JP2017026542A - Position estimation device, position estimation method, and position estimation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine the position of a device even in the state of being unable to use a service such as GPS or a wireless LAN.SOLUTION: A position estimation device 20 converts a measurement value obtained by a sensor of a terminal 10 and time when the measurement value was obtained into a feature vector, and estimates which the active state of the terminal 10 is, a riding state, a stopping state, or a walking state at each time on the basis of the feature vector. Then, the position estimation device 20 generates a departure and arrival sequence on the basis of estimated departure time when the terminal shifted from the stopping state to the riding state and estimated arrival time when the terminal shifted from the riding state to the stopping state, and searches for a route by checking up the departure and arrival sequence, a time table, and a route map.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a position estimation device, a position estimation method, and a position estimation program.

  In recent years, with the spread of mobile devices such as smartphones, technologies for specifying location information and services based on location information have appeared. For example, by using a Global Positioning System (GPS) that is a system for measuring the current position using an artificial satellite, it is possible to accurately know where the device is on the earth by coordinates.

  On the other hand, IPS (Indoor Positioning System) is known as a technique for specifying a position in a situation where GPS cannot be used (for example, see Non-Patent Document 1). IPS is intended to detect and guide objects and people indoors using sensor information other than GPS (radio waves, voice, light, etc.).

  Also, PlaceEngine, which is a location specifying service that does not use GPS, can estimate the current device location from the pre-registered location information of the wireless LAN access point (AP) and the radio wave intensity of the AP ( For example, refer nonpatent literature 2.).

Yanying Gu, Anthony Lo, Senior Member, IEEE, and Ignas Niemegeers “A Survey of Indoor Positioning Systems for Wireless Personal Networks” IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 11, NO. 1, FIRST QUARTER 2009 PlaceEngine, http://www.placeengine.com

  However, the conventional technology has a problem that the position of the device cannot be specified in a situation where services such as GPS and wireless LAN cannot be used.

  For example, in specifying the position using the GPS, the GPS cannot be measured when there is an obstacle between the device and the satellite in order to determine the GPS. That is, the position cannot be specified by GPS in the building or underground.

  Further, in the method of specifying the position from the AP position information and the radio wave intensity, the AP position information needs to be registered in the service database in advance, and the AP that is not registered cannot specify the position. In the first place, the service cannot be used in a place where no AP exists.

  Furthermore, in recent smart devices, the GPS sensor and the wireless LAN can be manually turned on and off by the user, so the position is specified when the GPS sensor is off or when the wireless LAN is off. I can't.

  Furthermore, when the position of the device cannot be specified or estimated, services such as a car navigation system that performs navigation based on the position of the device, a service that displays a weather forecast, a service that displays an advertisement, and the like cannot be performed.

  The position estimation apparatus of the present invention includes a numerical value measured by one or more sensors including at least an acceleration sensor provided in a terminal, a sensor information processing unit that converts a time when the numerical value is measured into a feature vector, The feature vector indicates whether the state of the terminal at the time is a state related to movement including a boarding state that is moving with a moving body that moves between points or a stopped state that is stopped Based on the state estimated by the activity state estimation unit, based on the state estimated by the activity state estimation unit, the estimated departure time estimated that the terminal has departed together with the mobile body, and the terminal And a departure / arrival sequence generation unit that generates a combination of estimated arrival times estimated to have arrived at other points together with the mobile body, and the movement at each point Based on a combination of the estimated departure time and the estimated arrival time from a storage unit that stores time information including the scheduled departure time and estimated arrival time and route information indicating which point each point is connected to. And a candidate route search unit that searches for a route traveled by the terminal and identifies the position of the terminal from the search result.

