JP6568774B2 - Position estimation device - Google Patents

Description

  The present invention relates to a position estimation device used for information distribution.

  In recent years, in commercial facilities and the like, information distribution systems that distribute contents such as store coupons in accordance with the position of a user's mobile terminal (hereinafter sometimes referred to as a user terminal) have become widespread (for example, patents). Reference 1). In such an information distribution system, it is important to estimate the position of the user terminal with high accuracy and distribute content (for example, a coupon at a store near the user's position) according to the position of the user terminal. .

JP 2013-186760 A

  The position of the user terminal is estimated based on the installation position information of the radio wave device that has received a radio wave such as a beacon and has received the radio wave. Here, when a radio wave device is used, radio waves from a radio wave device far away from the user terminal may be received by the user terminal due to radio wave reflection or the like. In this case, the position of the user terminal may be derived based on the position of the radio wave device far away, and in this case, content according to the actual position of the user terminal cannot be distributed.

  As a technique for estimating the position of the user terminal with high accuracy, there is a technique for determining whether or not the estimated position is an error based on the map information and the average walking speed of the user. In the method that considers the map information, for example, when the estimated position is not on a map passage (a place where the user can pass), the position can be determined as an erroneous estimation. Moreover, in the method which considers a user's average walking speed, the position which a user terminal cannot reach in consideration of a user's average walking speed can be determined as an erroneous estimation. However, for example, in a method that considers map information, if a position that should be determined to be erroneously estimated is on a map path, the position cannot be determined to be erroneously estimated. In addition, in the method that considers the average walking speed of the user, for example, it is considered that the walking speed of the user changes depending on the degree of congestion in the commercial facility. Further, even if the estimated position is not the actual position of the user terminal, the map information and the average walking speed described above are not used unless the position is extremely far from the position estimated so far. Even if it considers, it is difficult to determine the said position as an incorrect estimation. Thus, conventionally, it has not been possible to appropriately determine whether or not the estimated position is the actual position of the user terminal, and it has been difficult to estimate the position of the user terminal with high accuracy.

  This invention is made | formed in view of the said situation, and it aims at providing the position estimation apparatus which estimates the position of a user terminal with high precision.

  A position estimation apparatus according to an aspect of the present invention is a position estimation apparatus that estimates a position of a user terminal based on position information acquired in a user terminal, and includes position information of a node indicating an abstracted position, and Node link data storage means for storing a plurality of link costs for each user attribute information of links connecting nodes, and each link cost stored in the node link data storage means for correlation between the link and user attribute information Correction means for correcting based on, acquisition means for continuously acquiring position information of the user terminal acquired at the user terminal and user attribute information relating to the user terminal from the user terminal, and continuous by the acquisition means On the basis of the location information of the user terminal acquired by the node and the location information of the node stored in the node link data storage means The position estimation means for continuously estimating the node indicating the position of the user terminal and the link cost of the link between the nodes continuously estimated by the position estimation means according to the user attribute information acquired by the acquisition means Deriving means for deriving the link cost from the link cost stored in the node link data storage means, and when the link cost derived by the deriving means is a predetermined value, the estimation of the node related to the link cost is incorrect Determination means for determining that it is.

  In this position estimation device, a link cost is stored for each attribute information of each user for each link. Then, the link cost is derived according to the stored information, the link related to the node estimated based on the position information acquired from the user terminal, and the user attribute information acquired from the user terminal. That is, among the stored link costs of the link, the link cost corresponding to the user attribute information acquired from the user terminal is derived as the link cost of the link. Here, in this position estimation device, each link cost is corrected based on the correlation between the link and the user attribute information. When the corrected link cost is a predetermined value, it is determined that the estimation of the node related to the link cost is an error. That is, it is determined whether or not the node estimation is erroneous in consideration of the correlation between the link and the user attribute information. Accordingly, when a node related to a route (link) that is considered difficult to be selected by the user when the user attribute is considered, it is possible to determine that the estimation of the node is an error. That is, by taking into account user attributes, it is possible to determine with high probability whether or not the estimated position is the actual position of the user terminal, and it is possible to estimate the position of the user terminal with high accuracy. .

  The apparatus further includes a history storage unit that stores movement histories of a plurality of user terminals in each link, and the history storage unit stores user attribute information related to each user terminal as information included in the movement history, and a correction unit. May derive the number of movements for each user attribute information of each link from the movement history stored in the history storage means, and determine the correlation between each link and user attribute information based on the number of movements. Thereby, the correlation of each link and user attribute information is determined based on a user's actual movement history, and each link cost is corrected. Accordingly, it is possible to determine that the estimation of the node related to the link that is difficult for the user to select is erroneous based on the past actual data. Thereby, the position of the user terminal can be estimated with higher accuracy.

  Further, the correction means may further correct each link cost stored in the node link data storage means according to the degree of congestion determined based on the number of movements in the latest predetermined time. Users are difficult to move on a busy link. For this reason, the link cost is corrected in consideration of the degree of congestion, so that if the link is congested but it is passing in a short period of time, the estimation of the node related to the link is incorrect. It can be determined that This makes it possible to estimate the position of the user terminal with higher accuracy.

  Further, it further comprises history storage means for storing movement histories of a plurality of user terminals in each link, the history storage means stores date and time information passing through the link as information included in the movement history, and the correction means includes: The number of movements of each link within the latest predetermined time may be derived from the movement history stored in the history storage unit, and the degree of congestion described above may be specified. By considering the number of movements of each link within the latest predetermined time (for example, the latest 10 minutes), it is possible to appropriately grasp the congestion status of the links. This makes it possible to estimate the position of the user terminal with higher accuracy.

  The node link data storage means stores the link road width as information about the link, and the correction means further corrects the link cost of each link stored in the node link data storage means according to the road width. May be. On links with narrow roads, it is difficult for the user to move. For this reason, the link cost is corrected in consideration of the road width, so that when the road is narrow but the road is passing in a short period, the estimation of the node related to the link is incorrect. Can be determined. This makes it possible to estimate the position of the user terminal with higher accuracy.

