WO2011093447A1

WO2011093447A1 - Calculation device, control method for calculation device, control program, and recording medium

Info

Publication number: WO2011093447A1
Application number: PCT/JP2011/051748
Authority: WO
Inventors: 正克興梠; 武志蔵田; 隆一郎富永; 秀人嶌岡
Original assignee: 独立行政法人産業技術総合研究所; 三洋電機株式会社
Priority date: 2010-01-29
Filing date: 2011-01-28
Publication date: 2011-08-04
Also published as: JPWO2011093447A1; JP5565736B2

Abstract

In order to estimate the parameter of the motion state of a mobile object with high accuracy, a positioning device (1) is provided with a holding state estimation unit (10) which specifies the holding state of the device in a mobile object from measurement data, and a movement amount estimation unit (12) which performs calculation processing corresponding to the holding state specified by the holding state estimation unit (10) and calculates the movement vector of the mobile object from the measurement data, thereby being able to estimate the movement vector of the mobile object with high accuracy.

Description

COMPUTER DEVICE, COMPUTER DEVICE CONTROL METHOD, CONTROL PROGRAM, AND RECORDING MEDIUM

The present invention relates to a calculation device that calculates a parameter indicating a motion state of a moving body based on an output from a sensor held by the moving body, and more specifically, to specify the position of the moving body, The present invention relates to a calculation device that calculates the amount of movement of a moving object.

In the conventional positioning device that identifies the position of a moving body, there are two types of devices that use external infrastructure devices such as GPS (Global Positioning System) and cell-based positioning, and self-contained measuring devices (specifically In some cases, an autonomous navigation system that sequentially estimates the position and orientation of a moving object using an acceleration sensor or the like is used.

The present invention relates to the latter positioning device. In the case of an autonomous navigation system, a measuring device is often fixedly mounted on a moving object, and a single type / setting of a moving amount estimating means is used. Note that, as prior arts for specifying the movement state of a moving body by detection by a measuring device, the following Patent Documents 1 to 3, Non-Patent Documents 1 to 4 and the like can be cited.

Japanese Patent Gazette “Patent No. 4243684 (published on April 28, 2005)” Japanese Patent Publication “Japanese Laid-Open Patent Publication No. 2007-241867 (published September 26, 2007)” Japanese Patent Publication “Japanese Unexamined Patent Application Publication No. 2009-93440 (published April 30, 2009)”

As described above, in the conventional positioning device, the movement amount is estimated by exactly the same calculation process regardless of the mounting / holding posture of the measuring device. Then, the calculation process used for estimating the movement amount is a calculation process assuming a specific state. For this reason, when the mounting / holding posture is unexpected, there is a problem that the calculation of the movement amount cannot be performed or the accuracy of the calculated movement amount is lowered.

For example, when the moving body is a person, various states such as a state in which the measuring device is held by hand and a state in which it is in a pocket are assumed. And in each state, the kind and magnitude | size of the exercise | movement which a measuring device catches differ. For this reason, when the movement of the person holding the measuring device is captured, the type or setting of the optimum movement amount estimation means is also different.

Note that the above problem is not limited to a device that calculates the amount of movement, and is a problem that occurs in all devices that calculate parameters indicating the motion state of a moving body based on the detection result of a sensor.

The present invention has been made in view of the above-described problems, and an object of the present invention is to calculate a parameter indicating the motion state of the moving body by appropriate arithmetic processing according to the mounting / holding posture of the measuring apparatus with respect to the moving body. It is to provide a computing device or the like.

In order to solve the above-described problems, a calculation apparatus according to the present invention is a calculation apparatus that calculates a parameter indicating a moving state of a moving object using measurement data output from one or more sensors held by the moving object. Then, from the measurement data, holding state specifying means for specifying how the sensor is held by the moving body, and calculation processing according to the holding state specified by the holding state specifying means are performed, and the measurement is performed. And a parameter calculating means for calculating a parameter indicating the moving state of the moving body from the data.

In addition, in order to solve the above-described problem, the control method of the computing device of the present invention uses the measurement data output from one or more sensors held by the moving body to set a parameter indicating the moving state of the moving body. A control method of a computing device to calculate, according to a holding state specifying step for specifying how the sensor is held on a moving body from the measurement data, and a holding state specified in the holding state specifying step And a parameter calculating step of calculating a parameter indicating a moving state of the moving body from the measurement data.

According to the above configuration, since the parameter indicating the moving state of the moving body is calculated by the arithmetic processing according to how the sensor is held by the moving body, the reliability and accuracy of the parameter can be improved. .

Note that arithmetic processing is prepared in advance for each assumed holding state. The calculation processes corresponding to different holding states may be different or different in the calculation process (for example, mathematical formulas and procedures used). When the calculation process is the same, the parameter value corresponding to the holding state is calculated by weighting, switching of the scale factor, or the like.

Further, the sensor may be any sensor that outputs measurement data necessary for specifying the moving state of the moving body. For example, an acceleration sensor, a gyro sensor, a geomagnetic sensor, an atmospheric pressure sensor, and the like can be given.

Furthermore, in the above configuration, the parameter to be calculated is not particularly limited as long as it indicates the moving state of the moving body. Examples of the parameter include a moving direction, a moving speed, and a moving distance. Note that measurement data such as acceleration varies greatly depending on the holding state, and therefore the present invention is particularly suitable for calculating parameters using acceleration.

Further, in the above configuration, how the sensor is held (holding state) is determined based on the measurement data as to which of the holding states assumed in advance corresponds to none. Can do. For example, a holding state and a pattern of measurement data output in the holding state may be stored in association with each other.

As described above, the calculation apparatus according to the present invention responds to the holding state specifying means for specifying how the sensor is held by the moving body and the holding state specified by the holding state specifying means from the measurement data. It is a structure provided with the parameter calculation means which performs a calculation process and calculates the parameter which shows the movement state of the said mobile body from the said measurement data.

In addition, as described above, the control method of the computing device of the present invention is specified from the measurement data by the holding state specifying step for specifying how the sensor is held by the moving body and the holding state specifying step. And a parameter calculating step of performing a calculation process according to the holding state and calculating a parameter indicating the moving state of the moving body from the measurement data.

According to the above configuration, since the parameter indicating the moving state of the moving body is calculated by the arithmetic processing according to how the sensor is held by the moving body, the reliability and accuracy of the parameter can be improved. There is an effect.

Further objects, features, and superior points of the present invention will be fully understood from the following description. The benefits of the present invention will become apparent from the following description with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates an embodiment of the present invention, and is a block diagram illustrating a configuration of a main part of a positioning device. It is a figure which shows an example of the signal specific table used with the said positioning apparatus. It is a flowchart which shows an example of the process which the said positioning apparatus performs. It is a block diagram which shows the principal part structure of the positioning apparatus using an acceleration sensor as a measurement part. It is a block diagram which shows the principal part structure of a positioning apparatus provided with a posture angle estimation part. It is a block diagram which shows the principal part structure of a positioning apparatus provided with a gravity direction estimation part. It is a block diagram which shows the principal part structure of the positioning apparatus which calculates the movement vector of a pedestrian. It is a block diagram which shows the principal part structure of the positioning apparatus concerning other embodiment of this invention. It is a block diagram which shows the principal part structure of the highly accurate movement amount estimation part with which the said positioning apparatus is provided. It is a block diagram which shows the principal part structure of the simple movement amount estimation part with which the said positioning apparatus is provided. It is a block diagram which shows the principal part structure of the positioning apparatus which estimates a transition state. It is a block diagram which shows the principal part structure of the holding | maintenance state estimation part with which the said positioning apparatus is provided. It is a block diagram which shows the principal part structure of the simple movement amount estimation part with which the said positioning apparatus is provided. It is a flowchart which shows an example of the transition state detection process which the said positioning apparatus performs. It is a block diagram which shows the principal part structure of the positioning apparatus provided with two gravity direction estimation parts. It is a block diagram which shows the principal part structure of the holding | maintenance state estimation part with which the said positioning apparatus is provided. It is a block diagram which shows the principal part structure of the highly accurate movement amount estimation part to which a gravity direction vector is input. It is a block diagram which shows the principal part structure of the simple movement amount estimation part to which a gravity direction vector is input.

[Embodiment 1]
The positioning device of this embodiment will be described with reference to FIGS. Here, first, an outline of the positioning device will be described with reference to FIG. FIG. 1 is a block diagram showing a main configuration of a positioning device (calculation device) 1. As illustrated, the positioning device 1 includes a measurement unit (sensor) 2, a control unit 3, and a storage unit 4.

