JP5571027B2 - Portable device, program and method for correcting gravity vector used for autonomous positioning - Google Patents

Description

  The present invention relates to a technique for correcting a gravity vector used for autonomous positioning.

  Conventionally, there is a technique for correcting a gravity vector in order to estimate an accurate terminal posture (see, for example, Patent Document 1). According to this technique, the portable device has an angular velocity sensor and an acceleration sensor, and sequentially calculates the direction of gravity from the angular velocity data and the acceleration data. The gravity vector is calculated based on the angular velocity data. First, the gravity vector is corrected so as to approach the direction of the acceleration vector. Next, when the terminal performs a predetermined motion, the gravity vector is corrected so as to approach the direction of gravity estimated from the predefined relationship between the acceleration data and the angular velocity data. Finally, the gravity vector is corrected so as to approach the average direction of the acceleration vector in a predetermined period. An accurate terminal posture can be estimated using the gravity vector corrected in this way.

JP 2010-263930 A

  According to the technique described in Patent Document 1, it is necessary to mount at least an angular velocity sensor on a portable device. Moreover, it is necessary to define the relationship between acceleration data and angular velocity data in advance, and it is not possible to cope with individual differences in the manner of holding the portable device.

  Therefore, an object of the present invention is to provide a portable device, a program, and a method that can correct a gravity vector for a portable device having an autonomous positioning function regardless of individual differences in the manner of holding the device. .

According to the present invention, an acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
A mobile device having autonomous positioning means for outputting an estimated azimuth angle d autonomously measured using acceleration data and geomagnetic data,
A geomagnetic buffer means for accumulating a plurality of geomagnetic data within a predetermined period output from the geomagnetic sensor;
Estimated azimuth angle buffer means for accumulating a plurality of estimated azimuth angles d within a predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer means are divided for each predetermined angle section, and it is determined whether or not the observed frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution. Azimuth distribution verification means;
When it is determined that the distribution is uniform by the azimuth distribution verification means, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data stored in the geomagnetic buffer means is calculated, and the gravity vector is calculated as an autonomous positioning means. Gravity vector calculating means for outputting to
The autonomous positioning means corrects the estimated azimuth angle d using the gravity vector input from the gravity vector calculation means.

Ｅ＝Ｎ／ｋ
Ｎ：全観測度数
ｋ：所定角度区間の数
算出されたχ^２値が、基準χ^２値よりも小さい場合、一様分布であると判定することも好ましい。 According to another embodiment of the portable device of the present invention,
The azimuth distribution verification means is determined by χ ² test,
Reference χ ² values satisfying a predetermined significance level value are stored in advance when the number k of the predetermined angle interval i is a degree of freedom,
Χ ² value is calculated by the following formula,

E = N / k
N: total observed frequency k: number calculated chi ² value of a predetermined angle interval is smaller than the reference chi ² values, it is also preferable to determine that a uniform distribution.

Ｎ：全観測度数
Ｇ：重力ベクトル
ｍ：地磁気ベクトル
ことも好ましい。 According to another embodiment of the portable device of the present invention,
The gravity vector calculation means calculates the gravity vector from the geomagnetic data stored in the geomagnetic buffer means by the following formula.

It is also preferable that N: total observation frequency G: gravity vector m: geomagnetic vector.

According to the present invention,
An acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
A computer mounted on a portable device having a function,
A program that causes a computer to function as an autonomous positioning means that outputs an estimated azimuth angle d autonomously measured using acceleration data and geomagnetic data,
A geomagnetic buffer means for accumulating a plurality of geomagnetic data within a predetermined period output from the geomagnetic sensor;
Estimated azimuth angle buffer means for accumulating a plurality of estimated azimuth angles d within a predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer means are divided for each predetermined angle section, and it is determined whether or not the observed frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution. Azimuth distribution verification means;
When it is determined that the distribution is uniform by the azimuth distribution verification means, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data stored in the geomagnetic buffer means is calculated, and the gravity vector is calculated as an autonomous positioning means. Gravity vector calculating means for outputting to
The autonomous positioning means makes the computer function so as to correct the estimated azimuth angle d using the gravity vector input from the gravity vector calculating means.

Ｅ＝Ｎ／ｋ
Ｎ：全観測度数
ｋ：所定角度区間の数
算出されたχ^２値が、基準χ^２値よりも小さい場合、一様分布であると判定する
ようにコンピュータを更に機能させることも好ましい。 According to another embodiment of the program for the portable device of the present invention,
The azimuth distribution verification means is determined by χ ² test,
Reference χ ² values satisfying a predetermined significance level value are stored in advance when the number k of the predetermined angle interval i is a degree of freedom,
Χ ² value is calculated by the following formula,

E = N / k
N: total observed frequency k: number calculated chi ² value of a predetermined angle interval is smaller than the reference chi ² values, it is also preferred to further features a computer to determine that the uniform distribution.

