JP6494552B2 - Position estimating apparatus, program and method capable of correcting position based on transition between floors - Google Patents

Description

  The present invention relates to a technique for estimating a position / movement state of a mobile terminal.

  In recent years, pedestrian autonomous navigation (PDR, Pedestrian Dead Reckoning) for estimating the position and movement of a pedestrian has been actively developed. According to this PDR, for example, the position of a pedestrian can be measured even in an indoor environment where GPS (Global Positioning System) cannot be used.

  As a conventional technique for PDR, for example, Patent Document 1 stores correspondence information in which information on a human walking state is associated with a first position representing a position where the walking state occurs, and then autonomous navigation is performed. A technique is disclosed in which the first position is determined as the current position of the person according to the result of the above.

  Specifically, in this technique, when a walking state is estimated based on measurement information measured according to walking, the above-described position is calculated near the second position, which is a position calculated by autonomous navigation based on the measurement information. If the first position is determined to exist, the first position is determined as the current position. Further, as the above-mentioned walking state of the human, for example, “straddling”, “swinging”, “sidewalking”, “bending walking” and the like are cited as examples. Further, the correspondence information is generated from a floor plan or the like.

  In addition, as another conventional technology, a navigation technology that can be adopted for a mobile terminal possessed by a pedestrian, calculates a guide route to the destination, and corrects the terminal position by assuming that it is on the guide route while walking. There are also techniques to do this. In this technique, in order to correct the terminal position, map information is required, and in addition, a destination needs to be set.

Japanese Patent Laying-Open No. 2015-014587

  As described above, in the conventional technology as described above, the position of the pedestrian cannot be measured unless map information is prepared in advance.

  In general, in PDR, autonomous positioning is repeatedly performed and the position is tracked. Therefore, there is a problem that positioning errors accumulate with time. In order to suppress the accumulation of such positioning errors, the conventional technology as described in Patent Document 1 has a configuration with an excessive burden such as preparing map information in advance and setting a destination in some cases. It must be adopted.

  However, in situations where PDR is actually performed, map matching is often difficult. For example, the creation and maintenance of map information for all indoor buildings and underground malls where pedestrians can walk is difficult to actually adopt because map data owners and formats may differ from one another. On the other hand, for example, if the position of the pedestrian is estimated based on the output result from the sensor built in the portable device and the position can be corrected without using the map information, such a problem is solved. Can do.

  Therefore, an object of the present invention is to provide an apparatus, a program, and a method that can suppress accumulation of positioning errors without relying on map information.

According to the present invention, a position estimation device that estimates the position of a mobile subject in an area that can be moved between a plurality of floor areas by means of a predetermined transition , and holds the position in time series,
Floor transition detection means for detecting a transition of a floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
Identifies whether or not the position when a regression transition, which is a movement from another floor area to a floor area located in the past, is detected is the same location as the retained past position in the floor area. A place determination means;
A position having position correction means for calculating a correction amount based on information on the position determined to be the same place and information on the past position, and correcting the estimated position using the correction amount An estimation device is provided.

  As one embodiment of the position estimation apparatus according to the present invention, the same location determination means is configured to at least when the distance between the position where the regression transition is detected and the past position is equal to or less than a predetermined threshold. It is also preferable to determine that they are the same place.

Further, as another embodiment of the same location determination, the same place judging means includes at least a detected transition for means of the regression transition, the same transition for means of the detected transition immediately following the past position It is also preferable to determine that it is the same place.

  Further, the position estimation device further includes a corner detection unit that detects a corner in movement in the floor area, and the same location determination unit detects a position when the regression transition is detected. It is also preferable that the corner is detected later.

  As one embodiment of the position correction according to the present invention, the position correction means corrects the estimated position using at least the position difference between the position determined to be the same place and the past position. Is also preferable.

  Furthermore, as another embodiment of the position correction, the position correction means includes at least a traveling direction in the transition immediately before the position determined to be the same place and a traveling direction in the transition immediately after the past position. It is also preferable to correct the estimated position using the angle difference between

  As still another embodiment of the position correction, it is preferable that the position correction means also sets a held position that is past the position determined to be the same place as a correction target.

  Further, as one embodiment of the floor transition detection according to the present invention, the floor transition detection means is configured to detect the floor area based on the measurement result information of acceleration, angular velocity, geomagnetism and / or atmospheric pressure as information relating to the measured movement. It is also preferred to detect transitions.

