KR101989965B1 - Robot controll system - Google Patents

Abstract

The present invention relates to a robot control system having a robot behavior determining unit that detects an environment of a robot and selects a behavior of the robot according to the sensed environment.

Description

[0001] Robot control system [0002]

The present invention relates to a robot control system, and more particularly, to a robot control system having a robot behavior determining unit that detects an environment of the robot and selects a behavior of the robot according to the sensed environment.

The robot needs to be equipped with a system capable of recognizing the external environment, selecting actions based on the external environment, and executing the selected actions. In addition, in the case of a robot that operates autonomously, it is necessary to determine its own behavior according to internal and external environmental factors.

As a method for controlling the behavior of the existing robot, there has been a limitation in performing only an iterative operation according to an external command, or selecting an operation only on the basis of the environment of a specific element. Such an operation control system can not perform an operation considering various environmental factors.

Therefore, it is necessary to develop a robot control system that detects various environments and selects the behavior of the robot according to the sensed environment.

Published Patent 2006-0072597

An object of the present invention is to provide a robot behavior determination unit capable of detecting an environment of a robot such as a remaining battery level, light, motion, etc., converting the environment into a predetermined sensing value, and determining a behavior of the robot using the sensed amount, And to provide a robot control system in which the behavior of the robot can be determined according to the environment of the robot.

A robot control system according to an embodiment of the present invention includes: a sensor unit for sensing a current environment corresponding to at least one environment attribute; A memory for storing a behavior list of a robot including a plurality of behaviors, and a predetermined parameter value corresponding to each behavior included in the behavior list and corresponding to each environment attribute detected by each sensor unit, as data; A robot behavior determining unit for selecting any one of behaviors stored in the memory unit; And a robot behavior performing unit performing a behavior selected by the robot behavior determining unit. / RTI >

Wherein the sensor unit converts the environment of the sensed attribute into a predetermined value to derive a predetermined sensing value, and the robot behavior determination unit determines the robot behavior based on the parameter value stored in the memory unit and the sensing value A first algorithm for calculating a consumed energy value of each of the included actions, and a behavior selection algorithm having a second algorithm for comparing the energy consumption value of each behavior derived by the first algorithm and selecting a behavior, .

Preferably, the parameter value has a value corresponding to one of the actions included in the behavior list, and corresponding to one of the environment attributes detected by the respective sensor units, , The first algorithm substitutes the parameter value and the sensed amount value into Equation (1) to derive a consumed energy value of a current one environment attribute of a certain action,

The energy consumption value of the current work environment property of work = f (P, V) (Equation 1)

(P: parameter value, V: detection value value)

The energy consumption value under the current environment of the daily action is derived according to the following equation (2).

(식 2)

(Equation 2)

(B _j : energy consumption value under the current environment of action j,

P _ij : parameter value of environmental property i of behavior j,

V _i : detection value of environment attribute i)

Preferably, Formula 1 has the form of Formula 3 below.

(Equation 3)

(c ^h , c ^l , t ^h , t ^l : parameter value)

Preferably, the second algorithm compares the energy consumption values of each of the actions derived by the first algorithm and selects an action having the largest energy consumption value.

Preferably, the sensor unit includes at least one of a power amount sensor for detecting a remaining power amount of the robot, a sound sensor for detecting an external sound, a motion sensor for detecting an external motion, and an illuminance sensor for detecting an external illuminance do.

Preferably, the sensing amount value derived from the sensor unit is a value obtained by converting a measured value of the current environment into a percentage with respect to a predetermined reference value.

In the robot behavior control system according to the present invention, the robot selects a behavior among the behavior lists stored in the memory unit, performs the corresponding behavior, selects the behavior in consideration of the environmental factor, and performs the selected behavior.

That is, a parameter value is given to several environments and behaviors, a detected value is converted into a predetermined value, and then the parameter value and the detected value are substituted into an energy value derivation function, The energy consumption value of each action can be derived under an environment of a specific condition. Then, the energy consumption value of each behavior is derived, and by selecting a behavior according to a predetermined rule, the most appropriate behavior corresponding to the environment can be selected and performed.