  Further, the display device further includes display means for displaying the nodes continuously estimated by the position estimation means and the links between the nodes in association with the map information, and the display means is determined to be erroneous estimation by the determination means. The nodes may be displayed separately from other nodes. This makes it easier to visually grasp the node determined as an erroneous estimation.

  In addition, when it is determined that the estimation of the node related to the link cost is incorrect by the history storage unit that stores the movement history of a plurality of user terminals in each link and the determination unit, the estimation is determined to be incorrect. And a position correction unit that corrects the position indicated by the node according to the movement history stored in the history storage unit and outputs the corrected position as the position of the user terminal. As described above, when the estimated node is considered to be an error, the current position of the user terminal is output by outputting the position corrected based on the movement history of the user terminal (for example, the latest position) to the user terminal. Can be output to the user terminal.

  Here, although the above-described advantages are obtained by correcting the position based on the movement history of the user terminal, the above-described position correction may not be appropriately performed under a predetermined condition. That is, in the position correction by the position correction means, for example, since the position correction is performed based on the position (node) of the latest user terminal stored in the movement history, the position of the latest user terminal is incorrect, In addition, when the position of the latest portable user terminal and the position of the user terminal newly estimated by the position estimation means are greatly separated, the position newly estimated by the position estimation means is the correct position. Regardless, there is a possibility that the position of the user terminal is corrected to the wrong nearest position.

  In this regard, the position estimation apparatus according to an aspect of the present invention further includes a reliability calculation unit that calculates the reliability of the node estimated by the position estimation unit, and when the reliability is equal to or higher than a predetermined value, The link cost is not derived by the deriving means, and the node estimated by the position estimating means is output as the position of the user terminal. If the reliability is not equal to or higher than a predetermined value, the link cost is derived by the deriving means. Done. As described above, when the reliability is equal to or higher than the predetermined value, the link cost is not derived by the deriving unit, and the position correction by the position correcting unit is not performed. That is, when there is a high possibility that the position estimated by the position estimation unit is the correct position, position correction according to the movement history by the position correction unit is not performed, so the past position of the mobile terminal 20 in the movement history The correct position can be output as the position of the user terminal without being affected by the inaccuracy of the user terminal. For example, it is possible to prevent an incorrect position correction from being performed based on the position information of the erroneous user terminal stored most recently, and the position of the user terminal can be estimated with higher accuracy.

  According to the present invention, the position of the user terminal can be estimated with high accuracy.

It is a figure which shows the function structure of the position estimation apparatus which concerns on embodiment of this invention. It is a figure which shows the hardware constitutions of the position estimation apparatus shown in FIG. It is a figure which shows an example of the information memorize | stored in the node link data storage part, Fig.3 (a) has shown an example of node data, FIG.3 (b) has shown an example of link data, respectively. It is a figure which shows an example of the movement history data memorize | stored in the log | history memory | storage part. It is a figure which shows the display image in a display apparatus of the information output by the display part. It is a flowchart which shows the position estimation process of the position estimation apparatus which concerns on embodiment of this invention. It is a flowchart which shows the link cost update process which concerns on embodiment of this invention. It is a flowchart which shows the link cost update process which concerns on the modification of this invention. It is a figure which shows the function structure of the position estimation apparatus which concerns on the modification of this invention. It is a flowchart which shows the position estimation process of the position estimation apparatus which concerns on the modification of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same reference numerals are used for the same or equivalent elements, and redundant descriptions are omitted.

  FIG. 1 is a diagram illustrating a functional configuration of the position estimation apparatus according to the present embodiment. As shown in FIG. 1, the position estimation device 10 is included in the position estimation system 1. The position estimation system 1 includes a position estimation device 10, a portable terminal 20 (user terminal) that can communicate with the position estimation device 10, a radio wave device 30, and a display device 40. The position estimation system 1 is used, for example, for position estimation of the mobile terminal 20 in a store (commercial facility) such as a supermarket. The estimated position of the mobile terminal 20 is used, for example, for content distribution according to the position of the mobile terminal 20.

  The position estimation device 10 is a server that estimates the position of the mobile terminal 20 based on position information acquired by the mobile terminal 20. Details of the position estimation apparatus 10 will be described later. The display device 40 is a display that displays information output by the position estimation device 10. Details of the display device 40 will be described later.

  The radio wave device 30 is a radio wave device that transmits radio waves that can be received by the mobile terminal 20. For example, the radio wave device 30 transmits a BLE (Bluetooth (registered trademark) Low Energy) signal, an optical beacon signal, or a WiFi (Wireless Fidelity) (registered trademark) signal as a receivable radio wave. In the following description, it is assumed that the radio wave device 30 transmits a BLE signal. A plurality of radio wave devices 30 are provided in advance in a range where the mobile terminal 20 can receive a BLE signal. For example, in the case of a supermarket or the like, a plurality of radio wave devices 30 are provided on a merchandise display shelf or the like that is in the vicinity of a location where a user carrying the mobile terminal 20 passes. The BLE signal transmitted by the radio wave device 30 includes information for uniquely identifying the radio wave device 30 and information indicating the reception intensity of the radio wave. As information for uniquely identifying the radio wave device 30, a BSSID (Basic Service Set Identifier) preset in each radio wave device 30 can be used. BSSID is also generally called a MAC address. As information indicating the reception intensity of radio waves, RSSI (Received Signal Strength Indication, Received Signal Strength Indicator) can be used.

