JP2002049424A - Device for generating traveling object operation control data and method for the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a robot for performing a commanded movement even when a surrounding environment is relatively different. SOLUTION: A generating means 15 extracts serial data whose environment specifying words and driving commands are the same from among plural serial data, and generates operation commanding serial data constituted of common abstracted group environment data having directional abstracted environment data whose values are the same among the respective directional environment data of the serial data as operation commanding serial data corresponding to the environment specifying words by abandoning any directional abstracted environment data whose values are different. Therefore, an operator can perform operation control based on the operation commanding serial data only by providing the environment specifying words.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for generating motion control data of a moving object, and more particularly to a method for generating motion control data.

[0002]

2. Description of the Related Art
Various attempts have been made to simplify the operation programming of the robot. As one of them,
Japanese Patent Application Publication No. 214645 discloses a learning method of a robot that can learn an arbitrary traveling pattern and travel according to the pattern even if various obstacles are scattered on the way. A brief description will be given. The mode is switched to the learning mode, the robot is actually operated, and the distance to the obstacle and the instructed direction from the start of the operation until the stop key is pressed are stored in the storage device. Switching to the automatic traveling mode, the data in the storage device is sequentially read, and the traveling is performed while comparing the read direction and distance with the measurement direction and the distance measured by the distance measurement unit. Such a learning method eliminates the need for programming the robot.

However, in the above control method,
There were the following problems. If the sensing content and the operation instruction are stored as they are, the surrounding environment is slightly different, and the stored sensing content does not match, and the operation cannot be reproduced. In order to solve such a problem, it is conceivable to store all data for each different environment, but this requires an enormous storage capacity.

[0004] Such a problem arises not only when the robot runs by itself, but also when the character moves in a game.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method for learning the motion of a moving object or a device for deciding the motion of a moving object, which solves the above-described problem and can simplify the motion programming of the moving object.

[0006]

Means for Solving the Problems and Effects of the Invention 1) In the moving object motion control data generating apparatus according to the present invention, in order to control the motion of the moving object with an environment specifying word used by a human, A moving object operation control data generation device that generates data for operation control, wherein B) the moving object, when a movement command to move the moving object under the environment is given from an operator b1), Driving means for moving the moving object based on a moving command, b
2) a collective environment data measuring means for measuring collective environment data which is a set of surrounding environment data for each direction of the moving object at predetermined time intervals; C) a moving object operation control data generating device; C1) first storage means for storing the collective environment data for each predetermined time measured by the collective environment data measurement means together with the motion data of the moving object; c2) representing the environment from the operator during the movement When an environment specific word is given, additional means for adding and storing the word to the collective environment data, and c3) a first rule storage for storing a correspondence rule between the mutual relationship of the direction-specific environment data and the abstract environment data of the surrounding environment. Means, c4) an abstraction collection for specifying the abstraction environment data of the surrounding environment from the environment data for each predetermined time stored in the first storage means based on the correspondence rule. Environmental data specifying means, c5)
A second storage unit for storing the abstraction set environment data for each direction; c6) an abstraction set to which the environment identification word is added among the abstraction set environment data stored in the second storage unit Environment data is determined as the focused set environment data, and together with the focused set environment data, continuous data determination means for determining a predetermined number of set environment data before and after the focused set environment data as continuous data; c7) an environment identification word among the plurality of continuous data; Extracts the same linked data, discards different-valued abstracted environment data of different values for each direction of the linked data, and creates a common abstraction set environment having the same value of different-direction abstracted environment data. It is provided with a generator for generating linked data composed of data as action command linked data corresponding to the environment specific word.

[0007] In this way, the collective environment data for each predetermined time and the motion data of the moving object are stored together with the given environment specifying word, and the surrounding environment data are collected from the collective environment data for each predetermined time based on the correspondence rule. Identifying the abstracted collective environment data of the environment, of the abstracted collective environmental data, together with the set of interest collective data, determines a predetermined number of collective environmental data before and after the set of collective environmental data as connected data, and based on the plurality of connected data, The linked data composed of the abstract set environment data is generated as the operation command linked data corresponding to the environment specifying word. Therefore, the operator can perform the operation control based on the operation command consecutive data only by giving the environment specifying word. Further, since the operation of the moving object can be controlled by the environment specifying word, programming is also simplified.

[0008] 2) In the moving object motion control data generating device according to the present invention, the generating means extracts, from the plurality of continuous data, continuous data having the same environment specifying word and operation, and extracts each of the continuous data. Regarding the direction-specific abstraction environment data, discard the different-valued direction-specific abstraction environment data, and change the continuous data composed of the common abstraction set environment data having the same value-specific direction-based abstraction environment data to the environment specific word. Generated as corresponding operation command data. Therefore, it is possible to generate operation command consecutive data that generalizes consecutive data having the same environment specifying word and the same operation.

[0009] 3) In the moving object motion control data generating apparatus according to the present invention, the generating means may discard direction-specific abstraction environment data having different values for each direction of the extracted continuous data. Then, the values of the direction-specific abstraction environment data constituting the continuous data are mirror-symmetrically exchanged with respect to the focused environment set environment data. Therefore,
If only one is detected in the target situation in the time axis direction, the continuous data can be acquired without actually operating the other.

[0010] 4) In the moving object motion control data generating apparatus according to the present invention, the generation means may discard direction-specific abstraction environment data having a different value for each direction of the extracted continuous data. The values of the abstraction environment data for each direction constituting the continuous data are mirror-symmetrically exchanged in the direction orthogonal to the traveling direction.
Therefore, in the direction orthogonal to the traveling direction, if only one is detected in the target situation, the continuous data can be acquired without actually operating the other.

5) In the moving object motion control data generating device according to the present invention, when the rotation direction of the motion is different for the plurality of generated environment data, further, for each direction of the abstracted environment data, The direction-specific abstraction environment data having different values is discarded to form continuous data composed of common abstraction set environment data having direction-specific abstraction environment data having the same value. Therefore, it is possible to obtain generalized operation command data that looks into the left and right rotation directions.

6) In the moving object motion control data generating apparatus according to the present invention, when the operator operates a trial operation of the obtained plurality of motion command sequence data and temporarily applies it, another operation command sequence data is generated. Delete abstractions that are indistinguishable from data. In this way, by creating a plurality of candidates and excluding those that do not meet the conditions, it is possible to prevent a malfunction caused by excessively generalized candidates.

[0013] 7) In the method of using the moving object motion control data generating device according to the present invention, the operator can arbitrarily move the moving object under the environment to control the motion of the moving object. Is generated. Therefore, the operator can cause the moving object to learn the operation command sequence data corresponding to the environment specific word without analyzing and moving all the patterns.

8) The moving object motion control data generating apparatus according to the present invention is characterized in that A) controls the motion of a moving object with an environment specifying word used by a human in an environment composed of a passage and a passage intersection where each passage intersects. A moving object motion control data generating apparatus for causing the moving object to learn the environment specific word, wherein the moving object moves the object under the environment by an operator. When a command is given, based on the movement command, a driving unit for moving the moving object, b2) a set distance data measurement unit for measuring set distance data that is a set of direction-specific distance data at predetermined time intervals, C) The moving object motion control data generation device: c1) collects the set distance data for each predetermined time,
First storage means for storing together with the movement data of the moving object; c2) when an environment identification word representing the environment is given by the operator at the intersection of the paths during the movement, the first information is added to the set distance data. C3) first rule storage means for storing a correspondence rule between the mutual relationship between the values of the direction-specific distance data constituting the set distance data and the abstracted environment data of the surrounding environment; c4) the corresponding rule Abstraction set environment data specifying means for specifying abstracted set environment data of the surrounding environment from the set distance data for each predetermined time stored in the first storage means, based on c. Second storage means for storing the generalized set environment data; c6) Of the abstracted collective environment data stored in the second storage means, the abstracted set environment data to which the environment specifying word is added is focused on the abstract. Conversion Combined data, and together with the noted abstracted collective environment data, connected data determining means for determining a predetermined number of abstracted collective environment data before and after the set as continuous data, c7) an environment specifying word of the plurality of connected data And extract the same run data
Regarding each direction of the environment data of the continuous data, the abstracted environment data of the different direction is discarded, and the continuous data composed of the common abstraction set environment data having the same direction of the abstracted environment data of the same value is converted to the environment. There is provided generating means for generating as continuous data corresponding to the specific word.

As described above, the set distance data for each predetermined time and the motion data of the moving object are stored together with the given environment specifying word, and the group distance data for each predetermined time are stored based on the correspondence rule. Identifying the abstracted collective environment data of the environment, of the abstracted collective environmental data, together with the set of interest collective data, determines a predetermined number of collective environmental data before and after the set of collective environmental data as connected data, and based on the plurality of connected data, The linked data composed of the abstract set environment data is generated as the operation command linked data corresponding to the environment specifying word. Therefore, the operation control based on the operation command consecutive data can be performed by the environment specifying word given by the operator. Further, since the operation of the moving object can be controlled by the environment specifying word, programming is also simplified.

In the generation of the common abstraction collective environment data, connected data having the same environment specifying word and operation is extracted from the plurality of connected data, and for each direction of the connected data, Discard the direction-specific abstraction environment data of different values and generate continuous data composed of common abstraction set environment data having the same value of direction-specific abstraction environment data as operation command continuous data corresponding to the environment specific word. I do. Therefore, it is possible to generate operation command consecutive data that generalizes consecutive data having the same environment specifying word and the same operation.

9) In the moving object motion control data generating apparatus according to the present invention, the generating means discards different-direction abstraction environment data having different values for each direction of the extracted continuous data. In this case, the values of the direction-specific abstract environment data constituting the continuous data are mirror-symmetrically exchanged around the focused abstract set environment data.

The values are mirror-symmetrically exchanged around the focused environment set environment data. Therefore, in the time axis direction, if only one is detected in the target situation, the continuous data can be acquired without actually operating the other.

10) In the moving object motion control data generating apparatus according to the present invention, the generation means may discard direction-specific abstraction environment data having a different value for each direction of the extracted continuous data. The values of the abstraction environment data for each direction constituting the continuous data are mirror-symmetrically exchanged in the direction orthogonal to the traveling direction. Therefore, in the direction orthogonal to the traveling direction, if only one is detected in the target situation, the continuous data can be acquired without actually operating the other.

11) A moving object motion control data generating method according to the present invention is a moving object motion generating method for generating motion control data of the moving object in order to control the motion of the moving object with an environment specific word used by a human. The control data generation method, 1) while moving the moving object, measuring set environment data that is a set of surrounding environment data by direction of the moving object at predetermined time intervals, 2) set at every predetermined time Environment data is stored together with the movement data of the moving object.3) When an environment specifying word representing the environment is given by the operator during the movement, the environment data is added to the collective environment data and stored.
Based on a correspondence rule between the previously stored direction-specific environment data and the surrounding environment abstraction environment data,
From the stored collective environment data for each predetermined time,
Identify and store abstract environment data of the surrounding environment, 5)
Among the stored abstract collective environment data, the abstract collective environment data to which the environment specifying word is added is determined as the collective collective environment data. Is determined as connected data.6) Among the plurality of connected data, the connected data having the same environment specific word is extracted. Is discarded, and linked data composed of common abstraction set environment data having the same value of direction-specific abstraction environment data is generated as action command linked data corresponding to the environment specific word.

As described above, the collective environment data for each predetermined time and the motion data of the moving object are stored together with the given environment specifying word, and the surrounding environment data are collected from the collective environment data for each predetermined time based on the correspondence rule. Identifying the abstracted collective environment data of the environment, of the abstracted collective environmental data, together with the set of interest collective data, determines a predetermined number of collective environmental data before and after the set of collective environmental data as connected data, and based on the plurality of connected data, The linked data composed of the abstract set environment data is generated as the operation command linked data corresponding to the environment specifying word. Therefore, the operation control based on the operation command consecutive data can be performed by the environment specifying word given by the operator. Further, since the operation of the moving object can be controlled by the environment specifying word, programming is also simplified.

The relationship between the terms used in the claims and the embodiments will be described below.

"Environment specifying word": A word indicating the environment around a moving object when it moves. In the embodiment, it corresponds to an intersection, a three-way intersection, or the like.

"Aggregate environment data": A set of surrounding environment data for each direction. In the embodiment, left, front left, front, front right,
A set of surrounding environment data by direction of the five sensors on the right.
This is a concept that includes direction-specific surrounding environment data.

"Directional environment data": In the embodiment, the distance (40) of each sensor at each time shown in FIG. 11 corresponds to the concept including the direction-specific distance data.

"Motion data": Data indicating the moving state of a moving object. In the embodiment, the data corresponds to data based on a driving command given to the moving object by an operator.

"Abstract environment data": In the embodiment, "abstract in the direction" in the table shown in FIG.

"Abstract set environment data": In the embodiment, "surface" and "empty" at each time shown in FIGS.
And so on.

"Aggregating environment data": a set of surrounding environment data for each direction. In the embodiment, left, front left, front, front right,
It refers to a set of ambient environment data for each of the five sensors on the right.

"Attention set environment data": refers to abstract set environment data to which an environment specifying word given by the operator is added.

"Consecutive data": A set of collective environment data composed of target collective environment data and a predetermined number of collective environment data before and after the target collective environment data.

"Mirror symmetry exchange": In the embodiment, this corresponds to mirror symmetry exchange processing in the forward direction and / or mirror symmetry exchange processing in the time axis direction.

"Distance data by direction": In the embodiment, the distance (40) of each sensor at each time shown in FIG.

"Gathering distance data": A set of distance data for each direction.

"Path intersection": In the embodiment, this means a point at which paths intersect (be connected), such as an intersection, a three-way junction, a corner, or the like.

[0036]

BEST MODE FOR CARRYING OUT THE INVENTION Description of Functional Block Diagram One embodiment of the present invention will be described with reference to the drawings. The moving object motion control data generation device 10 shown in FIG. 1 is designed to operate under an environment composed of a passage and a passage intersection where each passage intersects.
An apparatus for causing the moving object 1 to learn the environment specifying word in order to control the movement of the moving object with the environment specifying word used by a human, the first storage unit 3, the addition unit 5, the first rule storage unit 7 The moving object 1 will be described first. The moving object 1 is provided with the abstract set environment data specifying means 9, the second storage means 11, the continuous data determining means 13, and the generating means 15.
The moving object 1 is a set of driving means 4 for moving itself based on the moving command when the moving command for moving the moving object 1 in the environment is given by the operator, and distance data for each direction at predetermined time intervals. The apparatus includes a collective distance data measuring unit 2 for measuring collective distance data.

The first storage means 3 stores the set distance data for each predetermined time together with the motion data of the moving object. When an environment specifying word representing the environment is given by the operator at the intersection of the paths during the movement, the adding unit 5 adds the environment specifying word to the collective distance data and stores the data.
The first rule storage means 7 stores a correspondence rule between the interrelationship between the values of the direction-specific distance data constituting the set distance data and the abstracted environment data of the surrounding environment. The abstract set environment data specifying unit 9 specifies the abstract set environment data of the surrounding environment from the set distance data for each predetermined time stored in the first storage unit 3 based on the correspondence rule. Second
Storage means 11 stores the above-mentioned abstracted collective environment data for each direction. The continuous data determination unit 13 determines, as the focused abstract set environment data, the abstract set environment data to which the environment specifying word is added, from the abstract set environment data stored in the second storage unit 11, Along with the noted abstract set environment data, a predetermined number of abstract set environment data before and after the set are determined as continuous data. Generating means 15
Extracting, from among the plurality of linked data, linked data having the same environment specifying word and operation, and discarding different-direction abstraction-based environment data having different values for each direction of the linked environment data to obtain the same value of the same value. Continuous data composed of common abstraction set environment data having direction-specific abstraction environment data is generated as continuous data corresponding to the environment specific word.

In this manner, the set distance data for each predetermined time and the motion data of the moving object are stored together with the given environment specifying word, and the group distance data for each predetermined time are stored based on the correspondence rule. Identifying the abstracted collective environment data of the environment, of the abstracted collective environmental data, together with the set of interest collective data, determines a predetermined number of collective environmental data before and after the set of collective environmental data as connected data, and based on the plurality of connected data, The linked data composed of the abstract set environment data is generated as the operation command linked data corresponding to the environment specifying word. Therefore, the operation control based on the operation command data can be performed only by the operator giving the environment specifying word. Further, since the operation of the moving object can be controlled by the environment specifying word, programming is also simplified.

In the generation of the common abstraction set environment data, connected data having the same environment specifying word and operation is extracted from the plurality of connected data. Discard the direction-specific abstraction environment data of different values and generate continuous data composed of common abstraction set environment data having the same value of direction-specific abstraction environment data as operation command continuous data corresponding to the environment specific word. I do. Therefore, it is possible to generate operation command consecutive data that generalizes consecutive data having the same environment specifying word and the same operation.

2. Hardware Configuration An operation plan determination robot (hereinafter abbreviated as a robot) 20 including the operation control data generation device 10 shown in FIG. 1 will be described.

As shown in FIG. 2, the robot 20 can run by a pair of front wheels 47 and a pair of rear wheels 48, and five sensors 40s1 to 40s5 are provided on the side wall. As shown in FIG. 3, the five sensors can measure the distance to the obstacle over the entire circumference by the sensors. The robot 20 receives a driving command from the remote controller via the antenna 21 and moves on its own based on the driving command.

The hardware configuration of the robot 20 will be described with reference to FIG. The robot 20 includes a CPU 23,
RAM 27, keyboard 28, interface 31,
A communication board 28 and a bus line 29 are provided. C
The PU 23 controls each unit via the bus line 29 according to a control program stored in the ROM 25.

The control program may be read from a recording medium such as a PC card and written into the ROM 25. In addition to the PC card, a flexible disk, a CD-ROM, a computer-readable recording medium in which a program such as an IC card is substantially integrated,
You may make it install in ROM25. Furthermore, you may make it download using a communication line.

The start command, the end command, and various drive commands given by the operator are given to the CPU 23 through the communication board 28. The interface 31 is connected to five sensors 40s1 to 40s5.

The interface 31 is connected to a motor drive unit 37 for driving and controlling the rear wheel drive motor M and a motor drive unit 38 for performing handling control. The motor drive unit 37 controls the rear wheels 47 and the motor drive unit 38 controls the front wheels 48 shown in FIG.

The RAM 27 also stores various calculation results and the like.

As shown in FIG. 5, the ROM 25 stores a correspondence table between environment data and abstracted environment data. With such a correspondence table, the robot 20
The surrounding environment can be grasped according to the sensor values extracted from the two sensors. In this sense, the environment data is a condition (specific numerical value), and the abstracted environment data is a result (data representing the detected environment).

3. Flowchart Next, a program stored in the ROM 25 will be described. The ROM 25 stores a plurality of programs shown in FIGS.

FIG. 7 is an overall flowchart. The operator gives an operation control command from a remote controller (not shown) so that the robot 20 moves along the locus 120 shown in FIG. Such an operation control command is received by the communication board 28 shown in FIG. CPU
23 performs a driving process based on the driving process program shown in FIG. Specifically, the CPU 23 determines whether or not a drive command as a movement command is given (step S11 in FIG. 8). When the drive command is given, the CPU 23 sends a signal for executing the drive command to the motor drive unit 37. , 39 (step S13). Thus, the motors M1 and M2 are driven, and the robot 20 moves along the locus 120 shown in FIG. The given drive command is temporarily stored in the RAM 27 (step S15).

When moving, the CPU 23 performs a sensor value detection process (step S1 in FIG. 7). Details of the detection processing will be described with reference to FIG. CPU23
Determines whether a predetermined time (in this case, Δt seconds) has elapsed (step S21 in FIG. 9), and stores the detection values of the five sensors 40s1 to 40s5 when the predetermined time has elapsed (step S23). The CPU 23 stores the data together with the current drive command in the RAM 27 as aggregate data (step S25). Thereby, the value of each sensor and the movement at that time at time t1 shown in FIG. 11A are stored.
By repeating the processing of steps S21 to S25, the group distance data at predetermined time intervals is stored in the plurality of RAMs 27, as shown in FIG. 11A.

Since each sensor is provided on the body (side wall) of the robot 20, it rotates together with the rotation of the robot 20. Therefore, each sensor value is not detected while the robot 20 is rotating. Also, in FIG. 11A, the sensor values before and after the rotation have completely different values. Thus, by detecting a numerical value in the coordinate system in which the traveling direction of the robot is the Y direction, it is possible to extract environmental data relative to the traveling direction of the robot.

While the robot 20 is moving, the operator inputs environment specific words and action specific words to be learned at appropriate timing. For example, the remote controller (not shown) is given to the robot 20 such as "passage / forward" at time t1, "intersection / right turn" at time t7, "passage / forward" at time t10. .

During the movement of the robot 20, the program shown in FIG. 10 is executed in addition to the program shown in FIG. The CPU 23 determines whether or not the environment specific word and the operation specific word are provided by the operator (step S31). When the environment specific word and the operation specific word are provided by the operator, the collective distance is determined at that timing. Append to data.

As a result, as shown in FIG. 11B, set distance data to which the environment specifying word and the action specifying word are added is generated.

Next, the CPU 23 generates abstracted environment data of the surrounding environment based on the set distance data (step S3 in FIG. 7). The abstract environment data generation process will be briefly described. As shown in FIG. 11A, five directional distance data are measured at each time. This is called collective distance data. About such collective distance data,
Using the correspondence table shown in FIG. 5, direction-specific environment data is generated from the direction-specific distance data. For example, time t1, t
In 2, the sensor value is "40" in the left direction,
In addition, the motions are both “forward”. Therefore, it is determined that there is a surface according to Rule 3. The same applies to the left front direction, the right front direction, and the right direction.

Since the sensor value in the forward direction is infinite, it is determined that the area is "empty". Also, from time t2 to time
At t3, the sensor value in the forward direction is significantly different from “infinity” to “85”. Therefore, it is determined by rule 4 that there is an end between them. The same applies to the left front direction.

In the present embodiment, since the distance in five directions is measured, there is no case where rules 1 and 2 can be applied. However, the distance in eight directions such as left rear, right rear, and back is measured. Is measured, the abstract environment data can be generated by applying rules 1 and 2. Further, sensors may be provided in more detailed directions, for example, 12 directions, 16 directions, and the like, and the values may be measured.

Thus, the abstract environment data as shown in FIGS. 12 to 16 is generated.

Next, the CPU 23 proceeds to step S5 in FIG.
The generation processing of the abstract connected data shown in FIG. The generation process of the abstract connected data will be described with reference to FIG. CP
U23 determines continuous data (step S4 in FIG. 17).
1). The continuous data refers to a set of a predetermined number of pieces of abstract set environment data before and after the set of abstract set environment data to which environment specific words such as “passage”, “intersection”, etc. are added. In the present embodiment, the abstract set environment data that is the center of the continuous data is referred to as the set environment data of interest. It was decided as one continuous data. As a result, a plurality of continuous data are determined based on the abstracted collective environment data shown in FIGS.

Next, the CPU 23 groups continuous data having the same environment specifying word and driving command (step S43 in FIG. 17). For example, the consecutive data that is the environment specific word “three-way intersection” driving instruction “clockwise” is consecutive data from t5 to t9 where t7 is the set environment data of interest, and t21 to t26 where t23 is the set environment data of interest. is there. Therefore, these are one group. Similarly, consecutive data of the environment specific word “passage” driving command “forward” are t10, t18, t29, t39, t47, t55, t64, t73, t82, t90, t100.
Is set as the attention set environment data. The consecutive data that is the environment specific word “intersection” driving instruction “counterclockwise”
t14 and t60. The consecutive data that is the environment specific word “intersection” driving instruction “clockwise” is t51 and t86. The consecutive data that is the environment specific word “three-way intersection” driving instruction “counterclockwise” is t7
7, t96. The consecutive data of the environment specific word “turning angle” driving instruction “clockwise” is t43. The consecutive data corresponding to the word “turning angle” driving instruction “counterclockwise” is t69. The linked data of the environment specific word “intersection” driving command “forward” is t35. The linked data that is the environment specific word “three-way intersection” drive command “straight ahead” is t105.

Next, the group number i to be processed is initialized (step S45 in FIG. 17), and generalization processing is performed for the i-th group (step S47). The generalization process refers to abstracting environmental data for each direction of continuous data belonging to each group. For example, in the group in which the environment specific word “intersection” driving command is “clockwise”, both are superimposed on the focused set environment data t51 and t87. In this case, as shown in FIGS. 18A and 18B, at t49 and t85, the left, front, right front, and right coincide, at t50 and t86, all of the left, front left, and right coincide. Front left, ~
Everything on the right matches, and on t53 and t89, everything on the left, front left, and right matches. Therefore, the environment specific word "intersection"
The group of the driving commands “clockwise” is generalized as shown in FIG. 18C.

Next, the CPU 23 determines whether or not the generalization processing has been completed for all groups (step S4).
9). If the generalization processing has not been completed for all groups, the processing target group number i is incremented by the CPU.
(Step S51), the process of step S47 is repeated.

As described above, the pair of the environment specifying word and the driving command is “tri-junction / clockwise”, “tri-junction / counterclockwise”,
Generated for each of "Trifurcation / straight", "passage / forward", "intersection, counterclockwise", "intersection / clockwise", "intersection / forward", "turn corner / clockwise", and "turn corner / counterclockwise". You. FIG. 19 shows a state in which a pair of an environment specifying word and a driving command is generated as “intersection / counterclockwise” from two consecutive data. In FIG. 20, the pair of the environment specifying word and the driving command is “tri-junction / clockwise” from the two consecutive data,
FIG. 21 shows a case in which the pair of the environment specifying word and the driving command is “triple-junction / counterclockwise” and “intersection / counterclockwise” from two consecutive data. FIG. 22 shows the abstract link data in which the pair of the environment specifying word and the driving command is “passage / straight”.

In this way, the abstract connected data is generated for each environment specifying word having a different driving instruction.

When the robot 20 stores the abstracted linked data as action command linked data in the robot 20 and gives the robot 20 an action specifying word along with an environment specifying word, the robot 20 in FIG. As for the trajectory, the robot 20 can cause the robot 20 to perform the movement intended by the operator.

For example, in FIG. 23A, the robot 20 is placed at
If the operator wants to move the
Give commands such as "Third junction / right turn" and "passage / straight".
The robot 20 is “passage / straight”, “triple-junction / right-turn”
Read out each operation command sequence data. The robot 20 periodically detects environmental data, determines whether the detected environmental data matches any one of the current operation command data, and if so, as it is, The drive command based on the operation command data is transmitted to the motor drive unit 37,
38. On the other hand, if the detected environment data does not match the current operation command data, it may be determined whether or not it matches the next operation command data. As a result, control is sequentially performed according to the operation command data.

For example, the first operation command data in FIG. 22 (left: null, front left: plane, front: empty, right: plane, front right: null)
l, operation: forward), the control information of the motor, “forward”, is read, and a control command is given to the motor drive units 37, 38.
When the robot starts to move, the CPU 23 determines that the sensor value after Δt seconds is the current operation command data (first operation command data).
It is determined whether or not they match. If they match, continue the current action "forward". On the other hand, if they do not match, the second operation command data (left: null, front left: plane, front: null, right: plane,
Right front: null, action: forward). If the two operation command data do not match, 3
Judge whether it matches the second operation command data ...
The operation is controlled in accordance with the arranged operation command data.

The operation command data is composed of a plurality of operation command data, and the operation command data at both ends (for example, +2 and -2) in each operation command data has a degree of generalization. high. Therefore, even if it should have originally shifted to the control based on the next operation command data, the shift may not have occurred. In such a case, it is sufficient to shift to the next operation command data as described below. For example, as shown in FIG.
In the state where the operation is controlled in step 1, when the control is switched to the operation command data R2 from the middle, and the moving object moves from the environment Cn to Cn + 5, the operation command data is Can be switched.

In the detection environment Cn + 3, the operation command data R
It is assumed that the operation command data coincides with the operation command data at -1 of -2, and then coincides with the operation command data of -1 of the operation command series data R1 in the detection environment Cn + 4. In this situation, when the operation proceeds to the detection environment Cn + 5, if the operation instruction data does not match the operation instruction data at 0 in the operation instruction data R1, whether the operation instruction data at the next operation instruction data R2 coincides with 0 is determined. Judge. If they match, it is determined whether the operation command data at -2 and -1 in the operation command data match the detection environments Cn + 4 and Cn + 3. May be switched to the operation control based on the operation command data R2.

It should be noted that the detection environment may be such that the time axis compression described later is performed even after the elapse of the time Δt.

As described above, in this embodiment,
The operation command sequence data is called in order, and it is determined whether the surrounding environment changes and the control is shifted to the next operation command sequence data. Therefore, the operator merely gives an environment specific word for each operation in the order in which control is desired, and the operation is controlled by the operation command linked data for the environment specific word of the robot.

4. Other Embodiments In the above-described embodiment, when the robot is moved under a certain environment and the extracted environment data has the same environment specifying word and driving command, generalization is performed to put them together. However, if the generalization is not appropriate, a malfunction may occur. That is, if the generalization is excessively generalized, the conditions will be matched even with other operation instruction linked data. Also, in the above embodiment, FIG.
In the environment described above, all the patterns may not be detected as targets for each group because they are arbitrarily moved.

Therefore, as described below, a plurality of candidates which have been subjected to mirror symmetry replacement processing with respect to the direction and / or the time axis are generated and actually moved to eliminate inappropriate candidates. Appropriate generalizations can be made.

For example, in the case of a three-way intersection, there are a right turn as shown in FIG. 24A and a right turn as shown in FIG. In this case, the linked data of both is data as shown in FIGS. 25A and 25B, and if generalized as it is, it becomes operation command linked data as shown in FIG. 25C. On the other hand, if any of them is subjected to mirror symmetry replacement processing in the forward direction to generate a new candidate and generalize it, operation command data as shown in FIG. 26C is generated. In this case, the data of Example 2 (the data of FIG. 25B) is replaced with the left and right data, and the left and right data are replaced.

When any one of them is subjected to mirror object replacement processing in the time axis direction and generalized, it becomes operation command consecutive data as shown in FIG. 27C. In this case, the data of Example 2 (the data of FIG. 25B) is set to t and t +
4, the data of t + 1 and t3 are replaced respectively.

Further, both the mirror object replacement processing for the time axis and the forward direction may be performed. This case is shown in FIG. 28C.

In the case of going straight on a three-way junction, FIG.
A, the case shown in FIG. 29B is conceivable. In this case, the continuous data of both is data as shown in FIGS. 30A and 30B, respectively.
The operation command data is as shown in FIG. On the other hand, similarly, mirror symmetric exchange processing in the forward direction,
The case where the mirror symmetric exchange process in the time axis direction and the case where the forward and mirror symmetric exchange processes in the time axis direction are performed are operation instruction linked data as shown in FIGS. 31C, 32C, and 33C, respectively.

The generated abstract data is stored as a candidate. Then, while moving the robot 20, among the candidates, candidates that malfunction with the conditions of the other groups are deleted assuming that excessive generalization has been performed, and each group is divided into small groups. It may be stored, and in an actual application, any one that satisfies may be selected. In this way, even if the combination of the environment specific word and the driving instruction is the same, instead of generalizing them all together, they are divided into a plurality of small groups and stored, and in the actual application stage, By specifying a suitable candidate, a malfunction due to excessive generalization can be avoided.

For example, assuming that the environment specific word and the driving command are the same, there are three candidates as “three-way intersection, right turn”, and by performing mirror symmetry replacement processing in the forward direction and mirror symmetry replacement processing in the time axis direction. , And seven candidates are obtained. If these seven are further memorized as a small group and tested,
If two of the small groups malfunction with other groups (for example, "intersection / turn right"),
This is deleted, and the remaining five small groups are stored as candidates for “three-way intersection, right turn”. If the operator actually gives an operation command such as “passage / forward” and “triple / right turn”, it is determined whether or not the current environment matches the data of the five small groups. I just need.

In the above embodiment, the generalization is made when the driving command and the environment specifying word are the same.
If the environment specific words are the same, a generalized candidate can be generated by ignoring the direction of rotation and setting clockwise and counterclockwise rotations. FIG. 34 shows such an example. In FIG. 34, "intersection / clockwise" shown in FIG. 34A and "intersection / counterclockwise" shown in FIG. 34B are generalized to "intersection / rotation" as shown in FIG. 34C except for the directionality. FIG.
An example in which a “three-way intersection / rotation” is generated from a “three-way intersection / clockwise” and a “three-way intersection / leftward” will be described. With regard to such generalized generalization, rules such as generalizing clockwise and counterclockwise as rotations may be stored in advance.

Further, even if the driving command and the environment specifying word are not the same, generalization may be performed to put together the same environment specifying word.

It should be noted that a plurality of candidates having different degrees of generalization are generated and inspected. If excessive generalization is performed, this is eliminated and the plurality of candidates are stored. By controlling the operation with any of the operation command data,
The desired movement can be performed by the robot.

It is also conceivable that the timing for inputting words is shifted. In such a case, the position of the focused set environment data may be shifted back and forth in the time direction, and the best match may be obtained.

The mirror symmetric exchange processing in the forward direction and the mirror symmetric exchange processing in the time axis direction have been described for two cases, but the same applies when three or more are combined. For example, in the case where there are three extracted consecutive data (continuous data 1, 2, 3), two are used to determine whether there is no mirror symmetry replacement processing and whether there is mirror symmetry replacement processing (forward direction and / or time axis direction). Generalization in the case where the combination is carried out, and in addition, when all three are combined without mirror symmetry exchange processing and with mirror symmetry exchange processing (forward direction and / or time axis direction), generalization is performed. Should be performed.

In the present embodiment, since the distance is used as the sensor value, the generalization process corresponds to a process of grasping the shape of the surrounding environment based on the value of each sensor and a predetermined rule. However, the present invention is not limited to this, and includes all processes of abstracting (generalizing) environment set data based on a predetermined rule. Fig. 3
Means data of combinations of environments in all directions based on the five sensors shown in FIG.

In the above embodiment, the environmental data is generated from the data extracted from the sensor. However, the data extracted from the sensor may be compressed in the time axis direction as follows. This is because if the time interval detected by the sensor is quite short compared to changes in the surrounding environment,
This is because, as in the above embodiment, the front two and the rear two of the focused set environment data may be equal.

Even if the data is compressed in the time axis direction as described above, there is no problem because each data has the meaning in the time series direction. Conversely, a desired operation can be performed regardless of the moving speed of the robot.

As for the end recognition, it may be determined whether the end is approaching or moving away. Thereby, the shape of the surrounding environment can be grasped more accurately. For example, between time tx and tx + 1, the sensor value in a certain direction changes from “infinity” to “100”.
When it becomes, it can be determined that the end is approaching. Also,
When the sensor value in a certain direction changes from “45” to “100” between the times tx and tx + 1, it can be determined that there is an end whose distance is from a short distance to a long distance. In the opposite case, for example, when the sensor value changes from “100” to “45”, it can be determined that there is an end from a far distance to a near distance.

In the above embodiment, two-dimensional distance data is employed, but the same can be applied to three-dimensional distance data. For example, the height of an object constituting the environment may be adopted. For example, if the height is less than a certain value (for example, 5 mm), it is possible to learn that the user rides up and moves, and if the height is equal to or more than the value, he or she detours. Further, an uphill or a downhill may be input as an environment specifying word.

Further, in the above embodiment, the case where the distance from the environment around the robot is used as the environment data has been described. A thing may be used.

Also, a case has been described where only the user himself / herself does not move, but when the opponent moves, different correction rules such as the moving speed of the opponent, the stored energy, and the shape of the object are stored. It should be corrected.

Further, in this embodiment, the case where the present invention is applied to a robot has been described, but the present invention can be similarly applied to a computer game. For example, in a game using a computer, the present invention can be applied to estimating a rule of movement of an object from a past movement of the object, and competing for a score to destroy the object in advance.

In the present embodiment, the functions shown in FIG. 1 are realized by using the CPU 23 and software. However, some or all of them may be realized by hardware such as a logic circuit.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a moving object motion control data generation device 10 according to the present invention.

FIG. 2 is an external view of a robot 20 including the moving object motion control data generation device shown in FIG.

FIG. 3 is a diagram showing detection directions of five sensors.

FIG. 4 is a diagram illustrating an example of a hardware configuration of the robot 20.

FIG. 5 shows a table for generating direction-specific environment data from direction-specific distance data.

FIG. 6 is a diagram showing a trajectory for operating the robot 20.

FIG. 7 is an overall flowchart.

FIG. 8 is a flowchart of a driving process.

FIG. 9 is a flowchart of a sensor value detection process.

FIG. 10 is a flowchart of an environment specific word input process.

FIG. 11 shows an example of a detected sensor value.

FIG. 12 shows abstracted environment data generated from the sensor values.

FIG. 13 shows abstraction environment data generated from the sensor values.

FIG. 14 shows abstraction environment data generated from the sensor values.

FIG. 15 shows abstracted environment data generated from the sensor values.

FIG. 16 shows abstracted environment data generated from the sensor values.

FIG. 17 is a flowchart of a process of generating abstract continuous data.

FIG. 18 is a diagram for describing a generalization process in each group.

FIG. 19 is a diagram for describing a generalization process in each group.

FIG. 20 is a diagram for explaining a generalization process in each group.

FIG. 21 is a diagram for describing a generalization process in each group.

FIG. 22 shows operation command data in a passage / forward movement.

FIG. 23 shows an environment in which the robot 20 operates.

FIG. 24 shows an operation example of a different pattern on a three-way intersection.

FIG. 25 shows data obtained from the sensor values detected in FIG.

FIG. 26 shows data when the sensor values detected in FIG. 24 are subjected to mirror symmetric exchange processing in the left-right direction.

FIG. 27 shows data in a case where the sensor values detected in FIG. 24 are subjected to mirror symmetry replacement processing in the time axis direction.

FIG. 28 shows data when the sensor values detected in FIG. 24 are subjected to mirror symmetric replacement processing in the left-right direction and the time axis direction.

FIG. 29 shows an operation example of a different pattern on a three-way intersection.

FIG. 30 shows data obtained from the sensor values detected in FIG.

FIG. 31 shows data when the sensor values detected in FIG. 29 are subjected to mirror symmetry replacement processing in the left-right direction.

FIG. 32 shows data in a case where the sensor values detected in FIG. 29 are subjected to mirror symmetric exchange processing in the time axis direction.

FIG. 33 shows data obtained when the sensor values detected in FIG. 29 are subjected to mirror symmetric replacement processing in the left-right direction and the time axis direction.

FIG. 34 shows data when the clockwise and counterclockwise rotations are collectively rotated on a three-way intersection.

FIG. 35 shows data when the clockwise and counterclockwise rotations are collectively rotated at an intersection.

[Explanation of symbols]

 23 CPU 27 Memory 40s1-40s5 Sensor

Claims (13)

[Claims] A) A moving object motion control data generating device for generating data for motion control of a moving object in order to control the motion of the moving object with an environment specifying word used by a human, The moving object, b1) When a moving command to move the moving object under the environment is given from the operator, based on the moving command,
Driving means for moving the moving object; b2) collective environment data measuring means for measuring collective environment data which is a set of surrounding environment data for each direction of the moving object at predetermined time intervals; C) controlling the moving object operation A data generating device, c1) first storage means for storing the collective environment data for each predetermined time measured by the collective environment data measuring means together with the motion data of the moving object, c2) from the operator during the movement When an environment specifying word representing the environment is given, the adding unit stores the word in addition to the collective environment data.c3) Stores a correspondence rule between the mutual relationship between the direction-specific environment data and the abstract environment data of the surrounding environment. First rule storage means, c4) based on the correspondence rules, extract the aggregated environment data of the surrounding environment from the aggregated environment data for each predetermined time stored in the first storage means. Abstract aggregate environmental data specifying means for, c5) a second for storing the direction-specific abstraction aggregate environmental data
C6) Among the abstracted collective environment data stored in the second storage means, the abstract collective environment data to which the environment specifying word is added is determined as the collective collective environment data, Continuous data determining means for determining a predetermined number of collective environmental data before and after the environmental data as continuous data, c7) extracting, from the plurality of continuous data, continuous data having the same environment specific word, in each direction of the continuous data; Regarding the different abstraction environment data, discard the direction-specific abstraction environment data of different values, and correspond the continuous data composed of the common abstraction set environment data having the same value of the direction-specific abstraction environment data to the environment specific word. A moving object motion control data generating device, comprising: generating means for generating motion command sequence data to be executed. 2. The moving object motion control data generating device according to claim 1, wherein the generating means extracts, from the plurality of data, continuous data having the same environment specifying word and motion, and each direction of the continuous data. Regarding the different abstraction environment data, discard the direction-specific abstraction environment data of different values, and correspond the continuous data composed of the common abstraction set environment data having the same value of the direction-specific abstraction environment data to the environment specific word. Generated as a series of operation command data. 3. The moving object motion control data generating apparatus according to claim 2, wherein said generating means discards direction-specific abstraction environment data having a different value for each direction of said extracted continuous data. And performing mirror-symmetric exchange of the values of the direction-specific abstraction environment data constituting the continuous data with the focus on the environment of interest set environment data. 4. The moving object motion control data generating apparatus according to claim 2, wherein said generating means generates direction-specific abstraction environment data having different values for each direction of said extracted continuous data. When discarding, for each direction of the abstracted environment data constituting the continuous data, the value thereof is mirror-symmetrically exchanged in a direction orthogonal to the traveling direction. 5. The moving object motion control data generation device according to claim 2, wherein, when the rotation direction of the motion is different for the plurality of generated environment data, further, the abstraction environment data for each direction is different. Discarding the abstraction environment data for each direction of the value to obtain continuous data composed of common abstraction set environment data having the abstraction environment data for the same direction of value. 6. The moving object motion control data generation device according to claim 2, wherein the operator operates a trial operation of the obtained plurality of motion command sequence data, and temporarily applies the operation command sequence data to another operation command sequence data. It is characterized by deleting abstracted objects so that they cannot be distinguished from. 7. The method of using the moving object motion control data generation device according to claim 2, wherein the operator controls the motion of the moving object by arbitrarily moving the moving object under the environment. Generating data. 8. A) In an environment composed of a passage and a passage intersection where each passage intersects, in order to control the movement of a moving object with an environment specifying word used by a human, the environment specifying word is added to the moving object. A moving object motion control data generation device to be learned, wherein: B) the moving object: b1) when an operator issues a moving command to move the moving object under the environment, based on the moving command,
Driving means for moving the moving object; b2) collective distance data measuring means for measuring collective distance data, which is a set of distance data for each direction at predetermined time intervals; C) the moving object operation control data generation device, c1) first storage means for storing the set distance data for each predetermined time together with the motion data of the moving object; c2) an environment identification word representing the environment from the operator at the intersection of the paths during the movement. Is given, the adding means for adding and storing the aggregate distance data, and c3) storing the correlation between the mutual relationship between the values of the direction-specific distance data constituting the aggregate distance data and the abstract environment data of the surrounding environment. C4) Identifying abstract set environment data of the surrounding environment from the set distance data for each predetermined time stored in the first storage means, based on the corresponding rule. Abstract aggregate environmental data specifying means, c5) a second for storing the direction-specific abstraction aggregate environmental data
C6) Among the abstraction set environment data stored in the second storage means, the abstraction set environment data to which the environment specifying word is added is determined as the attention abstraction set environment data, Continuous data determination means for determining a predetermined number of abstracted collective environment data before and after the abstracted collective environment data of interest as continuous data, c7) extracting continuous data having the same environment specifying word and operation from the plurality of continuous data Then, for each direction of the environment data of the continuous data, discarding the direction-specific abstraction environment data of different values, and connecting the continuous data composed of the common abstraction set environment data having the same value of the direction-specific abstraction environment data. Generating means for generating continuous data corresponding to the environment specific word. 9. The moving object motion control data generating apparatus according to claim 8, wherein said generating means discards direction-specific abstraction environment data having a different value for each direction of said extracted continuous data. And performing a mirror-symmetric exchange of the values of the direction-specific abstraction environment data constituting the continuous data with respect to the focused abstraction set environment data. 10. The moving object motion control data generation device according to claim 8, wherein the generation means includes a direction-based abstraction environment having a different value for each direction of the extracted continuous data. When discarding the data, the values of the abstraction environment data for each direction constituting the continuous data are mirror-symmetrically exchanged in a direction orthogonal to the traveling direction. 11. A moving object motion control data generating method for generating data for motion control of a moving object in order to control the motion of the moving object with an environment specific word used by a human, comprising: At the same time, collective environment data, which is a set of surrounding environment data for each direction of the moving object, is measured every predetermined time, and the collective environment data for each predetermined time is stored together with the operation data of the moving object. When an environment specifying word representing the environment is given by the operator, the word is added to the collective environment data and stored, and a correspondence rule between the pre-stored interrelationship of the environment data for each direction and the abstract environment data of the surrounding environment On the basis of the,
From the stored collective environment data for each predetermined time,
Identifying and storing the abstraction set environment data of the surrounding environment, and among the stored abstraction set environment data, determining the abstraction set environment data to which the environment identification word is added as the set data set of interest, Along with the noted set environment data, a predetermined number of set environment data before and after the set data are determined as connected data, and among the plurality of connected data, connected data having the same environment specific word is extracted, and abstraction for each direction of the connected data is performed. For the environment data, discard the direction-specific abstraction environment data having different values, and execute the operation command corresponding to the environment specific word by connecting the continuous data composed of the common abstraction set environment data having the direction-specific abstraction environment data having the same value. A method for generating moving object motion control data, characterized by generating as continuous data. 12. A moving object which generates data for controlling the movement of a moving object in order to control the movement of the moving object with an environment specific word used by a human, in a computer provided with an input device, a control device, and a storage device. A recording medium recording a program for functioning as an operation control data generation device, wherein the program is a program for causing the computer to execute the following processing, wherein the moving object is provided by an operator. Based on the received movement command, collective environment data, which is a set of direction-specific peripheral environment data of the moving object, is stored at predetermined time intervals when the mobile object is moved, together with the movement data of the moving object, When an environment specifying word representing the environment is given by the operator, the word is added to the collective environment data and stored, and the previously stored Based on the correspondence between the direction-specific environmental data and the surrounding environment abstraction environment data,
From the stored collective environment data for each predetermined time,
Identifying and storing the abstraction set environment data of the surrounding environment, and among the stored abstraction set environment data, determining the abstraction set environment data to which the environment identification word is added as the set data set of interest, Along with the noted set environment data, a predetermined number of set environment data before and after the set data are determined as connected data, and among the plurality of connected data, connected data having the same environment specific word is extracted, and abstraction for each direction of the connected data is performed. For the environment data, discard the direction-specific abstraction environment data having different values, and execute the operation command corresponding to the environment specific word by connecting the continuous data composed of the common abstraction set environment data having the direction-specific abstraction environment data having the same value. A recording medium recording a program characterized by generating as continuous data. 13. A recording medium generated by a computer and recording data for controlling the movement of the moving object in order to control the movement of the moving object using an environment specific word used by a human, wherein the movement of the moving object is controlled. The control data is generated by the following method. The moving object is based on a moving command given by an operator. A set of environment data, which is a set of environment data, is stored together with the movement data of the moving object.If an environment specifying word representing the environment is given by the operator during the movement, the set environment data is added to the set environment data and stored. , Based on the correspondence rules between the previously stored inter-direction environment data and the surrounding environment abstraction environment data,
From the stored collective environment data for each predetermined time,
Identifying and storing the abstraction set environment data of the surrounding environment, and among the stored abstraction set environment data, determining the abstraction set environment data to which the environment identification word is added as the set data set of interest, Along with the noted set environment data, a predetermined number of set environment data before and after the set data are determined as connected data, and among the plurality of connected data, connected data having the same environment specific word is extracted, and abstraction for each direction of the connected data is performed. For the environment data, discard the direction-specific abstraction environment data having different values, and execute the operation command corresponding to the environment specific word by connecting the continuous data composed of the common abstraction set environment data having the direction-specific abstraction environment data having the same value. A recording medium recording operation plan data, wherein the operation plan data is generated as continuous data.