JP6079350B2 - Robot control method, robot control apparatus, robot, and robot control program - Google Patents

Description

  The present invention relates to a robot control method, a robot control device, a robot, and a robot control program.

  Patent Document 1 adds a learning function for changing the setting of control target values and changing control system parameters to the control method of the conventional object gripping device, so that an object having the same degree of hardness as an object that has been gripped in the past can be obtained. It describes that the time required to achieve gripping is shortened when gripping is performed.

JP-A-10-249767

  The invention described in Patent Document 1 is characterized in that the number of trials until successful gripping is reduced using past gripping history. In the invention described in Patent Document 1, since only the gripping operation is an object, there are only one combination of parameters in the history stored in the database. Therefore, in the invention described in Patent Document 1, all past histories stored in the database can be used when parameters are compared in the learning function.

  However, when moving an object gripped using a robot, there are various parameters as compared to the case of simply gripping an object using a robot. For example, when an object is moved, there are a plurality of parameters such as a movement start position, a movement destination position, a relay point, and a speed. In addition, a series of operations on the object is defined by combining some or all of a plurality of parameters such as a movement start position, a movement destination position, a relay point, and a speed. Therefore, there are a plurality of parameter combinations in the data stored in the database as the past history.

  Therefore, when the parameters are compared in the learning function, the past history stored in the database cannot simply be used, and the invention described in Patent Document 1 is used when moving the object gripped using the robot. Cannot be applied.

  The present invention relates to a robot control method capable of selecting a scenario suitable for work and a scenario with a high success rate based on past execution results when causing the robot to perform work such as movement of the grasped object. An object is to provide a robot control device, a robot, and a robot control program.

  A first aspect of the present invention for solving the above problem is a robot control method for causing a robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined. Obtaining first execution information including a plurality of execution logs in which a scenario, a selected value of a first parameter defining the scenario, and an execution result of the scenario are associated, and the content of the target work; A first step of obtaining instruction information including a selected value of a second parameter that defines the target work; and from the first execution information, the work content indicated by the scenario is the content of the target work. A second step of extracting, as second execution information, an execution log that is the same and whose selection value of the first parameter is similar to the selection value of the second parameter; Including a third step of selecting a scenario having the highest success rate based on the execution result from a scenario included in the execution information of 2, and a fourth step of outputting the selected optimal scenario. It is characterized by.

  According to the first aspect, the first execution information including a plurality of execution logs associating the scenario executed by the robot, the selected value of the first parameter defining the scenario, and the execution result of the scenario. get. In addition, instruction information including the content of the target work of the robot and the selected value of the second parameter that defines the target work is acquired. Then, from the first execution information, an execution log in which the work content indicated by the scenario is the same as the content of the target work and the selection value of the first parameter is similar to the selection value of the second parameter is obtained as the second execution information. Extract as From the scenarios included in the execution log of the second execution information, a scenario that is optimal for the target work is selected, and the selected scenario is output. As a result, when the target work is to be performed by the robot, the work content indicated by the scenario in the past execution results is based on the same execution log as the content of the target work, and the scenario suitable for the work and the success rate A high scenario can be selected. Further, by selecting work results that have been successful in the past in an environment close to the target work, the robot can be controlled realistically, practically, universally, and flexibly.

  Here, in the first step, speed information that is information about a speed at which the robot is moved may be acquired as the first parameter and the second parameter. As a result, speed information that greatly affects the execution result can be used to extract an execution result in an environment close to the target work.

  Here, the execution log and the instruction information include speed information that is information related to a speed at which the robot is moved. In the second step, the speed information of the instruction information is the same as the speed information of the instruction information. An execution log associated with the speed information may be extracted. As a result, a more appropriate scenario can be selected by selecting a scenario using only the execution log that matches the speed information that greatly affects the execution result.

  Here, in the second step, based on the similarity calculated using a calculation formula that decreases as the difference between the selected value of the first parameter and the selected value of the second parameter increases. The second execution information may be extracted. As a result, it is possible to extract an execution result in an environment close to the target work.

  A second aspect of the present invention for solving the above-described problem is a robot control apparatus that causes the robot to perform a desired work using a scenario in which a plurality of unit operations of the robot are combined. A first acquisition unit that acquires first execution information including a plurality of execution logs that associate a scenario, a selected value of a first parameter that defines the scenario, and an execution result of the scenario; From the first acquisition information, the second acquisition unit that acquires the instruction information including the contents of the work and the selected value of the second parameter that defines the target work, and the work contents indicated by the scenario are A scenario group extraction that extracts, as second execution information, an execution log that is the same as the content of the target work and whose selected value of the first parameter is similar to the selected value of the second parameter. A scenario selection unit that selects a scenario with the highest success rate based on the execution result from scenarios included in the second execution information, and an output unit that outputs the selected optimal scenario. It is characterized by providing. As a result, when the robot performs an operation such as moving the gripped object, the work content indicated by the scenario in the past execution results is suitable for the work based on the same execution log as the content of the target work. Scenario with high success rate can be selected. Further, by selecting work results that have been successful in the past in an environment close to the target work, the robot can be controlled realistically, practically, universally, and flexibly.

