JP5696455B2 - State determination device, state determination method, and state determination program - Google Patents

Description

  The present invention relates to a state determination device, a state determination method, and a state determination program.

  In recent years, RTLS (Real Time Location System) is becoming popular as a technique for detecting the position of a person or an object. The RTLS is a system that grasps the location of an object in real time by using RFID (Radio Frequency IDentification) or GPS (Global Positioning System) technology. There are various types of RTLS, such as “a room in which a moving body is (existing)” or “a location where the moving body has passed”.

  As a technology related to RTLS, for example, a store management system is known in which an infrared sensor that detects infrared rays emitted from the human body is installed at the entrance of a store, and the detected information is used to determine whether the customer has entered or exited the store. ing. Further, a technique for grasping the location of an object using RFID is also known. With this technology, an RFID tag is attached to a moving body, a tag reader is installed on the environment side, and the location of the moving body is determined by determining whether the RFID tag has passed in front of the tag reader. To do. Furthermore, an interrogator is installed on the environment side, and a moving body detection method is known in which a moving body such as a person is provided with a transponder and the moving body is detected based on whether or not the transponder has reacted to the interrogator. ing. In addition to installing multiple infrared communication receivers where people pass, such as doors, carry infrared communication transmitters with people, and when a person passes the door, the receiver receives multiple ID information. The technology to do is known. In this technique, a human behavior mode (walking or running) is identified from the reception times of the plurality of ID information.

Japanese Patent Laid-Open No. 2004-133583 JP 2009-7157 A JP 2006-260254 A JP-A-8-194011 JP 2007-52474 A International Publication No. 2007/058302

  In the technique described above, the receiver receives information from the transmitter by RFID or infrared communication. In other words, the technique described above captures the movement of the detection target by using the receiver as a sensor. However, the sensor may malfunction due to characteristics of the sensor itself or environmental factors.

  For example, when using RFID, the detection area is determined by the directivity of the antenna of the tag reader and the reach of radio waves. For this reason, even if there is a tag in the detection area, the tag is not detected (not detected). On the other hand, the tag is detected even though there is no tag in the detection area ( (False detection). As an example in which the tag is not detected, there is a case where the tag is difficult to read because the tag of the moving body is at a right angle with respect to the antenna direction of the tag reader installed on the environment side. On the other hand, as an example of erroneous detection, the signal from the tag is reflected by some object in the environment, so that another tag reader different from the reader to be detected originally detects this signal. May end up.

  The disclosed technology has been made in view of the above, and an object thereof is to provide a state determination device, a state determination method, and a state determination program that can accurately determine the state of a moving body.

  In order to solve the above-described problems and achieve the object, a state determination device disclosed in the present application includes, in one aspect, a storage unit, an estimation unit, and a determination unit. The storage unit has a state transition probability that is a probability of transition from each state to another state along with a transition of the state of the moving object in the environment, and each of a plurality of detection information indicating detection or non-detection of the moving object in each state. Are stored in association with output probabilities that are the probabilities of being output. When a plurality of pieces of detection information are sequentially output for the moving object, the estimation unit estimates a state where the moving object has reached using a state transition probability and an output probability stored in the storage unit. The determination unit determines the passage of the moving object based on the state estimated by the estimation unit.

  According to one aspect of the state determination device disclosed in the present application, there is an effect that the state of the moving body can be accurately determined.

FIG. 1 is a diagram for explaining an overview of state determination in the first embodiment. FIG. 2 is a diagram illustrating a functional configuration of the state determination unit according to the first embodiment. FIG. 3 is a state transition diagram for explaining the state of the tag. FIG. 4 is a flowchart illustrating the operation of the estimation unit according to the first embodiment. FIG. 5 is a flowchart for explaining the passage determination process according to the first embodiment. FIG. 6 is a flowchart for explaining the turn-back determination process according to the first embodiment. FIG. 7 is a flowchart for explaining the subroutine 1 in the turn-back determination process according to the first embodiment. FIG. 8 is a flowchart for explaining the subroutine 2 in the turn-back determination process according to the first embodiment. FIG. 9 is a flowchart for explaining the stagnation determination process according to the first embodiment. FIG. 10 is a flowchart for explaining the first half of the subroutine 3 in the stagnation determination process according to the first embodiment. FIG. 11 is a flowchart for explaining the latter half of the subroutine 3 in the stagnation determination process according to the first embodiment. FIG. 12 is a flowchart for explaining subroutine A of subroutine 3 in the stagnation determination process according to the first embodiment. FIG. 13 is a flowchart for explaining the first half of the subroutine 4 in the stagnation determination process according to the first embodiment. FIG. 14 is a flowchart for explaining the latter half of the subroutine 4 in the stagnation determination process according to the first embodiment. FIG. 15 is a flowchart for explaining subroutine B of subroutine 4 in the stagnation determination process according to the first embodiment. FIG. 16 is a diagram for explaining the stochastic automaton applied in the first embodiment. FIG. 17 is a diagram illustrating a specific example of the probability automaton applied in the first embodiment. FIG. 18 is a diagram illustrating a functional configuration of the state determination unit according to the second embodiment. FIG. 19 is a diagram illustrating a configuration example of a state determination system having a state determination unit according to the first and second embodiments. FIG. 20 is an example of a state transition diagram of the position management server in the state determination system. FIG. 21 is an example of an environment map in which a plurality of state determination units are installed in the state determination system. FIG. 22 is a diagram illustrating a computer that executes a state determination program.

  Embodiments of a state determination device, a state determination method, and a state determination program disclosed in the present application will be described below in detail with reference to the drawings. The state determination device and the state determination method disclosed in the present application are not limited by the following embodiments.

  Hereinafter, embodiments of the state determination device disclosed in the present application will be described with reference to the drawings. Below, the example which implement | achieves the state determination apparatus which this application discloses by RFID is demonstrated. FIG. 1 is a diagram for explaining an overview of state determination in the first embodiment. In the environment 1, an antenna fixing pole P1 is installed. The pole P1 includes an antenna A1 of the tag reading device at a predetermined position. The antenna A1 enables tag reading in the detection area B1. Similarly, an antenna fixing pole P2 is installed in the vicinity of the pole P1 of the environment 1 so as to be adjacent to the pole P1. The pole P2 includes the antenna A2 of the tag reading device at a predetermined position. The antenna A2 enables tag reading in the detection area B2. In addition, a region B3 where the detection areas B1 and B2 overlap is formed around the boundary C located between the antennas A1 and A2. An RFID tag 10 is attached to a moving body M such as a person located in the environment 1.

  First, the configuration of a state determination unit according to an embodiment disclosed in the present application will be described. FIG. 2 is a diagram illustrating a functional configuration of the state determination unit 100 according to the present embodiment. As shown in FIG. 2, the state determination unit 100 includes the antennas A <b> 1 and A <b> 2, the antenna switching device 101, the reading device 102, the state determination device 103, and the output unit 104. Each of these components is connected so that signals and data can be input and output in one direction or in both directions.

