight to GPS satellites cannot be ensured.

[0003] One method for estimating true north with high accuracy without being affected by surrounding buildings is to detect the rotation of the Earth using a gyroscope (Non-Patent Document 1).

[0004] This will be explained using Figure 14. (a) shows a sphere that resembles the Earth, and the Earth moves at an angular velocity ω ie Consider an example of detecting true north at the position P of the black circle. The axes that form the NED (North-East-Down) coordinate system at the position P are indicated by arrows. (b) is a view of this from the side on a horizontal plane. The Earth rotates around an axis pointing upward in the figure. The position of P is at latitude L [rad]. At P, the angular velocity ω ie appears as an upward rotation in the diagram. This rotation is projected onto the rotation axis in the direction of the north vector North with a magnitude of cosL. In other words, if a gyro with an axis in the North direction is placed at position P, then ω ie The rotation of the Earth can be observed by the magnitude of cosL. (c) shows P viewed directly above in the NED coordinate system, that is, from above the horizon. The axis of the gyro is not north, but α from true north. i Consider the case where the angle is rotated by (i=1, 2) on the horizontal plane. The projection of the Earth's rotation to the north direction is ω ie cosL is further defined as α i The angular velocity of the Earth's rotation detected when the gyro axis is placed in each direction is ω iecosLcosα i ω ie cosL=ω iem Then, ω iem It becomes cosα, and when α is rotated from 0 to 2π, the angular velocity of the Earth's rotation changes sinusoidally. When rotated one full rotation, the angular velocity of the Earth's rotation is observed to be greatest when α is due north.

[0005] Therefore, true north can be determined by placing a highly sensitive gyro capable of measuring the Earth's rotation on a rotating platform and detecting the angle of the gyro axis that peaks in the waveform of angular velocity obtained by rotating the axis once on a horizontal plane. Because it detects the Earth's rotation, it can be detected regardless of the radio wave or magnetic environment. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Yongjian Zhang, Bin Zhou, Mingliang Song, Bo Hou, Haifeng Xing and Rong Zhang, “A Novel MEMS Gyro North Finder Design Based on the Rotation Modulation Technique,” Sensors (Switzerland) vol.17 (5), Article number 973, 2017 Summary of the Invention [Problem to be solved by the invention]

[0007] To accurately detect true north, the gyro axis must be rotated completely in a horizontal plane. If the rotating platform is tilted slightly east-west around the North axis, it will behave the same as if the surface of the rotating platform were placed at a different latitude and longitude point that is perfectly aligned with the horizontal plane. Therefore, the north angle will not indicate the longitude of P, but the north at the latitude and longitude of the other point where the surface of the rotating platform is aligned with the horizontal plane.

[0008] It depends on the required accuracy of true north, but if an accuracy of less than 1 degree is required, the turntable must be equally level, which is not easy to achieve.

[0009] Regarding this problem, it has been shown that even if the turntable is not placed perfectly horizontally, errors caused by not placing it perfectly horizontal can be corrected by installing an accelerometer with an axis on the turntable to detect the direction of gravity and measure the tilt (Non-Patent Document 1).

[0010] The tilt of the rotating platform is actually expressed as two angles, pitch and roll, relative to the axis of the gyro or accelerometer, but for simplicity's sake, we will explain it as a single tilt angle θ, assuming that the axis of the accelerometer is tilted by θ relative to the axis on the rotating platform surface that is perpendicular to it. If an accelerometer is installed with its axis horizontal on a tilted rotating platform as shown in Figure 15(a), and the rotating platform is rotated once, the projection of gravity onto the axis of the accelerometer will also change in a sinusoidal shape. θ can be determined from the amplitude of the detected acceleration sine wave.

[0011] However, this method has problems. Accelerometers detect acceleration in the form of voltage. This voltage is detected by multiplying the acceleration by a coefficient called the scale factor (SF). Accelerometers made from MEMS (Micro Electro Mechanical Systems) are highly temperature dependent, and the SF can fluctuate with temperature. There are limitations to calibration, and complete correction is difficult due to factors such as aging.

[0012] If there is SF fluctuation, the accuracy of the estimated tilt angle will deteriorate using the above method. As shown in Figure 15(b), when the accelerometer axis is tilted by θ, the component of gravity G detected is Gsinθ. Suppose there is SF fluctuation and the detected voltage is SF1*Gsinθ at one point and SF2*Gsinθ at another point. If the SF at that point is not recognized as SF2 and the acceleration and therefore the tilt angle are calculated using SF1, the angle detected will be θ', where SF2*Gsinθ=SF1*Gsinθ'.

[0013] In the worst case scenario, the difference between θ and θ' will directly appear as an error in the true north angle. Therefore, this method is highly dependent on SF and is not suitable for high-precision true north detection.