Ｎ：全観測度数
Ｇ：重力ベクトル
ｍ：地磁気ベクトル
ようにコンピュータを更に機能させることも好ましい。 According to another embodiment of the program for the portable device of the present invention,
The gravity vector calculation means calculates the gravity vector from the geomagnetic data stored in the geomagnetic buffer means by the following formula.

It is also preferable to further cause the computer to function as N: total observation power G: gravity vector m: geomagnetic vector.

According to the present invention,
An acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
A method for correcting a gravity vector in a portable device having:
Portable device
An autonomous positioning function that calculates a gravity vector using acceleration data and geomagnetic data, and outputs an estimated azimuth angle d autonomously measured;
A geomagnetic buffer function for storing a plurality of geomagnetic data output from a geomagnetic sensor within a predetermined period;
An estimated azimuth angle buffer function for accumulating a plurality of estimated azimuth angles d within a predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer function are divided for each predetermined angle section, and it is determined whether or not the observation frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution. A first step;
When it is determined that the distribution is uniform, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data accumulated in the geomagnetic buffer function is calculated, and the gravity vector is output to the autonomous positioning means. Steps,
The autonomous positioning function includes a third step of correcting the estimated azimuth angle d using the gravity vector input from the gravity vector calculation means.

  According to the portable device, program, and method of the present invention, the gravitational vector is calculated from the accumulated estimated azimuth at the timing when uniformly distributed geomagnetic data is obtained, and the autonomous positioning function is corrected using the gravitational vector. By doing so, the gravity vector can be corrected regardless of individual differences in the manner of holding the device. That is, according to the present invention, an angular velocity sensor is not required, and this can be corrected even if the error of the gravity vector differs due to individual differences in the manner in which the device is held.

It is a functional block diagram of the portable apparatus in this invention. This is a user coordinate system representing the relationship between the correct gravity vector G and geomagnetic vector M. It is a user coordinate system representing the relationship between an erroneous gravity vector G and geomagnetic vector M. It is explanatory drawing which divided the omnidirectional angle 360 degrees into 8 sections. It is explanatory drawing when the observation frequency Oi is not uniform distribution. It is explanatory drawing in case observation frequency Oi is uniform distribution.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a functional configuration diagram of a portable device according to the present invention.

  The mobile device 1 is possessed by a user and is, for example, a mobile phone or a smartphone. According to FIG. 1, the mobile device 1 includes an acceleration sensor 101 and a geomagnetic sensor 102.

  The acceleration sensor 101 detects a change in acceleration, that is, a speed per unit time. In the case of the three-axis type that can detect the tilt of the portable device, three-dimensional acceleration can be detected, and the measurement of the earth's gravity (static acceleration) can be supported.

  The geomagnetic sensor 102 measures the direction of geomagnetism in three axial directions (front-rear direction, left-right direction, and up-down direction). The geomagnetic sensor 102 separates the Hall elements and outputs values detected from the separated Hall elements.

  The portable device 1 also includes an autonomous positioning unit 111, an azimuth angle buffer unit 112, a geomagnetic buffer unit 113, an azimuth distribution test unit 114, a gravity vector calculation unit 115, and an application processing unit 116. The application processing unit 116 provides various services to the user using the estimated azimuth angle d derived from the autonomous positioning unit 111. These functional components are realized by executing a program that causes a computer mounted on the mobile device 1 to function. Moreover, the flow of processing of these functional components can also be understood as a gravity vector detection method in autonomous positioning.

  The autonomous positioning unit 111 calculates a gravity vector using acceleration data and geomagnetic data, and outputs an estimated azimuth angle d (°) measured autonomously. The autonomous positioning unit 111 is based on the existing dead reckoning (DR) technology (autonomous dead reckoning navigation technology), and even if it is in a place where it cannot receive positioning radio waves, such as indoors, its position is autonomous. You can measure the position. The dead reckoning technology is a relative self-position estimation technology using an encoder or inertial sensor, and generally adds geomagnetic data or angular velocity data (gyroscope) and acceleration data to known initial values (position and orientation). In addition, the subsequent position and orientation are calculated. Therefore, errors are accumulated only with the dead reckoning technology. That is, the gravitational vector error is also accumulated.

