KR100862190B1 - Apparatus and Method for detecting pose of robot - Google Patents

Abstract

The present invention relates to a robot posture detection device and a method for detecting a robot posture using values detected through a motor and a contact sensor. The apparatus for detecting a robot posture of the present invention includes a plurality of motors attached to a robot to detect movement of the robot, a plurality of contact sensors attached to the robot to distinguish a contact between a user and an object with the robot, and the motor; It includes a robot control unit for detecting the posture of the robot based on the information detected through the contact sensor, whereby the posture of the robot that can detect the posture of the robot in a variety of ways and interacts while contacting a person It can be grasped.

Robot, posture, detection, motor, sensor, contact, criteria, learning, person, user, object

Description

Apparatus and Method for detecting pose of robot

1 is a view showing an example of attachment of a motor and a sensor to a robot according to a preferred embodiment of the present invention,

2 is a view showing an example of the form attached to the robot control unit in the robot according to an embodiment of the present invention,

3 is a diagram illustrating an example of a network connection configuration of a robot control unit provided in a robot 100, a motor assembly including a motor unit, and a sensor unit according to an embodiment of the present invention;

4 and 5 illustrate an example of contact between a robot and a person according to an embodiment of the present invention;

FIG. 5 is a diagram illustrating another example for detecting a robot's interaction with a person and a robot's posture in a situation according to an embodiment of the present invention; FIG.

6 is a view for explaining the concept of the return to the reference posture from the detection posture and exceptional posture of the robot according to an embodiment of the present invention,

7 is a flowchart illustrating a method of detecting a posture of a robot according to a preferred embodiment of the present invention;

8 to 11 are views for explaining a configuration example for the posture detection of the robot according to the learning posture detection method according to an embodiment of the present invention, and

12 is a diagram showing an example of the configuration of the posture detecting means and the posture correcting means of the robot according to the program posture detection method according to an embodiment of the present invention.

The present invention relates to a robot capable of modifying a posture, and more particularly, to an apparatus and method for detecting a posture of a robot for detecting a corresponding posture according to a movement of a robot capable of modifying a posture.

In general, robots are often used in jobs that are difficult for humans to perform, for example, to transport heavy objects, to work that requires a high degree of precision, or in industrial sites where the work must be performed in high or low temperature working environments. .

Recently, non-industrial robots such as home robots, pet robots, lifesaving robots, and artificial human robots have emerged from the field of industrial robots performing industrial tasks specified in the industrial field. Among them, researches on interactive robots that can live with people in a residential environment and communicate with users or help users work are being actively conducted.

Compared to industrial robots, artificial robots are developed to have similar shapes and freedom of movement as humans. In particular, the artificial robot takes the form of a head, a torso, and arms and legs similar to a human like a human.

Since these artificial robots must move by themselves, many technologies are required, and in particular, a more accurate robot posture detection technique for controlling movement is required.

Hereinafter, known patents related to attitude control of a robot will be introduced.

Korean Patent Laid-Open No. 2004-0108526 (2004.12.24) "Motion control device and motion control method and robot device for an angle mobile robot" is a motion control device and motion control method and robot device for each mobile robot having a plurality of joint degrees of freedom. Describes the content of This patent invention provides a motion control device and control method for an angle mobile robot having an upright posture, and a robot device using the same. The acceleration movement, a posture sensor, and the like are attached to an upright walk using a (Zero Movement Point: ZMP) algorithm. It is to determine posture stability and use it for posture stabilization.

Korean Patent Laid-Open Publication No. 2005-0002302 (2005.01.07) "A real-time attitude control apparatus and method of biped walking robot" aims to control real-time posture by measuring the front and rear tilt and left and right tilt of biped walking robot. Disclosed is a method for real-time control of the posture of a biped walking robot using a single-axis or an inclined tilt sensor that detects the tilt of the walking robot in front of and behind the left and right tilts.

The two prior patents mainly target biped walking robots, and provide a method of using a acceleration sensor, a posture sensor, a tilt sensor, and the like to take a stable posture through mathematical calculations for posture control.

Korean Patent Laid-Open Publication No. 2001-0022664 (2001.03.26) "Robot device and its attitude control method" is an output of an acceleration sensor that aims to return to a normal state by detecting that the robot has a different posture than a normal state. According to the acceleration information obtained by the present invention, it suggests a method of retrieving and returning that the attitude state of the robot is different from the normal state, and states that the state information of the contact sensor can also be used for attitude estimation.