[0014] Instead of detecting tilt by rotating a single-axis accelerometer on a rotating platform, it is possible to return to the basics of tilt detection and use a two- or three-axis accelerometer. However, this problem cannot be solved unless the SFs of the multiple axes are always consistent or always known correctly. Furthermore, using multiple axes increases the cost of the device.

[0015] The present disclosure solves such problems and provides a sensor device, an orientation detection device, a sensor module, and an inclination angle estimation method that are capable of estimating an inclination angle without SF dependency. [Means for solving the problem]

[0016] The sensor device of this embodiment includes a rotational system that can rotate around a rotational axis, and an accelerometer having an axis that is not parallel to either a plane perpendicular to the rotational axis or the rotational axis is provided in the rotational system, and the sensor device includes a mechanism for rotating the rotational system, and a processing unit that acquires acceleration measured by the accelerometer corresponding to the rotation angle of three or more points in the rotated rotational system, separates the acceleration into a first component that follows the rotation of the rotational system and a second component that does not follow the rotation, and estimates the tilt angle of the rotational system based on the first component and the second component. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a diagram illustrating a sensor device for measuring tilt according to a first embodiment of the present invention. [Figure 2] FIG. 2 is a diagram for explaining the effect of the first embodiment of the present invention. [Figure 3] 1 is a block diagram showing an example of the configuration of a sensor device for measuring tilt according to a first embodiment of the present invention. [Figure 4] FIG. 3 is a diagram showing an example of the flow of a tilt estimation method according to the first embodiment of the present invention. [Figure 5]FIG. 10 is a diagram illustrating a sensor device for detecting true north according to a second embodiment of the present invention. [Figure 6] FIG. 10 is a block diagram showing an example of the configuration of a sensor device that detects true north according to a second embodiment of the present invention. [Figure 7] FIG. 10 is a diagram showing an example of the flow of a true north detection method according to a second embodiment of the present invention. [Figure 8] FIG. 2 is a diagram for explaining the range of μ for effectively operating the first or second embodiment of the present invention. [Figure 9] FIG. 2 is a diagram for explaining the range of μ for effectively operating the first or second embodiment of the present invention. [Figure 10] FIG. 10 is a diagram showing an example of the layout of a MEMS integrated circuit according to a third embodiment of the present invention. [Figure 11] 10A and 10B are diagrams illustrating an example of a sensor module according to a fourth embodiment of the present invention. [Figure 12] FIG. 10 is a diagram illustrating an example of a true north detecting device according to a fifth embodiment of the present invention. [Figure 13] FIG. 10 is a diagram illustrating an example of a true north detecting device according to a sixth embodiment of the present invention. [Figure 14] FIG. 1 is a diagram for explaining a method for estimating true north in the background art. [Figure 15] FIG. 1 is a diagram for explaining a method for estimating a tilt in the background art. DETAILED DESCRIPTION OF THE INVENTION

[0018] The following description will be made with reference to the accompanying drawings, in which only essential components of the present invention are shown, and components not directly related to the operation of the present invention are omitted from the drawings and description.

[0019] (First embodiment) Figure 1 shows one embodiment of the present invention. The sensor device 100 is a device that detects the tilt of a rotating platform 2, and includes the rotating platform 2 and an accelerometer 1. A detailed configuration is shown in Figure 3, which will be described later. The rotation mechanism of the rotating platform 2 and the lines for outputting data from the accelerometer 1 are omitted.

[0020] Accelerometer 1 is installed on a rotating table 2. In the figure, the rotation axis of the rotating table is z s The rotation table surface is z s is perpendicular to z s axis x in a plane perpendicular to s Accelerometer 1 is a single-axis accelerometer, and its detection axis is z s But z s It is not on a plane perpendicular to the plane of the sphere, but is at an angle intentionally tilted from both. s and x s is in the plane defined by z s The axis of the accelerometer tilted by μ is referred to as the μ axis from here on. The μ axis corresponds to the accelerometer axis direction 3. Note that x s Regarding z, s The plane with the μ axis is z s It is defined as the direction of a line that intersects with a plane perpendicular to the

[0021] Turntable 2 is z s The axis of rotation is the axis of rotation, and the rotation is performed in the direction of the arrow in the figure, but it can also be performed in the opposite direction. The rotation is performed in the direction of the arrow in the figure, and there are two main ways to measure the rotation and acceleration.

[0022] The first method is to measure acceleration while rotating, and the second is to measure by repeating the cycle of measuring while standing still, then rotating, then measuring while standing still, then rotating again. In the first method, the accelerometer is rotated at a constant speed. In either method, the accelerometer must be rotated at least approximately one full revolution. This is because the measured acceleration changes in a sinusoidal shape as the rotation occurs, and it is necessary to detect the amplitude and phase of this sinusoidal wave.