  The position estimation method of the present invention is a position estimation method executed by a position estimation device, and is a numerical value measured by one or more sensors including at least an acceleration sensor provided in a terminal, and the numerical value is measured. A sensor information processing step for converting time into a feature vector, and a state in which the state of the terminal at each time includes a riding state that is moving together with a moving body that moves between points, and a stopped state that is stopped Among the states related to the activity state estimation step for estimating which state was based on the feature vector, and based on the state estimated by the activity state estimation step, the terminal and the mobile unit A combination of an estimated departure time estimated to have departed from the terminal and an estimated arrival time estimated to have arrived at another point with the mobile unit. From a storage unit that stores a departure / arrival sequence generation step, time information including a scheduled departure time and an estimated arrival time of the mobile body at each point, and route information indicating which point each point is connected to, A candidate route search step of searching for a route traveled by the terminal based on a combination of the estimated departure time and the estimated arrival time, and identifying the location of the terminal from the search result. .

  According to the present invention, the position of a device can be specified even in a situation where services such as GPS and wireless LAN cannot be used.

FIG. 1 is a diagram illustrating an example of a configuration of a position estimation system and a position estimation apparatus according to the first embodiment. FIG. 2 is a diagram illustrating an example of data measured by a sensor. FIG. 3 is a diagram illustrating an example of a departure / arrival sequence of the position estimation device according to the first embodiment. FIG. 4 is a diagram illustrating an example of a candidate route search method of the position estimation apparatus according to the first embodiment. FIG. 5 is a diagram for explaining a position estimation method of the position estimation apparatus according to the first embodiment. FIG. 6 is a flowchart illustrating an example of processing of the position estimation device according to the first embodiment. FIG. 7 is a diagram illustrating an example of a configuration of a position estimation system and a position estimation apparatus according to another embodiment. FIG. 8 is a diagram illustrating an example of a configuration of a position estimation system and a position estimation apparatus according to another embodiment. FIG. 9 is a diagram illustrating an example of a computer in which the position estimation apparatus is realized by executing a program.

  Hereinafter, embodiments of a position estimation device, a position estimation method, and a position estimation program according to the present application will be described in detail with reference to the drawings. In addition, the position estimation apparatus, the position estimation method, and the position estimation program according to the present application are not limited by this embodiment.

[First Embodiment]
In the following embodiments, the configuration of the position estimation system and the position estimation device according to the first embodiment, the flow of processing of the position estimation device will be described in order, and finally the effect of the first embodiment will be described.

[Configuration of First Embodiment]
First, the configuration of the position estimation system and the position estimation apparatus according to the first embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of a configuration of a position estimation system and a position estimation apparatus according to the first embodiment. As illustrated in FIG. 1, the position estimation system 1 includes a terminal 10, a position estimation device 20, and a storage device 30. The terminal 10 includes sensors such as an acceleration sensor 101, a gyro sensor 102, and a magnetic force sensor 103.

  Specific examples of the terminal 10 include a smartphone, a wearable computer, a sensor, and a camera having a communication function. The acceleration sensor measures the acceleration in a predetermined direction of the terminal. Further, the gyro sensor measures an angular velocity when the terminal rotates. The magnetic sensor measures the magnitude and direction of the magnetic field.

  The position estimation apparatus 20 includes a sensor information processing unit 201, an activity state estimation unit 202, a departure / arrival sequence generation unit 203, and a candidate route search unit 204. The storage device 30 includes a route information storage unit 302 and a time information storage unit 301. Hereinafter, each part will be described in detail.

  The sensor information processing unit 201 converts a numerical value measured by at least one sensor including at least the acceleration sensor 101 provided in the terminal 10 and a time when the numerical value is measured into a feature vector. For example, the sensor information processing unit 201 handles a numerical value obtained from the sensor and a time when the numerical value is acquired as a time series.

  Further, the interval at which the sensor information processing unit 201 obtains a numerical value from the sensor, that is, the sampling rate is arbitrary. For example, the sampling rate may be 10 Hz. In this case, the sensor information processing unit 201 acquires a numerical value measured by the sensor 10 times per second, that is, at intervals of 0.1 second.

  In addition, the sensor information processing unit 201 acquires, for example, acceleration in the X-axis direction, acceleration in the Y-axis direction, and acceleration in the Z-axis direction measured by the acceleration sensor 101. The sensor information processing unit 201 collects a numerical value acquired from a sensor, or a collection of feature amounts including a magnitude, a maximum value, a minimum value, a slope, and the like when compared with an average, variance, and threshold value of each numerical value, that is, a feature vector. create.