  The antennas A1 and A2 are provided in the environment 1 shown in FIG. 1 and have the same frequency band. The antenna switching device 101 switches the antennas A1 and A2 at predetermined time intervals in order to enable tag detection in different areas by the antennas A1 and A2 having the same frequency band. The time interval is 50 ms, for example.

  The reading device 102 inputs the tag ID detected by the antennas A1 and A2 together with the detected antenna ID via the antenna switching device 101. Then, the reading device 102 outputs these associated IDs as detection results to the subsequent state determination device 103. This output is performed at a time interval (for example, 100 ms) that is twice the switching period of the antennas A1 and A2.

  The state determination device 103 determines the state of the moving body M present in the environment 1 based on the tag detection result input from the reading device 102, and outputs the determination result to the output unit 104. As illustrated in FIG. 2, the state determination device 103 includes an acquisition unit 113, a storage unit 123, an estimation unit 133, and a determination unit 143.

  The acquisition unit 113 acquires the tag detection result output from the reading device 102. The tag detection result is associated with the antenna ID and the tag ID detected by the antenna, for example, “A1: tag 10, A2: tag 20”. The acquisition unit 113 outputs the tag detection result to the estimation unit 133.

  The storage unit 123 stores a state transition probability 123a and an output probability 123b in association with each other. The state transition probability 123a is a probability of transition from each state to another state along with the transition of the state of the tag 10 in the environment 1. The output probability 123b is a probability that each of a plurality of pieces of detection information (output symbols k) indicating detection or non-detection of the tag 10 in each state is output.

Based on the tag detection result input from the acquisition unit 113, the estimation unit 133 estimates a state where the tag 10 is currently reached. That is, when a plurality of pieces of detection information are sequentially output for the tag 10, the estimation unit 133 uses the probability automaton defined by the state transition probability 123 a and the output probability 123 b stored in the storage unit 123. Estimate the reached state. The estimation unit 133 sequentially estimates the current state of the tag 10 each time detection information is detected. The estimation unit 133 outputs the state estimation results (for example, S ₁ and S ₅ described later) to the determination unit 143 at intervals of, for example, 100 ms. Note that the estimation unit 133 outputs the estimation result as the predetermined time elapses even when the tag 10 does not transition to another state.

  The determination unit 143 determines the state of the moving body M to which the corresponding tag is attached based on the detection state estimation result input from the estimation unit 133. The determination unit 143 includes a passage determination unit 143a, a turnback determination unit 143b, and a stagnation determination unit 143c. The determination unit 143 includes three determination units for each state type. The passage determination unit 143a determines whether or not the moving object M has passed in the environment 1. Further, the turnback determination unit 143b determines whether or not the moving object M has turned back in the environment 1. Further, the stagnation determination unit 143c determines whether or not the moving object M has stagnation in the environment 1. The stagnation determination unit 143c includes a counter 144c as an index for determining whether or not the moving object M is stagnation in an arbitrary state. The stagnation determination unit 143c determines that the moving object M is stagnation when the value of the counter 144c is equal to or greater than the threshold value T.

FIG. 3 is a state transition diagram for explaining the state of the tag. As shown in FIG. 3, there are nine states of S _{0 to} S ₈ as tag states in environment 1. Hereinafter, the nine states will be described. S ₀ is a state in which no tag is detected. S ₁ is a state in which the tag is detected in the detection area B1 (see FIG. 1). S ₂ is a state where the tag is detected in the detection area B3 (see FIG. 1). The detection area B3 is a region where the detection area B1 and the detection area B2 overlap. S ₃ is a state in which the tag is detected in the detection area B2 (see FIG. 1). S ₄ is a state where the tag has become undetected leaves the detection area B2. S ₅ is a _{state S 3} Similarly, the tag is detected in the detection area B2 (see FIG. 1). S ₆ is a _{state S 2} Similarly, the tag is detected in the detection area B3 (see FIG. 1). S ₇ is a _{state S 1} Similarly, the tag is detected in the detection area B1 (see FIG. 1). S ₈ is a state where the tag has become undetected leaves the detection area B1.

The arrow in FIG. 3 shows a state transition. For example, an arrow D1 in FIG. 3 indicates that the tag has transitioned from state S ₁ to state S _2. The arrow D2 indicates that the tag has transitioned from state _{S 2} to state _{S 1.} The arrow D3 is that the tag had remained state S _1, i.e., indicating that the tag has been stopped in the area of the antenna A1 shown in FIG. Further, arrows D4, D5 indicates that a state _S 4, _{S 8} is automatically reset to the initial state _{S 0} of the tag undetected respectively.

In FIG. 3, when a malfunction does not occur in the reading device 102, the tag state transitions with tag detection by the reading device 102 as an event. However, if the moving object M remains in the detection area B1, the event “the antenna A1 detects the tag with the same ID” repeatedly occurs every time the tag detection result is acquired from the reading device 102. In order to cope with such a situation, the state S _n (n is an integer of 0 to 8) has a self-loop (for example, the arrow D3 described above) that returns to its own state in response to an event. Further, assuming that the moving body M reciprocates between the antenna A1 and the antenna A2 or turns back halfway, each state S _n (n = 0, 1, 2, 3, 5, 6, 7) Has a path capable of bidirectional transition between two different states.

  The output unit 104 outputs the state (passing, turning back, stagnation) of the moving body M input from the determination unit 143 as a determination result.

  Note that the state determination apparatus 103 illustrated in FIG. 2 is physically realized by, for example, a personal computer. The acquisition unit 113, the estimation unit 133, and the determination unit 143 are physically realized by an integrated circuit such as a CPU (Central Processing Unit) or an ASIC (Application Specific Integrated Circuit). The storage unit 123 is physically realized by a semiconductor memory element such as a random access memory (RAM), a read only memory (ROM), and a flash memory. The output unit 104 is physically realized by a display device such as a liquid crystal display.

  Next, the operation of the state determination device 103 according to the first embodiment will be described.

  First, a series of processes executed by the estimation unit 133 will be described with reference to FIG. FIG. 4 is a flowchart illustrating the operation of the estimation unit 133 according to the first embodiment.

  First, the estimation unit 133 acquires a tag detection result output from the reading device 102 (step S1). The tag detection result includes at least the ID of the detected tag and the ID of the antenna that has detected the tag. Next, the estimation unit 133 estimates the tag state for the detected tag ID (step S2). This estimation is performed by applying the state transition probability and the output probability to the state transition of the tag in the environment. And the estimation part 133 outputs this estimation result to the determination part 143 (step S3).

  The series of steps S1 to S3 is repeatedly executed at regular intervals, such as a 100 ms period.

  Next, the operation of the determination unit 143 will be described with reference to FIGS.