  According to FIG. 1, the autonomous positioning unit 111 inputs the gravity vector G from the gravity vector calculation unit 115. The autonomous positioning unit 111 resets (resets) its own estimated gravity vector to the gravity vector G when the gravity vector G is input from the gravity vector calculation unit 115. Thereby, the error of the estimated gravity vector based on autonomous positioning can be corrected, and the accuracy of subsequent autonomous positioning can be improved.

  The estimated azimuth angle buffer unit 112 stores a plurality of estimated azimuth angles d output from the autonomous positioning unit 111 within a predetermined period. Within the predetermined period may be, for example, every time the gravity vector calculation unit 115 outputs the gravity vector G to the autonomous positioning unit 111, or may be every arbitrary predetermined time period. The accumulated estimated azimuth angle d is referred to by the azimuth angle distribution test unit 114.

  The geomagnetic buffer unit 113 accumulates a plurality of geomagnetic data output from the geomagnetic sensor 102 within a predetermined period. Here, it is preferable that the time period of the geomagnetic data accumulated in the geomagnetic buffer unit 113 and the time period of the estimated azimuth angle d accumulated in the estimated azimuth angle buffer unit 112 coincide. That is, the time period is preferably, for example, every time the gravity vector calculation unit 115 outputs the gravity vector G to the autonomous positioning unit 111.

  FIG. 2 is a user coordinate system representing the relationship between the correct gravity vector G and geomagnetic vector M.

  FIG. 2 shows the traveling azimuth angle θ estimated by the correct gravity vector G. Here, the gravity vector G is located on the central axis of a cone composed of a group of a large number of geomagnetic vectors m.

  The geomagnetism has come from the south to the north, and the geomagnetic vector is a vector indicating the north (magnetic north) direction with a specific dip angle φ depending on the region. The dip angle φ of the geomagnetic vector represents an angle with respect to the gravity vector G. A projection vector onto a plane (ground surface) perpendicular to the gravity vector G represents a north-facing vector. The rotation angle θ from the north-facing vector to the forward-facing vector represents the traveling direction azimuth. The angle is expressed clockwise from north.

  As is clear from FIG. 2, the geomagnetic vector is distributed on the conical surface with the gravity vector as an axis. That is, if a geomagnetic vector of 360 ° in all directions can be collected, the central axis of the cone can be determined to be a gravity vector.

  FIG. 3 is a user coordinate system representing the relationship between the erroneous gravity vector G and the geomagnetic vector M.

  According to FIG. 3, an offset is applied to the gravity vector, so that the group of geomagnetic vectors appears to rotate from the position where it should be in the coordinate system based on the gravity vector. The traveling azimuth angle θ estimated by such a gravity vector G is also erroneous. In particular, the gravity vector estimated by autonomous positioning causes an error with time. According to FIG. 3, since the gravity vector includes an error, the correct traveling azimuth angle θ cannot be estimated.

  However, as shown in FIG. 2, it is difficult to reliably collect 360-degree geomagnetic vectors. In general, it is necessary to cause a user holding the portable terminal to perform a predetermined action such that the user is directed to 360 ° in all directions (around the circumference). Forcing the user to perform such an action every time the gravity vector is corrected is extremely uncomfortable.

  Therefore, the azimuth distribution verification unit 114 detects the timing at which the distribution of the estimated azimuth angle d accumulated in the estimated azimuth angle buffer unit 112 is uniform with respect to all directions. At this timing, the gravity vector can be detected by calculating the central axis of the cone from the geomagnetic vector data m group stored in the geomagnetic buffer unit 113. The gravity vector is output to the autonomous positioning unit 111. The autonomous positioning unit 111 can correct the estimated gravity vector in autonomous positioning by resetting it to the input gravity vector G.

[Azimuth distribution test unit 114]
The azimuth distribution verification unit 114 determines whether or not the plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer unit 112 is a uniform distribution. In order to estimate the gravity vector, the geomagnetic vector needs to be distributed as uniformly as possible on the conical surface. The azimuth distribution verification unit 114 calculates whether or not sufficient geomagnetic data for buffering the gravity vector is buffered from the estimated azimuth distribution.

The azimuth distribution verification unit 114 instructs the gravity vector calculation unit 115 to calculate a gravity vector when the following two conditions are satisfied, for example.
(Condition 1) The number of estimated azimuth angles accumulated in the estimated azimuth angle buffer unit 112 is equal to or greater than a predetermined threshold N (for example, 300).
(Condition 2) The estimated azimuth angle accumulated in the estimated azimuth angle buffer unit 112 is a uniform distribution.
Hereinafter, the condition 2 will be described in detail.

  The azimuth distribution verification unit 114 divides the estimated azimuth angle d for each predetermined angle section, and determines whether or not the observation frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution.