  Here, an example of a feature vector creation method will be described. As shown in FIG. 2, the information obtained from the sensor is composed of a series of time and value pairs. FIG. 2 is a diagram illustrating an example of data measured by a sensor. 2A shows the acceleration of the terminal 10 at each time acquired from the acceleration sensor 101. FIG. Further, M in FIG. 2 indicates the magnetic force of the terminal 10 at each time obtained from the magnetic force sensor 103. Further, G in FIG. 2 indicates the angular velocity of the terminal 10 at each time acquired from the gyro sensor 102.

  The sensor information processing unit 201 creates a feature vector by dividing continuous information into block units. A block contains N consecutive data. For example, when N is 20, when data is collected from the sensor at intervals of 0.1 seconds, continuous data for 2 seconds is one block.

For example, when a feature vector corresponding to time i is created using a numerical value and time acquired from the acceleration sensor 101, the following equation (1) is obtained. For example, when the data acquisition interval is 0.1 seconds in equation (1), a ₁ ⁽ⁱ⁾ is the acceleration at time i, and a ₂ ⁽ⁱ⁾ is 0.1 seconds after a ₁ ^(i). It is acceleration. In this case, the acceleration may be an acceleration in one direction, a vector indicating the acceleration in the X direction, the Y direction, and the Z direction, or the magnitude of the acceleration vector. .

  The active state estimation unit 202 is in any state in which the terminal 10 is in each state at each time among the states relating to the movement including the riding state that is moving together with the moving body that moves between the points and the stopped state that is stopped. It is estimated based on the feature vector. In the first embodiment, the moving object, the point, the time information, and the route information correspond to a train, a station, a timetable, and a route map, respectively.

  The activity state estimation unit 202 estimates the activity state of the user who owns the terminal 10. In the first embodiment, the active state is a stopped state where the vehicle is stopped and not moving, a walking state where the vehicle is walking or running, and a riding state where the vehicle is on a moving train. Is included. The activity state estimation unit 202 estimates the state of the terminal 10 using the feature vector created by the sensor information processing unit 201. The estimation result of the activity state estimation unit 202 is indicated by whether the terminal 10 is in the riding state, the walking state, or the stopped state at each time.

  Even when the terminal 10 is on the train, it is stopped when the train is stopped. For example, when leaving the station A and arriving at the station C via the station B, it is in a boarding state while moving between the station A and the station B and while moving between the station B and the station C. The activity state while the train stops at the station A, the station B, and the station C is a stopped state.

  The activity state estimation unit 202 estimates the activity state using a machine learning algorithm such as a support vector machine. The machine learning algorithm performs learning by being provided with feature vectors in an active state such as a boarding state, a walking state, and a stopped state as learning data. Then, the activity state estimation unit 202 inputs the feature vector to be estimated to the machine learning algorithm that has learned, and estimates the activity state.

  Based on the state estimated by the activity state estimation unit 202, the departure / arrival sequence generation unit 203 estimates the estimated departure time when the terminal 10 departs from one point together with the mobile object, A combination with an estimated arrival time estimated to have arrived at the point is generated. For example, the estimated departure time is the time when the state of the terminal 10 estimated by the activity state estimation unit 202 has shifted from the stop state to the boarding state. Further, for example, the estimated arrival time is a time when the state of the terminal 10 estimated by the activity state estimation unit 202 has shifted from the boarding state to the stop state.

  Here, the departure / arrival sequence is a sequence of pairs of start time and end time of sections of time when the boarding state is continuous. The estimated departure time is the start time of one section. The estimated arrival time is the end time of one section. The estimated departure time and estimated arrival time can also be referred to as estimated values of train departure time and stop time at a certain station.

