JP2001157978A - Robot device, and control method thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To express natural growth in a pet robot. SOLUTION: In this robot device, and this control method for that, an action is generated by using a condition space in a part or all of an action model, and the condition space is changed to be enlarged or contracted. In the robot device and the control method for that, transition to a specified node in an action model comprising a probability condition transition model is described as transition to a supposed node comprising a supposable node, a first node group is alotted to the supposed node, and a group of nodes to be allotted to the supposed node are changed in order.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a robot apparatus and a control method therefor, and is suitably applied to, for example, a pet robot.

[0002]

2. Description of the Related Art In recent years, a pet robot of a four-legged walking type has been proposed and developed by the present applicant. Such a pet robot has a shape similar to a dog or cat bred in a general household, and can act autonomously in response to a user's action such as "hitting" or "patting" or the surrounding environment. It was done in.

The pet robot also has a learning function of changing the occurrence probability of a corresponding action based on a user's actions such as "hitting" and "stroking", and an action based on the accumulation of the actions and the elapsed time. It is equipped with a growth function that gradually changes the level of difficulty and complexity of the robot, thereby achieving high merchantability and amusement as a “pet robot”.

[0004]

By the way, in such a pet robot, an action model consisting of an individual stochastic state transition model is prepared for each growth stage (hereinafter, referred to as a "growth stage"), and a user receives an action model. "Growth" was expressed by switching the behavior model to the behavior model of the "growth stage" based on the accumulation of work and elapsed time. Further, in such a pet robot, the above-mentioned “learning” is expressed by changing the transition probability of a corresponding part of the behavior model in accordance with the action of the user.

However, according to this method, the behavior model is switched to a new behavior model each time the user grows up, so that a new behavior starts as if the personality has suddenly changed at each growth, The result of "learning" is discarded, and there is a problem that a user who is familiar with the behavior pattern up to that time feels unnatural.

According to this method, even if there is an overlapping behavior pattern, it is necessary to prepare a behavior model part for the behavior pattern for each "growth stage", and the operation of generating the behavior model is complicated accordingly. There was a problem of becoming.

Therefore, for example, if the behavior pattern and the learning result can be carried over to the next “growth stage” at the time of “growth”, the above-mentioned unnaturalness at the time of “growth” can be resolved, and It is thought that it is possible to express a more biological "growth" and to thereby improve the entertainment property.

The present invention has been made in view of the above points, and an object of the present invention is to propose a robot apparatus capable of improving entertainment characteristics and a control method thereof.

[0009]

According to the present invention, in order to solve the above-mentioned problems, in a robot apparatus, an action is generated using a holding means for holding an action model and a part or all of the state space of the action model. An action generation means is provided, and the action generation means changes a state space used for action generation in the action model while scaling the state space. As a result, in this robot device, since the state space used for action generation changes continuously, discontinuity of action output before and after the change of the state space used for action generation can be reduced.

According to the present invention, in a robot apparatus having an action model as a state transition model and generating an action based on the action model, a transition to a predetermined node in the action model is performed by a virtual node. Described as a transition to a virtual node, and a predetermined first
And a changing means for changing the node group to be assigned to the virtual node. As a result, in this robot device, since the basic behavior model is fixed, the output behavior can be made consistent.

Further, according to the present invention, in the method for controlling a robot apparatus, a first step of generating an action by using a part or all of the state space of the action model; And a second step of changing the space while scaling it. As a result, according to the control method of the robot apparatus, since the state space used for generating the action changes continuously, the discontinuity of the action output before and after the change of the state space used for generating the action is reduced. be able to.

Further, according to the present invention, in the control method of the robot apparatus, the transition to the predetermined node in the behavior model is described as a transition to a virtual node which is a virtual node, and the predetermined node is described in the virtual node. A first step of allocating a group and a second step of changing a node group to be allocated to a virtual node are provided. As a result, according to the control method of the robot device, since the basic behavior model is fixed, the output behavior can be made consistent.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

(1) First Embodiment (1-1) Configuration of Pet Robot According to First Embodiment In FIG. 1, reference numeral 1 denotes a pet robot according to the first embodiment as a whole, The leg units 3A to 3D are connected to the front, rear, left and right of the unit 2, respectively.
The head unit 4 and the tail unit 5 are connected to the front end and the rear end of the body unit 2, respectively.

As shown in FIG. 2, a CPU (Central Processing Unit) 10 and a DRA
M (Dynamic Random Access Memory) 11, Flash ROM (Read Only Memory) 12, PC (Personal Com
puter) A control unit 16 formed by connecting a card interface circuit 13 and a signal processing circuit 14 to each other via an internal bus 15 and a battery 17 as a power source of the pet robot 1 are housed therein. . The body unit 2 has an angular velocity sensor 18 for detecting the direction and the acceleration of the movement of the pet robot 1.
And an acceleration sensor 19 are also stored.

Further, the head unit 4 detects a CCD (Charge Coupled Device) camera 20 for capturing an image of an external situation, and a pressure received by a physical action such as "stroke" or "hit" from the user. A touch sensor 21, a distance sensor 22 for measuring a distance to an object located ahead, a microphone 23 for collecting external sound, and a speaker 24 for outputting a sound such as a cry. , An LED (Light Emitting Diode) (not shown) corresponding to the “eyes” of the pet robot 1 and the like are arranged at predetermined positions.

Further, the joints of the leg units 3A to 3D, the leg units 3A to 3D and the trunk unit 2
Each coupling portion of the head unit 4 and body unit connecting portion 2, and the tail unit actuator 25 _1, respectively, etc. The consolidated portion of the tail 5A of freedom minutes 5
To 25 _n and potentiometers 26 _{1 to} 26 _n are provided.

The angular velocity sensor 18, acceleration sensor 19, touch sensor 21, distance sensor 22, microphone 23, speaker 24 and potentiometer 26
Various sensors such as _{1 to} 26 _n , LEDs and actuators 25 _{1 to} 25 _n are respectively connected to the corresponding hubs 27.
_The CCD camera 20 and the battery 17 are connected to the signal processing circuit 14 of the control unit 16 through _{1 to} 27 _N.
Are directly connected to the signal processing circuit 14, respectively.