1 is a block diagram of a robot control system according to the present invention.
2 is a diagram showing a behavior decision structure of a robot control system according to the present invention.
FIG. 3 is a diagram showing parameters according to respective behavior and environment attributes built in the memory unit of the robot control system according to the present invention.
FIG. 4 is a graph illustrating energy consumption values when a predetermined sensing value and a parameter value are input to an energy value derivation function according to an embodiment of the robot control system according to the present invention.
5 is a graph showing the energy consumption values for the "remaining battery" environmental element of the "running" action derived according to the energy value derivation function according to the embodiment of the robot control system according to the present invention.
6 is a graph showing the energy consumption values for the "sound" environmental element of the "running" action derived according to the energy value derivation function according to an embodiment of the robot control system according to the present invention.
7 is a graph showing the energy consumption values for the "battery remaining amount" environmental element of the "sleep" action derived according to the energy value derivation function according to the embodiment of the robot control system according to the present invention.
FIG. 8 is a graph showing the energy consumption values for the "remaining battery" environmental element of the "sleep" action derived according to the energy value derivation function according to the embodiment of the robot control system according to the present invention.
9 is a view showing a behavior of a robot by the robot control system according to the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 shows a structure of a robot control system according to the present invention, and FIG. 2 shows a behavior decision structure of a robot control system according to the present invention.

A robot control system according to the present invention includes: a sensor unit (100) for sensing the inside and outside environment of a robot; A memory unit 200 in which a behavior list of the robot and parameter values corresponding to respective behavior and environment attributes are stored as data; A robot behavior determining unit (300) for selecting a behavior among a behavior list of the robot stored in the memory unit (200); And a robot behavior performing unit (400) for performing actions determined by the robot behavior determining unit (300); . ≪ / RTI >

The sensor unit 100 senses the internal and external environments of the robot. That is, the sensor unit 100 can detect environments inside and outside the robot corresponding to various environmental properties such as light, sound, motion, and heat. In addition, the sensor unit 100 may derive a predetermined value for each environmental attribute detected. Here, the numerical value derived from the value of each detected environmental attribute is referred to as the detected value of the corresponding environmental attribute. However, the sensor unit 100 does not necessarily need to have a predetermined calculation unit for deriving the sensing value, and the derivation of the sensing value may be performed by an algorithm built in the memory unit 200 It is possible.

The internal environment of the robot may be an operating element in various robots such as a recent behavior performance rate, a target operation time achievement rate, a power consumption of motion, and the like.

In addition, the external environment of the robot can be the amount of power of the robot, sound outside the robot, light, movement, and the like. However, the internal environment and the external environment of the robot do not necessarily have to be distinguished. Any environment related to the operation environment, condition, and the like of the robot may be detected by the sensor unit 100, have. For example, when the remaining battery level, sound, light, and external motion are all allocated as one environmental attribute, the sensor unit 100 may be provided to sense the remaining battery level, sound, light, and external motion. Accordingly, a plurality of environment attributes can be detected, and the environment of the detected attribute can be reflected in determining the behavior of the robot by the robot behavior determination unit 300 described later.

For example, the sensor unit 100 may include a power amount sensor for sensing the remaining battery power of the robot, a motion sensor for detecting movement, a brightness sensor for detecting brightness, and a sound sensor for sensing sound.

The power amount sensor can sense the remaining amount of the battery of the robot and is not limited to being integrally formed with the sensor unit 100, and may be included in a predetermined power source unit. When the amount of power in the power source unit is full charge is 100, the current amount of power is converted into a percentage unit, and the current amount of power is converted into the amount of watched amount value Vchar, .

For example, the light intensity sensor may be a predetermined light intensity sensor that detects only the brightness of a predetermined light, or an RGB camera, but is not limited thereto. For example, the current illuminance detected by the illuminance sensor may be converted into a percentage unit as shown in Equation 1 below, and may be derived as the illuminance detection value Vlux. In the following Equation 1, the lower limit brightness and the upper limit brightness are preset values to derive the luminance detection value as a percentage value. For example, when the set lower limit brightness is 50 LUX, the upper limit brightness is 150 LUX, and the current brightness is 120 LUX, the illuminance detection amount value can be derived as 70.

Vlux = {(curlux-minlux) / (maxlux-minlux)} x 100 [

(minlux: low brightness, curlux: current brightness, maxlux: high brightness)

The derived illuminance detection amount value is transmitted to a robot behavior determination unit 300 to be described later and can be used in a behavior selection algorithm.

The motion sensor is a sensor for detecting an external motion, and may be constituted by an RGB camera as an example. An algorithm for detecting an external motion through such a motion sensor may be as follows. First, an incoming image frame is obtained from an RGB camera, and each frame is divided into blocks of a certain area. For example, if a camera image is input with a resolution of 720 × 480, the image can be divided into blocks having a size of 72 × 48. A total of 100 blocks can be obtained by dividing the blocks. Next, a motion vector of each block is extracted. Then, if motion vectors of a specific block are detected to be larger than a certain size, and neighboring blocks have similar sizes and directions, motion can be determined to exist.

In addition, the motion value indicating the degree of motion of the outside can also be derived as the motion detection amount value Vmos by being converted into a predetermined percentage unit as shown in Equation 2 below. Also, the derived motion detection amount value may be transmitted to the robot behavior determination unit 300 to be described later and used for the behavior selection algorithm.

Vmos = (the number of blocks in which motion is detected / the total number of blocks) x 100 (Equation 2)

The sound sensor is a sound sensor.

Sound is always generated in a normal living environment. Therefore, preferably, a method of determining that a sound reflected in the behavioral selection occurs only when a sound other than the everyday noise is detected is necessary. A detailed algorithm for this may be as follows. However, the method is not limited to the following method, and various methods can be used.

1) The sound signal which is continuously detected is cut and stored in a predetermined time unit (for example, 0.1 second). This is called a sound chunk.

2) Average 10 consecutive sound chunks to identify daily noise.

3) Afterwards, averaging 10 consecutive sound chunks to derive the difference from the sound situation.

4) If it is judged that the difference is more than the reference value, it is recognized that sound is generated.

5) Measure the decibel of the sound generated.

The value measured by the noise sensor can also be derived by converting the value by a percentage as shown in Equation 3 below. The lower decibel value and the upper decibel value in Equation 3 are also similar to the lower limit brightness and the upper limit brightness previously set in Equation 1, and are preset values to derive a sound detection amount value converted into a percentage. The derived sound detection amount value may also be transmitted to the robot behavior determination unit 300 to be described later and used for the behavior selection algorithm.

Vsnd = {(spkDb - minDb) / (maxDb - minDb)} 100 [

(minDb: lower decibel value, spkDb: current decibel value, maxDb: upper decibel value)

In the above, a method of deriving the detection amount value only for the remaining battery level, illumination, motion, and sound has been disclosed, but it is also possible to derive the detection amount value for other environmental factors through a predetermined formula and algorithm. That is, it does not exclude other environmental factors.

It is also possible that other environmental factors such as odor, heat, etc., may also be assigned as work environment attributes. In this case, the sensor unit 100 may further include a sensor suitable for detecting odors, heat, and the like, and the odor and heat may also be converted into the sensed value. It is needless to say that a predetermined upper limit value and a lower limit value are set, and the detection value of each environmental attribute can be derived as a percentage value.

The memory unit 200 may include a predetermined storage device. The memory unit 200 may store a behavior list of the robot and parameter values corresponding to respective behavior and environment attributes. Meanwhile, as described above, an algorithm for deriving the detection amount values using the environmental elements sensed by the sensor unit 100 may be included.

For example, the behavior list stored in the memory unit 200 may include a plurality of individual actions such as running, walking, lying down, sleeping, and the like. In addition, the memory unit 200 may include an operation signal for appropriately operating the actuators included in the robot to implement respective behaviors.

The parameter value is a predetermined value and may be a value preliminarily stored by the developer. In addition, the parameter values may be modified and re-entered as needed.

The robot behavior determination unit 300 is a device that selects a specific behavior from a behavior list of the robot stored in the memory unit 200. [ The robot behavior determination unit 300 may be, for example, a CPU having an algorithm for performing a predetermined operation. The specific configuration of the robot behavior determination unit 300 will be described later.