First, the passage determination process will be described. FIG. 5 is a flowchart for explaining the passage determination process according to the first embodiment. In step S10, the passage determination unit 143a of the determination unit 143 waits until a new estimation result is input from the estimation unit 133. When the tag of the subject to the ID of the passage determination is detected (step S11; Yes), passage determination unit 143a, the state of the tag determines whether the transition to state _{S 4} (step S12). As a result of the determination, when a transition is made to state _{S 4} (step S12; Yes), passage determination unit 143a, in the environment 1, (step determines that the moving object M is "pass" from the antenna A1 to the direction of the antenna A2 S13). And it returns to step S10 and performs the process after it.

On the other hand, the result of determination in step S12, if not a transition to state _{S 4} (step S12; No), passage determination unit 143a, the tag determines whether a transition to state _{S 8} (step S14). As a result of the determination, when a transition is made to state _{S 8;} (step S14 Yes), passage determination unit 143a determines that the environment 1, the moving object M is "pass" from the antenna A2 in the direction of the antenna A1 (Step S15). And it returns to step S10 and performs the process after it.

Further, the result of determination in step S14, if not a transition to the state _{S 8} (Step S14; No), passage determination unit 143a, in the environment 1, the moving object M does not pass in either direction (not Pass state) (step S16). And it returns to step S10 and performs the process after it. In step S11, even when the tag with the ID that is the object of passage determination is not detected (step S11; No), the process returns to step S10, and the subsequent processing is executed.

Next, the turn-back determination process will be described with reference to FIGS. FIG. 6 is a flowchart for explaining the turn-back determination process according to the first embodiment. In step S <b> 20, the turnback determination unit 143 b of the determination unit 143 waits until a new estimation result is input from the estimation unit 133. Retrace determination unit 143b, the state of the tag determines whether the transition from the state _{S 0} to state _{S 1} (step S21). As a result of the determination, when a transition is made to state _{S 1;} (step S21 Yes), retrace determination unit 143b executes the subroutine 1 described later (step S22).

State; (No step S21), and retrace determination unit 143b, the tag state _S 0 to _{(S 4} is reset Meanwhile, the result of determination in step S21, if the tag is not a transition from the state _{S 0} to state _{S 1} ) it determines whether a transition to state _{S 5} from (step S23). As a result of the determination, when a transition is made to state _{S 5;} (step S23 Yes), retrace determination unit 143b executes the subroutine 2 which will be described later (step S24). As a result of the determination in step S23, if not a transition to state _{S 5;} (step S23 No), the process returns to step S20, and the subsequent processes are executed.

  Subsequently, the above-described subroutine 1 (step S22) will be described in detail with reference to FIG. FIG. 7 is a flowchart for explaining the subroutine 1 in the turn-back determination process according to the first embodiment.

In step S <b> 30, the turnback determination unit 143 b of the determination unit 143 waits until a new estimation result is input from the estimation unit 133. Retrace determination unit 143b, the state of the tag determines whether the transition from state _{S 1} to state _{S 2} (step S31). As a result of the determination, when a transition is made to state _{S 2;} (step S31 Yes), retrace determination unit 143b waits from the estimation unit 133 until a new estimation result is input (step S32). Retrace determination unit 143b, the state of the tag determines whether the transition from state _{S 2} to state _{S 1} (step S33). As a result of the determination, when a transition is made to state _{S 1;} (step S33 Yes), retrace determination unit 143b waits from the estimation unit 133 until a new estimation result is input (step S34). Retrace determination unit 143b, the state of the tag determines whether the transition from state _{S 1} to state _{S 0} (step S35). As a result of the determination, when a transition is made to state _{S 0;} (step S35 Yes), retrace determination unit 143b determines the "retrace" the state of the moving object M (step S36). Then, subroutine 1 ends. After the end of subroutine 1, the process returns to step S20 (return in FIG. 7).

As a result of the judgment in the step S31, if not a transition from the state _{S 1} to state _{S 2;} (step S31 No), retrace determination unit 143b, skip process in step S32~S34 described above, The process of step S35 is executed. Further, the result of determination in step S35, if not a transition from the state _{S 1} to state _{S 0;} (step S35 No), retrace determination unit 143b executes again the step S30 and subsequent steps.

Further, the result of determination in step S33, if not a transition to the state _{S 1;} or (step S33 No), retrace determination unit 143b continues, the state of the tag transitions from state _{S 2} to state _{S 3} It is determined whether or not (step S37). As a result of the determination, when a transition is made to state _{S 3;} (step S37 Yes), retrace determination unit 143b further waits until a new estimation result from the estimating unit 133 is input (step S38). Retrace determination unit 143b, the state of the tag determines whether the transition from state _{S 3} to the state _{S 2} (step S39). As a result of the determination, if not a transition to state S _2; (step S39 No), retrace determination unit 143b further state of the tag determines whether the transition from state S ₃ to the state S ₄ (Step S40). As a result of the determination, when a transition is made to state _{S 4;} (step S40 Yes), determines that the retrace determining unit 143b, the mobile M is to "pass" from detection area B1 side in the detection area B2 side ( Step S41). Then, subroutine 1 ends.

As a result of the judgment in the step S37, if not a transition from the state _{S 2} to state _{S 3;} (step S37 No), retrace determination unit 143b executes again the step S32 and subsequent steps. Further, as a result of the judgment in the step S39, when a transition from the state _{S 3} to the state _{S 2;} (step S39 Yes), retrace determination unit 143b executes again the step S32 and subsequent steps. Furthermore, as a result of the judgment in step S40, if not a transition from the state _{S 3} to the state _{S 4;} (step S40 No), retrace determination unit 143b executes again the step S38 and subsequent steps.

Next, the above-described subroutine 2 (step S24 shown in FIG. 6) will be described. FIG. 8 is a flowchart for explaining the subroutine 2 in the turn-back determination process according to the first embodiment. In FIG. 7, the turn-back determination process in the states S _{1 to} S ₄ (see FIG. 3) is assumed, but in FIG. 8, the turn-back determination process in the states S _{5 to} S ₈ (see FIG. 3) is assumed. 7 and 8 are the same except that the direction of the state transition is the opposite direction, and detailed description thereof is omitted, but steps S50 to S56 in FIG. 8 are the same as steps S30 to S36 in FIG. Correspond to each. Also, steps S57 to S61 in FIG. 8 correspond to steps S37 to S41 in FIG. 7, respectively. Furthermore, the state _S 5 to S ₈ respectively correspond to the states _S 1 to S _4. In the subroutine 1, the turn-back determining unit 143b determines the passage from the detection area B1 side to the detection area B2 side (step S41 in FIG. 7), but in the subroutine 2, the turn-back determining unit 143b is in the opposite direction. The passage is determined (step S61 in FIG. 8). The state _{S 0} at step S55 in FIG. 8 is a result of the state _{S 4} is reset.

Next, the stagnation determination process will be described with reference to FIGS. FIG. 9 is a flowchart for explaining the stagnation determination process according to the first embodiment. In step S70, the stagnation determination unit 143c of the determination unit 143 initializes the stagnation determination counter 144c to “C = 0”. The stagnation determination unit 143c waits until a new estimation result is input from the estimation unit 133 (step S71). Stagnation determining unit 143c, the state of the tag determines whether the transition from the state _{S 0} to state _{S 1} (step S72). As a result of the determination, when a transition is made to state _{S 1;} (step S72 Yes), stagnation determination unit 143c executes a subroutine 3 to be described later (step S73).

