KR20160150504A - Method and apparatus for localization mobile robot usign indoor magnetic field - Google Patents

Abstract

A method and an apparatus for recognizing a position of a mobile robot using an indoor magnetic field are disclosed. A method of recognizing a position of a mobile robot according to an embodiment of the present invention includes generating a loop closing constraint based on magnetic field values measured for nodes formed on a moving path of the mobile robot; And estimating a position of the mobile robot on the basis of the graph structure using the loop closing constraint. The method may further include generating a rotation constraint condition based on a magnetic field value measured while the mobile robot rotates And the step of estimating the position of the mobile robot may estimate the position of the mobile robot on the basis of the graph structure using the loop closing constraint and the rotational constraint condition.

Description

TECHNICAL FIELD [0001] The present invention relates to a method and an apparatus for locating a mobile robot using an indoor magnetic field,

Embodiments of the present invention relate to a method and apparatus for recognizing a position of a mobile robot in an indoor environment and a method of recognizing or estimating the position of the mobile robot in an indoor environment using a SLAM (simultaneous localization and mapping) And more particularly, to a method and apparatus for recognizing a position of a mobile robot.

Simultaneous Localization and Mapping (SLAM) technology, which is one of the position estimation and mapping techniques for mobile robots, can be divided into the following two types. The first is a method of recognizing a location in an environment with a specific marker (such as Landmark or Beacon) that the mobile robot can recognize. The second is a way to recognize the location in a typical environment without specific markings. In the second method, various sensors are used to recognize the position in an environment without a specific marker.

As a typical sensing method used at this time, a sensing method utilizing an image, a laser, and an ultrasonic wave can be mentioned. Conventionally, these sensing methods are used to find a landmark that is distinguishable from other locations, and these points are discontinuously generated and recognized.

As described above, the position can be estimated by using the camera image, the ultrasonic wave, the infrared ray, the Wi-Fi, etc. However, since the signal range and the estimation performance are limited, the ambient sensor value Is very useful. In case of the indoor magnetic field, the value distorted by the steel frame in the building has a stable value with time, and since the sensor for measuring this value is inexpensive, it can be an appropriate alternative as the positioning auxiliary signal.

However, since the magnetic field value has only three vector elements, a similar value may be generated depending on the space, thus causing an ambiguity problem. When a robot or a person navigates using simultaneous localization and mapping (SLAM) in the absence of a dictionary database of magnetic field maps, the ambiguity problem of the magnetic field can cause data association problems.

Embodiments of the present invention provide a method and apparatus for recognizing a position of a mobile robot that can solve the ambiguity problem of a magnetic field in an indoor environment and improve the accuracy of position estimation of the mobile robot.

Embodiments of the present invention provide a method and apparatus for locating a mobile robot capable of estimating a position of a mobile robot through a SLAM based on a graph structure using an indoor magnetic field.

A method of recognizing a position of a mobile robot according to an embodiment of the present invention includes generating a loop closing constraint based on magnetic field values measured for nodes formed on a moving path of the mobile robot; And estimating a position of the mobile robot based on a graph structure using the loop closing constraints.

Further, a method of recognizing a position of a mobile robot according to an embodiment of the present invention may further include generating a rotational constraint condition based on a magnetic field value measured while the mobile robot rotates, The position of the mobile robot can be estimated based on the graph structure using the loop closing constraint and the rotational constraint condition.

Wherein the generating the loop closing constraint comprises: grouping the formed nodes to form a node group; comparing and matching the magnetic field values for the predefined node group with the measured magnetic field values for the formed node group, Constraints can be created.

Wherein the step of generating the loop closing constraint condition comprises: forming the node group only when the mobile robot is traveling in a straight line section; performing matching with the predefined node group only when a new node group is formed; You can create a loop-closing constraint.

Wherein the generating the loop closing constraint comprises: calculating a matching score by obtaining a Euclidean distance between the vector of the formed node group and the vector of the predefined node group, and using the calculated matching score, Closing constraints can be created.

