WO2016135861A1 - Manipulator, motion planning method for manipulator, and control system for manipulator - Google Patents

Manipulator, motion planning method for manipulator, and control system for manipulator Download PDF

Info

Publication number
WO2016135861A1
WO2016135861A1 PCT/JP2015/055256 JP2015055256W WO2016135861A1 WO 2016135861 A1 WO2016135861 A1 WO 2016135861A1 JP 2015055256 W JP2015055256 W JP 2015055256W WO 2016135861 A1 WO2016135861 A1 WO 2016135861A1
Authority
WO
WIPO (PCT)
Prior art keywords
joint
arm
manipulator
unit
function
Prior art date
Application number
PCT/JP2015/055256
Other languages
French (fr)
Japanese (ja)
Inventor
潔人 伊藤
宣隆 木村
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to PCT/JP2015/055256 priority Critical patent/WO2016135861A1/en
Priority to JP2017501607A priority patent/JP6359756B2/en
Publication of WO2016135861A1 publication Critical patent/WO2016135861A1/en

Links

Images

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J5/00Manipulators mounted on wheels or on carriages
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J13/00Controls for manipulators
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00Programme-control systems
    • G05B19/02Programme-control systems electric
    • G05B19/42Recording and playback systems, i.e. in which the programme is recorded from a cycle of operations, e.g. the cycle of operations being manually controlled, after which this record is played back on the same machine

Definitions

  • the present invention relates to a method for generating a motion of a mobile manipulator that operates in an unknown environment, and an apparatus and system using the method.
  • Autonomous mobile manipulators are expected to be put to practical use as facilities that support or replace human activities in facilities such as warehouses and factories. Such an autonomous mobile manipulator is required to be as flexible and quick as humans.
  • a moving manipulator is attached to a moving carriage part that moves itself by moving forward, backward, or turning, an arm part that has one or more rotary joint mechanisms attached to the moving carriage part, and a tip of the arm part. And an end effector portion.
  • the end effector unit can hold or move a specific object. For example, a robot hand or a magnet. Such gripping or movement of the object by the end effector unit is referred to as an operation on the object.
  • the end effector unit is changed to a predetermined position and direction (hereinafter, the position and direction are adjusted by changing the rotation joint angle of the arm unit).
  • the position and direction are adjusted by changing the rotation joint angle of the arm unit.
  • a method of determining a time-series change (hereinafter referred to as an arm trajectory) of the joint angle value of the arm unit for moving from the current end effector posture to a target end effector posture is referred to as an operation plan.
  • Non-Patent Documents 1 and 2 In order to plan the motion of the arm trajectory from the current configuration to the target configuration, a technique called a sample-based search motion planning method using random numbers as disclosed in Non-Patent Documents 1 and 2 has been proposed. To explain this simply, using random numbers, it repeatedly generates various forms that the arm can take, and searches them for time-sequential forms based on graph theory. It is a method of generating.
  • the positional relationship between the manipulator and the target object is not limited, so it is not possible to teach in advance the target form of the arm that covers all situations. Therefore, in the mobile manipulator, it is necessary to autonomously generate the target form of the arm unit according to the situation and autonomously perform the operation plan from the current form to the target form.
  • Non-Patent Documents 1 and 2 can generate an arm trajectory by exhaustively searching for possible forms of all arm parts, including the target form of the arm part.
  • searching exhaustively including the target form in this way an enormous number of iterative processes are required. For this reason, it takes a long time to generate an appropriate movement path, or if the search is terminated in a short time, an unnatural arm trajectory may be generated that greatly reaches the target form. Is an issue. Therefore, in order to finish the arm trajectory search in a short time, it is effective to specify the target form uniquely.
  • Patent Document 1 determines the target arm form based on the degree of manipulability, the arm form cannot be uniquely determined. This is because an arm unit having a plurality of rotary joints may have a plurality of forms having the same operable degree.
  • a manipulator determines an appropriate target form of an arm part at the time of operation, and based on that, a manipulator control method for performing an operation plan capable of natural and quick operation of the arm part Is to provide.
  • a typical one is the most singular point form as a singular point form determined from a reference line, with a straight line connecting the rotation axis of the first joint of the manipulator and a predetermined reference point as a reference straight line.
  • An operation planning method for a manipulator characterized by searching for an arm trajectory from a predetermined form to a target form with a form at a short distance as a target form.
  • Another aspect of the present invention is executed by an information processing apparatus to position a reference point associated with an arm unit at a desired position in a manipulator including an arm unit including a plurality of rotary joints and an end effector unit.
  • This is an operation planning method.
  • a first process for setting a straight line connecting the rotation axis of the first joint of the manipulator and the reference point as a reference straight line a second process for setting the form of the arm portion determined from the reference straight line as a singular point form, Of the forms that the arm unit can take to position the reference point at a desired position, the third process for setting the form of the arm part closest to the singular point form as the target form, and the arm trajectory to the target form
  • a fourth process for searching is a fourth process for searching.
  • Another aspect of the present invention includes a movable carriage unit, an arm unit mounted on the movable carriage unit, and a control unit that controls the operation of the arm unit. And a second joint, and the control unit sets a reference point in association with the arm part on the opposite side of the moving carriage unit with respect to the second joint, and controls the movement of the arm unit that positions the reference point at a desired position.
  • the form of one arm part from a plurality of candidate forms A fourth function that is selected and set as the target form and a fifth function that determines the operation of the first joint and the second joint to reach the target form are provided.
  • Another aspect of the present invention is a manipulator having a movable carriage part and an arm part mounted on the movable carriage part, and the arm parts are a first joint, a second joint, and a third joint in order from the movable carriage side.
  • the arm portion is driven by a movement control device that positions a reference point set at a portion of the third joint opposite to the moving carriage portion at a desired position.
  • the movement control includes a first function for calculating, as a plurality of candidate forms, possible arm portions for positioning the reference point at a desired position, and a first straight line for connecting the reference point and the first joint.
  • the form of one arm part is selected from a plurality of candidate forms.
  • the fourth function that is selected and set as the target form and the fifth function that determines the operations of the first joint, the second joint, and the third joint to reach the target form are executed.
  • Another aspect of the present invention is an article operation system that automatically moves the movable carriage part to a desired position and controls the arm part to operate the article after reaching the desired position.
  • an article operation method is included. In such a system or method, the location where the article to be operated is placed is stored in advance, the moving carriage is moved to the vicinity thereof, and then the arm unit is controlled to operate the article.
  • the present invention can uniquely determine the form that the arm portion should take in performing a desired operation by using the criterion of the singular point form.
  • the number of joints may be three or more.
  • a mechanism for operating an article such as a robot hand, a magnet, or the like may be provided as an end effector at the tip of the arm portion or the like.
  • Various sensors may be provided. By detecting and controlling the position of the reference point and the arm with the sensor, more complicated operations such as avoiding an obstacle can be performed.
  • the distance to the singular point form there are various methods for calculating the distance to the singular point form, but as an example, the square of the difference between the angle of each rotary joint mechanism in the singular point form and the angle of each rotary joint mechanism in the form that the arm can take If the distance is evaluated using the square root of the sum of the values, it can be easily calculated using the data normally used by the system that controls the arm unit.
  • control parts which control an arm part may be mounted in a mobile trolley part. Alternatively, some or all of them may be outside the mobile carriage unit and connected by wireless or wired lines.
  • the manipulator according to the present invention can autonomously determine a target posture of an appropriate arm portion during operation, and can generate a trajectory of the arm portion capable of natural and quick operation based on the target posture.
  • FIG. 1 is a perspective view illustrating an example of a configuration of a picking robot system 1.
  • FIG. 5 is a flowchart showing an example of a series of operations for picking an article in the picking robot system 1.
  • 6 is a table showing an example of data included in a picking instruction 401.
  • FIG. 6 is a table showing an example of data included in a movement instruction 402.
  • FIG. 3 is a side view showing an example of a specific configuration of the autonomous mobile manipulator 100.
  • FIG. It is the block diagram which showed an example of the specific structure of the autonomous mobile manipulator control part.
  • 5 is a flowchart showing an example of an operation of moving to the front of the shelf of the autonomous mobile manipulator 100.
  • 5 is a flowchart showing an example of picking article posture observation operation of the autonomous mobile manipulator 100.
  • FIG. 4 is a flowchart illustrating an example of a picking operation of the autonomous mobile manipulator 100.
  • 6 is a block diagram showing an example of a specific configuration of an operation planning unit 1150.
  • FIG. 4 is a schematic diagram illustrating an example of a picking operation of the autonomous mobile manipulator 100.
  • FIG. 5 is a flowchart showing an example of the operation of an operation planning unit 1150. It is the schematic diagram which showed the example of the arm part form which grasps
  • FIG. 6 is a block diagram showing an example of a specific configuration of an operation planning unit 1150.
  • notations such as “first”, “second”, and “third” are attached to identify the constituent elements, and do not necessarily limit the number or order.
  • a number for identifying a component is used for each context, and a number used in one context does not necessarily indicate the same configuration in another context. Further, it does not preclude that a component identified by a certain number also functions as a component identified by another number.
  • a picking robot system that performs picking work of articles from a shelf by an autonomous mobile manipulator assuming the taking in and out of goods in a mail order warehouse will be specifically described.
  • the picking work refers to a work of taking out designated articles from the shelves in which they are stored.
  • FIG. 1 is a diagram showing an example of a configuration of a picking robot system 1 according to the present embodiment.
  • the picking robot system 1 includes one or more autonomous mobile manipulators 100, one or more shelves 200, one or more transfer robots 300, at least one management computer 400, and an input terminal 500.
  • the autonomous moving manipulator 100 is attached to a moving carriage 101 for moving itself by moving forward, backward, turning, etc., an arm 102 having one or more rotary joints attached to the moving carriage 101, and a tip of the arm.
  • the end effector unit 103 and the autonomous mobile manipulator control unit 110 are provided.
  • the end effector unit 103 can hold or move a predetermined object.
  • one or more articles 900 are stored on the shelf board 210 and installed in a predetermined rack arrangement area 2000 in the picking robot system 1.
  • the transfer robot 300 is configured to be able to move itself by moving forward, backward, turning, etc., like the moving carriage unit 101 of the autonomous moving manipulator 100. Further, the transfer robot 300 has a structure in which the upper part is a flat surface and the picking tray 310 can be loaded. The autonomous mobile manipulator 100 and the transfer robot 300 can be combined to form a single device.
  • the management computer 400 manages and operates the entire picking robot system 1.
  • the management computer 400 and the autonomous mobile manipulator 100 are configured to be able to communicate wirelessly, and have a configuration in which a transmission unit and a reception unit (not shown) are provided.
  • the management computer 400 and the transfer robot 300 are configured to be able to communicate wirelessly, and are each configured to include a transmission unit and a reception unit that are not illustrated.
  • the input terminal 500 is mounted as a PC or the like, and the input terminal 500 and the management computer 400 can communicate with each other by wire or wireless, and have a configuration in which a transmission unit and a reception unit (not shown) are provided.
  • the input terminal 500 has been described as being implemented by a PC or the like, the present invention is not limited to this, and any medium that can communicate with the management computer 400 by wire or wireless may be used.
  • various forms such as a mobile phone device and a tablet terminal can be used, and a server system including a plurality of computers may be used.
  • FIG. 2 is a flowchart showing an example of a series of operations for picking an article in the picking robot system 1 according to the present embodiment. This drawing shows an example in which one picking robot and one transfer robot operate for simplicity.
  • the part surrounded by the broken line A1 is related to the operation of the input terminal 500
  • the part surrounded by the broken line A2 is related to the operation of the management computer 300
  • the part surrounded by the broken line A3 is related to the autonomous mobile manipulator 100.
  • the operations surrounded by the broken line A4 indicate the operations applied to the transfer robot 300, respectively.
  • the operator uses the input terminal 500 to input an order for one or more items shipped to the management computer 400 (S11).
  • the management computer 400 can hold or access article management information data such as which article 900 is on which shelf 200 (or which shelf board 210). Further, the arrangement (position) information data of the shelf 200 can also be held or accessed.
  • the management computer 400 determines the order of picking the shipment items from the installation location of the rack 200 in which the shipment items are stored, and creates a picking item list (S21). Then, according to the order of the articles to be picked, the autonomous mobile manipulator 100 and the transport robot do not interfere with the travel path between the autonomous mobile manipulator 100 and the transport robot 300 or increase the picking time by taking a detour path. Each movement route is planned (S22). Then, a picking instruction 401 is transmitted to the autonomous mobile manipulator 100, and a movement instruction 402 is transmitted to the transfer robot 300 (S23 and S24).
  • an example of data included in the picking instruction 401 is shown in FIG. 3
  • an example of data included in the movement instruction 402 is shown in FIG.
  • the picking instruction 401 contains an order ID that is an identifier uniquely assigned to each order, a picking item ID that is an identifier uniquely assigned to each type of picking article, and a picking article.
  • transfer robot stop point data indicating a position where the transfer robot 300 is stopped.
  • the movement instruction 402 includes an order ID that is an identifier uniquely assigned to each order and movement route data for moving to the vicinity of the autonomous movement manipulator 100 to be picked.
  • the autonomous mobile manipulator 100 When receiving the picking instruction 401 from the management computer 400, the autonomous mobile manipulator 100 first moves to the front of the rack 200 in which the picking article is stored in accordance with the movement path data included in the picking instruction 401 (S31).
  • the front of the rack 200 is a range in which the picking article designated by the picking instruction 401 can be picked from the rack 200 by the arm unit 102 provided in the autonomous mobile manipulator 100.
  • the movement to the front of the rack 200 in which the picking article is stored is completed, the detailed posture (position and direction) of the article on the shelf plate 210 of the rack 200 is observed (S32), and the observed article position and Picking is performed based on the direction (S33). The details of the operation related to picking the article from S31 to S33 will be described later.
  • the transfer robot 300 when receiving the movement instruction 402 from the management computer 400, the transfer robot 300 is in the vicinity of the autonomous movement manipulator 100 that performs picking of the article according to the movement route data included in the movement instruction 402.
  • the vicinity is within the reach of the arm unit 102 provided in the autonomous mobile manipulator 100.
  • the management computer 400 transmits a storage instruction to the autonomous mobile manipulator 100 (S25).
  • the autonomous mobile manipulator 100 stores the picked article in the picking tray 310 on which the transport robot 300 is loaded (S33).
  • the management computer 400 checks whether all of the articles included in the picking article list created in step S21 have been picked (S26), and if not completed, repeats a series of processes from step S22 to S26. .
  • the management computer 400 transmits a movement instruction 402 including movement route data to the shipping area to the transport robot 300 (S27).
  • the transfer robot 300 moves to the shipping area according to the movement route data included in the movement instruction 402 (S42), and the picking operation is completed.
  • FIG. 5 is a diagram illustrating an example of a specific configuration of the autonomous mobile manipulator 100 according to the present embodiment.
  • the movable carriage unit 101 further includes two drive wheels 1011 attached to the left and right with respect to the traveling direction (left hand direction in the drawing) of the movable carriage, two casters 1012 provided in the front and rear, and an ambient environment measurement sensor 1013. . Assuming that the two drive wheels 1011 operate independently, by controlling the rotational speed of the drive wheels 1011, the movable carriage unit 101 can move in a predetermined plane such as forward, backward, and turn. is there.
  • the casters 1012 are provided at the front and rear in order to prevent the vehicle body from tilting back and forth as the movable carriage unit travels.
  • the traveling configuration of the autonomous mobile manipulator 100 is not limited to the configuration of the moving carriage unit 101, and may be any as long as it can move to any position and orientation on the floor surface.
  • a wheel having a steering for changing the wheel axis may be used, an omni wheel capable of omnidirectional movement may be used, or a leg type mechanism may be used.
  • the ambient environment measurement sensor 1013 is mounted as a laser distance sensor that measures the distribution of obstacle positions on a horizontal plane.
  • the ambient environment measurement sensor 1013 in the present invention is not limited to the laser distance sensor, and may be any sensor that can acquire sufficient information to recognize the position and orientation of the autonomous mobile manipulator 100.
  • the arm unit 102 is fixed to the movable carriage unit 101, and six rotary joint mechanisms 1021, 1022, 1023, 1024, 1025, and 1026 are connected in series from the fixed end.
  • the joint mechanisms 1021, 1022, 1023, 1024, 1025, and 1026 each have a function of rotating a portion that is in the extending direction from itself in the directions indicated by arrows J1, J2, J3, J4, J5, and J6 in the drawing. . That is, with the six degrees of freedom of rotation, the end effector portion 103 attached to the tip of the arm portion 102 can be moved to an arbitrary position and posture within a predetermined three-dimensional space range.
  • the end effector unit 103 includes two finger mechanisms 1031a and 1031b in the extending direction, with the direction fixed to the arm unit as a base end, and controls the finger mechanisms 1031a and 1031b so as to have an arbitrary size.
  • the object can be gripped.
