JP2021067494A - Gravity direction displacement calculation device - Google Patents

Abstract

To increase the number of acceleration sensors while preventing damage to and robbery of devices and suppressing investment for device maintenance, thereby increasing data counts and calculating the displacement or speed in a gravity direction with high accuracy.SOLUTION: The present disclosure is a gravity direction displacement calculation device G comprising: an acceleration acquisition unit 1 for acquiring information pertaining to acceleration in three axes from an acceleration sensor of a portable terminal M; an average calculation unit 2 for calculating the time average of acceleration in each axis direction; a direction conversion unit 3 for converting acceleration in each axis direction into acceleration in gravity and horizontal directions so as to set the composite vector direction of the time average of acceleration in each axis direction to the direction of gravity,; and a displacement calculation unit 4 for calculating displacement or speed in the gravity direction on the basis of acceleration in the gravity direction.SELECTED DRAWING: Figure 2

Description

The present disclosure relates to a technique for calculating displacement or velocity in the direction of gravity using an acceleration sensor.

The wave height can be calculated by using the buoy's accelerometer (see Patent Document 1). In general, accelerometers can be used to calculate displacement or velocity in the direction of gravity.

Patent No. 4892972

However, buoys are vulnerable to damage and theft and require investment in maintenance. Since it is difficult to increase the number of buoy accelerometers, it is difficult to increase the number of data and calculate the wave height with high accuracy. In general, if it is difficult to increase the number of acceleration sensors, it is difficult to increase the number of data and calculate the displacement or velocity in the direction of gravity with high accuracy.

Therefore, in order to solve the above-mentioned problems, the present disclosure increases the number of acceleration sensors by increasing the number of acceleration sensors while making the device less susceptible to damage and theft and suppressing the investment for maintenance of the device, thereby increasing the number of data and gravitational force. The purpose is to calculate the displacement or velocity in the direction with high accuracy.

In order to solve the above problem, the displacement or velocity in the direction of gravity is calculated by using the acceleration sensor of the mobile terminal. However, the accelerometer of the mobile terminal can be arranged independently of the direction of gravity and can receive not only the acceleration in the direction of gravity but also the acceleration of horizontal vibration. Therefore, the time average of the acceleration in each axial direction is calculated, the acceleration in the gravity direction is extracted, and the acceleration in the horizontal vibration is removed. Then, the composite vector direction of the time average of the acceleration in each axial direction is set in the gravity direction, and the instantaneous value of the acceleration in each axial direction is coordinate-converted into the instantaneous value of the acceleration in the gravity direction and the horizontal direction.

Specifically, in the present disclosure, an acceleration acquisition unit that acquires acceleration information in three axial directions from an acceleration sensor of a mobile terminal, an average calculation unit that calculates the time average of acceleration in each axial direction, and an average calculation unit in each axial direction. Synthesis of time average of acceleration The direction change unit that converts the acceleration in each axial direction into the acceleration in the gravity direction and the horizontal direction so as to set the vector direction in the gravity direction, and the gravity based on the acceleration in the gravity direction. It is a gravity direction displacement calculation device characterized by comprising a displacement calculation unit for calculating a directional displacement or a velocity.

According to this configuration, the mobile terminal is less susceptible to damage and theft than a dedicated device such as a buoy, maintenance investment is suppressed, and the number of data is increased by increasing the number of acceleration sensors of the mobile terminal. Therefore, the displacement or velocity in the direction of gravity can be calculated with high accuracy.

Then, as the acceleration data in the gravity direction, the acceleration data in the triaxial direction can be reduced to the acceleration data in the uniaxial direction. Furthermore, as the detection process of the gravity direction, instead of setting the composite vector direction of the acceleration in each axial direction one by one in the gravity direction at each time, the composite vector direction of the time average of the acceleration in each axial direction is once set in the gravity direction. Can be set to.

Further, in the present disclosure, the average calculation unit describes the time of acceleration in each axial direction over a period longer than that of the most dominant or longest fluctuation cycle of acceleration in each axial direction. It is a gravity direction displacement calculation device characterized by calculating an average.

According to this configuration, the acceleration of horizontal vibration can be removed with higher accuracy.

Further, in the present disclosure, the direction changing unit normalizes the acceleration in each axial direction by the gravitational acceleration so that the composite vector length of the time average of the acceleration in each axial direction is normalized by the gravitational acceleration. It is a characteristic gravity direction displacement calculation device.

According to this configuration, the displacement or velocity in the direction of gravity can be calculated with higher accuracy.

From the measurement result of the acceleration in the gravity direction, the gravity component which is a DC component and the DC drift component of the acceleration sensor are removed, and the displacement in the gravity direction or the vibration component of the velocity is extracted to the measurement limit on the low frequency side. Here, the low frequency removal processing in the time domain, the time from the time series data polynomial approximation result (e.g., gravitational acceleration × time for the time-series data in the gravity direction of displacement ^2/2.) Subtracts. On the other hand, in the low frequency removal process in the frequency domain, low frequency components (for example, around 0 Hz of the gravity component and 0 Hz of the DC drift component of the accelerometer) are removed from the frequency spectrum. Unlike the high-pass filter, the low-frequency removal process in the time domain and the frequency domain does not show a transient response, so that the vibration component of the displacement or velocity in the gravity direction can be extracted in a short time.

Specifically, the present disclosure presents the acceleration, velocity in the gravity direction by subtracting the polynomial approximation result of the acceleration, velocity or displacement in the gravity direction from the time series data of the acceleration, velocity or displacement in the gravity direction. Alternatively, the gravity direction displacement calculation device is further provided with a low frequency removing portion in a time region for removing the gravity component from the time-series data of the displacement, or removing the gravity component and the DC drift component of the acceleration sensor. Is.