The positioning device 1 is a device that is attached to or held by a moving body (a person or an object) and outputs a parameter indicating the motion state of the moving body. Specifically, the positioning device 1 calculates a movement vector of the moving body using the measurement data measured by the measuring unit 2 which is a sensor that detects the movement and state change of the moving body. Calculate and output the current position of the moving object.

The measurement unit 2 is a sensor that detects the movement state of the moving body, the position where the moving body exists, and the like and outputs the detected value as measurement data as described above. Moreover, the measurement part 2 outputs elapsed time (DELTA) T after outputting measurement data last with measurement data. Thereby, the time-dependent change of measurement data is grasped.

The measuring unit 2 may be configured by a sensor corresponding to the type of parameter output by the positioning device 1, the accuracy of the required parameter, and the like. For example, the measurement unit 2 can be configured by any one or a combination of an acceleration sensor, a gyro sensor, a magnetic sensor, and an atmospheric pressure sensor.

The control unit 3 controls the positioning device 1 in an integrated manner, and performs control to calculate and output the current position of the moving body using the measurement data output from the measurement unit 2. This control is realized by a holding state estimation unit (holding state identification unit) 10, an adjustment unit 11, a movement amount estimation unit (parameter calculation unit) 12, and a position determination unit 13 provided in the control unit 3.

The holding state estimation unit 10 estimates the mounting / holding posture of the positioning device 1 (more precisely, the measurement unit 2) with respect to the moving body (how it is held by the moving body) from the measurement data output by the measuring unit 2. And output state data indicating the estimated posture.

Here, the mounting / holding posture is specified using data that associates a presumed mounting / holding posture with a pattern of measurement data detected by the mounting / holding posture, and the mounting / holding posture is determined. It is assumed that the status data corresponding to is output.

For example, the mounting / holding posture is specified by using a machine learning framework by AdaBoost, for example, based on measurement data (acceleration, angular velocity, magnetism, atmospheric pressure, etc.) from the measurement unit 2. It can also be realized by configuring a discriminator for identifying the holding posture. In addition, since identification of the mounting / holding posture using Adaboost is a known technique as described in Non-Patent Document 1, description thereof is omitted here.

Therefore, in the positioning device 1, the measuring unit 2 is configured so as to be able to detect parameters necessary for identifying a mounting / holding posture assumed in advance. Then, it identifies which mounting / holding posture corresponds to the measurement data pattern of the measuring unit 2 and outputs state data indicating the identified mounting / holding posture.

The adjustment unit 11 controls the movement amount estimation unit 12 so that a parameter indicating the movement state of the moving body is calculated by a calculation process according to the wearing / holding posture estimated by the holding state estimation unit 10. Specifically, the adjustment unit 11 specifies a control signal corresponding to the state data output from the holding state estimation unit 10 with reference to the signal specification table 20 stored in the storage unit 4, and uses this control signal. The above control is performed by outputting to the movement amount estimation unit 12. The signal specification table 20 will be described later.

The movement amount estimation unit 12 calculates a parameter indicating the movement state of the moving body by a calculation process according to the wearing / holding posture estimated by the holding state estimation unit 10. Specifically, the movement amount estimation unit 12 receives the measurement data output from the measurement unit 2 and the control signal output from the adjustment unit 11, and moves the moving object from the measurement data by the arithmetic processing indicated by the control signal. The movement vector indicating the movement direction and the movement distance is calculated and output.

The position determination unit 13 determines the current position of the moving object from the movement vector output by the movement amount estimation unit 12. The determined current position is output, for example, by displaying an image indicating the current position of the positioning device 1.

The storage unit 4 is a device that stores various data used in the positioning device 1. As shown in the figure, the signal specifying table 20 is stored in the storage unit 4.

The signal specification table 20 is a table for the adjustment unit 11 to specify a control signal corresponding to the state data output from the holding state estimation unit 10. The signal specifying table 20 may be as shown in FIG. FIG. 2 is a diagram illustrating an example of the signal identification table 20.

As shown in FIG. 2, the signal specification table 20 is data in which the holding state and the control signal are associated with each other. Specifically, in the signal identification table 20 of FIG. 2, the control signals (1), (2), (3),... Are associated with the holding

states

1, 2, 3,.

That is, in the signal identification table 20 of FIG. 2, the holding state estimation unit 10 determines that the holding state of the positioning device 1 is any one of 1, 2, 3,. It is assumed that the state corresponds to the state, and the state data indicating the specified holding state is output to the adjustment unit 11. Note that it may be determined from the measurement data output by the measurement unit 2 that it does not correspond to any of the holding states shown in FIG. In this case, status data indicating the closest holding state may be output, or status data indicating that it is not applicable may be output.

Then, the adjustment unit 11 refers to the signal identification table 20, and the control signal associated with the output state data corresponds to any of the control signals (1), (2), (3),. You will specify what to do. In the illustrated example, one control signal is associated with one holding state, but a plurality of control signals may be associated. Thereby, it is possible to finely control the arithmetic processing performed by the movement amount estimation unit 12.

In addition, as described above, when the holding state estimation unit 10 is configured to output state data indicating that it does not correspond to any holding state, a control signal corresponding to a state that does not correspond to any holding state Are preferably associated with each other. Thereby, even if it does not correspond to any holding state, it becomes possible to calculate the movement vector by an appropriate calculation process.

Then, the adjustment unit 11 outputs the control signal identified with reference to the signal identification table 20 of FIG. 2 to the movement amount estimation unit 12, and the movement amount estimation unit 12 performs arithmetic processing according to the output control signal. The movement vector is calculated by going.

Note that the arithmetic processing corresponding to each holding state may be specified based on the result of actually measuring what measurement data is detected in the holding state, for example. Further, for example, it is possible to specify an appropriate arithmetic processing according to the holding state by using a machine learning framework by AdaBoost.

As described above, the signal identification table 20 is a table in which a presumed holding state is associated with a control signal for causing the movement amount estimating unit 12 to execute a calculation process according to the holding state.

[Process flow]
Next, the flow of processing executed by the positioning device 1 will be described with reference to FIG. FIG. 3 is a flowchart illustrating an example of processing executed by the positioning device 1.

The holding state estimation unit 10 confirms whether or not measurement data is input from the measurement unit 2 (S1), and when the measurement data is input (YES in S1), the holding state of the positioning device 1 is determined from the measurement data. (Mounting / holding posture) is estimated (S2). Then, state data indicating the estimated holding state is output to the adjustment unit 11.

Next, the adjustment unit 11 that has received the state data refers to the signal identification table 20 to identify a control signal corresponding to the received state data, and generates the identified control signal (S3). The adjustment unit 11 outputs the generated control signal to the movement amount estimation unit 12.

Then, the movement amount estimation unit 12 that has received the control signal performs a calculation process according to the received control signal and calculates a movement vector (S4). Thereafter, the movement amount estimation unit 12 transmits the calculated movement vector to the position determination unit 13, and calculates and outputs the current position of the positioning device 1 based on the movement vector received by the position determination unit 13 that has received the movement vector. Then, the process ends.

[Example using acceleration sensor]
Next, an example in which an acceleration sensor is used as the measurement unit 2 will be described with reference to FIG. FIG. 4 is a block diagram showing a main configuration of the positioning device 1 using an acceleration sensor as the measuring unit 2. The illustrated positioning apparatus 1 includes an acceleration sensor 2a as a measuring unit. The acceleration sensor 2a is a triaxial acceleration sensor that detects acceleration vectors in three axial directions perpendicular to each other and outputs the acceleration vectors as acceleration data.

It is known that a holding state can be specified and a movement vector can be calculated by using such a three-axis acceleration sensor. For example, Non-Patent Document 2 and Patent Document 2 describe whether a person is walking, running, or going up and down stairs from 3-axis acceleration data output by a 3-axis acceleration sensor held by a person. Is described.

In this technology, feature vectors are extracted from triaxial acceleration data using wavelet packet transformation, and each state (walking, running, etc.) is clustered using a self-organization method described in Non-Patent Document 3 or the like. Yes.

By applying such technology, the holding state can be estimated from the triaxial acceleration data. That is, the holding state estimation unit 10 in FIG. 4 determines which of holding states clustered in advance by a self-organization method from feature vectors extracted from time-series triaxial acceleration data using wavelet packet transformation. Identify.

Further, the movement amount estimation unit 12 in FIG. 4 calculates a movement vector by performing arithmetic processing according to the control signal received from the adjustment unit 11 using the triaxial acceleration data. In addition, since the calculation process which calculates a movement vector from triaxial acceleration data is well-known, description is abbreviate | omitted here.

[Example with posture angle estimation unit]
By calculating the attitude angle of the moving body with respect to the world coordinate system, it is possible to determine the gravity direction of the moving body (usually different from the gravity direction of the world coordinate system). can do.