The robot behavior performing unit 400 performs actions determined by the robot behavior determining unit 300, and may be a predetermined actuator, a joint, or the like. The specific action of the robot action performing unit 400 may be performed by a predetermined action method embedded in the memory unit 200. [

Hereinafter, a specific configuration of the robot behavior determination unit 300 will be described.

The robot behavior determination unit 300 may include a first algorithm that calculates a consumed energy value of each behavior included in the behavior list using the parameter value stored in the memory unit 200 and the sensed amount value, And a behavioral selection algorithm having a second algorithm for comparing any one of the behaviors of the behavior of the behavioral selection algorithm.

First, the first algorithm stores the sensed value of each environmental attribute sensed by the sensor unit 100 and the parameter value of each behavioral and environmental attribute stored in the memory unit 200 in a predetermined expression form The energy value of each action under the current environment can be derived by substituting the derived energy value for each environmental attribute by substituting the energy value derivation function. That is, the energy value derivation function of the work environment environment attribute of one action can be represented as a function of a predetermined form as shown in Equation 4 below. Here, the specific form of the energy value derivation function is not limited, and an example of the form will be described later.

The energy value of the workplace environmental attribute = f (P, V) (Equation 4)

Here, P on the left is a value stored in the memory unit 200 as a parameter value. Preferably, each parameter value corresponding to the one-environment attribute of one action may be stored in the memory unit 200.

V is a sensing value of the corresponding environmental attribute sensed by the sensor unit 100. For example, when the battery residual amount detection value detected by the remaining battery charge sensor is 55, the remaining battery charge detection amount V corresponding to the battery remaining amount environment attribute is 55. [ In addition, when the sound detection value detected by the sound sensor is 88, the sound detection value corresponding to the sound environment property is 88. That is, the sensing value of the corresponding environmental attribute sensed by each sensor unit 100 is input to the V value. Here, since the sensing value of the attribute detected by the sensor unit 100 is independent of the behavior, it is possible to assign V having the same value for each behavior under the same environment.

The value derived by the function is a consumed energy value corresponding to a behavior of one of a plurality of behaviors, that is, a circumstantial property of a behavior. The total consumed energy value of a single action can be derived by deriving the above function value for all environmental attributes and then summing it. That is, Equation 5 is shown below.

(식 5)

(Equation 5)

(B _j : energy consumption value of action j,

P _ij : parameter value of environmental property i of behavior j,

V _i : detection value of environment attribute i)

Here, if the parameter value is P _ij Because the parameter values are different according to each work behavior and each work environment property. That is, the parameter value may be a set of predetermined values or a vector as shown below.

= lower threshold(

= lower threshold

= upper threshold

= upper threshold

= lower clamping value

lower clamping value

= upper clamping value)

= upper clamping value)

In addition, the detection value of the environmental attribute is also different for each environment attribute, and is expressed by V _i .

With this formula, it is possible to derive the energy consumption value of one action under the current overall environment. However, this term is consistently referred to herein as the "energy consumption value of a behavior", but the term does not necessarily refer to the actual energy (power) to perform a particular action, Quot; value ". < / RTI > That is, in other words, the energy consumption value of the action can be understood to have the same meaning as a kind of score. For example, if the energy consumption value of the "running" action is derived as 85 and the energy consumption value of the "sleeping & Is 85, and it is understood that the score of the "sleep" action is 90 points.

In addition, the energy consumption values of other behaviors can be derived through the same process as above.

The second algorithm may then select any one action using the energy consumption values of each action derived by the first algorithm. As an example, it is possible to compare the energy consumption value of each behavior and select the behavior with the maximum energy consumption value. For example, as in the example above, if the energy consumption value of the "running" behavior under the current environment is derived as 85, and the energy consumption value of the sleeping behavior is derived as 90, then the first algorithm under the current environment is "sleeping" And the second algorithm can select the "sleeping" action by comparing the derived scores. Of course, the present invention is not limited to this. It is also possible to derive the energy consumption value of each action, then to select an action having an energy consumption value approximate to the average value by averaging, or to select an action having the smallest energy consumption value It is possible. For this purpose, an algorithm for comparing the energy consumption value of each behavior and selecting a specific action may be embedded in the robot behavior determination unit 300.

Hereinafter, the process of deriving the energy consumption value of each action using the first algorithm will be described in more detail.

For example, a process of obtaining a consumed energy value of a specific first action (e.g., running) will be described.

First, a parameter value corresponding to the first behavior and corresponding to a first environmental attribute (e.g., battery remaining amount) of the environmental attribute (running_battery remaining energy parameter value) and an environmental attribute detected by the sensor unit 100 The energy consumption value corresponding to the first behavior and simultaneously the energy consumption value corresponding to the first environment attribute can be derived by substituting the detection amount value (battery remaining amount detection amount value) corresponding to the environmental attribute into the energy value derivation formula.

This can be expressed as Equation 6 below.

The energy consumption value related to the battery residual environmental property of the running action = f (running_ battery remaining capacity parameter value, battery remaining capacity value) (Equation 6)

Equation 6 means substituting a parameter value corresponding to "running" and "battery remaining amount " and a" remaining battery amount detection value "into a function defined by f (P, V). That is, P is the value of the corresponding behavior_required environment attribute parameter, and V is the detection amount value of the corresponding property detected by the sensor unit 100. [

Then, a parameter value (running_ sound detection parameter) corresponding to the first environmental action (running) and corresponding to the second environmental attribute (sound) among the environmental attributes, and a parameter value The energy value corresponding to the first behavior and corresponding to the second environmental attribute can be derived by substituting the detection amount value (sound sensing amount value) corresponding to the attribute into the energy value derivation formula. That is, Equation 7 is shown below.

The energy consumption value related to the environmental property of the running action = f (the running sound detection parameter value, the sound detection amount value) (Equation 7)

Here, the form of the function f (P, V) is the same, and therefore, the same function is applied to all environment attributes to derive the energy consumption value for the environment attribute. However, each parameter value is changed according to each environment property, and the detection amount value of the environmental property also changes.

By repeating the above procedure, it is possible to derive all of the energy consumption values corresponding to the first behavior and corresponding to the respective environmental properties. If all of the derived energy values are summed up, the energy consumption value of the first action (running) under the current total environment can be ultimately derived. This can be expressed as Equation 8 below.

(Running_battery parameter value, battery residual quantity value) + f (running_ sound detection parameter value, sound detection value value) + f (running_motion detection parameter value, motion detection amount) Value) + f (run_luminance detection parameter value, illuminance detection value value) + ... (formula 8)

Then repeat the above process for the second action (eg walking). That is, a parameter value (walking_battery remaining amount parameter value) corresponding to the second behavior and corresponding to the first environmental attribute (e.g., battery remaining amount) and an environmental attribute detected by the sensor unit 100 correspond to the first environmental attribute (P, V) to the energy value derivation function f (P, V) to derive a consumed energy value corresponding to the second behavior and corresponding to the first environmental attribute.

Here again, as described above, the function f (P, V) has the same form for all other behavior and environment attributes and is applied equally. However, each parameter value is changed according to the behavior and environmental property. Therefore, in the case of the energy consumption value of the second action, the second behavior and the parameter value of each environmental attribute are applied to the parameter value of the function f (P, V). It goes without saying that the sensing value of the environmental factor sensed according to the change of the environment can also be changed.

Next, the energy consumption value corresponding to the second behavior and corresponding to the second and third environmental properties is derived. The sum of the above values is then used to derive the energy consumption value of the second action under the current environment.

Then repeat the above steps for the third and fourth actions. Finally, under the current environment, the energy consumption value of all the actions stored in the memory unit 200 can be obtained.

In the second algorithm, the energy consumption value of each action derived through the above process is compared, and the action having the maximum value can be selected and performed. As described above, the present invention is not limited to this, and it is also possible to select an action having the lowest energy consumption value or a value having a consumed energy value approximate to the average value.

In this way, the sensing value of each environmental element is reflected in a function for obtaining the energy consumption value of each behavior, and the energy consumption value of each derived behavior is compared to determine which behavior, The structure to be performed by the performance unit may be as shown in Fig.

Hereinafter, one form of the energy value derivation function f (P, V) will be described. The energy value derivation function may be, for example, as shown in Equation 9 below. However, the following Equation 9 is only an example, and it is needless to say that the energy value derivation function can be set differently and the parameter value type can be set differently.

(Equation 9)

In the above equation, the input variable is the V value corresponding to the detection value of the environmental property detected by the sensor unit. c ^h , c ^l , t ^h , and t ^l are parameter values, and c ^h , c ^l , t ^h , and t ^l are values previously stored in the memory unit 200, respectively. Therefore, the graph of Equation 9 is shown in FIG. FIG. 4 is a graph illustrating energy consumption values when a predetermined sensing value and a parameter value are input to an energy value derivation function according to an embodiment of the robot control system according to the present invention. The V value, which is the sensing value corresponding to the work environment property, is located on the X axis. If a predetermined V value is input in the above equation 9, the consumed energy value of the corresponding behavior of the environment attribute is derived as a value on the Y axis according to the parameter value corresponding to the behavior of the environment attribute.

In the above equation 9, the relationship between the output value and the input value is shown in the form of a sine function. Therefore, the range of the output value is limited by two parameters c ^h and c ^l . In addition, the sensitivity to the input value can be adjusted by t ^h , t ^l . c ^h , c ^l , t ^h , and t ^l are stored in the memory unit 200 for each environment attribute of a single action. Therefore, when deriving the energy consumption value of the one environment property of the one action, the parameter value in the corresponding environment property of the behavior is reflected and utilized in the energy value derivation function.

For example, the table as shown in FIG. 3 may be stored in the memory unit 200. 3 shows the respective parameter values of the "battery remaining amount" and "sound" among the environmental attributes for the "running" action and the "sleeping" action. In FIG. 2, lower thres, upper thres, lower clamp, and upper clamp respectively denote t ^l , t ^h , c ^l , and c ^h .

In addition, as described above, V is the sensing value of the environmental property sensed by the sensor unit 100.

For example, action belongs to the "run", an environment attribute is when to obtain the energy consumption value corresponding to the "battery remaining amount", the above function c ^h, c ^l, t ^h, t ^l Each "run" and The parameter value belonging to "remaining battery level" is substituted, and the current value of the remaining battery level detected by the power amount sensor is substituted into V. [

For example, with respect to the "running" action according to the above equation, the remaining energy level of the battery and the energy consumption value corresponding to sound are shown in FIG. 5 and FIG. 6, respectively. Therefore, when a specific V value is input, the current battery residual capacity environment attribute and the consumed energy value of the "running" behavior for the sound environment attribute are respectively derived from the specific value on the Y axis. Here, referring to the graphs of FIGS. 5 and 6, the consumed energy value of "running" in the current battery environment attribute is derived as about 1 to 3, and the consumed energy value of "running" To about -2 to about 2.

7 and 8 are graphs of energy values corresponding to the remaining battery level and sound among environmental attributes for the "sleeping" action. Here, referring to the graphs of FIGS. 7 and 8, the consumption energy value of "sleep" in the current battery environment property is derived as about -3 to 3. In the current sound environment property, Can be expected to be derived from about -0.5 to 1.

The energy value of one action is obtained from the energy value derivation function as described above, and then the energy values of other actions of the same action are summed up to obtain the energy value of the action. This process is as shown in Equation 5 above.

Hereinafter, the process of deriving the energy consumption value will be described in detail.

As a simple example, when determining the behavior of the robot with only the remaining battery power, For example, the sensing value of the remaining battery level detected by the sensor unit 100 is 70, and the behavior is compared only in the case of running and sleep. If the parameter value stored in advance in the energy value derivation function and the sensed value of the battery remaining amount are substituted, the consumed energy value of the running action and the consumed energy value of the sleeping behavior can be derived as follows. That is, it is a value derived by substituting only the parameter value relating to the remaining battery level and the detection amount value of the remaining battery level at f (P, V).

f (running _ battery remaining parameter value, 70) = 67

f (sleep_battery parameter value, 70) = 31

According to the above result, the robot behavior determining unit can perform the "running" action derived from the higher energy consumption value.

However, this is a case where only the remaining battery level is considered as an environmental factor, and other results may be derived when other environmental factors are combined. For example, when considering the illuminance in addition to the remaining amount of the battery, the energy consumption value corresponding to the illuminance of the "running" action may be lowered when the illuminance sensing amount detected by the illuminance sensor is too low (e.g. On the other hand, the energy consumption value corresponding to the illuminance of the "sleeping" action can be largely deduced. Therefore, when the sum of the energy consumption value corresponding to the battery remaining amount and the illumination energy value is larger than the total energy consumption value, sleeping can be selected even though the battery remaining amount is large.

According to the present invention, the robot selects a behavior among the behavior lists stored in the memory unit 200, and can select an action in consideration of the battery remaining amount and various other environmental factors. That is, a parameter value is embedded in the memory unit 200 for several environments and behaviors, and the external environment detected through the sensor unit 100 is derived as a sensing value converted into a predetermined numerical value, The value of the energy and the value of the sensed amount can be substituted into the energy value deriving function to derive the energy consumption value of each action under the environment of the specific condition. Then, the energy consumption value of each behavior for the entire environment can be derived by summing up the energy consumption values of the work done for all of the environmental factors. Then, by selecting an action having a maximum energy consumption value, for example, or by selecting an action according to another predetermined rule, the most appropriate action corresponding to the environment can be selected and performed. Therefore, it is possible to actively select and perform actions with respect to the surrounding environment.

9 is a view showing a behavior of a robot by the robot control system according to the present invention. The robot equipped with the robot control system according to the present invention detects environment elements according to various environmental attributes such as battery, sound, external motion, illumination, and the like. Then, based on each environment, It is possible to perform the running.

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, but, on the contrary, It should be understood that various modifications may be made by those skilled in the art without departing from the spirit and scope of the invention.

100:
200:
300: robot behavior decision unit
400: Robot action performing part

Claims (6)

A robot control system comprising:
A sensor unit for sensing a current environment corresponding to at least one environmental property;
A memory for storing a behavior list of a robot including a plurality of behaviors, and a predetermined parameter value corresponding to each behavior included in the behavior list and corresponding to each environment attribute detected by each sensor unit, as data;
A robot behavior determining unit for selecting any one of behaviors stored in the memory unit; And
A robot behavior performing unit performing a behavior selected by the robot behavior determining unit; / RTI >
The sensor unit includes:
The environment of the detected attribute is converted into a predetermined numerical value to derive a predetermined sensing value,
Wherein the robot behavior determination unit
A first algorithm for calculating a consumed energy value of each behavior included in the behavior list by using the parameter value stored in the memory unit and the sensed amount value,
And a behavioral selection algorithm having a second algorithm for comparing the energy consumption value of each behavior derived by the first algorithm to select any one action,
The parameter value may be,
Wherein the sensor has a value corresponding to one of the actions included in the behavior list and corresponding to one of the environmental attributes detected by the respective sensor units,
The first algorithm comprising:
The parameter value and the sensed amount value are substituted into Equation 1 below to derive a consumed energy value of a current one environment attribute of a certain action,
The consumed energy value under the current environment of the daily action is derived according to the following equation 2,
Equation (1) is a robot control system having the form of Equation (3).

The energy consumption value of the current work environment property of work = f (P, V) (Equation 1)
(P: parameter value, V: detection value value)

(Equation 2)
(B _j : energy consumption value under the current environment of action j,
P _ij : parameter value of environmental property i of behavior j,
V _i : detection value of environment attribute i)

(Equation 3)
(c ^h , c ^l , t ^h , t ^l : parameter value) delete delete The method according to claim 1,
The second algorithm comprising:
And selects a behavior having the largest energy consumption value by comparing energy consumption values of each behavior derived by the first algorithm. The method according to claim 1,
The sensor unit includes:
A power amount sensor for detecting the remaining power amount of the robot,
A sound sensor for detecting external sounds,
A motion sensor for detecting an external motion, and
And an illuminance sensor for detecting an external illuminance. The method according to claim 1,
The sensed amount value derived from the sensor unit,
Wherein the robot control system converts the measured value of the current environment into a percentage with respect to a predetermined reference value.

Images