According to this configuration, even if the measurement result of the acceleration in the gravity direction includes the gravity component which is a DC component and the DC drift component of the acceleration sensor, the displacement or velocity in the gravity direction is not used and the low frequency removal process showing the transient response is not used. The vibration component can be extracted in a short time up to the measurement limit on the low frequency side.

In the low frequency removal process in the time domain, the calculation amount cannot be reduced as compared with the low frequency removal process in the frequency domain, but the accuracy can be improved.

Further, the present disclosure discloses that the frequency spectrum of acceleration, velocity or displacement in the direction of gravity suppresses the DC drift component of the acceleration sensor, or suppresses the DC drift component and gravity component of the acceleration sensor. , The DC drift component of the acceleration sensor is removed from the frequency spectrum of the acceleration, velocity or displacement in the gravity direction, or the DC drift component of the acceleration sensor and the low frequency removing portion of the frequency region for removing the gravity component are removed. It is a displacement calculation device in the direction of gravity, which is further provided.

According to this configuration, even if the measurement result of the acceleration in the gravity direction includes the gravity component which is a DC component and the DC drift component of the acceleration sensor, the displacement or velocity in the gravity direction is not used and the low frequency removal process showing the transient response is not used. The vibration component can be extracted in a short time up to the measurement limit on the low frequency side.

In the low frequency removal process in the frequency domain, the accuracy cannot be improved as compared with the low frequency removal process in the time domain, but the amount of calculation can be reduced.

Further, the present disclosure is further provided with a low frequency measuring unit for measuring the DC drift frequency of the acceleration sensor based on the frequency spectrum of the acceleration, velocity or displacement in an arbitrary direction when the acceleration sensor is stationary. It is a displacement calculation device in the direction of gravity.

According to this configuration, it is possible to measure the DC drift frequency of the acceleration sensor by observing the stationary state of the acceleration sensor prior to the measurement process of the acceleration in the gravity direction.

Further, the present disclosure is characterized in that the low frequency measuring unit removes a gravity component from time series data of acceleration, velocity or displacement in the arbitrary direction when the acceleration sensor is stationary before frequency conversion. It is a directional displacement calculation device.

According to this configuration, the DC drift frequency of the acceleration sensor can be measured even when the DC drift component of the acceleration sensor is buried in the gravity component which is the DC component.

Further, the present disclosure is characterized in that the displacement calculation unit calculates the displacement or velocity in the gravity direction by integrating the time-series data of the acceleration in the gravity direction only twice or once in time. It is a displacement calculation device in the direction of gravity.

According to this configuration, the acceleration integration process in the time domain cannot reduce the amount of calculation as compared with the acceleration integration process in the frequency domain, but the accuracy can be improved.

Further, in the present disclosure, the displacement calculation unit divides the frequency spectrum of the acceleration in the direction of gravity by the square or the first power of 2πf (f is a frequency) or j2πf (j is an imaginary unit). It is a gravity direction displacement calculation device characterized by calculating a displacement or a velocity in the gravity direction.

According to this configuration, the acceleration integration process in the frequency domain cannot improve the accuracy as compared with the acceleration integration process in the time domain, but the amount of calculation can be reduced. Then, the body movement component or the like, which is a high frequency component, can be suppressed by dividing by 2πf or j2πf. By calculating 1 / (2πf) or 1 / (j2πf) in advance, division of 2πf or j2πf squared or 1st, which requires more calculation, can be performed with less calculation amount 1 / (2πf). Alternatively, it can be replaced with multiplication of 1 / (j2πf) squared or squared.

Further, the present disclosure discloses the time series of acceleration, velocity or displacement in the direction of gravity in converting the time series data of acceleration, velocity or displacement in the direction of gravity into the frequency spectrum of acceleration, velocity or displacement in the direction of gravity. It is a displacement calculation device in the direction of gravity, which further includes a frequency conversion unit that multiplies the data by a window function having a flat center.

According to this configuration, when the frequency spectrum is inversely transformed into the time series data, the suppression of the amplitude of the time series data can be reduced in the flat period in the center of the window function.

Further, in the present disclosure, when the frequency conversion unit converts the time-series data of the acceleration, velocity or displacement in the gravity direction into the frequency spectrum of the acceleration, velocity or displacement in the gravity direction, the acceleration in the gravity direction, It is a gravity direction displacement calculation device characterized by multiplying the time series data of velocity or displacement by a half overlap Han window function.

According to this configuration, in the half-overlap Han window function, the central period can be similarly flattened and the side rope level can be further reduced as compared with the rectangular window function.

Further, in the present disclosure, the displacement calculation unit calculates the wave height at the position where the ship exists based on the acceleration in the direction of gravity measured by using the acceleration sensor of the mobile terminal owned by the sailor. It is a gravity direction displacement calculation device characterized by.

According to this configuration, the wave height can be calculated with high accuracy by using a mobile terminal.

As described above, the present disclosure increases the number of data by increasing the number of accelerometers while reducing the damage and theft of the device and suppressing the investment in the maintenance of the device, thereby increasing the displacement or velocity in the direction of gravity. It can be calculated with high accuracy.