  • the configuration of the end effector portion 103 is not limited to this as long as it can grip the picking article.
  • a multi-finger gripper including three or more finger mechanisms 1031 or a suction gripper that grips an object by vacuum suction may be used.
  • the end effector unit 103 further includes an article measurement sensor 120.
  • the article measurement sensor 120 is implemented as a distance image sensor (also referred to as a depth camera) that measures the distribution of obstacle positions in the field of view.
  • the article measurement sensor 120 in the present invention is not limited to the distance image sensor, and may be any sensor that can acquire sufficient information for recognizing the posture of the article to be picked. It may be mounted as a combination of a plurality of sensors.
  • the article measurement sensor 120 is attached on the end effector 103 and the field of view is directed in the extending direction of the finger mechanism 1031.
  • the attachment method of the article measurement sensor 120 in the present invention is not limited to this. Not a thing. For example, an attachment method of preventing interference between the visual field of the article measurement sensor 120 and the finger mechanism 1031 by providing an elevation angle with respect to the extending direction of the finger mechanism 1031 may be used.
  • FIG. 6 is a diagram illustrating an example of a specific configuration of the autonomous mobile manipulator control unit 110.
  • the autonomous mobile manipulator control unit 110 includes, as functional blocks, an overall management unit 1100, a transmission unit 1101, a reception unit 1102, an image recognition unit 1110, an article database 1111, a space management unit 1120, a work procedure planning unit 1130, A work procedure database 1131, an operation execution unit 1140, an operation plan unit 1150, an arm control unit 1160, an end effector control unit 1170, a travel management unit 1180, an ambient environment map 1182, and a mobile cart control unit 1182 are provided.
  • an overall management unit 1100 includes, as functional blocks, an overall management unit 1100, a transmission unit 1101, a reception unit 1102, an image recognition unit 1110, an article database 1111, a space management unit 1120, a work procedure planning unit 1130, A work procedure database 1131, an operation execution unit 1140, an operation plan unit 1150, an arm control unit 1160, an end effector control unit 1170
  • the overall management unit 1100 has a function of giving a detailed processing command to each functional block included in the autonomous mobile manipulator control unit 110 based on the picking command 402 from the management computer 400 received by the receiving unit 1102. At the same time, the status of the autonomous mobile manipulator 100 is transmitted to the management computer 400 based on the processing result of each functional block.
  • the image recognition unit 1110 uses the distribution of obstacle positions in the field of view measured by the article measurement sensor 120, article information stored in the article database 1111, and a picking article ID given from the overall management unit 1100. It has a function of calculating the posture of an article to be picked or an obstacle in the field of view by predetermined image processing.
  • the space management unit 1120 stores the posture of an article or obstacle to be picked calculated by the image recognition unit 1110 (referred to as surrounding object posture information), and supplies a function to each functional block as necessary. Have.
  • the work procedure planning unit 1130 is a functional block that plans a picking work procedure.
  • the picking work procedure indicates a procedure that defines in which posture and in which order the arm unit 102 or the end effector unit 103 included in the autonomous mobile manipulator 100 is moved within a predetermined space range.
  • the work procedure database 1131 stores work procedure templates that are determined according to the shape of the article to be picked and a predetermined space range in which the article is stored such as the height of the shelf board 210.
  • the work procedure planning unit 1130 uses the picking article ID and shelf ID given from the overall management unit 1100, the work procedure template stored in the work procedure database 1141, and the surrounding object posture information stored in the entire space management unit 1120. Plan the picking work procedure.
  • the operation execution unit 1140 is a functional block that performs processing for performing a specific picking operation based on a processing command from the overall management unit 1100.
  • the operation performing unit 1130 performs a picking operation based on a processing command from the overall management unit 1100.
  • the operation planning unit 1150 causes the predetermined part of the autonomous mobile manipulator 100 such as the end effector unit 103 to be operated. Give the target posture.
  • the motion planning unit 1150 uses the target posture of the predetermined part of the autonomous mobile manipulator 100 given from the motion performing unit 1140 and the surrounding object posture information stored in the space management unit 1120 to plan the motion of the arm unit 103. It has a function to perform.
  • the operation plan is a predetermined time interval from the current configuration of the arm unit 102 to the configuration (target configuration) of the arm unit 102 that is the target posture of the end effector unit 103. This refers to processing for calculating the rotation angle of the joint mechanisms 1021, 1022, 1023, 1024, 1025, 1026.
  • a time-series change in the rotation angle of each joint mechanism of the arm unit 103 calculated in this way is referred to as an arm trajectory.
  • the motion planning function of the motion planning unit 1150 includes processing for calculating not only the arm unit 102 but also the rotation angle of the finger mechanism 1031 of the end effector unit 103 in the same manner. Similar to the arm trajectory, the time-series change in the rotation angle of the finger mechanism 1031 of the end effector is referred to as a finger trajectory.
  • the arm control unit 1160 performs control to change the rotation angle of each joint mechanism of the arm unit 102 according to the arm trajectory calculated by the motion planning unit 1150, and measures the actual rotation angle of each joint mechanism of the arm unit 102. It has the function to do.
  • the end effector control unit 1170 performs control to change the rotation angle of the finger mechanism 1031 of the end effector unit 103 in accordance with the finger trajectory calculated by the motion planning unit 1150, and the actual finger mechanism of the end effector unit 103. It has a function to measure the rotation angle.
  • the traveling management unit 1180 uses the measurement result of the ambient environment measurement sensor 1013 and the ambient environment map 1181 to recognize the position and orientation on its own plane from moment to moment. And based on the movement path
  • FIG. 1 The traveling management unit 1180 uses the measurement result of the ambient environment measurement sensor 1013 and the ambient environment map 1181 to recognize the position and orientation on its own plane from moment to moment. And based on the movement path
  • each functional block of the autonomous mobile manipulator control unit 110 may be implemented by predetermined software installed in the autonomous mobile manipulator control unit 110, or a part or all of the functional blocks may be an FPGA (FieldmProgrammable Gate Array). ), Hardware such as ASIC (Application Specific Specific Integrated Circuit).
  • FPGA FieldmProgrammable Gate Array
  • ASIC Application Specific Specific Integrated Circuit
  • FIG. 7 is a flowchart showing an example of the operation of the autonomous mobile manipulator 100 according to the present embodiment moving to the front of the shelf, and specifically illustrates step S31 in FIG.
  • the work procedure planning unit 1130 plans a stop position.
  • the planning of the stop position includes planning a stop position of the mobile carriage 101 suitable for picking an article stored in the shelf 210 specified by the shelf ID of the picking instruction 401.
  • the posture suitable for picking an article in the upward direction of the shelf differs depending on the height of the shelf 210 in the rack 200, and therefore, based on the shelf ID stored in the work procedure database 1131.
  • processing for calculating a stop position offset for the movement path of the picking instruction 401 is included.
  • the autonomous mobile manipulator control unit 1100 issues a movement processing command to the travel management unit 1180 (S3103).
  • the traveling management unit 1180 moves to the vicinity of the rack based on the movement route included in the picking command 401 (S3104), and autonomously moves to a position suitable for the picking operation based on the stop position offset planned by the work procedure planning unit 1130.
  • the stop position is finely adjusted so that the manipulator 100 is stopped.
  • FIG. 8 is a flowchart showing an example of the picking article posture observation operation of the autonomous mobile manipulator 100 according to the present embodiment, and specifically explains step S32 in FIG.
  • the overall management unit 1100 of the autonomous mobile manipulator control unit 110 searches the surrounding object posture information stored in the space management unit 1120 (S3201) and searches for the target picking target item. It is checked whether the posture is registered (S3202).
  • the posture of the target picking target item is registered, that is, when an item different from the target picking item is picked from the shelf 210 of the same rack 200 before the picking instruction, the picking target item is selected.
  • the posture of the target article is observed and stored in the space management unit 1120, it is reused. Thereby, the picking article posture observation operation can be skipped, and the operation time required for the picking operation of the autonomous mobile manipulator can be shortened.
  • the work procedure planning unit 1130 plans a shooting procedure (S3203).
  • the plan of the photographing procedure is to calculate a storage space in which articles can be stored in the shelf 210 specified by the rack ID and the shelf ID included in the picking instruction 401, and to measure the article measurement sensor with respect to the storage space.
  • the processing includes calculating the measurement posture and the number of times of measurement of the article measurement sensor 120 with respect to the rack 200 from the viewing angle of 120 and a predetermined resolution necessary for performing the article detection process in the image processing unit 1110.
  • the overall management unit 1100 issues a shooting operation command to the operation performing unit 1140 (S3204).
  • the motion execution unit 1140 converts one of the measurement postures of the article measurement sensor 120 planned by the photographing procedure into a posture based on the central point of the autonomous mobile manipulator 100, and sets the target posture as the target posture in 1150. Specify (S3205).
  • the motion planning unit 1150 plans an arm trajectory targeting the posture of the specified article measurement sensor 120 (S3206). Details of the operation plan will be described later.
  • the arm control unit 1160 moves the article measurement sensor 120 by controlling the arm unit 102 based on the planned arm trajectory (S3207).
  • the overall management unit 1100 issues an image processing command to the image processing unit 1110 (S3208).
  • the image processing unit 1110 acquires measurement data of the storage space of the shelf board 210 using the article measurement sensor 120 (S3209), and performs article and obstacle detection processing on the measurement data (S3210).
  • the detection processing refers to processing for recognizing the posture of an article by a predetermined method using the shape model of the article stored in advance in the object database 1111 and measurement data.
  • the article refers to not only the article to be picked but also all articles that can be arranged on the shelf board 210.
  • the posture of the article detected in step S3210 is stored as surrounding object posture information by the space management unit 1120 (S3211), and is reused as necessary as described above.
  • step S3210 if the picking target article is detected or the operation of the number of times for measuring the storage space planned in step S3203 is completed, the picking object posture observation operation is ended (S3210). If the picking target article has not been detected and the operation of the number of times for measuring the storage space has not been completed, the process returns to S3204 and the series of operations is repeated. When the measurement count operation is completed, but the picking target article is not detected, the overall management unit 1100 of the autonomous mobile manipulator 100 reports an article non-detection error to the management computer 400 (illustrated in FIG. 8). Not)
  • FIG. 9 is a flowchart illustrating an example of the picking operation of the autonomous mobile manipulator 100 according to the present embodiment, and specifically illustrates step S33 in FIG.
  • the picking target article is detected by the article search operation in FIG. 8, and the posture of the picking target article is stored in the space management unit 1120.
  • the work procedure planning unit 1120 plans a picking procedure (S3301).
  • the planning of the picking procedure includes planning the posture relationship of the end effector with respect to the picking target article for gripping the picking target article specified by the picking instruction 401 by the end effector unit 104. It is assumed that the gripping posture is stored in advance in the work procedure database 1131 in association with the picking article ID included in the picking instruction 401.
  • the overall management unit 1100 issues a picking command to the operation performing unit 1140 (S3302).
  • the motion performing unit 1140 searches the space management unit 1120 for the posture information of the picking target article and an object that can be an obstacle around it (S3303).
  • an end effector attitude with respect to the center point of the autonomous mobile manipulator 100 is calculated from the attitude of the article to be picked and the end effector attitude with respect to the picking article planned in the work procedure planning unit 1120, and the motion planning unit 1150.
  • the motion planning unit 1150 performs motion planning regarding the end effector posture designated as the target (S3305), and controls the movement of the arm unit (S3306).
  • the motion performing unit 1140 designates the posture of the finger mechanism for gripping the picking target article to the motion planning unit 1150 (S3307).
  • the motion planning unit 1150 plans a finger trajectory that realizes the posture of the designated finger mechanism (S3308).
  • the finger trajectory plan substantially corresponds to a process of interpolating the target finger posture from the designated current finger posture to the grasp in a predetermined time interval, and detailed description thereof is omitted.
  • the end effector control unit controls the finger mechanism 1031 of the end effector unit 103 based on the planned finger trajectory, and grips the picking article (S3309).
  • FIG. 10 is a block diagram illustrating an example of a specific configuration of the operation planning unit 1150 according to the present embodiment.
  • the motion planning unit 1150 includes a manipulator model 1151, an inverse kinematics calculation unit 1152, a reference posture calculation unit 1153, an arm posture selection unit 1154, a motion plan space construction unit 1155, an arm trajectory search unit 1156, and an object shape.
  • a model 1157 is a model 1157.
  • the manipulator model 1151 is model data including the geometric shape of each mechanism of the autonomous mobile manipulator 100 and the motion axis, motion direction, and motion range of motion parts such as a rotary joint mechanism.
  • the inverse kinematics calculation unit 1152 is a functional block that calculates possible values of each motion part of the autonomous mobile manipulator 100 using the manipulator model 1151 when a posture of a predetermined part of the autonomous mobile manipulator 100 is input. is there. For example, when the posture of the end effector unit 103 is input, a function of analytically or approximately calculating the rotation angle of the rotary joint mechanisms 1021, 1022, 1023, 1024, and 1025 is included.
  • the reference form calculation unit 1153 is a functional block for calculating a predetermined reference arm form, and will be described in detail later.
  • the arm configuration selection unit 1154 is a functional block that selects an appropriate target arm configuration from the arm configuration calculated by the inverse kinematics calculation unit 1152 based on the reference arm configuration calculated by the reference configuration calculation unit 1152. Detailed description will be given later.
  • the motion plan space construction unit 1155 is a functional block that constructs a motion plan space by combining the surrounding object posture information input from the space management unit 1120 and the shape model of the object stored in advance.
  • the arm trajectory search unit is a functional block that searches for an arm trajectory from the current configuration of the arm unit 103 to the target arm configuration selected by the arm configuration selection unit 1154 in the motion planning space.
  • FIG. 11 is a schematic diagram of an example of the picking operation of the autonomous mobile manipulator 100.
  • the operation of the operation planning unit 1150 will be described with respect to an example in which the picking target article 900 placed directly on the shelf board 210 is picked by the autonomous mobile manipulator 100.
  • the arm part 102 and the end effector part 103 are configured as a three-degree-of-freedom manipulator with the rotation axis facing the same direction. Therefore, the end effector unit 104 can take an arbitrary posture (position and direction) within a predetermined range within a two-dimensional plane indicated by the x-axis and the z-axis shown in the drawing.
  • FIG. 12 is a flowchart illustrating an example of the operation of the operation planning unit 1150 according to the present embodiment.
  • the end effector target posture is designated by the motion performing unit 1140 (P1).
  • the end effector target postures designated by the action performing unit 1140 are shown as finger mechanisms 1031xa and 1031xb in FIG. That is, the end effector target posture is specified as a posture in which the picking target article 900 is grasped by the grasping point P by the finger mechanisms 1031xa and 1031xb.
  • the grasping point P is a point planned by the work procedure planning unit 1120 as a posture suitable for grasping the picking object 900 by the finger mechanism 1031.
  • the motion planning unit 1150 uses the inverse kinematics calculation unit 1152 to calculate an arm configuration for realizing the end effector target posture for grasping the picking target article 900 from the grasping point P (P2).
  • the motion planning unit 1150 uses the inverse kinematics calculation unit 1152 to calculate an arm configuration for realizing the end effector target posture for grasping the picking target article 900 from the grasping point P (P2).
  • the grasping point P P2
  • FIG. 13 shows an example of a plurality of arm forms.
  • the arm portion form that realizes the end effector target posture for grasping the picking target article 900 by the grasping point P is configured in the order of the joint mechanisms 1022, 1023 a, and 1025 starting from the movable carriage portion 101.
  • the manipulability is a scalar amount calculated as shown in Equation 1.
  • J (q) is a Jacobian matrix of the arm unit 102, and the rotation angles of the joint mechanisms 1022 , 1023 , and 1025 of the arm unit 103 are ⁇ 1022 , ⁇ 1023 , and ⁇ 1025 (see FIG. 11), and the end effector unit
  • the position coordinates of the base point 103 are x, z, and the direction is ⁇ (see FIG. 11)
  • the relational expression shown in Equation 2 is given.
  • the maneuverability w of a predetermined arm configuration is obtained in the autonomous mobile manipulator 100 shown in FIG. Is calculated by the following equation (3).
  • the operable degree of the autonomous mobile manipulator 100 shown in FIG. 12 is determined only by the angle of the joint mechanism 1023.
  • the motion planning unit 1150 calculates a reference straight line uniquely determined from the target form specified by the motion performing unit 1140 in the reference form calculation unit 1153 in order to uniquely determine the arm part form.
  • P3 A specific example of the reference straight line is shown as a straight line L in FIG. That is, the reference straight line L is a straight line connecting the grasp point P and the rotation axis of the rotary joint 1022. And the arm part form in which each joint mechanism is located on the reference
  • this arm portion form is an arm form in which the rotary joint mechanisms 1022, 1023, and 1025 are arranged on the reference straight line L.
  • This form is referred to as a reference form in the following description.
  • the distance from the reference form is calculated for a plurality of arm part forms (upper elbow form and lower elbow form in the example of FIG. 12) calculated by the inverse kinematics calculation part.