At this time, the signal processing circuit 14 sequentially takes in the sensor data, image data and audio data supplied from each of the above-mentioned sensors,
, And sequentially stored in a predetermined position in the DRAM 11. In addition, the signal processing circuit 14 sequentially takes in remaining battery power data indicating the remaining battery power supplied from the battery 17 and stores the data at a predetermined position in the DRAM 11.

The sensor data, image data, sound data and remaining battery data stored in the DRAM 11 in this manner are used when the CPU 10 controls the operation of the pet robot 1 thereafter.

In practice, when the power of the pet robot 1 is turned on, the CPU 10 executes the control program stored in the memory card 28 or the flash ROM 12 inserted in the PC card slot (not shown) of the body unit 2 into the PC card interface. The data is read out via the circuit 13 or directly and stored in the DRAM 11.

Further, the CPU 10 thereafter, based on the sensor data, image data, audio data, and remaining battery power data sequentially stored in the DRAM 11 from the signal processing circuit 14 as described above, and from the user, Judge whether or not there is an instruction for and action.

Further, the CPU 10 determines the result of this determination and D
And it determines a subsequent action based on the control program stored in the RAM 11, by driving the actuator 25 ₁ to 25 _n required based on the determination result,
The head unit 4 can be swung up and down, left and right, the tail 5A of the tail unit 5 can be moved, and each of the leg units 3A to 3A.
D is driven to perform an action such as walking.

At this time, the CPU 10 generates audio data as required, and supplies the generated audio data to the speaker 24 via the signal processing circuit 14 as an audio signal, thereby outputting an audio based on the audio signal to the outside, LED above
Is turned on, off or blinking.

As described above, the pet robot 1 can autonomously act according to its own and surrounding conditions, and instructions and actions from the user.

(1-2) Software Configuration of Control Program The software configuration of the above-described control program in the pet robot 1 is shown in FIG. In FIG.
The device driver layer 30 is located at the lowest layer of the control program, and includes a device driver set 31 including a plurality of device drivers. In this case, each device driver is a CCD camera 2
0 (FIG. 2), an object such as a timer, which is allowed to directly access hardware used in a normal computer, and performs processing in response to an interrupt from the corresponding hardware.

The robotic server object 32 is located above the device driver layer 30 and includes, for example, the various sensors and actuators 25 ₁ described above.
A virtual robot 33 comprising a software group that provides an interface for accessing the hardware, such as to 25 _n, the power manager 34 made of a software suite for managing the power supply switching, the other various device drivers Device driver manager 35, which is a group of software to be managed, and pet robot 1
And a designed robot 36 which is a software group for managing the mechanism.

The manager object 37 includes an object manager 38 and a service manager 39.
It is composed of In this case, the object manager 38 has the robotic server object 32,
Middleware layer 40 and application
The service manager 39 is a software group that manages activation and termination of each software group included in the layer 41. The service manager 39 stores connection information between objects described in a connection file stored in the memory card 28 (FIG. 2). Is a group of software that manages the connection of each object based on.

The middleware layer 40 is located on the upper layer of the robotic server object 32 and is composed of a group of software which provides basic functions of the pet robot 1 such as image processing and sound processing. I have.
Further, the application layer 41 is located in a layer above the middleware layer 40, and
It is composed of a software group for determining the behavior of the pet robot 1 based on the processing result processed by each software group constituting the layer 40.

FIGS. 4 and 5 show specific software configurations of the middleware layer 40 and the application layer 41, respectively.

In the middleware layer 40,
As is clear from FIG. 4, for scale recognition, for distance detection,
Recognition system 5 including signal processing modules 50 to 55 for posture detection, touch sensor, motion detection, and color recognition, input semantics converter module 56, and the like.
7 and the output semantics converter module 57 and for attitude management, tracking, motion playback,
An output system 65 includes signal processing modules 58 to 64 for walking, falling back, turning on LEDs, and reproducing sounds.

In this case, the signal processing modules 50 to 55 of the recognition system 57 are connected to the DRAM 11 by the virtual robot 33 of the robotic server object 32.
The corresponding data among the sensor data, image data, and audio data read out from FIG. 2 (FIG. 2) is fetched, predetermined processing is performed based on the data, and the processing result is provided to the input semantics converter module 56.

The input semantics converter module 56, based on the processing results given from each of the signal processing modules 50 to 55, "detects a ball"
"Detected fall,""stroke,""beaten,"
"I heard the domeso scale,""Detected a moving object."
Alternatively, it recognizes the situation of itself and surroundings, such as "detected an obstacle", and commands and actions from the user, and outputs the recognition result to the application layer 41 (FIG. 2).

As shown in FIG. 5, the application layer 41 is composed of five modules: a behavior model library 70, a behavior switching module 71, a learning module 72, an emotion model 73, and an instinct model 74.

In this case, the behavior model library 70 includes
As shown in FIG. 6, "when the battery level is low", "when the vehicle returns to the fall", "when avoiding obstacles", "when expressing emotion", "when the ball is detected", and the like. Respectively correspond to several pre-selected condition items, and independent behavior models 70 _{1 to} 7 respectively.
0 _n is provided.

The behavior models 70 ₁ to 70 _1
n is stored in the emotion model 73 as described later as necessary when a recognition result is given from the input semantics converter module 56 or when a certain time has elapsed since the last recognition result was given. The next action is determined with reference to the corresponding emotion parameter value and the corresponding desire parameter value held in the instinct model 74, and the determination result is output to the action switching module 71.

In this embodiment, each of the behavior models 70 ₁ to 70 _n uses one node (state) NODE _{0 to} NO as shown in FIG.
Based whether to transition from DE _n to any other node NODE ₀ ~NODE _n the transition probability P ₁ to P _n which is set respectively arc ARC ₁ ~ARC _n1 to connect each node NODE ₀ ~NODE _n An algorithm called a probabilistic automaton that determines stochastically is used.

More specifically, each of the behavior models 70 ₁ to 70
_n corresponds to each of the nodes NODE _{0 to} NODE _n forming their own behavior models 70 ₁ to 70 _n , respectively, and FIG. 8 for each of these nodes NODE _{0 to} NODE _n
Has a state transition table 80 as shown in FIG.