State; (No Step S72), stagnation determination unit 143c, the tag state _S 0 to _{(S 4} is reset Meanwhile, the result of determination in step S72, the case where the tag is not a transition from the state _{S 0} to state _{S 1} ) it determines whether a transition to state _{S 5} from (step S74). As a result of the determination, when a transition is made to state _{S 5;} (step S74 Yes), stagnation determination unit 143c executes a subroutine 4 to be described later (step S75). As a result of the judgment in the step S74, the case where no transition to state _{S 5;} (step S74 No), the process returns to step S71, and the subsequent processes are executed.

  Subsequently, the above-described subroutine 3 (step S73) will be described in detail with reference to FIGS. FIG. 10 is a flowchart for explaining the first half of the subroutine 3 in the stagnation determination process according to the first embodiment.

In step S80, the stagnation determination unit 143c of the determination unit 143 waits until a new estimation result is input from the estimation unit 133. Stagnation determining unit 143c, the state of the tag determines whether the transition from state _{S 1} to state _{S 2} (step S81). As a result of the determination, when a transition is made to state _{S 2;} (step S81 Yes), stagnation determination unit 143c waits the estimating unit 133 until a new estimation result is input (step S82). Stagnation determining unit 143c, the state of the tag determines whether the transition from state _{S 2} to state _{S 1} (step S83). As a result of the determination, when a transition is made to state _{S 1;} (step S83 Yes), stagnation determination unit 143c further executes a subroutine A which will be described later (step S84).

Further, the result of determination in step S81, if not a transition to state _{S 2;} or (step S81 No), stagnation determination unit 143c continues, the state of the tag transitions from state _{S 1} to state _{S 0} It is determined whether or not (step S85). As a result of the determination, when a transition is made to state _{S 0;} (step S85 Yes), the subroutine 3 is terminated. Along with the end of this subroutine 3, the series of stagnation determination processing ends (END in FIG. 9). On the other hand, if no transition to state _{S 0;} (step S85 No), stagnation determination unit 143c, the mobile M is determined to be stagnant state _{S 1,} it executes a subroutine A which will be described later ( Step S86).

FIG. 11 is a flowchart for explaining the latter half of the subroutine 3 in the stagnation determination process according to the first embodiment. As a result of the judgment in the step S83, the case where no transition to state _{S 1;} (step S83 No), stagnation determination unit 143c determines that the moving object M is stagnant state _{S 2,} FIG. 11 The process after step S90 is executed. In step S90, stagnation determination unit 143c, the state of the tag determines whether the transition from state _{S 2} to state _{S 3.} As a result of the determination, when a transition is made to state _{S 3;} (step S90 Yes), stagnation determination unit 143c waits the estimating unit 133 until a new estimation result is input (step S91). Stagnation determining unit 143c, the state of the tag determines whether the transition from state _{S 3} to the state _{S 2} (step S92). As a result of the determination, if not a transition to state S _2; (step S92 No), the more stagnation determination unit 143c, determines that the moving object M is stagnant state S _3, the state of the tag determines whether the transition from state _{S 3} to the state _{S 4} (step S93). As a result of the determination, when a transition is made to state _{S 4;} (step S93 Yes), stagnation determination unit 143c, the mobile M is determined that was "pass" from detection area B1 side in the detection area B2 side ( Step S94). Then, subroutine 3 ends. Along with the end of this subroutine 3, the series of stagnation determination processing ends (END in FIG. 9).

Further, as a result of the judgment in the step S90, if not a transition to the state _{S 3;} or (step S90 No), stagnation determination unit 143c continues, the state of the tag transitions from state _{S 2} to state _{S 1} It is determined whether or not (step S95). As a result of the determination, when a transition is made to state _{S 1;} (step S95 Yes), stagnation determination unit 143c executes a subroutine A which will be described later (step S96). After execution, the stagnation determination unit 143c executes again the processing after step S80 in FIG. On the other hand, the result of the determination in step S95, if not a transition to the state _{S 1;} (step S95 No), the moving body M is determined that the stagnant state _{S 2,} stagnation determination unit 143c is A subroutine A described later is executed (step S97). Then, the stagnation determination unit 143c executes again the processing after step S82 in FIG.

As a result of the judgment in the step S92, when a transition from the state _{S 3} to the state _{S 2;} (step S92 Yes), stagnation determination unit 143c executes a subroutine A which will be described later (step S98). After the execution, the stagnation determination unit 143c executes the processes after step S82 in FIG. 10 again as in step S97 described above.

Furthermore, as a result of the judgment in the step S93, if not a transition to state _{S 4;} or (step S93 No), stagnation determination unit 143c continues, the state of the tag transitions from state _{S 3} to the state _{S 2} It is determined whether or not (step S99). As a result of the determination, when a transition is made to state _{S 2;} (step S99 Yes), stagnation determination unit 143c executes a subroutine A which will be described later (step S100). And the stagnation determination part 143c performs the process after step S82 of FIG. 10 again like step S97 and S98 mentioned above. On the other hand, the result of the determination in step S99, if not a transition to state _{S 2;} (step S99 No), the moving body M is determined that the stagnant state _{S 3,} stagnation determination unit 143c is A subroutine A described later is executed (step S101). And the stagnation determination part 143c performs the process after the above-mentioned step S91 again.

  FIG. 12 is a flowchart for explaining subroutine A of subroutine 3 in the stagnation determination process according to the first embodiment. In step S110 of FIG. 12, the stagnation determination unit 143c of the determination unit 143 increments the value of the stagnation determination counter 144c by one. The stagnation determination unit 143c compares the size of the counter 144c with the threshold T (step S111). If C ≧ threshold T (step S111; Yes), the stagnation determination unit 143c determines the state of the moving object M. It determines with "stagnation" (step S112). Then, subroutine A ends. On the other hand, if C <threshold value T as a result of the comparison in step S111 (step S111; No), the subroutine A is terminated as it is (step S113).

Next, the above subroutine 4 (step S75 shown in FIG. 9) will be described. FIG. 13 is a flowchart for explaining the first half of the subroutine 4 in the stagnation determination process according to the first embodiment. In FIG. 10, the stagnation determination process in the states S _{1 to} S ₄ (see FIG. 3) is assumed, but in FIG. 13, the stagnation determination process in the states S _{5 to} S ₈ (see FIG. 3) is assumed. 10 and 13 are the same except that the direction of state transition is the opposite direction, and detailed description thereof is omitted. However, steps S120 to S126 in FIG. 13 are the same as steps S80 to S86 in FIG. Correspond to each. The state _S 5 to S ₈ respectively correspond to the states _S 1 to S _4. Further, subroutine B is the same processing as subroutine A.