  • the distance there is a method of using the square root of a value obtained by summing the squares of the angular differences of the rotary joint mechanisms.
  • the calculated smallest distance is output to the arm trajectory searching unit 1156 as the target form of the arm unit (P5).
  • the lower elbow posture is selected as the target form of the arm unit and is output to the arm trajectory search unit 1156.
  • FIG. 13B illustrates the angular difference of the rotary joint mechanism.
  • the angle difference of the joint 1022 is A1-A2
  • the angle difference of the joint 1023b is B1-B2.
  • the angle difference is measured for other joints, and the square root of the sum of the squares of these is obtained.
  • the target arm configuration is an arm configuration that is closest to a straight line from the rotation axis of the rotary joint 1022 toward the grasp point P, and as is apparent from FIG. 13, the occupied space volume of the arm unit 103 is other arm configuration. Smaller than Here, the occupied space of the arm portion 103 is a volume of a predetermined ellipsoid that can enclose the arm portion 103. If the occupied space volume of the arm portion is small, the probability that the surrounding obstacle and the arm portion interfere with each other decreases, that is, it is suitable for the arm portion operation in a warehouse where articles are densely packed.
  • the area surrounded by the arm and the reference line in the two-dimensional plane indicated by the x axis and the z axis is used as a reference. For example, it is determined that the smaller area is closer.
  • the lower arm position is smaller than the upper arm position because the area surrounded by the arm and the reference line is smaller, so it is determined that the distance is shorter.
  • Another example of the distance calculation method is based on the number of intersections between the arm and the reference line in the two-dimensional plane indicated by the x-axis and the z-axis. For example, it is determined that a larger number of intersections is closer. In the example of FIG. 13B, since the number of intersections where the lower arm intersects the reference straight line (dotted line) is 1 and the number of intersections of the upper arm is 0, it is determined that the lower arm is closer in distance.
  • the arm unit trajectory searching unit 1156 detects the surrounding obstacles and the arm unit 103 out of the arm trajectory from the current configuration of the arm unit 103 to the target configuration determined by the arm unit posture selection unit 1054. Alternatively, an effective arm trajectory in which the end effector unit 104 does not interfere is searched (P6).
  • the arm part trajectory search unit 1156 can search for an arm trajectory using a conventional sample-based trajectory search algorithm as described in Non-Patent Document 1 or 2. Since the arm target form is uniquely determined, an effective arm trajectory can be searched for in a sufficiently short time even in the conventional sample-based trajectory search algorithm.
  • an arm portion trajectory search portion 1156 as shown in FIG. 15 can be configured.
  • the arm trajectory search unit 1156 is arranged in parallel with the conventional sample-based trajectory search algorithm execution unit 11563, and the arm trajectory from the current arm unit form to the target arm part form is obtained by simple linear interpolation of joint angle values.
  • a linear interpolation arm trajectory calculation unit 11561 to be calculated and an obstacle interference check unit 11562 for checking whether the arm unit 103 and the end effector unit 104 interfere with surrounding obstacles in the interpolated arm unit trajectory are executed. It is a configuration. In general, calculation of a linear interpolation trajectory enables a trajectory to be calculated in a short time with respect to a sample-based trajectory search algorithm, but cannot calculate a path that bypasses an obstacle.
  • the sample-based trajectory search algorithm takes time to calculate an effective trajectory, but can search for an arm trajectory that bypasses an obstacle. Therefore, in the interpolation trajectory calculated by the linear interpolation arm trajectory calculation unit 11561, if the obstacle interference check unit 11562 has not checked the obstacle interference, the trajectory search by the sample-based trajectory search algorithm execution unit 11563 is completed. In addition, the effective trajectory selection unit 11564 selects and outputs it. With such a configuration, it is possible to further shorten the time for searching for the arm trajectory.
  • the reference straight line L is a straight line connecting the grasp point P and the rotation axis of the rotary joint 1022, but the reference straight line is not limited to this, and may be a predetermined part of the autonomous mobile manipulator 100. That's fine.
  • FIG. 16 is an example in which the present method is applied to an operation plan (S3206 in FIG. 8) in the picking article posture observation operation of the autonomous mobile manipulator 100.
  • the posture of the article measurement sensor that is, the position and direction of the viewpoint C in FIG.
  • the autonomous mobile manipulator 100 having the mobile carriage unit 101 has been described.
  • the present invention is not limited to this.
  • the present invention can also be applied to a stationary manipulator that is fixed at a predetermined location and moves only the arm portion 103.

Abstract

Provided is a method for controlling a manipulator by determining a suitable target posture of an arm part during a movement in the manipulator and carrying out motion planning to enable the arm part to be naturally and quickly moved on the basis of the determination. The present invention searches for an arm trajectory from a predetermined posture to a target posture by setting a straight line that connects a predetermined rotating axis of the manipulator and a first predetermined reference point as a first reference line, setting a first specific point posture determined in accordance with the first reference line, and setting the posture having the shortest distance to the first specific point posture as the target posture.

Description

マニプレータ、マニプレータの動作計画方法、および、マニプレータの制御システムManipulator, manipulator operation planning method, and manipulator control system
 本発明は、未知の環境で動作する移動型マニプレータの動作を生成する方法、および当該方法を用いる装置とシステムに関する。 The present invention relates to a method for generating a motion of a mobile manipulator that operates in an unknown environment, and an apparatus and system using the method.
 自律移動マニプレータは、倉庫や工場などの施設において、ヒトの活動を支援したり代替したりするものとして、実用化が期待されている。こうした自律移動マニプレータには、ヒトと同等の柔軟性と迅速性が求められる。 Autonomous mobile manipulators are expected to be put to practical use as facilities that support or replace human activities in facilities such as warehouses and factories. Such an autonomous mobile manipulator is required to be as flexible and quick as humans.
 移動マニプレータは、一般に、前進・後退・旋回などにより自身を移動させる移動台車部と、移動台車部に取り付けられた、1つ以上の回転関節機構を有するアーム部と、アーム部の先端に取り付けられたエンドエフェクタ部とを有する。エンドエフェクタ部は、特定の物体を把持したり、移動させたりすることができる。例えば、ロボットハンドや磁石である。こうした、エンドエフェクタ部による物体の把持や移動を、物体に対する操作と呼称する。 In general, a moving manipulator is attached to a moving carriage part that moves itself by moving forward, backward, or turning, an arm part that has one or more rotary joint mechanisms attached to the moving carriage part, and a tip of the arm part. And an end effector portion. The end effector unit can hold or move a specific object. For example, a robot hand or a magnet. Such gripping or movement of the object by the end effector unit is referred to as an operation on the object.
 マニプレータによって、ある特定の物体に対して操作を与えるためには、アーム部の回転関節角度を変更することで、エンドエフェクタ部を所定の位置および方向(以下、位置および方向を合わせて、姿勢と称する)に移動させる。現在のエンドエフェクタの姿勢から、目標とするエンドエフェクタ姿勢まで移動させるための、アーム部の関節角度値の時系列変化(以下、アーム軌道と称する)を決定する方法を、動作計画と呼ぶ。 In order to give an operation to a specific object by the manipulator, the end effector unit is changed to a predetermined position and direction (hereinafter, the position and direction are adjusted by changing the rotation joint angle of the arm unit). To move. A method of determining a time-series change (hereinafter referred to as an arm trajectory) of the joint angle value of the arm unit for moving from the current end effector posture to a target end effector posture is referred to as an operation plan.
 ここで、複数の関節を有するマニプレータでは、あるエンドエフェクタの姿勢に対応する、アーム部関節角度の組み合わせ(以下、形態と称する)は複数存在する。従って、アーム部が周辺の障害物と衝突したり、エンドエフェクタが予期せぬ経路で移動したりしないように、適切なアーム部の形態を選択して、動作計画を行う必要がある。 Here, in a manipulator having a plurality of joints, there are a plurality of combinations (hereinafter referred to as forms) of arm joint angles corresponding to the posture of a certain end effector. Accordingly, it is necessary to select an appropriate arm portion form and perform an operation plan so that the arm portion does not collide with surrounding obstacles and the end effector does not move along an unexpected path.
 現在形態から目標形態に至るアーム軌道の動作計画を行うために、非特許文献1、2に開示されるような、乱数を利用したサンプルベース探索動作計画法という技術が提案されている。これは、平易に説明すれば、乱数を利用して、アーム部の取り得る様々な形態を繰り返し生成し、それらをグラフ理論に基づいて時系列に連続した形態を探索することで、移動経路を生成する方法である。 In order to plan the motion of the arm trajectory from the current configuration to the target configuration, a technique called a sample-based search motion planning method using random numbers as disclosed in Non-Patent Documents 1 and 2 has been proposed. To explain this simply, using random numbers, it repeatedly generates various forms that the arm can take, and searches them for time-sequential forms based on graph theory. It is a method of generating.
 また、特許文献1では、複数の把持形態に至るアーム部の移動経路における形態を各々計算し、可操作度と呼ばれる各々の形態の取り易さの評価値を算出する処理を行い、この評価結果に基づいた把持形態を選択する技術が公開されている。 Moreover, in patent document 1, the form in the movement path | route of the arm part which reaches a some holding | grip form is calculated, respectively, The process which calculates the evaluation value of the ease of taking of each form called an operable degree is performed, and this evaluation result A technique for selecting a grip form based on the above has been disclosed.
特開2009-172685JP 2009-172585 A
 移動型マニプレータにおいては、マニプレータと対象物の位置関係は限定されないため、全ての状況を網羅したアーム部の目標形態を予め教示することはできない。そのため、移動型マニプレータにおいては、状況に応じてアーム部の目標形態を自律的に生成し、現在形態から目標形態に至る動作計画も自律的に行う必要がある。 In the moving manipulator, the positional relationship between the manipulator and the target object is not limited, so it is not possible to teach in advance the target form of the arm that covers all situations. Therefore, in the mobile manipulator, it is necessary to autonomously generate the target form of the arm unit according to the situation and autonomously perform the operation plan from the current form to the target form.
 非特許文献1および2記載の技術は、アーム部の目標形態を含めた、全てのアーム部の取り得る形態を網羅的に探索して、アーム軌道を生成することが可能である。しかしながら、このように目標形態を含めて網羅的に探索する場合、膨大な数の繰り返し処理が必要となる。そのため、適切な移動経路が生成されるまでに長時間を要したり、探索を短時間で打ち切った場合は、目標形態まで大きく迂回して達するような不自然なアーム軌道が生成されたりすることが課題となる。従って、短時間でアーム軌道探索を終えるためには、目標形態を一意に指定することが有効な手段となる。 The techniques described in Non-Patent Documents 1 and 2 can generate an arm trajectory by exhaustively searching for possible forms of all arm parts, including the target form of the arm part. However, when searching exhaustively including the target form in this way, an enormous number of iterative processes are required. For this reason, it takes a long time to generate an appropriate movement path, or if the search is terminated in a short time, an unnatural arm trajectory may be generated that greatly reaches the target form. Is an issue. Therefore, in order to finish the arm trajectory search in a short time, it is effective to specify the target form uniquely.
 一方、特許文献1記載の技術は、目標アーム形態を可操作度によって決定しているため、アーム形態を一意に決めることができない。これは、複数の回転関節を有するアーム部においては、同一の可操作度を持つ形態が複数存在しうるためである。 On the other hand, since the technique described in Patent Document 1 determines the target arm form based on the degree of manipulability, the arm form cannot be uniquely determined. This is because an arm unit having a plurality of rotary joints may have a plurality of forms having the same operable degree.
 上記を鑑み、本発明が解決する課題は、マニプレータにおいて、動作時に適切なアーム部の目標形態を決定し、それに基づいてアーム部の自然かつ迅速な動作が可能な動作計画を行うマニプレータの制御方法を提供することにある。 In view of the above, the problem to be solved by the present invention is that a manipulator determines an appropriate target form of an arm part at the time of operation, and based on that, a manipulator control method for performing an operation plan capable of natural and quick operation of the arm part Is to provide.
 本願において開示される発明のうち代表的なものを挙げれば、マニプレータの第一関節の回転軸と所定の基準点とを結ぶ直線を基準直線とし基準直線より定まる特異点形態として特異点形態に最も近い距離にある形態を目標形態として、所定の形態から目標形態に至るアーム軌道を探索することを特徴とするマニプレータの動作計画方法である。 Among the inventions disclosed in this application, a typical one is the most singular point form as a singular point form determined from a reference line, with a straight line connecting the rotation axis of the first joint of the manipulator and a predetermined reference point as a reference straight line. An operation planning method for a manipulator characterized by searching for an arm trajectory from a predetermined form to a target form with a form at a short distance as a target form.
 本発明の他の側面は、複数の回転関節を備えたアーム部と、エンドエフェクタ部を備えたマニプレータにおいて、アーム部と関連付けられた基準点を所望の位置に位置づけるために情報処理装置で実行される動作計画方法である。この方法において、マニプレータの第一関節の回転軸と基準点とを結ぶ直線を基準直線として設定する第1の処理、基準直線より定まるアーム部の形態を特異点形態として設定する第2の処理、基準点を所望の位置に位置づけるためにアーム部が取り得る形態のうち、特異点形態に最も近い距離にあるアーム部の形態を目標形態として設定する第3の処理、目標形態に至るアーム軌道を探索する第4の処理、を備えることを特徴とする。 Another aspect of the present invention is executed by an information processing apparatus to position a reference point associated with an arm unit at a desired position in a manipulator including an arm unit including a plurality of rotary joints and an end effector unit. This is an operation planning method. In this method, a first process for setting a straight line connecting the rotation axis of the first joint of the manipulator and the reference point as a reference straight line, a second process for setting the form of the arm portion determined from the reference straight line as a singular point form, Of the forms that the arm unit can take to position the reference point at a desired position, the third process for setting the form of the arm part closest to the singular point form as the target form, and the arm trajectory to the target form And a fourth process for searching.
 本発明の他の側面は、移動台車部と、移動台車部に搭載されたアーム部と、アーム部の動作を制御する制御部を有し、アーム部は、移動台車側から順番に第一関節および第二関節を備え、制御部は、第二関節を基準にして移動台車部と反対側のアーム部に関連付けて、基準点を設定し、基準点を所望の位置に位置づけるアーム部の移動制御を行うマニプレータの制御システムである。このシステムは、基準点を所望の位置に位置づけるために取り得るアーム部の形態を、複数の候補形態として算出する第1の機能と、基準点と第一関節を結ぶ基準直線を算出する第2の機能と、基準直線に一致するアーム部の形態を、特異点形態として算出する第3の機能と、特異点形態との比較結果に基づいて、複数の候補形態から1つのアーム部の形態を選択し、目標形態として設定する第4の機能と、目標形態に至るための、第一関節および第二関節の動作を決定する第5の機能と、を有する。 Another aspect of the present invention includes a movable carriage unit, an arm unit mounted on the movable carriage unit, and a control unit that controls the operation of the arm unit. And a second joint, and the control unit sets a reference point in association with the arm part on the opposite side of the moving carriage unit with respect to the second joint, and controls the movement of the arm unit that positions the reference point at a desired position. It is the control system of the manipulator which performs. This system has a first function for calculating the form of the arm portion that can be taken to position the reference point at a desired position as a plurality of candidate forms, and a second function for calculating a reference straight line connecting the reference point and the first joint. Based on the comparison result between the singular point form and the third function that calculates the form of the arm part that matches the reference line and the form of the arm part, the form of one arm part from a plurality of candidate forms A fourth function that is selected and set as the target form and a fifth function that determines the operation of the first joint and the second joint to reach the target form are provided.
 本発明の他の側面は、移動台車部と、移動台車部に搭載されたアーム部を有するマニプレータであって、アーム部は、移動台車側から順番に第一関節、第二関節および第三関節を備え、アーム部は、第三関節の前記移動台車部と反対側の部位に設定された基準点を、所望の位置に位置づける移動制御装置により駆動されるものである。移動制御は、基準点を所望の位置に位置づけるために取り得るアーム部の形態を、複数の候補形態として算出する第1の機能と、基準点と前記第一関節を結ぶ基準直線を算出する第2の機能と、基準直線に重なるアーム部の形態を、特異点形態として算出する第3の機能と、特異点形態との比較結果に基づいて、複数の候補形態から1つのアーム部の形態を選択し、目標形態として設定する第4の機能と、目標形態に至るための、第一関節、第二関節および第三関節の動作を決定する第5の機能と、により実行される。 Another aspect of the present invention is a manipulator having a movable carriage part and an arm part mounted on the movable carriage part, and the arm parts are a first joint, a second joint, and a third joint in order from the movable carriage side. The arm portion is driven by a movement control device that positions a reference point set at a portion of the third joint opposite to the moving carriage portion at a desired position. The movement control includes a first function for calculating, as a plurality of candidate forms, possible arm portions for positioning the reference point at a desired position, and a first straight line for connecting the reference point and the first joint. Based on the comparison result between the second function, the third function that calculates the form of the arm portion overlapping the reference straight line as the singular point form, and the singular point form, the form of one arm part is selected from a plurality of candidate forms. The fourth function that is selected and set as the target form and the fifth function that determines the operations of the first joint, the second joint, and the third joint to reach the target form are executed.