It is a figure which shows the wave height calculation method using the mobile terminal of this disclosure. It is a figure which shows the structure of the gravity direction displacement calculation apparatus of this disclosure. It is a figure which shows the procedure of the gravitational direction displacement calculation processing of this disclosure. It is a figure which shows the procedure of the acceleration direction conversion processing of this disclosure. It is a figure which shows the concept of the acceleration direction conversion processing of this disclosure. It is a figure which shows the concept of the acceleration direction conversion processing of this disclosure. It is a figure which shows the specific example of the acceleration direction conversion process of this disclosure. It is a figure which shows the specific example of the acceleration direction conversion process of this disclosure. It is a figure which shows the procedure of the 1st displacement calculation processing of this disclosure. It is a figure which shows the specific example of the 1st displacement calculation process of this disclosure. It is a figure which shows the specific example of the 1st displacement calculation process of this disclosure. It is a figure which shows the procedure of the 2nd displacement calculation processing of this disclosure. It is a figure which shows the specific example of the 2nd displacement calculation process of this disclosure. It is a figure which shows the specific example of the 2nd displacement calculation process of this disclosure. It is a figure which shows the procedure of the DC drift frequency measurement processing of this disclosure. It is a figure which shows the specific example of the DC drift frequency measurement process of this disclosure. It is a figure which shows the specific example of the DC drift frequency measurement process of this disclosure. It is a figure which shows the structure of the half overlap Han window function of this disclosure. It is a figure which shows the effect of the half overlap Han window function of this disclosure.

Embodiments of the present disclosure will be described with reference to the accompanying drawings. The embodiments described below are examples of the embodiments of the present disclosure, and the present disclosure is not limited to the following embodiments.

(Structure of gravity direction displacement calculation device)
FIG. 1 shows a wave height calculation method using the mobile terminal of the present disclosure. The configuration of the gravity direction displacement calculation device of the present disclosure is shown in FIG. The procedure of the gravity direction displacement calculation processing of the present disclosure is shown in FIG.

The gravity direction displacement calculation device G includes an acceleration acquisition unit 1, an average calculation unit 2, a direction conversion unit 3, a displacement calculation unit 4, a low frequency removal unit 5, a frequency conversion unit 6, a frequency inverse conversion unit 7, and a displacement calculation unit 8. This can be realized by installing the program shown in FIGS. 3, 4, 9, 12, and 15 in a computer, which includes a gravity removing unit 9, a frequency conversion unit 10, and a low frequency measuring unit 11.

The gravity direction displacement calculation device G calculates the wave height at the position where the ship S exists based on the acceleration in the gravity direction measured by using the acceleration sensor of the mobile terminal M owned by the sailor. In the upper part of FIG. 1, the gravity direction displacement calculation device G is not mounted on the mobile terminal M and wirelessly communicates with the mobile terminal M. In the lower part of FIG. 1, the gravity direction displacement calculation device G is mounted on the mobile terminal M.

If the acceleration sensor of the mobile terminal M can receive not only the acceleration in the direction of gravity but also the acceleration of horizontal vibration, the present disclosure can be applied, for example, in the application of calculating the wave height.

The acceleration acquisition unit 1 acquires information on acceleration in the three axial directions from the acceleration sensor of the mobile terminal M (step S1). The direction changing unit 3 converts the acceleration in each axial direction into the acceleration in the gravitational direction and the horizontal direction (step S2). Details of steps S1 and S2 will be described with reference to FIGS. 4 to 8.

The displacement calculation unit 4 calculates the displacement or velocity in the gravity direction based on the acceleration in the gravity direction (step S3). The low frequency removing unit 5 removes the gravity component and the DC drift component of the acceleration sensor (step S3). Details of step S3 will be described with reference to FIGS. 9 to 19.

(Procedure of acceleration direction conversion process)
The procedure of the acceleration direction conversion process of the present disclosure is shown in FIG. The concept of the acceleration direction conversion process of the present disclosure is shown in FIGS. Specific examples of the acceleration direction conversion process of the present disclosure are shown in FIGS. 7 and 8.

In the acceleration direction conversion process, the displacement or velocity in the direction of gravity is calculated by using the acceleration sensor of the mobile terminal M. However, since the acceleration sensor of the mobile terminal M is put in a sailor's pocket or the like, it can be arranged regardless of the direction of gravity, and since it is separated from the center of gravity C of the ship S by a distance R, if only the acceleration in the direction of gravity is used. However, it can receive the acceleration of horizontal vibration.

As shown in the upper part of FIG. 5, the horizontal direction is the X and Y axis directions, the gravity direction is the Z axis direction, the in-plane direction of the mobile terminal M is the x and y axis directions, and the other directions of the mobile terminal M. Is the z-axis direction. As shown in the upper part of FIG. 5 and the upper middle part of FIG. 6, the Z-axis and the z-axis can be inconsistent.

In Patent Document 1, _{the combined vector direction of the instantaneous values a x} , a _y , and a _z of the acceleration in each axial direction is set in the direction of gravity. However, the instantaneous value _a x of the acceleration in each axis _direction, a y, a composite vector length of ^{_{a z √ (a x 2 +}} a y 2 + a z 2) is greater than the gravitational acceleration g, not only the acceleration in the direction of gravity Includes the acceleration of horizontal vibration. Therefore, the instantaneous value a _x of the acceleration in each axis _direction, a _y, be set combined vector direction of a _z direction of gravity, the instantaneous value a _x of the acceleration in each axis _direction, a _y, the direction of gravity a _z And the coordinates cannot be correctly converted to the instantaneous value of the acceleration in the horizontal direction.

In the present disclosure, the combined vector direction of the time averages of accelerations in each axial direction a _{x −} bar, a _{y −} bar, and a _{z −} bar is set in the direction of gravity. Then, time average _a x -bar acceleration in each axis _direction, a y _-bar, composite vector length of ^{_{a z -bar √ (a x -bar}} 2 + a y -bar 2 + a z -bar 2) is gravity It is equal to the acceleration g and includes only the acceleration in the direction of gravity and does not include the acceleration of horizontal vibration. Therefore, if the combined vector direction of the time average a _x- bar, a _y- bar, and a _z- bar of the acceleration in each axial direction is set in the gravity direction, the instantaneous values a _x , a _y , of the acceleration in each axial direction are set. The _{coordinates of az} can be correctly converted into instantaneous values of acceleration in the gravity direction and the horizontal direction.