  Similarly, FIG. 14 corresponds to FIG. 11, and FIG. 15 corresponds to FIG. FIG. 14 is a flowchart for explaining the latter half of the subroutine 4 in the stagnation determination process according to the first embodiment. Steps S130 to S141 in FIG. 14 correspond to steps S90 to S101 in FIG. 11, respectively. In the subroutine 3, the stagnation determining unit 143c determines the passage from the detection area B1 side to the detection area B2 side (step S94 in FIG. 11), but in the subroutine 4, the stagnation determining unit 143c is in the opposite direction. The passage is determined (step S134 in FIG. 14). FIG. 15 is a flowchart for explaining subroutine B of subroutine 4 in the stagnation determination process according to the first embodiment. Steps S150 to S153 in FIG. 15 correspond to steps S110 to S113 in FIG. 12, respectively.

As described above, the state determination device 103 performs a series of turn-back determination processing, so that the moving body M turns back from any of the states S ₁ , S ₂ , and S ₃ shown in FIG. Also, it is possible to make a turnback determination. Similarly, the state determination apparatus 103 can deal with a case where the moving body M is in a state of any of the states S ₁ , S ₂ , and S ₃ by executing a series of stagnation determination processes. .

Although the operation has been described above, the state determination device 103 according to the present exemplary embodiment uses the probability automaton illustrated in FIG. 16 in order to cope with a malfunction (not detected or erroneous detection of the tag 10) by the reading device 102 in the environment 1. . The stochastic automaton is, for example, a hidden Markov model (HMM). FIG. 16 is a diagram for explaining the stochastic automaton applied in the first embodiment. In FIG. 16, “a _ij ” represents the transition probability from the state S _i to the state S _j . “B _i (k)” represents the probability of outputting the symbol k in the state S _i . “K” represents an output symbol indicating that a tag has been detected in the corresponding detection area. For example, the output symbol when the tag 10 is detected in the detection area B1 is “k = A1”, and the output symbol when the tag 10 is detected in the detection area B2 is “k = A2”. The output symbol when the tag 10 is detected in the detection area B3 is “k = A12”, and the output symbol when the tag 10 is not detected in any detection area (undetected) is “k = Non”. ". Thus, for _example, b 1 (A1), in a state _{S 1,} indicates the probability of tag in the detection area B1 is output is _detected, b 2 (A12) is, in the state _{S 2,} the tag within the detection area B3 Indicates the probability of output when is detected. Further, b ₃ (Non) indicates the probability of output when no tag is detected in any detection area in the state S ₃ .

By using the probability automaton, for example, in the state _{S 1,} the output probability of the symbol tag is detected in each of the detection area _{is, "b 1 (A1) =} 0.8", "b 1 (A2 ) = 0.05 ”,“ b ₁ (A12) = 0.05 ”,“ b ₁ (Non) = 0.1 ”. As a result, the probability b ₁ (A12) that the detection (false detection) is detected when there is a tag in the detection area B3 despite the presence of the tag in the detection area B1 of the antenna A1 is defined as 0.05. . Similarly, the probability b ₁ (Non) that is not detected (undetected) even though there is a tag in the detection area B1 of the antenna A1 is defined as 0.1.

By using the probabilistic automaton, it is possible to estimate the state as described below. The time-series detection data from the first detection to the latest detection of a tag with a certain ID by the reading device 102 is represented as “observation symbol y”. For example, when the reading device 102 detects the tag 10 in the order of “not detected”, “detected by the antenna A1”, “detected by the antenna A1”, and “detected by the antenna A2”, the observation symbol is y = Non A1 A1. The time series data is A2. In the probability automaton shown in FIG. 16, the probability that the state S _i (i is an integer of 0 to 8) when such an observation symbol y is given is expressed by p (S _i | y). The estimation unit 133 of the state determination device 103 calculates p (S _i | y) for all the states S _i when a certain observation symbol y is given for the tag 10, and obtains the state S _i with the highest probability. Thus, the current state of the tag 10 is estimated. For example, when p (S ₄ | y) becomes the maximum among p (S _i | y), the estimation unit 133 estimates that the tag 10 has reached the state S ₄ . And the determination part 143 determines with the moving body M to which the tag 10 was attached having passed in the direction which goes to detection area B2 from detection area B1 by performing the passage determination process mentioned above.

FIG. 17 is a diagram illustrating a specific example of the probability automaton applied in the first embodiment for each of the states S _{0 to} S ₄ . In FIG. 17, the numerical values shown in the states S _{1 to} S ₃ are the output probabilities of the symbol k in each state. For example, in state _{S 1,} A1, A2, A12, the probability of outputting Non the (non-detection) are each 0.5,0.1,0.3,0.1. The numerical values attached to the arrows connecting between the states S ₀ to S ₄ represents the state transition probability. For example, the probability of transition from state S ₁ to state S ₁ (self-loop) is 0.4, the probability of transition from state S ₁ to state S ₀ is 0.3, and the probability of transition from state S ₁ to state S ₂ is 0.3.

Hereinafter, assuming that the estimation unit 133 observes the feature vector (A1, A12, A2, Non) related to the tag 10, a method for estimating the state (transition destination) reached by the tag 10 will be described. The state with the highest probability as the arrival state of the tag 10 is calculated as follows. A state in which the feature vector is input and reaches S ₀ is defined as ω ₀ . Similarly, the state that has reached S ₁ is ω ₁ , the state that has reached S ₂ is ω ₂ , the state that has reached S ₃ is ω ₃ , and the state that has reached S ₄ is ω ₄ . When the initial state is S ₀ , the probability when the feature vector is observed and transitions from S ₁ → S ₁ → S ₁ → S ₀ is calculated as follows.

_{p (A1│S 1) × p (} S 1 → S 1) × p (A12│S 1) × p (S 1 → S 1) × p (A2│S 1) × p (S 1 → S 0) = 0.5 x 0.4 x 0.3 x 0.4 x 0.1 x 0.3 = 0.00072

Here, p (A1 | S ₁ ) is a probability (output probability) that is output when A1 is observed in the state S ₁ , and p (S ₁ → S ₁ ) is changed from the state S ₁ to the state S ₁ . This is the probability of transition (state transition probability).

Similarly for the state transitions other than S ₁ → S ₁ → S ₁ → S ₀ , the above probabilities are obtained for all the states that may be reached when the feature vector (A1, A12, A2, Non) is observed. Is calculated as follows.