 本発明の他の側面は、上記の移動台車部を所望の位置まで自動的に移動させ、所望の位置に到達した後に、上述のアーム部の制御を行って、物品の操作を行う物品操作システムまたは物品操作方法を含む。このようなシステムまたは方法においては、予め操作対象とする物品を配置した個所を記憶しておき、その近傍まで移動台車を移動させ、しかるのち、物品を操作するためにアーム部を制御する。 Another aspect of the present invention is an article operation system that automatically moves the movable carriage part to a desired position and controls the arm part to operate the article after reaching the desired position. Or an article operation method is included. In such a system or method, the location where the article to be operated is placed is stored in advance, the moving carriage is moved to the vicinity thereof, and then the arm unit is controlled to operate the article.
 以上のように本発明は、特異点形態という基準を用いることにより、所望の動作を行うに当たってアーム部がとるべき形態を、一意に決定することができる。 As described above, the present invention can uniquely determine the form that the arm portion should take in performing a desired operation by using the criterion of the singular point form.
 以上の発明において、関節の数は3個以上あってもよい。また、アーム部の先端その他には、エンドエフェクタとしてロボットハンド、磁石その他の、物品を操作する機構を備えてもよい。また、各種のセンサを備えつけてもよい。センサにより、基準点やアームの位置を検知して制御することにより、障害物を避ける等さらに複雑な操作をすることができる。 In the above invention, the number of joints may be three or more. Further, a mechanism for operating an article such as a robot hand, a magnet, or the like may be provided as an end effector at the tip of the arm portion or the like. Various sensors may be provided. By detecting and controlling the position of the reference point and the arm with the sensor, more complicated operations such as avoiding an obstacle can be performed.
 特異点形態との距離を算出する方法は種々あり得るが、一例をあげれば、特異点形態における各回転関節機構の角度と、アーム部が取り得る形態における各回転関節機構の角度の差の二乗を総和した値の平方根を用いて、前記距離を評価するようにすれば、アーム部を制御するシステムが通常用いるデータを使って、容易に計算することができる。 There are various methods for calculating the distance to the singular point form, but as an example, the square of the difference between the angle of each rotary joint mechanism in the singular point form and the angle of each rotary joint mechanism in the form that the arm can take If the distance is evaluated using the square root of the sum of the values, it can be easily calculated using the data normally used by the system that controls the arm unit.
 なお、アーム部を制御する制御部は、全てが移動台車部に搭載されてもよい。あるいは、その一部または全部が移動台車部外にあって、無線もしくは有線回線によって接続されていてもよい。 In addition, all the control parts which control an arm part may be mounted in a mobile trolley part. Alternatively, some or all of them may be outside the mobile carriage unit and connected by wireless or wired lines.
 本発明によるマニプレータは、動作時に適切なアーム部の目標姿勢を自律的に決定され、それに基づいて自然かつ迅速な動作が可能なアーム部の軌道が生成できる。 The manipulator according to the present invention can autonomously determine a target posture of an appropriate arm portion during operation, and can generate a trajectory of the arm portion capable of natural and quick operation based on the target posture.
 上記した以外の課題、構成、及び効果は、以下の実施形態の説明により明らかにされる。 Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.
ピッキングロボットシステム1の構成の一例を示した斜視図である。1 is a perspective view illustrating an example of a configuration of a picking robot system 1. FIG. ピッキングロボットシステム1において、物品をピッキングする一連の動作の一例を示した流れ図である。5 is a flowchart showing an example of a series of operations for picking an article in the picking robot system 1. ピッキング指示401に含まれるデータの一例を示した表図である。6 is a table showing an example of data included in a picking instruction 401. FIG. 移動指示402に含まれるデータの一例を示した表図である。6 is a table showing an example of data included in a movement instruction 402. FIG. 自律移動マニプレータ100の具体的な構成の一例を示した側面図である。3 is a side view showing an example of a specific configuration of the autonomous mobile manipulator 100. FIG. 自律移動マニプレータ制御部110の具体的な構成の一例を示したブロック図である。It is the block diagram which showed an example of the specific structure of the autonomous mobile manipulator control part. 自律移動マニプレータ100の棚前に移動する動作の一例を示したフローチャートである。5 is a flowchart showing an example of an operation of moving to the front of the shelf of the autonomous mobile manipulator 100. 自律移動マニプレータ100のピッキング物品姿勢観測動作の一例を示すフローチャートである。5 is a flowchart showing an example of picking article posture observation operation of the autonomous mobile manipulator 100. 自律移動マニプレータ100のピッキング動作の一例を示すフローチャートである。4 is a flowchart illustrating an example of a picking operation of the autonomous mobile manipulator 100. 動作計画部1150の具体的な構成の一例を示したブロック図である。6 is a block diagram showing an example of a specific configuration of an operation planning unit 1150. FIG. 自律移動マニプレータ100のピッキング動作の一例の模式図である。4 is a schematic diagram illustrating an example of a picking operation of the autonomous mobile manipulator 100. FIG. 動作計画部1150の動作の一例を示したフローチャートである。5 is a flowchart showing an example of the operation of an operation planning unit 1150. 自律移動マニプレータ100において物品を把握するアーム部形態の例を示した模式図である。It is the schematic diagram which showed the example of the arm part form which grasps | stores articles | goods in the autonomous mobile manipulator 100. FIG. 自律移動マニプレータ100において物品を把握するアーム部形態の例を示した模式図である。It is the schematic diagram which showed the example of the arm part form which grasps | stores articles | goods in the autonomous mobile manipulator 100. FIG. 自律移動マニプレータ100において物品を把握するアーム部形態の例を示した模式図である。It is the schematic diagram which showed the example of the arm part form which grasps | stores articles | goods in the autonomous mobile manipulator 100. FIG. 動作計画部1150の具体的な構成の一例を示したブロック図である。6 is a block diagram showing an example of a specific configuration of an operation planning unit 1150. FIG. 自律移動マニプレータ100において物品を把握するアーム部形態の例を示した模式図である。It is the schematic diagram which showed the example of the arm part form which grasps | stores articles | goods in the autonomous mobile manipulator 100. FIG.
 以下、本発明の実施の形態について添付図面を参照して説明する。なお、全ての図面において、実施形態が異なる場合であっても、同一または相当する部材については同一の符号を付し、共通する説明は繰り返さない。 Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In all the drawings, even if the embodiments are different, the same or corresponding members are denoted by the same reference numerals, and the common description will not be repeated.
 本明細書等における「第1」、「第2」、「第3」などの表記は、構成要素を識別するために付するものであり、必ずしも、数または順序を限定するものではない。また、構成要素の識別のための番号は文脈毎に用いられ、一つの文脈で用いた番号が、他の文脈で必ずしも同一の構成を示すとは限らない。また、ある番号で識別された構成要素が、他の番号で識別された構成要素の機能を兼ねることを妨げるものではない。 In this specification and the like, notations such as “first”, “second”, and “third” are attached to identify the constituent elements, and do not necessarily limit the number or order. In addition, a number for identifying a component is used for each context, and a number used in one context does not necessarily indicate the same configuration in another context. Further, it does not preclude that a component identified by a certain number also functions as a component identified by another number.
 図面等において示す各構成の位置、大きさ、形状、範囲などは、発明の理解を容易にするため、実際の位置、大きさ、形状、範囲などを表していない場合がある。このため、本発明は、必ずしも、図面等に開示された位置、大きさ、形状、範囲などに限定されない。 The position, size, shape, range, etc. of each component shown in the drawings and the like may not represent the actual position, size, shape, range, etc. in order to facilitate understanding of the invention. For this reason, the present invention is not necessarily limited to the position, size, shape, range, and the like disclosed in the drawings and the like.
 <ピッキングロボットシステム、およびその動作>
 本実施例では、通販倉庫における商品の出し入れなどを想定した、自律移動マニプレータによって棚からの物品のピッキング作業を行う、ピッキングロボットシステムについて具体的に説明する。ここで、ピッキング作業とは、指定された物品を、それらが収められている棚から取りだす作業を指す。
<Picking robot system and its operation>
In the present embodiment, a picking robot system that performs picking work of articles from a shelf by an autonomous mobile manipulator assuming the taking in and out of goods in a mail order warehouse will be specifically described. Here, the picking work refers to a work of taking out designated articles from the shelves in which they are stored.
 図1は、本実施例にかかるピッキングロボットシステム1の構成の一例を示した図面である。ピッキングロボットシステム1は、1台以上の自律移動マニプレータ100と、1台以上の棚200と、1台以上の搬送ロボット300と、少なくとも1つの管理コンピュータ400と、入力端末500とから構成される。 FIG. 1 is a diagram showing an example of a configuration of a picking robot system 1 according to the present embodiment. The picking robot system 1 includes one or more autonomous mobile manipulators 100, one or more shelves 200, one or more transfer robots 300, at least one management computer 400, and an input terminal 500.
 自律移動マニプレータ100は、前進・後退・旋回などにより自身を移動させる移動台車部101と、移動台車部101に取り付けられた1つ以上の回転関節を有するアーム部102と、アーム部の先端に取り付けられたエンドエフェクタ部103と、自律移動マニプレータ制御部110とを備える。エンドエフェクタ部103により、所定の物体を把持したり、移動させたりすることができる構成である。 The autonomous moving manipulator 100 is attached to a moving carriage 101 for moving itself by moving forward, backward, turning, etc., an arm 102 having one or more rotary joints attached to the moving carriage 101, and a tip of the arm. The end effector unit 103 and the autonomous mobile manipulator control unit 110 are provided. The end effector unit 103 can hold or move a predetermined object.
 ラック200には、1つ以上の物品900が棚板210の上に収納され、ピッキングロボットシステム1内の所定のラック配置エリア2000に設置される。 In the rack 200, one or more articles 900 are stored on the shelf board 210 and installed in a predetermined rack arrangement area 2000 in the picking robot system 1.
 搬送ロボット300は、自律移動マニプレータ100の移動台車部101と同様に、前進・後退・旋回などにより自身を移動させることが可能な構成となっている。また、搬送ロボット300は、その上部が平面となっており、ピッキングトレイ310を積載可能な構造を備える。なお、自律移動マニプレータ100と搬送ロボット300を合体させ、1つの装置として構成することもできる。 The transfer robot 300 is configured to be able to move itself by moving forward, backward, turning, etc., like the moving carriage unit 101 of the autonomous moving manipulator 100. Further, the transfer robot 300 has a structure in which the upper part is a flat surface and the picking tray 310 can be loaded. The autonomous mobile manipulator 100 and the transfer robot 300 can be combined to form a single device.
 管理コンピュータ400は、ピッキングロボットシステム1の全体を管理し運用する。管理コンピュータ400と、自律移動マニプレータ100との間は、無線での通信が可能な構成であり、それぞれに図示されない送信部、受信部とを設けた構成となっている。同様に、管理コンピュータ400と、搬送ロボット300との間も、無線での通信が可能な構成であり、それぞれに図示されない送信部、受信部とを設けた構成となっている。 The management computer 400 manages and operates the entire picking robot system 1. The management computer 400 and the autonomous mobile manipulator 100 are configured to be able to communicate wirelessly, and have a configuration in which a transmission unit and a reception unit (not shown) are provided. Similarly, the management computer 400 and the transfer robot 300 are configured to be able to communicate wirelessly, and are each configured to include a transmission unit and a reception unit that are not illustrated.
 入力端末500はPCなどとして実装され、入力端末500と、管理コンピュータ400とは、有線ないし無線での通信が可能であり、それぞれに図示されない送信部、受信部を設けた構成となっている。なお、入力端末500は、PCなどで実装されると説明したが、これに限定するものではなく、有線ないし無線で管理コンピュータ400との通信可能な媒体であればよい。たとえば、携帯電話機器やタブレット端末など様々な形態も取り得るし、複数のコンピュータから構成されるサーバーシステムであってもよい。 The input terminal 500 is mounted as a PC or the like, and the input terminal 500 and the management computer 400 can communicate with each other by wire or wireless, and have a configuration in which a transmission unit and a reception unit (not shown) are provided. Although the input terminal 500 has been described as being implemented by a PC or the like, the present invention is not limited to this, and any medium that can communicate with the management computer 400 by wire or wireless may be used. For example, various forms such as a mobile phone device and a tablet terminal can be used, and a server system including a plurality of computers may be used.
 図2は、本実施例にかかるピッキングロボットシステム1において、物品をピッキングする一連の動作の一例を示したフロー図面である。本図面は、簡単のため1台のピッキングロボットと、1台の搬送ロボットが動作する例を示す。 FIG. 2 is a flowchart showing an example of a series of operations for picking an article in the picking robot system 1 according to the present embodiment. This drawing shows an example in which one picking robot and one transfer robot operate for simplicity.
 図2において、破線A1で囲まれた部分は入力端末500にかかる動作を、破線A2で囲まれた部分は管理コンピュータ300にかかる動作を、破線A3で囲まれた部分は自律移動マニプレータ100にかかる動作を、破線A4で囲まれた部分は搬送ロボット300にかかる動作を、それぞれ示す。 In FIG. 2, the part surrounded by the broken line A1 is related to the operation of the input terminal 500, the part surrounded by the broken line A2 is related to the operation of the management computer 300, and the part surrounded by the broken line A3 is related to the autonomous mobile manipulator 100. The operations surrounded by the broken line A4 indicate the operations applied to the transfer robot 300, respectively.
 まず、オペレータは入力端末500を用いて、1つ以上の物品からなる出荷物品のオーダーを管理コンピュータ400に入力する(S11)。管理コンピュータ400は、どの物品900がどの棚200(あるいはさらにどの棚板210)にあるか等の物品管理情報データを、保持またはアクセス可能である。また、棚200の配置(位置)情報データも、保持またはアクセス可能である。 First, the operator uses the input terminal 500 to input an order for one or more items shipped to the management computer 400 (S11). The management computer 400 can hold or access article management information data such as which article 900 is on which shelf 200 (or which shelf board 210). Further, the arrangement (position) information data of the shelf 200 can also be held or accessed.
 管理コンピュータ400は、出荷物品が収められているラック200の設置場所などから、効率的にピッキングを行うために、出荷物品をピッキングする順番を定め、ピッキング物品リストを作成する(S21)。そして、ピッキングする物品の順番に従い、自律移動マニプレータ100と、搬送ロボット300との移動経路が干渉したり、迂回経路をとることでピッキング時間が増大したりしないように、自律移動マニプレータ100と搬送ロボットそれぞれの移動経路を計画する(S22)。そして、自律移動マニプレータ100に対してはピッキング指示401を、搬送ロボット300に対しては移動指示402を、それぞれ送信する(S23およびS24)。ここで、ピッキング指示401に含まれるデータの一例を図3、移動指示402に含まれるデータの一例を図4に、それぞれ示す。 The management computer 400 determines the order of picking the shipment items from the installation location of the rack 200 in which the shipment items are stored, and creates a picking item list (S21). Then, according to the order of the articles to be picked, the autonomous mobile manipulator 100 and the transport robot do not interfere with the travel path between the autonomous mobile manipulator 100 and the transport robot 300 or increase the picking time by taking a detour path. Each movement route is planned (S22). Then, a picking instruction 401 is transmitted to the autonomous mobile manipulator 100, and a movement instruction 402 is transmitted to the transfer robot 300 (S23 and S24). Here, an example of data included in the picking instruction 401 is shown in FIG. 3, and an example of data included in the movement instruction 402 is shown in FIG.
 図3において、ピッキング指示401には、オーダー毎に固有に割り当てられた識別子であるオーダーIDと、ピッキング物品の種類毎に固有に割り当てられた識別子であるピッキング物品IDと、ピッキング物品が収められているラック200に個別に割り当てられたラックIDと、ラック200において物品が収納されている棚板210を個別に示す棚板IDと、ラック200の前にピッキングできる位置に移動するための移動経路データと、搬送ロボット300を停止させる位置を示す搬送ロボット停止地点データを含む。 In FIG. 3, the picking instruction 401 contains an order ID that is an identifier uniquely assigned to each order, a picking item ID that is an identifier uniquely assigned to each type of picking article, and a picking article. A rack ID individually assigned to the rack 200, a shelf ID individually indicating the shelf 210 in which the articles are stored in the rack 200, and movement path data for moving to a position where the rack 200 can be picked. And transfer robot stop point data indicating a position where the transfer robot 300 is stopped.
 図4において、移動指示402には、オーダー毎に固有に割り当てられた識別子であるオーダーIDと、ピッキングする自律移動マニプレータ100の近傍に移動するための移動経路データとを含む。 4, the movement instruction 402 includes an order ID that is an identifier uniquely assigned to each order and movement route data for moving to the vicinity of the autonomous movement manipulator 100 to be picked.