Furthermore, a method of recognizing a position of a mobile robot according to an embodiment of the present invention includes acquiring odometry information of the mobile robot and generating odometry constraint conditions based on the obtained odometry information Wherein the step of estimating the position of the mobile robot can estimate the position of the mobile robot on the basis of the graph structure using the loop closing constraints and the omometry constraints.

The step of estimating the position of the mobile robot can estimate the position of the mobile robot using SLAM (Simultaneous Localization and Mapping) based on a graph structure.

According to another aspect of the present invention, there is provided a method of recognizing a position of a mobile robot, the method comprising: generating a constraint based on a magnetic field value measured while the mobile robot rotates; And estimating a position of the mobile robot on the basis of a graph structure using the rotational constraint condition.

The step of generating the rotational restraint condition may generate a rotational restraint condition for the rotational angle of the mobile robot using the relationship between the magnetic field vectors measured before and after the rotation of the mobile robot.

Further, a method of recognizing a position of a mobile robot according to another embodiment of the present invention includes acquiring odometry information of the mobile robot and generating odometry constraint conditions based on the acquired odometry information Wherein the step of estimating the position of the mobile robot can estimate the position of the mobile robot on the basis of the graph structure using the rotational constraint condition and the odometry constraint condition.

The apparatus for recognizing a position of a mobile robot according to an embodiment of the present invention includes a constraint for creating a loop closing constraint based on measured magnetic field values for nodes formed on a moving path of the mobile robot, Generating unit; And a position estimator for estimating a position of the mobile robot based on a graph structure using the loop closing constraints.

Wherein the constraint condition generator generates a rotation constraint condition based on a magnetic field value measured while the mobile robot is rotating, and the position estimator generates a constraint condition based on the loop constraint condition and the rotational constraint condition, The position can be estimated.

The constraint condition generating unit forms the node group by grouping the formed nodes and compares the magnetic field values measured for the formed node group with the magnetic field values for the predefined node group to generate the loop closing constraint .

Wherein the constraint condition generator forms the node group only when the mobile robot travels a straight line section and performs matching with the predefined node group only when a new node group is formed, Can generate

The constraint condition generator may calculate a matching score by obtaining a Euclidean distance between the vector of the formed node group and the vector of the predefined node group, and generate the loop closing constraint using the calculated matching score can do.

Further, the apparatus for recognizing a position of a mobile robot according to an embodiment of the present invention may further include an encoder for obtaining odometry information of the mobile robot, and the constraint condition generator may generate an odometry information based on the obtained odometry information. And the position estimator can estimate the position of the mobile robot based on the graph structure using the loop closing constraints and the omometry constraints.

According to the embodiments of the present invention, it is possible to solve the ambiguity problem of the magnetic field in the indoor environment and improve the accuracy of the position estimation of the mobile robot.

In particular, embodiments of the present invention solve the ambiguity problem of the magnetic field in an indoor environment by grouping a series of consecutive values rather than a single measurement value, thereby accurately locating the mobile robot Can be estimated.

According to embodiments of the present invention, the position of the mobile robot can be estimated through the graph structure-based SLAM using the indoor magnetic field, and the position of the mobile robot can be estimated using only one sensor for measuring the magnetic field.

FIG. 1 shows a configuration of a position recognition apparatus for a mobile robot according to an embodiment of the present invention.
Fig. 2 shows an example for explaining rotational constraint conditions.
Fig. 3 shows an example for explaining the loop closing constraint.
FIG. 4 shows an example of a process of optimizing a position graph.
FIG. 5 shows an example of the path estimation result according to the present invention.
6 is a flowchart illustrating a method of recognizing a position of a mobile robot according to an embodiment of the present invention.
7 is a flowchart illustrating a method of recognizing a position of a mobile robot according to another embodiment of the present invention.
8 is a flowchart illustrating a method of recognizing a position of a mobile robot according to another embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following, although limited embodiments are described, these embodiments are examples of the present invention and those skilled in the art can easily modify these embodiments.