In this state transition table 80, the node NO
Input events (recognition results) as transition conditions in DE _{0 to} NODE _n are listed in order of priority in the row of “input event name”, and further conditions for the transition conditions are described in the rows of “data name” and “data range”. It is described in the corresponding column.

Therefore, in the node NODE ₁₀₀ represented by the state transition table 80 of FIG. 8, "ball detected (BALL)"
When the recognition result is given, the "size (SIZE)" of the ball given together with the recognition result is in the range of "0 to 1000", or the recognition of "obstacle detected (OBSTACE)" is given. When the result is given, the condition for transition to another node is that the "distance" to the obstacle given together with the recognition result is in the range of "0 to 100". .

In the node NODE ₁₀₀ , even when the recognition result is not input, the behavior models 70 _{1 to} 70 ₁
0 _n of the parameter values of each emotion and each desire held in the emotion model 73 and the instinct model 74 that are periodically referred to, the “joy (JO)” held in the emotion model 73.
When the parameter value of any of "Y)", "Surprise", or "Sadness" is in the range of "50 to 100", it is possible to make a transition to another node.

Further in the state transition table 80, with the node name that can transition from the node NODE ₀ ~NODE _n in the column of "transition destination node" in the column of "probability of transition to another node" is listed, the "input Other nodes NODE _{0 to} N that can transition when all the conditions described in the rows of “event name”, “data value”, and “data range” are met
The transition probability to ODE _n is described in a corresponding part in the column of “transition probability to another node”, and the node N
The action to be output when transitioning from ODE _{0 to} NODE _n is “output action” in the column of “transition probability to another node”.
Is described in the line. Note that the sum of the probabilities of the respective rows in the column of “transition probability to another node” is 100 [%].

Therefore, in the node NODE ₁₀₀ represented by the state transition table 80 in FIG.
L) "and the" SIZE "of the ball is" 0 ".
Is given in the range of “node NODE” with a probability of “30 [%]”.
₁₂₀ (node 120) ”, and then“ ACTIO ”
The action of "N1" will be output.

Each of the behavior models 70 ₁ to 70 _n is configured such that a number of nodes NODE _{0 to} NODE _n described as such a state transition table 80 are connected, and the input semantics converter module 56 For example, when a recognition result is given, the next action is stochastically determined using the state transition table 80 of the corresponding nodes NODE ₀ to NODE _n , and the determination result is output to the action switching module 71. ing.

The action switching module 71, among the actions respectively output from the behavior model 70 ₁ to 70 _n of the action model library 70, output from the predetermined high priority behavior model 70 ₁ to 70 _n A command for selecting an action and executing the action (hereinafter, referred to as an action command) is transmitted to the middleware layer 40.
To the output semantics converter 57. Incidentally in this embodiment, is denoted behavioral models 70 ₁ more to 70 _n priority lower is set higher in FIG.

Further, based on the action completion information provided from the output semantics converter 57 after the action is completed, the action switching module 71 notifies the learning module 72, the emotion model 73, and the instinct model 7 that the action has been completed.
Notify 4.

On the other hand, the learning module 72 inputs the recognition result of the teaching received from the user, such as “hit” or “stroke”, among the recognition results given from the input semantics converter 56.

Then, based on the recognition result and the notification from the action switching module 71, the learning module 72 lowers the probability of occurrence of the action when “beaten (scorched)” and “strokes (praise)”. was) "sometimes to increase the expression probability of that action, behavior model 70 _1-7 corresponding in behavioral model library 70
Change the corresponding transition probability of 0 _n .

On the other hand, the emotion model 73 indicates “joy (joy)
) "," Sadness "," anger "
) "," Surprise "," disgust "
For a total of six emotions of “fear” and “fear”, a parameter indicating the strength of the emotion is stored for each emotion. Then, the emotion model 73 converts the parameter values of each of these emotions from the specific recognition results such as “hit” and “stroke” given from the input semantics converter module 56 and the elapsed time and action switching module 71, respectively. The notification is sequentially updated based on the notification and the like.

Specifically, the emotion model 73 includes the recognition result from the input semantics converter 56 and the degree (preset) at which the behavior of the pet robot 1 acts on the emotion at that time. The held parameter values of each desire and the degree at which the behavior of the pet robot 1 acts on the emotion (preset), the degree of suppression and stimulation received from other emotions, the elapsed time ΔE [t], the current parameter value of the emotion E [t], and the rate of change of the emotion according to the recognition result, etc. hereinafter, the coefficient representing the same is referred to as sensitivity) as k _e, the following expression in a predetermined cycle

[0051]

(Equation 1)

Is used to calculate the parameter value E [t + 1] of the emotion in the next cycle.

The emotion model 73 updates the parameter value of the emotion by replacing the calculation result with the parameter value E [t] of the emotion. It is to be noted that the parameter value of the emotion to be updated in response to each recognition result or the notification from the action switching module 71 is determined in advance. For example, when a recognition result such as “hit” is given, “anger” is given. The emotion parameter value has increased,
When a recognition result such as “stroke” is given, the parameter value of the emotion of “joy” increases.

On the other hand, the instinct model 74 is composed of “exersise”, “affection”, “appetite” (ap
For each of the four independent desires "petite" and "curiocity", a parameter indicating the strength of the desire is held for each of the desires. Then, the instinct model 74 converts the parameter values of the desire into the recognition result given from the input semantics converter module 56, the elapsed time and the action switching module 71, respectively.
The information is sequentially updated based on a notification or the like.

More specifically, the instinct model 74 is “movement desire”,
Regarding “affection desire” and “curiosity”, the change amount of the desire calculated by a predetermined arithmetic expression based on the action output, elapsed time, and recognition result of the pet robot 1 is ΔI
[K], I [k] of the current parameter value of the desire, the coefficient representing the sensitivity of the desire as k _i, the following expression in a predetermined cycle

[0056]

(Equation 2)

Is used to calculate the parameter value I [k + 1] of the desire in the next cycle, and replaces the result of this operation with the current parameter value I [k] of the desire to update the parameter value of the desire. I do. It is to be noted that the parameter value of the desire to be changed with respect to the behavior output, the recognition result, and the like is determined in advance. Parameter value decreases.