 図2に戻り説明を続ける。自律移動マニプレータ100は、管理コンピュータ400から、ピッキング指示401を受信すると、まずピッキング指示401に含まれる移動経路データに従い、ピッキング物品が収められているラック200の前方まで移動する(S31)。ここで、ラック200の前方とは、自律移動マニプレータ100が備えるアーム部102によって、ピッキング指示401によって指定されたピッキング物品を、ラック200よりピッキング可能な範囲である。そして、ピッキング物品が収められているラック200の前まで移動が完了すると、ラック200の棚板210における物品の詳細な姿勢(位置および方向)を観測する(S32)そして、観測された物品位置および方向に基づき、ピッキングを実施する(S33)。なお、S31からS33に至る物品のピッキングにかかる動作の詳細は後述する。 Referring back to FIG. When receiving the picking instruction 401 from the management computer 400, the autonomous mobile manipulator 100 first moves to the front of the rack 200 in which the picking article is stored in accordance with the movement path data included in the picking instruction 401 (S31). Here, the front of the rack 200 is a range in which the picking article designated by the picking instruction 401 can be picked from the rack 200 by the arm unit 102 provided in the autonomous mobile manipulator 100. When the movement to the front of the rack 200 in which the picking article is stored is completed, the detailed posture (position and direction) of the article on the shelf plate 210 of the rack 200 is observed (S32), and the observed article position and Picking is performed based on the direction (S33). The details of the operation related to picking the article from S31 to S33 will be described later.
 一方、搬送ロボット300は、管理コンピュータ400から、移動指示402を受信すると、移動指示402に含まれる移動経路データに従い、物品のピッキングを実施する自律移動マニプレータ100の近傍である。ここで、近傍とは、自律移動マニプレータ100が備えるアーム部102の到達範囲内である。 On the other hand, when receiving the movement instruction 402 from the management computer 400, the transfer robot 300 is in the vicinity of the autonomous movement manipulator 100 that performs picking of the article according to the movement route data included in the movement instruction 402. Here, the vicinity is within the reach of the arm unit 102 provided in the autonomous mobile manipulator 100.
 ここで、ピッキング指示から自律移動マニプレータ100による物品のピッキングに至る動作(S23、S31、S32)と、移動指示から搬送ロボット300による自律移動マニプレータ近傍までの移動に至る動作(S24、S41)とは並列に実施される。そして、S32とS41の動作の両方が完了すると、管理コンピュータ400は、自律移動マニプレータ100に収納指示を送信する(S25)。自律移動マニプレータ100は、収納指示を受信すると、ピッキングした物品を、搬送ロボット300が積載するピッキングトレイ310に収納する(S33)。 Here, the operation from the picking instruction to picking the article by the autonomous mobile manipulator 100 (S23, S31, S32) and the action from the moving instruction to the vicinity of the autonomous mobile manipulator by the transfer robot 300 (S24, S41) Implemented in parallel. Then, when both the operations of S32 and S41 are completed, the management computer 400 transmits a storage instruction to the autonomous mobile manipulator 100 (S25). When receiving the storage instruction, the autonomous mobile manipulator 100 stores the picked article in the picking tray 310 on which the transport robot 300 is loaded (S33).
 管理コンピュータ400は、ステップS21にて作成したピッキング物品リストに含まれる物品の全てがピッキングされたかを確認し(S26)、完了していない場合は、再びステップS22からS26に至る一連の処理を繰り返す。 The management computer 400 checks whether all of the articles included in the picking article list created in step S21 have been picked (S26), and if not completed, repeats a series of processes from step S22 to S26. .
 一方、ピッキング物品リストに含まれる物品の全てがピッキングされた場合、管理コンピュータ400は、搬送ロボット300に、出荷エリアまでの移動経路データを含む移動指示402を送信する(S27)。搬送ロボット300は、管理コンピュータ400より、移動指示402を受信すると、移動指示402に含まれる移動経路データに従い、出荷エリアへと移動することで(S42)、ピッキング動作は完了となる。 On the other hand, when all of the articles included in the picking article list have been picked, the management computer 400 transmits a movement instruction 402 including movement route data to the shipping area to the transport robot 300 (S27). When receiving the movement instruction 402 from the management computer 400, the transfer robot 300 moves to the shipping area according to the movement route data included in the movement instruction 402 (S42), and the picking operation is completed.
 <自律移動マニプレータの構成>
 図5は、本実施例にかかる自律移動マニプレータ100の具体的な構成の一例を示した図面である。
<Configuration of autonomous mobile manipulator>
FIG. 5 is a diagram illustrating an example of a specific configuration of the autonomous mobile manipulator 100 according to the present embodiment.
 移動台車部101は、移動台車の進行方向(図面左手方向)に対して左右に取り付けられる2つの駆動車輪1011と、前後に2つ設けられたキャスタ1012と、周囲環境計測センサ1013とを更に備える。二つの駆動車輪1011は、独立して動作するもとのとして、駆動車輪1011の回転速度を制御することで、移動台車部101は、前進・後退・旋回などの所定の平面移動を出来る構成である。キャスタ1012は、移動台車部の走行に伴い車体が前後に傾くのを防ぐために前後に設けられる。なお、本発明において自律移動マニプレータ100の走行構成は、この移動台車部101の構成に限るものではなく、床面上の任意の位置姿勢に移動可能であれば良い。例えば、車輪軸を変更するステアリングを有する車輪を用いても良く、全方位移動可能なオムニホイールを用いても良く、脚型機構であっても良い。 The movable carriage unit 101 further includes two drive wheels 1011 attached to the left and right with respect to the traveling direction (left hand direction in the drawing) of the movable carriage, two casters 1012 provided in the front and rear, and an ambient environment measurement sensor 1013. . Assuming that the two drive wheels 1011 operate independently, by controlling the rotational speed of the drive wheels 1011, the movable carriage unit 101 can move in a predetermined plane such as forward, backward, and turn. is there. The casters 1012 are provided at the front and rear in order to prevent the vehicle body from tilting back and forth as the movable carriage unit travels. In the present invention, the traveling configuration of the autonomous mobile manipulator 100 is not limited to the configuration of the moving carriage unit 101, and may be any as long as it can move to any position and orientation on the floor surface. For example, a wheel having a steering for changing the wheel axis may be used, an omni wheel capable of omnidirectional movement may be used, or a leg type mechanism may be used.
 周囲環境計測センサ1013は、水平面上の障害物位置の分布を計測するレーザ距離センサとして実装される。ただし、本発明における周囲環境計測センサ1013は、レーザ距離センサに限ったものではなく、自律移動マニプレータ100の位置及び向きを認識するために十分な情報を取得可能とするものであれば良い。 The ambient environment measurement sensor 1013 is mounted as a laser distance sensor that measures the distribution of obstacle positions on a horizontal plane. However, the ambient environment measurement sensor 1013 in the present invention is not limited to the laser distance sensor, and may be any sensor that can acquire sufficient information to recognize the position and orientation of the autonomous mobile manipulator 100.
 アーム部102は、移動台車部101に固定され、固定された端から順に、6個の回転関節機構1021、1022、1023、1024、1025、1026が、直列に接続される。関節機構1021、1022、1023、1024、1025、1026は、それぞれ図中にJ1,J2,J3,J4,J5,J6によって示される矢印方向に、自身より延伸方向にある部位を回転させる機能を持つ。すなわち、この6つの回転自由度により、アーム部102の先端に取り付けられたエンドエフェクタ部103を、所定の3次元空間範囲内で任意の位置および姿勢に移動させることが可能である。 The arm unit 102 is fixed to the movable carriage unit 101, and six rotary joint mechanisms 1021, 1022, 1023, 1024, 1025, and 1026 are connected in series from the fixed end. The joint mechanisms 1021, 1022, 1023, 1024, 1025, and 1026 each have a function of rotating a portion that is in the extending direction from itself in the directions indicated by arrows J1, J2, J3, J4, J5, and J6 in the drawing. . That is, with the six degrees of freedom of rotation, the end effector portion 103 attached to the tip of the arm portion 102 can be moved to an arbitrary position and posture within a predetermined three-dimensional space range.
 エンドエフェクタ部103は、アーム部に固定される方向を基端として、更に延伸方向に2つの指機構1031aと1031bとを備え、指機構1031aおよび1031bを制御することにより、所定の大きさの任意の物体を把持可能である。なお、本発明において、エンドエフェクタ部103の構成は、ピッキング物品を把持可能な構成であれば、これに限るものではない。たとえば、3つ以上の指機構1031を備える多指グリッパや、真空吸着により物体を把持する吸着グリッパなどであってもよい。 The end effector unit 103 includes two finger mechanisms 1031a and 1031b in the extending direction, with the direction fixed to the arm unit as a base end, and controls the finger mechanisms 1031a and 1031b so as to have an arbitrary size. The object can be gripped. In the present invention, the configuration of the end effector portion 103 is not limited to this as long as it can grip the picking article. For example, a multi-finger gripper including three or more finger mechanisms 1031 or a suction gripper that grips an object by vacuum suction may be used.
 エンドエフェクタ部103には、更に物品計測センサ120を備える。物品計測センサ120は、視野内の障害物位置の分布を計測する距離画像センサ(デプスカメラとも呼ばれる)として実装される。ただし、本発明における物品計測センサ120は、距離画像センサに限ったものではなく、ピッキングする物品の姿勢を認識するために十分な情報を取得可能とするものであれば良く、カメラと距離画像センサなど複数のセンサの組み合わせとして実装されても良い。本実施例では、エンドエフェクタ部103の上に物品計測センサ120が取り付けられており、指機構1031の延伸方向に視野を向けているが、本発明における物品計測センサ120の取付け方法はこれに限ったものではない。たとえば、指機構1031の延伸方向に対して、仰角を設けることで、物品計測センサ120の視野と指機構1031の干渉を防ぐといった取付け方法であってもよい。 The end effector unit 103 further includes an article measurement sensor 120. The article measurement sensor 120 is implemented as a distance image sensor (also referred to as a depth camera) that measures the distribution of obstacle positions in the field of view. However, the article measurement sensor 120 in the present invention is not limited to the distance image sensor, and may be any sensor that can acquire sufficient information for recognizing the posture of the article to be picked. It may be mounted as a combination of a plurality of sensors. In this embodiment, the article measurement sensor 120 is attached on the end effector 103 and the field of view is directed in the extending direction of the finger mechanism 1031. However, the attachment method of the article measurement sensor 120 in the present invention is not limited to this. Not a thing. For example, an attachment method of preventing interference between the visual field of the article measurement sensor 120 and the finger mechanism 1031 by providing an elevation angle with respect to the extending direction of the finger mechanism 1031 may be used.
 図6は、自律移動マニプレータ制御部110の具体的な構成の一例を示した図面である。本実施例にかかる自律移動マニプレータ制御部110は、機能ブロックとして、全体管理部1100、送信部1101、受信部1102、画像認識部1110、物品データベース1111、空間管理部1120、作業手順計画部1130、作業手順データベース1131、動作遂行部1140、動作計画部1150、アーム制御部1160、エンドエフェクタ制御部1170、走行管理部1180、周囲環境地図1182、および移動台車制御部1182を備える。
FIG. 6 is a diagram illustrating an example of a specific configuration of the autonomous mobile manipulator control unit 110. The autonomous mobile manipulator control unit 110 according to the present embodiment includes, as functional blocks, an overall management unit 1100, a transmission unit 1101, a reception unit 1102, an image recognition unit 1110, an article database 1111, a space management unit 1120, a work procedure planning unit 1130, A work procedure database 1131, an operation execution unit 1140, an operation plan unit 1150, an arm control unit 1160, an end effector control unit 1170, a travel management unit 1180, an ambient environment map 1182, and a mobile cart control unit 1182 are provided.
 全体管理部1100は、受信部1102によって受信される管理コンピュータ400からのピッキング指令402に基づき、自律移動マニプレータ制御部110が具備する各機能ブロックへの詳細な処理指令を与える機能を有する。同時に、各機能ブロックの処理結果に基づき、管理コンピュータ400へと、自律移動マニプレータ100の状況を送信する。 The overall management unit 1100 has a function of giving a detailed processing command to each functional block included in the autonomous mobile manipulator control unit 110 based on the picking command 402 from the management computer 400 received by the receiving unit 1102. At the same time, the status of the autonomous mobile manipulator 100 is transmitted to the management computer 400 based on the processing result of each functional block.
 画像認識部1110は、物品計測センサ120によって計測される視野内の障害物位置の分布と、物品データベース1111に格納される物品情報と、全体管理部1100から与えられるピッキング物品IDとを用いて、所定の画像処理により、ピッキングする物品もしくは視野内の障害物の姿勢を算出する機能を有する。 The image recognition unit 1110 uses the distribution of obstacle positions in the field of view measured by the article measurement sensor 120, article information stored in the article database 1111, and a picking article ID given from the overall management unit 1100. It has a function of calculating the posture of an article to be picked or an obstacle in the field of view by predetermined image processing.
 空間管理部1120は、画像認識部1110によって算出されたピッキングする物品や障害物の姿勢(周囲物体姿勢情報と称する)を記憶しておき、必要に応じて、各機能ブロックへと供給する機能を有する。 The space management unit 1120 stores the posture of an article or obstacle to be picked calculated by the image recognition unit 1110 (referred to as surrounding object posture information), and supplies a function to each functional block as necessary. Have.
 作業手順計画部1130は、ピッキング作業手順を計画する機能ブロックである。ここで、ピッキング作業手順とは、自律移動マニプレータ100が備えるアーム部102もしくはエンドエフェクタ部103を、所定の空間範囲内において、どの姿勢に、どの順番で移動させるかを規定した手順のことを示す。作業手順データベース1131には、ピッキングする物品の形状と、棚板210の高さなど物品の収納されている所定の空間範囲に応じて決定される、作業手順テンプレートが格納されている。作業手順計画部1130は、全体管理部1100から与えられるピッキング物品IDおよび棚板IDと、作業手順データベース1141が記憶する作業手順テンプレートと、全空間管理部1120が記憶する周囲物体姿勢情報とを用いて、ピッキング作業手順を計画する。 The work procedure planning unit 1130 is a functional block that plans a picking work procedure. Here, the picking work procedure indicates a procedure that defines in which posture and in which order the arm unit 102 or the end effector unit 103 included in the autonomous mobile manipulator 100 is moved within a predetermined space range. . The work procedure database 1131 stores work procedure templates that are determined according to the shape of the article to be picked and a predetermined space range in which the article is stored such as the height of the shelf board 210. The work procedure planning unit 1130 uses the picking article ID and shelf ID given from the overall management unit 1100, the work procedure template stored in the work procedure database 1141, and the surrounding object posture information stored in the entire space management unit 1120. Plan the picking work procedure.
 動作遂行部1140は、全体管理部1100からの処理指令に基づいて、具体的なピッキング動作を遂行するための処理を行う機能ブロックである。動作遂行部1130は、全体管理部1100からの処理指令に基づいて、ピッキング動作を遂行する。具体的には、作業手順計画部1130によって計画された手順に従ってアーム部102もしくはエンドエフェクタ103を動作させために、動作計画部1150に、エンドエフェクタ部103などの自律移動マニプレータ100の所定の部位の目標姿勢を与える。 The operation execution unit 1140 is a functional block that performs processing for performing a specific picking operation based on a processing command from the overall management unit 1100. The operation performing unit 1130 performs a picking operation based on a processing command from the overall management unit 1100. Specifically, in order to operate the arm unit 102 or the end effector 103 in accordance with the procedure planned by the work procedure planning unit 1130, the operation planning unit 1150 causes the predetermined part of the autonomous mobile manipulator 100 such as the end effector unit 103 to be operated. Give the target posture.
 動作計画部1150は、動作遂行部1140から与えられる自律移動マニプレータ100の所定の部位の目標姿勢と、空間管理部1120に記憶される、周囲物体姿勢情報とを用いて、アーム部103の動作計画を行う機能を有する。ここで、動作計画とは、アーム部102の現在形態から、エンドエフェクタ部103の目標姿勢となるアーム部102の形態(目標形態)に至るまでの、所定の時間刻みで、アーム部103が有する関節機構1021、1022、1023、1024、1025、1026の回転角を算出する処理を指す。このように算出されたアーム部103の各関節機構の回転角の時系列変化を、アーム軌道と呼称する。また、動作計画部1150が有する動作計画機能は、アーム部102だけではなく、エンドエフェクタ部103の指機構1031の回転角も同様に算出する処理も含む。アーム軌道と同様に、エンドエフェクタの指機構1031の回転角の時系列変化を、指軌道と呼称する。 The motion planning unit 1150 uses the target posture of the predetermined part of the autonomous mobile manipulator 100 given from the motion performing unit 1140 and the surrounding object posture information stored in the space management unit 1120 to plan the motion of the arm unit 103. It has a function to perform. Here, the operation plan is a predetermined time interval from the current configuration of the arm unit 102 to the configuration (target configuration) of the arm unit 102 that is the target posture of the end effector unit 103. This refers to processing for calculating the rotation angle of the joint mechanisms 1021, 1022, 1023, 1024, 1025, 1026. A time-series change in the rotation angle of each joint mechanism of the arm unit 103 calculated in this way is referred to as an arm trajectory. Further, the motion planning function of the motion planning unit 1150 includes processing for calculating not only the arm unit 102 but also the rotation angle of the finger mechanism 1031 of the end effector unit 103 in the same manner. Similar to the arm trajectory, the time-series change in the rotation angle of the finger mechanism 1031 of the end effector is referred to as a finger trajectory.