Embodiments of the present invention solve the ambiguity problem of the magnetic field in an indoor environment by using a magnetic field value by grouping a series of consecutive values rather than a single measurement value and accurately estimating the position of the mobile robot That is the point.

Embodiments of the present invention can be applied to a magnetic field value measured during rotation while the mobile robot is stationary, for example, rotational constraint conditions generated based on a magnetic field vector, and magnetic field values measured for nodes formed on the movement path The SLAM based on the graph structure using the loop closing constraints generated based on the position of the mobile robot can be estimated.

FIG. 1 shows a configuration of a position recognition apparatus for a mobile robot according to an embodiment of the present invention.

1, a mobile robot position recognition apparatus 100 according to an embodiment of the present invention includes a magnetic field sensor 110, an encoder 120, a constraint condition generation unit 130, and a position estimation unit 140 .

The magnetic field sensor 110 is mounted on a mobile robot and measures a geomagnetic field. One magnetic field sensor is mounted on the mobile robot to measure a geomagnetic field at a position of the mobile robot.

The encoder 120 acquires odometry data of the mobile robot and provides the obtained odometry information to the position estimator 140. [

At this time, the encoder 120 can detect the rotational speed and the rotational direction of the electric wheel of the mobile robot, acquire the odometry information of the mobile robot using the detected rotational speed and the rotational direction, May include the x-axis coordinate, the y-axis coordinate, and the angle of the traveling direction of the mobile robot.

The constraint condition generation unit 130 is a configuration for generating a constraint based on a graph structure, and generates an ellometry constraint, a rotation constraint condition, and a loop closing constraint.

Here, the odomometry constraint means a constraint created based on the odometry data acquired by the encoder, and the rotational constraint condition is a constraint created based on the magnetic field value measured while the mobile robot rotates And the loop closing constraint means a constraint created based on the measured magnetic field values for the nodes formed on the movement path of the mobile robot.

The constraint condition generation unit 130 may generate only the omometry constraint and the loop closing constraint according to the situation, or may generate only the omometry constraint and the rotation constraint condition, You can also create both constraints and rotational constraints.

At this time, the constraint condition generation unit 130 can generate a loop closing constraint condition based on the measured magnetic field values for the nodes formed on the movement path of the mobile robot. The constraint condition generation unit 130 groups the formed nodes to form a node group , And comparing the magnetic field values for the formed node group with the magnetic field values for the predefined node group to create a loop closing constraint.

The constraint condition generation unit 130 forms a node group only when the mobile robot travels in a straight line section, performs matching with a predefined node group only when a new node group is formed, Can be generated. Here, the constraint condition generator 130 calculates a matching score by obtaining a Euclidean distance between the vector of the formed node group and the vector of the predefined node group, and calculates the matching score using the calculated matching score to calculate the loop closing constraint condition Can be generated.

The constraint condition generation unit 130 can generate a rotation constraint condition when the mobile robot rotates while the mobile robot is stationary while the mobile robot rotates while the mobile robot is stationary. Can be generated in a state where it is assumed to be constant on the base station. That is, since the constraint condition generator 130 rotates together with the magnetic field sensor when the mobile robot rotates, the constraint condition for the rotation angle of the robot can be generated using the relationship between the magnetic field vectors measured before and after the rotation.

That is, when the mobile robot moves on a two-dimensional plane and the mobile robot rotates without any translational displacement, the magnetic field vector appearing in the global coordinate frame does not change. Therefore, By using the measurement characteristics of the magnetic field in the rotor, a rotational restraint condition can be generated.

The constraint condition generator 130 may generate a loop closing constraint condition by considering the magnetic field as one fingerprint and comparing and matching the currently observed magnetic field value with the previously observed value. However, since the magnetic field vector has only three elements, ambiguity occurs in the matching process. To solve this problem, the nodes generated while the robot travels a certain distance are grouped, and the measured magnetic field values for the corresponding nodes are defined as one vector and used for matching. In the case of node grouping, since the magnetic field sensing value is sensitive to the rotation angle of the robot, grouping is performed only when the robot travels in a certain straight line section. The matching is performed only when a new node group is formed, and the matching score is calculated considering shifting between the node groups. The matching score can be obtained by the Euclidean distance between each vector, and the sequence of such matching is as shown in FIG.