Here, a positioning device including an attitude angle estimation unit that calculates an attitude angle will be described with reference to FIG. FIG. 5 is a block diagram illustrating a main configuration of a positioning device (calculation device) 30 including a posture angle estimation unit. In addition, about the structure similar to the positioning apparatus 1 of FIG. 1, FIG. 4, the same reference number is attached | subjected and the description is abbreviate | omitted.

As illustrated, the positioning device 30 includes an acceleration sensor 2a and a gyro sensor 2b as measurement units. The gyro sensor 2b detects an angular velocity vector and outputs angular velocity data. Here, it is assumed that the gyro sensor 2b outputs angular velocity data in the triaxial direction, but it may be biaxial or less.

The posture angle estimation unit (posture angle calculation means) 14 calculates the posture angle of the moving body holding the positioning device 30 using the acceleration data output from the acceleration sensor 2a and the angular velocity data output from the gyro sensor 2b. To do. The posture angle estimation unit 14 outputs the calculated posture angle to the holding state estimation unit 10 and the movement amount estimation unit 12.

By acquiring the posture angle as time-series data, it is possible to measure the change of the posture angle in addition to acceleration, deceleration, movement of the center of gravity, etc., regarding the movement created by the moving body. As a result, the state of the moving body can be estimated more accurately and finely. That is, by providing the posture angle estimation unit 14, it is possible to increase the estimation accuracy of the holding state, increase the variation of the holding state that can be detected, and the like.

Also, the movement amount estimation unit 12 specifies the movement direction of the moving body using the posture angle output from the posture angle estimation unit 14, and calculates a movement vector based on this. In addition, since the method of calculating | requiring a moving direction from an attitude angle is well-known, description is abbreviate | omitted here.

[Example with gravity direction estimation unit]
The motion state of the moving body can be specified with higher accuracy also by calculating the gravity direction of the moving body. Note that, as described above, the gravity direction of the moving body is usually different from the gravity direction of the world coordinate system.

Here, a positioning device including a gravity direction estimation unit for calculating the gravity direction will be described with reference to FIG. FIG. 6 is a block diagram illustrating a main configuration of a positioning device (calculation device) 40 including a gravity direction estimation unit. In addition, about the structure similar to the positioning apparatus 1 of FIG. 1, FIG. 4, the same reference number is attached | subjected and the description is abbreviate | omitted.

The gravity direction estimation unit (gravity direction identification means) 15 calculates the gravity direction vector of the positioning device 40 (more precisely, the measurement unit 2) using the measurement data output from the measurement unit 2. Further, the gravity direction estimation unit 15 outputs the calculated gravity direction vector to the holding state estimation unit 10 and the movement amount estimation unit 12.

The gravity direction vector can be specified by tracking the gravity acceleration vector using the acceleration vector and the angular velocity vector. For this reason, the positioning device 40 needs to include at least an acceleration sensor and a gyro sensor as the measurement unit 2. Non-Patent Document 2 describes that a gravitational acceleration vector is tracked using an input of an acceleration vector and an angular velocity vector using a Kalman filter framework. The gravity direction estimation unit 15 can also be realized by such a technique.

Since the gravity direction vector represents the attitude angle with respect to gravity, it is a great hint for estimating the holding state. That is, the holding state estimation unit 10 can estimate the holding state with high accuracy by using the gravity direction vector output by the gravity direction estimation unit 15, and increase the number of detectable holding states. It becomes possible.

Further, the movement amount estimation unit 12 can decompose the measurement data (acceleration data, angular velocity data, etc.) into the gravity direction component and the other by using the gravity direction vector output from the gravity direction estimation unit 15. The estimation accuracy of the movement vector can be improved.

[Configuration example when the moving body is a pedestrian]
Next, an example of a positioning device configured on the assumption that the moving body is a pedestrian will be described with reference to FIG. FIG. 7 is a block diagram showing a main configuration of a positioning device that calculates a pedestrian movement vector. In addition, about the structure similar to the positioning apparatus 1 of FIG. 1, FIG. 4, the same reference number is attached | subjected and the description is abbreviate | omitted.

As illustrated, the positioning device 50 (calculation device) includes a walking motion detection unit (parameter calculation unit) 12a, a stride estimation unit (parameter calculation unit) 12b, a movement direction detection unit 12c, and a movement vector as a movement amount estimation unit. A calculation unit 12d is provided. Among these, the adjustment part 11 transmits a control signal to the walking motion detection part 12a and the stride estimation part 12b, and performs the arithmetic processing which exhibits the highest performance according to the holding | maintenance state.

Here, it is assumed that the measurement unit 2 includes a three-axis acceleration sensor, a three-axis angular velocity sensor, and a three-axis geomagnetic sensor. Here, it is assumed that the movement vector of the pedestrian is calculated and the current position of the pedestrian is output.

The walking motion detection unit 12a identifies whether or not a walking motion is being performed from the measurement data output by the measurement unit 2 by a calculation process according to the control signal received from the adjustment unit 11. That is, the walking motion detection unit 12a can detect the walking motion by a plurality of arithmetic processes, and detects the walking motion by the arithmetic process specified by the control signal among these arithmetic processes.

Specifically, the walking motion detection unit 12a compares the pattern of change with time of the acceleration data output from the measurement unit 2 with the pattern of change with time of the acceleration data generated when the walking motion is performed. Identifies that a walking action is taking place. In addition, the pattern of change with time of the acceleration data is a pattern in which the acceleration changes with a constant period and amplitude, and a technique for calculating the walking speed from this amplitude is known (for example, Patent Document 3). For this reason, the walking motion detection unit 12a specifies the amplitude of the pattern and notifies the stride estimation unit 12b of the specified amplitude. Note that the walking motion detection method, the movement direction detection method, and the movement distance calculation method are not limited to the above examples.

The stride length estimation unit 12b calculates a walking speed from the amplitude output by the walking motion detection unit 12a by a calculation process according to the control signal received from the adjustment unit 11. And the stride (distance walked at the elapsed time) is calculated by multiplying the calculated walking speed by the elapsed time.

Here, it is empirically known that the measurement data output from the measurement unit 2 such as acceleration data and angular velocity data has its signal intensity increased or decreased depending on the holding state. The signal intensity can also be expressed as an amplitude in a waveform (sine curve or the like) drawn by plotting measurement data in time series, for example. Patent Document 1 describes that whether or not a walking motion is performed is determined based on whether or not such signal intensity exceeds a predetermined threshold.

For this reason, the control signal transmitted to the walking motion detector 12a may detect the walking motion with a threshold corresponding to the holding state. In other words, a holding signal with a low signal strength is associated with a control signal that detects a walking motion with a small threshold, and a holding signal with a high signal strength is associated with a control signal that detects a walking motion with a large threshold. By canceling the influence of the state on the signal intensity, it is possible to reliably detect the walking motion. Moreover, as a control signal transmitted to the walking motion detection unit 12a, for example, a walking motion may be detected with a sensitivity (scale factor) corresponding to the holding state. Also with this configuration, it is possible to cancel the influence of the holding state on the signal intensity and reliably detect the walking motion.

And as a control signal transmitted to the stride length estimation unit 12b, a signal or the like for converting measurement data into amplitude with a sensitivity (scale factor) according to the holding state is assumed. Thereby, even when the value of the measurement data to be output changes due to the influence of the holding state, it is possible to calculate an accurate walking speed.

The moving direction detection unit 12c detects the direction in which the pedestrian is moving using the measurement data. Note that a known method can be applied to the detection of the moving direction, and for example, it can be detected from acceleration data and angular velocity data.

The movement vector calculation unit 12d calculates a movement vector by multiplying the stride (walking distance) output from the step estimation unit 12b by the direction output from the movement direction detection unit 12c, and outputs the calculated movement vector to the position determination unit 13. . Thereby, the position determination part 13 can pinpoint the present position of the pedestrian who has the positioning apparatus 50. FIG.

[Embodiment 2]
Next, another embodiment of the present invention will be described with reference to FIGS. The positioning device of this embodiment has two movement amount estimation units with different estimation accuracy of the movement amount, and the main point is that these movement amount estimation units are selectively used according to the posture estimated by the holding posture estimation unit 10. It is a special feature point. In addition, about the structure similar to the said embodiment, the same reference number is attached | subjected and the description is abbreviate | omitted.

[Main components of positioning device]
First, the configuration of the positioning device of the present embodiment will be described with reference to FIG. FIG. 8 is a block diagram showing a main configuration of the positioning device (calculation device) 60. As illustrated, the positioning device 60 includes a control unit 3 and a storage unit 4. Moreover, the acceleration sensor 2a, the gyro sensor 2b, and the geomagnetic sensor 2c are provided as a measurement part. That is, the positioning device 60 estimates the movement vector using the acceleration, angular velocity, and geomagnetism as input data.