According to the present invention, the computer mounted in the apparatus that estimates the position of the portable subject in the area that can be moved between the plurality of floor areas by the predetermined transition means , and holds the position in time series functions. A position estimation program for
Floor transition detection means for detecting a transition of a floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
Identifies whether or not the position when a regression transition, which is a movement from another floor area to a floor area located in the past, is detected is the same location as the retained past position in the floor area. A place determination means;
The computer functions as a position correction unit that calculates a correction amount based on the information related to the position determined to be the same place and the information related to the past position, and corrects the estimated position using the correction amount. A position estimation program is provided.

According to the present invention, the position of a portable subject in an area that can be moved between a plurality of floor areas with a predetermined means for transition and that is installed in a device that holds the position in time series is further estimated. An estimation method,
Detecting a transition of the floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
A step of determining whether or not a position when a regression transition that is a movement from another floor area to a floor area located in the past is detected is the same place as the retained past position in the floor area. When,
A position estimation method comprising: calculating a correction amount based on information related to the position determined to be the same place and information related to the past position, and correcting the estimated position using the correction amount Is provided.

  According to the position estimation device, program, and method of the present invention, accumulation of positioning errors can be suppressed without relying on map information.

It is the schematic for demonstrating the condition where a user walks carrying the position estimation apparatus by this invention. It is a functional block diagram which shows the function structure in one Embodiment of the position estimation apparatus by this invention. It is a table which shows one Example of the position estimation result recorded on the position and floor transition information storage part. It is a schematic diagram for demonstrating various embodiment in the same place determination process performed in the same place determination part. It is a schematic diagram for demonstrating other embodiment about the correction process of the estimated position in a position correction part. It is a graph for demonstrating one Embodiment of the azimuth | direction correction and position correction | amendment in an azimuth | direction correction part.

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

  FIG. 1 is a schematic diagram for explaining a situation where a user walks with the position estimation device according to the present invention.

According to FIG. 1 (A), the user is walking in a building as a portable subject while carrying the portable terminal 1 as the position estimating device according to the present invention. This building is a building having three floors F _a , F _b and F _c having different heights as a floor area. Moreover, between these floors, transition (movement) is possible with a predetermined transition type (steps and elevators in FIG. 1). Note that the portable subject in the sense that the terminal is capable of positioning and is the target subject of position estimation is not limited to a person as in this embodiment. For example, an animal such as a pet or a moving body such as a walking robot can be a portable subject.

  The mobile terminal 1 includes an acceleration sensor 101 that can measure acceleration as a vector quantity. Based on the acceleration information output from the acceleration sensor 101, the mobile terminal 1 estimates its own position using, for example, a known PDR (Pedestrian Dead Reckoning) technique in the position estimation unit 113, and further calculates the estimated position information. Hold in time series.

  The mobile terminal 1 may further include a gyro sensor 102, a geomagnetic sensor 103, and / or an atmospheric pressure sensor 104. By using measurement information from such a sensor, it is possible to estimate a position with higher accuracy. For example, the output from the acceleration sensor 101 is used to determine the number of steps and the travel distance (path) in walking, the output from the acceleration sensor 101 and the geomagnetic sensor 103 are used to determine the direction in which the user travels, and the gyro The amount of change in the user's direction of travel can be determined using the output from the sensor 102. Furthermore, it is possible to estimate the current position with high accuracy by taking into consideration the acquired travel distance information and travel direction change information.

  However, the portable terminal 1 does not hold the map information about this building in this embodiment. Therefore, it is not assumed in the present embodiment that the estimated position is corrected by using a known map matching technique to improve the position estimation accuracy.

In order to improve the position estimation accuracy without relying on such map information, the mobile terminal 1 specifically,
(A) a floor transition detection unit 111 that detects a floor transition based on the measured “information related to movement”, that is, a movement from one floor of the mobile terminal 1 to another floor;
(B) The same place determination unit 112 that determines whether or not the position when the return transition to the floor located in the past is detected is the same place as the past position held in the floor area;
(C) A position correction unit 113c that calculates a correction amount based on the information related to the position determined to be the same place and the information related to the past position, and corrects the estimated position using the correction amount. And have.

  Here, “information relating to movement” in (A) will be described in detail later, but output information from the sensor 101 or the like provided in the mobile terminal 1, that is, measurement of acceleration, angular velocity, geomagnetism and / or atmospheric pressure. It can be the result information.

  As described above, the mobile terminal 1 detects the floor transition based on the information measured each time, and calculates the correction amount for the estimated position based on the position information related to the detected floor transition. Further, it is possible to determine a more accurate estimated position using this correction amount. As a result, it is possible to suppress accumulation of positioning errors, which is a problem in PDR, without relying on map information.

  Furthermore, according to the mobile terminal 1, since the correction amount can be calculated during walking with floor transition, the estimated position can be corrected as appropriate during walking. As a result, an estimated position with higher accuracy can be determined every moment.

  In addition, although the position estimation apparatus of this invention has taken the form of the portable terminal in this embodiment, of course, it is not limited to it. For example, instead of a communication terminal, a portable device dedicated to PDR may be used. Or it can also be set as the information processing apparatus which can communicate between terminals, such as a smart phone with a built-in sensor which a user carries. In this case, the information processing apparatus may be a server that acquires necessary information from terminals owned by a plurality of users and accurately estimates the position of each user.

  In any case, according to the present invention, it is possible to suppress or eliminate the accumulation of the error of the estimated position accompanying the movement in the floor (area) by the information related to the floor transition.

  FIG. 2 is a functional block diagram showing a functional configuration in an embodiment of the position estimation apparatus according to the present invention.

  According to FIG. 2, the portable terminal 1 as a position estimation apparatus according to the present invention includes an acceleration sensor 101, a position / floor transition information storage unit 105, a display unit 106, and a processor memory. Moreover, you may have the gyro sensor 102, the geomagnetic sensor 103, and / or the atmospheric | air pressure sensor 104. FIG. Furthermore, it is preferable to have the communication unit 107. Here, the processor memory realizes the position estimation function by executing a program that causes the computer of the mobile terminal 1 to function.

  In addition, the processor memory includes a floor transition detection unit 111, a same location determination unit 112, and a position estimation unit 113 including a position correction unit 113a as functional components. Here, it is preferable that the position estimation unit 113 further includes a corner detection unit 113b. The position correction unit 113a preferably includes an azimuth correction unit 113c. Note that the processing flow shown by connecting the functional components of the mobile terminal 1 in FIG. 2 with arrows is also understood as an embodiment of the position estimation method according to the present invention.

  Similarly, in FIG. 2, the acceleration sensor 101 can be an acceleration meter that measures acceleration as a vector quantity. For example, a three-axis type measuring instrument using a capacitance method or a piezoresistive method formed using a MEMS (Micro Electro Mechanical Systems) technique may be used. Further, as the acceleration sensor 101, an accelerometer that can measure gravitational acceleration and count the number of steps of the user who has the mobile terminal 1 can be used.

  The gyro sensor 102 is a gyroscope that detects an angular velocity, and can be a direction change meter that detects and measures a change in direction. For example, it may be a three-axis vibration gyro sensor that detects an angular velocity from a deformation of a structure due to Coriolis force acting on a vibrating arm. Further, as the gyro sensor 102, a fluid type gyro sensor, an optical gyro sensor, or the like can be adopted.

  The geomagnetic sensor 103 is, for example, a three-axis type geomagnetism meter that measures the geomagnetism using the magnetoresistive (AMR, GMR or TMR) effect, the magnetic impedance (MI) effect, the fluxgate (FG) method, or the Hall effect. can do. Note that the geomagnetic sensor 103 and the acceleration sensor 101 may be combined to form an orientation measuring unit that measures the orientation of the mobile terminal 1.

  The barometric sensor 104 can be a highly sensitive barometer that can detect minute changes in barometric pressure corresponding to an altitude difference of several centimeters to several tens of centimeters. For example, it may be a diaphragm type or cantilever type pressure gauge formed by using MEMS technology. By using such an atmospheric pressure sensor 104, it is possible to measure the amount of movement in the vertical direction (vertical direction) due to an elevator or a staircase.

Similarly in FIG. 2, the position / floor transition information storage unit 105
(A) Information related to the position of the terminal 1 estimated (positioned) by the position estimation unit 113 described in detail later;
(B) Information related to the floor transition determined by the floor transition detection unit 111, which will be described later in detail, for example, the fact that the transition is detected and the transition type of the transition are stored in association with the corresponding time. In addition, when the corner detection part 113b mentioned later is set, the information which concerns on the position of said (a) can also be only the position of the point determined to be a corner among the estimated positions.

  FIG. 3 is a table showing an example of the position estimation result recorded in the position / floor transition information storage unit 105.

According to FIG. 3, the position / floor transition information storage unit 105 stores a position estimation result (positioning result) that is information related to the position of the terminal 1 estimated (positioned) by the position estimating unit 113. The stored position estimation result is obtained at each position estimation time.
(A) Direction (for example, azimuth angle) of the mobile terminal 1 at the time
(B) The distance (movement distance) that the terminal 1 has moved from the previous time to the time, and (c) the x coordinate value in the horizontal plane coordinate system of the location of the mobile terminal 1 at the time (Position coordinate x) and y coordinate value (Position coordinate y)
Is recorded in association with each other.

  Here, the direction of (a) may be calculated by a known method in the position estimation unit 113 based on measurement result information output from the acceleration sensor 101, the geomagnetic sensor 103, or the gyro sensor 102, for example. Further, the moving distance (b) can be calculated from a predetermined step length after determining the number of steps using, for example, acceleration information output from the acceleration sensor 101. Further, the position coordinates of (c) may be calculated by a known method by combining the calculation results of (a) and (b), for example.

[Floor transition detection processing]
Returning to FIG. 2, the floor transition detection unit 111 detects the transition of the floor, that is, the movement of the mobile terminal 1 from one floor to another floor based on the sensor measurement result information. Here, the sensor measurement result information is “information related to movement” and is output information from at least one of the sensors 101 to 104, that is, information related to measurement values of acceleration, angular velocity, geomagnetism, and / or atmospheric pressure.

  For example, a change in the position (position coordinate z) in the vertical height direction (vertical direction) is determined from the time change in the atmospheric pressure information output from the atmospheric pressure sensor 104, and the change corresponds to transition (movement) between floors. If a predetermined threshold value for the amount of change is exceeded, a floor transition can be detected. It is also possible to specify the type of floor transition, that is, the type of means for moving between floors, such as stairs, spiral stairs, elevators, escalators, and slopes, from the speed and profile of the atmospheric pressure change over time.

  Furthermore, the floor transition is detected based on the profile of the time change of the acceleration vector output from the acceleration sensor 101 and the acceleration change in the gravitational acceleration direction (vertical direction) in the acceleration information. That is, it is possible to specify the type of means for moving between floors as described above. Further, based on the measurement result information output from the gyro sensor 102 or the geomagnetic sensor 103, for example, it can be specified that the spiral staircase is used as the transition type.

  Thus, in the present embodiment, the floor transition detection unit 111 determines not only the detection of the floor transition but also the transition type of the detected transition. Further, information on the detected floor transition and the determined transition type is recorded in the position / floor transition information storage unit 105 in association with the occurrence time of the transition, for example.

[Identical location judgment processing]
Similarly, in FIG. 2, the same location determination unit 112 determines the position when the regression transition to the floor located in the past (the movement to return from another floor to the previously located floor) is detected in the past. It is determined whether or not it is “the same place” as the past position held on the floor.

  Here, the condition “same place” can take various forms as a predetermined criterion. Hereinafter, the conditions will be described with reference to the drawings.

  FIG. 4 is a schematic diagram for explaining various embodiments in the same place determination process performed by the same place determination unit 112.

4A to 4C, the mobile terminal 1 (the user who owns it) makes a floor transition from the position p of the floor F _a to the floor F _b using a transition type of staircase, and then the same staircase ( FIG. 4 (a)), another staircase (FIG. 4 (B)), or elevator manner leading to a position q and return transition to the floor F _a with (FIG. 4 (C)) are shown.

Here, the same location determination unit 112 has, as one determination form,
(A) If it can be confirmed that the floor located in the past and the floor located after the regression transition are the same, it may be determined that they are “the same place”.

In this case, the same location determination unit 112, specifically, from the position and floor transition information storage unit 105, when acquiring the floor transition detection information from the floor F _b to the floor F _a, floor serving as its previous transition By determining and acquiring the floor transition detection information from F _a to the floor F _b , the above determination (a) can be made. It may be determined that they are “the same place” on the condition that the floor transition detection information from the past F _a to F _b is searched. Such determination (A), as in case of FIG. 4 (A), and when the transition means between the floor F _a and F _b is only one, the doorway of the different transition means at floor F _a Effective when they are in the vicinity of each other.

As another determination form, the same location determination unit 112
(A) At least, when the distance between the position q when the regression transition is detected and the past position p is equal to or less than a predetermined threshold value, it may be determined as “the same place”.

In this case, the same location determination unit 112, specifically, from the position and floor transition information storage unit 105, and the floor transition detection information from the floor F _b to the floor F _a, after floor transition completion position q (now) When the information is acquired, the floor transition detection information from the floor F _a to the floor F _{b as} the previous transition, and the information of the position p stored immediately before the floor transition from the F _a to the F _b The above determination (a) can be made by searching for and acquiring.

Here, for example, the distance between the position p and the position q may be calculated, and if the distance is equal to or less than a predetermined threshold value, it may be determined that they are “the same place”. At this time, as a predetermined threshold,
(A) A fixed value (for example, 10 m) may be used,
(B) It may be a predetermined percentage (for example, 20%) of the total travel distance up to the present (total travel distance at a past time in a predetermined period),
(C) You may employ | adopt the larger one among the values calculated | required by said (a) and (b).

According to such determination (A), in the case of FIG. 4A, for example, the position q is determined to be within the predetermined distance range as viewed from the position p and is determined to be “the same place” as the position p. However, in the case of FIG. 4B, for example, the position q is determined to be outside the predetermined distance range as viewed from the position p, and is determined not to be “the same place” as the position p. That is, according to the determination (a), it is possible to make an effective determination even when there are a plurality of transition means between the floors F _a and F _b as shown in FIG. 4B. .

Furthermore, as another form of the same place determination process, the same place determination unit 112
(C) at least, it is a "place" if the transition type of the detected (to the floor F _a) Regression transition, a transition type of transition detected immediately after the last location p is the same May be determined.

In this case, the same location determination unit 112, specifically, from the position and floor transition information storage unit 105, and the floor transition detection information from the floor F _b to the floor F _a, and obtains the transition type information of the same floor transitions In this case, the above determination (c) can be made by searching and acquiring the floor transition detection information from the floor F _a to the floor F _b and the transition type information of the floor transition as the previous transition. it can.

That is, when the acquired transition type of the floor transition from the floor F _b to the floor F _a is the same type as the searched transition type of the floor transition from the previous floor F _a to the floor F _b (for example, both stairs If the transition type is not the same type (for example, as shown in FIG. 4C, one is a staircase and the other is an elevator). ), It can be determined that they are not “the same place”.

Furthermore, the same place determination unit 112 is, as still another embodiment,
(D) When both the distance condition (B) and the transition type condition (C) are satisfied, it may be determined that the “same place” is set.

  Which of the same location determination processes described above is adopted can be determined according to a building or the like that is actually a location estimation place. Alternatively, the same location determination process with high estimated position accuracy may be selected by repeating trials of position estimation.

  When a corner detection unit 113b described later is set, the same location determination unit 112 is determined as a corner as the detection position q and / or the past position p of the regression transition in the above (a) and (d). It is also possible to adopt only the location of the spot. In the first place, the position recorded in the position / floor transition information storage unit 105 may be only the position of the point determined to be a corner. Furthermore, the same place determination unit 112 can set the detection position q of the regression transition as the position of the corner point detected immediately after the regression transition is detected. Further, the position p immediately before the past floor transition may be the position of the corner point detected immediately before the floor transition is detected.

[Position estimation / correction processing]
Returning to FIG. 2, the position estimation unit 113 is based on the measurement result information output from the acceleration sensor 101, the gyro sensor 102, the geomagnetic sensor 103, and / or the atmospheric pressure sensor 104, for example, using a known PDR technique. Further, the estimated position information is recorded in the position / floor transition information storage unit 105 as the position data of the terminal 1 in time series.

  For example, the position estimation unit 113 may detect an acceleration variation that is a walking candidate based on the measurement result information output from the acceleration sensor 101, and calculate the number of steps based on the detection result. If the number of steps is known, the moving distance is determined. Here, whether or not the acceleration variation is a candidate for walking can be determined based on, for example, the detected time interval of acceleration variation or its standard deviation.

  Further, the corner detection unit 113b of the position estimation unit 113 detects a corner in movement within the floor. Here, the corner is a turning position in the traveling direction (which forms a corner) on the movement path in the floor. Specifically, this corner (position) is determined based on whether angular velocity change corresponding to a turning angle is detected based on angular velocity measurement result information output from the gyro sensor 102, or an acceleration sensor. It is possible to detect and determine the turning point of the direction of travel calculated using the output from the 101 and the geomagnetic sensor 103. Further, it may be determined by detecting whether or not the moving route is bent based on time-series data of the positioning result (estimated position) in the horizontal direction (in-plane direction).

  In the embodiment provided with such a corner detection unit 113b, the estimated position recorded in the position / floor transition information storage unit 105 can be limited to the position determined to be a corner. In general, the corner position often has to bend at that position when passing through the floor, and when moving in the vicinity of the corner, it is generally used as the reference position for calculating the correction amount. It is an easy position.

  Further, the position correction unit 113a of the position estimation unit 113, when the same place determination unit 112 determines that the detection position q of the regression transition and the past position p are the same place, these positions q (or position q) Information) and position p (or information related to position p), a correction amount is calculated, and the estimated position is corrected using this correction amount (a variable for correcting the position is updated).

Specifically, position correction unit 113a, in the present embodiment, by using the difference in position between the position p and the position q, the offset O _P position as the correction amount, the following equation (1) O _P = q- p
Calculate and update with Wherein each q and p of the equation (1) is a position vector representing a position q and a position p, and therefore, the offset O _P is the amount of correction vectors.

Then, the position correcting unit 113a, the offset O _P upon calculating the 'a, for example, the following equation (2) q' position q obtained by correcting the position q = q-O _P
Calculate and update with In this case, the corrected position q ′ information is recorded in the position / floor transition information storage unit 105.

  Further, as a change mode, the position correction unit 113a has not only the position q after the regression transition determined to be the same place, but also a position that is past (held in the storage unit 105) than the position q. It is also preferable to make it a correction target. Hereinafter, an embodiment in which correction is performed retroactively to a past position will be described with reference to the drawings.

  FIG. 5 is a schematic diagram for explaining another embodiment of the correction process of the estimated position in the position correction unit 113a.

In FIG. 5, the mobile terminal 1 (the user who owns it) makes a floor transition from the position p1 of the floor F _{a to} the floor F _b by using the staircase 1 (using the transition type of staircase) and passes the position p1 _p . Next, a floor transition is made from the position p2 of the floor F _{b to} the floor F _c by using the staircase 2 (using a transition type called staircase), and then the floor F is moved from the floor F _c by the staircase 3 (using a transition type called staircase). and floor transitions to _b passed the position q1, Thereafter, how to reach the floor F by stairs 1 from a position q1 _p of _b (with a transition type of stairs) positioned floor transitions to the floor F _a q2 is shown Has been. These transition information and position information are recorded in the position / floor transition information storage unit 105 in association with the corresponding time.

Here, the same place judging unit 112, the distance between the position p1 and the position q2 at floor F _a is both positions p1 and q2 equal to or less than the predetermined threshold value is the same location, that is determined. In this case, the position correcting unit 113a, the offset O _P = q2-p1 is calculated by, for example, the above equation (1), by using the offset O _P, estimated position q2 is q2 '= q2-O corrected as _P Is done. In the PDR, the estimated position q after the position q2 ′ is calculated by accumulating displacement (direction and moving distance) from q2 ′. Therefore, if carried out position correction in q2, the correction effect is then the O _P are spread to all future estimated position until they are updated. Thus, the offset O _P and the estimated position will be updated every time a determination that the same place is performed. Thereby, accumulation of errors, which is an important problem in PDR, is appropriately suppressed.

Further, in the present embodiment, the offset O _P may not only be used as a correction amount for the future position q2 after the position q2 previous historical position, relative to the position q1 _p and position q1 in FIG. 5, for example Is also applicable. For example, the position q1 in the floor F _b, when return transition to the floor F _b is detected, the receive determines not at the same location between the past position p2, be corrected by the determination There is no. However, according to the present embodiment, based on the regression transition to the immediately following floor F _a, it is possible to obtain the opportunity of correcting retroactively. Thereby, it is possible to retroactively suppress the error accumulation.

Here, the correction of the position q1 is, for example, the following equation (3a) q1 ′ = q1−O (O _P , Δt)
It is also preferable that this is performed. Here, O (O _P , Δt) is an offset applied to the position q1. Δt is a time difference between the time at the position q2 and the time at the position q1. The offset O (O _P , Δt) is smaller than the offset O _P calculated at the position q2 by the amount retroactive (depending on Δt), on the assumption that the error increases with time in the PDR. The amount of correction is a value. Such consideration regarding the offset makes it possible to perform an appropriate correction suitable for the actual situation of the PDR. Further, the following equation as a modification, based on the distance rather than time (3b) q1 '= q1- O (O P, Δd)
It is also preferable that correction is performed by the above. Here, Δd is a total moving distance from the position q1 to the position q2. The offset O (O _P , Δd) in the above equation (3b) is moved to a position earlier than the offset O _P calculated at the position q2 on the assumption that the error does not accumulate even if time does not move. The correction amount takes a smaller value (depending on Δd) by the amount of retroactiveness.

  As described above, instead of applying the position correction to the present and subsequent (future), it is possible to correct (recalculate) the position estimated retroactively. Which form or both forms is taken may be appropriately selected according to the degree of real-time property required for the position estimation process and the importance of the past position.

  Returning to FIG. 2, the azimuth correction unit 113 c of the position correction unit 113 a proceeds in the floor transition immediately before the position q determined to be the same place and the floor transition immediately after the past position p. The estimated position is corrected using the angle difference with the direction.

  Here, the “advancing direction at the floor transition immediately before the position q” can be the direction of a vector that connects the position immediately before the floor transition read out from the storage unit 105 and the position q. In addition, “advancing direction at the floor transition immediately after the position p” can also be a vector direction connecting the position p and the position immediately after the floor transition read out from the storage unit 105 in the same manner. Alternatively, each of the change modes may be the direction (for example, the azimuth angle) of the terminal 1 at the position q and the position p. Furthermore, the direction of the vector connecting the position q and the position at the time immediately before the position q and the direction of the vector connecting the position p and the position at the time immediately after the position p are respectively set to the position q. Further, the traveling direction may be the floor transition at the position p.

  FIG. 6 is a graph for explaining an embodiment of azimuth correction and position correction in the azimuth correction unit 113c.

6, compared past locations p in the floor F _a, position q immediately after the return transition to the floor F _a with the stairs, the situation that received a determination that the same location is shown. Here, in the present embodiment, the azimuth correcting unit 113c is exactly opposite to the “traveling direction at the floor transition immediately before the position q” as the “traveling direction at the floor transition immediately after the position p”. as the difference between the azimuth angle) is 180 degrees, the azimuth offset O _omega, the following equation _{(4) O ω = θ q} - (θ p + π)
Calculate and update with Here, θ _q and θ _p in the above equation (4) are respectively an azimuth angle of “traveling direction in floor transition immediately before position q” and “immediately after position p”, as shown in FIG. It is the azimuth of the “traveling direction at the floor transition”.

Then, (the position correction unit in 113a) heading correction unit 113c is when the calculated azimuth offset O _omega, azimuth theta _q obtained by correcting the azimuth angle theta _q ', for example, the following equation (5) theta _q' = Θ _q −O _ω
Calculate and update with Next, using the corrected azimuth angle θ _q ′, the position q ^ (two-dimensional vector) next to the position q is expressed by, for example, the following equation (6) q ^ = q + [l · sin (θ _q ′) ), l ・ cos (θ _q ')]
Calculate and update with Here, l in the above equation (6) is the movement distance at the time at the position q. This expression can also be regarded as an expression in which the next position is calculated from the azimuth angle and the movement distance by the PDR, in which the azimuth angle θq is simply replaced with θ _q ′. That is, the position q itself is not corrected by this equation. Unlike the position offset O _P , the azimuth offset O _ω is not obtained cumulatively. Therefore, the correction of the azimuth according to the above equation (5) is performed after the calculation of the azimuth offset O _ω . Need to be applied to the location.

Furthermore, in the present embodiment, the azimuth offset O _omega, not only used as a correction amount of the future, similarly to the offset O _P position, may be applied retrospectively. For example, using the position q2 azimuthal offset O _omega calculated in shown in FIG. 5, if the position q1 which retroactively correct the azimuth correction of the azimuth angle, for example, the following equation (7a ) Θ _q1 '= θ _q1 −O (O _ω , Δt)
Can be implemented. Here, Δt is a time difference between the time at the position q2 and the time at the position q1 retroactive. The azimuth offset O (O _ω , Δt) in the above equation (7a) is also smaller than O _ω by the amount retroactive from the time at the position q2 where the azimuth offset O _ω is calculated (depending on Δt). The amount of correction is Further, as a change mode, the azimuth angle is corrected by the following equation (7b) θ _q1 ′ = θ _q1 −O (O _ω , Δd)
You may carry out using. Here, Δd is the total movement distance from the retroactive position q1 to the position q2. The azimuth offset O (O _ω , Δd) in the above equation (7b) is also a correction amount that takes a value smaller than O _ω by the amount traced back from the position q2 to the previous position (depending on Δd). .

  In general, as a major factor for accumulating positioning errors in the PDR, there is an error in the estimated value of the direction of the positioning object, and the error in the estimated position increases according to the magnitude of the error in the direction as the positioning object progresses. The phenomenon is mentioned. On the other hand, as described above, by correcting the azimuth angle of the terminal 1 and correcting the estimated position of the terminal 1 using the corrected azimuth angle, an increase in position error accompanying such progress is prevented or It can be suppressed.

The position correction unit 113a
(A) using the offset O _P position, for example, be corrected position q by the above equation (2),
(B) Using the azimuth offset _Oω , the position q may be corrected by, for example, the above equations (5) and (6), or
It is also possible to correct the position q using both the offset O _P and azimuth offset O _omega of (c) position.

As described above, according to the position correction unit 113a, when the mobile terminal 1 (the pedestrian who owns the mobile terminal 1) is moving, when the floor transition is detected and it is determined that it is the same place, the floor correction is performed. the position or information related to the position according (traveling direction during transition), may be performed to correct the estimated position to generate an offset O _P and / or azimuth offset O _omega position. As a result, it is possible to suppress accumulation of positioning errors, which is a problem in PDR, without relying on map information.

  Returning to FIG. 2, information on the position of the mobile terminal 1 with higher accuracy estimated and corrected by the position estimation unit 113 and information on the past movement route (time-series history information on the estimated position) are displayed on, for example, a display. May be displayed in a predetermined format on the display unit 106. Alternatively, in order to process these pieces of information by an external information processing apparatus, it is also preferable that the information is transmitted to the information processing apparatus via the communication unit 107.

  As described above in detail, according to the present invention, the transition of the floor is detected based on the information measured each time, and the correction amount for the estimated position based on the position information related to the detected floor transition. Is calculated. Further, it is possible to determine an estimated position with higher accuracy every moment using this correction amount. As a result, it is possible to suppress accumulation of positioning errors, which is a problem in PDR, without relying on map information. In addition, since the correction amount can be calculated as appropriate during movement involving floor transition, for example, during walking, it is possible to perform dynamic correction of the estimated position during movement.

  Furthermore, the present invention makes it possible to appropriately correct the estimated position of the apparatus using, for example, only a sensor built in the apparatus as one embodiment. Therefore, as a position estimation device according to the present invention, for example, a general-purpose user interface device such as a smartphone, a wearable terminal, a tablet computer, an electronic book, or a PDA (Personal Digital Assistant), which has an acceleration sensor, a geomagnetic sensor, etc. It is also possible to do.

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

1 Mobile terminal (position estimation device)
Reference Signs List 101 acceleration sensor 102 gyro sensor 103 geomagnetic sensor 104 barometric pressure sensor 105 position / floor transition information storage unit 106 display unit 107 communication unit 111 floor transition detection unit 112 same location determination unit 113 position estimation unit 113a position correction unit 113b corner detection unit 113c Correction unit

Claims (10)

A position estimation device that estimates a position of a mobile subject in an area movable between a plurality of floor areas with a predetermined means for transition , and holds the position in time series,
Floor transition detection means for detecting a transition of a floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
Identifies whether or not the position when a regression transition, which is a movement from another floor area to a floor area located in the past, is detected is the same location as the retained past position in the floor area. A place determination means;
A position correction unit that calculates a correction amount based on information on the position determined to be the same place and information on the past position, and corrects the estimated position using the correction amount; A position estimation device characterized by the above.   The same location determination means determines that the location is the same when at least the distance between the location where the regression transition is detected and the past location is less than or less than a predetermined threshold. The position estimation apparatus according to claim 1. Determining the same position determining unit includes at least a detected transition for means of the regression transition, and the transition for means of the detected transition immediately following the past position is the same place if the same The position estimation apparatus according to claim 1 or 2, wherein A corner detecting means for detecting a corner in movement in the floor area;
The said same place determination means makes the position when the said regression transition is detected the corner detected after the said regression transition is detected, The any one of Claim 1 to 3 characterized by the above-mentioned. Position estimation device.   The position correction unit corrects the estimated position using at least a position difference between the position determined to be the same place and the past position. The position estimation apparatus according to item 1.   The position correction means calculates an estimated position using at least an angular difference between a traveling direction in a transition immediately before the position determined to be the same place and a traveling direction in a transition immediately after the past position. The position estimation apparatus according to claim 1, wherein correction is performed.   7. The position correction unit according to any one of claims 1 to 6, wherein the position to be corrected is also a held position that is past the position determined to be the same place. Position estimation device.   The floor transition detection means detects a transition of a floor area based on measurement result information of acceleration, angular velocity, geomagnetism, and / or atmospheric pressure as information relating to the measured movement. 8. The position estimation apparatus according to any one of 7 above. A position estimation program for estimating a position of a mobile subject in an area movable between a plurality of floor areas with a predetermined means for transition , and causing a computer mounted on a device to hold the position in time series to function.
Floor transition detection means for detecting a transition of a floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
Identifies whether or not the position when a regression transition, which is a movement from another floor area to a floor area located in the past, is detected is the same location as the retained past position in the floor area. A place determination means;
The computer functions as a position correction unit that calculates a correction amount based on the information related to the position determined to be the same place and the information related to the past position, and corrects the estimated position using the correction amount. A position estimation program characterized in that A position estimation method in a computer mounted on a device that estimates a position of a mobile subject in an area movable between a plurality of floor areas with a predetermined means for transition , and holds the position in time series,
Detecting a transition of the floor area, which is a movement to a different floor area , based on the measured information relating to the movement ;
A step of determining whether or not a position when a regression transition that is a movement from another floor area to a floor area located in the past is detected is the same place as the retained past position in the floor area. When,
A step of calculating a correction amount based on the information related to the position determined to be the same place and the information related to the past position, and correcting the estimated position using the correction amount. A position estimation method.

Images