  The portable terminal 20 is a device that acquires position information of the portable terminal 20 based on radio waves from the radio wave device 30. The portable terminal 20 is used (carried) by the user. The mobile terminal 20 is, for example, a mobile phone such as a smart phone or a tablet terminal. The portable terminal 20 acquires position information continuously (continuously) repeatedly. Specifically, the mobile terminal 20 receives the BLE signal from the radio wave device 30 and acquires the location information of the terminal itself. For example, the portable terminal 20 acquires the BSSID included in the BLE signal as the position information (information indicating the position) of the own terminal. That is, the portable terminal 20 can acquire the position of the radio wave device 30 specified by the BSSID as the position information of the own terminal by acquiring the BSSID. Note that the mobile terminal 20 may receive BLE signals from a plurality of radio wave devices 30 simultaneously (or continuously at a short time interval). In this case, the mobile terminal 20 may use, for example, the BSSID related to the BLE signal having the largest RSSI as the position information of the own terminal. The mobile terminal 20 estimates the acquired location information (ie, BSSID) together with information indicating the date and time when the location information was acquired, and a user ID that uniquely identifies the user who is the owner of the mobile terminal 20. Transmit to device 10. The transmission may be performed every time position information is acquired, or may be performed at predetermined time intervals.

  In addition, the mobile terminal 20 stores demogram (demographic) information (user attribute information) of the user who is the owner. Demogram information is information indicating demographic attributes, and is socio-economic characteristics data possessed by the user, such as gender, age, area of residence, income, occupation, educational background, and family composition. The portable terminal 20 transmits the demogram information of the user to the position estimation device 10 together with the position information described above. More specifically, the mobile terminal 20 transmits a demogram ID, which is an ID number for uniquely identifying the demogram information indicating the demogram information, to the position estimation device 10.

  Next, details of the function of the position estimation apparatus 10 will be described with reference to FIG. As shown in FIG. 1, the position estimation apparatus 10 includes a node link data storage unit 11 (node link data storage unit), a correction unit 12 (correction unit), a history storage unit 13 (history storage unit), a user An information acquisition unit 14 (acquisition unit), a position estimation unit 15 (position estimation unit), a derivation unit 16 (derivation unit), a determination unit 17 (determination unit), and a display unit 18 (display unit) are provided. ing.

  FIG. 2 is a diagram illustrating a hardware configuration of the position estimation device 10 illustrated in FIG. 1. As shown in FIG. 2, the position estimation apparatus 10 physically includes one or a plurality of CPUs 101, a RAM 102 and a ROM 103 which are main storage devices, an input device 104 such as a keyboard and a mouse which are input devices, and an output such as a display. The computer 105 includes a device 105, a communication module 106 that is a data transmission / reception device such as a network card, an auxiliary storage device 107 such as a semiconductor memory, and the like.

  Each function of the position estimation apparatus 10 is configured such that one or a plurality of predetermined computer software is read on hardware such as the CPU 101 and the RAM 102 shown in FIG. 2, so that the input apparatus 104 and the output apparatus 105 are controlled under the control of the CPU 101. This is realized by operating the communication module 106 and reading and writing data in the RAM 102 and the auxiliary storage device 107.

  Returning to FIG. 1, the node link data storage unit 11 is a database that stores a plurality of pieces of node position information indicating the abstracted positions and a plurality of pieces of information related to links connecting the nodes. A node is an abstraction of a position in a building having a certain range or an outdoor position. For example, in the store as shown in FIG. 5, the area in front of the shelf at each product corner is one node N. One node N is a position associated with the position of one radio wave device, for example. A link (connection relationship, connection) L is set between the nodes N. The link L is set based on whether the user can move, for example. That is, a path that is a path between the nodes N and allows the user to move is set as the link L.

  FIG. 3 is a diagram illustrating an example of information stored in the node link data storage unit 11, FIG. 3A is an example of node data 111, and FIG. 3B is an example of link data 112. Show. As shown in FIG. 3A, in the node data 111, a node ID, a central latitude / longitude, a radius (m), and a BSSID are associated with each node. The node ID is an ID number for uniquely identifying the node. The center latitude / longitude is information (node position information) indicating the latitude and longitude of the position of the center of the node. The radius (m) is information indicating the radius from the center of the node indicating the range of the node. BSSID is the BSSID of the radio wave device associated with the node. Note that the radio wave device is not necessarily associated with the node. In this case, in the node data 111, there is a node ID that is not associated with the BSSID.

  As shown in FIG. 3B, in the link data 112, a link ID, a node ID, an adjacent node ID, a demogram ID, an Euclidean distance (m), a correlation weight, and a link cost are associated with each link. It has been. The link ID is an ID number for uniquely identifying the link. The node ID is an ID number for uniquely identifying one node related to the link. The adjacent node ID is an ID number for uniquely specifying the other node related to the link that is adjacent to the node indicated by the node ID. The demogram ID is an ID number for uniquely identifying demogram information. The Euclidean distance (m) is information indicating the distance of the link indicated by the link ID, that is, the distance (m) between the central latitude and longitude of the two nodes indicated by the node ID and the adjacent node ID. The correlation weight is a value indicating the correlation between the link indicated by the link ID and the demogram information indicated by the demogram ID. The correlation weight is set to a larger value as the correlation between the link and the demogram information is higher. The link cost is a value indicating the difficulty of selecting a link, which is determined according to the Euclidean distance and the correlation weight (details will be described later).

  As described above, the link cost for each demogram information of each link is determined according to the Euclidean distance. More specifically, the link cost for each demogram information of each link is determined according to the Euclidean distance and the correlation weight. The link cost is a value indicating the difficulty of selecting a link. The link cost increases as the Euclidean distance increases. This is because the node indicated by the node ID and the adjacent node ID related to the link continues in the environment where the position information is continuously acquired as the Euclidean distance is large, that is, the length of the link is long. It is based on the idea that it is difficult to become a position. The link cost increases as the correlation weight is smaller. This is based on the idea that the lower the correlation weight, that is, the lower the correlation between the link and the demogram information, the lower the probability that the user related to the demogram information will select the link. As described above, the node link data storage unit 11 stores the link cost determined according to the Euclidean distance and the correlation weight of the link for each demogram information for each link as information about the link.

