KR101094688B1

KR101094688B1 - Indoor moving object database updating method

Info

Publication number: KR101094688B1
Application number: KR20090114613A
Authority: KR
Inventors: 임재걸; 정병윤
Original assignee: (주)대산; 동국대학교 경주캠퍼스 산학협력단
Priority date: 2009-11-25
Filing date: 2009-11-25
Publication date: 2011-12-20
Also published as: KR20110057967A

Abstract

A method of calculating the velocity of a moving object based on the measured position of the indoor moving object and updating the indoor moving object database is provided. The mobile terminal measures its current location. Based on a series of indoor positioning results, the moving speed and the current position of the moving object are predicted. The difference between the measured current position of the mobile terminal and the predicted current position of the mobile terminal is compared with an allowable error. According to the comparison result, it is determined whether to update the database of the measured current position of the mobile terminal.

Indoor moving object database update, Kalman filter, oblivion rate

Description

Indoor moving object database updating method

The present invention relates to a database update, and more particularly, to a method for updating an indoor mobile object database, which is a key element of an indoor location based service system.

There is a moving object database update technique in the prior art. In the conventional moving object database, it is assumed that the moving object carries a GPS receiver. Since GPS provides not only the position of the moving object but also the current speed of the moving object, the conventional moving object database update technique does not need to study the method of obtaining the current speed of the moving object.

GPS signals cannot be received indoors. Therefore, the indoor mobile object is required to carry a mobile terminal such as a portable PC or PDA capable of wireless Internet, not a GPS receiver, to enable indoor location-based services. Indoor location-based services provide services based on the location of indoor moving objects, so a moving object database containing location information according to the time of indoor moving objects is essential, but the results of research on indoor moving object databases are still insignificant.

In the conventional moving object database, the moving object is equipped with a GPS receiver, so that the speed of the moving object provided by the GPS can be used, but there is a problem that GPS cannot be used because the GPS signal is too weak indoors.

Since GPS cannot be used indoors, various studies on the method of measuring the location of indoor mobile terminals have been conducted, but there is no research on an efficient method of storing the location information of indoor mobile terminals in a mobile object database. It only records the location and time of measurement on a regular basis.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and an object thereof is to provide a method of calculating a moving object's speed based on a measured position of an indoor moving object and updating the indoor moving object database based on the measured position. have.

Another object of the present invention is to estimate the position of the current point of view of the current position by a series of position measurement of the moving object in the process of recording the position information of the moving object of the indoor moving object database for indoor location-based services The present invention provides a method for updating an indoor moving object database that controls whether a measurement position of a current time point is included in a moving object database according to a similar comparison result.

In order to achieve the above object, the indoor mobile object database update method according to the present invention comprises the steps of (i) the mobile terminal measuring its current location; (ii) predicting the moving speed and the current position of the moving object based on the series of indoor positioning results; (iii) comparing the difference between the current position of the mobile terminal measured in step (i) and the current position of the mobile terminal predicted in step (ii) with a tolerance; And (iv) determining whether to update the database with respect to the measured current position of the mobile terminal according to the comparison result of step (iii).

The present invention saves communication costs by estimating the position of the current point of time with past measurement positions of the portable mobile terminal, and omitting transmitting the current measurement position to the moving object database if the estimated position is not very different from the current measurement position. It also saves on database storage costs.

The present invention provides an efficient update strategy of an indoor mobile object database. The present invention is fundamentally different from the determination of whether to perform a location update or a handoff decision in an existing broadband wireless communication system in the sense of an update strategy of an indoor mobile object database for an indoor location based service. In broadband wireless communication, the error range is determined by the base station unit to which the signal reaches, and the error range is several tens of kilometers. For example, if an area covered by a base station A is 10 km in radius, the terminal determines whether it is within 100 meters or 1 km of the base station A within the base station A area.

Hereinafter, an indoor moving object database update method according to an embodiment of the present invention will be described in detail with reference to the accompanying example drawings.

The mobile object has a portable terminal, and the terminal transmits its location to the server computer via the wireless Internet, and the server computer stores the location of the received portable mobile terminal in the mobile object database. The server computer not only provides a location-based service to a portable mobile terminal but also provides a useful service based on the location information of the mobile object to other PCs connected online or via the Internet. The server computer generally uses a drawing database to indicate the position of the moving object on the drawing. In general, knowing the measurement positions p _i and p _j at time points t _i and t _j

In the case of obtaining the furnace speed but having a large error in the measurement position, the calculated speed is also large, so the present invention provides a method using the Kalman filter method.