  A third aspect of the present invention for solving the above problem is a robot that causes a movable part of the robot to perform a desired work using a scenario in which a plurality of unit operations of the robot are combined. A first acquisition unit that acquires first execution information including a plurality of execution logs in which the scenario, the selected value of the first parameter that defines the scenario, and the execution result of the scenario are associated with each other; From the first acquisition information, the second acquisition unit that acquires the instruction information including the content of the work and the selected value of the second parameter that defines the target work, the work content indicated by the scenario is A scenario group extraction that extracts, as second execution information, an execution log that is the same as the content of the target work and whose selected value of the first parameter is similar to the selected value of the second parameter. A scenario selection unit that selects a scenario with the highest success rate based on the execution result from the scenario included in the second execution information, and controls the movable unit using the selected optimal scenario. And a drive control unit. As a result, when the robot performs an operation such as moving the gripped object, the work content indicated by the scenario in the past execution results is suitable for the work based on the same execution log as the content of the target work. Scenario with high success rate can be selected. Further, by selecting work results that have been successful in the past in an environment close to the target work, the robot can be controlled realistically, practically, universally, and flexibly.

  A fourth aspect of the present invention for solving the above-described problem is a robot control program for causing a robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined. Obtaining first execution information including a plurality of execution logs in which a scenario, a selected value of a first parameter defining the scenario, and an execution result of the scenario are associated, and the content of the target work; A first step of obtaining instruction information including a selected value of a second parameter that defines the target work; and from the first execution information, the work content indicated by the scenario is the content of the target work. A second step of extracting, as second execution information, an execution log that is the same and whose selection value of the first parameter is similar to the selection value of the second parameter; A third step of selecting a scenario having the highest success rate based on the execution result from scenarios included in the second execution information; and a fourth step of outputting the selected optimal scenario. It is characterized by being executed by an arithmetic unit. As a result, when the robot performs an operation such as moving the gripped object, the work content indicated by the scenario in the past execution results is suitable for the work based on the same execution log as the content of the target work. Scenario with high success rate can be selected. Further, by selecting work results that have been successful in the past in an environment close to the target work, the robot can be controlled realistically, practically, universally, and flexibly.

1 is a diagram illustrating an example of a schematic configuration of a robot system 1 according to a first embodiment of the present invention. It is a figure which shows an example of a function structure of a control apparatus. It is a figure which shows an example of the information stored in a library memory | storage part. It is a figure which shows typically the movement path | route of the hand in case a parameter differs with respect to each scenario. It is a figure which shows an example of a structure of an execution log table. It is a figure which shows an example of the hardware constitutions of a control apparatus. It is a flowchart which shows an example of the scenario determination process of the robot system. It is a flowchart which shows an example of the scenario group A extraction process of the robot system. It is a flowchart which shows an example of a scenario S selection process. It is a flowchart which shows an example of the scenario determination process of the robot system 2 which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows an example of the scenario group A extraction process of the robot system.

<First Embodiment>
A first embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a system configuration diagram showing an example of a configuration of a robot system 1 according to an embodiment of the present invention. The robot system 1 in this embodiment mainly includes a robot 10, a control device 20, a first imaging unit 30, and a second imaging unit 31.

The robot 10 is an arm type robot having two arms. In the present embodiment, a dual-arm robot having two arms 11 will be described as an example, but the number of arms 11 of the robot 10 may be one.
The arm 11 includes a plurality of joints (joints) 12 and a plurality of links 13.

  At the tip of the arm 11, a hand 14 (so-called end effector or so-called end effector that can hold a workpiece A that is an operation target of the robot 10 or can perform a predetermined operation on the target by holding a tool. A hand effector) is provided. The arm 11 and the hand 14 correspond to a movable part of the present invention.

  The joint 12 and the hand 14 are provided with an actuator (not shown) for operating them. The actuator includes, for example, a servo motor and an encoder. The encoder value output by the encoder is used for feedback control of the robot 10 by the control device 20.

  A force sensor (not shown) is provided inside the hand 14 or the tip of the arm 11. The force sensor detects a force applied to the hand 14. As the force sensor, for example, a six-axis force sensor that can simultaneously detect six components, ie, a force component in the translational three-axis direction and a moment component around the three rotation axes, can be used. The physical quantities used in the force sensor are current, voltage, charge amount, inductance, strain, resistance, electromagnetic induction, magnetism, air pressure, light, and the like. The force sensor can detect six components by converting a desired physical quantity into an electrical signal. The force sensor is not limited to six axes, and may be, for example, three axes. The position where the force sensor is provided is not particularly limited as long as the force applied to the hand 14 can be detected.