[0023] For example, four points can be defined at 90-degree intervals within one revolution, and measurements can be taken from 0 degrees → 90 degrees → 180 degrees → 270 degrees. In this case, there is no need to return to 0 degrees at the end, so it does not make a full revolution, but it does make enough of a revolution to determine the amplitude and phase of the sine wave, i.e., it makes almost one revolution.

[0024] The amplitude and phase of a sine wave can be calculated using the Fourier transform. According to the Fourier transform and sampling theorem, the amplitude of a sine wave can be estimated from measurements as long as there are at least three points within a 360-degree range. The point with the maximum amplitude does not necessarily need to be included in the measurements.

[0025] In the case of a three-point example, measurements should be taken at three points: 0 degrees, 120 degrees, and 240 degrees.

[0026] Basically, the measured acceleration is Fourier transformed, so it does not necessarily have to end with one revolution, and multiple revolutions are also acceptable. It is not necessary to perform a Fourier transform for all well-known frequencies. A specified frequency, or in this case, a specified phase, for example, 0 degrees → 90 degrees → 180 degrees → 270 degrees, can be used, and a DFT (Discrete Fourier Transform) can be used to estimate the coefficients. In other words, each measurement value is multiplied by exp(-jα), where α is the rotation angle, and the multiplication results for all points are added together.

[0027] The acceleration measured by an accelerometer with μ axis can be divided into a component that changes sinusoidally with rotation and a component that does not change with rotation (DC component). Note that the accelerometer is properly calibrated, and the bias component of the accelerometer is removed from the detected DC component before use.

[0028] The tilt angle is calculated from the DC component and the amount including the phase of the rotational component, but since both are separated from measurements on one axis, the scale factor (SF) is always the same for both, so no errors depending on the SF occur.

[0029] A specific explanation will be given below using formulas.

[0030] The above explanation was given using only one tilt angle for ease of explanation, but tilt is usually defined using two angles: pitch angle θ and roll angle γ. Another definition of rotation is yaw angle, which in this case is the deviation from true north, or north angle φ.

[0031] Hereafter, we use the NED (North East Down: XYZ are ordered northeast-down, and for Z-axis rotation, right rotation is positive when viewed from above) system as the global frame. The system with the tilted turntable is the sensor frame, which is defined as a right-handed system with the z-axis pointing down, following the NED system.

[0032] TIFF2025118327000002.tif106170

[0033] TIFF2025118327000003.tif39170

[0034] TIFF2025118327000004.tif29170

[0035] TIFF2025118327000005.tif72170

[0036] TIFF2025118327000006.tif72170

[0037] TIFF2025118327000007.tif23170

[0038] Figure 2 shows the results of a simulation of the effects of this embodiment. (a) is the method of the background art, and (b) is this embodiment. The top row is the pitch angle error, and the bottom row is the roll angle error. One revolution is divided into four parts, and acceleration is measured at 90-degree intervals. Noise is added, but it is very small. In (a), the roll angle is fixed at -10 degrees, and the pitch angle θ is varied on the horizontal axis from 0 degrees to 10 degrees. Each line represents the case where the SF error is changed from 0.001 to 0.05. When the horizontal axis θ is large and when the scale factor error is large, the error increases, reaching an error of around 0.5 degrees at its largest.

[0039] (b) shows the present embodiment. Because this embodiment is not affected by the scale factor, the horizontal axis parameter is the absolute value of the residual bias of the μ-axis accelerometer, and the parameter is the μ angle. The pitch angle is fixed at 10 degrees, and the roll angle is fixed at -10 degrees, which corresponds to the condition where the horizontal axis θ = 10 degrees in (a), i.e., the condition where the error is maximized in (a). As will be described later, if the accelerometer contains residual bias that was not fully calibrated, an error occurs, and the magnitude of the error varies depending on μ. Even if μ is changed in various ways, if the residual bias is around 10 mG, the error is at most around 0.15 degrees. Considering that the residual bias of a high-precision accelerometer is usually less than 1 mG, it can be seen that this is sufficiently small compared to the error in (a) when θ = 10 degrees.

[0040] In this way, a single axis accelerometer can be rotated to obtain tilt that is independent of SF.