Furthermore, the positioning device 60 includes a display unit 5. The display unit 5 is a device that displays an image according to the control of the control unit 3. Here, it is assumed that the positioning device 60 has a function of displaying the current position of the user who owns the positioning device 60 on a map using the estimated movement vector. Therefore, the display unit 5 displays a map image, information indicating the current position of the user, and the like.

The control unit 3 includes a holding state estimation unit 10, an adjustment unit 11, a position determination unit (parameter calculation means) 13, and a display control unit 18. In addition, a high-precision movement amount estimation unit 16 and a simple movement amount estimation unit (simple parameter calculation means) 17 are provided as the movement amount estimation unit.

The holding state estimation unit 10 estimates the holding state of the positioning device 60 using the measurement data output from the measurement unit, generates state data indicating the estimated holding state, and outputs the state data to the adjustment unit 11. The holding state estimation unit 10 outputs state data indicating that the positioning device 60 is in an unexpected state when the input measurement data does not correspond to any of the holding states assumed in advance.

Note that the holding state is not limited to the above example, and can be easily applied as long as it can be identified using a machine learning framework such as AdaBoost. Further, it is not always necessary to use a machine learning framework, and any holding state that can be identified based on measurement data may be used. However, when adding / changing the holding state, the adjusting unit 11, the high-precision moving amount estimating unit 16, and the simple moving amount estimating unit are configured so that an optimal movement vector according to the holding state is estimated. There is a need to.

The adjustment unit 11 controls the movement amount estimation unit 12 so that a parameter indicating the motion state of the moving body is calculated by a calculation process according to the state data output from the holding state estimation unit 10. Specifically, the adjustment unit 11 specifies a control signal corresponding to the state data using the signal specification table 20 of the storage unit 4, and uses the specified control signal for the high-precision movement amount estimation unit 16 and the simple movement amount estimation unit. The above control is performed by outputting to 17.

The high-accuracy movement amount estimation unit 16 has a configuration corresponding to the movement amount estimation unit 12 in FIG. 1 and estimates a movement vector from measurement data based on a control signal output from the adjustment unit 11. More specifically, the high-precision movement amount estimation unit 16 estimates a pedestrian movement vector.

Here, a more detailed configuration of the high-precision movement amount estimation unit 16 will be described with reference to FIG. FIG. 9 is a block diagram showing a main configuration of the high-precision movement amount estimation unit 16. As shown in the figure, the high-precision movement amount estimation unit 16 includes a walking motion detection unit (parameter calculation unit) 16a, a stride estimation unit (parameter calculation unit) 16b, a movement direction detection unit (parameter calculation unit) 16c, and a movement vector calculation. Unit (parameter calculation means) 16d. These have the same functions as the walking motion detection unit 12a, the stride estimation unit 12b, the movement direction detection unit 12c, and the movement vector calculation unit 12d in FIG.

Specifically, the walking motion detection unit 16a uses the triaxial acceleration data output from the acceleration sensor 2a and the triaxial angular velocity data output from the gyro sensor 2b in accordance with the control signal received from the adjustment unit 11. Perform arithmetic processing. Thereby, the walking motion detection unit 16a detects the walking motion, calculates the amplitude, and outputs the amplitude to the stride estimation unit 16b.

The stride length estimation unit 16b calculates the walking speed from the amplitude output by the walking motion detection unit 16a by a calculation process according to the control signal received from the adjustment unit 11. Then, the stride (movement distance) is calculated from the calculated walking speed, and is output to the movement vector calculation unit 16d.

The moving direction detection unit 16c performs arithmetic processing according to the control signal received from the adjustment unit 11, using the triaxial angular velocity data output from the gyro sensor 2b and the triaxial geomagnetic data output from the geomagnetic sensor 2c. The moving direction of the positioning device 60 is specified. Thus, the detection of the moving direction may also be performed by a calculation process according to the control signal received from the adjustment unit 11. For example, the contribution of the measurement data output by the measurement unit in the holding state where the reliability of the output measurement data is considered to be low is reduced or excluded from the measurement data output by the measurement unit in the other holding state The azimuth may be specified by the arithmetic processing.

The movement vector calculation unit 16d calculates a movement vector (referred to as a movement vector (high accuracy)) by multiplying the stride (walking distance) output from the stride estimation unit 16b by the direction output from the movement direction detection unit 16c. The movement vector (high accuracy) is output to the position determination unit 13.

Similar to the high-precision movement amount estimation unit 16, the simple movement amount estimation unit 17 performs a calculation process according to the control signal transmitted by the adjustment unit 11, thereby moving the movement vector (movement vector) from the measurement data output by the measurement unit. (Referred to as “simple”) is calculated, and the calculated movement vector (simple) is output to the position determining unit 13.

However, the simple movement amount estimation unit 17 is different from the high accuracy movement amount estimation unit 16 in that the amount of data used for the arithmetic processing is smaller than that of the high accuracy movement amount estimation unit 16. Specifically, the simple movement amount estimation unit 17 performs calculation processing using all of the three-axis acceleration data, the three-axis angular velocity data, and the three-axis geomagnetic data, while the high-precision movement amount estimation unit 16 performs the calculation process. Axial acceleration data and triaxial geomagnetic data are used, and triaxial angular velocity data is not used.

In the case of an unexpected holding state, measurement data having a value that is not assumed by the high-precision movement amount estimation unit 16 can be input, so that the movement vector ( The movement vector (simple) output from the simple movement amount estimation unit 17 that performs arithmetic processing with a small amount of data may be more appropriate than (high accuracy). For this reason, when it is in an unexpected holding state or the like, the calculation accuracy of the movement vector can be increased by performing the calculation process by the simple movement amount estimation unit 17.

Here, a more detailed configuration of the simple movement amount estimation unit 17 will be described with reference to FIG. FIG. 10 is a block diagram illustrating a main configuration of the simple movement amount estimation unit 17. As illustrated, the simple movement amount estimation unit 17 includes a walking motion detection unit (simple parameter calculation unit) 17a, a stride estimation unit (simple parameter calculation unit) 17b, a movement direction detection unit (simple parameter calculation unit) 17c, and a movement. A vector calculation unit (simple parameter calculation means) 17d is provided. These have the same functions as the walking motion detection unit 16a, the stride length estimation unit 16b, the movement direction detection unit 16c, and the movement vector calculation unit 16d of the high-precision movement amount estimation unit 16.

Specifically, the walking motion detection unit 17a performs arithmetic processing according to the control signal received from the adjustment unit 11 using the triaxial acceleration data output from the acceleration sensor 2a. Thereby, the walking motion detection unit 17a detects the walking motion, calculates the amplitude, and outputs the amplitude to the stride estimation unit 17b. That is, the walking motion detection unit 17a is different from the walking motion detection unit 16a in that the triaxial angular velocity data is not used for the calculation process.

The stride length estimation unit 17b calculates a walking speed from the amplitude output by the walking motion detection unit 17a by a calculation process according to the control signal received from the adjustment unit 11. Then, the stride (movement distance) is calculated from the calculated walking speed, and is output to the movement vector calculation unit 17d.

The moving azimuth detecting unit 17c performs arithmetic processing according to the control signal received from the adjusting unit 11, and specifies the moving azimuth of the positioning device 60 from the three-axis geomagnetic data output from the geomagnetic sensor 2c. That is, the moving direction detection unit 17c is different from the moving direction detection unit 16c in that the direction is specified without using the triaxial angular velocity data.

The movement vector calculation unit 17d calculates a movement vector (simple) by multiplying the stride (walking distance) output by the stride estimation unit 17b by the azimuth output by the movement azimuth detection unit 17c, and sets the calculated movement vector (simple) as the position. The data is output to the determination unit 13.

The simple movement amount estimation unit 17 is not limited to the above example as long as the amount of data used for the arithmetic processing is smaller than that of the high-precision movement amount estimation unit 16. For example, when the high-accuracy movement amount estimation unit 16 uses triaxial data, only the biaxial or uniaxial data may be used. For example, the amount of data used may be reduced while using the same kind of data, for example, by making the period for acquiring data from the measurement unit longer than that of the high-precision movement amount estimation unit 16.

In the positioning device 60, both the movement vector (high accuracy) output from the high-accuracy movement amount estimation unit 16 and the movement vector (simple) output from the simple movement amount estimation unit 17 are input to the position determination unit 13. Then, the position determination unit 13 determines whether to use the movement vector (high accuracy), the movement vector (simple), or both based on the control signal transmitted by the adjustment unit (switching unit) 11. Based on this determination, position data is calculated.