 アーム制御部1160は、動作計画部1150によって算出されたアーム軌道に従い、アーム部102の各関節機構の回転角を変化させる制御を行うと共に、アーム部102の各関節機構の実際の回転角を計測する機能を有する。同様に、エンドエフェクタ制御部1170は、動作計画部1150によって算出された指軌道に従い、エンドエフェクタ部103の指機構1031の回転角を変化させる制御を行うと共に、エンドエフェクタ部103の指機構の実際の回転角を計測する機能を有する。 The arm control unit 1160 performs control to change the rotation angle of each joint mechanism of the arm unit 102 according to the arm trajectory calculated by the motion planning unit 1150, and measures the actual rotation angle of each joint mechanism of the arm unit 102. It has the function to do. Similarly, the end effector control unit 1170 performs control to change the rotation angle of the finger mechanism 1031 of the end effector unit 103 in accordance with the finger trajectory calculated by the motion planning unit 1150, and the actual finger mechanism of the end effector unit 103. It has a function to measure the rotation angle.
 走行管理部1180は、周囲環境計測センサ1013の計測結果と、周囲環境地図1181とを用いて、自身の平面上の位置および向きを時々刻々認識する。そして、全体管理部1100から与えられる移動経路と、自身の位置及び向きに基づき、移動台車部101の左右の車輪を制御する指令を算出する機能を有する。 The traveling management unit 1180 uses the measurement result of the ambient environment measurement sensor 1013 and the ambient environment map 1181 to recognize the position and orientation on its own plane from moment to moment. And based on the movement path | route given from the general management part 1100, and the own position and direction, it has the function to calculate the instruction | command which controls the left-right wheel of the mobile trolley part 101. FIG.
 なお、本発明にかかる自律移動マニプレータ制御部110の各機能ブロックは、自律移動マニプレータ制御部110に実装される所定のソフトウェアによって実装されてもよいし、一部もしくは全部がFPGA(Field Programmable Gate Array)、ASIC(Application Specific Integrated Circuit)などのハードウェアによって実装されてもよい。また以上の構成は、単体のコンピュータで構成してもよいし、あるいは、入力装置、出力装置、処理装置、記憶装置の任意の部分が、自律移動マニプレータ100の内部または外部に配置され、ネットワークで接続された他のコンピュータで構成されてもよい。 Note that each functional block of the autonomous mobile manipulator control unit 110 according to the present invention may be implemented by predetermined software installed in the autonomous mobile manipulator control unit 110, or a part or all of the functional blocks may be an FPGA (FieldmProgrammable Gate Array). ), Hardware such as ASIC (Application Specific Specific Integrated Circuit). The above configuration may be configured by a single computer, or any part of the input device, the output device, the processing device, and the storage device may be arranged inside or outside the autonomous mobile manipulator 100 and connected to the network. You may comprise with the other computer connected.
 <自律移動マニプレータの動作>
 次に、図7ないし図9を参照して、本実施例にかかる自律移動マニプレータ100において、一つの物品をピッキングする動作について説明する。当該動作においてピッキングの対象となる物品を、以降の説明では、ピッキング対象物品と呼称する。
<Operation of autonomous mobile manipulator>
Next, with reference to FIG. 7 thru | or FIG. 9, the operation | movement which picks one article | item in the autonomous mobile manipulator 100 concerning a present Example is demonstrated. In the following description, an article that is a target of picking in the operation is referred to as a picking target article.
 図7は、本実施例にかかる自律移動マニプレータ100の、棚前に移動する動作の一例を示したフローチャートであって、図2におけるステップS31を具体的に説明したものである。 FIG. 7 is a flowchart showing an example of the operation of the autonomous mobile manipulator 100 according to the present embodiment moving to the front of the shelf, and specifically illustrates step S31 in FIG.
 自律移動マニプレータ制御部の全体管理部1100は、管理コンピュータ400からピッキング指示401を受信すると(S3101)、作業手順計画部1130において、停止位置を計画する。ここで、停止位置の計画とは、ピッキング指示401の棚板IDにより指定される棚板210に収納される物品をピッキングするために適した移動台車101の停止位置を計画することを含む。具体的には、ラック200における前記棚板210の高さにより、前記棚板上方向にある物品をピッキングするために適した姿勢は異なるため、作業手順データベース1131に記憶される棚板IDに基づいて、ピッキング指示401の移動経路に対する停止位置オフセットを算出する処理などが含まれる。次に、自律移動マニプレータ制御部1100は、走行管理部1180に対して、移動処理指令を発行する(S3103)。走行管理部1180は、ピッキング指令401に含まれる移動経路に基づきラック付近まで移動し(S3104)、作業手順計画部1130によって計画された前記停止位置オフセットに基づき、ピッキング動作に適した位置に自律移動マニプレータ100を停止させるように停止位置を微調整する。
When the overall management unit 1100 of the autonomous mobile manipulator control unit receives the picking instruction 401 from the management computer 400 (S3101), the work procedure planning unit 1130 plans a stop position. Here, the planning of the stop position includes planning a stop position of the mobile carriage 101 suitable for picking an article stored in the shelf 210 specified by the shelf ID of the picking instruction 401. Specifically, the posture suitable for picking an article in the upward direction of the shelf differs depending on the height of the shelf 210 in the rack 200, and therefore, based on the shelf ID stored in the work procedure database 1131. Thus, processing for calculating a stop position offset for the movement path of the picking instruction 401 is included. Next, the autonomous mobile manipulator control unit 1100 issues a movement processing command to the travel management unit 1180 (S3103). The traveling management unit 1180 moves to the vicinity of the rack based on the movement route included in the picking command 401 (S3104), and autonomously moves to a position suitable for the picking operation based on the stop position offset planned by the work procedure planning unit 1130. The stop position is finely adjusted so that the manipulator 100 is stopped.
 図8は、本実施例にかかる自律移動マニプレータ100のピッキング物品姿勢観測動作の一例を示すフローチャートであって、図2におけるステップS32を具体的に説明したものである。 FIG. 8 is a flowchart showing an example of the picking article posture observation operation of the autonomous mobile manipulator 100 according to the present embodiment, and specifically explains step S32 in FIG.
 自律移動マニプレータ制御部110の全体管理部1100は、ピッキング対象物品の姿勢を観測する前に、空間管理部1120に記憶される周囲物体姿勢情報を検索し(S3201)、対象となるピッキング対象物品の姿勢が登録されているかチェックを行う(S3202)。ここで、対象となるピッキング対象物品の姿勢が登録されていた場合、すなわち、当該ピッキング指示前に、同一のラック200の棚板210から当該ピッキング対象物品とは異なる物品をピッキングした時に、当該ピッキング対象物品の姿勢を観測し、空間管理部1120に記憶させていた場合は、それを再利用する。これにより、ピッキング物品姿勢観測動作をスキップし、自律移動マニプレータのピッキング動作にかかる動作時間を短縮することができる。 Before observing the posture of the picking target article, the overall management unit 1100 of the autonomous mobile manipulator control unit 110 searches the surrounding object posture information stored in the space management unit 1120 (S3201) and searches for the target picking target item. It is checked whether the posture is registered (S3202). Here, when the posture of the target picking target item is registered, that is, when an item different from the target picking item is picked from the shelf 210 of the same rack 200 before the picking instruction, the picking target item is selected. When the posture of the target article is observed and stored in the space management unit 1120, it is reused. Thereby, the picking article posture observation operation can be skipped, and the operation time required for the picking operation of the autonomous mobile manipulator can be shortened.
 空間管理部1120に対象となるピッキング対象物品の姿勢が登録されていない場合、作業手順計画部1130において撮影手順を計画する(S3203)。ここで、撮影手順の計画とは、ピッキング指示401に含まれるラックIDおよび棚IDにより指定される棚板210において、物品を収納しうる収納空間を算出し、前記収納空間に対して物品計測センサ120の視野角と、画像処理部1110において物品検出処理を行うために必要な所定の解像度とから、ラック200に対する物品計測センサ120の計測姿勢と計測回数とを算出する処理を含む。 When the posture of the target picking target article is not registered in the space management unit 1120, the work procedure planning unit 1130 plans a shooting procedure (S3203). Here, the plan of the photographing procedure is to calculate a storage space in which articles can be stored in the shelf 210 specified by the rack ID and the shelf ID included in the picking instruction 401, and to measure the article measurement sensor with respect to the storage space. The processing includes calculating the measurement posture and the number of times of measurement of the article measurement sensor 120 with respect to the rack 200 from the viewing angle of 120 and a predetermined resolution necessary for performing the article detection process in the image processing unit 1110.
 次に、全体管理部1100は、動作遂行部1140に、撮影動作指令を発行する(S3204)。動作遂行部1140は、前記撮影手順によって計画された物品計測センサ120の計測姿勢のうち1つ、自律移動マニプレータ100の中心点を基準とした姿勢に変換し、動作計画部に1150に目標姿勢として指定する(S3205)。そして、動作計画部1150は、指定された物品計測センサ120の姿勢を目標とした、アーム軌道を計画する(S3206)。なお、動作計画の詳細は後述する。アーム制御部1160は、計画された前記アーム軌道に基づきアーム部102を制御することで、物品計測センサ120を移動させる(S3207)。 Next, the overall management unit 1100 issues a shooting operation command to the operation performing unit 1140 (S3204). The motion execution unit 1140 converts one of the measurement postures of the article measurement sensor 120 planned by the photographing procedure into a posture based on the central point of the autonomous mobile manipulator 100, and sets the target posture as the target posture in 1150. Specify (S3205). Then, the motion planning unit 1150 plans an arm trajectory targeting the posture of the specified article measurement sensor 120 (S3206). Details of the operation plan will be described later. The arm control unit 1160 moves the article measurement sensor 120 by controlling the arm unit 102 based on the planned arm trajectory (S3207).
 物品計測センサ120の移動が実質的に完了すると、全体管理部1100は、画像処理部1110に画像処理指令を発行する(S3208)。画像処理部1110は、物品計測センサ120を用いた棚板210の収納空間の計測データを取得し(S3209)、計測データに対して物品および障害物の検出処理を行う(S3210)。検出処理では、予め物体データベース1111に記憶させた物品の形状モデルと、計測データを用いて所定の方法で、物品の姿勢を認識する処理を指す。なお、ここで前記物品とはピッキング対象物品だけではなく、棚板210上に配置されうる全ての物品を指す。ステップS3210において検出された物品の姿勢は、空間管理部1120によって、周囲物体姿勢情報として記憶され(S3211)、前述のように必要に応じて再利用される。 When the movement of the article measurement sensor 120 is substantially completed, the overall management unit 1100 issues an image processing command to the image processing unit 1110 (S3208). The image processing unit 1110 acquires measurement data of the storage space of the shelf board 210 using the article measurement sensor 120 (S3209), and performs article and obstacle detection processing on the measurement data (S3210). The detection processing refers to processing for recognizing the posture of an article by a predetermined method using the shape model of the article stored in advance in the object database 1111 and measurement data. Here, the article refers to not only the article to be picked but also all articles that can be arranged on the shelf board 210. The posture of the article detected in step S3210 is stored as surrounding object posture information by the space management unit 1120 (S3211), and is reused as necessary as described above.
 ステップS3210において、ピッキング対象物品が検出されるか、ステップS3203によって計画された収納空間を計測するための計測回数の動作が完了された場合、ピッキング物体姿勢観測動作を終了する(S3210)。ピッキング対象物品が検出されておらず、且つ、収納空間を計測するための計測回数の動作が未完了の場合は、再びS3204に戻り一連の動作を繰り返す。なお、計測回数動作が完了したが、ピッキング対象物品が検出されていない場合、自律移動マニプレータ100の全体管理部1100は、管理コンピュータ400に、物品未検出エラーの報告を行う(図8には図示されない)。 In step S3210, if the picking target article is detected or the operation of the number of times for measuring the storage space planned in step S3203 is completed, the picking object posture observation operation is ended (S3210). If the picking target article has not been detected and the operation of the number of times for measuring the storage space has not been completed, the process returns to S3204 and the series of operations is repeated. When the measurement count operation is completed, but the picking target article is not detected, the overall management unit 1100 of the autonomous mobile manipulator 100 reports an article non-detection error to the management computer 400 (illustrated in FIG. 8). Not)
 図9は、本実施例にかかる自律移動マニプレータ100のピッキング動作の一例を示すフローチャートであって、図2におけるステップS33を具体的に説明したものである。ここでは、図8における物品探索動作により、ピッキング対象物品が検出され、前記ピッキング対象物品の姿勢が、空間管理部1120に記憶されているものとする。 FIG. 9 is a flowchart illustrating an example of the picking operation of the autonomous mobile manipulator 100 according to the present embodiment, and specifically illustrates step S33 in FIG. Here, it is assumed that the picking target article is detected by the article search operation in FIG. 8, and the posture of the picking target article is stored in the space management unit 1120.
 まず、作業手順計画部1120において、ピッキング手順を計画する(S3301)。ここで、ピッキング手順の計画とは、ピッキング指示401により指定されるピッキング対象物品を、エンドエフェクタ部104により把持するための、ピッキング対象物品に対するエンドエフェクタの姿勢関係を計画することを含む。前記把持姿勢は、ピッキング指示401に含まれるピッキング物品IDに紐づけて、予め作業手順データベース1131に記憶されているものとする。 First, the work procedure planning unit 1120 plans a picking procedure (S3301). Here, the planning of the picking procedure includes planning the posture relationship of the end effector with respect to the picking target article for gripping the picking target article specified by the picking instruction 401 by the end effector unit 104. It is assumed that the gripping posture is stored in advance in the work procedure database 1131 in association with the picking article ID included in the picking instruction 401.
 次に、全体管理部1100は、動作遂行部1140にピッキング指令を発行する(S3302)。動作遂行部1140は、空間管理部1120から、ピッキング対象物品と、その周囲の障害物となり得る物体の姿勢情報を検索する(S3303)。そして、ピッキング対象の物品の姿勢と、作業手順計画部1120において計画されたピッキング物品に対するエンドエフェクタ姿勢とから、自律移動マニプレータ100の中心点を基準としたエンドエフェクタ姿勢を算出し、動作計画部1150に目標として指定する(S3304)。そして、動作計画部1150は、目標として指定されたエンドエフェクタ姿勢に関して動作計画を行い(S3305)、アーム部を移動制御する(S3306)。 Next, the overall management unit 1100 issues a picking command to the operation performing unit 1140 (S3302). The motion performing unit 1140 searches the space management unit 1120 for the posture information of the picking target article and an object that can be an obstacle around it (S3303). Then, an end effector attitude with respect to the center point of the autonomous mobile manipulator 100 is calculated from the attitude of the article to be picked and the end effector attitude with respect to the picking article planned in the work procedure planning unit 1120, and the motion planning unit 1150. Is designated as a target (S3304). Then, the motion planning unit 1150 performs motion planning regarding the end effector posture designated as the target (S3305), and controls the movement of the arm unit (S3306).
 ステップS3306のアーム部の移動が実質的に完了すると、動作遂行部1140は、ピッキング対象物品を把持するための指機構の姿勢を、動作計画部1150に指定する(S3307)。動作計画部1150は、指定された指機構の姿勢を実現する指軌道を計画する(S3308)。指軌道の計画とは、実質的には指定した現在の指姿勢から、把持に至る目標指姿勢を、所定の時間刻みで補間する処理に相当し、詳細な説明は割愛する。そして、エンドエフェクタ制御部は、計画された指軌道に基づき、エンドエフェクタ部103の指機構1031を制御し、ピッキング物品の把持を行う(S3309)。 When the movement of the arm unit in step S3306 is substantially completed, the motion performing unit 1140 designates the posture of the finger mechanism for gripping the picking target article to the motion planning unit 1150 (S3307). The motion planning unit 1150 plans a finger trajectory that realizes the posture of the designated finger mechanism (S3308). The finger trajectory plan substantially corresponds to a process of interpolating the target finger posture from the designated current finger posture to the grasp in a predetermined time interval, and detailed description thereof is omitted. Then, the end effector control unit controls the finger mechanism 1031 of the end effector unit 103 based on the planned finger trajectory, and grips the picking article (S3309).