The probability of transition from S ₁ → S ₂ → S ₁ → S ₀ is 0.000675
The probability of transition from S ₁ → S ₁ → S ₁ → S ₁ is 0.000096
The probability of transition from S ₁ → S ₁ → S ₂ → S ₁ is 0.000135
The probability of transition from S ₁ → S ₂ → S ₁ → S ₁ is 0.00009
The probability of the transition from S ₁ → S ₂ → S ₂ → S ₁ is 0.000225
The probability of a transition from S ₁ → S ₁ → S ₁ → S ₂ is 0.00072
The probability of transition from S ₁ → S ₁ → S ₂ → S ₂ is 0.00018
The probability of transition from S ₁ → S ₂ → S ₁ → S ₂ is 0.0000675
The probability of transition from S ₁ → S ₂ → S ₂ → S ₂ is 0.0003
The probability of transition from S ₁ → S ₂ → S ₃ → S ₂ is 0.0003375
The probability of the transition from S ₁ → S ₁ → S ₂ → S ₃ is 0.000135
The probability of transition from S ₁ → S ₂ → S ₂ → S ₃ is 0.000225
The probability of a transition from S ₁ → S ₂ → S ₃ → S ₃ is 0.00045
The probability of transition from S ₁ → S ₂ → S ₃ → S ₄ is 0.003375

Since these are independent events that cannot occur at the same time, the probabilities of ω _{0 to} ω ₄ are calculated as follows.

p (A1, A12, A2, Non | ω ₀ ) = 0.00072 + 0.000675 = 0.001395
p (A1, A12, A2, Non | ω ₁ ) = 0.00096 + 0.000135 + 0.00009 + 0.000225 = 0.000546
p (A1, A12, A2, Non | ω ₂ ) = 0.000072 + 0.00018 + 0.0000675 + 0.0003 + 0.0003375 = 0.000957
p (A1, A12, A2, Non | ω ₃ ) = 0.000135 + 0.000225 + 0.00045 = 0.00081
p (A1, A12, A2, Non | ω ₄ ) = 0.003375

Since the prior probabilities of ω _{0 to} ω ₄ are equal, p (ω ₀ ) = p (ω ₁ ) = p (ω ₂ ) = p (ω ₃ ) = p (ω ₄ ) = 0.2.

  Therefore, the arrival state with the highest probability when the feature vector (A1, A12, A2, Non) is observed is obtained as follows from Bayes' theorem.

_{p (ω 0 │A1, A12,} A2, Non) = p (A1, A12, A2, Non│ω 0) × p (ω 0) = 0.000279
_{p (ω 1 │A1, A12,} A2, Non) = p (A1, A12, A2, Non│ω 1) × p (ω 1) = 0.000109
_{p (ω 2 │A1, A12,} A2, Non) = p (A1, A12, A2, Non│ω 2) × p (ω 2) = 0.000191
_{p (ω 3 │A1, A12,} A2, Non) = p (A1, A12, A2, Non│ω 3) × p (ω 3) = 0.000162
_{p (ω 4 │A1, A12,} A2, Non) = p (A1, A12, A2, Non│ω 4) × p (ω 4) = 0.000675

From the above, the estimation unit 133, the probability automaton of FIG. 17, when the feature vector (A1, A12, A2, Non ) was observed, with a high probability of tag 10 is most arrival status is estimated state _{S 4} .

In the above description, the estimation method using the stochastic automaton has been described for the states S _{0 to} S ₄ (upper side in FIG. 3). However, the estimation unit 133 has the states S ₀ and S _{5 to} S ₈ (FIG. 3). The arrival state of the tag 10 can be estimated for the lower side) by the same calculation method.

As described above, the state determination device 103 in the present embodiment includes the storage unit 123, the estimation unit 133, and the determination unit 143. The storage unit 123 stores a state transition probability 123a and an output probability 123b in association with each other. State transition probability 123a, together with the transition of the state of the tag 10 in the environment 1 is the probability of transition from each state _S 0 to S ₈ to the other state. Output probability 123b is the probability that each of the plurality of symbols k indicating the detection or non-detection of the tag 10 is output at each state S ₀ to S _8. When a plurality of symbols k are sequentially output for the tag 10, the estimation unit 133 estimates the state that the tag 10 has reached using the state transition probability 123 a and the output probability 123 b stored in the storage unit 123. The determination unit 143 determines the passage of the tag 10 based on the state estimated by the estimation unit 133. Determination device 103 uses the state transition probability 123a and output probability 123b and is defined probability automaton, from the state S ₀ to S _8, it obtains the highest state probability person has arrived, from the state Determine the passage of people. In other words, the state determination device 103 estimates the state in consideration of malfunctions such as undetected and erroneous detection of the tag 10. Therefore, the state determination apparatus 103 can perform the state estimation of the moving object M carrying the tag 10 with the highest probability even when a malfunction occurs. As a result, the accuracy of state determination is improved.

  The state determination apparatus 103 in the present embodiment uses a state transition diagram having a self-loop and a path capable of bidirectionally transitioning between two states as shown in FIG. 3, FIG. 16, or FIG. The state of the tag is estimated by a state transition whose detection is an event. For this reason, the state determination device 103 does not need to prepare an infinite number of pattern tables for determining the state of the tag in accordance with the malfunction occurrence pattern of the reading device 102. That is, the state determination apparatus 103 can estimate the state that the tag has reached by using a probability automaton with a simple configuration.

  Furthermore, according to the state determination apparatus 103 in the present embodiment, the determination unit 143 determines whether the tag 10 is turned back or stagnated based on the state estimated by the estimation unit 133. Thereby, the state determination apparatus 103 can determine not only passing but also turning back and stagnation as the arrival state of the tag 10 and thus the state of the moving body M. Therefore, the state determination device 103 accurately determines this even when the moving body M returns midway without passing through the detection area, or when the moving body M stops near the detection area. be able to. As a result, the behavior of the moving body M can be grasped in more detail and with high accuracy.

  Next, the configuration of the state determination unit in the second embodiment will be described. FIG. 18 is a diagram illustrating a functional configuration of the state determination unit according to the second embodiment. FIG. 18 is a diagram illustrating a functional configuration of the state determination unit 200 according to the present embodiment. As illustrated in FIG. 18, the state determination unit 200 includes the antennas A <b> 1 and A <b> 2, the antenna switching device 201, the reading device 202, the state determination device 203, and the output unit 204. Each of these components is connected so that signals and data can be input and output in one direction or in both directions.

  The state determination device 203 determines the state of the moving body M belonging to the environment 1 based on the tag detection result input from the reading device 202, and outputs the determination result to the output unit 204. As shown in FIG. 18, the state determination device 203 includes an acquisition unit 213, a storage unit 223, an estimation unit 233, a determination unit 243, and an adjustment unit 253. The storage unit 223 stores the state transition probability 223a and the output probability 223b. The determination unit 243 further includes a passage determination unit 243a, a turnback determination unit 243b, and a stagnation determination unit 243c having a counter 244c.