When both the movement vector (high accuracy) and the movement vector (simple) are used, the movement vector is calculated by a simple average or a weighted average of the movement vector (high accuracy) and the movement vector (simple). Even if the calculation process by the high-accuracy movement amount estimation unit 16 is performed, in the holding posture where it is assumed that a calculation result with high reliability cannot be obtained, the movement vector (simple) is replaced with the movement vector (simple). By taking into account partial elements of (high accuracy), the estimation accuracy of the movement vector can be increased.

When the posture data output from the holding posture estimation unit 10 indicates any of the holding postures assumed in advance, the adjustment unit 11 moves the movement vector (high accuracy) or the movement vector (high accuracy) and the movement vector ( A simple control signal is output. On the other hand, if the posture data output by the holding posture estimation unit 10 does not correspond to any of the holding postures assumed in advance, a control signal for outputting a movement vector (simple) is transmitted. Then, the position determination unit 13 calculates the current position using a movement vector corresponding to the control signal from the adjustment unit 11.

The display control unit 18 performs control to display an image on the display unit 5. Specifically, the display control unit 18 displays a map image on the display unit 5 based on the map data 21 stored in the storage unit 4. Further, the display control unit 18 displays a mark indicating that the user exists at the position indicated by the position data received from the position determination unit 13 on the displayed map.

[Example of detecting transition status]
Here, for example, when the operation of changing the positioning device is performed, the acceleration or the like generated by this operation is detected by the measurement unit 2. Thus, there is a possibility that the moving direction and moving distance of the user may not be accurately calculated from the measurement data detected when the user holding the positioning device changes the positioning device.

Therefore, here, an example in which a transition state that is a state from a state in which a holding state is changed to another state is estimated will be described with reference to FIG. FIG. 11 is a block diagram illustrating a main configuration of a positioning device (calculation device) 70 that estimates a transition state. In addition, about the structure similar to the positioning apparatus 60 of FIG. 8, the same reference number is attached | subjected and the description is abbreviate | omitted.

As shown in the figure, the positioning device 70 is configured such that the drive control signal is input from the holding state estimation unit 10 to the measurement unit, and the measurement data input from the measurement unit to the simple movement amount estimation unit 17 is a biaxial angular velocity. It differs from the positioning device 60 in that it is data, triaxial acceleration data, and triaxial geomagnetic data.

First, details of the holding state estimation unit 10 will be described with reference to FIG. FIG. 12 is a block diagram illustrating a main configuration of the holding state estimation unit 10. As shown in the figure, the holding state estimation unit 10 includes a holding state identification unit 10a, a transition state detection unit 10b, and a holding state determination unit 10c.

The holding state identifying unit 10a identifies the holding state of the positioning device 70 from the measurement data output by the measuring unit, and outputs state identification data indicating the identified holding state to the holding state determining unit 10c. Specifically, the holding state identification unit 10a corresponds to any one of the holding states assumed in advance using the triaxial acceleration data, the triaxial angular velocity data, and the triaxial geomagnetic data. State identification data indicating the identification result is output.

The transition state detection unit 10b performs a transition state detection process to detect that the positioning device 70 is in the transition state from the measurement data output by the measurement unit. And when it detects that it exists in a transition state, that is output to the holding | maintenance state determination part 10c. For example, when it is detected that the holding posture has changed significantly in a short time, or when it is detected that the detected holding posture is unstable, it is estimated that the state is a transition state. By detecting a simple state, a transition state can be detected. In addition, since the direction detected at the time of transition changes abruptly, it is possible to detect the transition state using the angular velocity data and the geomagnetic data used for specifying the direction. The details of the transition state detection process executed by the transition state detection unit 10b will be described later.

The holding state determination unit (measurement control means) 10c determines the holding state of the positioning device 70 based on the outputs of the holding state identification unit 10a and the transition state detection unit 10b. Then, state data indicating the determined holding state is generated and output to the adjustment unit 11.

In addition, when the transition state detection unit 10b outputs that the transition state detection unit 10b is in the transition state, the holding state determination unit 10c outputs a drive control signal to the measurement unit 2 and stops some data measurement by the measurement unit 2. The output destination of the measurement data is switched from the high-precision movement amount estimation unit 16 to the simple movement amount estimation unit 17. Note that the drive control by the holding state determination unit 10c is not limited to the above example, and is used by the simple movement amount estimation unit 17 by, for example, lowering the drive frequency of the measurement unit 2, in other words, lowering the output frequency of measurement data. The amount of data to be performed may be smaller than that of the high-precision movement amount estimation unit 16.

That is, the positioning device 70 does not perform the movement vector calculation by the high-precision movement amount estimation unit 16 in the transition state, and only performs the movement vector calculation by the simple movement amount estimation unit 17 while reducing the data amount output by the measurement unit 2. . Thereby, the power consumption in the measurement part 2 can be reduced.

Here, as shown in FIG. 11, the high-accuracy movement amount estimation unit 16 receives triaxial acceleration data, triaxial angular velocity data, and triaxial geomagnetic data, whereas the simple movement amount estimation unit. 17, it is assumed that triaxial acceleration data, biaxial angular velocity data, and triaxial geomagnetic data are input.

That is, when the transition state detection unit 10b outputs that the transition state detection unit 10b is in the transition state, the holding state determination unit 10c outputs a drive control signal to the measurement unit 2 to change the angular velocity measured by the gyro sensor 2b from three axes to two axes. Switch to. Then, the block for calculating the movement vector is switched from the high-precision movement amount estimation unit 16 to the simple movement amount estimation unit 17.

Also, the holding state determination unit 10c performs the same processing as that in the transition state even when the state identification data indicating that it does not correspond to any of the holding postures assumed in advance is received from the holding state identification unit 10a. . This is because when the holding posture assumed in advance does not correspond, even if the high-accuracy movement amount estimation unit 16 is used, improvement in calculation accuracy cannot be expected.

The simple movement amount estimation unit 17 of the positioning device 70 will be described with reference to FIG. FIG. 13 is a block diagram illustrating a main configuration of the simple movement amount estimation unit 17. As shown in FIG. 13, the walking motion detector 17a calculates the amplitude from the triaxial acceleration data and the biaxial angular velocity data, and the stride estimation unit 17b calculates the stride from this amplitude. Further, the moving direction detection unit 17c specifies the direction from the triaxial geomagnetic data and the biaxial angular velocity data, and the movement vector calculation unit 17d calculates a movement vector (simple) from the stride and the direction. Thus, the simple movement amount estimation unit 17 uses less angular velocity data (from three axes to two axes) than the high-precision movement amount estimation unit 16 shown in FIG.

Note that the walking motion detection unit 17a, the stride length estimation unit 17b, the movement direction detection unit 17c, and the movement vector calculation unit 17d are output by the adjustment unit 11 according to a calculation process corresponding to the control signal corresponding to the transition state, The stride, azimuth, and movement vector (simple) are calculated.

More specifically, in the period when the holding state determination unit 10c determines that the holding state determination unit 10c is in the transition state, or the period when the adjustment unit 11 determines that it does not correspond to any of the holding postures assumed in advance, the orientation and position are changed. Control is performed so as not to change, or to make the change smaller than periods other than those described above. Accordingly, it is possible to prevent the current position from being estimated to be a position greatly deviated from the actual position by the measurement data output in the transition state or the period of the unexpected holding posture.

For example, the walking motion detection unit 17a detects a walking motion using time-series acceleration data. For this reason, the adjustment unit 11 transmits a control signal to the walking motion detection unit 17a, and contributes the acceleration data output by the measurement unit during a period that does not correspond to either the transition state or the holding posture assumed in advance. The walking motion is detected by a calculation process that is smaller than the acceleration data output by the measurement unit during other periods. Of course, the contribution of the acceleration data output by the measurement unit during a period that does not correspond to any of the transition state or the presumed holding posture may be completely removed.

Similarly, the adjustment unit 11 also contributes the acceleration data output by the measurement unit during a period other than the transition state or the presumed holding posture to the stride length estimation unit 17b. The step length may be detected by a calculation process that is smaller than the acceleration data output by the measuring unit or is excluded.

Further, the adjustment unit 11 similarly applies the measurement data (angular velocity data and geomagnetic data) output by the measurement unit during the period in which the moving direction detection unit 17c does not correspond to either the transition state or the presumed holding posture. May be made smaller than the acceleration data output by the measurement unit during other periods, or the orientation may be specified by a calculation process that is excluded.

Note that since the azimuth error is particularly affected by the positioning error, only the moving azimuth may be controlled by the adjustment unit 11. That is, if the moving direction is estimated incorrectly, a position that is far from the actual position is estimated. Therefore, it is preferable to perform control so that at least the moving direction is not changed in the transition state or the like.