  Returning to FIG. 1, the correction unit 12 corrects each link cost stored in the node link data storage unit 11 based on the correlation between the link and the demogram information. Specifically, the correction unit 12 corrects each link cost so that the link cost of the demogram information having a low correlation with the link becomes higher. Specifically, the correction unit 12 derives the number of movements (total number of movements) for each demogram information of each link from the movement history stored in the history storage unit 13. Further, the correction unit 12 regards the combination of the link and the demogram information having a smaller number of movements as having a lower correlation with each other, and reduces the correlation weight stored in the node link data storage unit 11. As a result, each link cost is corrected so as to increase as the link cost of the demogram information having a low correlation with the link. For example, the correction unit 12 performs the above correction every time the movement history of the history storage unit 13 is updated. Note that the correction unit 12 may perform the correction at predetermined time intervals.

  The history storage unit 13 stores movement histories of a plurality of mobile terminals 20 in each link. The history storage unit 13 stores a demogram ID indicating demogram information related to each mobile terminal 20 as information included in the travel history. The history storage unit 13 stores, as the movement history, information related to the node where the mobile terminal 20 is stored, which is stored by the position estimation unit 15 (details will be described later).

  FIG. 4 is a diagram illustrating an example of the movement history data 131 stored in the history storage unit 13. As shown in FIG. 4, in the movement history data 131, a user ID, a demogram ID, a node ID, a previous node ID, a link ID, a date and time, and an abnormal value flag are associated with each movement history. The user ID is an ID number for uniquely identifying the user. The demogram ID is an ID number for uniquely identifying demogram information related to the user indicated by the user ID. The node ID is an ID number for uniquely identifying a node (details will be described later) estimated by the position estimation unit 15 as a node where the mobile terminal 20 exists. The immediately preceding node ID is an ID number for uniquely identifying the node estimated by the position estimating unit 15 immediately before (one before) as the node where the mobile terminal 20 exists. The link ID is an ID number for uniquely specifying the link between the nodes indicated by the node ID and the immediately preceding node ID. The date and time is the date and time when the position information of the mobile terminal 20 used for the node estimation by the position estimation unit 15 is acquired. The abnormal value flag is a flag indicating whether or not the node estimated by the position estimation unit 15 indicated by the node ID is an erroneously estimated node. An error in node estimation means that the position information of the mobile terminal 20 used for the estimation of the node is not the actual position of the mobile terminal 20. As the abnormal value flag, “1” is input when the node is erroneously estimated. The abnormal value flag is updated by the determination unit 17 (details will be described later).

As described above, the correction unit 12 derives the number of movements for each demogram information of each link from the movement history stored in the history storage unit 13, and the link of the link and the demogram information with the smaller number of movements indicates a node link. The correlation weight stored in the data storage unit 11 is reduced. For example, in the movement history data 131 shown in FIG. 4, the number of movements (number of movement histories) is “3” for the combination of the link ID “15” and the demogram ID “3”. The number of movements is “2” for the combination of the link ID “16” and the demogram ID “3”. In this case, the correction unit 12 sets the correlation weight of the combination of the link ID “16” having the movement number “2” and the demogram ID “3”, the link ID “15” having the movement number “3”, and the demogram. The correlation weight of the combination of ID “3” is made smaller. Specifically, the correction unit 12 derives the correlation weight according to the following equation (1). In the following equation (1), h is the correlation weight of a certain link and demogram information combination, and Ty is the total number of movements in the movement history data 131 of FIG. LDy indicates the number of movements of the combination of the link and demogram information, and α indicates a weighting coefficient that can be arbitrarily changed.
h = 1 + Ty / LDy * α (1)

The correction unit 12 inputs the correlation weight derived from the above-described equation (1) to the link data 112 in FIG. Further, the correction unit 12 updates the link cost for each demogram information of each link from the Euclidean distance and the correlation weight shown in FIG. Specifically, the correction unit 12 updates the link cost according to the following equation (2). In the following equation (2), Lc represents the link cost of a certain link and demogram information combination, Ed represents the Euclidean distance of the above link, and h represents the correlation weight of the certain link and demogram information combination. Yes.
Lc = Ed * h (2)

  The user information acquisition unit 14 continuously receives the position information (that is, BSSID) of the mobile terminal 20 acquired in the mobile terminal 20 and the demogram information (that is, demogram ID) related to the mobile terminal 20 from the mobile terminal 20. get. More specifically, the user information acquisition unit 14 includes, as the information (user information) of the user who is the owner of the mobile terminal 20, information indicating the date and time when the position information was acquired, in addition to the position information and demogram information, and User IDs are acquired continuously. The user information acquisition unit 14 outputs the user information to the position estimation unit 15.

  The position estimation unit 15 is based on the user information of the mobile terminal 20 continuously acquired by the user information acquisition unit 14 and the node data 111 and link data 112 stored in the node link data storage unit 11. The node where 20 is located (exists) is continuously estimated. The position estimation unit 15 may perform node estimation every time user information is input from the user information acquisition unit 14, or may collectively estimate a plurality of nodes at predetermined time intervals. In the following description, it is assumed that the position estimation unit 15 estimates a node every time user information is input.

  When receiving the user information input from the user information acquisition unit 14, the position estimation unit 15 refers to the node data 111 in the node link data storage unit 11. Then, the position estimation unit 15 identifies a node ID associated with a BSSID that matches the BSSID included in the user information in the node data 111. Further, the position estimation unit 15 refers to the movement history data 131 of the history storage unit 13 and, based on the user ID and information indicating the date and time included in the user information, as the node ID of the user indicated by the user ID The immediately preceding node ID indicating the node specified immediately before (immediately before) is specified. Then, the position estimation unit 15 refers to the link data 112 of the node link data storage unit 11, and indicates a link related to the combination of the node ID and the adjacent node ID that matches the above-described combination of the node ID and the immediately preceding node ID. Specify the ID. The position estimation unit 15 displays a movement history in which information indicating the user ID, demogram ID, and date / time included in the user information is associated with the identified node ID, immediately preceding node ID, and link ID. Stored in the movement history data 131. In the movement history data 131, the abnormal value flag is associated with the link ID or the like. However, at the stage where the movement history is stored by the position estimation unit 15, the abnormal value flag is set to a null value. . After the above processing, the position estimating unit 15 outputs the identified link ID and the demogram ID included in the user information to the deriving unit 16.