 <自律移動マニプレータの動作計画方法>
 図10ないし図15を参照して、本実施例にかかる自律移動マニプレータ100の動作計画方法について説明する。ここでは、図9におけるステップS3305に相当するピッキング動作にかかる動作計画を具体的に説明する。
<Operation planning method for autonomous mobile manipulator>
With reference to FIG. 10 thru | or FIG. 15, the operation | movement planning method of the autonomous mobile manipulator 100 concerning a present Example is demonstrated. Here, the operation plan for the picking operation corresponding to step S3305 in FIG. 9 will be specifically described.
 図10は、本実施例にかかる動作計画部1150の具体的な構成の一例を示したブロック図である。動作計画部1150は、マニプレータモデル1151と、逆運動学計算部1152と、基準姿勢算出部1153と、アーム姿勢選択部1154と、動作計画空間構築部1155と、アーム軌道探索部1156と、物体形状モデル1157とを具備する。 FIG. 10 is a block diagram illustrating an example of a specific configuration of the operation planning unit 1150 according to the present embodiment. The motion planning unit 1150 includes a manipulator model 1151, an inverse kinematics calculation unit 1152, a reference posture calculation unit 1153, an arm posture selection unit 1154, a motion plan space construction unit 1155, an arm trajectory search unit 1156, and an object shape. A model 1157.
 マニプレータモデル1151は、自律移動マニプレータ100の、各機構の幾何学形状と、回転関節機構など動作部位の動作軸・動作方向・動作範囲を含んだモデルデータである。逆運動学計算部1152は、自律移動マニプレータ100の所定の部位の姿勢が入力された場合、前記マニプレータモデル1151を用いて、自律移動マニプレータ100の各動作部位の取り得る値を算出する機能ブロックである。たとえば、エンドエフェクタ部103の姿勢を入力した場合、回転関節機構1021、1022、1023、1024、1025の回転角度を解析的もしくは近似的に算出する機能などが含まれる。基準形態算出部1153は、所定の基準となるアーム形態を算出する機能ブロックであり、詳細な説明は後述する。アーム形態選択部1154は、基準形態算出部1152が算出した基準アーム形態を基に、逆運動学計算部1152によって算出されたアーム部形態から、適切な目標アーム形態を選択する機能ブロックであり、詳細な説明は後述する。動作計画空間構築部1155は、空間管理部1120から入力される周囲物体姿勢情報と、予め記憶された物体の形状モデルとを組み合わせて、動作計画空間を構築する機能ブロックである。アーム軌道探索部は、前記動作計画空間において、現在のアーム部103の形態から、アーム形態選択部1154が選択した目標アーム形態に至るまでの、アーム軌道を探索する機能ブロックである。 The manipulator model 1151 is model data including the geometric shape of each mechanism of the autonomous mobile manipulator 100 and the motion axis, motion direction, and motion range of motion parts such as a rotary joint mechanism. The inverse kinematics calculation unit 1152 is a functional block that calculates possible values of each motion part of the autonomous mobile manipulator 100 using the manipulator model 1151 when a posture of a predetermined part of the autonomous mobile manipulator 100 is input. is there. For example, when the posture of the end effector unit 103 is input, a function of analytically or approximately calculating the rotation angle of the rotary joint mechanisms 1021, 1022, 1023, 1024, and 1025 is included. The reference form calculation unit 1153 is a functional block for calculating a predetermined reference arm form, and will be described in detail later. The arm configuration selection unit 1154 is a functional block that selects an appropriate target arm configuration from the arm configuration calculated by the inverse kinematics calculation unit 1152 based on the reference arm configuration calculated by the reference configuration calculation unit 1152. Detailed description will be given later. The motion plan space construction unit 1155 is a functional block that constructs a motion plan space by combining the surrounding object posture information input from the space management unit 1120 and the shape model of the object stored in advance. The arm trajectory search unit is a functional block that searches for an arm trajectory from the current configuration of the arm unit 103 to the target arm configuration selected by the arm configuration selection unit 1154 in the motion planning space.
 図11は、自律移動マニプレータ100のピッキング動作の一例の模式図である。ここで、自律移動マニプレータ100によって、棚板210の上の直置きされるピッキング対象物品900を、ピッキングする例について、動作計画部1150の動作を説明する。図11では、説明を簡単にするために、自律移動マニプレータ100のアーム部102の関節機構は1022、1023、1025の3つだけとする。すなわち、アーム部102およびエンドエフェクタ部103は、回転軸が同方向を向いた3自由度のマニプレータとして構成される。従って、エンドエフェクタ部104は図中に示すx軸およびz軸によって示される2次元平面内で、所定の範囲内における任意の姿勢(位置および方向)をとり得る。 FIG. 11 is a schematic diagram of an example of the picking operation of the autonomous mobile manipulator 100. Here, the operation of the operation planning unit 1150 will be described with respect to an example in which the picking target article 900 placed directly on the shelf board 210 is picked by the autonomous mobile manipulator 100. In FIG. 11, to simplify the description, there are only three joint mechanisms 1022, 1023, and 1025 of the arm unit 102 of the autonomous mobile manipulator 100. That is, the arm part 102 and the end effector part 103 are configured as a three-degree-of-freedom manipulator with the rotation axis facing the same direction. Therefore, the end effector unit 104 can take an arbitrary posture (position and direction) within a predetermined range within a two-dimensional plane indicated by the x-axis and the z-axis shown in the drawing.
 図12は、本実施例にかかる動作計画部1150の動作の一例を示すフローチャートである。前述のように、動作計画部1150が動作計画を実施するにあたり、動作遂行部1140よりエンドエフェクタ目標姿勢が指定される(P1)。動作遂行部1140より指定される前記エンドエフェクタ目標姿勢を、図11では指機構1031xaおよび1031xbとして示す。すなわち、エンドエフェクタ目標姿勢は、ピッキング対象物品900を、指機構1031xaおよび1031xbによって把持点Pによって把握する姿勢として指定されている。ここで把握点Pは、ピッキング対象物900を、指機構1031によって把握するために適した姿勢として、作業手順計画部1120により計画された点である。 FIG. 12 is a flowchart illustrating an example of the operation of the operation planning unit 1150 according to the present embodiment. As described above, when the motion planning unit 1150 executes the motion planning, the end effector target posture is designated by the motion performing unit 1140 (P1). The end effector target postures designated by the action performing unit 1140 are shown as finger mechanisms 1031xa and 1031xb in FIG. That is, the end effector target posture is specified as a posture in which the picking target article 900 is grasped by the grasping point P by the finger mechanisms 1031xa and 1031xb. Here, the grasping point P is a point planned by the work procedure planning unit 1120 as a posture suitable for grasping the picking object 900 by the finger mechanism 1031.
 次に動作計画部1150は、逆運動学計算部1152を用いて、把握点Pによってピッキング対象物品900を把握するエンドエフェクタ目標姿勢を実現するための、アーム形態を算出する(P2)。一般に、同一方向に関節回転軸を有するマニプレータでは、所定のエンドエフェクタの姿勢に対応する、アーム部形態は複数存在する。 Next, the motion planning unit 1150 uses the inverse kinematics calculation unit 1152 to calculate an arm configuration for realizing the end effector target posture for grasping the picking target article 900 from the grasping point P (P2). Generally, in a manipulator having joint rotation axes in the same direction, there are a plurality of arm part forms corresponding to a predetermined end effector posture.
 図13に、複数のアーム形態の一例を示す。図13に示すように、把握点Pによってピッキング対象物品900を把握するエンドエフェクタ目標姿勢を実現するアーム部形態は、移動台車部101を始点として、関節機構1022、1023a、1025の順に構成される形態(上肘形態と呼称する)と、同様に移動台車部101を始点として、関節機構1022、1023b、1025の順に構成される形態(下肘姿勢と呼称する)の2種類が存在する。 FIG. 13 shows an example of a plurality of arm forms. As shown in FIG. 13, the arm portion form that realizes the end effector target posture for grasping the picking target article 900 by the grasping point P is configured in the order of the joint mechanisms 1022, 1023 a, and 1025 starting from the movable carriage portion 101. There are two types: a form (referred to as an upper elbow form) and a form (referred to as a lower elbow posture) configured in the order of joint mechanisms 1022, 1023b, and 1025 starting from the moving carriage 101.
 ここで、下記数1に示すアーム形態の取り易さを示す可操作度を計算する。なお、可操作度は、数1のように計算されるスカラ量である。 Here, the operability indicating the ease of taking the arm form shown in the following equation 1 is calculated. The manipulability is a scalar amount calculated as shown in Equation 1.
Figure JPOXMLDOC01-appb-M000001
Figure JPOXMLDOC01-appb-M000001
 ただし、J(q)は、アーム部102のヤコビ行列であり、アーム部103の関節機構1022、1023、1025の回転角度をθ1022、θ1023、θ1025として(図11参照)、エンドエフェクタ部103の基点の位置座標をx、z、方向をψとしたとき(図11参照)、数2に示される関係式であらわされるものである。 However, J (q) is a Jacobian matrix of the arm unit 102, and the rotation angles of the joint mechanisms 1022 , 1023 , and 1025 of the arm unit 103 are θ 1022 , θ 1023 , and θ 1025 (see FIG. 11), and the end effector unit When the position coordinates of the base point 103 are x, z, and the direction is ψ (see FIG. 11), the relational expression shown in Equation 2 is given.
Figure JPOXMLDOC01-appb-M000002
Figure JPOXMLDOC01-appb-M000002
 ここで、関節機構1022と1023間の直線距離をL、関節機構1023と1025の間の直線距離をLすると、図12に示す自律移動マニプレータ100において、所定のアーム形態の可操作度wは下記数3によって算出される。 Here, when the linear distance between the joint mechanisms 1022 and 1023 is L 2 , and the linear distance between the joint mechanisms 1023 and 1025 is L 3 , the maneuverability w of a predetermined arm configuration is obtained in the autonomous mobile manipulator 100 shown in FIG. Is calculated by the following equation (3).
Figure JPOXMLDOC01-appb-M000003
Figure JPOXMLDOC01-appb-M000003
 数3より明らかなように、図12に示す自律移動マニプレータ100の可操作度は、関節機構1023の角度だけによって決まる。 As is clear from Equation 3, the operable degree of the autonomous mobile manipulator 100 shown in FIG. 12 is determined only by the angle of the joint mechanism 1023.
 図13Aを参照して、上肘姿勢および下肘姿勢の可操作度に着目すると、上肘姿勢の関節機構1023が成す角度θ1023と下肘姿勢の関節機構1023が成す角度θ1023とでは、θ1023=-θ1023が成立し、同一の可操作度となる。従って、特許文献1記載の技術のように、可操作度に着目しても、アーム部形態は一意に決定されない。 Referring to FIG. 13A, focusing on the manipulability of the upper elbow posture and the lower elbow posture, the angle a θ 1023 formed by the joint mechanism 1023 in the upper elbow posture and the angle b θ 1023 formed by the joint mechanism 1023 in the lower elbow posture Then, a θ 1023 = −b θ 1023 is established, and the same manipulability is obtained. Therefore, as in the technique described in Patent Document 1, even if attention is paid to the degree of manipulability, the arm portion form is not uniquely determined.
 そこで、本実施例にかかる動作計画部1150は、アーム部形態を一意に決定するために、基準形態算出部1153において、動作遂行部1140から指定される目標形態から一意に決まる基準直線を算出する(P3)。基準直線の具体的な一例を、図13に直線Lとして示す。すなわち、基準直線Lは、把握点Pと、回転関節1022の回転軸を結ぶ直線である。そして、基準直線L上に各関節機構が位置するアーム部形態を算出する(P4)。 Therefore, the motion planning unit 1150 according to the present embodiment calculates a reference straight line uniquely determined from the target form specified by the motion performing unit 1140 in the reference form calculation unit 1153 in order to uniquely determine the arm part form. (P3). A specific example of the reference straight line is shown as a straight line L in FIG. That is, the reference straight line L is a straight line connecting the grasp point P and the rotation axis of the rotary joint 1022. And the arm part form in which each joint mechanism is located on the reference | standard straight line L is calculated (P4).
 図14に示されるように、このアーム部形態は、回転関節機構1022、1023、1025が基準直線L上に配置されたアーム形態である。この形態を、以降の説明では基準形態と呼称する。 As shown in Fig. 14, this arm portion form is an arm form in which the rotary joint mechanisms 1022, 1023, and 1025 are arranged on the reference straight line L. This form is referred to as a reference form in the following description.
 次に、アーム部形態選択部1154において、逆運動学計算部で算出された複数のアーム部形態(図12の例では上肘形態と下肘形態)について、基準形態との距離を算出する。ここで距離の一例として、各回転関節機構の角度差の二乗を総和した値の平方根を用いる方法などがある。そして、算出された前記距離の最も小さいものを、アーム部の目標形態として、アーム軌道探索部1156に出力する(P5)。従って、図12の例においては、下肘姿勢がアーム部の目標形態として選択され、アーム軌道探索部1156に出力される。 Next, in the arm part form selecting part 1154, the distance from the reference form is calculated for a plurality of arm part forms (upper elbow form and lower elbow form in the example of FIG. 12) calculated by the inverse kinematics calculation part. Here, as an example of the distance, there is a method of using the square root of a value obtained by summing the squares of the angular differences of the rotary joint mechanisms. Then, the calculated smallest distance is output to the arm trajectory searching unit 1156 as the target form of the arm unit (P5). Accordingly, in the example of FIG. 12, the lower elbow posture is selected as the target form of the arm unit and is output to the arm trajectory search unit 1156.
 図13Bに回転関節機構の角度差を図示する。下側のアーム部形態の場合、関節1022の角度差はA1-A2であり、関節1023bの角度差はB1-B2である。同様に他の関節についても角度差を測定し、それらの二乗を総和した値の平方根を求める。これを、上側のアーム部形態と下側のアーム部形態について計算し、比較すると、図13Bの場合は、下側のアーム部形態のほうの距離が小さいことが分かる。回転関節機構の回転角度は、アーム姿勢の制御のためにシステムが通常用いるデータであるため、容易に計算することができる。このようにして決定されるアーム部の目標形態の特徴は下記の通りである。 FIG. 13B illustrates the angular difference of the rotary joint mechanism. In the case of the lower arm configuration, the angle difference of the joint 1022 is A1-A2, and the angle difference of the joint 1023b is B1-B2. Similarly, the angle difference is measured for other joints, and the square root of the sum of the squares of these is obtained. When this is calculated and compared for the upper arm configuration and the lower arm configuration, in the case of FIG. 13B, it can be seen that the lower arm configuration has a smaller distance. Since the rotation angle of the rotary joint mechanism is data normally used by the system for controlling the arm posture, it can be easily calculated. The characteristics of the target form of the arm portion determined in this way are as follows.
 目標となるアーム部形態は、回転関節1022の回転軸から把握点Pに向かう直線に最も近いアーム形態であり、図13から明らかなように、アーム部103の占有空間容積が他のアーム部形態と比較して小さくなる。ここで、アーム部103の占有空間とは、アーム部103を内包しうる所定の楕円体の体積である。アーム部の占有空間容積が小さければ周囲の障害物とアーム部が干渉する確率が低下し、即ち物品が密集したような倉庫内でのアーム部動作に適する。 The target arm configuration is an arm configuration that is closest to a straight line from the rotation axis of the rotary joint 1022 toward the grasp point P, and as is apparent from FIG. 13, the occupied space volume of the arm unit 103 is other arm configuration. Smaller than Here, the occupied space of the arm portion 103 is a volume of a predetermined ellipsoid that can enclose the arm portion 103. If the occupied space volume of the arm portion is small, the probability that the surrounding obstacle and the arm portion interfere with each other decreases, that is, it is suitable for the arm portion operation in a warehouse where articles are densely packed.
 距離の算出法の他の例としては、x軸およびz軸によって示される2次元平面内で、アームと基準直線で囲まれる面積を基準とすることが考えられる。例えば、面積が小さいほうが近いと判定する。図13Bの例では、下側のアームの位置のほうが、上側のアームの位置よりも、アームと基準直線で囲まれる面積が小さいので、距離が近いと判定する。 As another example of the method for calculating the distance, it is conceivable that the area surrounded by the arm and the reference line in the two-dimensional plane indicated by the x axis and the z axis is used as a reference. For example, it is determined that the smaller area is closer. In the example of FIG. 13B, the lower arm position is smaller than the upper arm position because the area surrounded by the arm and the reference line is smaller, so it is determined that the distance is shorter.
 距離の算出法の他の例としては、x軸およびz軸によって示される2次元平面内で、アームと基準直線の交点の数を基準とすることが考えられる。例えば、交点の数が多いほうが近いと判定する。図13Bの例では、下側のアームが基準直線(点線)と交わる交点の数は1で、上側のアームの交点の数は0なので、下側のアームのほうが距離が近いと判定する。 Another example of the distance calculation method is based on the number of intersections between the arm and the reference line in the two-dimensional plane indicated by the x-axis and the z-axis. For example, it is determined that a larger number of intersections is closer. In the example of FIG. 13B, since the number of intersections where the lower arm intersects the reference straight line (dotted line) is 1 and the number of intersections of the upper arm is 0, it is determined that the lower arm is closer in distance.