Also, it is desirable to adjust the moving direction and the current position in the transition state or the like with reference to the moving direction and the position at the time before or after the time at which the moving direction and the position are to be obtained. Of course, it is still more preferable to adjust with reference to the moving azimuth | direction and position in the time around.

That is, the moving direction detection unit 17c may store a change history of the direction, and specify the most likely direction from the change history as the current direction. Similarly, the position determination unit 13 may store a position change history, and specify the most likely position as the current position from the change history. As a result, the noise component of the posture identification information (various data used for identifying the holding state) is reduced, and stable posture identification becomes possible.

[Transition state detection process flow]
Next, the flow of the transition state detection process performed by the transition state detection unit 10b will be described based on FIG. FIG. 14 is a flowchart illustrating an example of the transition state detection process. In the illustrated process, a transition state is specified by detecting a pattern in which the acceleration changes abruptly based on the triaxial acceleration data using the fact that the acceleration changes abruptly at the time of transition. Note that the data used for specifying the transition state may not be the measurement data itself output by the measurement unit 2. For example, it is naturally possible to use data obtained by filtering measurement data with a low-pass filter or the like.

First, the transition state detection unit 10b acquires triaxial acceleration data from the acceleration sensor 2a (S10). And the transition state detection part 10b calculates the difference with the acceleration of 1 step before about each axis | shaft (S11). That is, the transition state detection unit 10b acquires triaxial acceleration data at regular time intervals and stores the values. When acceleration data is newly acquired, the difference between the acceleration data and the last stored acceleration data is calculated.

Next, the transition state detection unit 10b calculates a difference average value (dif_a_ave) for a certain period for each axis using the difference calculated in S11 (S12). Further, the transition state detection unit 10b sets a threshold value (DFTH) for determining whether or not the transition state is set (S13).

As the threshold value (DFTH), a predetermined value may be used. However, since the magnitude of the acceleration detected in the transition state is proportional to the magnitude of the walking speed at that time, it is possible to set a threshold (DFTH) according to the walking speed or acceleration from a predetermined time before to the present. preferable. That is, the transition state detection unit 10b sets a threshold value (DFTH) proportional to the walking speed or acceleration.

In addition, the value of the measurement data output from the measurement unit 2 increases as the walking speed increases. For this reason, it may replace with said structure and may employ | adopt the structure which sets a big threshold value, so that the value of measurement data becomes large. Further, the threshold value to be set may be changed continuously according to the value of the measurement data (for example, the average value of the values in the three axis directions) or may be changed stepwise.

Subsequently, the transition state detection unit 10b checks whether or not the difference average value (dif_a_ave) calculated in S12 is smaller than the threshold value (DFTH) set in S13 (S14). When the transition state detection unit 10b determines that the difference average value (dif_a_ave) is smaller than the threshold value (DFTH) (YES in S14), the transition state detection unit 10b determines that it is not the transition state (S15), and holds that state. It outputs to the determination part 10c and returns to the process of S10.

On the other hand, if the transition state detection unit 10b determines that the difference average value (dif_a_ave) is equal to or greater than the threshold value (DFTH) (NO in S14), the transition state detection unit 10b determines that the transition state is present (S15) and retains that effect. It outputs to the state determination part 10c, and returns to the process of S10.

[Example of multiple gravity direction estimation units]
Next, an example in which a plurality of gravity direction estimation units are provided will be described with reference to FIG. FIG. 15 is a block diagram illustrating a main configuration of a positioning device 80 including two gravity direction estimation units. In addition, about the structure similar to the positioning apparatus 60 of FIG. 8, the same reference number is attached | subjected and the description is abbreviate | omitted.

As shown in the figure, the positioning device 80 includes a high-precision gravity direction estimation unit 15a and a simple gravity direction estimation unit (simple gravity direction specifying means) 15b as gravity direction estimation units.

The high precision gravity direction estimation unit 15a has the same function as the gravity direction estimation unit 15 provided in the positioning device 40 of FIG. That is, the high-precision gravity azimuth estimation unit 15a calculates the gravity azimuth vector using the triaxial acceleration data output from the acceleration sensor 2a and the triaxial angular velocity data output from the gyro sensor 2b. Here, the gravity direction vector calculated by the high-precision gravity direction estimation unit 15a is referred to as a gravity direction vector (high accuracy). Further, the high-precision gravity azimuth estimation unit 15 a outputs the calculated gravity azimuth vector (high accuracy) to the holding state estimation unit 10 and the high-precision movement amount estimation unit 16. The gravity direction vector (high accuracy) may also be output to the simple movement amount estimation unit 17.

The simple gravity azimuth estimation unit 15b calculates the gravity azimuth vector using the acceleration data and the angular velocity data in the same manner as the high precision gravity azimuth estimation unit 15a, but the amount of data used is larger than that of the high precision gravity azimuth estimation unit 15a. There are few differences. In the illustrated example, the simple gravity direction estimation unit 15b calculates the gravity direction vector using the biaxial angular velocity data and the triaxial acceleration data. Here, the gravity direction vector calculated by the simple gravity direction estimation unit 15b is referred to as a gravity direction vector (simple). Further, the simple gravity direction estimation unit 15 b outputs the calculated gravity direction vector (simple) to the holding state estimation unit 10 and the simple movement amount estimation unit 17. Note that the gravity direction vector (simple) may also be output to the high-precision movement amount estimation unit 16.

The holding state estimation unit 10 of the positioning device 80 estimates the holding state using the gravity direction vector (high accuracy) or the gravity direction vector (simple) output as described above. Depending on the holding state, measurement data having a value that is not assumed by the high-precision gravity azimuth estimation unit 15a is input, and therefore, less than the gravity azimuth vector (high accuracy) output by the high-precision gravity azimuth estimation unit 15a. The gravity direction vector (simple) output from the simple gravity direction estimation unit 15b that performs arithmetic processing with the amount of data may be a more appropriate value. For this reason, the estimation accuracy of the gravity direction vector can be improved by performing the calculation process by the simple gravity direction estimation unit 15b in a transition state or an unexpected holding state.

The use of the gravity direction vector in the holding state estimation unit 10 will be described with reference to FIG. FIG. 16 is a block diagram illustrating a main configuration of the holding state estimation unit 10. As shown in the drawing, in addition to the triaxial acceleration data, the triaxial angular velocity data, and the triaxial geomagnetic data, the gravity direction is input to the holding state identifying unit 10a and the transition state detecting unit 10b. This gravity direction is a gravity direction vector (high accuracy) and a gravity direction vector (simple).

By using the gravity azimuth vector (high accuracy) or the gravity azimuth vector (simple), the traveling direction component of the positioning device 80 and the direction perpendicular to the three-axis acceleration data, three-axis angular velocity data, and three-axis geomagnetic data are obtained. It can be decomposed into components, whereby the holding state can be identified with high accuracy.

The same applies to the detection of the transition state. That is, since the acceleration changes abruptly at the time of transition, it is possible to detect a transitional state by detecting a rapidly changing pattern based on the vertical / traveling direction acceleration. Further, for example, the transition state may be detected when the traveling direction has changed abruptly. In addition, although advancing direction changes also when a mobile body (pedestrian) changes direction, since the degree of the change is larger in the transition state, it is possible to distinguish the direction change and the transition state.

In addition, the high-precision movement amount estimation unit 16 and the simple movement amount estimation unit 17 of the positioning device 80 also use the gravity direction vector (high accuracy) or the gravity direction vector (simple) output as described above. Estimate.

This will be described with reference to FIG. FIG. 17 is a block diagram showing a main configuration of the high-precision movement amount estimation unit 16 to which a gravity direction vector is input. As illustrated, in addition to the triaxial acceleration data, the gravity direction is input to the walking motion detector 16a. In addition to the triaxial geomagnetic data, the gravity direction is input to the moving direction detector 16c. These gravity directions are gravity direction vectors (high accuracy).

As described above, the high-precision movement amount estimation unit 16 of the positioning device 80 detects the walking motion, specifies the movement direction, and the like using the gravity direction vector (high accuracy), and thus estimates the movement vector with high accuracy. be able to.

Similarly, the simple movement amount estimation unit 17 detects the walking motion, specifies the movement direction, and the like using the gravity direction vector (simple). This will be described with reference to FIG. FIG. 18 is a block diagram illustrating a main configuration of the simple movement amount estimation unit 17 to which a gravity direction vector is input.

As illustrated, the simple movement amount estimation unit 17 of the positioning device 80 includes a walking motion detection unit 17a, a stride estimation unit 17b, a movement direction detection unit 17c, and a movement vector calculation unit 17d, as in FIG. .