However, the existing inventions as described above only provide the grasp of the posture in a state in which the robot moves independently, and do not mention the grasp of the posture of the robot while the robot is in contact with each other. For example, there is a need for a method of identifying a robot's posture in a situation in which a robot posture detection is important, such as a human holding a robot.

Korean Patent Publication No. 2001-0022664 mentions that a touch sensor can be used to determine posture, but it can only know where it has been touched, and grasp the robot posture in a situation where a human and a robot interact with each other. Has the problem that is impossible.

A first object of the present invention for solving the above problems is to provide a robot posture detection device and method that can detect the posture of the robot with respect to external contact more easily.

It is a second object of the present invention to provide a robot posture detection device and a method for allowing a robot such as a state in which a person is holding a robot to interact with each other more easily.

A third object of the present invention is to provide an apparatus and method for detecting a posture of a robot for easier interaction with an external contact with which the robot is in contact.

An apparatus for detecting a posture of a robot according to an embodiment of the present invention for achieving the above object includes: a plurality of motors attached to the robot to detect movement of the robot; A plurality of contact sensors attached to the robot to distinguish a contact between a user and an object with respect to the robot; And a robot controller configured to detect the attitude of the robot based on information detected by the motor and the contact sensor.

Preferably, the motor, the processing unit for controlling the operation of the motor; And a driver that provides a communication interface between the motor and the robot controller to detect a movement of the robot according to the operation of the motor and provide the robot to the robot controller.

The touch sensor may include a first contact sensor having a function of sensing heat and a sensing function of humidity; A second contact sensor having a heat sensing function; And a third touch sensor having a sensing function of whether or not the external touch is applied.

The first touch sensor detects the contact by classifying a case where the human body and the object are in contact, when the object is in contact, when the human body is in proximity but not in contact, and when the human body and the object are neither in proximity nor contact. Perform The second touch sensor performs the touch detection by classifying a case in which a human body is in contact, an object in contact, a case in which the human body is in proximity but not in contact with the human body, and a case in which the human body is not in proximity or in contact with the body. The third touch sensor distinguishes the case where the human body or the thing is in contact with the case where the human body or the thing is not in contact and performs the touch sensing.

The robot controller may detect the posture of the robot by using only the value of the motor, using only the value of the sensor, or using the value of the motor and the sensor together. Detect.

The robot control unit may include: a program posture detection method for detecting a posture of the robot by comparing a pre-programmed posture value with a currently detected value when detecting the posture of the robot; And a learning posture detection method for detecting a posture of the robot through posture learning of the motor and sensor values.

The robot control unit may include: a state information collecting unit collecting information detected from the motor and the contact sensor when using the program posture detection method; And a posture detection unit configured to detect postures of the robots corresponding to the collected information based on preset posture information of the robots corresponding to information detected from the motor and the contact sensor.

The robot controller may detect a posture of the robot by the posture detector when the posture of the robot detected in response to the collected information by the posture detector is not a reference posture of the preset robot. Correct the reference posture set in the similar posture group of the preset posture information of the robot included.

The robot controller may include: a storage unit configured to receive and store a learning result model of the robot obtained through learning from a posture detection learning device when using the learning pose detection method; A state information collecting unit collecting information detected from the motor and the contact sensor; And a posture detection unit that detects a posture of the robot corresponding to the collected information based on the stored learning result model.

The robot control unit, when the posture of the robot detected by the posture detection unit is not a reference posture, the posture similar group of the posture information of the preset robot including the posture of the robot detected by the posture detection unit And a posture correcting unit configured to correct the reference posture set therein.

The learning result model includes: the robot posture at the time of inputting state values of the motor; The robot posture at the time of inputting both the state values of the motor and the values of the contact sensor; And the robot attitude model at that time with the values of the touch sensor as inputs.

The state value of the motor includes a state value of the motor itself, a value of symbolizing a state value of the motor, and a value of symbolizing a group unit state value of the motor.

The posture correcting unit, when the postures of the robot detected by the posture detecting unit, are inconsistent with each other in comparison with the reference posture, change the state value of the motor to move to the reference posture with respect to the robot.