In the instinct model 74, the “appetite” is calculated based on the remaining battery level data provided via the input semantics converter module 56, by setting the remaining battery level to B _L at a predetermined cycle as follows:

[0059]

(Equation 3)

The parameter value I [k] of “appetite” is calculated by the
The parameter value of the “appetite” is updated by replacing with [k].

In the present embodiment, the parameter value of each emotion and each desire is regulated to fluctuate in the range of 0 to 100, and the coefficients k _e and k _i
Is also set individually for each emotion and each desire.

On the other hand, as shown in FIG. 4, the output semantics converter module 57 of the middleware layer 40 provides “forward” and “pleased” provided by the action switching module 71 of the application layer 41 as described above. , "Sound" or "tracking (follow the ball)" to the corresponding signal processing modules 58 to 64 of the output system 65.

The signal processing modules 58 to 6
4, given the behavior command based on the action command, and servo command value to be applied to the actuator 25 ₁ to 25 _n (Fig. 2) corresponding to perform that action, the output from the speaker 24 (FIG. 2) Generating sound data of the sound to be played and / or driving data to be given to the LED of the “eye”,
These data are transferred to the virtual robot 33 of the robotic server object 32 and the signal processing circuit 14.
The corresponding actuators 25 ₁ to 25 ₁ through (FIG. 2)
25 _n , sequentially to the speaker 24 or the LED.

As described above, the pet robot 1 can perform an autonomous action according to its own and surrounding conditions, and instructions and actions from the user, based on the control program. .

(1-3) Growth Model of Pet Robot 1 Next, a growth function mounted on the pet robot 1 will be described. The pet robot 1 is provided with a growth function that changes behavior as if it were "growing" in response to a user's action or the like.

That is, the pet robot 1 is provided with five “growth stages” as growth processes, “birth period”, “childhood period”, “childhood period”, “adolescence period”, and “adult period”. . Then, in the behavior model library 70 (FIG. 5) of the application layer 41, among the condition items such as “when the remaining battery power is low”,
As shown in FIG. 9, for all condition items related to “growth” such as “motion” and “behavior” (hereinafter referred to as “growth-related condition items”), the behavior model 70
_{As k} , behavior models 70 _{k (1)} to 70 _{k (5)} corresponding to “birth”, “childhood”, “childhood”, “adolescence” and “adult” are provided, respectively. . In the behavior model library 71, the next behavior is initially determined for these growth-related condition items by using the behavior model _{70k (1)} of the “birth period”.

In this case, each behavior model 70 of “birth period”
_{k (1)} has a small number of nodes NODE _{0 to} NODE _n (FIG. 7), and the behavior output from these behavior models 70 _{k (1)} is “pattern 1 (walking pattern for“ birth period ”)”. The action or action content corresponds to the "birth period", such as "sound in pattern 1" (a crying pattern for "birth period").

Thus, in this pet robot 1, initially, for example, “walk” for “motion”, according to each behavior model 70 _{k (1)} of “birth”,
The action is performed so as to be “monotonous” by repeatedly performing the same action so that the movement becomes “simple” movement such as “stand” and “sleep”.

At this time, the application layer 4
The first learning module 72 (FIG. 5) holds therein a parameter indicating the degree of “growth” (hereinafter, referred to as a growth parameter), and the recognition result and the elapsed time given from the input semantics converter module 56. Based on the information and the like, the value of the growth parameter is sequentially updated in accordance with the number of times (teaching) from the user such as “stroke” or “hit”, the elapsed time, and the like.

The learning module 72 evaluates the value of the growth parameter each time the power is turned on to the pet robot 1, and when the value exceeds a threshold value set in advance corresponding to “childhood”. Notifies the behavior model library 70 of this. In addition, the behavior model library 70
When this notification is given, the behavior model to be used is changed to the behavior model 70 _{k (2)} of “childhood” for each of the above-mentioned growth-related condition items.

At this time, each behavior model 70 of “childhood”
_{k (2)} is the node N more than the behavior model 70 “birth” of _{k (1).}
The number of ODE _{0 to} NODE _n is large, and the content of the behavior output from the behavior model 70 _{k (2)} also has a higher level of difficulty and complexity (growth level) than the behavior in “childhood”. ing.

Thus, in the pet robot 1, after that, according to the behavior model _{70k (2)} , for example, “motion” becomes “slightly sophisticated and complicated” by increasing the number of actions. In addition, regarding the "action", the action is performed so as to be "with a little purpose".

Further, the learning module 74 thereafter sets the thresholds set in advance so that the values of the growth parameters correspond to “childhood”, “adolescence”, and “adulthood”, respectively, in the same manner as described above. This is notified to the behavior model library 71 each time it exceeds. Behavior model library 7
1. Each time this notification is given, the behavior model used for each of the above growth-related condition items is _{defined as} a behavior model 70 _{k (3) of} "childhood", "adolescence" and "adult".
Change sequentially to ~ 70k ₍₅₎ .

At this time, each of the behavior models 70 _{k (3)} to 70 _{k (5)} of “childhood”, “adolescence” and “adult” respectively includes nodes NODE ₀ to NODE ₀ to
The number of NODE _n increases, and these behavior models 70
The content of the action output from _{k (3)} to 70k ₍₅₎ also increases in the level of difficulty and complexity of the action as the “growth stage” rises.

As a result, in this pet robot 1, the “growth stage” is raised (ie, “birth stage” to “childhood”, “childhood” to “boyhood”, “childhood” to “adolescence”, “ "Adolescent" to "adult"), "motion" changes from "simple" to "advanced / complex"
Also, the “action” changes stepwise from “monotonous” to “action with purpose”.

As described above, in the pet robot 1, the actions and actions are changed to “birth period”, “childhood”, “boyhood”, “adolescence”, It is designed to "grow" in five stages of "adult".

In the case of this embodiment, the growth model of the pet robot 1 is a model that branches off after the "boyhood" as shown in FIG.