  FIG. 4 is an explanatory diagram in which the omnidirectional angle 360 ° is divided into eight sections.

  According to FIG. 4, the omnidirectional angle 360 ° is divided into 8 sections, and the predetermined angle section is, for example, 45 °. Then, the number of estimated azimuth angles accumulated in the estimated azimuth angle buffer unit 112 is tabulated as “observation frequency” for each predetermined angle section.

  FIG. 5 is an explanatory diagram when the observation frequency Oi is not uniformly distributed.

  According to FIG. 5, since the observation frequency Oi for each section is not uniformly distributed, the gravity vector serving as the central axis of the cone in the geomagnetic vector is also incorrect.

  FIG. 6 is an explanatory diagram when the observation frequency Oi has a uniform distribution.

  FIG. 6 shows the number of estimated azimuth angles (observation frequency) accumulated in the estimated azimuth angle buffer unit 112 for each predetermined angle section. Since the observation frequency Oi for each section has a uniform distribution, the gravity vector serving as the central axis of the cone in the geomagnetic vector is also correct.

Ｅ＝Ｎ／ｋ
Ｎ：全観測度数
ｋ：所定角度区間の数
これは、「観測度数Ｏiを、平均μｉで分散σｉの正規分布に従う、ｋ個の独立なランダム変数とすると、統計量はχ^２に従う」ことを意味する。χ^２検定は、ｋという１個の母数をもつ。これは観測度数Ｏiの自由度に等しい正の整数である。観測度数Ｏiと平均Ｅ（期待度数）とのずれが偶然であるならば、χ^２値は、自由度ｋ−１のχ^２検定に従う。 The azimuth distribution verification unit 114 performs determination by χ ² test. The “χ ² test” is a kind of probability distribution in estimated statistics. According to the present invention, the χ ² value is calculated by the following equation.

E = N / k
N: total observed frequency k: number which the predetermined angle interval, "the observed frequency Oi, follows a normal distribution of variance σi in average .mu.i, When k independent random variables, statistics follows the chi ^2" that means. The χ ² test has a single parameter k. This is a positive integer equal to the degree of freedom of the observation frequency Oi. If the deviation between the observed frequency Oi and the mean E (expected frequency) is a coincidence, the χ ² value follows the χ ² test with k−1 degrees of freedom.

Here, the azimuth distribution test unit 114 stores in advance a reference χ ² value that satisfies a predetermined significance level value when the number k of the predetermined angle section i is a degree of freedom. It is determined calculated chi ² value is less than the reference chi ² value, to be a uniform distribution.
Calculated χ ² value <reference χ ² value: uniform distribution

This will be described using specific numerical values.
In the example of FIG. 6, k = 8, N = 361, and χ ² = 5.21.
Further, it is assumed that the following values (χ ² distribution table) are stored as reference χ ² values.
When the degree of freedom is 7 and the significance level is 0.05, the reference χ ² _0.05 = 14.067
Then, the null hypothesis that the data has a uniform distribution cannot be rejected as follows.
χ ² <χ ² _0.05
Therefore, the group of estimated azimuth angles accumulated in the estimated azimuth angle buffer unit 112 is determined to have a uniform distribution.

[Gravity vector calculation unit 115]
The gravity vector calculation unit 115 calculates a gravity vector serving as the central axis of the cone from the geomagnetic data stored in the geomagnetic buffer unit 113 when the azimuth distribution verification unit 114 determines that the distribution is uniform. .

Ｇ：重力ベクトル
ｍ：地磁気ベクトル The gravity vector calculation unit 115 calculates the gravity vector from the geomagnetic data m stored in the geomagnetic buffer unit 113 by the following formula.

G: gravity vector m: geomagnetic vector

  Then, the gravity vector calculation unit 115 feeds back the calculated gravity vector G to the autonomous positioning unit 111. Thereby, the autonomous positioning unit 111 replaces the gravity vector estimated by itself with the gravity vector G fed back.

  As described above in detail, according to the portable device, program, and method of the present invention, the gravity vector is calculated at the timing when the uniformly distributed geomagnetic data is obtained, and the autonomous positioning function is performed using the gravity vector. By correcting, the gravity vector can be corrected regardless of individual differences in the manner of holding the device. That is, according to the present invention, an angular velocity sensor is not required, and this can be corrected even if the error of the gravity vector differs due to individual differences in the manner in which the device is held.