  The state determination unit 200 has the same configuration as the state determination unit 100 according to the first embodiment except that the state determination unit 200 further includes an adjustment unit 253. Therefore, the same components are denoted by the same reference numerals at the end, and detailed description thereof is omitted. Specifically, the antenna switching device 201, the reading device 202, the state determination device 203, and the output unit 204 in the second embodiment are the antenna switching device 101, the reading device 102, and the state determination device in the first embodiment. 103 and the output unit 104. In addition, the acquisition unit 213, the storage unit 223, the estimation unit 233, and the determination unit 243 in the second embodiment are the acquisition unit 113, the storage unit 123, the estimation unit 133, and the determination unit 143 in the first embodiment. Correspond to each. Furthermore, the state transition probability 223a and the output probability 223b of the storage unit 223 in the second embodiment correspond to the state transition probability 123a and the output probability 123b in the first embodiment, respectively. In addition, the passage determination unit 243a and the turnback determination unit 243b of the determination unit 243 according to the second embodiment are components corresponding to the passage determination unit 143a and the turnback determination unit 143b of the determination unit 143 according to the first embodiment, respectively. . The stagnation determination unit 243c having the counter 244c in the second embodiment is a component corresponding to the stagnation determination unit 143c having the counter 144c in the first embodiment.

  Hereinafter, the adjustment unit 253 which is a main difference from the first embodiment will be described. The adjustment unit 253 learns parameters in the probability automaton (see FIGS. 16 and 17). The adjustment unit 253 automatically adjusts the state transition probability and the output probability described above to values (parameters) suitable for the environment 1 in response to changes in the environment 1. For example, the BaumWelch algorithm, which is a parameter learning method for a hidden Markov model, can be used for parameter learning.

  The state determination unit 200 according to the second embodiment has the following effects in addition to the effects achieved by the state determination unit 100 according to the first embodiment. That is, if a new object is placed near the antennas A1 and A2 installed in the environment 1 or an existing object is placed in a different manner, the reading apparatus 202 may malfunction due to radio wave shielding or multipath effects. May occur. According to the state determination unit 200 in the second embodiment, the adjustment unit 253 of the state determination device 203 appropriately updates parameters such as the existing state transition probability and output probability in accordance with such a change in the environment 1. As a result, the parameter is always tuned to an optimum value. Therefore, even if the environment 1 changes, the state determination apparatus 203 can estimate the state that the tag 10 reaches with high accuracy. As a result, a state determination that is robust against environmental changes is realized.

  According to the state determination unit 200 in the second embodiment, the adjustment unit 253 reflects the observation result by applying the observation symbol y to the state transition probability and the output probability value (parameters) set as initial values. It will be corrected to the value. The parameter is autonomously updated at any time whenever the state with the highest probability is estimated as the arrival state. The state transition probability and output probability can be statistically predicted in advance.For example, when a shielding object is placed around the antenna and the way radio waves are reflected, the state transition probability and output probability change accordingly. Probability also varies. The adjustment unit 253 automatically adjusts the value of these probabilities according to the environment by the learning effect of the hidden Markov model. In this adjustment, for example, the above BaumWelch algorithm can be applied. This makes it possible to optimally estimate the state where the tag has arrived, regardless of changes in the environment.

  Subsequently, a state determination system E configured to include the state determination units 100 and 200 in the first and second embodiments will be described. FIG. 19 is a diagram illustrating a configuration example of a state determination system having a state determination unit according to the first and second embodiments. As shown in FIG. 19, the state determination system E includes state determination units 100 and 200 and state determination units 100 and 200-2 to state determination units 100 and 200 -n (n is a natural number) having the same configuration as this. And have. These state determination units 100, 200 and 100, 200-2 to n are connected to each other via a network N so that information can be transmitted and received between them. To the same network N, a location management server E1, a tag database E2, a map database E3, and a location display device E4 are connected.

  In the state determination system E, the state determination units 100, 200 and 100, 200-2 to n are installed in a place suitable for performing a state determination of a person via a tag in a predetermined environment. The state determination results by the state determination units 100, 200 and 100, 200-2 to n are collected by the location management server E1.

  FIG. 20 is an example of a state transition diagram of the position management server E1 in the state determination system E. The location management server E1 events the reception of the state determination results transmitted from the state determination units 100, 200 and 100, 200-2 to n, and is tagged from the state transition diagram as shown in FIG. Determine the location of the moving object. In the tag database E2, the ID of the tag is associated with the information (identification number, name, etc.) of the moving object to which the corresponding tag is attached. With this association, the state of the moving object can be determined by determining the state of the tag as the moving object.

  Map data corresponding to the state transition diagram shown in FIG. 20 is stored in the map database E3. FIG. 21 is a map F of an environment in which nine state determination units 100, 200 and 100, 200-2 to 9 are installed in the state determination system E. As shown in FIG. 21, for example, state determination units 100 and 200 are installed at doors on the boundary with the rooms R1 and R2. Note that the rooms R1 and R2 in FIG. 21 correspond to the detection areas B2 and B1 shown in FIG. 1, respectively. Thereby, the state determination system E can determine the state (passing, turning back, stagnation) of a person who goes back and forth between the room R1 and the room R2. Further, the state determination units 100 and 200-7 located at the boundary between the stairs and the hallway can determine the state of a person who has finished going up the stairs and arrived at the hallway, or a person going down the stairs. Furthermore, since the state determination units 100 and 200-8 are installed near the entrance of the elevator, it is possible to determine the state of the person who goes up and down the elevator (e.g., once getting down and getting on again, waiting for the elevator). it can.

  The position display device E4 visualizes who is where based on information transmitted from each of the position management server E1, the tag database E2, and the map database E3. That is, when the position display device E4 receives the tag state determination result from the position management server E1, the position display device E4 refers to the tag database E2 to identify the person with the tag and reads the state from the map database E3. Display on map F. On the map F, the name, number, symbol, etc. that can identify the person are displayed together with the state of the moving object, so the user can see who is the determination target and where (current location) is at a glance. Can be grasped. Since the state determination result is obtained by a probabilistic automaton adapted to the environment, the display information is accurate information with little influence of undetected or erroneous detection.

  Even within the same environment, the way objects are placed and the radio field intensity of the antenna differ depending on the location. However, as described above, by installing multiple status determination units in one environment, differences in the surrounding conditions within the same environment It is possible to deal with fine grain. In addition, when state determination is performed without using a stochastic automaton, the accuracy of state determination is likely to vary depending on the location of the antenna, but in this embodiment, relative calculation is performed by using the state transition probability and the output probability. It is estimated that the state with the highest probability is the arrival state. Therefore, if any of the plurality of states has a higher probability than any other state, that state will surface as an arrival state. Thereby, the variation in the precision by the installation location of an antenna is reduced.

[Status determination program]
Various processes of the state determination apparatuses 103 and 203 described in the above embodiments can be realized by executing a prepared program on a computer system such as a personal computer or a workstation. Therefore, in the following, an example of a computer that executes a state determination program having the same function as the state determination apparatuses 103 and 203 described in the above embodiment will be described with reference to FIG. FIG. 22 is a diagram illustrating a computer that executes a state determination program.

  As shown in FIG. 22, the computer 300 in the above embodiment includes a CPU (Central Processing Unit) 310, a ROM (Read Only Memory) 320, an HDD (Hard Disk Drive) 330, a RAM (Random Access Memory) 340, Have These units 300 to 340 are connected via a bus 400.