In addition to the triaxial acceleration data, the gravity direction vector (simple) is input to the walking motion detection unit 17a, and the gravity direction vector (simple) is input to the movement direction detection unit 17c in addition to the triaxial geomagnetic data. The

As described above, the simple movement amount estimation unit 17 of the positioning device 80 uses the gravity direction vector (simple) to detect walking motion, specify the movement direction, and the like.

As described above, whether the reliability of the gravity direction vector (high accuracy) or the reliability of the gravity direction vector (simple) is high depends on the holding state. For this reason, you may output both a gravity direction vector (high precision) and a gravity direction vector (simple) to the high precision movement amount estimation part 16 and the simple movement amount estimation part 17. FIG. Then, the high-precision movement amount estimation unit 16 and the simple movement amount estimation unit 17 determine which one of the gravity direction vector (high accuracy) and the gravity direction vector (simple) to use according to the state data, and determine the determined gravity A movement vector may be calculated from the orientation vector. Note that the adjustment unit 11 may control whether to use a gravity direction vector (high accuracy) or a gravity direction vector (simple).

In the above example, both the gravity direction vector (high accuracy) and the gravity direction vector (simple) are always output. However, depending on the holding state, the gravity direction vector (high accuracy) or the gravity direction vector (simple ) May be output.

[Modification]
In the above embodiment, the measurement device including the measurement unit 2 has been described. However, the measurement unit 2 may be configured separately from the measurement device. That is, the measuring device of the present invention is not limited to the one incorporating the measuring unit 2, and may be one that is communicably connected to the measuring device. In this case, the measurement data output from the measuring device is received, and the parameter relating to the moving state of the moving body that holds the measuring device is calculated.

Moreover, in the said embodiment, the simple movement amount estimation part 17 and the simple gravity direction estimation part 15b process using a part of measurement data which the high precision movement amount estimation part 16 and the high precision gravity direction estimation part 15a use. Although the example to perform was demonstrated, it is not restricted to this example. For example, a high-precision sensor for the high-precision movement amount estimation unit 16 and the high-precision gravity direction estimation unit 15a and a low-precision sensor for the simple movement amount estimation unit 17 and the simple gravity direction estimation unit 15b may be mounted. Good.

Furthermore, in the above-described embodiment, one block (for example, the movement amount estimation unit 12) can execute a plurality of calculation processes, and is designated by a control signal output from the adjustment unit 11 among the plurality of calculation processes. Although the example which performs a calculation process was demonstrated, it is not restricted to this example. For example, the movement amount estimation unit 12 may be divided into a plurality of blocks that execute one type of calculation process, and the block specified by the control signal output from the adjustment unit 11 may be calculated.

In the above-described embodiment, the example in which the arithmetic processing is changed via the adjustment unit 11 has been described. However, the holding state estimation unit 10 outputs the state data directly to the movement amount estimation unit 12 and the like to estimate the movement amount. The unit 12 or the like may determine and execute a calculation process according to the state data.

The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims, and embodiments obtained by appropriately combining technical means disclosed in different embodiments. Is also included in the technical scope of the present invention.

As described above, the calculation device of the present invention is a calculation device that calculates a parameter indicating the moving state of the moving body using measurement data output from one or more sensors held by the moving body. From the measurement data, a holding state specifying unit that specifies how the sensor is held by the moving body, and a calculation process according to the holding state specified by the holding state specifying unit are performed. It is a structure provided with the parameter calculation means which calculates the parameter which shows the movement state of a moving body.

The measurement data includes acceleration data indicating the magnitude and direction of the acceleration detected by the sensor, and angular velocity data indicating the magnitude and direction of the angular velocity detected by the sensor. Posture angle calculating means for calculating the posture angle of the moving object using the data and the acceleration data, and the holding state specifying means uses the posture angle calculated by the posture angle calculating means and the measurement data. It is preferable to specify how the sensor is held by the moving body.

重力 By calculating the attitude angle, the gravity direction of the moving object can be specified. By specifying the gravity direction, the holding state can be specified with higher accuracy. Therefore, according to the above configuration for specifying the holding state using the posture angle and the measurement data, it is possible to improve the specifying accuracy of the holding state, thereby increasing the accuracy with which appropriate arithmetic processing is performed. .

Further, by improving the specific accuracy of the holding state, it becomes possible to distinguish and detect a finer holding state. For this reason, according to said structure, it is also possible to increase the variation of the holding | maintenance state assumed beforehand and to calculate a parameter by more suitable arithmetic processing.

The measurement data includes acceleration data indicating the magnitude and direction of the acceleration detected by the sensor, and the calculation device uses the acceleration data to specify a gravity direction specifying means for specifying the gravity direction of the moving body. It is preferable that the holding state specifying unit specifies how the sensor is held by the moving body using the gravity direction specified by the gravity direction specifying unit and the measurement data.

By specifying the gravity direction, it becomes possible to specify the holding state with higher accuracy. Therefore, according to the above configuration that specifies the holding state using the gravity direction and the measurement data, it is possible to improve the specifying accuracy of the holding state, thereby increasing the accuracy with which appropriate calculation processing is performed. . Note that the gravity azimuth specified by the gravity azimuth specifying means is the gravity azimuth of the moving object, not the gravity azimuth in the world coordinate system.

Here, the holding state of the sensor is not always constant and may change. For example, when the moving body is a person, it is conceivable that the holding state changes by changing the sensor. The measurement data detected by the sensor during a period in which the holding state changes from one holding state to another holding state has a value that is significantly different from normal. For this reason, when the parameter is calculated using the measurement data in this period as it is, there is a possibility that a parameter greatly different from the actual motion state is calculated.

Therefore, it is preferable that the holding state specified by the holding state specifying unit includes a transition state in which the holding state of the sensor transitions to another holding state.

According to the above configuration, the transition state is specified as one of the holding states, and the parameter is calculated by the arithmetic processing according to the transition state, so that it is possible to prevent a parameter greatly different from the actual motion state from being calculated. be able to.

For example, when the sensor held in the pocket is taken out, the holding state changes from the holding state held in the pocket to the holding state held in the hand. In the above configuration, the holding state during the period in which the holding state is changing is detected as the transition state.

Here, in the transition state, the value of measurement data (for example, acceleration data, angular velocity data, geomagnetism data, etc.) output by the sensor rises sharply due to a large change in the position and orientation of the sensor in a short time. Often to do. For this reason, it is possible to detect the transition state by calculating the rate of change of the measurement data over time and determining whether the rate of change exceeds a predetermined threshold.

However, since the value of the measurement data output by the sensor depends on the motion state of the moving object holding the sensor, the method of always detecting the transition state with the same threshold value increases the detection accuracy of the transition state. There is a limit. For example, if the moving speed of the moving body increases, the value of measurement data output by the sensor increases accordingly.

For this reason, the holding state specifying means sets a larger threshold value as the value of the measurement data output from the sensor is larger, and the transition is performed when the change rate of the measurement data exceeds the set threshold value. It is preferable to identify the state.

According to the above configuration, since the transition state is specified by setting a threshold value having a large value of the measurement data output from the sensor, it is possible to improve the detection accuracy of the transition state.

In addition, the calculation device includes a simple parameter calculation unit that calculates a parameter indicating a moving state of the moving body using a part of the measurement data, and the parameter according to the holding state specified by the holding state specifying unit. The parameter calculated by the calculation means is output, the parameter calculated by the simple parameter calculation means is output, or the parameter calculated by the parameter calculation means is combined with the parameter calculated by the simple parameter calculation means Preferably, switching means for switching whether to output the selected parameter is provided.

According to the above configuration, since the parameter calculation unit is switched according to the holding state, the parameter can be calculated by an appropriate unit according to the holding state. For example, in a state where the value of measurement data is unstable such as a transition state, the parameter calculation accuracy by the parameter calculation means may be reduced. For this reason, in the state where the value of the measurement data is unstable, the parameter can be calculated by a simple calculation process using a small amount of measurement data by switching to the simple parameter calculation means.

Note that calculating a parameter using a part of measurement data means calculating the parameter with a data amount smaller than the measurement data. For example, when the measurement data includes acceleration data in the triaxial direction and angular velocity data in the triaxial direction, the parameter is calculated using only the acceleration data in the triaxial direction, the acceleration data in the triaxial direction and 2 This refers to calculating a parameter using the angular velocity data in the axial direction. In addition, for example, a case where the parameter is calculated by reducing the data amount by increasing the period for acquiring the measurement data from the sensor or by increasing the period at which the sensor outputs the measurement data is included.