  The derivation unit 16 determines the link cost of the link between nodes continuously estimated by the position estimation unit 15 and the link cost according to the demogram information acquired by the user information acquisition unit 14, as the node link data storage unit 11. Derived based on the link data 112. That is, when receiving the link ID and the demogram ID from the position estimation unit 15, the deriving unit 16 refers to the link data 112 and identifies the link cost associated with the combination of the link ID and the demogram ID. The deriving unit 16 outputs the link cost specified from the combination of the link ID and the demogram ID to the determination unit 17.

  When the link cost specified (derived) by the derivation unit 16 is equal to or greater than a predetermined threshold, the determination unit 17 determines that the estimation of the node related to the link cost is an error. That is, when the determination unit 17 receives an input of the link cost from the derivation unit 16, the determination unit 17 determines whether the link cost is equal to or higher than a predetermined threshold value. When the link cost is equal to or higher than a predetermined threshold, the determination unit 17 refers to the movement history data 131 in the history storage unit 13 and sets the abnormal value flag of the movement history with the latest date to “1” (abnormal value flag). Stand up). The determination unit 17 may update the abnormal value flag in response to the input of information (for example, date and time) specifying the movement history for updating the abnormal value flag from the position estimation unit 15. The determination unit 17 indicates each piece of information (user ID, demogram ID, node ID, previous node ID, link ID, and date / time) related to the movement history with the abnormal value flag set to “1”, and the node ID in the movement history. Information indicating that the estimated node is incorrect is transmitted to the mobile terminal 20 and an external device (not shown). The external device is a device that distributes content according to the position of the mobile terminal 20, for example. When it is determined that the node estimation is incorrect, in the mobile terminal 20, for example, the most recent node that is determined to have no node estimation error is set as the position of the mobile terminal 20. On the other hand, when the link cost is not equal to or higher than the predetermined threshold, the determination unit 17 does not update the abnormal value flag of the movement history with the latest date and time. And the determination part 17 transmits each information (user ID, demogram ID, node ID, immediately preceding node ID, link ID, and date and time) regarding the said movement history to the portable terminal 20 and an external device (not shown). Thereby, for example, in the external device, the content associated with the node indicated by the node ID is distributed.

  The display unit 18 displays the nodes continuously estimated by the position estimation unit 15 and the links between the nodes on the display device 40 in association with a map (map information). For example, the display unit 18 acquires a map of each product corner in the store from a map DB (not shown), and displays the position of the node and the link in association with the map. When the display unit 18 displays the nodes in association with the map, the display unit 18 may display the nodes determined to be erroneous estimation by the determination unit 17 separately from the other nodes. The display unit 18 performs the distinguished display by referring to the movement history data 131 of the history storage unit 13.

  FIG. 5 is a diagram showing a display image of the information output by the display unit 18 on the display device 40. The display unit 18 receives designation of the start point and end point of the node and demogram information (demogra ID) from, for example, an operator of the position estimation apparatus 10. The display unit 18 that has received the designation refers to the movement history data 131 in the history storage unit 13, and the node through which the user associated with the designated demogram ID passes from the start point to the end point of the designated node To identify the most frequent node combinations. For example, FIG. 5 shows an example in which a node N3 is designated as the node start point, a node N9 is designated as the node end point, and a man in his 30s is designated as the demogram ID. The display unit 18 refers to the movement history data 131, and the nodes N4, N5, N6 are the nodes through which the user of the demogram ID indicating a male in his 30s passes most frequently from the node N3 to the node N9. N7 and N8 are specified. In this case, the display unit 18 displays the nodes N3, N4, N5, N6, N7, N8, and N9 and the links L15, L16, L17, L18, L19, and L20 that are links between the nodes on the map Mi. Highlight.

  Further, the display unit 18 refers to the movement history data 131 and identifies a node related to a movement history in which an abnormal value flag is set among movement histories of males in their thirties from the node N3 toward the node N9. Further, the display unit 18 is a node corresponding to the movement history in which the abnormal value flag is set, from the movement history before and after the user of the movement history in which the abnormal value flag is set, and is assumed to be a node Is identified. The display unit 18 displays the node in distinction from other nodes when an operator selects a node that is assumed to exist as described above, and an abnormal value flag is set. The nodes related to the movement history are further distinguished and displayed. Thereby, when the user is located at a certain node, it is possible to distinguish and display a node that is likely to be estimated as incorrect position information. In the example shown in FIG. 5, when the node N4 is selected by the operator, the node N4 is displayed separately from the other nodes N3, N5, N6, N7, N8, and N9, and a user is displayed on the node N4. Nodes N12 and N15, which are nodes that are likely to be estimated as erroneous position information when positioned, are further distinguished and displayed. The node N12 is a cosmetic corner node, and is a node that is unlikely to be passed by a man in his thirties. The node N15 is a node that is unlikely to be passed by the user in consideration of the Euclidean distance from the preceding and following nodes. The nodes N12 and N15 may be further distinguished from each other according to the frequency of erroneous estimation.

  Next, the position estimation process of the position estimation apparatus 10 will be described. FIG. 6 is a flowchart showing the position estimation process of the position estimation apparatus 10.

  In the position estimation process, first, user information is acquired from the mobile terminal 20 by the user information acquisition unit 14 (step S1). The user information includes location information of the mobile terminal 20, user demogram information, information indicating the date and time when the location information was acquired, and a user ID.

  Subsequently, the position estimation unit 15 estimates the position (node) of the mobile terminal 20 (step S2). The position estimation unit 15 is based on the user information of the mobile terminal 20 continuously acquired by the user information acquisition unit 14 and the node data 111 and link data 112 stored in the node link data storage unit 11. The node where 20 is located (exists) is continuously estimated.

  Subsequently, the link cost corresponding to the demogram information acquired by the user information acquisition unit 14 that is the link cost of the link between nodes continuously estimated by the position estimation unit 15 by the deriving unit 16 is derived (Step S1). S3). And it is determined by the determination part 17 whether a link cost is more than a predetermined threshold value (step S4).