[0041] FIG. 3 shows a block diagram of a sensor device 100 according to this embodiment. The sensor device 100 includes an accelerometer 1, a control unit 71, an acceleration acquisition and pre-processing unit 72, and an inclination angle calculation unit 73. The acceleration acquisition and pre-processing unit 72 and the inclination angle calculation unit 73 correspond to the processing units that perform the processing of this embodiment. The accelerometer 1 is fixed on a rotating table 2. The control unit 71 instructs the rotating table 2 or the rotation mechanism 2A that rotates the rotating table 2 to set an angle, and also issues an acceleration acquisition instruction to the acceleration acquisition and pre-processing unit 72 in accordance with the angle of the rotating table. The acceleration output from the accelerometer 1 is pre-processed as necessary by the acceleration acquisition and pre-processing unit 72. For example, when measurements are performed at four angles at 90-degree intervals, the control unit 71 rotates the rotating table and the rotating table comes to a stable standstill at each specified angle. Upon receiving an acceleration acquisition instruction from the control unit 71, the acceleration acquisition and pre-processing unit 72 acquires acceleration for a predetermined period of time. The acceleration acquisition and pre-processing unit 72 applies temperature correction and other adjustments to the acquired acceleration as necessary, and calculates the average, sum, or low-pass filtered value (hereinafter collectively referred to as the average) over the acquisition period. It performs similar pre-processing for the four angles and outputs the resulting pre-processed acceleration to the inclination angle calculation unit 73. The inclination angle calculation unit 73 separates the pre-processed angular velocity into a rotational component and a DC component using the above algorithm, estimates SF from these components, and uses the estimated SF to estimate and output the inclination angles (pitch angle and roll angle). The output inclination angles are input as necessary to a display unit (not shown), other devices, the north angle calculation unit of a true north detection device, a PC, etc.

[0042] As for the blocks other than the accelerometer 1, some or all of them may be built into the accelerometer module, or they may be located outside the turntable, and the acceleration data may be extracted and processed outside the turntable via wired or wireless communication.

[0043] FIG. 4 shows an example of the flow of an inclination estimation method according to one embodiment of the present invention. The control unit 71 instructs the rotation platform to set an angle (S200) and issues an instruction to acquire acceleration at three or more known rotation angles (S201). The acceleration output from the accelerometer 1 is pre-processed as necessary by the acceleration acquisition and pre-processing unit 72. The inclination angle calculation unit 73 separates the pre-processed acceleration into a rotational component and a DC component using the above algorithm (S202), estimates the SF from these components (S203), and uses the estimated SF to estimate and output the inclination angles (pitch angle and roll angle) (S204).

[0044] It should be noted that what is truly necessary to implement the present invention is not a rotating table in the literal sense. s Any mechanism other than a turntable can be used as long as it rotates around the axis of rotation. Accelerometer 1 is assumed to be made of MEMS, but it can also be made of other materials such as quartz.

[0045] Note that when measurements are taken while rotating, α is continuous, but when stationary measurements and rotation are repeated, α becomes discrete. Nowadays, even with continuous measurements, values are always output discretely according to the ODR (Output Data Rate), so there is no difference in the formula between the two methods when it comes to inclinometer applications. On the other hand, with the true north detection device described below, when measurements are taken while rotating, if the gyro axis is not perfectly perpendicular to the rotation axis, the rotation of the rotating table will leak onto the gyro axis and add to the Earth's rotation, degrading the phase detection accuracy. When measuring while rotating, a highly accurate rotating table is required.

[0046] (Second embodiment) Next, as a second embodiment, a sensor device including the above-described inclinometer for detecting true north will be described with reference to FIG.

[0047] 5, a sensor unit 4 including a gyro and an accelerometer is mounted on the rotating table 2. Of course, the gyro and the accelerometer do not have to be contained in a single module, and individual modules may be mounted on the rotating table 2.

[0048] The accelerometer axis is the same as in Figure 1. The gyro axis is x s In effect, the gyro axis direction 5 is aligned with the x axis. s The principle of true north detection is as described in the background art. As with tilt measurement, the turntable 2 is rotated almost one revolution or more, and the angular velocity at each rotation angle is measured. The magnitude of the angular velocity changes sinusoidally in response to changes in the angle of the gyro axis. If there is no tilt, the angle of the axis that shows the maximum value is the north angle. Of course, since the angle that shows the minimum value is due south, it is also possible to detect the minimum value.

[0049] Assuming that the tilt is not 0 degrees, we will explain the north angle correction method using the tilt measured by an inclinometer. Note that in the method in Non-Patent Document 1, the north angle correction also has an error due to the SF of the gyro, but the following method is not affected by the SF error of the gyro.

[0050] TIFF2025118327000008.tif109170

[0051] TIFF2025118327000009.tif46170

[0052] TIFF2025118327000010.tif34170

[0053] TIFF2025118327000011.tif51170

[0054] TIFF2025118327000012.tif27170

[0055] TIFF2025118327000013.tif22170

[0056] Ambiguity occurs because the equation is squared when solving. This is resolved using the following algorithm. When R is calculated, it should be solved using the four-quadrant arctangent, but since it becomes a simple arctangent, the sign before making it into a ratio is used.