As shown in the lower middle part of FIG. 6, the x-axis, y-axis and z-axis are coordinate-converted to the x'-axis, y'-axis and z'-axis. As shown in the upper and lower rows of FIG. 6, the Z axis and the z'axis coincide with each other.

_{The acceleration acquisition unit 1 acquires information on accelerations a x} , a _y , and a _{z in} the three axial directions from the acceleration sensor of the mobile terminal M (step S11). In the upper part of FIG. 7, the y-axis and the Z-axis are substantially the same, and the xz plane and the XY plane are substantially the same. _{The time average a y-} bar of the acceleration in the y-axis direction is significantly different from the original 0 [G], and the time average a _x- bar of the acceleration in the x-axis direction is slightly different from the original 0 [G]. _{The time average az-} bar of acceleration in the z-axis direction is significantly different from the original 1 [G]. The time average is calculated over several tens of seconds.

The average calculation unit 2 performs acceleration in each axial direction over a period longer than the most dominant or longest fluctuation cycle among the fluctuation periods _{of acceleration a x} , a _y , and _{az in each axial direction.} The time averages of a _{x −} bar, a _{y −} bar, and a _{z −} bar are calculated (step S12). Here, the most dominant fluctuation period means a fluctuation period having the largest amplitude aside from the period. The longest fluctuation period means the fluctuation period having the longest period aside from the amplitude. In other words, we want to extract the acceleration in the direction of gravity with high accuracy while removing the acceleration of horizontal vibration with high accuracy.

Before normalization in the lower part of FIG. 7, the time average a _x- bar of the acceleration in the x-axis direction is 0.015 [G], and the time average a _y- bar of the acceleration in the y-axis direction is 0.998. It is [G], and the time average _az −bar of the acceleration in the z-axis direction is −0.181 [G]. Then, time average _a x -bar acceleration in each axis _direction, a y _-bar, composite vector length of ^{_{a z -bar √ (a x -bar}} 2 + a y -bar 2 + a z -bar 2) is 1 It is .015 [G].

Direction changing unit 3, time average _a x -bar acceleration in each axis _direction, a y _-bar, composite vector length of ^{_{a z -bar √ (a x -bar}} 2 + a y -bar 2 + a z -bar 2 ) Is normalized by the gravitational acceleration g, and the accelerations a _x , a _y , and a _z in each axial direction are normalized by the gravitational acceleration g (step S13). In other words, we want to calculate the acceleration in the direction of gravity with high accuracy.

After normalization in the lower part of FIG. 7, the time average a _x- bar of the acceleration in the x-axis direction is 0.015 [G], and the time average a _y- bar of the acceleration in the y-axis direction is 0.984. It is [G], and the time average _az −bar of the acceleration in the z-axis direction is −0.178 [G]. Then, time average _a x -bar acceleration in each axis _direction, a y _-bar, composite vector length of ^{_{a z -bar √ (a x -bar}} 2 + a y -bar 2 + a z -bar 2) is 1 It is .000 [G].

The direction changing unit 3 sets the combined vector direction of the time averages of accelerations in each axial direction a _{x −} bar, a _{y −} bar, and a _{z −} bar in the direction of gravity, so that the acceleration a _x , a in each axial direction is set. _{Convert y} and _az into accelerations in the gravity direction and the horizontal direction (step S14).

In the upper part of FIG. 8, when the x-axis, y-axis and z-axis are coordinate-converted to the x'axis, y'-axis and z'-axis, the roll rotation angle r (x-axis is the rotation axis) is represented by Equation 1. The pitch rotation angle p (the y-axis is the rotation axis) is represented by the equation 2, and the acceleration az'in the _z'axis direction after the roll rotation and the pitch rotation is represented by the equation 3. Incidentally, x after the roll rotation pitch rotation ', y' axis direction of the acceleration a _{x ',} a y' does not need to be particularly calculated for the purpose of measuring the acceleration in the gravity direction.

In the lower part of FIG. 8, the z'axis and the Z axis substantially coincide with each other, and the x'y'plane and the XY plane substantially coincide with each other. Then, -bar 'time average _{a z} of acceleration in the axial direction' z is the original 1 [G] and approximately equal, x -bar 'time average _{a x} of the axial acceleration' is the original 0 [G] When substantially equal, -bar 'time average _{a y} acceleration of the axial' y is approximately equal to the original 0 [G]. The time average is calculated over several tens of seconds.

In this way, the mobile terminal M is less susceptible to damage and theft than a dedicated device such as a buoy, maintenance investment is suppressed, and the number of data is increased by increasing the number of acceleration sensors of the mobile terminal M. Therefore, the displacement or velocity in the direction of gravity can be calculated with high accuracy.

Then, as the acceleration data in the gravity direction, the acceleration data in the triaxial direction can be reduced to the acceleration data in the uniaxial direction. Furthermore, as the detection process of the gravity direction, instead of setting the composite vector direction of the acceleration in each axial direction one by one in the gravity direction at each time, the composite vector direction of the time average of the acceleration in each axial direction is once set in the gravity direction. Can be set to.

Further, by calculating the time average of the acceleration in each axial direction, the acceleration of the horizontal vibration can be removed with higher accuracy. Further, by normalizing the composite vector length of the time average of the acceleration in each axial direction, the displacement or velocity in the gravitational direction can be calculated with higher accuracy.

(Procedure of the first displacement calculation process)
The procedure of the first displacement calculation process of the present disclosure is shown in FIG. Specific examples of the first displacement calculation process of the present disclosure are shown in FIGS. 10 and 11. The first displacement calculation process is just an example.

Then, the gravity component which is a DC component and the DC drift component of the acceleration sensor are removed from the measurement result of the acceleration in the gravity direction, and the displacement in the gravity direction or the vibration component of the velocity is extracted to the measurement limit on the low frequency side. Here, as the low-frequency removal processing in the time domain, the time from the time series data polynomial approximation result (e.g., gravitational acceleration × time ^2/2 with respect to time-series data in the gravity direction of the displacement.) Subtracts. Therefore, unlike the high-pass filter, the low-frequency removal process in the time domain does not show a transient response, so that the vibration component of the displacement or velocity in the gravity direction can be extracted in a short time.