1 is a block diagram of an indoor location-based service system for an embodiment of the present invention. As shown in FIG. 1, the server computer 30 stores the measurement position of the portable mobile terminal transmitted by the portable mobile terminal 40, which is a mobile object, in the mobile object database 20, and then the portable mobile terminal 40. May be provided to the portable terminal 40 itself or to other computers 50 and 60 in consideration of its current indoor location.

The data transmitted by the portable mobile terminal 40 is composed of its own identifier (MoId), measurement time (time), and measurement location (location), which is a mobile object as shown in FIG. The moving object database 20 stores location information of a plurality of moving objects, and distinguishes each other by assigning a unique identifier to each moving object. Each portable mobile terminal 40 periodically measures its indoor location, for example every second.

The method of measuring the indoor location of the portable mobile terminal 40 is a reception received from an access point (not shown) installed in a wireless local area network using a LAN card (not shown) as shown in FIG. 3. The received signal strength indication (hereinafter referred to as 'RSSI') is used as an input to the decision tree 31 to find its indoor location.

As shown in FIG. 4, the discretization process is performed using the discretization program 42 as the training data 41. By executing the decision tree building program using the discretized training data 43 as input to the decision tree building program 44, the decision tree 31 of FIG. 3 is generated. The training data consists of RSSI values read multiple times at each candidate point as shown in FIG. The candidate point is a point where the moving object may be located, and the application area is divided into a checkerboard pattern, and the point where the segment meets in a cross pattern in the checkerboard pattern.

For example, in the first row of FIG. 5, the first RSSI reading result from candidate point A shows that the RSSI value of router 1 (AP1) is -50, the RSSI value of router 2 (AP2) is -48, and so on. , n indicates that the RSSI value of router n is equal to -52, and the next row indicates that the RSSI values from AP1 to APn at the same candidate point A are -47, -50,... , -51. If the area marked Micro Lab in FIG. 6 is an application area, all points shown in FIG. 6 are candidate points.

In the decision tree method, the process 32 of measuring the current indoor location of the portable mobile terminal 40 substitutes the RSSI values read from the current indoor location into the decision tree 31 and goes down to the child node. Repeat until you reach the leaf node of tree 31. The label of the leaf node of the decision tree 31 at which the repetition step ends is determined as the current position of the portable mobile terminal 40.

An example of such a decision tree is shown in FIG. If the RSSI values 33 read at the current position are as shown in FIG. 8, the decision process 32 is performed by the root node of the decision tree 31 as AP2, so that the RSSI value of AP2 shown in FIG. The next child node to be visited is determined and descended to node 2 of the decision tree 31. Since the label of the node 2 is AP4, the decision process 32 determines the child node as the fourth term of FIG. 8, that is, -58. In the case of the decision tree 31 shown in Fig. 7, the procedure goes down to the node 16. Since the label of the node 16 is AP3, the third node of FIG. 8 determines the child node. This process is repeated until it reaches the leaf node. The label of the leaf node of the decision tree 31 at which the repetition step ends is determined as the current position of the portable mobile terminal 40.

Referring to FIG. 9, the portable mobile terminal 40 of FIG. 1 measures its current position by the decision tree method of FIG. 3 and based on a series of past recorded positions that have already been recorded, as shown in FIG. 10. Similarly, estimate the current position (step 91). Then, it is determined whether the distance between the measured value and the estimated value is smaller than the allowable error (step 92). It is preferable to make the said tolerance into the average error of a measured value.

As a result of the determination in step 92, if the distance between the measured value and the estimated value is smaller than the allowable error, the counter and the flag are set to 0, respectively, and the recording and transmission, that is, updating of the moving object database 20 is omitted (step 93). Only when the measurement position is transmitted to the server computer 30 in the form of FIG. 2, namely, the identifier of the moving object, the measurement time, and the measurement position. At this time, in order to deal with the case where the measured value is abnormal, it is transmitted only when the distance between the measured value and the estimated value is not smaller than the allowable error several times (th3) or more. Otherwise, it is stored in the buffer and continuously more than th3 times. When the difference between the measured value and the estimated value is larger than the tolerance, the measured values stored in the buffer are transmitted at once. That is, if the difference between the measured current indoor position and the predicted current position is less than the tolerance, the counter is incremented (step 94). It is determined whether less than a predetermined number th3 has occurred when the difference between the counter value, that is, the measured current indoor position and the predicted current position is greater than the allowable error (step 95). According to the embodiment of the present invention, in order to ignore the outliers using 3 m as the tolerance, the update is resumed when the difference between the predicted value and the measured value is greater than or equal to th3 times the tolerance. The th3 is an upper limit value for the outliers to be continuous. When the outliers are successively more than th3 times, 3 is used as a value to conclude that the measured value is no longer the outlier. However, the present invention is not limited thereto. The smaller the tolerance 92, the greater the update frequency, and the more accurate the position information of the moving object recorded in the database. On the contrary, increasing the value may reduce the update frequency, thereby reducing the communication cost and the data storage cost. Instead, the location information of the moving object recorded in the database may be inaccurate.

As a result of the determination in step 95, when the difference between the measured current indoor position and the predicted current position is more than the allowable error less than a predetermined number th3, the measured value which is the measured current position is added to the buffer (step 96). If the difference between the measured current indoor position and the predicted current position is greater than or equal to the allowable error, it is determined whether the flag is set to 0 (step 97).

If the flag in step 97 is not set to 0, only the measurement is stored in its own database and transmitted to the server computer 30 (step 98). If the flag is set to 0, the buffer and measurements are stored in its own database and sent to the server computer 30 to update the moving object database (step 99).

10 is a flowchart illustrating a process of estimating a current position at a current time t _k by applying a previously recorded measurement position to a Kalman filter process. The Kalman filter repeats the prediction of the state and the update of the prediction value using the measured value in the process of obtaining a value indicating the state (Welch and Bishon, 2007). For example, in the case of indoor positioning, the process of predicting the position of the terminal to an arbitrary value and updating the measured value is reflected, and based on this, the process of predicting the current position and reflecting the measured value is updated. By repeating, the position of the terminal is determined. This process is applied to a Kalman filter on the measurement position of the terminal contained in the database of its own, t _k before the time set at the time _{t o, t 1, ...,} t i by the speed of the moving object from the time point t _i It estimates and estimates the position of the time point t _k from this velocity and the measurement position of the time point t _i . The measurement positions of a series of time points t _o , t ₁ , ..., t _i before the time point t _k are also recorded in the database of the terminal itself and the moving object database 20 of the server computer 30. 10 first retrieves the measurements of portable mobile terminal 40 at time points t _o , t ₁ , ..., t _i from the terminal's own database (not shown) (step 101), the position of the Kalman filter process (step 102) an application to calculate the speed of the moving object at the time t _i, and applies the speed of the mobile object to a measuring position at the time point t _i portable mobile at time t _k terminal 40 Estimate (step 103).

In the Kalman filter process, as shown in FIG. 11, the initialization process (step 111) initializes the X_current vector to the measurement of the time t _o , initializes the P_matrix representing the error covariance to a large value, and predicts the current state from the last state. The Phi_matrix used for this purpose, the Q_matrix indicating the error of the predicted value, and the R_matrix indicating the error of the measured value are initialized as shown in FIG. 12 (111). In step 112, the prediction and correction processes are repeated with the measurements from t ₁ to t _i to obtain the state of the moving object, that is, the position and velocity at X_current at time t _i .

FIG. 13 is a flowchart for explaining the position estimation step (step 103) of FIG.

Referring to FIG. 13, it is determined whether all terms of P_matrix representing error covariance are smaller than th2 (step 1132).

As a result of the determination in step 1132, if all terms of P_matrix are smaller than th2, it is assumed that X_current is correct, and the result of multiplying the velocity (x _v ) in the x-axis direction by the elapsed time (t _k -t _i ) by the x coordinate In addition, the x coordinate at the time point t _k is obtained, and in the same manner, the y coordinate at the time point t _k is also obtained by adding the result of multiplying the velocity (y _v ) in the y axis direction by the elapsed time (t _k -t _i ) to the y coordinate. . Based on the X_current, an estimate of the portable mobile terminal 40 at time t _k is calculated by the following Equation 1 (step 1133).

Estimate =

However, as any one of the preceding even not a small value for the estimated position in the Kalman filter process, assume that the mobile object status at the time point t _i report inaccurate reliable the X_current and the time t _k, obtained by the just origin of P_matrix ( Step 1134).

In the present invention, as shown in Fig. 9, an estimate of the current position is obtained and used based on past measurements. In addition to using the Kalman filter shown in FIG. 10 as a method for obtaining the estimated value, there may be various methods such as the method using the forgetting rate shown in FIG. 15. Similar to the Kalman filter method of FIG. 10, the method of using the forgetting rate of FIG. 15 is a series of time points t _o and t before the time point t _k , which is first stored in the database of the terminal itself in order to estimate the current position at the current time t _k . _1, ..., and it searches the measuring position _{m o, m 1, ...,} m i t _i on (step 151). Speed of the moving object at the time t _i is a _i m - can be expected to be similar to m _i-1. It can also be expected that the velocity of the moving object at time t _i will also be related to m _i-1 -m _i-2 . However, m _i - m _i-1 value if the influence on the moving object speed at the time point t _i that _{α m i-1 - m i} -2 values effect on the moving object speed at the time point t _i Can be said to be 1-α. Therefore, the method using the oblivion ratio α to obtain the velocity of the moving object at the time t _i is determined the velocity at every point, and then (step 152), obtaining the total sum by applying the forgetting rate α to those them at the time t _i one of a speed of the mobile object, 153, of the mobile object at the current time point t _k estimated position is calculated by adding the product of the speed and the elapsed time calculated in step 153 for the measurement of the mobile object at the time t _i where m _i ( Step 154).

14 shows a process in which the portable mobile terminal 40 transmits the current position to the mobile object database 20 of the server computer 20. The indoor positioning module (FIG. 3) executed in the portable mobile terminal 40 measures the current position and sends it to the update program (FIG. 9) (step 141). The update program estimates the current position with a recent measurement as described in FIG. 9 and then discards it if the difference between the estimate and the measurement is within tolerance, otherwise records the measurement in the database of the terminal itself (not shown). And also to the communication module (step 142). The communication module bundles the terminal identifier, the measurement time, and the measurement position as shown in FIG. 2 and transmits it to the server computer 30 (step 143). The communication module of the server computer 30 receives the packet sent by the portable mobile terminal 40 and delivers it to the mobile object database 20 (step 144). The mobile object database 20 may send content or useful information considering the current location of the portable mobile terminal 40 to the portable mobile terminal 40 (step 145).

Although the present invention has been described as a specific preferred embodiment, the present invention is not limited to the above-described embodiments, and the present invention is not limited to the above-described embodiments without departing from the gist of the present invention as claimed in the claims. Anyone with a variety of variations will be possible.

Indoor location-based services include a service that grasps the current location of visitors or customers in a museum or a department store and provides a variety of information considering a location in real time, and a service that extracts useful information by analyzing already recorded location data. Examples of information provided in real time include cultural content related to the exhibits currently being viewed, the nearest teacher, and the path to the exhibits of interest. In addition, by analyzing the recorded position data can be seen the exhibitor's favorite exhibits, the frequency of rest of the viewer. These indoor location-based services have countless applications such as subway station management and factory management. Although there are few indoor location-based services, indoor location-based services will soon become a part of daily life as GPS-based location-based services are widely used, and the demand for the present invention will explode accordingly.

1 is an exemplary configuration diagram of an indoor location based service system according to an exemplary embodiment of the present invention.

2 is a data format transmitted by a mobile terminal to a server computer according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating a process of measuring the current location of the portable mobile terminal illustrated in FIG. 1.

4 is an exemplary diagram illustrating a process of creating a decision tree from training data in the process of FIG. 3.

5 is a diagram showing an example of training data created by RSSI read from a router of a wireless local area network according to an embodiment of the present invention.

6 illustrates candidate points according to an embodiment of the present invention.

FIG. 7 is a diagram illustrating an example of a decision tree created from the training data shown in FIG. 5.

8 is a diagram illustrating an example of RSSI signals read from a current position by a portable terminal according to an exemplary embodiment of the present invention.

9 is a flowchart illustrating a process of determining whether to update a moving object database according to an embodiment of the present invention.

10 is a flowchart for explaining a current position estimation step shown in FIG.

FIG. 11 is a diagram illustrating an algorithm of the Kalman filter process of FIG. 10.

12 is a diagram illustrating initial values of various variables used in the Kalman filter processing algorithm of FIG.

FIG. 13 is a flowchart for explaining an example of the position estimation step in FIG. 10.

FIG. 14 is a diagram illustrating a process of transmitting to the server computer by the portable mobile terminal illustrated in FIG. 1.

15 is a flowchart illustrating a process of obtaining an estimate of a current position using a forgetting rate according to an embodiment of the present invention.

Claims

(i) the mobile terminal measuring its current location;

(ii) predicting the moving speed and the current position of the moving object based on the series of indoor positioning results;

(iii) comparing the difference between the current position of the mobile terminal measured in step (i) and the current position of the mobile terminal predicted in step (ii) with a tolerance; And

(iv) determining whether to update the database of the measured current position of the mobile terminal according to the comparison result of step (iii),

Step (i) is

(i-1) discretizing training data consisting of RSSI values of each router read multiple times at each candidate point where the mobile object is likely to be located;

(i-2) generating the decision tree by inputting the discretized training data to the decision tree building program 44 to execute the decision tree building program;

(i-3) repeating the process of descending to the child node by substituting the RSSI values read from the current position into the decision tree until the leaf node of the decision tree arrives; And

(i-4) determining the label of the leaf node at which the repetition step ends as the current position of the portable mobile terminal.

The indoor mobile object database of claim 1, wherein when the mobile terminal receives a signal from a router installed in a wireless local area network, step (i) determines the location of its own indoor home based on the strength of the received signal. Update method.

The method of claim 2, wherein step (i) converts the strengths (RSSI) of the signals received from the at least three routers into distances, respectively, to determine the coordinates of the at least three routers and the distance between the portable mobile terminal and each router. The method of updating the indoor moving object database which obtains the coordinates of the portable mobile terminal as the position of the portable mobile terminal by substituting the triangulation method into the triangulation method.

delete

The method of claim 1, wherein step (ii)

(ii-1) retrieving measurement positions at a time point t _o , t ₁ , ..., t _i from a database of the mobile terminal itself;

(ii-2) calculating a speed of the moving object at the time t _i by applying the Kalman filter process to the detected measurement position at the time _{t o, t 1, ...,} t i; And

(ii-3) estimating the position of the moving object at time t _k by applying the velocity of the moving object to the measurement position at the time point t _i .

The method of claim 6, wherein the Kalman filter process,

Initialize the X_current vector to the measurement of time t _o , initialize the P_matrix representing the error covariance to a large value, and use the Phi_matrix used to predict the current state from the last state, the Q_matrix representing the error of the prediction, and the error of the measurement. Initializing the R_matrix; And

Repeating the prediction and correction process with measurements from time t ₁ to t _i to obtain the position and velocity of the moving object in X_current at time t _i .

8. The method of claim 7, wherein step (ii-3)

Determining whether all terms of P_matrix representing the error covariance are less than th2;

If all terms of the P_matrix are smaller than th2, the X_current is assumed to be correct, and the result of multiplying the velocity (x _v ) in the x-axis direction by the elapsed time (t _k -t _i ) to the x coordinate is added to the x coordinate at the time point t _k . Obtaining the x coordinate of, and calculating the y coordinate at the time point t _k by adding the result of multiplying the velocity y _{v in the} y-axis direction by the elapsed time (t _k -t _i ) to the y coordinate;

Based on the X_current, an estimate of the portable mobile terminal at time t _k is expressed by Equation:

Obtaining as; And

A small value as set forth any of the P_matrix or it's just the origin of the presumed location of the report inaccurate, the mobile object's state at the time t _i, and assuming no reliable the X_current time t _k, obtained by the Kalman filter process And determining the indoor moving object database.

The method of claim 1, wherein step (ii)

(ii-a) retrieve the measurement positions (m _o , m ₁ , ..., m _i ) of the moving object at a series of time points t _o , t ₁ , ..., t _i before time t _k Making;

(ii-b) obtaining a velocity of the moving object at the time point t _k ;

the effect on the speed of the moving object at the point in time _i-1 and m is a value t _i - (ii-c) above to a series of time points t _o, t _1, ..., mobile object speed at the t _i m _i Obtaining the sum obtained by applying the forgetting rate α as the velocity of the moving object at the time point t _i ; And

(ii-d) the mobile addition product of the speed and the elapsed time of movement of the object from the time point t _i obtained in step (ii-3) to the moving object measuring position m _i at the time t _i at the present time t _k A method of updating an indoor moving object database, comprising: obtaining an estimated position of an object.

The method of claim 1, wherein after performing step (iii), it is determined whether less than a predetermined number of cases in which the difference between the current position measured in step (i) and the current position predicted in step (ii) is greater than or equal to the tolerance. More steps,

In step (iv), if the difference between the current position measured in step (i) and the current position predicted in step (ii) is greater than or equal to the tolerance, the current position is provided to the moving object database and updated. And skipping updating the moving object database when the difference between the current position measured in step (i) and the current position predicted in step (ii) is greater than or equal to the allowable error. How to update the object database.

11. The method of claim 10, wherein the tolerance is an error average and the predetermined number is three.