That is, in the case of the pet robot 1, the behavior model library 70 of the application layer 41 (FIG. 5) stores
Behavior model 7 for "childhood", "adolescence" and "adult"
A plurality of action models are prepared for each of 0 _{k (3)} to 70 _{k (5)} .

Actually, as the behavior model 70 _{k (3)} of each growth-related condition item, for example, “childhood”, a behavior model (CHILD ₎ for performing an action with a rough and fast movement that is rough and fast. 1) and an action model (CHILD 2) for causing the user to perform an action of a “soft” character whose movement is smoother and slower than this.

The behavior model 70 _{k (4)} of the “adolescent” includes a behavior that is more sloppy and faster than the “rough” personality of the “boyhood” and has an “irritated” behavior. An action model (YOUNG 1) and an action model (YOUNG 2) for performing actions and actions of a “normal” character that is slower and smoother than this, and a slower action and an action amount An action model (YOUNG 3) is provided to allow the user to perform an action with a small "unseen" personality.

Further, the behavior model 70 _{k (5)} of the “adult period” has a more aggressive “aggressive” character that moves more swiftly and is more angry than the “irritable” character of the “adolescent period”. Behavior model (ADULT 1)
An action model (ADULT) that allows the user to perform a behavior that is smoother, slower, and more angry and more angry.
2) and a behavior model (ADULT 3) for making the behavior “slow” with smoother and slower movement and less movement, and a movement that is even slower and lower than this An action model (ADULT 4) is provided to allow the user to perform actions with less "quiet" personality.

When the learning module 72 (FIG. 5) of the application layer 41 notifies the behavior model library 70 to raise the “growth stage” as described above, the learning module 72 (FIG. 5) , "Strapped" and "stroked" in its "growth stage"
Based on the frequency, etc., the behavior model of any “character” (CHILD 1, CHILD 2, YOUNG 1 to YOUNG 3, ADULT 1) as the behavior model of the “growth stage” next to each growth-related condition item
~ ADULT 4) Specify whether to use.

As a result, the behavior model library 70
Based on this designation, for each growth-related condition item,
Behavior model 70 _{k (3)} -70 used after “childhood”
_{k (5)} designated behavior model of “personality” (CHILD 1, CH
ILD2, YOUNG1 to YOUNG3, ADULT1 to ADULT4).

In this case, after “childhood”, when moving to the next “growth stage”, “character” in the next “growth stage” is determined by “character” in the current “growth stage”. Only the transition between the "characters" connected by arrows in FIG. 10 can be made. Therefore, for example, when the behavior model (CHILD 1) with the character “rough” in “childhood” is used, the behavior model (YOUNG 3) with the character “decoyed” in “adolescence” is used. Can not do.

As described above, in this pet robot 1, as if a real animal forms a character according to the manner of breeding by the owner, etc., the pet robot 1 grows in accordance with the action of the user and the like. "Personality" is also changed. (1-4) Specific Configuration of Behavior Model 70 _k Next, the behavior model 70 _k of each of the growth-related condition items described above.
The specific configuration of FIG. 9 will be described.

In the case of the pet robot 1, the behavior model 70 _k of each growth-related condition item has a vast state space in which all behavior patterns that can be expressed by the pet robot 1 are stored.

The behavior model 70 _k is based on a state space portion that generates basic behaviors of the pet robot such as “walk”, “sleep”, and “stand” in the state space as a core. In the “period”, only a very small part including the core is used, and each time the “growth” is performed, a state space part to be newly added (a state that creates a new action or a series of action patterns) The transition of the behavior model 70 in each “growth stage” is performed by permitting the transition to the “space stage” and separating the state space portion that is no longer used (the state space portion that generates a behavior or a series of behavior patterns that is not used). _{k (1)} to 70 _{k (5)} are generated.

In the pet robot 1, as a method of permitting a transition to a state space part to be newly added or separating an unnecessary state space, the transition probability to the state space is determined according to “growth”. Change method.

For example, in FIG. 11, the node NODE _A
The transition condition from the node NODE _B to "node found" is that the series of nodes 90 from node NODE _B perform a series of action patterns such as "approach the ball and kick it" Then
Although behavior pattern PA ₁ of "kicking it chasing a ball" in the transition probability P ₁ is when finding ball at node NODE _A is expressed, the transition probability P ₁
Is “0”, such an action pattern PA
₁ is never expressed.

Therefore, in this pet robot 1,
It is when the turned "growth stage" to express such behavior pattern PA ₁ is the initial time may be set this transition probability P ₁ to "0", when it is the "growth stage" the transition probability P ₁ so as to change to a larger preset number than "0". On the contrary to this, in the case of forget this behavior pattern PA ₁ when it becomes a "growth stage", the node NODE _A from the node N when they become the "growth stage"
The transition probability to the ODE _B are to be changed to "0".

In this pet robot, as a specific method for updating the transition probabilities of the necessary portions, the behavior models of the growth-related condition items include “childhood”, “childhood”, Files corresponding to the “growth stages” of “adolescence” and “adult”, respectively, as shown in FIG.
1A to 91D are provided.

The difference files 91A to 91D correspond to the nodes whose transition probabilities are to be changed as described above (to correspond to the node A in FIG. 11) in order to express a new action as described above when going up to the “growth stage”. ) And the node state transition table 80 (FIG. 8)
In which the transition probability at which the transition probability should be changed and the transition probability after the change at the location are stored.

The behavior model 7 for each growth-related condition item
0 _k is initially the behavior model 70 _{k (1)} for “birth period”
On the other hand, when the learning module 72 (FIG. 5) later gives a notification that "growth" has been performed as described above, the corresponding difference file 91 for "growth stage" is generated.
For each of the nodes described in the difference files 91 to 91D, the transition probabilities of the designated portions are changed to the designated numerical values based on 91D.

For example, in the case of the examples shown in FIGS. 8 and 12, when the child grows up in “childhood”, the behavior model 70 _k of each growth-related condition item has a transition probability in the state transition table 80 of the node NODE _100. "1st row" and "1" in the described area (in FIG. 8, below the column of "output action" and on the right side of the row of "data range")
The transition probability of the “column” is changed to “20” [%], and the transition probability of the “first row” and “n-th column” of the state transition table is “30” [%],. To each. At the same time, the behavior model 70 _k of each growth-related condition item is:
The corresponding transition probabilities are similarly changed for the other nodes NODE ₃₂₀ , NODE ₇₂₀ ,... Described in the difference file 91A for “childhood”.

