JP2021056001A - Arithmetic system, arithmetic device and arithmetic program - Google Patents

Abstract

To suppress the increase in current consumption and the lengthening of processing time.SOLUTION: An arithmetic system includes: an object; an acceleration sensor that is placed on the object and outputs acceleration data; an angular velocity sensor that is placed on the object and outputs angular velocity data; an output unit that outputs posture information at a predetermined cycle; an arithmetic unit that calculates the posture information by an operation including a matrix operation in a part of periods of the predetermined cycle based on the acceleration data and the angular velocity data; and an interpolation unit that calculates the posture information by interpolation calculation based on the calculation result of the arithmetic unit in some other periods of the predetermined cycle.SELECTED DRAWING: Figure 2

Description

The present disclosure relates to arithmetic systems, arithmetic units and arithmetic programs.

There is known a technique that uses Euler angles (an example of posture information) based on quaternion (quaternion) operations to express the posture of a device.

Japanese Unexamined Patent Publication No. 2016-109608

However, as in the above-mentioned conventional technique, if a matrix operation such as a quaternion operation is performed every processing cycle, there is a problem that the current consumption increases and the processing time becomes long.

Therefore, in one aspect, it is an object of the present invention to suppress an increase in current consumption and an increase in processing time.

On one side, the object and
An acceleration sensor that is placed on the object and outputs acceleration data,
An angular velocity sensor that is placed on the object and outputs angular velocity data,
An output unit that outputs posture information representing the posture of the object at predetermined intervals, and
Based on the acceleration data and the angular velocity data, the attitude information is calculated by an operation including a matrix operation in a part of the predetermined period, and the other part of the period in the predetermined period. An arithmetic system is provided that includes an interpolation unit that calculates the attitude information by interpolation calculation based on the calculation result of the arithmetic unit.

On one side, according to the present invention, it is possible to suppress an increase in current consumption and a long processing time.

It is a system block diagram of the arithmetic system 1 by one Example. It is a figure which shows an example of the functional block of the microcomputer 4. It is explanatory drawing of an example of the interpolation operation. It is a figure which shows the simulation result (without interpolation) of Euler angle output using a sensor log. It is a figure which shows the simulation result (the number of thinnings = 2) of the Euler angle output using a sensor log. It is a figure which shows the simulation result (the number of thinnings = 5) of the Euler angle output using a sensor log. It is a figure which shows the relationship between the thinning number N3 and the current consumption of a microcomputer 4. It is a schematic flowchart which shows the operation example of the arithmetic system 1.

Hereinafter, each embodiment will be described in detail with reference to the accompanying drawings.

FIG. 1 is a system configuration diagram of an arithmetic system 1 according to an embodiment.

The arithmetic system 1 is mounted on a housing 10 (an example of an object). The arithmetic system 1 includes an angular velocity sensor 2, an acceleration sensor 3, and a microcomputer 4 (hereinafter, referred to as “microcomputer 4”).

The angular velocity sensor 2 is arranged in the housing 10 and outputs angular velocity data. In this embodiment, as an example, the angular velocity sensor 2 detects, for example, the angular velocity around three axes.

The acceleration sensor 3 is arranged in the housing 10 and outputs acceleration data. In this embodiment, as an example, the acceleration sensor 3 detects acceleration around three axes, for example. It is assumed that the three axes related to the acceleration sensor 3 are the same as the three axes related to the angular velocity sensor 2.

The microcomputer 4 is arranged in, for example, a housing 10. The microcomputer 4 has a CPU (Central Processing Unit) connected by a data bus, a RAM (Random Access Memory), a ROM (Read Only Memory), and the like.

FIG. 2 is a diagram showing an example of a functional block of the microcomputer 4. FIG. 2 also shows an angular velocity sensor 2 and an acceleration sensor 3 connected to the microcomputer 4. The angular velocity sensor 2 and the acceleration sensor 3 ^{may be connected to the microcomputer 4 via I 2} C (I-squared-C).