  A specific example of the departure / arrival sequence will be described with reference to FIG. FIG. 3 is a diagram illustrating an example of a departure / arrival sequence of the position estimation device according to the first embodiment. A section with a low value indicates that the terminal 10 is in a stopped state. A section with a high value indicates that the terminal 10 is in the boarding state. The time when the activity state of the terminal 10 changes from the stop state to the boarding state is the estimated departure time. The time when the activity state of the terminal 10 changes from the boarding state to the stop state is the estimated arrival time. FIG. 3 shows that, for example, a train carrying a user holding the terminal 10 leaves the station at time d (10:57), and the train arrives at the next station at time a (10:58). ing.

  Candidate route search unit 204 searches for a route traveled by terminal 10 from time information storage unit 301 and route information storage unit 302 based on a combination of estimated departure time and estimated arrival time. Specifically, the candidate route search unit 204 searches the time information storage unit 301 for a combination of the scheduled departure time and the estimated arrival time that matches the combination of the estimated departure time and the estimated arrival time, and the departure schedule obtained as a result of the search. By acquiring a point corresponding to the combination of the time and the estimated arrival time from the route information storage unit 302, the route on which the terminal 10 has moved is searched. And based on a search result, the position of the user who owns the terminal 10 and the terminal 10 is specified.

  The time information storage unit 301 stores time information including scheduled departure time and arrival time at each point of the moving body. The route information storage unit 302 stores route information indicating which point each point is connected to.

  The candidate route search unit 204 searches, for example, a list of stations through which a train that is estimated to have been taken by a user who has the terminal 10 has passed. Further, the time information storage unit 301 and the route information storage unit 302 may store route data in which route maps and timetables of each route are converted into a database in advance.

  A method of searching for a route will be specifically described with reference to FIG. FIG. 4 is a diagram illustrating an example of a candidate route search method of the position estimation apparatus according to the first embodiment. First, it is assumed that the departure and arrival sequences generated by the departure and arrival sequence generation unit 203 are {Td1, Ta1}, {Td2, Ta2}, and {Td3, Ta3}. For example, in {Td1, Ta1}, Td1 and Ta1 indicate an estimated departure time and an estimated arrival time, respectively. The candidate route search unit 204 receives these departure / arrival sequences.

  In addition, S1, S2, etc. of FIG. 4 have shown the station of a train. The search result by the candidate route search unit 204 is expressed in a format such as {S1, S2, S3}. For example, {S1, S2, S3} indicates a route that leaves the S1 station and arrives at the S3 station via the S2 station.

  When a departure / arrival sequence {Td1, Ta1}, {Td2, Ta2}, {Td3, Ta3} is input, first, a station where the departure time and the stop time match {Td1, Ta1} is searched from the route map and timetable. To do. At this time, as shown in FIG. 4, {T1, S4}, {S1, S5}, {S2, S6}, {S3, S7} are routes that match {Td1, Ta1} with the departure time and stop time. Are searched.

  Further, a route that has the arrival station of the already searched route as the departure station and that matches {Td2, Ta2} with the departure time and the stop time is searched. As a result, three routes {S1, S4, S8}, {S1, S5, S9}, {S3, S7, S10} are searched. Note that {S2, S6} is excluded from candidates at this point because there is no route that satisfies the condition.

  Similarly, a route in which {Td3, Ta3} matches the departure time and the stop time is searched. Finally, {S1, S4, S8, S11}, {S3, S7, S10, S12} are output as candidate routes.

  The outline of the position estimation method will be described in the position estimation device 20 with reference to FIG. FIG. 5 is a diagram for explaining a position estimation method of the position estimation apparatus according to the first embodiment. The sensor data in FIG. 5 is data that the sensor information processing unit 201 acquires from each sensor of the terminal 10. And the activity state of FIG. 5 has shown the state of the terminal in each time which the activity state estimation part 202 estimates based on sensor data. As shown in FIG. 5, the activity state includes three states: a riding state, a walking state, and a stopped state.