In this case, among the transition probabilities that change the numerical value in this way, the transition probability up to that point is “0” (that is, transition to a node serving as a starting point of a series of action patterns is prohibited). And the transition probability after change is “0” (that is, the transition to the node serving as the starting point of the series of action patterns is prohibited). By changing from "0" to a predetermined numerical value, or by changing the transition probability to "0", the series of behavior patterns comes to appear in a new "growth stage", or A series of behavior patterns may not be expressed.

Even when the necessary transition probabilities are changed in this manner, each difference file is set so that the sum of the transition probabilities included in the corresponding row in the changed state transition table 80 becomes 100 [%]. The values of the respective transition probabilities in 91A to 91D are selected.

(1-5) Operation and Effect of the Embodiment In the above configuration, the pet robot 1 performs basic actions in a vast state space in which all action patterns are stored. With the state space part as the core,
In the "birth period", only a very small part including the core is used, and each time "growth" thereafter, the unused state space other than the core is separated and the transition to the state space that you want to increase newly Allow each "growth stage" to allow
, The action models 70 _{k (1)} to 70 _{k (n)} are generated, and the action is performed according to the generated action models 70 _{k (1)} to 70 _{k (n)} .

Therefore, in the pet robot 1, the state space of the behavior models 70 _{k (1)} to 70 _{k (n)} in each “growth stage” changes continuously, so that the output behavior before and after “growth” is obtained. Can be reduced, and "growth" can be expressed more naturally. Further, in this pet robot 1, the state space portion for generating the basic action is shared by all the “growth stages”, so that the learning result of the basic action is sequentially carried over to the next “growth stage”. be able to.

Further, in the pet robot 1, since the state space part for generating the basic behavior is shared by all the “growth stages”, the behavior model 70 _{k (1) of} each “growth stage” is used. It is easy to create up to 70 _{k (n)} , and it is also possible to reduce the data amount of the entire behavior model as compared with the case where an individual behavior model is prepared for each "growth stage" as in the past. .

Further, in the pet robot 1, as described above, the state space of a series of unnecessary action patterns is cut off in accordance with the "growth", and the transition of the series of necessary action patterns to the state space is permitted. Each "Growth Stage"
Since the behavior models 70 _{k (1)} to 70 _{k (n)} in the above are generated, each series of behavior patterns can be made into parts, and the behavior model 70
The operation of generating _K can be further facilitated.

According to the above configuration, the state space portion for performing basic actions in the vast state space in which all action patterns are stored is used as a core, and the "birth period" includes the core. Using only a very small part, each time "growth" thereafter, separate the unused state space parts other than the core and allow transition to the state space part that you want to newly increase, each "growth stage" By generating the behavior models 70 _{k (1)} to 70 _{k (n)} in
Behavior model 70 _{k (1)} to 70 in each “growth stage”
By continuously changing the state space of _{k (n)} , it is possible to reduce the discontinuity of the output behavior before and after “growth”.
Thus, "growth" can be more naturally expressed, and thus a pet robot that can improve entertainment properties can be realized.

[0102] (2) As shown in the second embodiment (2-1) Principle Figure 11, the transition to a series of behavior patterns PA ₁ to get new by "growth" is a particular state (node NODE _A If the only occur from), it can control the expression of the behavior patterns by simply changing the transition probabilities P _1. However, as shown in FIG. 13, this transition has a plurality of states (nodes NODE _{A1 to} NOD
E _A3 ), all corresponding transition probabilities P
To Kotororu the ₁₀ to P ₁₂ it is not easy.

Therefore, in such a case, as shown in FIG. 14, a virtual node (hereinafter, referred to as a virtual node) NODE _K is provided in the action model, and the nodes NODE _A1 to NODE _{A3 are connected} to each other. A series of action patterns PA
Replacing the transition to the node NODE _B as the _first starting point to the transition to the virtual node NODE _K, the virtual node NODE _K
And the node N at the starting point of the above-described series of action patterns PA ₁
What is necessary is just to make it correspond to ODE _B.

In this way, the transition probability can be easily controlled, and this series of action patterns PA ₁ is replaced with another series of action patterns P
In the case of switching to the A ₂ also, the virtual node NODE
The behavior pattern PA before the correspondence of the real node to _K
From the node NODE _{B at} the starting point of ₁ to the next action pattern PA
This can be done easily only by changing the node NODE _C for ₂ origin.

(2-2) Configuration of the Pet Robot 100 According to the Present Embodiment Here, 100 in FIG. 1 shows the pet robot according to the second embodiment as a whole, and each of the pet robots related to “growth”. The pet robot 1 according to the first embodiment, except that the configuration of the action model 70 _k (FIG. 9) of the condition item is different.
It is configured similarly to.

That is, in this pet robot 100, the behavior model 70 _k of each growth-related condition item includes
As shown in FIG. 5, an action model for generating a basic action common to each “growth stage” such as “stand”, “sit”, and “walk” (hereinafter referred to as a basic action model) 101, some virtual nodes NODE _{K1 to} NODE are included in the basic behavior model 101.
DE _Kn is provided.

The behavior model 70 for each growth-related condition item
_{In k} , behavior pattern files 102A to 102E are provided corresponding to the respective “growth stages”. These action pattern files 102A to 102E
Are, as shown in FIG. 16, each virtual node N in the basic behavior model 101 in the “growth stage”.
This is a file in which state transition tables of a group of nodes that respectively generate a series of action patterns PA _{1 to} PA _n associated with ODE _{K1 to} NODE _Kn are collected.

Further, the behavior model 7 of each growth-related condition item
0 The _k, as shown in FIG. 17, behavior pattern file 102 for each virtual node NODE _K1 ~NODE _Kn and the actual node in the "growth stage" (the "growth stage"
Any of the action patterns PA stored in A to 102E
There is provided a file (hereinafter, referred to as a difference file 103) in which correspondence tables 103A to 103E each representing a correspondence relationship with a node at a starting point of _{1 to} PA _n (hereinafter the same).