[0057] TIFF2025118327000014.tif24170

[0058] By doing so, it is possible to estimate a tilt-corrected north angle that is independent of both the SF of the accelerometer and the SF of the gyro.

[0059] FIG. 6 shows a block diagram of a sensor device 101 according to this embodiment. The sensor device 101 includes a sensor unit 4, a control unit 71, an acceleration / angular velocity acquisition / pre-processing unit 74, an inclination angle calculation unit 73, and a north angle calculation unit 75. The acceleration / angular velocity acquisition / pre-processing unit 74, the inclination angle calculation unit 73, and the north angle calculation unit 75 correspond to the processing units that perform the processing of this embodiment. The sensor unit 4 is fixed on the rotating table 2. The control unit 71 instructs the rotating table 2 or the rotation mechanism 2A that rotates the rotating table 2 to determine an angle, and also issues an instruction to the acceleration / angular velocity acquisition / pre-processing unit 74 to acquire acceleration and angular velocity according to the angle of the rotating table. The acceleration and angular velocity output from the sensor unit 4 are subjected to pre-processing as necessary by the acceleration / angular velocity acquisition / pre-processing unit 74. For example, if the rotating platform is to perform measurements at four angles spaced 90 degrees apart, after the rotating platform has stabilized at each angle, the acceleration / angular velocity acquisition / pre-processing unit 74 receives an acceleration / angular velocity acquisition command from the control unit 71 and acquires acceleration and angular velocity for a predetermined period. For each of these, temperature corrections and other adjustments are made as necessary, and the average value for that period is calculated. The same pre-processing is performed for each of the four angles, and the resulting pre-processed acceleration is output to the tilt angle calculation unit 73, and the pre-processed angular velocity is output to the north angle calculation unit 75. The tilt angle calculation unit 73 estimates the tilt angle (pitch angle and roll angle). The north angle calculation unit 75 extracts the rotation component from the input pre-processed angular velocity using a process similar to the Fourier transform described above, and either calculates the north angle (in the case of a method in which the pre-corrected north angle is obtained and then tilt correction is performed), or extracts the phase calculation parameters before that. The tilt angle (pitch angle and roll angle) estimated from the tilt angle calculation unit 73 is input, and the tilt-corrected north angle is calculated and output using these tilt angles and phase calculation parameters. The output north angle is input to a display unit 76, or, if necessary, to another device (not shown), such as a PC.

[0060] The blocks other than the sensor unit 4 are located outside the turntable, and the acceleration and angular velocity data may be extracted and processed outside the turntable via wired or wireless communication, or some or all of them may be built into the sensor unit 4.

[0061] FIG. 7 shows an example of the flow of a true north detection method according to one embodiment of the present invention. The control unit 71 instructs the rotation platform to set an angle (S300) and issues an instruction to acquire acceleration and angular velocity at three or more known rotation angles (S301). The acceleration and angular velocity output from the sensor unit 4 are pre-processed as necessary by the acceleration and angular velocity acquisition and pre-processing unit 74. The pre-processed acceleration and angular velocity are output to the tilt angle calculation unit 73 and the north angle calculation unit 75, respectively. The north angle calculation unit 75 extracts the Earth's rotation component from the input pre-processed angular velocity using a process similar to the Fourier transform described above (S302), and extracts phase calculation parameters from the phase of the extracted rotation component (S303). The tilt angle calculation unit 73 estimates the tilt angle (pitch angle and roll angle) from the acceleration (S304). The north angle calculation unit 75 calculates and outputs the tilt-corrected north angle using the phase calculation parameters and the estimated tilt angles (pitch angle and roll angle) (S305).

[0062] TIFF2025118327000015.tif41170

[0063] TIFF2025118327000016.tif29170

[0064] Thus, there exists an optimal value of μ, which will be clarified below.

[0065] With both the pitch angle and roll angle set to 5 degrees, the total estimated angle error for the bias that could not be completely removed is (θ error^2 + γ error^2) 1 / 2 When plotting [deg] with μ as a parameter, the result is shown in Figure 8(a). However, the noise σ is 0.0069 [m / s 2] and is assumed to be filtered by a low-pass filter equivalent to the Nyquist frequency determined by the sampling frequency. In other words, the noise is assumed to be Gaussian noise and independent for each sample.