  5 is a departure / arrival sequence generated by the arrival / departure sequence generation unit. The candidate routes in FIG. 5 are routes searched by the candidate route search unit 204 based on the arrival / departure sequence. As shown in FIG. 5, the candidate route search unit 204 acquires a plurality of candidate routes as search results. Then, for example, the candidate route search unit 204 may estimate the arrival point of the obtained candidate route as the position of the terminal 10 and the user who owns the terminal 10. In the example of FIG. 5, “Okachimachi”, “Shineicho”, or “Ringtsu Toto” is estimated as the position of the terminal 10 and the user.

[Process of First Embodiment]
The processing of the first embodiment will be described with reference to FIG. FIG. 6 is a flowchart illustrating an example of processing of the position estimation device according to the first embodiment. First, the sensor information processing unit 201 collects information from various sensors of the terminal 10 (step S101). Next, the sensor information processing unit 201 converts the collected information into a feature vector (step S102). Next, the activity state estimation unit 202 estimates the activity state from the feature vector (step S103).

  The departure / arrival sequence generation unit 203 generates a departure / arrival sequence from the activity state (step S104). Then, the candidate route search unit 204 lists candidate routes by matching the departure / arrival sequence with the timetable stored in the time information storage unit 301 and the route map stored in the route information storage unit 302 (step S105).

[Effect of the first embodiment]
The sensor information processing unit 201 of the position estimation device 20 converts a numerical value measured by at least one sensor including at least the acceleration sensor 101 provided in the terminal 10 and a time when the numerical value is measured into a feature vector.

  In addition, the activity state estimation unit 202 of the position estimation device 20 includes, at each time, a state relating to movement including a riding state that is moving together with a moving body that moves between points and a stopped state that is stopped. The terminal 10 is estimated based on the feature vector. For example, the activity state estimation unit 202 performs estimation using a machine learning algorithm that performs machine learning using the state of the terminal 10 and feature vectors in each state as learning data.

  Further, the departure / arrival sequence generation unit 203 of the position estimation device 20 is based on the state estimated by the activity state estimation unit 202, the estimated departure time estimated that the terminal 10 has departed with the moving object, and the terminal 10 generates a combination with an estimated arrival time that is estimated to have arrived at another point together with the moving object. For example, the estimated departure time may be a time when the state of the terminal 10 estimated by the activity state estimation unit 202 shifts from the stop state to the boarding state. Further, for example, the estimated arrival time is the time when the state of the terminal 10 estimated by the activity state estimation unit 202 has shifted from the boarding state to the stop state.

  In addition, the candidate route search unit 204 of the position estimation device 20 searches the route on which the terminal 10 has moved from the time information storage unit 301 and the route information storage unit 302 based on the combination of the estimated departure time and the estimated arrival time. Specifically, for example, the candidate route search unit 204 searches the time information storage unit 301 for a combination of the scheduled departure time and the estimated arrival time that matches the combination of the estimated departure time and the estimated arrival time, and the search result is obtained. By acquiring a point corresponding to the combination of the scheduled departure time and the estimated arrival time from the route information storage unit 302, the route traveled by the terminal 10 is searched, and the position of the terminal 10 is specified from the search result.

  The time information storage unit 301 stores time information including scheduled departure time and arrival time at each point of the moving body. The route information storage unit 302 stores route information indicating which point each point is connected to.

  Thus, in the first embodiment, the current position of the terminal is specified using information obtained from the sensor provided in the terminal. Thereby, the position of the device can be specified even in a situation where services such as GPS and wireless LAN cannot be used.

  For example, the sensor information processing unit 201 may acquire, for example, acceleration in the X-axis direction, acceleration in the Y-axis direction, and acceleration in the Z-axis direction measured by the acceleration sensor 101. Thus, the estimation accuracy can be improved by using the acceleration in each direction obtained from the acceleration sensor for estimation.

[Other Embodiments]
The sensors provided in the terminal 10 are not limited to those described in the above embodiment, and may include an atmospheric pressure sensor, a temperature sensor, and the like. Moreover, although the measured value of the acceleration sensor is considered to be effective as a parameter for estimating the activity state, a configuration without the acceleration sensor may be used.

  In addition, the measurement value obtained by the magnetic force sensor can be used as an activity state estimation parameter as in the above embodiment, and the candidate path is verified by matching with the information indicating the strength of the magnetic field at each point. be able to.

  In the above-described embodiment, an example in which the moving body is a train has been described. However, the present invention is not limited to a train, and may be, for example, an airplane, a bus, or a ship that moves according to a predetermined route and time. The present invention can be applied.

  Further, the candidate route as shown in FIG. 5 is obtained from the estimated departure time and estimated arrival time, route information and time information. As a method for further narrowing down the obtained candidate routes, for example, priority may be given to routes with a large number of users. For example, when route A and route B are searched as candidate routes, it is known in advance that the number of users per hour of route A is 10,000 and the number of users of route B per hour is 100. For example, it is more likely that the route A as the first candidate matches the actual movement route of the user. In addition, it is conceivable to narrow down candidate routes using a known method.

  In addition, the train may not move according to the timetable due to delay or the like. In this case, by referring to the delay information, the estimation accuracy can be improved by correcting the time information.

  Further, the configuration of the present invention is not limited to that shown in FIG. 1, and may be, for example, a configuration as shown in FIG. 7 or FIG. 7 and 8 are diagrams illustrating an example of the configuration of a position estimation system and a position estimation apparatus according to another embodiment.

  For example, as illustrated in FIG. 7, a measurement unit 10 a and a control unit 20 a may be provided in a terminal 15 a that also functions as a position estimation device. In this case, the measurement unit 10a includes various sensors. The control unit 20a has the same function as the position estimation device 20 in the first embodiment.

  In addition, as illustrated in FIG. 8, the terminal 15b that also functions as a position estimation device may include a measurement unit 10b, a control unit 20b, and a storage unit 30b. In this case, the measurement unit 10b includes various sensors. Further, the control unit 20b has the same function as the position estimation device 20 in the first embodiment. Furthermore, the storage unit 30b has the same function as the storage device 30 in the first embodiment. It is not always necessary to store all timetables and route maps in the storage unit 30b, but temporarily store the timetables and route maps necessary for the candidate route search unit 204b to search for routes. It may be.

[System configuration, etc.]
Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. Further, all or any part of each processing function performed in each device is realized by a CPU (Central Processing Unit) and a program analyzed and executed by the CPU, or hardware by wired logic. Can be realized as

  In addition, among the processes described in the present embodiment, all or part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed can be performed. All or a part can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, and information including various data and parameters shown in the above-described document and drawings can be arbitrarily changed unless otherwise specified.

[program]
FIG. 9 is a diagram illustrating an example of a computer in which the position estimation apparatus is realized by executing a program. The computer 1000 includes a memory 1010 and a CPU 1020, for example. The computer 1000 also includes a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These units are connected by a bus 1080.

  The memory 1010 includes a ROM (Read Only Memory) 1011 and a RAM (Random Access Memory) 1012. The ROM 1011 stores a boot program such as BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1090. The disk drive interface 1040 is connected to the disk drive 1100. For example, a removable storage medium such as a magnetic disk or an optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to a mouse 1110 and a keyboard 1120, for example. The video adapter 1060 is connected to the display 1130, for example.

  The hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. That is, the program that defines each process of the position estimation apparatus is implemented as a program module 1093 in which code executable by a computer is described. The program module 1093 is stored in the hard disk drive 1090, for example. For example, a program module 1093 for executing processing similar to the functional configuration in the position estimation device is stored in the hard disk drive 1090. The hard disk drive 1090 may be replaced by an SSD (Solid State Drive).

  The setting data used in the processing of the above-described embodiment is stored as program data 1094 in, for example, the memory 1010 or the hard disk drive 1090. Then, the CPU 1020 reads the program module 1093 and the program data 1094 stored in the memory 1010 and the hard disk drive 1090 to the RAM 1012 and executes them as necessary.