  The ROM 320 stores in advance state determination programs that exhibit the same functions as the acquisition units 113 and 213, the estimation units 133 and 233, the determination units 143 and 243, and the adjustment unit 253 described in the above embodiment. That is, the ROM 320 stores a state determination program 320a as shown in FIG. Note that the state determination program 320a may be separated as appropriate.

  Then, the CPU 310 reads the state determination program 320a from the ROM 320 and executes it.

  The HDD 330 stores a state transition probability 330a and an output probability 330b. The state transition probability 330a and the output probability 330b correspond to the state transition probability 123a and the output probability 123b shown in FIG. 2, respectively.

  Then, the CPU 310 reads the state transition probability 330a and the output probability 330b. Then, the CPU 310 stores these in the RAM 340. The CPU 310 executes the state determination program 320a using the state transition probability 340a and the output probability 340b stored in the RAM 340. In addition, the CPU 310 executes the state determination program 320a by using the state transition probability 340a and the output probability 340b stored in the RAM 340. In addition, as for each data (state transition probability 340a and output probability 340b) stored in the RAM 340, it is not always necessary to store all data in the RAM 340, and only data necessary for processing may be temporarily stored in the RAM 340. .

  The state determination program 320a does not necessarily need to be stored in the HDD 330 from the beginning.

  For example, the computer 300 stores the program in a “portable physical medium” such as a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, and an IC card inserted into the computer 300. Then, the computer 300 may read and execute the program from these media.

  Furthermore, the program is stored in “another computer (or server)” connected to the computer 300 via a public line, the Internet, a LAN, a WAN, or the like. Then, the computer 300 may read and execute the program from these.

  In the above embodiment, an example in which the state determination device disclosed in the present application is realized by an RFID tag has been described. However, the communication method is not limited to this. For example, infrared communication may be used. In other words, an infrared light receiver is installed on the environment side, an infrared light emitter is provided on the mobile body side, and when the mobile body enters the detection area of the light receiver, an ID signal transmitted by the light emitter as a moving object is received. It may be a system that the receiver receives. Further, the communication method is not limited to wireless, but may be wired communication.

  In the above embodiment, although two antennas are installed in the environment, three or more antennas may be used. Moreover, although the said Example demonstrated on the premise that the radio wave arrival range (detection area) of several antennas overlaps, these do not need to overlap.

  The following additional notes are further disclosed with respect to the embodiments including Examples 1 and 2 above.

(Supplementary note 1) Along with the transition of the state of a moving object in the environment, a state transition probability that is a probability of transition from each state to another state, and each of a plurality of detection information indicating detection or non-detection of the moving object in each state A storage unit that stores an output probability that is an output probability in association with each other;
When a plurality of detection information is sequentially output for the moving object, using the state transition probability and the output probability stored in the storage unit, an estimation unit that estimates the state reached by the moving object;
A determination unit that determines the passage of the moving object based on the state estimated by the estimation unit;
A state determination device comprising:

(Additional remark 2) The said determination part determines the return and stagnation of the said moving body based on the state estimated by the said estimation part, The state determination apparatus of Additional remark 1 characterized by the above-mentioned.

(Supplementary note 3) The state determination apparatus according to Supplementary note 1 or 2, further comprising an adjustment unit that adjusts the state transition probability and the output probability according to a change in the environment.

(Appendix 4) A state determination method executed by a computer,
Probability of transition from one state to another with the transition of the state of the moving object when a plurality of detection information indicating detection or non-detection of the moving object is sequentially output in each state for the moving object in the environment Using the state transition probability and the output probability that is the probability that each of the plurality of detection information is output in each state, the state that the moving object has reached is estimated,
A state determination method, wherein the passage of the moving object is determined based on the estimated state.

(Appendix 5)
Probability of transition from one state to another with the transition of the state of the moving object when a plurality of detection information indicating detection or non-detection of the moving object is sequentially output in each state for the moving object in the environment Using the state transition probability and the output probability that is the probability that each of the plurality of detection information is output in each state, the state that the moving object has reached is estimated,
A state determination program for executing a process of determining the passage of the moving object based on the estimated state.

1 environment 10 tag 100, 200 state determination unit 100, 200-2 to n state determination unit 102, 202 reading device 103, 203 state determination device 113, 213 acquisition unit 123 storage unit 123a, 223a state transition probability 123b, 223b output probability 133, 233 Estimation unit 143, 243 Determination unit 143a, 243a Passage determination unit 143b, 243b Turnback determination unit 143c, 243c Stagnation determination unit 144c, 244c Counter 104, 204 Output unit 253 Adjustment unit 300 Computer 310 CPU
320 ROM
320a State determination program 330 HDD
340 RAM
330a, 340a State transition probability 330b, 340b Output probability 400 Bus A1, A2 Antenna B1, B2, B3 Detection area E State determination system E1 Position management server E2 Tag database E3 Map database E4 Position display device M Mobile body N Network S ₀ , _{S 1, S 2, S 3} , S 4, S 5, S 6, S 7, S 8 state

Claims (5)

Along with the state transition of the moving object in the environment, a state transition probability that is a probability of transition from each state to another state, and a probability that each of the plurality of detection information indicating detection or non-detection of the moving object in each state is output A storage unit that stores the output probabilities in association with each other,
When a plurality of pieces of detection information are sequentially output for the moving object, the state at which the moving object has arrived is estimated using the state transition probability and the output probability stored in the storage unit, and the event is self-timed. The state of the moving object is estimated by the state transition using the detection of the moving object as an event, using a state transition diagram having a self-loop that returns to the state and a path that can bidirectionally transition between the two states. An estimation unit;
A determination unit that determines the passage of the moving object based on the state estimated by the estimation unit;
A state determination device comprising:   The state determination apparatus according to claim 1, wherein the determination unit determines whether the moving object is turned back or stagnated based on the state estimated by the estimation unit.   The state determination apparatus according to claim 1, further comprising an adjustment unit that adjusts the state transition probability and the output probability according to a change in the environment. A state determination method executed by a computer,
Probability of transition from one state to another with the transition of the state of the moving object when a plurality of detection information indicating detection or non-detection of the moving object is sequentially output in each state for the moving object in the environment The state transition probability and the output probability that is the probability that each of the plurality of pieces of detection information is output in each state are used to estimate the state that the moving object has arrived and Using a state transition diagram having a returning self-loop and a path capable of bi-directionally transitioning between two states, estimating the state of the moving object by state transition using the detection of the moving object as an event ,
A state determination method, wherein the passage of the moving object is determined based on the estimated state. On the computer,
Probability of transition from one state to another with the transition of the state of the moving object when a plurality of detection information indicating detection or non-detection of the moving object is sequentially output in each state for the moving object in the environment The state transition probability and the output probability that is the probability that each of the plurality of pieces of detection information is output in each state are used to estimate the state that the moving object has arrived and Using a state transition diagram having a returning self-loop and a path capable of bi-directionally transitioning between two states, estimating the state of the moving object by state transition using the detection of the moving object as an event ,
A state determination program for executing a process of determining the passage of the moving object based on the estimated state.

Images