[0066] The number of samples n used for estimation is 60,000. In estimation theory, the effective noise decreases in inverse proportion to the square root of the number of samples. In this example, it is σ / √n. However, if a low-pass filter with a cutoff frequency significantly lower than the Nyquist frequency is used, the degree of effective noise reduction relative to the number of samples decreases and becomes larger than σ / √n. Taking this into consideration, the effective noise due to n samples is σ e In the example of Figure 8(a), σ / √n=2.8*10 -5 [m / s 2 ], but it will be larger if the cutoff frequency is low.

[0067] In FIG. 8(a), in the vicinity of 0 on the horizontal axis, the smaller μ is, the larger the error, but as the horizontal axis increases, the larger μ is, the larger the error becomes.

[0068] TIFF2025118327000017.tif29170

[0069] TIFF2025118327000018.tif33170

[0070] These are plotted in Figure 8(b). The horizontal line is determined by noise. n , the line with an upward sloping b , which depends on the bias. Figure 8(a) shows the root-sum-square values of these.

[0071] TIFF2025118327000019.tif29170

[0072] Figure 9 (a) shows the total error (θ error^2 + γ error^2) at a residual bias of 1 mg for the case of 0.1 degrees, and (b) for the case of 20 degrees. 1 / 2The horizontal axis is the angle μ, with the dashed line representing the error due to noise, the dashed line representing the error due to residual bias, and the red line representing the sum of these (square root of the sum of squares). In either case, there is a μ at which the error is minimal. In the example of Figure 9(a), the μ:k1 at which the error is minimal is 61 degrees, and in the example of Figure 9(b), the μ:k2 at which the error is minimal is 7 degrees.

[0073] If there is no tilt at all, there will be no minimum point and the minimum will be 90 degrees. However, since it is virtually impossible to place the device perfectly horizontally, the minimum value is set to the expected value of the tilt due to installation error when placed horizontally. Here, this is set to 0.1 degrees. Also, the maximum tilt angle is set to 20 degrees as an operating condition of the device.

[0074] The optimum μ varies depending on the conditions, but since the optimum value can be estimated depending on the conditions of use, it is advisable to set it to an optimum value within the range of the conditions of use.

[0075] TIFF2025118327000020.tif91170

[0076] By doing so, it is possible to estimate the slope with a small estimation error.

[0077] (Third embodiment) Another embodiment of the present invention will be described, which relates to a MEMS integrated circuit in which a gyro for detecting true north and an accelerometer for diagonal axes are integrated on a single chip.

[0078] An embodiment of the present invention is shown in Figure 10. Only the locations and axial orientations of the gyroscope 50 and accelerometer 51 are shown. Of course, in reality, masses constituting the sensors, their control circuits, and physical quantity detection circuits are also implemented, but these are not relevant to the essence of the present invention and are therefore omitted here.

[0079] A single-axis gyro 50 and a single-axis accelerometer are integrated on a single chip in the MEMS integrated circuit 200. The axis of the gyro 50 is horizontal in the drawing, and the axis of the accelerometer 51 is tilted by π / 2-μ [rad] relative to the direction of the axis of the gyro 50. In other words, the gyro 50 and the accelerometer 51 are laid out so that the axis of the accelerometer 51 is at an angle of μ with respect to the direction perpendicular to the axis of the gyro 50 (shown by a horizontal dashed line in the drawing).

[0080] Even if a gyro and accelerometer are used on separate modules or chips and are mounted at the appropriate angle, mounting errors occur. By integrating them on a single chip as in this embodiment, the μ angle can be accurately determined by the layout during IC design, and the device can be manufactured almost exactly as designed. Of course, deviations will never be completely eliminated, but the error can be made much smaller than when separate gyros and accelerometers are mounted to have the same μ angle. Because μ angle errors strongly affect the magnitude of the components that follow rotation and those that do not, which are important when estimating tilt, it is extremely important to be able to mount them at the intended μ. This embodiment makes it possible to accurately achieve the designed angle.

[0081] (Fourth embodiment) The next embodiment relates to a package mounting form of the MEMS integrated circuit in FIG. 10. FIG. 11 shows an embodiment of a sensor module 201 in which a MEMS integrated circuit 200 is mounted in a package 54. The single-chip integrated IC is ultimately mounted so that the axis of the accelerometer forms an angle μ with respect to the rotation axis of the rotating table. Therefore, it is necessary to mount it in a package that is easy to mount on the rotating table. In FIG. 11, the package 54 is mounted perpendicular to the mounting substrate 53, and the MEMS integrated circuit 200 is mounted inside it. The MEMS integrated circuit 200 is mounted perpendicular to the mounting substrate 53. In addition, it is mounted so that the axis of the gyro 50 is parallel to the surface of the mounting substrate.

[0082] The package 54 may also contain drive circuits and data output circuits for the gyro and accelerometer, as well as a circuit for performing true north detection processing, a battery, etc., but these are not shown.