  The program module 1093 and the program data 1094 are not limited to being stored in the hard disk drive 1090, but may be stored in, for example, a removable storage medium and read out by the CPU 1020 via the disk drive 1100 or the like. Alternatively, the program module 1093 and the program data 1094 may be stored in another computer connected via a network (LAN (Local Area Network), WAN (Wide Area Network), etc.). The program module 1093 and the program data 1094 may be read by the CPU 1020 from another computer via the network interface 1070.

DESCRIPTION OF SYMBOLS 1 Position estimation system 10 Terminal 20 Position estimation apparatus 30 Storage apparatus 101 Acceleration sensor 102 Gyro sensor 103 Magnetic sensor 201 Sensor information processing part 202 Activity state estimation part 203 Departure / arrival sequence generation part 204 Candidate path search part 301 Time information storage part 302 Path information Memory

Claims (7)

A numerical value measured by one or more sensors including at least an acceleration sensor provided in the terminal, and a sensor information processing unit that converts a time when the numerical value is measured into a feature vector;
Whether the state of the terminal at each time is a state related to movement including a boarding state that is moving together with a moving body that moves between points or a stopped state that is stopped is the state described above An activity state estimator that estimates based on a vector;
Based on the state estimated by the activity state estimation unit, the estimated departure time when the terminal departs from one point with the mobile object, and the terminal arrives at another point together with the mobile object A departure and arrival sequence generation unit that generates a combination of estimated arrival time and
From the storage unit that stores time information including the scheduled departure time and estimated arrival time of the mobile body at each point, and route information indicating which point each point is connected to, the estimated departure time and the Based on a combination of estimated arrival times, search for a route traveled by the terminal, and a candidate route search unit that identifies the position of the terminal from the search result;
A position estimation apparatus comprising: The departure / arrival sequence generation unit uses the time when the state of the terminal estimated by the activity state estimation unit shifts from the stop state to the boarding state as the estimated departure time, and the terminal estimated by the activity state estimation unit The time when the state has shifted from the riding state to the stopped state is the estimated arrival time,
The candidate route search unit searches the time information for a combination of the estimated departure time and the estimated arrival time that matches the combination of the estimated departure time and the estimated arrival time, and the estimated departure time obtained as a result of the search 2. The position estimation apparatus according to claim 1, wherein a route traveled by the terminal is retrieved by acquiring a point corresponding to the combination of the estimated arrival time from the route information. The activity state estimation unit estimates the state of the terminal as a moving body that moves between stations, a train that moves between stations,
The position estimation device according to claim 1, wherein the candidate route search unit searches from a storage unit that stores a timetable and a route map as the time information and the route information.   The sensor information processing unit converts the acceleration in the X-axis direction, the acceleration in the Y-axis direction, and the acceleration in the Z-axis direction measured by the acceleration sensor into the feature vector as the numerical values. Item 4. The position estimation device according to any one of Items 1 to 3.   2. The activity state estimation unit estimates a state of a terminal at each time using a machine learning algorithm that performs machine learning using a state of the terminal and a feature vector in the state as learning data. 5. The position estimation device according to any one of 4 to 4. A position estimation method executed by a position estimation device,
A numerical value measured by one or more sensors including at least an acceleration sensor provided in the terminal, and a sensor information processing step for converting the time when the numerical value was measured into a feature vector;
Whether the state of the terminal at each time is a state related to movement including a boarding state that is moving together with a moving body that moves between points or a stopped state that is stopped is the state described above An activity state estimation process based on a vector;
Based on the state estimated by the activity state estimation step, the estimated departure time when the terminal departs from one point with the mobile object, and the terminal arrives at another point together with the mobile object A departure and arrival sequence generation step for generating a combination of the estimated arrival time and
From the storage unit that stores time information including the scheduled departure time and estimated arrival time of the mobile body at each point, and route information indicating which point each point is connected to, the estimated departure time and the Based on a combination of estimated arrival times, a route that the terminal has moved is searched, and a candidate route search step that specifies the position of the terminal from the search result;
The position estimation method characterized by including.   A position estimation program for causing a computer to function as the position estimation apparatus according to claim 1.

Images