The behavior model 7 for each growth-related condition item
0 _k, together with the initial time to add it reads the data of the behavior pattern file 102A for the "birth stage" to the basic behavior model 101, the corresponding table 103A for "Birth-life" stored in the difference file 103 Based on each virtual node NODE _{K1 to} NOD in the basic behavior model 101.
An action model 70 _{k (1)} for “birth period” is generated by converting E _Kn into a real node, and an action is generated based on the action model 70 _{k (1)} .

The behavior model 70 for each growth-related condition item
_{k then} "grows" from learning module 72 (FIG. 5).
When the notification is given, the data of the behavior pattern file 102B for "childhood" is added to the basic behavior model 101 in place of the data of the behavior pattern file 102A for "birth" and the difference file 103 Each virtual node NODE _{K1 to} NO in the basic behavior model 101 based on the stored “childhood” correspondence table 103B
A behavior model 70 _{k (2)} for “childhood” is generated by converting DE _Kn into a real node, and the behavior model 70 _{k (2)}
Generate an action based on

Further, the behavior model 7 for each growth-related condition item
0 _k is then set to the learning module 72
Each time a notification of “growth” is given from the user, the behavior pattern file 102A added to the basic behavior model 101
The data of “-102E” are sequentially changed to those for “childhood”, “adolescence” and “adult”, and the “growth stage” correspondence table 1 stored in the difference file 103 is shown in FIG.
Each of the virtual nodes NODE _{K1 to} NODE _Kn in the basic behavior model 101 is converted into a real node based on the basic behavior model 03C to 103E, so that the behavior model 70 for “childhood”, “adolescence”, and “adult” is obtained. _{k (3)} to 70 _{k (5)} are sequentially generated,
An action is generated based on the action models 70 _{k (3)} to 70 _{k (5)} .

Thus, the pet robot 100
Then, each virtual node NODE in the basic behavior model 101
Behavior pattern PA corresponding to each of _{K1 to} NODE _Kn
And a _{1 ~PA n} to sequentially changed in accordance with the "growth", it has been made to alter act accordingly "growth".

(2-3) Operation and Effect of the Embodiment In the above configuration, the pet robot 100
Each virtual node NODE _{K1 to} N in the basic behavior model 101
Behavior patterns PA _{1 to} P corresponding to ODE _Kn
An is changed sequentially with “growth”, and the behavior is changed according to “growth”.

Therefore, in the pet robot 100, the basic behavior model 10 for generating a basic behavior model
1 is shared in all "Growth Stages", so that behavior can be consistent throughout all "Growth Stages" and learning results of basic behaviors are sequentially carried over to the next "Growth Stage" be able to.

Further, in this pet robot 100, the basic behavior model 101 is thus converted into all the “growth stages”.
This makes it easy to create behavior models, and also reduces the data volume of the entire behavior model as compared to the case where individual behavior models are prepared for each "growth stage" as in the past. it can.

Further, in the pet robot 100, the virtual nodes NODE _{K1 to} NODE in the basic behavior model 101
Each of the behavior patterns PA _{1 to} PA _n associated with DE _Kn can be made into parts, and the task of generating the behavior model 70 _K of each growth-related condition item can be further facilitated.

Further, in the pet robot 100, in addition to the same functions and effects as those obtained in the first embodiment, as described above, the virtual node NO
In order to use DE _K1- NODE _Kn ,
For example a series of behavior patterns node NODE _A1 transition is here and there to PA ₁ in as shown in FIG. 13 ~NODE _A3
When that occurs from also it can be made to the behavior model 70 _K of the growth-related condition item can readily generated.

According to the above configuration, the basic behavior model 10
1, virtual nodes NODE _{K1 to} NODE _Kn are provided, and each of the virtual nodes NODE _{K1 to} NODE _Kn has a series of behavior patterns PA _{1 to} PA for “birth period”.
_n (originating nodes) are associated with each other to generate the behavior model 70 _{k (1)} for each “birth period”, and thereafter, with the “growth”, the virtual nodes NODE _{K1 to} NODE _Kn
Each of a series of action patterns PA _{1 to} P
For the "childhood" of the A _n, "childhood" for, as Yuku is switched for the for the "adolescence" and "adulthood", for these "childhood", for the "childhood", "adolescence "And" Adult "
By generating the behavior models 70 _{k (2)} to 70 _{k (5)} , the behavior can be made consistent throughout all “growth stages”. Thus, "growth" can be more naturally expressed, and thus a pet robot that can improve entertainment characteristics can be realized.

(3) Other Embodiments In the above-described first and second embodiments, a case has been described in which the present invention is applied to the quadruped walking pet robots 1 and 100. However, the present invention is not limited to this, and can be widely applied to various other types of robots.

In the above-described first and second embodiments, the state space of the behavior models 70 _{k (1)} to 70 _{k (5)} of each “growth stage” is sequentially enlarged with “growth”. However, the present invention is not limited to this, and the state space of the behavior models 70 _{k (1)} to 70 _{k (5)} of each “growth stage” is sequentially reduced or changed during the expansion. At the "growth stage", the behavior model 70 _{k (1)} -7
The state space of 0 _{k (5)} may be reduced.

Furthermore, in the above-described first and second embodiments, the case where the pet robots 1 and 100 “grow” in five stages has been described. However, the present invention is not limited to this, and the present invention is not limited thereto. It may be made to "grow" in the number of steps of.

Further, in the above-described first embodiment, the holding means for holding the behavior model (the behavior model including all behavior patterns that can be performed by the pet robot 1), and the state space of part or all of the behavior model Has been described in which the action generation means for generating an action by using one action model 70 _k and the CPU 10 has been described. However, the present invention is not limited to this, and various other configurations can be widely applied. Can be.

Further, in the above-described second embodiment, a case has been described in which the changing means for changing the node group assigned to the virtual nodes NODE _{K1 to} NODE _Kn is constituted by one action model 70 _k and the CPU 10. However, the present invention is not limited to this, and various other configurations can be widely applied.