[0083] After the MEMS integrated circuit 200 is mounted on the package 54, it is sealed with a package lid 52. At this time, the inside is evacuated or filled with nitrogen or an inert gas, etc., as necessary.

[0084] In the figure, there are screw holes 59 at the four corners of the mounting board 53, which allow screws to pass through and be attached to the screw holes in the rotating table. The underside of the mounting board 53, especially the underside of the part where the screw holes are located, is manufactured with a high degree of parallelism.

[0085] The mounting board 53 also has a positioning hole 60. The hole is nearly a perfect circle, and its inner surface is machined to have minimal irregularities. The rotating table also has a similar hole, and the orientation of the rotating table and the sensor module is aligned using a pin or similar that fits the hole.

[0086] However, the method of attachment to the rotating table is not limited to this, and the rotating table may have a clamp-like device and be fixed to the rotating table with the clamp.

[0087] By doing so, it becomes possible to mount an accelerometer having an oblique axis with high precision.

[0088] (Fifth embodiment) Next, an embodiment of a true north detection device 202 in which a sensor module 201 is attached to a rotary table 2 will be described with reference to FIG.

[0089] The sensor module 201 is attached to the rotating table 2 with screws 55. The rotating table surface, and at least the multiple parts (four in the figure) where the screws are attached, are made with a high degree of parallelism, so that the angle of the axes of the gyro 50 and accelerometer 51 is small with respect to the set value. Also, as mentioned above, the sensor module is attached to the rotating table in the correct orientation with the positioning pins 61.

[0090] The turntable 2 has a rotation mechanism (not shown). The rotation mechanism causes the turntable 2 to rotate at a constant speed or to move to a specified angle and then stop. When measuring while the turntable is stationary, the turntable is rotated so that the angle at which it stops has a sufficiently small error compared to the set angle, or the error is known.

[0091] The true north angle is detected by the above-mentioned method from the outputs of the gyro 50 and accelerometer 51 at each rotation angle.

[0092] By doing this, it becomes possible to detect true north with high accuracy using a MEMS sensor with a small number of axes.

[0093] The shape of the rotating table or rotating object does not necessarily have to be circular.

[0094] (Sixth embodiment) Figure 13 shows another embodiment of the true north detection device, an example that does not use a rotating table. The true north detection device 203 has a rotation device 56 with a built-in rotation mechanism, and a rotation plate 57 is attached to the rotation device 56. The rotation plate 57 serves as a module mounting surface that is sufficiently parallel to the rotation axis (not shown) of the rotation device 56.

[0095] The sensor module 204 has a MEMS integrated circuit 200 mounted inside a package 58, similar to that shown in FIG. 11, but the package 58 does not have a mounting substrate and is simply attached to a rotating plate 57 with screws 55. At least the surfaces of the package 58 and the rotating plate 57 that come into contact via the screws are made sufficiently parallel. Positioning holes or pins may be provided to ensure that the axis of the gyro is perpendicular to the axis of rotation. These are mounted on the surface where the rotating plate 57 and package 58 come into contact and are not shown.

[0096] The procedure for detecting true north is as described above.

[0097] By doing this, it becomes possible to detect true north with high accuracy using a MEMS sensor with a small number of axes.

[0098] The present invention is not limited to the above-described embodiments, and the components can be modified and embodied in practice without departing from the spirit of the invention. Furthermore, various inventions can be created by appropriately combining multiple components disclosed in the above-described embodiments. For example, some components may be omitted from all the components shown in the embodiments. Furthermore, components from different embodiments may be appropriately combined.