The posture correcting unit requests the user to return the robot to the reference posture through a user interface when the posture of the robot does not become the reference posture by changing the robot posture.

According to an aspect of the present invention, there is provided a method for detecting a posture of a robot, the method including: detecting a movement of the robot based on a state value of the motor through a plurality of motors attached to the robot; Collecting contact information of a user and an object with respect to the robot detected through a plurality of contact sensors attached to the robot; And detecting a posture of the robot based on the detected motion information of the robot and the contact information.

In accordance with another aspect of the present invention, there is provided a posture detecting method of a robot, the method comprising: determining whether the detected posture of the robot is an exception posture; Calculating a reference posture of the similar posture group including the detected posture of the robot when the posture of the detected robot is an exceptional posture; And changing a state value of the motor according to the detected attitude of the robot to a value corresponding to the reference posture.

According to an embodiment of the present invention, a method of detecting a robot includes: determining whether a state of a motor corresponding to the reference posture and the detected robot's posture coincides if the detected robot's posture is not an exception posture; Changing the state value of the motor according to the detected attitude of the robot to a value corresponding to the reference attitude, if the state value of the motor according to the attitude of the detected robot does not match. .

According to the present invention, by detecting the position of the robot through a motor for detecting the movement of the robot and a contact sensor that can distinguish the contact between the human contact and the object, it is possible to detect the posture of the robot in a variety of ways It is also possible to grasp the attitude of the interacting robot while making contact.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that the same elements in the figures are represented by the same numerals wherever possible. In addition, detailed descriptions of well-known functions and configurations that may unnecessarily obscure the subject matter of the present invention will be omitted.

The present invention proposes a method for determining the pose of a robot using a motor and a sensor.

Accordingly, the present invention provides a contact sensor capable of detecting heat and humidity (which is emitted from a human body) and a method of determining a posture of the robot using the same, whereby a part of the robot is simply in contact with an object. Make sure you can tell if something is wrong. Accordingly, unlike the conventional method, the posture of the robot, which may occur during interaction with the robot, may be detected.

In addition, the present invention provides a method of detecting a posture by learning motor / sensor data as well as a method of simply programming a posture to be detected in a memory. Through this, even the posture that is difficult to be determined as a specific posture can be detected through the most similar posture through learning. When using this method, a posture which should generally be detected is programmed in memory in advance, and when a posture detection is performed, it is possible to detect a posture that is difficult to detect or has an ambiguity.

In addition, the attitude of the biped walking robot was detected through mathematical calculations using a motor value, an acceleration sensor, an inclination sensor, and the like, but in the present invention, the robot and the person are controlled by a motor or a (contact) sensor or a mixture of the two. Even when this contact is made, the robot's posture can be detected.

In the present invention, the robot is not limited to a special form such as 2 or 4 groups, but for convenience of description, two forefoot (left and right), two forefoot (left and right), and a head (two eyelids controlled by a motor) , Four ears) and a torso robot with torso. In addition, the motor and the motor is connected to the frame, and each part should also be connected, but this is a part of the attitude detection method to be achieved in the present invention can be dealt with separately, so the motor and motor connection, the connection of each part of the robot, etc. It does not, but it is assumed that they are connected to enable operation.

In the present invention, the motor and sensor attached to the robot device, the means for detecting the position of the robot through the information provided through the motor and the sensor and a mixture thereof, and the most stable stable posture predefined from the detected position of the robot And a robot control unit for performing a means for correcting the posture of the robot apparatus, a means for detecting the posture of the robot, and a means for correcting the posture of the robot.

1 is a diagram illustrating an example of attachment of a motor and a sensor to a robot according to a preferred embodiment of the present invention. In Figure 1 (a) shows the front of the robot 100, Figure 1 (b) shows the rear of the robot 100.

As shown, the motor units 151 to 155 are drivers for performing a communication function between the motor itself performing physical movement, the control of the motor, and the robot control unit (200 of FIG. 2) controlling the overall operation of the robot. Consists of The motors 151 to 155 may provide their own position value information such as how much the robot 100 is moved from the reference position.

Motor assembly unit 110, 120, 130, 140 is a simplified diagram of a plurality of configurations, such as the motor unit (151 ~ 155) is connected. Therefore, the motor assembly parts 110, 120, 130, and 140 may be regarded as connected to two or more frames having the same configuration as the motor parts 151 to 155.