  Various changes, modifications, and omissions of the above-described various embodiments of the present invention can be easily made by those skilled in the art. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

DESCRIPTION OF SYMBOLS 1 Portable apparatus 101 Acceleration sensor 102 Geomagnetic sensor 103 GPS positioning part 111 Autonomous positioning part 112 Estimated azimuth angle buffer part 113 Geomagnetic buffer part 114 Azimuth angle distribution verification part 115 Gravity vector calculation part 116 Application processing part

Claims (7)

An acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
An autonomous positioning means for outputting an estimated azimuth angle d autonomously measured using the acceleration data and the geomagnetic data,
A geomagnetic buffer means for accumulating a plurality of geomagnetic data within a predetermined period output from the geomagnetic sensor;
Estimated azimuth angle buffer means for accumulating a plurality of estimated azimuth angles d within the predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer means are divided for each predetermined angle section, and it is determined whether or not the observed frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution. Azimuth distribution verification means to
When the azimuth distribution test means determines that the distribution is uniform, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data stored in the geomagnetic buffer means is calculated, and the gravity vector is calculated as the gravity vector. A gravity vector calculating means for outputting to the autonomous positioning means,
The mobile device according to claim 1, wherein the autonomous positioning unit corrects the estimated azimuth angle d using the gravity vector input from the gravity vector calculation unit. The azimuth distribution verification means is determined by χ ² test,
When the number k of the predetermined angle section i is the degree of freedom, a reference χ ² value that satisfies a predetermined significance level value is stored in advance,
Χ ² value is calculated by the following formula,

E = N / k
N: total observed frequency k: number calculated chi ² value of a predetermined angle interval, if the reference chi less than ² values, a mobile according to claim 1, wherein determining that a uniform distribution apparatus. The gravity vector calculation means calculates a gravity vector from the geomagnetic data m stored in the geomagnetic buffer means by the following formula.

The mobile device according to claim 1 or 2 , wherein N: total observation frequency G: gravity vector m: geomagnetic vector. An acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
A computer mounted on a portable device having a function,
A program that causes a computer to function as an autonomous positioning means that outputs an estimated azimuth angle d autonomously measured using the acceleration data and the geomagnetic data,
A geomagnetic buffer means for accumulating a plurality of geomagnetic data within a predetermined period output from the geomagnetic sensor;
Estimated azimuth angle buffer means for accumulating a plurality of estimated azimuth angles d within the predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer means are divided for each predetermined angle section, and it is determined whether or not the observed frequency Oi of the estimated azimuth angle d for each predetermined angle section i is a uniform distribution. Azimuth distribution verification means to
When the azimuth distribution test means determines that the distribution is uniform, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data stored in the geomagnetic buffer means is calculated, and the gravity vector is calculated as the gravity vector. A gravity vector calculating means for outputting to the autonomous positioning means,
The autonomous positioning program causes a computer to function so as to correct the estimated azimuth angle d using the gravity vector input from the gravity vector calculation unit. The azimuth distribution verification means is determined by χ ² test,
When the number k of the predetermined angle section i is the degree of freedom, a reference χ ² value that satisfies a predetermined significance level value is stored in advance,
Χ ² value is calculated by the following formula,

E = N / k
N: total observed frequency k: number calculated chi ² value of a predetermined angle interval, if the reference chi less than ² values, characterized in that the computer is further caused to function so as to determine that the uniform distribution The program for portable devices according to claim 4 . The gravity vector calculation means calculates a gravity vector from the geomagnetic data m stored in the geomagnetic buffer means by the following formula.

6. The program for a portable device according to claim 4 , wherein the computer further functions as N: total observation frequency G: gravity vector m: geomagnetic vector. An acceleration sensor that outputs triaxial acceleration data;
A geomagnetic sensor that outputs triaxial geomagnetic data;
A method for correcting a gravity vector in a portable device having:
The portable device is:
An autonomous positioning function that calculates a gravity vector using the acceleration data and the geomagnetic data, and outputs an estimated azimuth angle d autonomously measured;
A geomagnetic buffer function for accumulating a plurality of geomagnetic data within a predetermined period output from the geomagnetic sensor;
An estimated azimuth angle buffer function that accumulates a plurality of estimated azimuth angles d within the predetermined period output from the autonomous positioning means;
A plurality of estimated azimuth angles d accumulated in the estimated azimuth angle buffer function are divided for each predetermined angle interval, and it is determined whether or not the observation frequency Oi of the estimated azimuth angle d for each predetermined angle interval i is a uniform distribution. A first step to:
When it is determined that the distribution is uniform, a gravity vector serving as a central axis of a cone composed of a group of geomagnetic data accumulated in the geomagnetic buffer function is calculated, and the gravity vector is output to the autonomous positioning means. Two steps,
The autonomous positioning function includes a third step of correcting the estimated azimuth angle d using the gravity vector input from the gravity vector calculating means.

Images