The displacement calculation unit 4 calculates the time-series data of the displacement z in the gravity direction by integrating the time-series data _{of the acceleration az in the gravity direction only twice in time (step S21).} From the upper part of FIG. 10 to the middle part of FIG. 10, the process of integration is executed twice.

The low frequency removing unit 5 removes the gravity component from the time series data of the displacement z in the gravity direction by subtracting the polynomial approximation result of the displacement z in the gravity direction from the time series data of the displacement z in the gravity direction ( Step S22). Toward the lower part of FIG. 10 from the middle of FIG. 10, the polynomial approximation results in the gravity direction of displacement z as gt ^2/2, the processing of gravity removal is performed.

When converting the time-series data of the displacement z in the gravity direction into the frequency spectrum of the displacement z in the gravity direction, the frequency conversion unit 6 converts the time-series data of the displacement z in the gravity direction into a half-overlap han window function (FIG. 18). , 19, which will be described later), is multiplied (step S23). The window function multiplication process is executed from the lower part of FIG. 10 to the upper part of FIG.

The low frequency removing unit 5 removes the DC drift component of the acceleration sensor from the frequency spectrum of the displacement z in the gravity direction by suppressing the DC drift component of the acceleration sensor in the frequency spectrum of the displacement z in the gravity direction (step). S24). From the upper part of FIG. 11 to the middle part of FIG. 11, the frequency conversion process and the drift removal process are executed. Since the frequency threshold value of the DC drift component of the accelerometer (described later using FIGS. 15 to 17) is set to 0.05 Hz, the fluctuation of the displacement z in the gravity direction in which the fluctuation cycle is 20 seconds or less is measured. Can be done.

The frequency inverse transformation unit 7 inversely transforms the frequency spectrum of the displacement z in the gravity direction into time-series data of the displacement z in the gravity direction (step S25). The frequency inverse transformation process is executed from the middle stage of FIG. 11 to the lower stage of FIG. Since the half-overlap Han window function (described later with reference to FIGS. 18 and 19) is flat at time t = 20s-60s, the time-series data of the displacement z in the gravity direction has an amplitude at time t = 20s-60s. Has not decreased.

The first displacement calculation process is just an example. Displacement calculating unit 4, the low-frequency removal unit 5, a frequency converting unit 6 and the frequency inverse transform unit 7, if the time-series data in the gravity direction of displacement z or velocity v _z can be correctly calculated, what order You may execute the process with.

Low-frequency removing section 5, the direction of gravity acceleration a _z, from the time-series data of the velocity v _z or displacement z, the direction of gravity acceleration a _z, by subtracting the polynomial approximation results of velocity v _z or displacement z, gravity The gravity component may be removed from the time-series data of the acceleration _az , the velocity _vz or the displacement z in the direction, or the gravity component and the DC drift component of the acceleration sensor may be removed.

Displacement calculating unit 4, by integrating only twice or once in a time series data time direction of gravity acceleration a _z, may be calculated displacements z or velocity v _z direction of gravity. Frequency converting unit 6, the direction of gravity acceleration a _z, time-series data of the velocity v _z or displacement z, the direction of gravity acceleration a _z, Upon converting to the frequency spectrum of the velocity v _z or displacement z, the direction of gravity acceleration The half-overlap Han window function may be multiplied by the time-series data of _{a z} , velocity v _{z or displacement z.}

In this way, even if the measurement result of the acceleration in the gravity direction includes the gravity component which is a DC component and the DC drift component of the acceleration sensor, the displacement or velocity vibration in the gravity direction is not used without using the low frequency removal process which shows a transient response. The components can be extracted in a short time up to the measurement limit on the low frequency side.

In the low frequency removal process in the time domain, the calculation amount cannot be reduced as compared with the low frequency removal process in the frequency domain, but the accuracy can be improved. Further, the acceleration integration process in the time domain cannot reduce the amount of calculation as compared with the acceleration integration process in the frequency domain, but the accuracy can be improved.

(Procedure for second displacement calculation process)
The procedure of the second displacement calculation process of the present disclosure is shown in FIG. Specific examples of the second displacement calculation process of the present disclosure are shown in FIGS. 13 and 14. The second displacement calculation process is just an example.

Then, the gravity component which is a DC component and the DC drift component of the acceleration sensor are removed from the measurement result of the acceleration in the gravity direction, and the displacement in the gravity direction or the vibration component of the velocity is extracted to the measurement limit on the low frequency side. Here, as a low frequency removal process in the frequency domain, a low frequency component (for example, near 0 Hz of the gravity component or 0 Hz of the DC drift component of the acceleration sensor) is removed from the frequency spectrum. Therefore, unlike the high-pass filter, the low-frequency removal process in the frequency domain does not show a transient response, so that the vibration component of the displacement or velocity in the gravity direction can be extracted in a short time.

Frequency converting unit 6, a time series data of acceleration a _z direction of gravity, when converted into a frequency spectrum of the acceleration a _z direction of gravity, the time-series data of the direction of gravity acceleration a _z, half overlap Han window function (See below with reference to FIGS. 18 and 19) is multiplied (step S31). The window function multiplication process is executed from the upper part of FIG. 13 to the lower part of FIG.

The low frequency removing unit 5 suppresses the DC drift component and the gravity component of the acceleration sensor in the frequency spectrum of the acceleration _{az in} the gravity direction, and thereby suppresses the DC drift component of the acceleration sensor and the DC drift component of the acceleration sensor from the frequency spectrum of the acceleration _{az in the gravity direction.} And the gravity component is removed (step S32). From the lower part of FIG. 13 to the upper part of FIG. 14, a frequency conversion process, a drift removal process, and a gravity removal process (including a double integration process described later) are executed. Since the frequency threshold value of the DC drift component of the accelerometer (described later using FIGS. 15 to 17) is set to 0.05 Hz, the fluctuation of the displacement z in the gravity direction in which the fluctuation cycle is 20 seconds or less is measured. Can be done.

The displacement calculation unit 4 _{divides the frequency spectrum of the acceleration az in} the gravity direction by the square of 2πf (f is the frequency) or j2πf (j is the imaginary unit) to obtain the frequency spectrum of the displacement z in the gravity direction. Calculate (step S33). When dividing by the square of j2πf, the sign of the amplitude of the frequency spectrum of the displacement z in the gravity direction is only inverted as compared with the case of dividing by the square of 2πf. From the lower part of FIG. 13 to the upper part of FIG. 14, the double integration process (including the above-mentioned frequency conversion process, drift removal process, and gravity removal process) is executed.

The frequency inverse transformation unit 7 inversely transforms the frequency spectrum of the displacement z in the gravity direction into time-series data of the displacement z in the gravity direction (step S34). From the upper part of FIG. 14 to the lower part of FIG. 14, the frequency inverse transformation process is executed. Since the half-overlap Han window function (described later with reference to FIGS. 18 and 19) is flat at time t = 20s-60s, the time-series data of the displacement z in the gravity direction has an amplitude at time t = 20s-60s. Has not decreased.

The second displacement calculation process is just an example. Displacement calculating unit 4, the low-frequency removal unit 5, a frequency converting unit 6 and the frequency inverse transform unit 7, if the time-series data in the gravity direction of displacement z or velocity v _z can be correctly calculated, what order You may execute the process with.

The low frequency removing unit 5 suppresses the DC drift component of the acceleration sensor or suppresses the DC drift component and the gravity component of the acceleration sensor in the frequency spectrum of the acceleration _az , the velocity _{vz or the displacement z in the direction of gravity.} By doing so, _{the DC drift component of the acceleration sensor may be removed from the frequency spectrum of the acceleration az} , the velocity _vz or the displacement z in the direction of gravity, or the DC drift component and the gravity component of the acceleration sensor may be removed.

Displacement calculating unit 4, by dividing the frequency spectrum of the direction of gravity acceleration a _z in 2πf or square or 1 square of J2paif, may be calculated displacements z or velocity v _z direction of gravity. At the time of division by first power of J2paif, compared to the time division by first power of 2 [pi] f, the frequency spectrum of the gravity direction of the velocity v _z is the complex number rather than real. However, even when division by first power of J2paif, if you take the real part of the complex number, as in the case of division by first power of 2 [pi] f, the resulting frequency spectrum of the gravitational direction of the velocity v _z. Frequency converting unit 6, the direction of gravity acceleration a _z, time-series data of the velocity v _z or displacement z, the direction of gravity acceleration a _z, Upon converting to the frequency spectrum of the velocity v _z or displacement z, the direction of gravity acceleration The half-overlap Han window function may be multiplied by the time-series data of _{a z} , velocity v _{z or displacement z.}

In this way, even if the measurement result of the acceleration in the gravity direction includes the gravity component which is a DC component and the DC drift component of the acceleration sensor, the displacement or velocity vibration in the gravity direction is not used without using the low frequency removal process which shows a transient response. The components can be extracted in a short time up to the measurement limit on the low frequency side.

In the low frequency removal process in the frequency domain, the accuracy cannot be improved as compared with the low frequency removal process in the time domain, but the amount of calculation can be reduced. Further, the acceleration integration process in the frequency domain cannot improve the accuracy as compared with the acceleration integration process in the time domain, but the amount of calculation can be reduced. Further, the body movement component which is a high frequency component can be suppressed by dividing by 2πf or j2πf. By calculating 1 / (2πf) or 1 / (j2πf) in advance, division of 2πf or j2πf squared or 1st, which requires more calculation, can be performed with less calculation amount 1 / (2πf). Alternatively, it can be replaced with multiplication of 1 / (j2πf) squared or squared.

(Procedure for DC drift frequency measurement processing)
The procedure of the DC drift frequency measurement process of the present disclosure is shown in FIG. Specific examples of the DC drift frequency measurement process of the present disclosure are shown in FIGS. 16 and 17. This process is just an example.

The DC drift frequency of the accelerometer varies from individual to individual and changes over time. Therefore, the DC drift frequency of the accelerometer is measured by observing the stationary state of the accelerometer.

_{The acceleration acquisition unit 1 acquires information on accelerations a x} , a _y , and a _z in the three axial directions when the acceleration sensor is stationary (step S41). In the upper part of FIG. 16, since the mobile terminal M is stationary on a desk or the like on land, the acceleration a _z in the z axis direction is close to the gravitational acceleration g, x, the y-axis direction acceleration a _x, a _y is 0 Become closer to. However, if the mobile terminal M is stationary in a sailor's pocket or the like, the accelerations a _x , a _y , and a _z in each axial direction can each take a finite value.

Gravity removing section 9, an arbitrary direction of the acceleration during the acceleration sensor rest (in this case, the acceleration a _z in the z axis _direction) from the time-series data, removing the gravity component before frequency conversion (step S42). From the top of FIG. 16 toward the middle of Figure 16, from the time series data of acceleration _{a z} in the z axis direction, the time-averaged -9.86 [m / ^{s 2]} of the acceleration _{a z} of the z-axis direction is subtracted.

Frequency converter 10, (in this case, the acceleration a _z in the z axis _direction) any direction of the acceleration time series data, when converted into a frequency spectrum of an arbitrary direction of the acceleration, the time-series data of an arbitrary direction of the acceleration, Multiply any window function (here, a normal Han window function) (step S43). From the middle of FIG. 16 to the bottom of FIG. 16, the window function multiplication process is executed. The frequency conversion process is executed from the lower part of FIG. 16 to the upper part of FIG.

Displacement calculating unit 8 (in this case, the acceleration a _z in the z axis _direction) any direction of the acceleration by (the j of imaginary unit.) Or J2paif (frequency. The f) frequency spectrum 2πf of dividing the square of the , The frequency spectrum of the displacement in the arbitrary direction (here, the displacement z in the z-axis direction) is calculated (step S44). From the upper part of FIG. 17 to the lower part of FIG. 17, the process of integration is executed twice.

The low frequency measuring unit 11 measures the DC drift frequency of the accelerometer based on the frequency spectrum of the displacement of the accelerometer in an arbitrary direction (here, the displacement z in the z-axis direction) when the accelerometer is stationary (step S45). In the lower part of FIG. 17, since the frequency threshold of the DC drift component of the acceleration sensor is set to 0.025 Hz, the gravity has a fluctuation period of 40 seconds or less within a range of an error of 0.05 m in the displacement z in the gravity direction. The fluctuation of the displacement z in the direction can be measured.

Here, in the upper part of FIG. 17, (in this case, the acceleration a _z in the z axis _direction) any direction of the acceleration in the frequency spectrum of, for not running division of (2 [pi] f) ² or (J2paif) ^2, the high frequency component Is not sufficiently suppressed, and it is difficult to set the frequency threshold of the DC drift component of the accelerometer.

However, in the lower part of FIG. 17, in the frequency spectrum of the displacement in the arbitrary direction (here, the displacement z in the z-axis direction ^{), the division of (2πf) 2} or (j2πf) ² is executed, so that the high frequency component is sufficient. It is easy to set the frequency threshold of the DC drift component of the acceleration sensor.

The above processing is just an example. The displacement calculation unit 8, the gravity removal unit 9, the frequency conversion unit 10, and the low frequency measurement unit 11 execute the processing in any order if the frequency threshold value of the DC drift component of the acceleration sensor can be set correctly. You may.

The low frequency measuring unit 11 may measure the DC drift frequency of the accelerometer based on the frequency spectrum of the acceleration, velocity or displacement in an arbitrary direction when the accelerometer is stationary. The gravity removing unit 9 may remove the gravity component from the time-series data of the acceleration, velocity, or displacement in an arbitrary direction when the acceleration sensor is stationary before frequency conversion. The displacement calculation unit 8 is almost the same as the displacement calculation unit 4. The frequency conversion unit 10 is almost the same as the frequency conversion unit 6.

In this way, prior to the measurement process of the acceleration in the direction of gravity, the DC drift frequency of the acceleration sensor can be measured by observing the stationary state of the acceleration sensor. Then, even when the DC drift component of the acceleration sensor is buried in the gravity component which is the DC component, the DC drift frequency of the acceleration sensor can be measured by removing the gravity component before frequency conversion.

(Structure of half overlap Han window function)
The configuration of the half overlap Han window function of the present disclosure is shown in FIG. The effect of the half-overlap Han window function of the present disclosure is shown in FIG.

Frequency converting unit 6 converts the direction of gravity acceleration a _z, time-series data of the velocity v _z or displacement z, the direction of gravity acceleration a _z, the frequency spectrum of the velocity v _z or displacement z. At this time, the frequency conversion unit 6 _{multiplies the time-series data of the acceleration az} , the velocity _vz, or the displacement z in the direction of gravity by a window function having a flat center. Better yet, the frequency converter 6 _{multiplies the time-series data of acceleration az} , velocity _vz or displacement z in the direction of gravity by a half-overlap Han window function.

The upper part of FIG. 18 shows a half-overlap Han window function wn (t) in the time domain. wna (t) is a Han window function at a normalization time of -1.0 to 0.0 and is represented by the equation 4, and wnb (t) is a normalization time of -0.5 to 0.5. It is a Han window function, represented by the equation 5, wnc (t) is a Han window function with a normalization time of 0.0 to 1.0, is represented by the equation 6, and wn (t) is wna ( It is represented by the equation 7 as the sum of t), wnb (t) and wnc (t). wn (t) has a flat portion at a normalization time of −0.5 to 0.5.

The lower part of FIG. 18 shows the half-overlap Han window function Wnb (f) in the frequency domain. Wnb (f) is represented by Equation 8 as a Fourier transform of wn (t). The lower part of FIG. 18 also shows a normal Han window function in the frequency domain. A normal Han window function in the time domain is a Han window function with a normalization time of -1.0 to 1.0 and does not have a flat part with a normalization time of -0.5 to 0.5. .. In the half-overlap Han window function Wnb (f) in the frequency domain, the side rope level can be further reduced as compared with the normal Han window function in the frequency domain.

In FIG. 19, when the time-series data of the displacement z in the gravity direction is converted into the frequency spectrum of the displacement z in the gravity direction, the half-overlap Han window function and the normal Han window function are converted into the time-series data of the displacement z in the gravity direction. Or the rectangular window function is multiplied. Then, after removing the DC drift component and the gravity component of the acceleration sensor from the frequency spectrum of the displacement z in the gravity direction, the frequency spectrum of the displacement z in the gravity direction is inversely converted into the time-series data of the displacement z in the gravity direction.

When the half-overlap Han window function is applied to the time-series data of the displacement z in the gravity direction, the normalization time-compared to the case where the time-series data of the displacement z in the gravity direction is multiplied by the normal Han window function or the rectangular window function. In the time t = 20s to 60s corresponding to 0.5 to 0.5, the suppression of the amplitude of the time-series data of the displacement z in the gravity direction after the inverse frequency conversion can be reduced.

In this way, _{when the frequency spectrum is inversely converted into time-series data for acceleration az} , velocity _vz or displacement z in the direction of gravity, the amplitude suppression of the time-series data is suppressed in the flat period at the center of the window function. Can be reduced. Then, with respect to the acceleration _az , velocity _vz or displacement z in the direction of gravity, the half-overlap Han window function can similarly flatten the central period and make the side rope level more flat compared to the rectangular window function. It can be reduced.

When wna (t), wnb (t) and wnc (t) are separately multiplied by the time series data and then the time series data after the frequency inverse conversion is added, the time series data after the frequency inverse conversion is added. Discontinuities can occur. However, when wn (t), which is the sum of wna (t), wnb (t), and wnc (t), is multiplied by the time series data at once, and then the time series data after the inverse frequency transformation is output, the frequency is used. No discontinuity occurs in the time series data after the inverse transformation.

Also, when the half-overlap Han window function is multiplied by the time series data, compared to when the normal Han window function is multiplied by the time series data, in the flat period in the center of the window function, the time after frequency inverse conversion Since the suppression of the amplitude of the series data can be reduced, it is not necessary to acquire the time series data before the frequency conversion for a long time in order to output the time series data after the frequency inverse conversion with the same accuracy.

The gravitational direction displacement calculation device of the present disclosure can be applied, for example, in an application for calculating wave height, if the acceleration sensor of the mobile terminal can receive not only the acceleration in the gravitational direction but also the acceleration of horizontal vibration. ..

S: Ship M: Mobile terminal C: Center of gravity G: Gravity direction displacement calculation device 1: Acceleration acquisition unit 2: Average calculation unit 3: Direction conversion unit 4: Displacement calculation unit 5: Low frequency removal unit 6: Frequency conversion unit 7: Frequency inverse conversion unit 8: Displacement calculation unit 9: Gravity removal unit 10: Frequency conversion unit 11: Low frequency measurement unit

Claims (12)

An acceleration acquisition unit that acquires acceleration information in the three axial directions from the acceleration sensor of the mobile terminal,
An average calculation unit that calculates the time average of acceleration in each axial direction,
Combining time averages of acceleration in each axial direction A direction changing unit that converts acceleration in each axial direction into acceleration in the gravity direction and horizontal direction so as to set the vector direction in the direction of gravity.
A displacement calculation unit that calculates the displacement or velocity in the direction of gravity based on the acceleration in the direction of gravity.
Gravity direction displacement calculation device characterized by being provided with. The average calculation unit calculates the time average of the acceleration in each axial direction over a period longer than the most dominant or longest fluctuation cycle of the acceleration in each axial direction. The gravitational direction displacement calculation device according to claim 1. The direction changing unit is characterized in that the acceleration in each axial direction is normalized by the gravitational acceleration so that the composite vector length of the time average of the acceleration in each axial direction is normalized by the gravitational acceleration. The gravitational direction displacement calculation device according to 1 or 2. Gravity from the time-series data of acceleration, velocity or displacement in the direction of gravity by subtracting the polynomial approximation result of acceleration, velocity or displacement in the direction of gravity from the time-series data of acceleration, velocity or displacement in the direction of gravity. The direction of gravity according to any one of claims 1 to 3, further comprising a low frequency removing portion in a time region for removing the component or removing the gravity component and the DC drift component of the acceleration sensor. Displacement calculator. Acceleration in the direction of gravity by suppressing the DC drift component of the acceleration sensor or by suppressing the DC drift component and gravity component of the acceleration sensor in the frequency spectrum of acceleration, velocity or displacement in the direction of gravity. , The accelerometer's DC drift component is removed from the velocity or displacement frequency spectrum, or the accelerometer's DC drift component and a low frequency removing portion in a frequency region for removing the gravity component are further provided. The gravity direction displacement calculation device according to any one of claims 1 to 4. 4 or 5 according to claim 4, further comprising a low frequency measuring unit for measuring the DC drift frequency of the accelerometer based on the frequency spectrum of the acceleration, velocity or displacement in an arbitrary direction of the accelerometer at rest. The gravity direction displacement calculation device described in. The sixth aspect of claim 6, wherein the low-frequency measuring unit removes a gravity component from time-series data of acceleration, velocity, or displacement in the arbitrary direction when the acceleration sensor is stationary, before frequency conversion. Gravity direction displacement calculation device. The displacement calculation unit is characterized in that the displacement or velocity in the gravity direction is calculated by integrating the time-series data of the acceleration in the gravity direction only twice or once in time. Gravity direction displacement calculation device according to any one of. The displacement calculation unit divides the frequency spectrum of the acceleration in the gravity direction by the square or the first power of 2πf (f is a frequency) or j2πf (j is an imaginary unit), thereby causing the displacement or velocity in the gravity direction. The gravity direction displacement calculation device according to any one of claims 1 to 7, wherein the calculation is performed. When converting the time-series data of acceleration, velocity or displacement in the gravity direction into the frequency spectrum of acceleration, velocity or displacement in the gravity direction, the center is flat in the time-series data of acceleration, velocity or displacement in the gravity direction. The gravity direction displacement calculation device according to claim 5 or 9, further comprising a frequency conversion unit for multiplying a window function. The frequency conversion unit converts the time-series data of acceleration, speed or displacement in the direction of gravity into the frequency spectrum of acceleration, speed or displacement in the direction of gravity, and the time-series of acceleration, speed or displacement in the direction of gravity. The gravity direction displacement calculation device according to claim 10, wherein the data is multiplied by a half overlap Han window function. The displacement calculation unit calculates the wave height at the position where the ship is present based on the acceleration in the direction of gravity measured by using the acceleration sensor of the mobile terminal owned by the sailor. Item 4. The gravity direction displacement calculation device according to any one of Items 1 to 11.

Images