According to an embodiment of the present invention, the sensor used in the sensor units 160 to 176 provided in the robot 100 has any one of the following characteristics.

First, the sensor used for the sensor units 160 to 176 is a touch sensor (hereinafter, referred to as touch sensor 1) having a heat sensing function and a humidity sensing function. The touch sensor 1 may detect heat and humidity emitted from a human body.

Second, the sensor used for the sensor unit 160 to 176 is a touch sensor (hereinafter referred to as touch sensor 2) having a heat sensing function. The touch sensor 2 may detect heat emitted from a human body.

Third, the sensor used for the sensor units 160 to 176 is simply a touch sensor (hereinafter referred to as touch sensor 3) that can know that something has been touched.

Sensors used in the sensor units 160 to 176 having any one of the three characteristics, such as the robot with a driver that provides a function that can exchange the detected value with the robot control unit 200, etc. 100).

The touch sensor 1 of the sensor unit 160 to 176 detects a function such as when the human body is in contact, when the object is in contact, when the human body is in proximity but not in contact, and when the human body and the object are not in proximity or in contact with each other. Can be distinguished.

Like the touch sensor 1, the touch sensor 2 of the sensor unit 160 to 176 is a case in which the human body is in contact, when the object is in contact, when the human body is in proximity but not in contact, and when the human body is neither in proximity nor in contact, that is, In four cases, detection functions can be distinguished.

Finally, the contact sensor 3 of the sensor unit 160 to 176 may distinguish a sensing function into a case in which the human body or an object is in contact with the person or the object is not in contact.

In the present invention provides a sensing function through the sensor having the characteristics of the contact sensor 1 and the contact sensor 2 of the sensor unit (160 ~ 176).

Existing touch sensor (contact sensor 3) can simply know that something has been touched, but the contact sensor 1 and the contact sensor 2 can distinguish whether a person touches or something touched, the robot to achieve in the present invention ( 100) can be used in the posture detection method.

Even when the touch sensor 1 and the touch sensor 2 of the sensor units 160 to 176 are used, the sensed sensor output value of the part hidden by one step by a person such as clothes exposed to the outside of the body may be different. However, even in such a case, these can be distinguished by experimental values, and thus the present invention will not be described in detail.

As illustrated in FIG. 1, the motor units 151 to 155 may be configured in any number to configure the robot 100 device.

For example, the motor unit 152 is for controlling the right eyelid, and the motor unit 153 is for controlling the left eyelid. In addition, the motor unit 151 and the motor unit 154 are for controlling the right ear and the left ear, respectively.

As described above, the present invention may be used to attach various parts for grasping the position of the motor unit attached to all positions such as the eyelid or the ear as well as the sensor unit as well as the legs of the robot 100. For example, a state in which the left ear is bent forward, a state in which the right eyelid is closed, and the like are also objects for the posture detection of the robot 100 in the present invention.

FIG. 1B illustrates an example in which sensor units 172 to 176 are attached to the back of the robot 100. The attachment of the sensor parts 172 to 176 may be attached at any position of the body of the robot 100, similarly to the motor parts 151 to 155, and is generally attached to a part contacting the bottom surface.

However, in the present invention, the sensor parts 172 to 176 are attached not only to the bottom surface but also to portions easily contacting the person. For example, since the sensor parts 172 to 176 are parts in which the back of the head of the robot 100, the ear, the forehead, the waist, the side, and the hip are frequently in contact with a person, the sensor parts 172 to 176 are also included in these parts. ) Can be attached.

2 is a view showing an example of the form attached to the robot control unit 200 in the robot 100 according to an embodiment of the present invention.

As shown, the robot control unit 200 is attached to the inside of the robot 100, Figure 2 shows an example in which the robot control unit 200 is attached to the abdomen of the robot 100, the robot 100 It can also be attached to other parts of).

In the present invention, the part for supplying power to the robot 100 is not shown separately, but since the power may be supplied from a battery or the like, it will not be described separately.

3 is a robot control unit 200 provided in the robot 100 according to an embodiment of the present invention, a motor assembly unit 110 to 140 including a motor unit 151 to 155, and a sensor unit 160 to 176. Is a diagram illustrating an example of a network connection configuration.

As shown, the motor units 151 to 155, 111 and 112 and the sensor units 160 to 176 are connected to the CPU (Central Processing Unit) 210 and the memory 220 through the bus 230 in the robot control unit 200. The structure is connected. Since the motor assembly parts 110 to 140 briefly display the set in which the motor parts 151 to 155, 111 and 112 are gathered, the motor assembly parts 110 to 140 are connected through the bus 230 in the same manner as the individual motor parts 151 to 155, 111 and 112 are connected.

4 and 5 are diagrams showing examples of contact between the robot 100 and a person according to an embodiment of the present invention.

4 is a view for explaining an example of the interaction between the robot 100 and the person 500 according to an exemplary embodiment of the present invention, and the attitude detection of the robot 100 in the situation.

As shown in FIG. 4, the robot 100 has two rear legs (specifically, 160 and 162 contact sensors) contacting the abdomen B2 and B3 of the person 500, and the right hand B1 of the person 500. ) Is a diagram illustrating a form in which the robot 100 contacts the hip (specifically, the contact sensor of 161).

In the case of using a sensor (contact sensor 3) which simply determines whether or not the existing contact, the robot 100 is holding the front head with the hips on the floor and the two rear feet in contact with the floor. In FIG. 4, the shape of the robot 100 is not technically distinguishable.

Therefore, in the present invention, the above two cases can be distinguished by using a touch sensor having the same characteristics as the touch sensor 1 or the touch sensor 2.

In the case of the touch sensor 1 capable of detecting heat and humidity, information about heat and humidity emitted from the body of the person 500 is stored in advance, and when the person 500 comes into contact with the contact sensor 1, it is determined.

In the case of the touch sensor 2 capable of detecting heat, since it is a touch sensor capable of detecting heat emitted from the body of the person 500, the person 500 may be distinguished from simply touching the floor or an object. Can be.

As described above, in the present invention, by using the touch sensor 1 capable of sensing heat and humidity and the touch sensor 2 capable of sensing heat, as well as the posture when the robot 100 operates independently, the person 500 and In the case of interaction, the attitude of the robot can be determined.

FIG. 5 is a diagram illustrating another example for detecting a robot's interaction with a person and a robot's posture in a situation according to an embodiment of the present invention.

As shown in the figure, the robot 100 has the two rear feet touching the floor with the hips attached to the floor, and the person 500 holds the two front feet of the robot 100 by hand.

Here, if it is assumed that the robot 100 moves independently without interacting with the person 500, it is impossible to determine the posture of the robot 100. However, according to the present invention, since the robot 100 can be distinguished from the part where the robot 100 is in contact with the person 500 by simply touching the bottom surface, this state is the state where the robot 100 puts the rear part of the body on the floor. As can be seen that the person 500 and the two forefoot are touching.

According to the present invention through FIG. 5, it can be seen that not only the robot 100 operates independently but also the posture when the robot 100 is interacting with the person 500. .

According to the present invention, the posture detection using the motor and the sensor of the robot 100 may be performed using only the value of the motor, only the value of the sensor, or both the value of the motor and the sensor. .

6 is a view for explaining the concept of the return to the reference posture from the detection posture and exceptional posture of the robot 100 according to an embodiment of the present invention.

As shown, the similar posture groups 610 and 620 of the robot 100 do not completely match the values of the motor or sensor that are set corresponding to the posture, but the values are within the same range with the same posture. A posture group that can be handled.

That is, even when the robot 100 sits, the motor value or the sensor value may be slightly different, so that various similar posture groups may be defined by the user. In the present invention, the determined posture of the robot is referred to as detection postures 612 and 622. In addition, the posture detected in the present invention is a very special posture, that is, an unstable state such as the robot 100 is turned upside down is called an exception posture 613, 623.

The posture correcting means according to the present invention refers to a correction so that the detection postures 612 and 622 and the exception postures 613 and 623 become the reference postures 611 and 621. Here, the reference postures 611 and 621 refer to a posture that becomes a standard for the similar posture groups 610 and 620, that is, the most stable sitting posture of the robot 100.

In the present invention, finding the reference postures 611, 621 closest to the detection postures 612 and 622 or the exception postures 613 and 623 is a method of finding the closest distance between two vectors using a vector of a motor value and a sensor value. It is available.

7 is a flowchart illustrating a method of detecting a posture of a robot according to an exemplary embodiment of the present invention.

As described above, the robot 100 detects the attitude of the robot 100 based on the information collected through the attached motors and the sensors (S110).

After detecting the posture, the robot 100 determines whether the detected posture information is an exceptional posture (S120).

As a result of the exceptional posture determination, if the detected attitude information is the exceptional attitude, the robot 100 detects the reference posture closest to the detected attitude information, which is the exceptional attitude, from the memory 220 (S130).

If it is determined that the detected posture information in step S120 is not an exceptional posture, the robot 100 compares the motor value of the reference posture with the motor value of the detected posture information (S150).

At this time, the robot 100 determines whether both the set motor value of the reference posture and the motor value of the detected posture information match (S160).

When the set motor value of the reference posture and the motor value of the detected posture information all match, the robot 100 determines that the posture detected in step S110 is the correct posture of the robot 100 and ends the posture detecting operation performed by the performing master ( S170).

On the other hand, after detecting the reference position closest to the posture information detected in the exceptional position in step S130, or if the motor value of the detected posture information and the set motor value of the reference posture in step S160 does not match any one, the robot 100 ) Changes the motor value stored in the memory 220 and performs the posture detection step S110 again.

At this time, such a step may be repeated. For example, the robot 100 attempts to return to the nearest reference posture while the robot 100 is upside down. The user may request that the user return to the standard posture through the user interface such as the speaker, LED, and LCD. In this case, since the user may return to a stable posture but may not be a complete reference posture, the user may complete the return to the reference posture by repeating the above steps. In general, since the robot 100 has an external output device such as a speaker, an LCD, and an LED, in the present invention, a case in which the robot 100 has a speaker, an LCD, and an LED is not illustrated separately.

According to the embodiment of the present invention, the motor or sensor values may be stored in advance in the memory 220, and the attitude may be judged when the motion or movement of the robot 100 is similar to or coincide with this value. (Hereinafter referred to as program posture detection method). In addition, according to another embodiment of the present invention, the posture may be determined through learning without storing a value corresponding to the posture in advance (hereinafter, referred to as a learning posture detection method).

12 is a configuration example of a posture detecting means and a posture correcting means of a robot according to a program posture detecting method according to an embodiment of the present invention.

In FIG. 12, the state collection program 221 through the sensor information in the memory 220 of the robot control unit 200 reads a motor or sensor value attached to the robot 100 and outputs it to the posture detection program 224. .

The posture detection program 224 has a combination of motor and sensor values and a posture of the combination in a memory beforehand, and determines a posture by comparing the motor and sensor values input from the state collection program 221.

In this case, when the posture is not the reference posture, the robot control unit 200 of the robot 100 attempts to return to the reference posture through the posture correcting program 222. The posture correction means will be described later in detail.

In this embodiment, the posture information of the robot 100 stored in advance in the memory 220 may be (motor value vector, posture), (sensor value vector, posture), (motor value vector, sensor value vector, posture), or the like. Can be. That is, in this embodiment, the posture using only the motor value attached to the robot 100, the posture using only the sensor value, and the posture using both the motor value and the sensor value can be freely configured.

For example, when the state of FIG. 4 is programmed using only the sensor value, (sensor value vector, posture) is represented by (160 state value, 161 state value, 162 state value,…, 175 posture value, posture value). Can be. When using the contact sensor 1 or the contact sensor 2, (sensor value vector, attitude) can be expressed as (HUMAN_TOUCH, HUMAN_TOUCH, HUMAN_TOUCH, NO_TOUCH, ..., NO_TOUCH, POSE1).

In addition, in the case of a motor value or a sensor value used in this case, the values may be symbolized as in the above example. In addition, in the case of a motor, it is also possible to symbolize the value in groups. For example, the position of each foot may be divided into ten, and one of the ten may be treated as a symbolized value. have.

In the case of a motor, it can basically use the same value as the angle of movement. For example, a test may be performed that a detected value, such as (angle1 <value <angle2), is between two values, angle1 and angle2. As described above, in the case of a motor or a sensor value, the detected value itself may be used, the value may be symbolized, or the value may be symbolized in a group unit.

8 to 11 are diagrams for explaining a configuration example for the posture detection of the robot according to the learning posture detection method according to an embodiment of the present invention.

8 is a conceptual diagram illustrating a robot posture detection method using learning according to the present invention.

As shown, in the learning posture detection method, when performing posture learning, the robot control unit 200 and the posture detection learning computer 300 should be connected to the interface cable 310.

At this time, the motor and sensor values are collected in the posture detection learning program 400 of the posture detection learning computer 300 shown in FIGS. 8 and 9 by collecting the motor and sensor values in the state collection program 221 shown in FIG. 11. Convey them.

9 is a view showing in more detail the robot posture detection means using the learning according to the invention of FIG.

As shown, the posture detection learning program 400 of the posture detection learning computer 300 to communicate with the robot control unit 200 of the robot 100, the data collection unit 410, the learning data 420 , A learning algorithm 430, and a learning result model 440.

The data collecting unit 410 stores the data transmitted from the robot control unit 200 as the learning data 420, and the learning algorithm 430 uses the learning data 420 to perform posture learning and the result. Save as learning result model 440.

In the process of learning, a user (person, 500) is involved, which is the step of collecting data. Rather than simply reading and storing motor and sensor values in the data collection phase, the user must enter the robot posture for these values. This step is usually an intervention in the machine learning phase, so this step is not described in detail.

10 is a diagram illustrating an example of a configuration of training data for detecting a robot posture according to an exemplary embodiment of the present invention.

As shown in the figure, the motor state values 421, 422, and 423 and the sensor state values 424 and 425 are used as data for learning, and as a result, the pose Pose 426 of the robot 100 is obtained. It is.

The present invention does not limit the learning data configuration simply as shown in FIG. The learning data may use the value of the motor or sensor itself, symbolize the values, or symbolize the values in groups. This has been described above.

11 is a view showing an example of the configuration of the posture detection means and posture correction means using the learning result model according to an embodiment of the present invention.

As shown, when the learning result model 440 of FIG. 9 is made through learning in FIG. 9, a copy 441 of this is copied to the memory 220 in the robot control unit 200 and used for posture detection.

At this time, when the state acquisition program 221 reads the value of the motor or sensor and transfers it to the posture detection program 223, the posture detection program 223 uses a copy 441 of the learning result model 440 to provide a robot ( 100 is detected. At this time, if the detected posture is not the reference posture, the posture correction program 222 returns to the reference posture of the robot 100. Since the posture correcting means and the posture correcting method are illustrated in FIGS. 6 and 7, redundant description thereof will be omitted.

In the above, specific preferred embodiments of the present invention have been illustrated and described. However, the present invention is not limited to the above-described embodiments, and any person having ordinary skill in the art to which the present invention pertains may make various modifications and other equivalents without departing from the gist of the present invention attached to the claims. Implementation will be possible. Therefore, the true technical protection scope of the present invention should be defined only by the appended claims.

According to the present invention described above, by detecting the position of the robot through a motor for detecting the movement of the robot and a contact sensor that can distinguish the contact between the human contact and the object, it is possible to detect the attitude of the robot in more various ways It is also possible to grasp the position of the interacting robot while in contact with a person.

In addition, according to the robot posture detection device and method of the present invention, the robot and the person mutual as well as the robot's posture in a state in which the robot operates independently through a contact sensor that can distinguish the contact between the person and the object. The attitude of the robot in the contact state can also be detected. This makes it possible to accurately understand the attitude of the robot even when the pet / animal robot is regarded as a doll and the children make a lot of contact with the robot, thus enabling more intelligent robot application development.

In addition, according to the present invention, if the posture of the robot is detected after the posture of the robot is unstable, it is possible to return to a stable reference posture. By allowing the robot to request help for returning to the standard posture through the output means attached to the robot, it is possible to prevent a malfunction, failure or accident of the robot that may be issued due to the unstable state of the robot. .

Claims (19)

A plurality of motors disposed in a robot and configured to correct movement of the robot and provide movement information; A plurality of touch sensors distributed in the robot and detecting contact states of a user or an object with the robot and providing contact information; And The posture of the robot is detected by comparing a pre-programmed posture value with the motion information and the contact information, or by analyzing the motion information and the contact information according to a posture learning model provided from an external device. A robot position detection apparatus comprising a robot control unit for detecting. The method of claim 1, The motor, A processing unit controlling an operation of the motor; And And a driver for providing a communication interface between the motor and the robot controller to detect a movement of the robot according to the operation of the motor and to provide the driver to the robot controller. The method of claim 1, The contact sensor, A first contact sensor having a function of sensing heat and a sensing function of humidity; A second contact sensor having a heat sensing function; And Apparatus for detecting the attitude of the robot, characterized in that any one of the third contact sensor having a sensing function for the external contact is applied. The method of claim 3, wherein The first contact sensor, When the human body is in contact, when the object is in contact, the human body is in close proximity but not in contact, and the human body and the object is not in close proximity or not in contact with the robot characterized in that for performing the touch detection Posture detection device. The method of claim 3, wherein The second contact sensor, When the human body is in contact, when the object is in contact, the human body is close but not in contact with the body, the human body is not close to the contact is not detected by detecting the posture of the robot, characterized in that Device. The method of claim 3, wherein The third contact sensor, And detecting the contact by classifying a case where a human body or an object is touched and a case where the human body or an object is not touched. The method of claim 1, The robot controller, when detecting the posture of the robot, And the position of the robot is detected by any one of using only the value of the motor, using only the value of the sensor, and using both the value of the motor and the sensor. The method of claim 1, The robot controller, when detecting the posture of the robot, A program posture detection method for detecting a posture of the robot by comparing the pre-programmed posture value with a value currently detected through the motor and the contact sensor; And Posture detection of the robot, characterized in that using any one of the learning posture detection method for detecting the posture of the robot by analyzing the current value detected by the motor and the contact sensor according to the posture learning model provided from the external device Device. The method of claim 8, When the robot control unit uses the program attitude detection method, A state information collecting unit collecting information detected from the motor and the contact sensor; And And a posture detection unit configured to detect the posture of the current robot by analyzing the information collected by the state information collection unit according to the preset posture information of the robot. The method of claim 9, The robot control unit, And a posture correcting unit configured to correct the posture of the robot to a reference posture most similar to the posture of the posture or the exceptional posture, when the posture detection unit detects the posture of the posture. The method of claim 8, When the robot control unit uses the learning posture detection method, A storage unit for receiving and storing a learning result model of the robot obtained through learning from a posture detection learning device; A state information collecting unit collecting information detected from the motor and the contact sensor; And And a posture detection unit configured to detect postures of the robots corresponding to the collected information based on the stored learning result models. The method of claim 11, The robot control unit, And a posture correcting unit configured to modify the posture of the robot to a reference posture most similar to the posture of the posture or the exceptional posture, when the posture of the posture detecting unit is an exceptional posture. The method of claim 12, The learning result model, The robot posture at the time of inputting state values of the motor; The robot posture at the time of inputting both the state values of the motor and the values of the contact sensor; And And the at least one model of the robot posture at that time as values of the touch sensor as inputs. The method of claim 13, The state value of the motor, And at least one of a state value of the motor itself, a value symbolized by the state value of the motor, and a value symbolized by the group unit state value of the motor. The method of claim 10 or 12, The posture correction unit, Posture detection device of the robot, characterized in that for modifying the attitude of the robot by controlling the operation of the motor through the state value of the motor. The method of claim 15, The posture correction unit, If the posture of the robot is not corrected to the reference position, the posture detection device of the robot, characterized in that for requesting the user to manually modify the posture of the robot through a user interface. Detecting movement information of the robot based on a state value of the motor through a plurality of motors distributed in the robot; Distributedly disposed on the robot to detect contact information of a user or an object with respect to the robot; And The posture of the robot is detected by comparing a pre-programmed posture value with the motion information and the contact information, or by analyzing the motion information and the contact information according to a posture learning model provided from an external device. Posture detection method of a robot comprising the step of detecting. The method of claim 17, Determining whether the detection posture of the robot is an exception posture; If the detection posture is the exceptional posture, calculating the reference posture of the similar posture group including the detection posture or the exceptional posture; And And changing the state value of the motor according to the detection posture to a value corresponding to the reference posture. The method of claim 18, If the detection posture is not the exceptional posture, determining whether the reference posture matches the state value of the motor according to the posture posture; And If the state value of the motor according to the reference posture and the detection posture do not match, changing the state value of the motor according to the detected posture of the robot to a value corresponding to the reference posture. Robot posture detection method.

Images