는 기존에 구축된 i번째 노드 그룹 시퀀스를 의미하며,

는 스퀀스 간의 매칭 스코어를 의미한다.3, s _new denotes a generated node group sequence,

Denotes an i-th node group sequence established in the past,

Means a matching score between sequences.

The constraint condition generation unit 130 will be described as follows: 1) a method of generating a rotational constraint condition and 2) a method of generating a loop-closed constraint condition.

1) Method of creating rotational constraint condition

The rotational constraint can be generated by using the measurement characteristics of the magnetic field under specific conditions. If the mobile robot moves on a two-dimensional plane, and the mobile robot rotates without any translational displacement, the magnetic field vector appearing in the global coordinate frame does not change.

FIG. 2 illustrates such a situation. As shown in FIG. 2, the rotational constraint condition can be generated based on the magnetic field value measured while the mobile robot is rotating, and satisfies the following condition (1) The rotational constraint can be added as a constraint of the SLAM based on the graph structure.

&Quot; (1) "

와

는 병진 모션과 회전 모션 각각에 대한 문턱 값(threshold)을 의미하며,

와

는 i 프레임과 j 프레임 각각에서 이동 로봇의 방향각을 의미한다.Here, i, j denotes position indexes,

Wow

Denotes a threshold for each of the translational motion and the rotational motion,

Wow

Denotes the direction angle of the mobile robot in each of the i frame and the j frame.

는 아래 <수학식 2>와 같이 나타낼 수 있다.
Cost function for rotational constraint

Can be expressed by Equation (2) below.

&Quot; (2) &quot;

는 회전 구속 조건에 의해 구속된 위치 쌍의 집합을 의미하고,

는 회전 측정에 대한 정보 매트릭스를 의미하며,

는 아래 <수학식 3>과 같이 나타낼 수 있다.
here,

Denotes a set of positional pairs constrained by rotational constraint conditions,

Means an information matrix for the rotation measurement,

Can be expressed as Equation (3) below.

&Quot; (3) &quot;

는 회전에 대한 상대 위치 측정(relative pose measurement)을 의미하고,

는 회전에 대응하는 관찰 모델을 의미한다.here,

Refers to relative pose measurement for rotation,

Means an observation model corresponding to the rotation.

는 0으로 처리될 수 있고,

는 아래 <수학식 4>와 같이 정의될 수 있다.
Based on the relationship shown in Figure 1

Lt; RTI ID = 0.0 &gt; 0, &lt;

Can be defined as Equation (4) below.

&Quot; (4) &quot;

와

는 i번째 프레임과 j번째 프레임 각각에 기초한 자기장 측정을 묘사하는 벡터들을 의미한다.here,

Wow

Refers to vectors describing magnetic field measurements based on the i-th frame and the j-th frame, respectively.

는 아래 <수학식 5>와 같이 계산될 수 있다.
Since this observation model is a direction function of the mobile robot, heading correction can be performed from rotational constraint conditions. Information matrix

Can be calculated as Equation (5) below.

Equation (5)

는 자기장 측정에 대한 표준 편차를 의미한다.
here,

Means the standard deviation for the magnetic field measurement.

2) Loop closing constraint generation method

To assess the applicability of magnetic field measurements for loop closing, we check the similarity of the actual datasets and magnetic field measurements. In other words, the similarity matrix should be constructed to determine whether the magnetic field changes along the path. Similarity measures can be used as a basis for loop closure.

Cosine similarity and Euclidean distance can be used to calculate similarity based on a single measurement and a sequential group of measurements. The cosine similarity Scos between the two vectors p and q can be calculated by Equation (6) below, and the Euclidean distance D _E between the two vectors p and q can be calculated by Equation (7) below .

&Quot; (6) &quot;

&Quot; (7) &quot;

In the present invention, the ambiguity problem can be solved by using the sequential measurement, and the ambiguity problem can be solved through the sequential based loop closing algorithm.

The path of the mobile robot includes linear segments by the indoor environment. Each time the mobile robot moves a predetermined number of positions or nodes along a linear path, the magnetic field values of the three measured elements are converted into a single sequential (or sequence) vector Group. The grouped nodes are grouped into super nodes and the indexes for the nodes can be stored and managed.

Constraint generation unit 130 may detect loop closing using sequence-based matching whenever a new node group sequence is generated.

At this time, the constraint condition generation unit 130 can detect the loop closing by shifting the generated node group sequence in a state in which the predetermined node group sequence is fixed, and calculating the matching score. As for example, shown in Figure 3, when forming a new node group sequence (s _new), by being overlapped to have matching Nmin subnode while shifting to the right of point b, starting from the left point a, the similarity measure (D _E) Is calculated. Here, the matching may be indicated by partial overlap, and the connection line between the two node group sequences shown in FIG. 3 represents the constraint.

The Nmin subnodes in FIG. 3 may refer to the minimum nodes to prevent false positive loop closing during shifting. Since the matching may be performed in the reverse direction, the matching score for the reverse matching using the reverse sequence may be calculated. Match score may be evaluated by using Euclidean distance (D _E) is defined in Equation (7), while the one of node groups the node sequence of the two group sequence shifting.

At this time, it can be determined that the loop closing occurs when the Euclidean distance or similarity measure D _E is smaller than the preset threshold value T ^LC and the matching is in the reverse direction.

Through the above process, the constraint generator can generate the rotational constraints and the loop closing constraints used in the SLAM based on the graph structure.

The position estimation unit 140 estimates the position of the mobile robot through the SLAM based on the graph structure using the constraint conditions generated by the constraint condition generation unit 130. [

In this case, the position estimating unit 140 may estimate the position of the mobile robot through the SLAM based on the graph structure using the Odometry constraint and the rotational constraint, and may also calculate the position of the robot using the Odometry constraint and the loop closing constraint The position of the mobile robot can be estimated through the SLAM based on the structure, and the position of the mobile robot can be estimated through the graph structure based SLAM using the odomometry constraint, the rotational constraint condition, and the loop closing constraint. Of course, the constraint used for estimating the position of the mobile robot can be determined differently depending on the situation.

That is, the position estimating unit 140 corrects the estimated position value of the mobile robot by graph optimization on the generated graph structure, calculates the final displacement of the mobile robot through the SLAM based on the graph structure, The position can be estimated.

The graph structure may include a node indicating a position of the mobile robot and an edge indicating a constraint condition between the nodes.

As described above, the position recognition apparatus according to the embodiment of the present invention solves the ambiguity problem of the magnetic field in the indoor environment through the SLAM based on the graph structure using at least one constraint of the rotational constraint condition and the loop closing constraint, The estimation accuracy can be improved.

FIG. 4 illustrates an example of a process of optimizing a position graph. The solid line represents a trajectory by odometry information, the triangle represents a trajectory by a SLAM based on a graph structure, Denotes a super node.

As shown in FIG. 4, the rotation angles of the mobile robot are corrected (FIG. 4A) by using only rotational constraints without loop closing, and red and blue shown in FIG. 4 represent group nodes, And the loop is closed through FIGS. 4B to 4I, and the overall path is corrected.

FIG. 5 shows an example of the path estimation result according to the present invention.

5, the floor map is superimposed with the trajectory of the mobile robot. As shown in FIG. 5, the path of the mobile robot is largely deviated from the reference cyan path by the magenta, According to the invention, it can be seen that the path of the mobile robot by the graph structure base (blue) closely follows the reference path.

Hereinafter, embodiments of a position recognition method of a mobile robot will be described, and a method according to embodiments of the present invention may be performed in the position recognition apparatus shown in FIG.

FIG. 6 is a flowchart illustrating an operation of a mobile robot according to an embodiment of the present invention. Referring to FIG. 6, an operation for estimating the position of a mobile robot using a rotational constraint condition, a loop closing constraint condition, Fig.

Referring to FIG. 6, a method of recognizing a position of a mobile robot according to an embodiment of the present invention generates a rotational constraint condition based on a magnetic field value measured while the mobile robot rotates (S610).

At this time, the step S610 may generate the rotational restraint condition for the rotation angle of the robot using the relation between the magnetic field vectors measured before and after the rotation while the mobile robot is rotating in the stationary state.

To form a node group by grouping the nodes formed on the movement path in order to generate a loop closing constraint independently of generation of the rotational constraint, and to perform loop closure based on the magnetic field value measured for the nodes of the formed node group Constraint conditions are generated (S620, S630).

At this time, step S620 may form a node group only when the mobile robot travels in a straight line section. Step S630 compares the magnetic field values measured for the formed node group with the magnetic field values for the predefined node group , You can create a loop-closing constraint.

At this time, in step S630, a loop closing constraint can be generated by performing matching with a predefined node group only when a new node group is formed, and it is possible to generate a loop closing constraint on the vector for the formed node group and the predefined node group The Euclidean distance between the vectors can be calculated to calculate the matching score, and the calculated matching score can be used to generate the loop closing constraint.

Step S630 may create a loop closing constraint by shifting the generated node group sequence with the predetermined node group sequence fixed to calculate the matching score.

In order to generate the odometry constraint condition, the odometry information of the mobile robot is acquired, and the odometry constraint condition is generated based on the obtained odometry information (S640, S650).

The position of the mobile robot is estimated through the SLAM based on the graph structure using the rotational constraint condition generated in step S610, the loop closing constraint condition generated in step S630, and the omometry constraint condition generated in step S650 (step S660) .

FIG. 7 is a flowchart illustrating a method of recognizing a position of a mobile robot according to another embodiment of the present invention, and is a flowchart illustrating an operation of estimating a position of a mobile robot using rotational constraint conditions and odometry constraint conditions .

Referring to FIG. 7, a method for recognizing a position of a mobile robot according to another embodiment of the present invention generates a rotational constraint condition based on a magnetic field value measured while the mobile robot rotates (S710).

At this time, in step S710, the rotational constraint condition for the rotation angle of the robot can be generated using the relationship between the magnetic field vectors measured before and after the rotation while the mobile robot is rotating in the stationary state.

In order to generate the odometry constraint condition, the odometry information of the mobile robot is obtained, and the odometry constraint condition is generated based on the obtained odometry information (S720, S730).

In operation S740, the position of the mobile robot is estimated through the SLAM based on the graph structure using the rotational constraint conditions generated in step S710 and the odometry constraint conditions generated in step S730.

FIG. 8 is a flow chart illustrating a method of recognizing a position of a mobile robot according to another embodiment of the present invention, and is an operation flowchart for estimating a position of a mobile robot using a loop closing constraint and an odometry constraint .

Referring to FIG. 8, a method of recognizing a position of a mobile robot according to another embodiment of the present invention includes: forming a node group by grouping nodes formed on the movement path of the mobile robot; And generates a loop closing constraint based on the magnetic field value (S810, S820).

At this time, step S810 may form a node group only when the mobile robot is traveling in a straight line section. Step S820 compares the magnetic field values measured for the formed node group with the magnetic field values for the predefined node group , You can create a loop-closing constraint.

At this time, in step S820, a loop closing constraint can be generated by performing matching with a predefined node group only when a new node group is formed, and a loop closing constraint can be generated. The Euclidean distance between the vectors can be calculated to calculate the matching score, and the calculated matching score can be used to generate the loop closing constraint.

Step S820 may create a loop closing constraint by shifting the generated node group sequence with the predetermined node group sequence fixed and calculating the matching score.

In order to generate the odometry constraint condition, the odometry information of the mobile robot is acquired, and the odometry constraint condition is generated based on the acquired odometry information (S830, S840).

The position of the mobile robot is estimated through the SLAM based on the graph structure using the loop closing constraint created in step S820 and the OLT matrix constraint created in step S840 (S850).

The system or apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the systems, devices, and components described in the embodiments may be implemented in various forms such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array ), A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to embodiments may be implemented in the form of a program instruction that may be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI &gt; or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (16)

Generating a loop closing constraint based on the measured magnetic field values for the nodes formed on the moving path of the mobile robot; And
Estimating a position of the mobile robot based on a graph structure using the loop closing constraint
And the position of the mobile robot is detected.
The method according to claim 1,
Generating a rotational constraint condition based on a magnetic field value measured while the mobile robot is rotating
Further comprising:
The step of estimating the position of the mobile robot
And estimating a position of the mobile robot based on a graph structure using the loop closing constraint and the rotational constraint.
The method according to claim 1,
The step of generating the loop closing constraint comprises:
The mobile node is grouped to form a node group, and magnetic field values measured for the formed node group are compared with magnetic field values for a predefined node group, Way.
The method of claim 3,
The step of generating the loop closing constraint comprises:
Wherein the mobile robot is configured to form the node group only when the mobile robot travels a straight line section and to perform matching with the predefined node group only when a new node group is formed, Location recognition method.
The method of claim 3,
The step of generating the loop closing constraint comprises:
Calculating a matching score by obtaining a Euclidean distance between the vector of the formed node group and the vector of the predefined node group, and calculating a matching score based on the position of the mobile robot which generates the loop closing constraint using the calculated matching score Recognition method.
The method according to claim 1,
Acquiring odometry information of the mobile robot, and generating an odometry constraint condition based on the obtained odometry information
Further comprising:
The step of estimating the position of the mobile robot
And estimating a position of the mobile robot based on the graph structure using the loop closing constraint and the omometry constraint.
The method according to claim 1,
The step of estimating the position of the mobile robot
A position recognition method for estimating a position of the mobile robot using SLAM (Simultaneous Localization and Mapping) based on a graph structure.
Generating a rotation constraint based on a magnetic field value measured while the mobile robot rotates; And
Estimating a position of the mobile robot based on a graph structure using the rotational constraint condition
And the position of the mobile robot is detected.
9. The method of claim 8,
The step of generating the rotational constraint condition
And a rotation constraint condition for the rotation angle of the mobile robot is generated using a relationship between magnetic field vectors measured before and after rotation of the mobile robot.
9. The method of claim 8,
Acquiring odometry information of the mobile robot, and generating an odometry constraint condition based on the obtained odometry information
Further comprising:
The step of estimating the position of the mobile robot
And estimating a position of the mobile robot based on a graph structure using the rotational restraint condition and the odometry constraint condition.
A constraint generation unit that generates a loop closing constraint based on the measured magnetic field values for the nodes formed on the movement path of the mobile robot; And
A position estimating unit for estimating a position of the mobile robot based on a graph structure using the loop closing constraints,
And the position information of the mobile robot.
12. The method of claim 11,
The constraint condition generation unit
Generating a rotation constraint condition based on a magnetic field value measured while the mobile robot rotates,
The position estimating unit
And estimating a position of the mobile robot based on a graph structure using the loop closing constraint and the rotational constraint.
12. The method of claim 11,
The constraint condition generation unit
The mobile node is grouped to form a node group, and magnetic field values measured for the formed node group are compared with magnetic field values for a predefined node group, Device.
14. The method of claim 13,
The constraint condition generation unit
Wherein the mobile robot is configured to form the node group only when the mobile robot travels a straight line section and to perform matching with the predefined node group only when a new node group is formed, Position recognition device.
14. The method of claim 13,
The constraint condition generation unit
Calculating a matching score by obtaining a Euclidean distance between the vector of the formed node group and the vector of the predefined node group, and calculating a matching score based on the position of the mobile robot which generates the loop closing constraint using the calculated matching score Recognition device.
12. The method of claim 11,
An encoder for obtaining the odometry information of the mobile robot
Further comprising:
The constraint condition generation unit
Generating an odometry constraint condition based on the acquired odometry information,
The position estimating unit
And estimates the position of the mobile robot based on the graph structure using the loop closing constraint and the odometry constraint.

Images