When the holding state specified by the holding state specifying unit is a holding state in which the parameter calculated by the simple parameter calculating unit is output, the sensor is controlled, and measurement data not used by the simple parameter calculating unit Among them, it is preferable to include a measurement control means for stopping at least a part of the measurement or reducing the output frequency of the measurement data.

According to the above configuration, when the parameter is calculated by the simple parameter calculation means, the measurement by the sensor is partially stopped or the output frequency of the measurement data is lowered, so that the power consumption of the sensor is reduced. can do.

The gravity direction specifying means includes simple gravity direction specifying means for specifying the gravity direction of the moving body using a part of measurement data used for specifying the gravity direction of the moving body, and the parameter calculating means includes: Preferably, the parameter is calculated using the gravity azimuth specified by the gravity azimuth specifying means or the simple gravity azimuth specifying means according to the holding state specified by the holding state specifying means.

Depending on the holding state, measurement data with a value that is not assumed by the gravity azimuth specifying means is input, so it is more appropriate to obtain the gravity azimuth with a smaller amount of data than the gravity azimuth specified by the gravity azimuth specifying means. A value may be calculated.

Therefore, according to the above configuration that switches which one of the gravity azimuth specifying means and the simple gravity azimuth specifying means to use according to the holding state, the parameter estimation accuracy can be increased.

The parameter calculation means does not identify the contribution of the measurement data output from the sensor during the period when the holding state specifying means is in the transition state as the holding state specifying means is in the transition state. It is preferable to calculate the parameter by a calculation process that is smaller than the contribution of measurement data output from the sensor during the period.

As described above, the measurement data output in the transition state is unstable and has low reliability. Therefore, in the above configuration, in the calculation processing in the transition state, the contribution of the measurement data output during the transition state period is relatively small. Thereby, the accuracy of parameters can be improved.

The computer may be realized by a computer. In this case, a control program for realizing the computer by the computer by operating the computer as each unit of the computer, and recording the program Such computer-readable recording media also fall within the scope of the present invention.

Each block of the

positioning devices

1, 30, 40, 50, 60, and 70 (hereinafter referred to as the positioning device 1 and the like), in particular, the control unit 3 may be configured by hardware logic, as follows. Alternatively, it may be realized by software using a CPU.

That is, the positioning device 1 or the like includes a CPU (central processing unit) that executes instructions of a control program that realizes each function, a ROM (read only memory) that stores the program, and a RAM (random access memory) that expands the program. And a storage device (recording medium) such as a memory for storing the program and various data. An object of the present invention is a recording medium in which program codes (execution format program, intermediate code program, source program) of a control program such as the positioning device 1 which is software for realizing the functions described above are recorded so as to be readable by a computer. This can also be achieved by supplying to the positioning device 1 and the like and reading and executing the program code recorded on the recording medium by the computer (or CPU or MPU).

Examples of the recording medium include tapes such as magnetic tapes and cassette tapes, magnetic disks such as floppy (registered trademark) disks / hard disks, and disks including optical disks such as CD-ROM / MO / MD / DVD / CD-R. Card system such as IC card, IC card (including memory card) / optical card, or semiconductor memory system such as mask ROM / EPROM / EEPROM / flash ROM.

Alternatively, the positioning device 1 or the like may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited. For example, the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication. A net or the like is available. Further, the transmission medium constituting the communication network is not particularly limited. For example, even in the case of wired such as IEEE 1394, USB, power line carrier, cable TV line, telephone line, ADSL line, etc., infrared rays such as IrDA and remote control, Bluetooth ( (Registered trademark), 802.11 wireless, HDR, mobile phone network, satellite line, terrestrial digital network, and the like can also be used. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

In addition, the specific embodiments or examples made in the section for carrying out the invention are merely to clarify the technical contents of the present invention, and are limited to such specific examples. The present invention should not be construed in a narrow sense but can be implemented with various modifications within the spirit of the present invention and the scope of the following claims.

According to the present invention, the parameter indicating the moving state of the moving object can be estimated with high accuracy. For example, since the moving direction, moving distance, etc. of a person who moves on foot can be estimated with high accuracy, the present invention can also be suitably applied to a navigation apparatus that displays a person's current location on a map.

1, 30, 40, 50, 60, 70 Positioning device (calculation device)
2 Measurement unit (sensor)
10 Holding state estimation unit (holding state specifying means)
10c Holding state determination unit (measurement control means)
11 Adjustment unit (switching means)
12 Movement amount estimation unit (parameter calculation means)
12a Walking motion detection unit (parameter calculation means)
12b Stride estimation unit (parameter calculation means)
13 Position determination unit (parameter calculation means)
14 Attitude angle estimation unit (Attitude angle calculation means)
15 Gravity orientation estimation unit (gravity orientation identification means)
15b Simple gravity direction estimation unit (simple gravity direction specifying means)
16a Walking motion detection unit (parameter calculation means)
16b Stride estimation unit (parameter calculation means)
16c Moving direction detection unit (parameter calculation means)
16d movement vector calculation unit (parameter calculation means)
17 Simple movement amount estimation unit (simple parameter calculation means)
17a Walking motion detection unit (simple parameter calculation means)
17b Stride estimation unit (simple parameter calculation means)
17c Moving direction detection unit (simple parameter calculation means)
17d Movement vector calculation unit (simple parameter calculation means)

Claims

A calculation device that calculates a parameter indicating a moving state of a moving body using measurement data output from one or more sensors held by the moving body,
From the measurement data, holding state specifying means for specifying how the sensor is held by the moving body,
A calculation apparatus comprising: parameter calculation means for performing a calculation process according to the holding state specified by the holding state specifying means and calculating a parameter indicating a moving state of the moving body from the measurement data.
The measurement data includes acceleration data indicating the magnitude and direction of acceleration detected by the sensor, and angular velocity data indicating the magnitude and direction of angular velocity detected by the sensor,
A posture angle calculating means for calculating a posture angle of the moving body using the angular velocity data and the acceleration data;
2. The holding state specifying unit specifies how the sensor is held by a moving body using the posture angle calculated by the posture angle calculating unit and the measurement data. The computing device described in 1.
The measurement data includes acceleration data indicating the magnitude and direction of acceleration detected by the sensor,
Using the acceleration data, comprising gravity direction specifying means for specifying the gravity direction of the moving body,
2. The holding state specifying unit specifies how the sensor is held by a moving body using the gravity direction specified by the gravity direction specifying unit and the measurement data. The computing device described in 1.
4. The holding state specified by the holding state specifying means includes a transition state in which the holding state of the sensor is transitioning to another holding state. The computing device described.
The holding state specifying means sets a threshold value that is larger as the value of the measurement data output from the sensor is larger, and is in the transition state when the rate of change of the measurement data exceeds the set threshold value. The calculation apparatus according to claim 4, characterized by:
Simple parameter calculating means for calculating a parameter indicating a moving state of the moving body using a part of the measurement data;
According to the holding state specified by the holding state specifying unit, a parameter calculated by the parameter calculating unit, a parameter calculated by the simple parameter calculating unit, or a parameter calculated by the parameter calculating unit 6. The calculation according to claim 1, further comprising: a switching unit that switches whether to output a parameter calculated by combining the parameter calculated by the simple parameter calculation unit. apparatus.
When the holding state specified by the holding state specifying unit is a holding state in which the parameter calculated by the simple parameter calculating unit is output, the sensor is controlled, and measurement data not used by the simple parameter calculating unit The calculation apparatus according to claim 6, further comprising a measurement control unit that stops at least a part of the measurement or lowers the output frequency of the measurement data.
The gravity direction specifying means comprises simple gravity direction specifying means for specifying the gravity direction of the moving body using a part of measurement data used for specifying the gravity direction of the moving body,
The parameter calculating means calculates the parameter using the gravity direction specified by the gravity direction specifying means or the simple gravity direction specifying means according to the holding state specified by the holding state specifying means. The calculation apparatus according to claim 3.
The parameter calculation means does not identify the contribution of the measurement data output from the sensor during the period when the holding state specifying means is in the transition state as the holding state specifying means is in the transition state. The calculation apparatus according to claim 4, wherein the parameter is calculated by an arithmetic process that is smaller than a contribution of measurement data output from the sensor during a period.
A control method for a computing device that uses a measurement data output from one or more sensors held by a moving body to calculate a parameter indicating a moving state of the moving body,
From the measurement data, a holding state specifying step for specifying how the sensor is held by the moving body;
And a parameter calculation step of calculating a parameter indicating a movement state of the moving body from the measurement data, performing a calculation process according to the holding state specified in the holding state specifying step. .
A control program for operating the computing device according to any one of claims 1 to 9, wherein the control program causes a computer to function as each of the means.
A computer-readable recording medium on which the control program according to claim 11 is recorded.