  Note that the configuration of the robot 10 is not limited to the above-described configuration because the main configuration has been described in describing the features of the present embodiment. This does not exclude the configuration of a general gripping robot. For example, although a 6-axis arm is shown in FIG. 1, the number of axes (number of joints) may be further increased or decreased. The number of links may be increased or decreased. In addition, the shape, size, arrangement, structure, and the like of various members such as the arm, hand, link, and joint may be appropriately changed. The end effector is not limited to the hand 14.

  The control device 20 performs processing for controlling the entire robot 10. The control device 20 may be installed at a location away from the main body of the robot 10 or may be built in the robot 10 or the like. When the control device 20 is installed at a location away from the main body of the robot 10, the control device 20 is connected to the robot 10 by wire or wirelessly.

  The first imaging unit 30 and the second imaging unit 31 are units that generate image data by imaging the vicinity of the work area of the robot 10 from different angles. The first imaging unit 30 and the second imaging unit 31 include, for example, a camera and are provided on a work table, a ceiling, a wall, or the like. In the present embodiment, the first imaging unit 30 and the second imaging unit 31 are provided on the work table. As the first imaging unit 30 and the second imaging unit 31, a visible light camera, an infrared camera, or the like can be employed.

  Next, a functional configuration example of the robot system 1 will be described. FIG. 2 is an example of a functional block diagram of the control device 20. The control device 20 mainly includes an information acquisition unit 201, a scenario selection unit 202, a robot drive unit 203, a library storage unit 204, and an execution log storage unit 205.

  The information acquisition unit 201 acquires the goal G and the speed level input from the input device 25 (see FIG. 6) or the like. The goal G is information indicating the content of work that the robot 10 wants to perform. The goal G includes work contents and variables (parameters) that define the work contents. Here, the work content has a format such as “Move A from XS to XE”, for example, and “Move A from XS to XE” means that the work A is moved from XS to XE. The parameters include, for example, parameters (work A, position XS, position XE) included in the work content and a speed level that is another parameter. The speed level is a speed at which the hand 14 is moved. For example, when the hand 14 is holding the work A, the speed level is a speed at which the work A is moved. The information acquisition unit 201 corresponds to the first acquisition unit of the present invention. The goal G corresponds to the instruction information of the present invention. The parameter included in the goal G corresponds to the second parameter of the present invention.

  The scenario selection unit 202 selects one arbitrary scenario based on the information stored in the library storage unit 204 (detailed later) and the execution log stored in the execution log storage unit 205 (detailed later). . In this embodiment, a group of operation commands that can be interpreted by the robot drive unit 203 is referred to as a “scenario”. A scenario is a combination of a plurality of robot unit motions necessary to realize the motion. For example, when the operation is “place work A at target position D”, the operation command group includes “move the hand to work A (to the position of work A)”, “hold work A”, “position “Move the hand to D”, “Release work A”, and the like. The processing performed by the scenario selection unit 202 will be described in detail later. The scenario selection unit 202 corresponds to the second acquisition unit, scenario group extraction unit, and scenario selection unit of the present invention.

  The robot drive unit 203 acquires information for specifying the current state of the robot 10 (for example, the joint angle, the position of the hand, the output value of a sensor such as a force sensor) from the robot 10, and the surroundings of the robot 10 Information for specifying the current state of the environment (for example, the position of the work A and the area where the work A occupies space, the position of the obstacle, the area where the obstacle occupies space, etc.) Obtain from the imaging unit 31. Then, the robot drive unit 203 controls the robot 10 to move the workpiece A in the scenario selected by the scenario selection unit 202 based on the acquired information. Since the process of the robot drive unit 203 can be realized by a known technique, a detailed description thereof is omitted. The robot drive unit 203 corresponds to the drive control unit of the present invention.

  The library storage unit 204 stores a plurality of scenarios composed of a plurality of operation instruction groups. FIG. 3 is a diagram illustrating an example of a data configuration recorded in the library storage unit 204.

  As illustrated in FIG. 3, the library storage unit 204 stores a scenario name, which is scenario identification information, work contents, and a group of operation commands that constitute the work contents, in association with each scenario. .

  Here, the scenario name, work content, and operation instruction group will be described by taking the scenario name as scenario 1 as an example. The work content of scenario 1 is “Move A from XS to XE”, which is the content of moving work A from XS to XE. The work content corresponds to the work content indicated by the scenario of the present invention.

  Work contents of scenario 1 are “Move A from XS to XE”, “Move to XS”, “Grasp A”, “Move from XS to X1”, “Move from X1 to XE”, and “Release A”. It consists of operation instructions (operation instruction group).

The operation command “Move to XS” is an operation command to move the position of the hand 14 to the position XS.
The operation command “Grasp A” is an operation command for gripping the workpiece A with the hand 14.
The operation command “Move from XS to X1” is an operation command for moving the position of the hand 14 from the position XS to the position X1.
The operation command “Move from X1 to XE” is an operation command to move the position of the hand 14 from the position X1 to the position XE.
The operation command “Release A” is an operation command for releasing the workpiece A held by the hand.

  Here, the workpiece A, the position XS, the position X1, and the position XE displayed in capital letters in the scenario are changeable parameters. This parameter corresponds to the first parameter of the present invention.

  FIGS. 4A, 4 </ b> B, and 4 </ b> C are diagrams schematically illustrating movement paths of the hand 14 when parameters are different for a plurality of scenarios. Numbers such as [1] and [2] in FIGS. 4A, 4B, and 4C indicate execution log numbers (detailed later, see FIG. 5). 4A, 4B, and 4C, the case where the hand 14 is moved in the scenario 1 is indicated by a solid line, and the case where the hand 14 is moved in the scenario 2 (see FIG. 3) is one point. This is indicated by a chain line, and a case where the hand 14 is moved in scenario 3 (see FIG. 3) is indicated by a two-dot chain line.

  As shown in FIG. 4, if the scenario is different, the hand 14 moves along a different route. For example, as shown in FIG. 3, the work contents of scenario 1, scenario 2, and scenario 3 are all “Move A from XS to XE”. However, scenario 1, scenario 2, and scenario 3 have different operation instruction groups. Therefore, as shown in FIG. 4, when hand 14 is moved in scenario 1, when hand 14 is moved in scenario 2, When the hand 14 is moved in the scenario 3, the hand 14 moves along a different route.

  Even in the same scenario, if the parameters are different, the hand 14 moves along a different route. For example, even when the hand 14 is moved in the scenario 1, if the parameters XS, X1, and XE are different, the hand 14 moves along different routes.

  3 and 4 show only an example of the scenario and the movement route of the hand 14, and are not limited to the cases shown in FIGS. 3 and 4.

  Returning to the description of FIG. The execution log storage unit 205 stores an execution log that is a result of the robot 10 executing in the past. As shown in FIG. 5, the execution log storage unit 205 mainly includes a number storage area 2051 for storing an execution log number for specifying an execution log, a scenario name storage area 2052 for storing a scenario name for specifying a scenario, An execution log table is stored in which a parameter storage area 2053 for storing a parameter for defining a scenario and an execution result storage area 2054 for storing a scenario execution result are associated with each other. The execution log is stored in the execution log table. ing.

  The parameter storage area 2053 has a storage area for storing parameter information for all parameters appearing in the scenario stored in the library storage unit 204. In the present embodiment, the parameter storage area 2053 mainly stores an A storage area 2053a in which information on the workpiece A is stored, an XS storage area 2053b in which information on the position XS is stored, and information on the position XE. An XE storage area 2053c, an X1 storage area 2053d for storing information about the position X1, an X2 storage area 2053e for storing information about the position X2, and a speed storage area for storing a speed level indicating the moving speed of the hand 14 2053f. Other parameters are not shown. The value stored in the parameter storage area 2053 corresponds to the selected value of the first parameter of the present invention. The speed level corresponds to the speed information of the present invention.

  Since the parameter storage area 2053 has an area for storing all parameters, there are parameters that are not used depending on the scenario. For parameters that are not used, for example, “−” indicating that they are not used is stored in the parameter storage area 2053.

  FIG. 6 is a block diagram illustrating an example of a schematic configuration of the control device 20. As shown in the figure, a control device 20 composed of, for example, a computer includes a CPU (Central Processing Unit) 21 that is an arithmetic device, a RAM (Random Access Memory) that is a volatile storage device, and a nonvolatile storage device. A memory 22 composed of a certain ROM (Read only Memory), an external storage device 23, a communication device 24 for communicating with an external device such as the robot 10, an input device 25 such as a mouse and a keyboard, and an output device such as a display 26, and an interface (I / F) 27 for connecting the control device 20 and other units.

  Each functional unit is realized, for example, when the CPU 21 reads out and executes a predetermined program stored in the external storage device 23 or the like, and the CPU 21 uses the memory 22 or the external storage device 23. . The predetermined program may be installed in advance in the external storage device 23 or the like, or may be downloaded from the network via the communication device 24 and installed or updated.

  The configuration of the robot system 1 described above is not limited to the above-described configuration because the main configuration has been described in describing the features of the present embodiment. The present invention is not limited by the way of classification and names of the constituent elements. The configuration of the control device 20 can be classified into more components depending on the processing content. Moreover, it can also classify | categorize so that one component may perform more processes. Further, the processing of each component may be executed by one hardware or may be executed by a plurality of hardware.

  Further, the robot 10 may include at least a part of functions of the control device 20, the first imaging unit 30, and the second imaging unit 31. Further, the configuration of a general robot system is not excluded.

  Next, a characteristic process of the robot system 1 having the above configuration in the present embodiment will be described.

  FIG. 7 is a flowchart showing a flow of scenario selection processing. This process is started at an arbitrary timing, for example, by inputting a scenario selection start instruction via a button or the like (not shown).

  The information acquisition unit 201 acquires the goal G input via the input device 25 or the like, and outputs it to the scenario selection unit 202 (step S100, corresponding to the first step of the present invention).

  The scenario selection unit 202 acquires the contents of the scenario from the library storage unit 204 and acquires the execution log stored in the execution log storage unit 205 (corresponding to the first step of the present invention). Then, the scenario selection unit 202 extracts the scenario group A from the execution log stored in the execution log storage unit 205 (step S102, corresponding to the second step of the present invention). Here, the process of step S102 will be described in detail.

FIG. 8 is a flowchart showing the flow of the scenario group A extraction process.
The scenario selection unit 202 sets J to 1 (step S1021). J is a numerical value (here, a natural number) indicating an execution log number stored in the execution log storage unit 205.

  The scenario selection unit 202 determines whether or not there is a Jth execution log in the execution log storage unit 205 (step S1022).

  If there is a Jth execution log in the execution log storage unit 205 (YES in step S102), the scenario selection unit 202 determines that the work content of the scenario in the Jth execution log is the goal G acquired in step S100. It is determined whether or not it matches the work content (step S1023). The process in step S1023 will be described by taking as an example a case where J = 1 and the work content of the goal G acquired in step S100 is “Move A from XS to XE”.

  Since J = 1, the scenario selection unit 202 acquires the first execution log from the execution log storage unit 205. Since the scenario name of the first execution log is scenario 1, the scenario selection unit 202 acquires the work contents of scenario 1 from the library storage unit 204. The work content of scenario 1 is “Move A from XS to XE”.

  In this embodiment, since only the scenario name is included in the execution log storage unit 205 and the details of the scenario are stored in the library storage unit 204, the scenario selection unit 202 includes the library storage unit 204 and the execution Information is acquired from the log storage unit 205. However, when the details of the scenario are stored in the execution log storage unit 205, the scenario selection unit 202 only needs to acquire information from the execution log storage unit 205, and acquires information from the library storage unit 204. There is no need.

  The scenario selection unit 202 compares the work content “Move A from XS to XE” of the scenario 1 with the work content “Move A from XS to XE” of the goal G. In this case, the work content “Move A from XS to XE” of scenario 1 matches the work content “Move A from XS to XE” of goal G. It is determined that the work content matches the work content of the goal G acquired in step S100.

  When the work content of the Jth execution log matches the goal G (YES in step S1023), the scenario selection unit 202 calculates the similarity between the Jth execution log and the goal G (step S1024). . Regarding the processing of step S1024, J = 1, the work content of the goal G is “Move A from XS to XE”, the parameters of the goal G are A for the sphere 11, XS for (15, 50, 0), A case where XE is (60, 0, 0) and the speed level is 2 will be described as an example. The sphere 11 of A indicates the type of the object to be grasped, and (15, 50, 0) of XS and (60, 0, 0) of XE indicate coordinates in a three-dimensional space. The contents of the sphere 11 are stored in the memory 22 or the like. The value of the goal G parameter corresponds to the selected value of the second parameter of the present invention.

The scenario selection unit 202 calculates the similarity between the first execution log and the goal G by substituting the parameter of the goal G and the parameter of the first execution log into Equation (1).
[Equation 1]
Similarity = Σw [i] / (Goal.Param [i] −Log.Param [i]) (1)

  Here, w [i] is a weighted fixed value of each parameter i, and can be an arbitrary value for each parameter. Goal. Param [i] is the value of the parameter i of the goal G, and Log. Param [i] is the value of parameter i of the scenario. Thus, Equation (1) takes the reciprocal of the difference for each parameter, and adds them with a predetermined weight.

  The parameters of the No. 1 execution log are A for sphere 10, XS (50, 10, 0), XE (20, 50, 0), and speed level 2 (see FIG. 5). In this embodiment, since all w [i] are 1, the reciprocal of the difference between the sphere 11 and the sphere 10 and the reciprocal of the distance between the coordinates (15, 50, 0) and the coordinates (50, 10, 0). And the reciprocal of coordinates (60, 0, 0) and coordinates (20, 50, 0) and the reciprocal of the difference between speed level 2 and speed level 2 (that is, 0) is calculated as the similarity. The Here, as the values of the sphere 10 and the sphere 11, for example, values such as the size of the sphere and the weight of the sphere stored in the memory 22 can be used for calculating the similarity. For the speed level, for example, a speed value stored in the memory 22 or the like can be used.

  The formula (1) is an example of a formula in which the similarity decreases as the difference increases, and the formula used for calculating the similarity is not limited thereto. As the formula used for calculating the similarity, various types of formulas can be used as long as the difference in parameters is large and the similarity becomes small.

  In the formula (1), w [i] is not limited to 1. For example, for the speed level, w [i] may be set so that the weighting is heavy with respect to other parameters.

  The scenario selection unit 202 determines whether the similarity between the Jth execution log and the goal G is equal to or greater than a threshold value (arbitrary value) (step S1025).

  If the similarity between the Jth execution log and the goal G is equal to or greater than the threshold (YES in step S1025), the scenario selection unit 202 adds the Jth execution log to the scenario group A (step S1026).

  When step S1026 is performed, the work content of the scenario of the Jth execution log does not match the goal G (NO in step S1023), and the similarity between the Jth execution log and the goal G is a threshold (optional) If the value is not greater than or equal to (NO in step S1025), the scenario selection unit 202 adds 1 to J (step S1027), and performs step S1022 again.

  If there is no J-th execution log in the execution log storage unit 205 (NO in step S1022), the scenario selection unit 202 ends the scenario group A extraction process (step S102).

  In the present embodiment, the goal G is acquired first, and then the information in the library storage unit 204 and the execution log storage unit 205 is acquired. However, the information acquisition order is not limited to this. The goal G may be acquired after the information of the library storage unit 204 and the execution log storage unit 205 is acquired first, or all the information may be acquired simultaneously.

  The scenario selection unit 202 selects a scenario S from the scenario group A selected in step S102 (step S104, corresponding to the third step of the present invention). Here, the process of step S104 will be described in detail.

FIG. 9 is a flowchart showing the flow of the scenario S selection process.
The scenario selection unit 202 sets K to 1 and sets the maximum success rate to 0 (step S1041). Here, K is a number indicating the scenario name of the scenario stored in the library storage unit 204. For example, the case where K is 1 is the case of scenario 1.

  The scenario selection unit 202 determines whether there is a scenario whose scenario name is scenario K in the execution logs included in the scenario group A (step S1042).

  When there is an execution log of scenario K in scenario group A (YES in step 1042), scenario selection unit 202 calculates the success rate of scenario K (step S1043). The process of step S1043 will be described using a specific example.

  For example, when K is 1 and the execution logs with numbers 1 to 10, 32, 33, 87, and 146 are included in the scenario group A in the execution log table shown in FIG. 202 extracts, from the scenario group A, numbers 1, 4, 7, and 10 whose scenario name is scenario 1 from the execution logs of numbers 1 to 10, 32, 33, 87, and 146.

  The scenario selection unit 202 calculates the success rate of the execution logs of numbers 1, 4, 7, and 10 extracted from the scenario group A. Referring to the execution result storage area 2054 of the execution log table, the result of the execution log of number 7 is successful, but the results of numbers 1, 4, and 10 are failure. Therefore, the scenario selection unit 202 calculates the success rate as ¼.

  The scenario selection unit 202 determines whether or not the success rate calculated in step S1043 is larger than the maximum success rate (step S1044).

  If the success rate calculated in step S1043 is larger than the maximum success rate (YES in step S1044), the scenario selection unit 202 sets scenario K as scenario S and the success rate of scenario K as the maximum success rate (step S1044). S1045). When steps S1044 and S1045 are performed first, the maximum success rate is set to 0 in step S1041. Therefore, unless the success rate of scenario 1 is 0, scenario 1 is set as scenario S, and the success rate of scenario 1 is Maximum success rate.

  When step S1045 is performed and when the success rate calculated in step S1043 is not greater than the maximum success rate (NO in step S1044), the scenario selection unit 202 adds 1 to K (step S1046). Step S1042 is performed again.

  If there is no execution log for scenario K in scenario group A (NO in step S1042), scenario selection unit 202 determines whether or not the success rate has been calculated for all scenarios (step S1047). For example, if the scenarios stored in the library storage unit 204 are scenarios 1 to 10, the scenario selection unit 202 determines that all scenarios have been processed when K is greater than 10. You may do it.

  If processing has not been completed for all scenarios (NO in step S1047), scenario selection unit 202 performs step S1042 again. When the processing is completed for all scenarios (YES in step S1047), the scenario selection unit 202 ends the scenario S selection processing (step S104).

  The scenario selection unit 202 outputs the scenario S selected in step S104 to the robot drive unit 203. The robot drive unit 203 controls the robot 10 to move the robot 10 in the scenario S output from the scenario selection unit 202 (step S106, corresponding to the fourth step of the present invention).

  The robot drive unit 203 determines whether or not the result of controlling the robot 10 in step S106 is successful, and stores the result and the scenario S in association with each other in the execution log storage unit 205 (step S108). For example, when the operation command group of the scenario S can be executed to the end, the robot drive unit 203 can determine that the execution of the scenario S is successful.

  According to the present embodiment, when a robot performs a task such as movement of a grasped object, a scenario suitable for the task and a scenario with a high success rate can be selected based on a past execution result. . In addition, it is possible to select and execute work results with a high success rate according to the situation (workpiece size set as parameters, start and goal positions, etc.) from past execution results, improving process reliability. Can be made.

  In practice, it is rare that the work is performed again in an exactly matched environment (work size, start and goal positions, etc.). Therefore, the robot can be controlled realistically, practically, versatilely and flexibly by selecting work results that have been successful in the past in a closer environment as in the present embodiment.

  Further, in this embodiment, speed information that has a large influence on execution is used as a parameter, and is used for calculation of similarity, so that it can be used for extraction of execution results in an environment close to the target work.

<Second Embodiment>
In the first embodiment of the present invention, the similarity including the speed level is calculated. However, the speed level has a greater influence on the work result than other parameters.

In the second embodiment of the present invention, the scenario S is selected using only execution logs having the same speed level. Less than. A robot system 2 according to the second embodiment will be described. Since the configuration of the robot system 2 is the same as that of the robot system 1, description thereof is omitted. Further, the same parts as those of the first embodiment are denoted by the same reference numerals, and description thereof is omitted.
A characteristic process of the robot system 2 configured as described above in the present embodiment will be described.

  FIG. 10 is a flowchart showing the flow of scenario selection processing. This process is started at an arbitrary timing, for example, by inputting a scenario selection start instruction via a button or the like (not shown).

  The information acquisition unit 201 acquires the goal G input via the input device 25 or the like, and outputs it to the scenario selection unit 202 (step S100).

  The scenario selection unit 202 extracts the scenario group A from the execution log stored in the execution log storage unit 205 (step S103). Here, the process of step S103 will be described in detail.

FIG. 11 is a flowchart showing the flow of the scenario group A extraction process.
The scenario selection unit 202 sets J to 1 (step S1031). The scenario selection unit 202 determines whether or not there is a Jth execution log in the execution log storage unit 205 (step S1032). When there is a Jth execution log in the execution log storage unit 205 (YES in step S1032), the scenario selection unit 202 determines that the work content of the scenario of the Jth execution log is the goal G acquired in step S100. It is determined whether or not the content matches the work content (step S1033). Note that the process in step S1031 is the same as the process in step S1021, the process in step S1032 is the same as the process in step S1022, and the process in step S1033 is the same as the process in step S1023.

  When the work content of the Jth execution log matches the goal G (YES in step S1033), the scenario selection unit 202 refers to the speed storage area 2053f of the execution log table stored in the execution log storage unit 205. Then, it is determined whether or not the speed level of the No. J execution log matches the speed level of the goal G (step S1034).

  If the speed level of the Jth execution log matches the speed level of the goal G (YES in step S1034), the scenario selection unit 202 calculates the similarity between the Jth execution log and the goal G (step S1034). S1035). The process of step S1035 is not described in detail because the speed level is not included in the parameter of the similarity calculation formula (1) with respect to the process of step S1024.

  The scenario selection unit 202 determines whether the similarity between the Jth execution log and the goal G is equal to or greater than a threshold (arbitrary value) (step S1036). The process of step S1036 is the same as the process of step S1025.

  If the similarity between the Jth execution log and the goal G is equal to or greater than a threshold (arbitrary value) (YES in step S1036), the scenario selection unit 202 adds the Jth execution log to the scenario group A. (Step S1037). The process of step S1037 is the same as that of step S1026.

  When step S1026 is performed, if the work content of the scenario of the Jth execution log does not match the goal G (NO in step S1033), the speed level of the Jth execution log does not match the speed level of the goal G If this is the case (NO in step S1034), and if the similarity between the Jth execution log and the goal G is not equal to or greater than a threshold (arbitrary value) (NO in step S1036), the scenario selection unit 202 sets 1 to J. It adds (step S1038) and performs step S1032 again. The processing in step S1038 is the same as that in step S1027.

If there is no J-th execution log in the execution log storage unit 205 (NO in step S1032), the scenario selection unit 202 ends the scenario group A extraction process (step S103).
The scenario selection unit 202 selects a scenario S from the scenario group A selected in step S102 (step S104).

  The scenario selection unit 202 outputs the scenario S selected in step S104 to the robot drive unit 203. The robot drive unit 203 controls the robot 10 to move the robot 10 in the scenario S output from the scenario selection unit 202 (step S106).

  The robot drive unit 203 determines whether or not the result of controlling the robot 10 in step S106 is successful, and stores the result and the scenario S in association with each other in the execution log storage unit 205 (step S108).

  According to the present embodiment, since the scenario S is selected using only the execution log whose speed level matches the speed level of the goal G, a more appropriate scenario can be selected.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be made to the above embodiment. In addition, it is apparent from the scope of the claims that the embodiments added with such changes or improvements can be included in the technical scope of the present invention. In particular, in the first and second embodiments, a case where a robot system in which a robot and a control device are separately provided is illustrated, but the present invention is a robot system in which a robot and a control device are separately provided. Or a robot including a control device or the like in the robot, or a robot control device including the control device. The present invention can also be provided as a program for controlling a robot or the like or a storage medium storing the program.

1, 2: Robot system, 10: Robot, 11: Arm, 12: Joint, 13: Link, 14: Hand, 20: Control device, 21: CPU, 22: Memory, 23: External storage device, 24: Communication device 25: input device, 26: output device, 30: first imaging unit, 31: second imaging unit, 201: information acquisition unit, 202: scenario selection unit, 203: robot drive unit, 204: library storage unit, 205: execution log storage unit, 2051: number storage area, 2052: scenario name storage area, 2053: parameter storage area, 2053a: A storage area, 2053b: XS storage area, 2053c: XE storage area, 2053d: X1 storage area, 2053e: X2 storage area, 2053f: Speed storage area, 2054: Execution result storage area

Claims (7)

A robot control method for causing a robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined.
Obtaining first execution information including a plurality of execution logs in which a scenario executed by the robot, a selected value of a first parameter defining the scenario, and an execution result of the scenario are associated; A first step of obtaining instruction information including a content of the work and a selected value of a second parameter that defines the target work;
Based on the first execution information, an execution log in which the work content indicated by the scenario is the same as the content of the target work and the selected value of the first parameter is similar to the selected value of the second parameter is stored in the second execution log. A second step of extracting as execution information of
A third step of selecting a scenario with the highest success rate based on the execution result from the scenarios included in the second execution information;
A fourth step of outputting the selected scenario;
A robot control method comprising: The robot control method according to claim 1,
In the first step, speed information that is information about a speed at which the robot is moved is acquired as the first parameter and the second parameter. The robot control method according to claim 1,
The execution log and the instruction information include speed information that is information about a speed at which the robot moves.
In the second step, an execution log associated with the same speed information as the speed information of the instruction information is extracted from the first execution information. The robot control method according to any one of claims 1 to 3,
In the second step, based on the similarity calculated using a calculation formula that decreases as the difference between the selected value of the first parameter and the selected value of the second parameter increases, A robot control method characterized by extracting execution information. A robot control device for causing the robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined.
A first acquisition unit that acquires first execution information including a plurality of execution logs that associate a scenario executed by the robot, a selected value of a first parameter that defines the scenario, and an execution result of the scenario. When,
A second acquisition unit that acquires instruction information including the content of the target work and a selection value of a second parameter that defines the target work;
Based on the first execution information, an execution log in which the work content indicated by the scenario is the same as the content of the target work and the selected value of the first parameter is similar to the selected value of the second parameter is stored in the second execution log. A scenario group extraction unit that extracts the execution information of
A scenario selection unit that selects a scenario having the highest success rate based on the execution result from the scenarios included in the second execution information;
An output unit for outputting the selected optimum scenario;
A robot control device comprising: A robot that causes a movable part of the robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined,
A first acquisition unit that acquires first execution information including a plurality of execution logs that associate a scenario executed by the robot, a selected value of a first parameter that defines the scenario, and an execution result of the scenario. When,
A second acquisition unit that acquires instruction information including the content of the target work and a selection value of a second parameter that defines the target work;
Based on the first execution information, an execution log in which the work content indicated by the scenario is the same as the content of the target work and the selected value of the first parameter is similar to the selected value of the second parameter is stored in the second execution log. A scenario group extraction unit that extracts the execution information of
A scenario selection unit that selects a scenario having the highest success rate based on the execution result from the scenarios included in the second execution information;
A drive control unit that controls the movable unit using the selected optimal scenario;
A robot characterized by comprising: A robot control program for causing a robot to perform a desired work using a scenario in which a plurality of robot unit operations are combined.
Obtaining first execution information including a plurality of execution logs in which a scenario executed by the robot, a selected value of a first parameter defining the scenario, and an execution result of the scenario are associated; A first step of obtaining instruction information including a content of the work and a selected value of a second parameter that defines the target work;
Based on the first execution information, an execution log in which the work content indicated by the scenario is the same as the content of the target work and the selected value of the first parameter is similar to the selected value of the second parameter is stored in the second execution log. A second step of extracting as execution information of
A third step of selecting a scenario with the highest success rate based on the execution result from the scenarios included in the second execution information;
A fourth step of outputting the selected optimal scenario;
A robot control program that causes an arithmetic device to execute the above.

Images