The microcomputer 4 includes a data acquisition unit 40, a data buffer unit 41, a normal calculation unit 42 (an example of a calculation unit), a calculation result storage unit 43, an interpolation unit 44, and an output unit 46. The data acquisition unit 40, the normal calculation unit 42, the interpolation unit 44, and the output unit 46 can be realized, for example, by the CPU executing one or more programs in the ROM. The data buffer unit 41 and the calculation result storage unit 43 can be realized by, for example, RAM. The data buffer unit 41 and / or the calculation result storage unit 43 may be realized by a flash memory or the like.

The data acquisition unit 40 acquires the angular velocity data from the angular velocity sensor 2 and acquires the acceleration data from the acceleration sensor 3. In this embodiment, as an example, the data acquisition unit 40 acquires the angular velocity data and the acceleration data of each cycle at each synchronized predetermined cycle.

The data acquisition unit 40 stores (buffers) the acquired angular velocity data and acceleration data in the data buffer unit 41. The angular velocity data and the acceleration data may be stored in a FIFO (First-in, First-out) format for a predetermined number of cycles. The data acquisition unit 40 may store the angular velocity data as an integrated value (a value obtained by integrating the value of the current cycle with the value of the previous cycle). Further, the data acquisition unit 40 may store the acceleration data as an average value (an average value for a predetermined number of cycles before this cycle).

Based on the acceleration data and the angular velocity data in the data buffer unit 41, the normal calculation unit 42 calculates Euler angles (an example of posture information) representing the posture of the housing 10 by a calculation including a quaternion calculation. The calculated Euler angles may be expressed in a coordinate system fixed on the ground, for example, in pitch, roll, and yaw. The normal calculation unit 42 stores the calculation result in the calculation result storage unit 43. The calculation result may be stored in the FIFO format for a predetermined number of cycles.

The operation including the quaternion operation includes a conversion process for converting the angular velocity data and the acceleration data into quaternions, a process for performing a quaternion operation on the quaternion-converted data, and a conversion process for converting the result of the quaternion operation into Euler angles representation.

For example, if the acceleration A = (Ax, Ay, Az), the angular velocity G = (Gx, Gy, Gz), and the immediately preceding (previous) quaternion is Q = (Qw, Qx, Qy, Qz), the normal calculation unit 42 , (Ax, Ay, Az) and (Qw, Qx, Qy, Qz) are cross-producted to calculate the vector B. Specifically, it is as follows.
B = A × Q
Here, let B = (Bx, By, Bz).
Next, the normal calculation unit 42 quaternions the vector B to calculate the vector C.
C = (0, Bx, By, Bz)
Next, the normal calculation unit 42 calculates the vector D by adding the vector C and the vector G. Specifically, it is as follows.
D = (0, Bx + Gx, By + Gy, Bx + Gz)
Next, the normal calculation unit 42 calculates the vector F by multiplying the vector D by the immediately preceding quaternion Q (quaternion calculation). Specifically, it is as follows.
F = D × Q
In general, the quota operation between q = s + iu + jv + kw and q'= s'+ iu' + jv'+ kw'is expressed by the following equation.

次いで、通常演算部４２は、以下のようにして、ベクトルFに対してオイラー変換を行うことで、ピッチ、ロール、及びヨーを算出する。尚、F = (Quat.w,Quat.x,Quat.y,Quat.z)とすると、ピッチE_pitch、ロールE_roll、及びヨーE_yawを算出するためのオイラー変換は、以下の通りである。尚、記号"*"は、乗算を表す。
E_pitch = Arcsin (2 * Quat.w * Quat.y - 2 * Quat.x * Quat.z)
E_roll = Arctan｛(2 * Quat.y * Quat.z + 2 * Quat.w * Quat.x) / (-Quat.w * Quat.w + Quat.x * Quat.x + Quat.y * Quat.y - Quat.z * Quat.z)｝
E_yaw = Arctan｛(2 * Quat.x * Quat.y + 2 * Quat.w * Quat.z) / (Quat.w * Quat.w + Quat.x * Quat.x - Quat.y * Quat.y - Quat.z * Quat.z)｝

Next, the normal calculation unit 42 calculates the pitch, roll, and yaw by performing an oiler conversion on the vector F as follows. If F = (Quat.w, Quat.x, Quat.y, Quat.z), the Euler transformations for calculating _{pitch E pitch} , roll E _roll , and yaw E _{yaw are as follows.} .. The symbol "*" represents multiplication.
E _pitch = Arcsin (2 * Quat.w * Quat.y --2 * Quat.x * Quat.z)
E _roll = Arctan {(2 * Quat.y * Quat.z + 2 * Quat.w * Quat.x) / (-Quat.w * Quat.w + Quat.x * Quat.x + Quat.y * Quat .y --Quat.z * Quat.z)}
E _yaw = Arctan {(2 * Quat.x * Quat.y + 2 * Quat.w * Quat.z) / (Quat.w * Quat.w + Quat.x * Quat.x --Quat.y * Quat. y --Quat.z * Quat.z)}

The interpolation unit 44 calculates Euler angles by interpolation calculation based on the calculation result of the normal calculation unit 42. Euler angles calculated by interpolation operations may be expressed in a coordinate system fixed on the ground, for example, in pitch, roll, and yaw. The interpolation operation may be a method using linear interpolation or non-linear interpolation (exponent, logarithmic, moving average, weighted interpolation, etc.). For example, in FIG. 3, the pitch P5 at the current time t5 is an approximation obtained by linear interpolation using the calculation results P1 to P4 (calculation results of the normal calculation unit 42) of the pitches at the five time points (t0 to t4) before the current time. Calculated on the straight line C1. The same calculation can be performed for rolls and yaws. In this example, the calculation results at the latest 5 time points are used, but the calculation results at a number of time points other than the latest 5 time points (the calculation result of the normal calculation unit 42) may be used. For example, assuming that the previous calculation results of pitch, roll, and yaw are Ep1, Er1, and Ey1, and the results of the previous two times are Ep2, Er2, and Ey2, the interpolation calculation may be executed as follows.
Ep = (Ep1-Ep2) / 2 + Ep1
Er = (Er1-Er2) / 2 + Er1
Ey = (Ey1-Ey2) / 2 + Ey1

The calculation load of the interpolation calculation is significantly lower than that of the calculation by the normal calculation unit 42. This is because the calculation by the normal calculation unit 42 includes a quaternion calculation having a relatively high calculation load. Therefore, by providing the interpolation unit 44 that calculates Euler angles by interpolation calculation without performing quaternion calculation, it is possible to reduce the calculation load for calculating Euler angles. The interpolation unit 44 calculates Euler angles by interpolation calculation at a cycle in which the normal calculation unit 42 does not normally perform. Either one of the normal calculation unit 42 and the interpolation unit 44 calculates the Euler angles, so that the Euler angles are calculated every predetermined cycle as a whole.

In this embodiment, as an example, the interpolation unit 44 determines whether or not a predetermined switching condition is satisfied for each predetermined cycle, and calculates Euler angles by interpolation calculation in a cycle in which the predetermined switching condition is satisfied. In this case, the normal calculation unit 42 does not perform the calculation in the cycle.

The predetermined switching condition is arbitrary, but may be satisfied, for example, when the output of the acceleration sensor 3 and the output of the angular velocity sensor 2 are smaller than the predetermined level. The predetermined level may correspond to the level of output that occurs when the housing 10 is in a substantially stationary state. In this case, since the interpolation unit 44 performs the calculation during the period when the Euler angles do not fluctuate significantly, the calculation load can be efficiently reduced as compared with the case where the normal calculation unit 42 always performs the calculation.

In another example, the predetermined switching condition is when the number of cycles in which the normal calculation unit 42 continuously performs the calculation reaches the predetermined number N1 regardless of the output of the acceleration sensor 3 and the output of the angular velocity sensor 2. May be satisfied. In this case, the predetermined number N1 may be any number of 2 or more.

In this embodiment, as an example, when the predetermined switching condition is satisfied, the interpolation unit 44 subsequently determines whether or not the predetermined return condition is satisfied at each predetermined cycle, and in the cycle in which the predetermined return condition is satisfied. , Stop the interpolation operation. In this case, the normal calculation unit 42 performs the calculation in the cycle. After that, the interpolation unit 44 similarly determines whether or not the predetermined switching condition is satisfied for each predetermined cycle, and calculates the Euler angles by the interpolation calculation in the cycle in which the predetermined switching condition is satisfied.

The predetermined return condition is arbitrary, but for example, in a configuration in which the predetermined switching condition is satisfied when the output of the acceleration sensor 3 and the output of the angular velocity sensor 2 are smaller than the predetermined level, the predetermined return condition is the acceleration sensor. It may be satisfied when the output of 3 and the output of the angular velocity sensor 2 are equal to or higher than a predetermined level.

In another example, the predetermined return condition is when the number of cycles in which the interpolation unit 44 continuously performs the interpolation calculation reaches the predetermined number N2 regardless of the output of the acceleration sensor 3 and the output of the angular velocity sensor 2. May be satisfied. In this case, the predetermined number N2 may be any number of 2 or more.

In this embodiment, as an example, the predetermined switching condition is satisfied when the normal calculation unit 42 performs the calculation, and the predetermined return condition is the number of cycles in which the interpolation unit 44 continuously performs the interpolation calculation. Satisfied when the number N3 is reached. That is, the interpolation unit 44 performs the interpolation calculation in the next predetermined number N3 cycle for each cycle in which the normal calculation unit 42 performs the calculation. For example, when the predetermined number N3 = 3, the interpolation unit 44 continuously performs the interpolation calculation in the next three cycles for each cycle in which the normal calculation unit 42 performs the calculation. In this case, only one of the four cycles is the cycle in which the normal calculation unit 42 performs the calculation. Hereinafter, the predetermined number N3 is also referred to as “thinning number N3”.

In the modified example, the predetermined switching condition may be realized by a combination of a plurality of conditions. Similarly, the predetermined return condition may be realized by a combination of a plurality of conditions.

In this way, in the present embodiment, in the state where the normal calculation unit 42 performs the calculation (hereinafter, also referred to as “normal calculation state”), it is determined whether or not the predetermined switching condition is satisfied at each predetermined cycle. When a predetermined switching condition is satisfied, the state in which the normal calculation unit 42 performs the calculation is switched to the state in which the interpolation unit 44 performs the interpolation calculation. On the other hand, when the predetermined switching condition is not satisfied, the state in which the normal calculation unit 42 performs the calculation continues. Further, in the state where the interpolation unit 44 performs the interpolation calculation (hereinafter, also referred to as “interpolation calculation state”), it is determined whether or not the predetermined return condition is satisfied at each predetermined cycle, and when the predetermined return condition is satisfied. Switches from the state in which the interpolation unit 44 performs the interpolation calculation to the state in which the normal calculation unit 42 performs the calculation. On the other hand, if the predetermined return condition is not satisfied, the state in which the interpolation unit 44 performs the interpolation calculation continues.

The output unit 46 outputs Euler angles at predetermined intervals while the arithmetic system 1 is operating. For example, the output unit 46 outputs Euler angles to a display (not shown) or the like that can be provided in the housing 10. Alternatively, the output unit 46 may output Euler angles to an external device via a wired or wireless communication network. The external device may be, for example, a PC (Personal Computer), a server, or the like. The wireless communication network may be based on Bluetooth (registered trademark) or Wi-Fi (Wireless Fidelity), and includes a wireless communication network for mobile phones, the Internet, World Wide Web, WAN (Wide Area Network), and the like. It may be.

Next, the effect of this embodiment will be described with reference to FIGS. 4A and after.

4A to 4C are diagrams showing the simulation results of Euler angles output (here, roll as an example) using sensor logs, and the horizontal axis is time. FIG. 4A shows the simulation result without interpolation (the configuration in which the normal calculation unit 42 always performs the calculation), and FIGS. 4B and 4C show the simulation result with interpolation. FIG. 4B shows a case where the number of thinnings N3 = 2, and FIG. 4C shows a case where the number of thinnings N3 = 5. FIG. 5 is a diagram showing the relationship between the number of thinnings N3 and the current consumption of the microcomputer 4. The horizontal axis represents the number of thinnings N3, and the vertical axis represents the current consumption. The current consumption is normalized as "1" when the number of thinnings N3 = 0 (that is, no interpolation).

When the number of thinnings N3 = 2, as shown in the Q1 part of FIG. 4B, the Euler angles output fluctuates slightly as compared with the case shown in FIG. 4A, but there is almost no influence of the interpolation calculation. Similarly, when the number of thinnings N3 = 5, as shown in the Q2 part of FIG. 4C, the Euler angle output fluctuates slightly as compared with the case shown in FIG. 4A, but there is almost no influence of the interpolation calculation.

Further, as shown in FIG. 5, as the number of thinnings N3 increases, the current consumption decreases. For example, when the number of thinnings N3 = 3, the current consumption when the number of thinnings N3 = 0 can be reduced by about 30%.

In this way, according to this embodiment, it is possible to reduce the current consumption of the microcomputer 4 for calculating Euler angles without significantly reducing the accuracy of Euler angles output. The current consumption is reduced because the calculation load is reduced. When the calculation load is reduced, the processing time is also reduced. Therefore, according to this embodiment, it is possible to suppress an increase in current consumption and a long processing time.

Next, an operation example of the arithmetic system 1 of this embodiment will be described with reference to FIG.

FIG. 6 is a schematic flowchart showing an operation example of the arithmetic system 1. The process shown in FIG. 6 is executed at predetermined intervals during the operation of the arithmetic system 1.

In step S600, the arithmetic system 1 acquires the angular velocity data and the acceleration data of the current cycle and stores them in the data buffer unit 41.

In step S601, the arithmetic system 1 determines whether or not the arithmetic state is a normal arithmetic state. As described above, the normal calculation state is a state in which the normal calculation unit 42 performs a calculation. The calculation state includes a normal calculation state and an interpolation calculation state. The interpolation calculation state is a state in which the interpolation unit 44 performs the interpolation calculation as described above. If the determination result is "YES", the process proceeds to step S602, and if not, the process proceeds to step S614.

In step S602, the arithmetic system 1 determines whether or not a predetermined switching condition is satisfied. If the determination result is "YES", the process proceeds to step S610, and if not, the process proceeds to step S604.

In step S604, the normal calculation unit 42 of the calculation system 1 executes the sensor unit conversion process. The sensor unit conversion process includes a conversion process for converting angular velocity data and acceleration data into quarters.

In step S606, the normal calculation unit 42 of the calculation system 1 executes the noise determination process. If the noise is greater than the predetermined level, the data of this cycle may be discarded. In this case, the process returns to step S600 in the next cycle. Note that step S606 may be executed before step S604.

In step S608, the normal calculation unit 42 of the calculation system 1 performs a quaternion calculation.

In step S609, the normal calculation unit 42 of the calculation system 1 executes a conversion process for converting the quaternion calculation result into Euler angles representation. As a result, Euler angles of this cycle can be obtained.

In step S610, the calculation system 1 sets the calculation state to the interpolation calculation state.

In step S612, the interpolation unit 44 of the calculation system 1 calculates Euler angles by interpolation calculation based on the calculation result of the normal calculation unit 42 before the current cycle. As a result, Euler angles of this cycle can be obtained.

In step S614, the arithmetic system 1 determines whether or not the predetermined return condition is satisfied. If the determination result is "YES", the process proceeds to step S616, and if not, the process proceeds to step S612.

In step S616, the calculation system 1 sets the calculation state to the normal calculation state. In this case, when step S616 is completed, the process proceeds to step S604.

In step S618, the output unit 46 of the arithmetic system 1 outputs the Euler angles obtained in step S608 or step S612. The output unit 46 may only output (store) Euler angles to a predetermined storage device at predetermined intervals. In this case, the Euler angle data in the storage device may be output to an external device or the like later.

Although each embodiment has been described in detail above, the present invention is not limited to a specific embodiment, and various modifications and changes can be made within the scope of the claims. It is also possible to combine all or a plurality of the components of the above-described embodiment.

For example, in the above-described embodiment, only the calculation result of the normal calculation unit 42 is used in the interpolation calculation, but the calculation is not limited to this. For example, when performing an interpolation calculation in a certain cycle, the calculation result of the normal calculation unit 42 before the cycle and the calculation result of the interpolation calculation before the cycle may be used.

Further, in the above-described embodiment, the housing 10 is an example of an object, but the arithmetic system 1 can be mounted on any object. For example, the arithmetic system 1 may be worn by the user (for example, shoes) or may be worn by the user himself / herself.

Further, in the above-described embodiment, the normal calculation unit 42 uses quaternion operations (an example of matrix operations) when calculating Euler angles representing the postures of the housing 10, but uses other matrix operations. May be good. For example, the normal calculation unit 42 may calculate Euler angles representing the posture of the housing 10 using the following rotation matrix.

尚、ここで回転の方向は、Ｒｘはｙ軸をｚ軸に向ける方向であり、Ｒｙはｚ軸をｘ軸に向ける方向であり、Ｒｚはｘ軸をｙ軸に向ける方向である。

Here, as for the direction of rotation, Rx is the direction in which the y-axis is directed to the z-axis, Ry is the direction in which the z-axis is directed to the x-axis, and Rz is the direction in which the x-axis is directed to the y-axis.

1 Calculation system 2 Angle speed sensor 3 Accelerometer 4 Microcomputer 10 Housing 40 Data acquisition unit 41 Data buffer unit 42 Normal calculation unit 43 Calculation result storage unit 44 Interpretation unit 46 Output unit

Claims (8)

With an object
An acceleration sensor that is placed on the object and outputs acceleration data,
An angular velocity sensor that is placed on the object and outputs angular velocity data,
An output unit that outputs posture information representing the posture of the object at predetermined intervals, and
Based on the acceleration data and the angular velocity data, the attitude information is calculated by an operation including a matrix operation in a part of the predetermined period, and the other part of the period in the predetermined period. A calculation system including an interpolation unit that calculates the attitude information by an interpolation calculation based on the calculation result of the calculation unit. The arithmetic system according to claim 1, wherein the matrix operation includes a quaternion operation. The arithmetic system according to claim 2, wherein the other partial cycle includes a cycle in which a predetermined switching condition is satisfied. The calculation system according to claim 3, wherein the predetermined switching condition is satisfied when the output of the acceleration sensor and the output of the angular velocity sensor are smaller than a predetermined level. The calculation system according to claim 2, wherein the interpolation unit performs the interpolation calculation in the next predetermined number of cycles for each cycle in which the calculation unit performs the calculation. The arithmetic system according to any one of claims 1 to 5, wherein the posture information is expressed by Euler angles. A data acquisition unit that acquires acceleration data and angular velocity data at predetermined intervals from an acceleration sensor and an angular velocity sensor placed on an object.
Based on the acceleration data and the angular velocity data, a calculation unit that calculates posture information representing the posture of the object in a part of each predetermined cycle by a calculation including a matrix calculation, and a calculation unit.
An arithmetic unit including an interpolation unit that calculates the posture information by interpolation calculation based on the calculation result of the calculation unit in a part of other cycles for each predetermined cycle. Acceleration data and angular velocity data are acquired at predetermined intervals from the accelerometer and angular velocity sensor placed on the object.
Based on the acceleration data and the angular velocity data, posture information representing the posture of the object is calculated by a calculation including a matrix calculation in a part of the predetermined cycles.
The posture information is calculated by interpolation calculation based on the calculation result of the calculation in some other cycles for each predetermined cycle.
An arithmetic program that causes a computer to perform processing.

Images