[0099] The embodiment of the present invention can also be configured as follows. [Item 1] a mechanism for rotating a rotation system rotatable around a rotation axis, the mechanism including an accelerometer having an axis that is not parallel to either a plane perpendicular to the rotation axis or the rotation axis; a processing unit that acquires accelerations measured by the accelerometers corresponding to rotation angles of three or more points of the rotated rotation system, separates the accelerations into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, and estimates an inclination angle of the rotation system based on the first component and the second component; a processing unit that acquires accelerations measured by the accelerometers corresponding to rotation angles of three or more points of the rotated rotation system, separates the accelerations into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, and estimates an inclination angle of the rotation system based on the first component and the second component; A sensor device comprising: [Item 2] The first component is a component parallel to the plane perpendicular to the rotation axis, and the second component is a component perpendicular to the plane. Item 1. The sensor device according to item 1. [Item 3] a gyroscope having an axis parallel to the plane of the rotational system is provided in the rotational system; the processing unit acquires angular velocities measured by the gyroscope corresponding to rotation angles of three or more points of the rotated rotation system, and estimates a direction of a specific orientation based on the angular velocities and the estimated tilt angle of the rotation system. Item 3. The sensor device according to item 1 or 2. [Item 4] The specific direction is due north or due south. Item 3. The sensor device according to item 3. [Item 5] TIFF2025118327000021.tif55170[Item 6] the first value is 61 degrees, the second value is 7 degrees; Item 6. The sensor device according to item 5. [Item 7] Item 3 is an integrated circuit provided in the sensor device, in which the gyroscope and the accelerometer are integrated into a single chip. [Item 8] A package mounting the integrated circuit according to item 7; a mechanism capable of attaching the package to the rotation system; A sensor module comprising: [Item 9] a rotation system rotatable around a rotation axis, the rotation system including an accelerometer having an axis that is not parallel to either the surface or the rotation axis; Acquire accelerations measured by the accelerometers corresponding to rotation angles of three or more points of the rotated rotation system; separating the acceleration into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, and estimating the tilt angle of the rotation system based on the first component and the second component; Slope estimation method. [Item 10] A gyroscope is provided in a rotation system rotatable around a rotation axis, and has an axis parallel to a plane perpendicular to the rotation axis. An accelerometer is provided in the rotation system, and has an axis that is not parallel to either the plane or the rotation axis. Acquiring angular velocities and speeds measured by the gyroscope and the accelerometer corresponding to rotation angles of three or more points of the rotated rotation system; Separating the acceleration into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, estimating a tilt angle of the rotation system based on the first component and the second component; estimating a direction of a specific orientation based on the angular velocity and the estimated tilt angle of the rotation system; Orientation detection method. [Explanation of symbols]

[0100] 1 accelerometer 2 turntables 3 Accelerometer axis direction 4 Sensor Unit 5 Gyro axis direction 50 Gyro 51 Accelerometer 52 Package lid 53 Circuit Board 54 packages 55 screws 56 Rotating Device 57 Rotating Plate 58 packages 59 screw holes 60 holes 61 pin 71 Control Unit 72 Acceleration acquisition and preprocessing section 73 Inclination angle calculation section 74 Acceleration / Angular Velocity Acquisition / Pre-processing Section 75 North angle calculation section 76 Display section 100 Sensor Device 101 Sensor Device 200 MEMS Integrated Circuits 201 Sensor Module 202 True North Detection Device 203 True North Detection Device 204 Sensor Module

Claims

1. a mechanism for rotating a rotation system rotatable around a rotation axis, the mechanism including an accelerometer having an axis that is not parallel to either a plane perpendicular to the rotation axis or the rotation axis; a processing unit that acquires accelerations measured by the accelerometers corresponding to rotation angles of three or more points of the rotated rotation system, separates the accelerations into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, and estimates an inclination angle of the rotation system based on the first component and the second component; A sensor device comprising:

2. The first component is a component parallel to the plane perpendicular to the rotation axis, and the second component is a component perpendicular to the plane. The sensor device of claim 1 .

3. a gyroscope having an axis parallel to the plane of the rotational system is provided in the rotational system; the processing unit acquires angular velocities measured by the gyroscope corresponding to rotation angles of three or more points of the rotated rotation system, and estimates a direction of a specific orientation based on the angular velocities and the estimated tilt angle of the rotation system. The sensor device of claim 1 .

4. The specific direction is due north or due south. The sensor device according to claim 3 .

5.

6. the first value is 61 degrees, the second value is 7 degrees; The sensor device according to claim 5 .

7. 8. An integrated circuit in which the gyroscope and the accelerometer are integrated into a single chip, the integrated circuit being included in the sensor device according to claim 7.

8. The method according to claim 3 a package in which the integrated circuit is mounted; a mechanism capable of attaching the package to the rotation system; A sensor module comprising:

9. a rotation system rotatable around a rotation axis, the rotation system including an accelerometer having an axis that is not parallel to either the surface or the rotation axis; Acquire accelerations measured by the accelerometers corresponding to rotation angles of three or more points of the rotated rotation system; separating the acceleration into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation of the rotation system, and estimating an inclination angle of the rotation system based on the first component and the second component; Slope estimation method.

10. A gyroscope is provided in a rotation system rotatable around a rotation axis, and has an axis parallel to a plane perpendicular to the rotation axis. An accelerometer is provided in the rotation system, and has an axis that is not parallel to either the plane or the rotation axis. acquiring angular velocities and speeds measured by the gyroscope and the accelerometer corresponding to rotation angles of three or more points of the rotated rotation system; Separating the acceleration into a first component that follows the rotation of the rotation system and a second component that does not follow the rotation, estimating a tilt angle of the rotation system based on the first component and the second component; estimating a direction of a specific orientation based on the angular velocity and the estimated tilt angle of the rotation system; Orientation detection method.

Images