[0124]

As described above, according to the present invention, in a robot apparatus and a control method thereof, an action is generated using a part or all of a state space of an action model, and the state space is enlarged and reduced. By changing the state space, the state space used for generating the action can be changed continuously, and the discontinuity of the action output before and after the change of the state space used for generating the action can be reduced.
As a result, the type of output action and the like can be smoothly and naturally sequentially changed, and thus a robot apparatus and a control method thereof capable of improving entertainment characteristics can be realized.

Further, in the robot apparatus and the control method therefor, a transition to a predetermined node in the action model composed of the stochastic state transition model is described as a transition to a virtual node composed of a virtual node. By assigning one node group and sequentially changing the node group assigned to the virtual node, the basic behavior model is fixed, so that the output behavior is sequentially changed while maintaining consistency. be able to.
As a result, the type of output action and the like can be smoothly and naturally sequentially changed, and thus a robot apparatus and a control method thereof capable of improving entertainment characteristics can be realized.

[Brief description of the drawings]

FIG. 1 is a perspective view illustrating an external configuration of a pet robot according to first and second embodiments.

FIG. 2 is a block diagram illustrating a circuit configuration of the pet robot.

FIG. 3 is a conceptual diagram showing a software configuration of a control program.

FIG. 4 is a conceptual diagram showing a software configuration of a middleware layer.

FIG. 5 is a conceptual diagram showing a software configuration of an application layer.

FIG. 6 is a conceptual diagram serving to explain an action model library.

FIG. 7 is a schematic diagram illustrating a stochastic automaton.

FIG. 8 is a table showing a state transition table.

FIG. 9 is a conceptual diagram showing a detailed configuration of an action model library.

FIG. 10 is a conceptual diagram showing a growth model of a pet robot.

FIG. 11 is a conceptual diagram for describing acquisition and forgetting of a behavior pattern accompanying growth.

FIG. 12 is a conceptual diagram serving to explain a difference file according to the first embodiment.

FIG. 13 is a conceptual diagram explaining transition from a plurality of nodes to a starting node of one action pattern.

FIG. 14 is a conceptual diagram explaining the use of a virtual node.

FIG. 15 is a conceptual diagram showing a configuration of an action model of each action-related condition item according to the second embodiment.

FIG. 16 is a conceptual diagram serving to explain an action pattern file.

FIG. 17 is a conceptual diagram serving to illustrate a difference file according to the second embodiment.

[Explanation of symbols]

1, 100 ... pet robot, 10 ... CPU, 16
... Control unit 33... Virtual robot, 40... Middleware layer, 41, 100... Applicationware 70... Behavior model library, 70 ₁ to 70 _n ,
70 _k , 70 _{k (1)} to 70 _{k (5)} ,..., Behavior model, 91
A~91D, 103 ...... difference file, 101 ...... basic behavior model, 102A~102E ...... behavior pattern file, PA ₁ ~PA _n ...... behavior patterns, NODE _K1 ~
NODE _Kn ...... Virtual node.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2C150 CA02 DA05 DA24 DA26 DA27 DA28 DF02 DF04 EF16 EF23 EF29 3F059 AA00 BA00 BB06 DA07 DB04 DC00 DC01 DC08 FC00 5H004 GA26 GB16 HA07 HB07 HB08 HB09 HB15 JB29 KD55 KD56 MA63 BB03 FF04 GG05 HH05 HH21 JJ07 KK29 KK31 KK32

Claims (12)

[Claims] 1. An apparatus comprising: holding means for holding an action model; and action generating means for generating an action using a part or all of the state space of the action model, wherein the action generating means includes: Wherein the state space used for generating the action is changed while being enlarged or reduced. 2. The behavior model is a stochastic state transition model. The behavior generation means sets the transition probability to a state in which transition is prohibited by setting the transition probability of the behavior model to 0. The robot apparatus according to claim 1, wherein the state space used for generating the action in the action model is expanded by changing the state space to a predetermined value larger than 0. 3. The behavior model is a probability state transition model, and the behavior generation means sets a transition probability to a target state to 0,
The robot apparatus according to claim 1, wherein the state space used for generating the action in the action model is reduced. 4. A growth model that grows in a stepwise manner, wherein the action generation means expands or contracts a state space used for the action generation in the action model in accordance with the growth of the growth model. The robot apparatus according to claim 1, wherein the change is performed while changing. 5. An action model comprising a state transition model,
In the robot device that generates an action based on the action model, in the action model, a transition to a predetermined node is described as a transition to a virtual node that is a virtual node, and a predetermined node is assigned to the virtual node. A robot apparatus to which a group is assigned, comprising changing means for changing the node group assigned to the virtual node. 6. A growth model having a growth model that grows in a stepwise manner, wherein the changing means includes changing the node group assigned to the virtual node in accordance with the growth of the growth model. Item 6. The robot device according to item 5. 7. A control method of a robot apparatus having an action model and generating an action based on the action model, wherein a first or second action space is generated using a part or all of the state space of the action model. A control method for a robot apparatus, comprising: a step; and a step of changing the state space of the behavior model used for generating the behavior while changing the state space. 8. The behavior model is a stochastic state transition model. In the second step, the transition probability to a state where transition is prohibited by setting the transition probability of the behavior model to 0 is set. The control method for the robot apparatus according to claim 7, wherein the state space used for generating the action in the action model is expanded by changing the value of the action model to a predetermined value larger than 0. 9. The behavior model is a probability state transition model. In the second step, a transition probability to a target state is set to 0,
The method according to claim 7, wherein the state space used for generating the action in the action model is reduced. 10. The robot apparatus has a growth model that grows in a stepwise manner, and in the second step, a state used for generating the behavior in the behavior model in accordance with the growth of the growth model. The method according to claim 7, wherein the space is changed while being enlarged or reduced. 11. A control method for a robot apparatus having an action model that is a state transition model and generating an action based on the action model, wherein transition to a predetermined node in the action model is performed by a virtual node. A first description that describes a transition to a virtual node and allocates a predetermined node group to the virtual node
And a second step of changing the node group assigned to the virtual node. 12. The robot apparatus has a growth model that grows in stages, and in the second step, changing the node group assigned to the virtual node in accordance with the growth of the growth model. The method for controlling a robot device according to claim 11, wherein: