JP4096826B2 - Robot motion program manufacturing method, motion control system, and robot control system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an operation program creation support apparatus, an operation program manufacturing method, a control system, an operation program manufacturing method, and a computer program for an apparatus that operates according to an operation command, and is particularly suitable for an entertainment-oriented robot used in a home environment. The present invention relates to a robot motion program creation support apparatus, a robot motion program manufacturing method, a motion control system, a robot control system, a robot motion program manufacturing method, and a computer program.
[0002]
[Prior art]
In recent years, entertainment-oriented robots aimed at penetrating homes have attracted attention. Especially in the field of pet robots and toy robots, various products have been developed and are forming a large market.
[0003]
In addition, robots that place importance on communication with humans through dialogue have been studied. For example, NEC's “PaPeRo” has been published at shows and the like. Such an entertainment robot moves according to an operation program created in advance by a user or a developer, and the creation of the operation program is regarded as important.
[0004]
In order to manufacture a robot that does not bore users under such circumstances, it is necessary to create an enormous amount of robot operation programs. In the conventional technology, in order to create the operation program, the user or developer uses a text editor or the like on the screen of the robot development apparatus to use a script format (for example, a high-level language format such as C language). ) To create an operation control program.
[0005]
Japanese Laid-Open Patent Publication No. 7-200288 discloses a method of displaying a graphic corresponding to a program part on a screen and visually programming using the graphic.
[0006]
[Patent Document 1]
JP-A-7-200288
[Patent Document 2]
JP-A-9-91017
[0007]
[Problems to be solved by the invention]
However, in the above conventional technology, a user or developer must manually create and edit a robot operation program in a script format using a text editor or the like on the screen. In order to realize the above, there is a problem that a huge amount of programs must be created by humans, which requires a lot of time and labor.
[0008]
In Japanese Patent Laid-Open No. 7-200288, if it is assumed that a part is selected and a program is created with all the parts without editing, the parts assuming all cases must be prepared. As a result, there is a problem that it is meaningless to create a part because a human creates a vast program as much as an operation program and requires a lot of time and labor.
[0009]
SUMMARY OF THE INVENTION An object of the present invention is to solve the above-described problems, and to reduce a burden on a program creator by providing a visual program environment.
[0010]
Another object of the present invention is to provide an environment for automatically generating a program in view of the above problems.
[0011]
Another object of the present invention is to provide an environment of a robot control system for executing a created program in view of the above problems.
[0012]
[Means for Solving the Problems]
  The robot operation program manufacturing method of the present invention includes an automatic generation means for generating an operation program for designating an operation by each operation unit of the robot for producing the sound by using the sound to be emitted by the robot as an input. Features.
[0013]
  The automatic generation means is included in a terminal and / or robot connected to the robot via a line. The automatic generation means is characterized in that the robot operation is displayed as a time chart on the screen, and an operation command is arranged on the time chart to create a robot operation program.
[0014]
  The automatic generation means includes a storage unit that stores a plurality of component programs including designation of the operation of each operation unit of the robot and sound corresponding to the operation. Further, the automatic generation means refers to the part program, and determines a part program to be used as a part of the operation program according to a comparison result between the voice to be uttered by the robot and the voice included in the part program..
[0015]
  The automatic generation means limits the component program to be referred to to a component program selected using a predetermined random number. The automatic generation means generates the operation program by expanding or contracting each execution time of one or more operations included in the component program in accordance with the execution time of the voice to be uttered by the robot. In addition, the automatic generation means limits the designation of the operation to be expanded and contracted to an operation that does not involve an actual operation.
[0016]
  The operation control system of the present invention controls an operation by an operation program, generates an operation program based on the sound or sound to be output, and executes the operation program generated by the automatic generation unit to execute the sound or And execution control means for executing an operation generated in accordance with the reproduction of the sound.
[0017]
  The robot control system of the present invention controls the operation of each operation unit of the robot by an operation program, and performs an operation of designating the operation by each operation unit of the robot to produce the sound by inputting the sound to be generated by the robot Automatic generation means for generating a program, and execution control for executing an operation by sending a motion designation generated in accordance with the reproduction of sound to each motion part of the robot by executing an operation program generated by the automatic generation means Means.
[0019]
  Further, the present invention is further characterized by further having a state return means for returning each operation unit to an initial state after completion of the operation of the robot by the execution control means.
[0020]
  Also, when the status recovery means detects a person's face,Without returning to the initial state,It is directed to the person.
[0047]
DETAILED DESCRIPTION OF THE INVENTION
In the following, embodiments of the present invention will be described with reference to the drawings as examples applied to a robot.
[0048]
A first embodiment of the present invention will be described below with reference to the drawings. First, FIG. 1 is a diagram showing a configuration in the first embodiment of the present invention. Referring to FIG. 1, the first embodiment of the present invention includes a terminal device 1, a terminal storage unit 2, a robot control unit 3, a robot storage unit 4, and a robot operation unit 5. The terminal side includes a terminal device 1 and a terminal storage unit 2, and the robot includes a robot control unit 3, a robot storage unit 4, a robot control unit 3, and a robot operation unit 5. Each interface unit (not shown) for exchanging data between the robot side and the robot side is provided, and each interface unit is connected by a communication network (not shown) (not shown). At this time, data may be exchanged using a recording medium such as a semiconductor memory card instead of the communication circuit network.
[0049]
The terminal device 1 includes an operation program creation unit 11 that operates under program control by a processor (CPU) (not shown). The operation program creation unit 11 includes a manual creation unit 111 and an automatic generation unit 112. The manual creation means 111 expresses the motion of the robot as a visual time chart and arranges motion commands on the time chart, thereby enabling creation of a robot motion program. The automatic generation means 112 automatically generates a designation of an operation by the robot operation unit 5 of the robot for producing the sound by using the sound or sound to be generated by the robot as an input.
[0050]
The terminal storage unit 2 is a non-volatile recording medium device (for example, a magnetic disk) that can be read and written, and stores a plurality of component programs 21 that are used as components when an operation program is automatically generated.
[0051]
The robot control unit 3 includes an operation program execution unit 31, a head control unit 32, a leg control unit 33, a hand control unit 34, a face, A facial expression control unit 35, a utterance control unit 36, a sensor monitoring unit 37, and a face recognition unit 38 are provided. Further, the operation program execution unit 31 includes an execution control unit 311 and a state return unit 312. The robot control unit 3 is connected to the terminal device 1 via the communication line 6 or the like, and accepts program transmission from the terminal device 1. The execution control means 311 controls each movable part of the robot simultaneously in parallel, and determines whether to continue the operation of each operation part according to the degree of influence when a failure occurs. The state return means 312 returns the robot operation unit 5 to the initial state after the operation of the robot is completed.
[0052]
The robot storage unit 4 is a non-volatile recording medium device (for example, a magnetic disk) that can be read and written, and stores a plurality of operation programs 41 created by the terminal device 1 and transmitted to the robot side. . The operation program 41 in this case is a program that controls the operation of the operation units 51 to 55 (described later) of the robot.
[0053]
The robot operation unit 5 includes a head actuator 51, a leg actuator 52, a hand actuator 53, a facial expression LED 54, a utterance unit 55, a sensor unit 56, and a camera 57. The facial expression LED 54 is mounted around the eyes of the robot and the mouth / cheek positions, and the camera 57 is mounted at the positions of the eyes. In the following description, the head actuator 51, the leg actuator 52, the hand actuator 53, the facial expression LED 54, the generator 55, the sensor 56, and the camera 57 are collectively referred to as an operation unit ( Or it may be called a movable part).
[0054]
In addition, a microcomputer is attached to each of the operation units 51 to 57 of the robot operation unit 5, and each of the operation units 51 to 57 operates under the control of each microcomputer. In addition, you may make it control each operation | movement part 51-57 with one microcomputer or a some microcomputer. Therefore, the operation units 51 to 55 perform an operation based on an operation instruction based on a command generated from the robot control unit 3, and the operation units 56 and 57 notify the situation by transmitting the command to the robot control unit 3. .
[0055]
The operation program creating means 11 of the terminal device 1 creates the operation program 41 by a user's manual operation or automatic generation in accordance with an instruction from the user according to the content displayed on the screen. When manual creation is instructed, the manual creation unit 111 is activated, and the user is caused to create the operation program 41 using the GUI tool. When automatic generation is instructed, the automatic generation unit 112 is activated, and the operation program 41 is automatically generated in accordance with the reference voice or sound designated by the user while using the component program 21 of the storage unit 2.
[0056]
The operation program execution means 31 of the robot control unit 3 controls each operation unit of the robot according to the operation program 41 transmitted from the terminal device 1. That is, the operation program execution means 31 of the robot controller 3 serves as a control tower for controlling the entire robot. First, the robot control unit 3 activates the robot execution control unit 311, analyzes the operation program 41, and according to the contents, the head control unit 32, the leg control unit 33, the hand control unit 34, and the facial expression unit control unit. 35, an operation instruction is given to the generator control means 36. The head control means 32, the leg control means 33, the hand control means 34, the facial expression control means 35, and the generation control means 36 are a head actuator 51, a leg actuator 52, a hand according to a given operation instruction. The actuator 53, the facial expression LED 54, and the generator 55 are controlled by commands. Next, the robot control unit 3 activates the state return means 312 to return each operation unit of the robot to the initial state.
[0057]
The sensor monitoring means 37 monitors image information such as obstacles and steps such as stairs detected by the sensor 56, voice information such as human conversation, and supplies such information to the execution control means 311. Further, the face recognition unit 38 analyzes the image information of the camera 57 to determine whether there is a human face, and supplies the information to the execution control unit 311 as well.
[0058]
When a failure occurs during the execution of the operation, the execution control unit 311 determines the influence of the failure for each of the operation units 51 to 55, and if there is an influence, instructs the corresponding control unit 32 to 36 to stop the operation.
[0059]
Next, the operation of the first exemplary embodiment of the present invention will be described with reference to the drawings. 2 to 6 are flowcharts showing the operation of the first embodiment of the present invention. First, the operation of the first embodiment of the present invention will be described with reference to FIG. For ease of explanation, in the flowchart showing the overall image of FIG. 2, the processing on the terminal side is described on the left side, and the processing on the robot side is described on the right side.
[0060]
When the user requests creation of an operation program, the terminal device 1 first detects the request (step A1 in FIG. 2) and determines whether the user has instructed manual creation or automatic generation (step A2). . When manual creation is instructed, the terminal device 1 activates the manual creation unit 111 of the operation program creation unit 11 and causes the user to create an operation program with the support of the GUI tool (step A3). When the automatic generation is instructed, the terminal device 1 activates the automatic generation unit 112 of the operation program generation unit 11 and causes the user to specify the audio data used as a reference. A program for the operation unit is automatically generated (step A4).
[0061]
When the creation of the operation program ends and a user program transmission request is detected (step A5), the terminal device 1 transmits the created operation program to the robot side via the communication line (step A6).
[0062]
The robot control unit 3 on the robot side monitors the presence / absence of a transmission program from the terminal side (step A7), receives the program if it exists, stores it in the robot storage unit 4 (step A8), and executes the operation program The execution control means 311 of the means 31 is activated and the operation units 51 to 55 are controlled according to the contents of the operation program (step A9). When the operation is completed, the robot control unit 3 activates the state return unit 312 of the operation program execution unit 31 to return the operation units 51 to 55 to the initial state (step A10).
[0063]
FIG. 7 is an example of a GUI tool used in step A3 in the flowchart shown in FIG. This tool corresponds to the manual creation means 111 of the present invention.
[0064]
Referring to FIG. 7, on the left side of the GUI tool displayed on the screen of the terminal device 1, there is a list F <b> 1 of operation command buttons corresponding to the operation units 51 to 55, and a pointing device such as a mouse not illustrated from here. , The operation command button F1 is arranged on the right time chart F2 by drag and drop. A scale F4 for representing the length of time is provided at the top of the time chart F2. In the example of FIG. 7, the length of one scale is 0.1 seconds.
[0065]
FIG. 8 is an explanatory diagram showing the button expansion / contraction operation and display method, and is performed by the manual creation means 111 in accordance with the user's button operation. For example, when the operation command button F3 arranged along the time chart F2 in FIG. 7 is clicked, a small rectangle G1 for expansion and contraction is displayed on the left and right of the button as shown in FIG. When dragged to, the operation command button expands and contracts. When the center of the arranged operation command button is dragged, the button moves left and right. Further, when the buttons are overlapped with each other, hatching display is performed to call the user's attention as shown by the button G2 in FIG.
[0066]
Further, when one of the operation command buttons F3 on FIG. 7 is double-clicked, an operation setting dialog as shown in FIG. 9 is displayed on the screen of the terminal device 1. In the case of the buttons of the operation units 51 to 53 accompanied by an actuator (robot action), either an angle designation or a speed designation can be designated on the dialog. For example, if the horizontal angle is set to “20 ° to the left” and the horizontal width of the button is set to “0.5 seconds”, this means an operation of “turning to the left by 20 ° for 0.5 seconds”. Is set to “20 ° / second left”, it means an operation of “continuing to turn left at a speed of 20 ° / second for 0.5 seconds”. Depending on the settings, the performance limit of the actuator may be exceeded. At this time, the button is displayed in red to alert the user. In the case of the button of the facial expression part 54 with an LED, the light emission color of each LED is designated on the dialog. In the case of the button of the utterance unit 55, a voice file name (* .mp3, * .wav, etc.) or text for voice synthesis is designated on the dialog. The audio file may be transmitted together with the operation program when it is transmitted, and stored in the robot storage unit 4, or all necessary files may be stored in the robot storage unit 4 in advance. When the setting is completed and the dialog is closed, the setting contents are displayed on the button on the screen of the terminal device 1 as shown in FIG.
[0067]
In the display of the operation command with the operation command button in FIG. 7, there is a mode indicating the state at the end of the operation. For example, after the head operation command “Left 20 °, Right 40 °, Left 40 °, Right 20 °, Left 40 °, Right 20 °, Left 160 ° / sec, Right 160 ° / sec” The head states are “left 20 °, right 20 °, left 20 °, front, left 40 °, left 20 °, right behind, left 20 °”, respectively. Therefore, in the display of the operation command on the operation command button, “left 20 °, right 40 °, left 40 °, right 20 °, left 40 °, right 20 °, left 160 ° / sec, right 160 ° / sec. ”Is displayed instead of“ 20 ° left, 20 ° right, 20 ° left, front, 40 ° left, 20 ° left, right behind, 20 ° left ”.
[0068]
When the above operation is repeated and programming on the GUI tool is completed, the manual creation unit 111 converts this into an operation program that can be recognized by the robot. FIG. 10 is an example of an operation program corresponding to the screen of FIG. Referring to FIG. 10, the operation program is divided into program sections I1 to I5 corresponding to the operation units 51 to 55. When the operation program is executed, the programs in the sections I1 to I5 are simultaneously executed in parallel. The operation command I6 in the program corresponds to the operation command button F3 in FIG. 7 on a one-to-one basis. Here, the operation command refers to a basic unit of operation that can be executed by the control means 32 to 36 of the robot control unit 3. In order to facilitate comparison with FIG. 7, the width of the rectangle representing the operation command is matched with the execution time. When there are blank areas (areas where buttons are not arranged) in the time chart of FIG. 7, these are converted into non-operational instructions such as “still” and “off” as shown in the instruction I7 in FIG.
[0069]
This GUI tool can greatly improve the efficiency of programming. In the conventional robot programming environment, the operation instructions had to be written in a script language such as "LEFT = 20 TIME = 0.5", so it was impossible to grasp how each operation unit synchronized, It was necessary to repeat the operation of fine adjustment after executing the trial. If the above GUI tool is used, it is possible to visually grasp the state of synchronization of the respective operation units as shown in FIG.
[0070]
FIG. 3 is a flowchart showing detailed processing for step A4 in the flowchart shown in FIG. This flowchart corresponds to the process of the automatic generation unit 112.
[0071]
When the user designates reference voice data (step B1 in FIG. 3), the automatic generation means 112 first generates an operation program having a voice command for the reference voice in the program section I5 in FIG. 10 (step B2). Here, when a text for voice synthesis is designated as voice data, the automatic generation means 112 once records the synthesized voice and uses the recorded data as a reference voice after step B3. Next, the automatic generation means 112 initializes the identifier i for identifying the operation units 51 to 55 (step B3), determines whether the identifier i indicates the utterance unit (step B4), and the utterance unit If not, the user is inquired whether or not to generate a program for the operation unit pointed to by i (step B5). Here, when the user designates omission, the automatic generation means 112 moves to the processing of the next operation unit (step B10), and when the generation is designated, the automatic generation means 112 starts collation with the part program. After the point is set at the head of the reference voice, a program corresponding to one part program is generated (step B7). After that, the automatic generation means 112 moves the reference voice collation start point to the head of the ungenerated section (step B8), and repeats the processing from step B7 until the generation of the entire reference voice is completed (step B9). . When the program generation of the current operation unit is completed, the automatic generation unit 112 updates the identifier i to indicate the next operation unit (step B10), and repeats the processing from step B4 until the program generation of all operation units is completed. (Step B11).
[0072]
FIG. 11 is an example of setting information of the part program 21. In the component program, dedicated component groups such as “for the head” and “for the leg” are prepared corresponding to the operation units 51 to 55, and one component program includes the operation unit as shown in FIG. Is composed of a pair of one or a plurality of operation instructions J1 related to the above and a collation voice J2 for calculating the similarity between the reference voice and the reference voice. The part program is adjusted in advance so that the robot behaves naturally when the operation command J1 and the collation voice J2 are simultaneously executed. Generally, a part of the operation program created by the GUI tool is used. Further, the total time of the operation command J1 and the utterance time of the collation voice J2 are adjusted to coincide. The operation command J1 becomes a part of the generated operation program, but the collation voice J2 is only used for similarity calculation, and does not become a part of the operation program.
[0073]
FIG. 4 is a flowchart showing detailed processing of the automatic generation means 112 for the processing of step B7 in the flowchart shown in FIG. When the operation unit pointed to by the identifier i and the reference start point on the reference voice are determined, the automatic generation unit 112 first selects N candidates from the part program by random numbers (step C1 in FIG. 4). The reason why all parts programs are not candidates here is that in the world of entertainment robots, where the present invention is important, it is difficult for the user to get bored with the valuation of motion, so the motion of a specific reference voice is unique. This is because it is preferable to make a slight change each time, rather than to
[0074]
When the candidates are determined, the automatic generation means 112 initializes the identifier j for identifying them and the identifier k of the adopted part (step C2), and initializes the maximum similarity to 0 (step C3). Next, the automatic generation means 112 extracts the collation voice from the part program indicated by the identifier j (step C4), cuts out the voice data having the same length as the reference voice collation start point (step C5), and the degree of similarity between them. Is calculated (step C6). Here, when the collation voice is longer than the remaining interval of the reference voice, an extra second half of the collation voice is omitted, and the similarity is calculated from the voice extracted from the reference voice.
[0075]
Note that step C6 does not specify a method of calculating the similarity, and a speech spectrum may be used, or speech vector quantization may be used.
[0076]
When the similarity can be calculated, the automatic generation means 112 compares it with the maximum similarity of the past part program (step C7), and if the similarity is higher, stores it as the maximum similarity. At the same time, the identifier k is updated to indicate the current part program (step C8). After that, the automatic generation means 112 updates the identifier j to point to the next part program (step C9), and repeats the processing from step C4 until all the part programs are checked (step C10).
[0077]
When the check is completed, the operation instruction group of the component indicated by the identifier k is copied to the end of the program partition related to the operation unit. If the total time of the operation command group does not fit in the total time of the reference voice, only the amount that fits is copied.
[0078]
FIG. 12 is an explanatory diagram showing the progress of the automatic generation processing of the automatic generation means 112 in FIGS. First, the first candidate is extracted from the part program group K0 for the head using a random number, and each of the collation voices is collated with the voice K5a having the same length starting from the collation start point p1 of the reference voice K5, and then the most similar degree is obtained. The high-part K1 is employed, and the operation instruction group K1a is copied to the program partition K4 at the head. Next, the collation start point is updated to p2, and after the component K2 is adopted in the same procedure, the operation instruction group K2a is copied to the partition K4. Finally, the collation start point is updated to p3, and the component K3 is adopted. At this time, the collation voice K3b of K3 is longer than the remaining section K5c of the reference voice K5, and therefore, the reference voice K3b and the first half section K3c The similarity of the remaining voice segment K5c is calculated. When all of the operation command group k3a is copied, the end of the reference voice K5 is exceeded, so only the first two instructions that fall within the reference voice section are copied. When the movement generation of the head is completed in this way, the automatic generation means 112 generates the movement of the leg part to the facial expression part in the same procedure.
[0079]
With this automatic generation processing, it is not necessary to prepare a part program that assumes every case, and a large amount of operation programs corresponding to combinations of part programs can be generated, thereby further improving programming efficiency.
[0080]
FIG. 5 is a flowchart showing detailed processing for step A9 in the flowchart shown in FIG. This flowchart corresponds to the process of the execution control unit 311.
[0081]
The execution control means 311 first saves the initial state before execution of the operation units 51 to 54 (step D1 in FIG. 5). The initial state here refers to, for example, horizontal and vertical orientations for the head, bending angle of the joint for the hand, and volume value for the utterance. This initial state is used in the processing of the state return means 312 described later.
[0082]
Next, the execution control means 311 decomposes the operation program into individual programs corresponding to the program sections I1 to I5 (see FIG. 10) (step D2), and initializes the identifier i of the operation unit (step D3). If there is an individual program corresponding to the identifier i, the first operation command is taken out (steps D4 to D5), and the corresponding unit of the control units 32 to 36 of the robot control unit 3 is instructed to execute the operation command (step D6). ). After that, the execution control means 311 updates the identifier i to point to the next operation unit (step D7), and repeats the processing from step D4 until all the operation units are processed (step D8).
[0083]
Here, the control means 32 to 36 returns control to the execution control means 311 without waiting for completion of the designated operation command. For this reason, each operation | movement part starts operation | movement almost simultaneously.
[0084]
When all the operation units start to move, the execution control unit 311 monitors completion notifications and failure notifications from the control units 32 to 36 (steps D9 to D10). If there is a failure notification, the execution control unit 311 proceeds to failure countermeasures from step D16. . If there is a completion notification, the execution control means 311 determines which operation unit it corresponds to (step D11). If there is an operation command remaining in the operation unit, the execution control unit 311 extracts the next command ( Steps D12 to D13) and the corresponding means of the control means 32 to 36 are instructed to execute the operation command (Step D14). If all the operation commands of the operation unit have been completed, the execution control unit 311 checks whether the other operation units have been completed in the same manner (step D15). If not, the process from step D9 is repeated.
[0085]
If a failure notification is detected during operation, the execution control means 311 determines whether or not it is a fatal failure (eg battery shortage) (step D16), and if it is fatal, stops all operations. Then, the process returns to the process of FIG. 2 (step D23). Otherwise, the execution control unit 311 determines whether or not the failure affects each operation unit, and instructs only the affected operation unit to stop the operation (steps D18 to D22).
[0086]
This method makes it possible to appropriately cope with the situation instead of unconditionally stopping all operations when a failure occurs.
[0087]
For example, if the sensor 56 detects an obstacle in the direction of movement of the robot and the sensor monitoring means transmits it to the execution control means 311 as an obstacle notification, the movement of the leg stops because the collision occurs when the sensor advances further. Since there is no direct effect on the head or vocalization, they continue to operate.
[0088]
In addition, when the operation units 51 to 55 are stopped due to a failure, a completion notification related to the operation unit is not issued, and thus the problem that the stopped operation unit restarts operation due to the completion notification does not occur.
[0089]
FIG. 6 is a flowchart showing detailed processing for the processing of step A10 in the flowchart shown in FIG. This flowchart corresponds to the processing of the state return means 312.
[0090]
When the execution of the operation program is completed, the state returning means 312 initializes the identifier i of the operation unit (step E1 in FIG. 6), and then determines whether the identifier i indicates the head (step E2). If the part is indicated, the process proceeds to the head-only process from step E7. Otherwise, the state return means 312 acquires the information of the initial state saved at step D1 in FIG. 5 (step E3), and gives an instruction to return the initial state to the corresponding control means. After that, the state return means 312 updates the identifier i to point to the next operating unit (step E5), and repeats the processing from step E2 until the processing of all the operating units is completed (step E6). When i indicates the head, the state returning means 312 determines whether or not a human face is reflected on the camera 57 via the face recognition means 38 (step E7). The head control means 32 is instructed to move the head following the face (step E8). The head control means 32 repeats the process of calculating the deviation between the center of the camera image and the center of the human face and moving the head so as to correct the deviation. If the face is not shown, the state return means 312 proceeds to the process of step E3, and performs the return process to the initial state (initial position in the case of the head) in the same manner as the other operation units.
[0091]
With this method, for example, even when an operation program that ends with the head facing backward is executed, the operation program is returned to an appropriate position after the operation is completed, so that the next operation program can be executed without any problem. Also, if you are talking to a person, you will turn to that person after completing the action.
[0092]
Next, a second embodiment of the present invention will be described with reference to the drawings. Referring to FIG. 1, the automatic generation unit 112 described in the first embodiment is different from the automatic generation unit 112 in the second embodiment. Differences will be described in the operation of the second embodiment.
[0093]
Next, the operation of the second exemplary embodiment of the present invention will be described with reference to the drawings.
In the first embodiment, in the partial program generation process in step B7 of FIG. 3, the automatic generation unit 112 extracts a plurality of component programs as candidates when the operation program is automatically generated. The generation unit 112 extracts a single component program with a random number or allows the user to specify it. Further, after extracting only one part program, the automatic generation means 112 does not use the verification voice, expands and contracts the total time of the operation commands in the part program in accordance with the time of the reference voice, and operates the operation time of each operation unit. To match the time of the reference voice.
[0094]
FIG. 13 is an explanatory diagram showing the progress of the automatic generation processing of the automatic generation means 112 in the second embodiment. Here, as an example, the ratio of execution times of operation instructions included in the extracted component program is 1: 2: 3, and the ratio of execution times on the operation program obtained by copying and decompressing the operation instructions is also 1: 2. 2: 3. Thus, in the second embodiment, expansion and contraction is performed so as to keep the distribution of the execution time of the original part program.
[0095]
This method is based on an empirical rule that the movement of the robot looks natural when the end timing of the voice and other movements are matched.
[0096]
Next, a third embodiment of the present invention will be described with reference to the drawings. Referring to FIG. 1, the automatic generation unit 112 described in the second embodiment is different from the automatic generation unit 112 in the third embodiment. Differences will be described in the operation of the third embodiment.
[0097]
Next, the operation of the third exemplary embodiment of the present invention will be described with reference to the drawings. In the second embodiment, in the partial program generation process in step B7 in FIG. 3, all operation instructions in the part program are subject to expansion / contraction. In the third embodiment, no operation such as “still” or “off” is performed. Only the operation command representing is expanded or contracted.
[0098]
FIG. 14 is an explanatory diagram showing the progress of the automatic generation processing of the third embodiment. Here, as an example, only the instruction corresponding to no operation is expanded while maintaining the ratio of 1: 2, and the execution time of the instruction with the actual operation does not change.
[0099]
In the case of the second embodiment, if the degree of extension is large, it will feel that the movement is extended, but in the third embodiment, only the non-operation section is extended, so that natural operation is performed without impairing sharpness. Can be generated.
[0100]
The no-operation instruction corresponds to a blank area (an area where buttons are not arranged) in the time chart of FIG.
[0101]
Next, a fourth embodiment of the present invention will be described with reference to the drawings. Referring to FIG. 1, the operation program generated by the automatic generation means described in the first, second, or third embodiment is displayed on a time chart on a screen for manual editing such as fine adjustment. This is different in that a function for allowing the user to edit is added. Differences will be described in the operation of the fourth embodiment.
[0102]
Next, the operation of the fourth exemplary embodiment of the present invention will be described with reference to the drawings. In the fourth embodiment, the automatic generation of the operations described in the first to third embodiments is visually performed on the time chart of FIG. That is, the automatic generation unit 112 displays the created part program on the screen as shown in FIG. Since the method of changing the content displayed on the screen is the same as the operation performed by the manual creation unit 111 in the first embodiment, the description thereof is omitted. With this method, the user can manually fine-tune the automatically generated operation, so that it is possible to finish the operation more appropriately.
[0103]
Next, a fifth embodiment of the present invention will be described with reference to the drawings. Referring to FIG. 1, the first embodiment is based on the point that the manual creation unit 111 of the terminal device 1 is provided in the robot control unit and the part program 21 of the terminal storage unit 2 is stored in the robot storage unit 4. It differs from the form. In this case, an interface circuit for connecting the display device and the keyboard from the robot is provided, and when the operation program is generated, the display device and the keyboard are connected. Differences will be described in the operation of the fifth embodiment.
[0104]
Next, the operation of the fifth exemplary embodiment of the present invention will be described with reference to the drawings. In the first embodiment, the operation program is automatically generated on the terminal device side, but in the fifth embodiment, this is performed on the robot side. That is, the manual creation unit 111 in FIG. 1 becomes a part of the operation program execution unit 31 on the robot side. When the user designates the reference voice and instructs the generation and execution of the operation program, the robot side performs the execution control unit 311. Prior to execution, the manual creation unit 111 is activated, and an operation program is automatically generated and then executed. Since the operation of the manual creation unit 111 is the same as that of the first embodiment, description thereof is omitted. The part program 21 used at this time is also stored on the robot storage unit 4 side.
[0105]
In the above description, the robot is taken as an example. However, the application target is not limited to a so-called robot, and the present invention can be applied to a general apparatus that operates according to an operation command. For example, in a game machine in which input / output is an input switch, screen output, and sound output, the face display unit LED 54 is a screen output unit, the utterance unit 55 is a sound output unit, and a sensor 56 is input in the configuration of the robot operation unit 5 of FIG. It can be regarded as a switch. The head actuator 51, leg actuator 52, hand actuator 53, facial expression LED 54, and camera 57 are not provided. An entertainment studio device in which a person sits inside and enjoys the world of virtual reality with a movable chair and powerful images and sound. In this case, instead of the leg actuator 52, the entire chair on which a person is seated may be operated. Further, instead of the hand actuator 53 and the head actuator 51, a device for spraying water or a device for generating fog using dry ice or the like may be used. An image may be displayed on the screen instead of the desired expression section LED54.
[0106]
Each step of FIGS. 3, 4, 5 and 6 can be realized by a processing program in which instructions of a computer are arranged.
[0107]
FIG. 15 is a diagram showing a recording medium 91 in which the processing program according to the embodiment of the present invention is recorded, and a computer 92 that loads and executes the processing program from the recording medium 91.
[0108]
【The invention's effect】
As described above, the first effect of the present invention is that the conventional program method in which trial execution is repeated because the operation tool can be visually grasped by the GUI tool when the operation program is manually created. Can save time and improve production efficiency.
[0109]
The second effect of the present invention is that an operation program can be automatically generated simply by designating a reference voice by the user, so that a large amount of programs can be created efficiently, which is a problem of an entertainment robot, A robot that can't get bored of can be easily realized.
[0110]
The third effect of the present invention is that when a failure occurs during the execution of the operation program, it is determined for each operation unit whether the failure affects, so the leg is stopped even if an obstacle is detected. As the operation continues, an optimum response can be made according to the situation.
[0111]
According to the fourth effect of the present invention, since the return process to the initial state is performed after the execution of the operation program, even if the operation program is executed such that the operation unit ends in an inappropriate state, the next operation program can be performed without any problem. Can be executed. Further, when the user is in a conversation with the person, the person faces the person after the operation program is executed, so that the conversation can be continued without a sense of incongruity.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration in a first embodiment of the present invention.
FIG. 2 is a flowchart showing the operation of the first exemplary embodiment of the present invention.
3 is a flowchart showing the automatic program generation operation of FIG. 2 by the automatic generation means of FIG.
4 is a flowchart showing the operation of the partial program generation of FIG. 3 by the automatic generation means of FIG.
5 is a flowchart showing an operation program execution operation of FIG. 2 by the execution control means of FIG.
6 is a flowchart showing the operation of returning the initial state of FIG. 2 by the state returning means of FIG.
FIG. 7 is a diagram showing a screen configuration of a GUI tool according to the first embodiment of the present invention.
FIG. 8 is a diagram illustrating an extension / contraction operation and a display method of an operation command button according to the first embodiment of the present invention.
FIG. 9 is a diagram showing an operation setting dialog according to the first embodiment of the present invention.
FIG. 10 is a diagram showing a configuration of an operation program according to the first embodiment of the present invention.
FIG. 11 is a diagram showing a configuration of a part program according to the first embodiment of the present invention.
FIG. 12 is a diagram illustrating a process of automatic generation processing according to the first embodiment of the present invention.
FIG. 13 is a diagram illustrating a process of automatic generation processing according to the second embodiment of the present invention.
FIG. 14 is a diagram illustrating a process of automatic generation processing according to the third embodiment of the present invention.
FIG. 15 is a diagram illustrating a recording medium on which a computer program according to an embodiment of the present invention is recorded, and a computer that loads and executes the computer program from the recording medium.
[Explanation of symbols]
1 Terminal device
11 Operation program creation means
111 Manual creation means
112 Automatic generation means
2 Terminal storage
21 Parts program
3 Robot controller
31 Operation program execution means
311 Execution control means
312 State return means
32 Head control means
33 Leg control means
34 Hand control means
35 Facial expression control means
36 Voice control means
37 Sensor monitoring means
38 Face recognition means
4 Robot storage
41 Operation program
5 Robot movement part
51 Head actuator
52 Leg actuator
53 Hand actuator
54 Facial expression LED
55 Voice part
56 sensors
57 camera
6 Communication lines
91 Recording media
92 computers

Claims (19)

  A robot operation program manufacturing method comprising: automatic generation means for generating an operation program for designating an operation by each operation unit of the robot for producing the sound by inputting an audio to be generated by the robot.   The robot operation program manufacturing method according to claim 1, wherein the automatic generation means is included in a terminal connected to the robot via a line, the robot, or both.   The automatic generation means displays the operation of the robot as a time chart on a screen, and arranges an operation command on the time chart to create an operation program for the robot. Item 3. The robot operation program manufacturing method according to Item 2.   4. The automatic generation unit includes a storage unit that stores a plurality of component programs including designation of movement of each movement unit of the robot and sound corresponding to the movement. The robot operation program manufacturing method according to claim 1.   The automatic generation means refers to the part program and determines a part program to be used as a part of an operation program according to a comparison result between a voice to be uttered by the robot and a voice included in the part program. The robot operation program manufacturing method according to claim 4.   6. The robot operation program manufacturing method according to claim 5, wherein the automatic generation means limits the component program to be referred to to a component program selected using a predetermined random number.   The automatic generation means generates an operation program by expanding or contracting each execution time of one or more operations included in the component program in accordance with an execution time of a voice to be uttered by the robot. Item 5. The robot operation program manufacturing method according to Item 4.   8. The robot motion program manufacturing method according to claim 7, wherein the automatic generation means limits the designation of the motion to be expanded and contracted to a motion not accompanied by an actual motion. In an operation control system that controls operation by an operation program,
Automatic generation means for generating an operation program based on voice or sound to be output;
An operation control system comprising: an execution control unit that executes an operation generated in accordance with reproduction of voice or sound by executing the operation program generated by the automatic generation unit. In a robot control system that controls the motion of each motion part of a robot by an motion program,
Automatic generation means for generating an operation program for designating an operation by each operation unit of the robot for producing the sound, with the audio to be emitted by the robot as an input;
An execution control unit that executes an operation program generated by the automatic generation unit, and sends an operation designation generated in accordance with the reproduction of the sound to each operation unit of the robot to execute the operation. Characteristic robot control system.   11. The robot control system according to claim 10, wherein the automatic generation unit is included in a terminal connected to the robot via a line, the robot, or both.   11. The automatic generation means displays an operation of the robot as a time chart on a screen and arranges an operation command on the time chart to create an operation program for the robot. Or the robot control system of 11.   The said automatic generation means is provided with the memory | storage part which stored several component programs containing the designation | designated operation | movement of each operation | movement part of a robot, and the audio | voice corresponding to operation | movement. The robot control system according to claim 1.   The automatic generation means refers to the component program, and determines a component program to be used as a component of the operation program according to a comparison result between a sound to be generated by the robot and a sound included in the operation component. The robot control system according to claim 13.   15. The robot control system according to claim 14, wherein the automatic generation means limits the component program to be referenced to a component program selected using a predetermined random number.   The automatic generation means generates an operation program by expanding or contracting each execution time of one or more operations included in the component program in accordance with an execution time of a voice to be uttered by the robot. Item 14. The robot control system according to Item 13.   The robot control system according to claim 16, wherein the automatic generation unit limits the designation of an operation to be expanded and contracted to an operation not involving an actual operation.   The robot control system according to claim 10, further comprising a state return unit that returns each operation unit to an initial state after completion of the operation of the robot by the execution control unit. 19. The robot control system according to claim 18 , wherein the state returning means does not return the head to the initial state when detecting a human face, but directs the head toward the person.

Images