  If it is determined in step S4 that the link cost is equal to or higher than a predetermined threshold, information indicating that the estimation of the node related to the link cost is incorrect is output (transmitted) to the mobile terminal 20 and the external device. (Step S5). On the other hand, when it is determined in step S4 that the link cost is not equal to or higher than the predetermined threshold, the node ID indicating the node related to the link cost is output (transmitted) to the mobile terminal 20 and the external device (step S6). ). The above is the position estimation process.

  Next, the link cost update process of the position estimation device 10 will be described. FIG. 7 is a flowchart illustrating an example of the link cost update process.

  In the link cost update process, first, node data 111 and link data 112 stored in the node link data storage unit 11 are created (step S11). The node data 111 and the link data 112 are created, for example, when an operator manually inputs each data. In the link data 112, the correlation weight is set to “1”, for example, and the link cost is set according to the Euclidean distance (step S12).

  Subsequently, the correcting unit 12 determines whether or not the movement history of the movement history data 131 has been updated (step S13). If the movement history is not updated in step S13, the determination in step S13 is performed again after a predetermined time has elapsed. On the other hand, if the movement history is updated in step S13, the correction unit 12 derives the number of movements for each demogram information of each link based on the movement history (step S14). Then, the correction unit 12 updates the link cost for each demogram information of each link based on the number of movements (step S15). That is, the correction unit 12 decreases the correlation weight of the link data 112 for the combination of the link and the demogram information with a smaller number of movements, and the link for each demogram information of each link from the Euclidean distance and the correlation weight of the link data 112. Deriving the cost. When the process of step S15 is completed, the process of step S13 is performed again. The above is the link cost update process.

  Next, the effect of the position estimation apparatus 10 mentioned above is demonstrated.

  In the position estimation device 10, the link cost determined according to the Euclidean distance of the link is stored in the node link data storage unit 11 for each attribute information of the user for each link. Then, the link cost is derived according to the stored information, the link related to the node estimated based on the position information acquired from the mobile terminal 20, and the demogram information acquired from the mobile terminal 20. . That is, among the link costs of the link stored in the node link data storage unit 11, the link cost corresponding to the demogram information acquired from the mobile terminal 20 is derived as the link cost.

  Here, in this position estimation device 10, each link cost is corrected by the correction unit 12 so that the link cost of the demogram information having a low correlation with the link becomes higher. Then, when the link cost derived by the deriving unit 16 is equal to or greater than a predetermined threshold, the determination unit 17 determines that the estimation of the node related to the link cost is an error. That is, a node related to a link having a low correlation with the demogram information is likely to be determined as an erroneous estimation. Thereby, considering demogram information, it is possible to determine that a node related to a route (link) that is considered difficult for the user to select is erroneous estimation. That is, by considering the demogram information, it is possible to determine with high probability whether or not the estimated position is the actual position of the mobile terminal 20, and to estimate the position of the mobile terminal 20 with high accuracy. it can.

  In addition, a history storage unit 13 that stores movement histories of a plurality of portable terminals 20 in each link is further provided, and the history storage unit 13 stores demogram information related to each portable terminal 20 as information included in the movement history. The correction unit 12 derives the number of movements for each demogram information of each link from the movement history stored in the movement history data 131 of the history storage unit 13, and the combination of the link and the demogram information with the smaller number of movements indicates the mutual number. Is considered to have low correlation. As a result, the link cost related to the combination of the link having a small number of movements and the demogram information becomes higher, and the node related to the link that is difficult for the user to select can be determined to be erroneous estimation according to the past actual data. This makes it possible to estimate the position of the mobile terminal 20 with higher accuracy.

  The display unit 18 further includes a display unit 18 that displays the nodes continuously estimated by the position estimation unit 15 and the links between the nodes in association with a map (map information). Nodes determined to be erroneous estimation are displayed separately from other nodes. This makes it easier to visually grasp the node determined as an erroneous estimation. Further, by recognizing a node that is easily determined as an erroneous estimation, it is possible to appropriately and easily perform re-installation and re-tuning of the radio wave device 30 (for example, a beacon).

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment. For example, the correction unit 12 derives the number of movements for each demogram information of each link from the movement history stored in the movement history data 131 of the history storage unit 13, and the combination of the link and the demogram information having a smaller number of movements, Although it has been described that the correlation is low, the correlation between the link and the demogram information may be set in advance.

  Moreover, although the portable terminal 20 was demonstrated as acquiring position information based on the radio wave from the radio wave apparatus 30, it is not limited to this, By other positioning methods, such as GPS (Global Positioning System) or base station positioning, You may acquire position information.

  In addition, the correction unit 12 has been described as correcting so that the link cost of demogram information having a low correlation with the link increases, but the link cost may be further corrected (updated) according to another standard. . That is, the correction unit 12 assumes that the number of movements within the most recent predetermined time (for example, 10 minutes to 1 hour) is congested for each link cost stored in the link data 112 of the node link data storage unit 11. Further correction may be made so as to increase the link cost of the link to be performed. Users are unlikely to move on a busy link. For this reason, if the link cost of a link that is assumed to be congested is increased, the link will be added to the link when it is passing through the link in a short period of time despite being a congested link. It can be determined that the node concerned is an erroneous estimation. This makes it possible to estimate the position of the mobile terminal 20 with higher accuracy. In addition, when the link cost is corrected in consideration of congestion, the determination unit 17 does not move from the same node for a predetermined time, for example, even if the link cost is equal to or higher than a predetermined threshold (that is, In a case where the link cannot be moved in a short period due to the influence of congestion), it may be determined that the node related to the link is not an erroneous estimation.

  In the case of considering the congestion as described above, the correction unit 12 uses the movement history stored in the movement history data 131 of the history storage unit 13 for each link within the latest predetermined time (for example, 10 minutes to 1 hour). The number of movements may be derived and corrected so that the link cost of a link with a large number of movements increases. By considering the number of movements of each link within the most recent predetermined interval, it is possible to appropriately grasp the link congestion status. This makes it possible to estimate the position of the mobile terminal 20 with higher accuracy. In addition, when considering congestion, the degree of correction according to the congestion state of each link may be set in advance.

  FIG. 8 is a flowchart showing link cost update processing when congestion is considered. The processes in steps S111 to S115 in FIG. 8 are the same as the processes in steps S11 to S15 in FIG. When considering the congestion, after the link cost for each demogram information of each link is updated in step S115, the link cost is updated based on the movement history of the latest predetermined time (for example, 10 minutes) (step S200). . When the process of step S200 is completed, the process of step S113 is performed again. The above is the link cost update process when the congestion is taken into consideration.

  Further, the node link data storage unit 11 stores the road width of the link in the link data 112 as information relating to the link, and the correction unit 12 determines the road width for the link cost of each link stored in the node link data storage unit 11. You may correct | amend further so that the link cost of a narrow link may become high. On links with narrow roads, it is difficult for the user to move. For this reason, by increasing the link cost of a link with a narrow road width, if the link passes through the link in a short period of time even though it is a narrow link, the node related to the link may be erroneously estimated. It can be determined that there is. This makes it possible to estimate the position of the mobile terminal 20 with higher accuracy.

  Next, in order to estimate the position of the mobile terminal 20 with higher accuracy, a modification example in which the necessity of position correction is determined and position correction using a predetermined position correction method is performed only when it is determined to be necessary will be described. FIG. 9 is a diagram illustrating a functional configuration of the position estimation device 10A of the position estimation system 1A according to the modification. As illustrated in FIG. 9, the position estimation device 10 </ b> A includes a position correction unit 31 and a reliability calculation unit 32 in addition to the configuration of the position estimation device 10 described in the above embodiment.

  When the determination unit 17 determines that the estimation of the node related to the link cost is an error, the position correction unit 31 stores the position indicated by the node determined to be an error in the history storage unit 13. The correction is performed according to the movement history data 131 thus corrected, and the corrected position is set as the position of the mobile terminal 20.

  Specifically, the position correction unit 31 refers to the movement history data 131 and identifies the node (the latest node) in which the mobile terminal 20 is located before the node that is determined to be erroneous by the determination unit 17. To do. Then, the position correction unit 31 specifies the nearest node as the position of the mobile terminal 20 instead of the node determined to be erroneous in estimation. Finally, the position correction unit 31 transmits the specified corrected position of the mobile terminal 20 to the mobile terminal 20 and an external device (not shown).

  As described above, when the estimated node is considered to be an error, by outputting the position corrected based on the movement history of the mobile terminal 20 (specifically, the latest position) to the mobile terminal 20 or the like, A position that seems to be close to the current position of the mobile terminal 20 can be output to the mobile terminal 20.

  Here, although the above-mentioned advantage is obtained by using the position closest to the mobile terminal 20, the above-described position correction may not be appropriately performed under a predetermined condition. That is, in the position correction by the position correction unit 31, position correction is performed based on the position (node) of the latest mobile terminal 20 stored in the movement history data 131, so that the position of the latest mobile terminal 20 is incorrect. And when the position of the latest mobile terminal 20 and the position of the mobile terminal 20 newly estimated by the position estimation unit 15 are greatly separated, the position newly estimated by the position estimation unit 15 is In spite of the correct position, the position of the mobile terminal 20 may be corrected to the wrong nearest position. For these reasons, the position estimation device 10 </ b> A includes a reliability calculation unit 32.

なお、上記（３）式で示される関数は一例であり、信頼度算出のために用いる関数はこれに限定されない。また、推定結果のノードに設置された無線電波装置から発信された電波強度のみを変数として持つ関数を例に挙げたが、他の値を変数として用いてもよい。 The reliability calculation unit 32 calculates the reliability of the node estimated by the position estimation unit 15. For example, the reliability calculation unit 32 calculates the reliability of the node using a function having an RSSI value used for node estimation as a variable. For example, when the radio wave intensity transmitted from the radio wave device installed in the estimated node is x, the estimated reliability f (x) can be calculated by the following function.

In addition, the function shown by said Formula (3) is an example, and the function used for reliability calculation is not limited to this. Moreover, although the function having only the radio wave intensity transmitted from the radio wave device installed at the node of the estimation result as an example is given, other values may be used as the variable.

  If the reliability is equal to or higher than a predetermined value, the link cost is not derived by the deriving unit 16, and the node estimated by the position estimating unit 15 is the position of the portable terminal 20 and the external terminal 20. Sent to the device. On the other hand, when the reliability is not equal to or higher than the predetermined value, the derivation unit 16 derives the link cost and the determination unit 17 determines.

  As described above, when the reliability is equal to or higher than the predetermined value, the derivation unit 16 does not derive the link cost, and the position correction unit 31 does not perform the position correction. That is, when there is a high possibility that the position estimated by the position estimation unit 15 is the correct position, position correction according to the movement history by the position correction unit 31 is not performed, so the past of the mobile terminal 20 in the movement history is not performed. The correct position can be output as the position of the portable terminal 20 without being affected by the inaccuracy of the position of the mobile terminal 20. For example, it is possible to prevent erroneous position correction from being performed based on the position information of the wrong mobile terminal 20 stored most recently, and to estimate the position of the mobile terminal 20 with higher accuracy. it can.

  Finally, with reference to FIG. 10, the flow of the position estimation process according to the modification will be described.

  In the position estimation process, first, user information is acquired from the mobile terminal 20 by the user information acquisition unit 14 (step S201). The user information includes location information of the mobile terminal 20, user demogram information, information indicating the date and time when the location information was acquired, and a user ID.

  Subsequently, the position estimation unit 15 estimates the position (node) of the mobile terminal 20 (step S202). The position estimation unit 15 is based on the user information of the mobile terminal 20 continuously acquired by the user information acquisition unit 14 and the node data 111 and link data 112 stored in the node link data storage unit 11. The node where 20 is located (exists) is continuously estimated.

  Subsequently, the reliability of the node estimated by the position estimation unit 15 is calculated by the reliability calculation unit 32 (step S203). As described above, the reliability calculation is performed by, for example, a function having an RSSI value as a variable.

  Subsequently, the reliability calculation unit 32 determines whether or not the reliability is equal to or higher than a predetermined value (step S204). If it is determined in S204 that the reliability is equal to or higher than the predetermined value, processing by the derivation unit 16, the determination unit 17, and the position correction unit 31, which will be described later, is not performed, and the position calculation unit 15 is performed by the reliability calculation unit 32. Is output (transmitted) to the portable terminal 20 or the like as the position of the portable terminal 20 (step S205).

  On the other hand, if it is determined in S204 that the reliability is not equal to or greater than the predetermined value, the derivation unit 16 indicates the link cost of the link between nodes continuously estimated by the position estimation unit 15 and is determined by the user information acquisition unit 14. A link cost corresponding to the acquired demogram information is derived (step S206). And it is determined by the determination part 17 whether a link cost is more than a predetermined threshold value (step S207).

  If it is determined in S207 that the link cost is not equal to or higher than the predetermined threshold, the node ID indicating the node related to the link cost is output (transmitted) to the mobile terminal 20 or the like (step S205).

  On the other hand, if it is determined in S207 that the link cost is equal to or higher than the predetermined threshold, information indicating that the estimation of the node related to the link cost is incorrect is output from the determination unit 17 to the position correction unit 31. (Step S208). Then, the position indicated by the node determined to be erroneous by the position correction unit 31 is corrected according to the movement history data 131 stored in the history storage unit 13, and the corrected position is the mobile terminal 20. The position of

  Specifically, the position correction unit 31 refers to the movement history data 131 and identifies the node (the latest node) in which the mobile terminal 20 is located before the node that is determined to be erroneous by the determination unit 17. To do. Then, the position correction unit 31 specifies the nearest node as the position of the mobile terminal 20 instead of the node determined to be erroneous in estimation. Finally, the position correction unit 31 outputs (transmits) the specified corrected position of the mobile terminal 20 to the mobile terminal 20 and an external device (not shown). The above is the position estimation process according to the modification.

  In the above modification, the reliability is calculated when determining the presence or absence of correction by the position correction unit 31, and the position correction is performed only when the reliability does not exceed the threshold. It is also possible to set a threshold value for the radio wave intensity transmitted from the radio wave device installed in the resulting node and perform correction only when the radio wave intensity does not exceed the threshold value.

  DESCRIPTION OF SYMBOLS 1 ... Position estimation system, 10 ... Position estimation apparatus, 11 ... Node link data storage part, 12 ... Correction | amendment part, 13 ... History storage part, 14 ... User information acquisition part, 15 ... Position estimation part, 16 ... Derivation part, 17 ... determining unit, 18 ... display unit, 20 ... portable terminal, 31 ... position correcting unit, 32 ... reliability calculating unit.

Claims (8)

A position estimation device that estimates the position of a user terminal based on position information acquired in the user terminal,
Node link data storage means for storing a plurality of link cost information for each of the user attribute information of the link between the nodes and the position information of the nodes indicating the abstracted positions;
Correction means for correcting each link cost stored in the node link data storage means based on the correlation between the link and user attribute information;
Acquisition means for continuously acquiring, from the user terminal, location information of the user terminal acquired in the user terminal, and user attribute information related to the user terminal;
Based on the position information of the user terminal continuously acquired by the acquisition means and the position information of the nodes stored in the node link data storage means, the node indicating the position of the user terminal is continuously estimated. Position estimation means for
A link stored in the node link data storage unit is a link cost of a link between nodes continuously estimated by the position estimation unit and corresponding to the user attribute information acquired by the acquisition unit. Derivation means derived from the cost,
A position estimation apparatus comprising: a determination unit that determines that the estimation of a node related to the link cost is an error when the link cost derived by the deriving unit is a predetermined value. A history storage means for storing movement histories of a plurality of user terminals in each link;
The history storage means stores user attribute information related to each user terminal as information included in the movement history,
The correction means derives the number of movements for each user attribute information of each link from the movement history stored in the history storage means, and determines the correlation of each link and user attribute information based on the number of movements. The position estimation apparatus according to claim 1.   The position estimation according to claim 1, wherein the correction unit further corrects each link cost stored in the node link data storage unit according to a degree of congestion determined based on the number of movements in the latest predetermined time. apparatus. A history storage means for storing movement histories of a plurality of user terminals in each link;
The history storage means stores date and time information that has passed the link as information included in the movement history,
4. The position estimating apparatus according to claim 3, wherein the correction unit derives the number of movements of each link within the latest predetermined time from the movement history stored in the history storage unit, and specifies the degree of congestion. The node link data storage means stores a link width as information about the link,
The position estimating device according to any one of claims 1 to 4, wherein the correction unit further corrects the link cost of each link stored in the node link data storage unit according to the road width. Further comprising display means for displaying the nodes continuously estimated by the position estimating means and links between the nodes in association with map information;
The position estimation device according to claim 1, wherein the display unit displays the node determined to be erroneous estimation by the determination unit in distinction from other nodes. History storage means for storing movement histories of a plurality of user terminals in each link;
When it is determined by the determination means that the estimation of the node related to the link cost is incorrect, the position indicated by the node for which the estimation is determined to be incorrect is indicated in the movement history stored in the history storage means. The position estimation apparatus according to any one of claims 1 to 6, further comprising position correction means that corrects the position in accordance with the position and outputs the corrected position as the position of the user terminal. A reliability calculation means for calculating the reliability of the node estimated by the position estimation means;
If the reliability is greater than or equal to a predetermined value, the link cost is not derived by the deriving means, but the node estimated by the position estimating means is output as the position of the user terminal,
The position estimation device according to claim 7, wherein, when the reliability is not equal to or higher than a predetermined value, the derivation unit derives a link cost.

Images