 図12を再度参照する。動作計画部1150では、アーム部軌道探索部1156で、現在のアーム部103の形態から、アーム部姿勢選択部1054により決定された目標形態に至るアーム軌道のうち、周囲の障害物とアーム部103もしくはエンドエフェクタ部104が干渉しない有効なアーム軌道を探索する(P6)。 Refer again to FIG. In the motion planning unit 1150, the arm unit trajectory searching unit 1156 detects the surrounding obstacles and the arm unit 103 out of the arm trajectory from the current configuration of the arm unit 103 to the target configuration determined by the arm unit posture selection unit 1054. Alternatively, an effective arm trajectory in which the end effector unit 104 does not interfere is searched (P6).
 アーム部軌道探索部1156では、非特許文献1もしくは2記載のような従来のサンプルベース軌道探索アルゴリズムを用いてアーム軌道を探索することができる。アーム部目標形態が一意に定まるため、従来のサンプルベース軌道探索アルゴリズムにおいても、十分短い時間で有効なアーム軌道を探索することが可能である。 The arm part trajectory search unit 1156 can search for an arm trajectory using a conventional sample-based trajectory search algorithm as described in Non-Patent Document 1 or 2. Since the arm target form is uniquely determined, an effective arm trajectory can be searched for in a sufficiently short time even in the conventional sample-based trajectory search algorithm.
 一方、更に探索時間を短縮するために、図15に示すようなアーム部軌道探索部1156を構成できる。 On the other hand, in order to further shorten the search time, an arm portion trajectory search portion 1156 as shown in FIG. 15 can be configured.
 図15においてアーム部軌道探索部1156は、従来のサンプルベース軌道探索アルゴリズム実行部11563と並列して、現在アーム部形態から目標アーム部形態に至るアーム軌道を、単純な関節角度値の線形補間によって算出する線形補間アーム軌道算出部11561と、補間されたアーム部軌道においてアーム部103やエンドエフェクタ部104が周囲の障害物と干渉していないかチェックする障害物干渉チェック部11562とが実行される構成である。一般に線形補間軌道の算出は、サンプルベース軌道探索アルゴリズムに対して、短時間で軌道の算出を可能とするが、障害物を迂回する経路を算出することができない。一方、サンプルベース軌道探索アルゴリズムは、有効な軌道の算出には時間を要するが、障害物を迂回するアーム部軌道を探索することが可能である。そこで、線形補間アーム軌道算出部11561によって算出された補間軌道において、障害物干渉チェック部11562によって障害物干渉がチェックされていなければ、サンプルベース軌道探索アルゴリズム実行部11563による軌道探索が完了するより先に、有効軌道選択部11564により選択されて出力する構成をとる。このような構成とすることで、よりアーム軌道を探索する時間を短縮することが可能となる。 In FIG. 15, the arm trajectory search unit 1156 is arranged in parallel with the conventional sample-based trajectory search algorithm execution unit 11563, and the arm trajectory from the current arm unit form to the target arm part form is obtained by simple linear interpolation of joint angle values. A linear interpolation arm trajectory calculation unit 11561 to be calculated and an obstacle interference check unit 11562 for checking whether the arm unit 103 and the end effector unit 104 interfere with surrounding obstacles in the interpolated arm unit trajectory are executed. It is a configuration. In general, calculation of a linear interpolation trajectory enables a trajectory to be calculated in a short time with respect to a sample-based trajectory search algorithm, but cannot calculate a path that bypasses an obstacle. On the other hand, the sample-based trajectory search algorithm takes time to calculate an effective trajectory, but can search for an arm trajectory that bypasses an obstacle. Therefore, in the interpolation trajectory calculated by the linear interpolation arm trajectory calculation unit 11561, if the obstacle interference check unit 11562 has not checked the obstacle interference, the trajectory search by the sample-based trajectory search algorithm execution unit 11563 is completed. In addition, the effective trajectory selection unit 11564 selects and outputs it. With such a configuration, it is possible to further shorten the time for searching for the arm trajectory.
 なお、図13においては、基準直線Lを、把握点Pと回転関節1022の回転軸とを結ぶ直線としたが、基準直線はこれに限るものではなく、自律移動マニプレータ100の所定の部位であればよい。 In FIG. 13, the reference straight line L is a straight line connecting the grasp point P and the rotation axis of the rotary joint 1022, but the reference straight line is not limited to this, and may be a predetermined part of the autonomous mobile manipulator 100. That's fine.
 図16は、自律移動マニプレータ100のピッキング物品姿勢観測動作における動作計画(図8におけるS3206)に、本方法を適用した例である。図16において、動作遂行部からは物品計測センサの姿勢、即ち図16における視点Cの位置および方向が指定される。この視点Cを実現するアーム部形態は、関節機構1022、1023c、1025に至る形態と、関節機構1022、1023d、1025に至る形態の2種類が存在する。そこで、基準直線として、回転関節1022の回転軸と、視点Cを結ぶ直線L’をとる。そして、この基準直線L’によって決まるアーム部基準形態を算出し、それに近いアーム部形態として、関節機構1022、1023d、1025に至る形態を選択することができる。 FIG. 16 is an example in which the present method is applied to an operation plan (S3206 in FIG. 8) in the picking article posture observation operation of the autonomous mobile manipulator 100. 16, the posture of the article measurement sensor, that is, the position and direction of the viewpoint C in FIG. There are two types of arm part forms that realize this viewpoint C: forms reaching joint mechanisms 1022, 1023 c, and 1025, and forms reaching joint mechanisms 1022, 1023 d, and 1025. Therefore, a straight line L ′ connecting the rotation axis of the rotary joint 1022 and the viewpoint C is taken as a reference straight line. Then, an arm part reference form determined by the reference straight line L 'is calculated, and a form reaching the joint mechanisms 1022, 1023d, and 1025 can be selected as an arm part form close thereto.
 以上、本発明の具体的実施形態を詳細に説明したが、これらは開示の発明の明確な理解のために提示された実施の例である。従って、特許請求の範囲から逸脱することなく、上述の実施形態には、さらに各種態様、変形例、および組み合わせが形成できることは言うまでもない。ある実施例の構成の一部を他の実施例の構成に置き換えることが可能であり、また、ある実施例の構成に他の実施例の構成を加えることが可能である。また、各実施例の構成の一部について、他の実施例の構成の追加・削除・置換をすることが可能である。 Although specific embodiments of the present invention have been described in detail above, these are examples of implementations presented for a clear understanding of the disclosed invention. Accordingly, it goes without saying that various aspects, modifications, and combinations can be further formed in the above-described embodiment without departing from the scope of the claims. A part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace the configurations of other embodiments with respect to a part of the configurations of the embodiments.
 たとえば、図11ないし図13、および図14では、3自由度に限定し平面上での動作について説明したが、3以上の自由度を持つアーム部103に対しても本質的に同様の考え方で、本発明を適用できることは明らかである。 For example, in FIGS. 11 to 13 and FIG. 14, the operation on the plane is limited to three degrees of freedom, but the arm portion 103 having three or more degrees of freedom is essentially the same way of thinking. It is clear that the present invention can be applied.
 また、以上の実施形態では移動台車部101を持つ自律移動マニプレータ100について説明したが、本発明はこれに限るものではない。たとえば、所定の箇所に固定された、アーム部103だけを移動させるような据え付け型マニプレータに対しても適用できる。 In the above embodiment, the autonomous mobile manipulator 100 having the mobile carriage unit 101 has been described. However, the present invention is not limited to this. For example, the present invention can also be applied to a stationary manipulator that is fixed at a predetermined location and moves only the arm portion 103.
 各種のマニプレータの制御に利用することができる。 Can be used to control various manipulators.
1    ピッキングロボットシステム
100  自律移動マニプレータ
101  移動台車部
102  アーム部
1021、1022、1023、1024、1025、1026 回転関節機構
103  エンドエフェクタ部
1031 指機構
110  自律移動マニプレータ制御部
120  手先カメラ
200  ラック
210  棚板
2000 ラック配置エリア
300  搬送ロボット
310  ピッキングトレイ
400  管理コンピュータ
500  入力端末
900  ピッキング対象物品
DESCRIPTION OF SYMBOLS 1 Picking robot system 100 Autonomous movement manipulator 101 Moving cart part 102 Arm part 1021, 1022, 1023, 1024, 1025, 1026 Rotating joint mechanism 103 End effector part 1031 Finger mechanism 110 Autonomous movement manipulator control part 120 Hand camera 200 Rack 210 Shelf board 2000 Rack placement area 300 Transport robot 310 Picking tray 400 Management computer 500 Input terminal 900 Picking target article

Claims (12)

  1.  複数の回転関節を備えたアーム部と、エンドエフェクタ部を備えたマニプレータにおいて、前記アーム部と関連付けられた基準点を所望の位置に位置づけるために、情報処理装置で実行される動作計画方法であって、
     前記マニプレータの第一関節の回転軸と前記基準点とを結ぶ直線を基準直線として設定する第1の処理、
     前記基準直線より定まる前記アーム部の形態を特異点形態として設定する第2の処理、
     前記基準点を所望の位置に位置づけるために前記アーム部が取り得る形態のうち、前記特異点形態に最も近い距離にある前記アーム部の形態を目標形態として設定する第3の処理、
     前記目標形態に至るアーム軌道を探索する第4の処理、
     を備えることを特徴とする、マニプレータの動作計画方法。
    An operation planning method executed by an information processing apparatus to position a reference point associated with the arm unit at a desired position in a manipulator having an arm unit having a plurality of rotary joints and an end effector unit. And
    A first process for setting a straight line connecting the rotation axis of the first joint of the manipulator and the reference point as a reference straight line;
    A second process of setting the form of the arm portion determined from the reference straight line as a singular point form;
    3rd process which sets the form of the said arm part in the distance nearest to the said singular point form as a target form among the forms which the said arm part can take in order to position the said reference point in a desired position,
    A fourth process for searching for an arm trajectory that reaches the target form;
    An operation planning method for a manipulator, comprising:
  2.  請求項1記載のマニプレータの動作計画方法であって、
     前記第一関節は前記マニプレータの基端に最も近い回転関節であり、
     前記第一関節と回転軸が並行であり前記第一関節よりアーム部延伸側に位置する第二関節を備え、
     前記第一関節機構と回転軸が並行であり前記第二関節よりアーム部延伸側に位置する第三関節を備え、
     前記基準点は前記エンドエフェクタ部の一部であり、
     前記特異点形態は、前記基準直線上に前記第一関節と前記第二関節と、前記第三関節とが位置する特異点形態であることを特徴とする、マニプレータの動作計画方法。
    A manipulator operation planning method according to claim 1,
    The first joint is a rotary joint closest to the proximal end of the manipulator;
    The first joint and the rotation axis are parallel, and includes a second joint located on the arm extension side from the first joint,
    The first joint mechanism and the rotation axis are parallel, and includes a third joint located on the arm part extension side from the second joint,
    The reference point is part of the end effector;
    The singularity form is a singularity form in which the first joint, the second joint, and the third joint are located on the reference straight line.
  3.  請求項1記載のマニプレータの動作計画方法であって、
     前記エンドエフェクタ部に物品計測センサを更に備え、
     前記物品計測センサの計測データにより、前記エンドエフェクタ部の姿勢を算出処理し、
    算出されたエンドエフェクタ部の姿勢から、前記基準点を算出する処理を備えることを特徴とするマニプレータの動作計画方法。
    A manipulator operation planning method according to claim 1,
    The end effector unit further includes an article measurement sensor,
    Based on the measurement data of the article measurement sensor, calculate the attitude of the end effector unit,
    A manipulator motion planning method, comprising: processing for calculating the reference point from the calculated attitude of the end effector unit.
  4.  請求項1記載のマニプレータの動作計画方法であって、
     前記第3の処理において、
     前記特異点形態における各回転関節機構の角度と、前記アーム部が取り得る形態における各回転関節機構の角度の差の二乗を総和した値の平方根を用いて、前記距離を評価することを特徴とする、マニプレータの動作計画方法。
    A manipulator operation planning method according to claim 1,
    In the third process,
    The distance is evaluated using a square root of a value obtained by summing a square of a difference between angles of the rotary joint mechanisms in the singular point form and angles of the rotary joint mechanisms in the form that the arm unit can take. A manipulator operation planning method.
  5.  移動台車部と、
     前記移動台車部に搭載されたアーム部と、
     前記アーム部の動作を制御する制御部を有し、
     前記アーム部は、前記移動台車側から順番に第一関節および第二関節を備え、
     前記制御部は、前記第二関節を基準にして前記移動台車部と反対側の前記アーム部に関連付けて、基準点を設定し、前記基準点を所望の位置に位置づけるアーム部の移動制御を行うマニプレータの制御システムであって、
     前記基準点を所望の位置に位置づけるために取り得る前記アーム部の形態を、複数の候補形態として算出する第1の機能と、
     前記基準点と前記第一関節を結ぶ基準直線を算出する第2の機能と、
     前記基準直線に一致する前記アーム部の形態を、特異点形態として算出する第3の機能と、
     前記特異点形態との比較結果に基づいて、前記複数の候補形態から1つのアーム部の形態を選択し、目標形態として設定する第4の機能と、
     前記目標形態に至るための、前記第一関節および第二関節の動作を決定する第5の機能と、
     を有する、マニプレータの制御システム。
    A mobile carriage section;
    An arm part mounted on the movable carriage part;
    A control unit for controlling the operation of the arm unit;
    The arm portion includes a first joint and a second joint in order from the movable carriage side,
    The control unit sets a reference point in association with the arm unit on the side opposite to the movable carriage unit with respect to the second joint, and performs movement control of the arm unit that positions the reference point at a desired position. A manipulator control system,
    A first function that calculates the form of the arm portion that can be taken to position the reference point at a desired position as a plurality of candidate forms;
    A second function for calculating a reference straight line connecting the reference point and the first joint;
    A third function for calculating the form of the arm portion that matches the reference straight line as a singular point form;
    Based on the comparison result with the singular point form, a fourth function of selecting a form of one arm part from the plurality of candidate forms and setting as a target form;
    A fifth function for determining movements of the first joint and the second joint to reach the target form;
    A manipulator control system.
  6.  前記第二関節を基準にして前記移動台車部と反対側の前記アーム部に、第三関節を備え、
     前記第三関節を基準にして前記移動台車部と反対側の前記アーム部に、エンドエフェクタ部を備え、
     前記エンドエフェクタ部の一部に前記基準点を設定し、
     前記第5の機能は、
     前記目標形態に至るための、前記第一関節、第二関節、および第三関節の動作を決定する機能を有する、請求項5記載のマニプレータの制御システム。
    The arm part on the opposite side of the movable carriage part with respect to the second joint is provided with a third joint,
    The arm part on the side opposite to the movable carriage part with respect to the third joint is provided with an end effector part,
    Set the reference point in a part of the end effector section,
    The fifth function is:
    The manipulator control system according to claim 5, which has a function of determining movements of the first joint, the second joint, and the third joint to reach the target form.
  7.  前記第4の機能は、
     前記特異点形態における各回転関節機構の角度と、前記複数の候補形態における各回転関節機構の角度の差の二乗を総和した値の平方根を用いて、前記距離を評価する、請求項5記載のマニプレータの制御システム。
    The fourth function is:
    The distance is evaluated using a square root of a value obtained by summing a square of a difference between angles of the rotary joint mechanisms in the singular point form and angles of the rotary joint mechanisms in the plurality of candidate forms. Manipulator control system.
  8.  前記制御部は、
     前記移動台車部に搭載されるか、あるいは、その一部または全部が前記移動台車部外にあって、無線もしくは有線回線によって接続されている、請求項5記載のマニプレータの制御システム。
    The controller is
    The control system for a manipulator according to claim 5, wherein the control system is mounted on the movable carriage unit, or a part or all of the movable carriage unit is outside the movable carriage unit and connected by a wireless or wired line.
  9.  移動台車部と、前記移動台車部に搭載されたアーム部を有するマニプレータであって、
     前記アーム部は、前記移動台車側から順番に第一関節、第二関節および第三関節を備え、
     前記アーム部は、
     前記第三関節の前記移動台車部と反対側の部位に設定された基準点を、所望の位置に位置づける移動制御装置により駆動されるものであり、
     前記移動制御装置は、
     前記基準点を所望の位置に位置づけるために取り得る前記アーム部の形態を、複数の候補形態として算出する第1の機能と、
     前記基準点と前記第一関節を結ぶ基準直線を算出する第2の機能と、
     前記基準直線に重なる前記アーム部の形態を、特異点形態として算出する第3の機能と、
     前記特異点形態との比較結果に基づいて、前記複数の候補形態から1つのアーム部の形態を選択し、目標形態として設定する第4の機能と、
     前記目標形態に至るための、前記第一関節、第二関節および第三関節の動作を決定する第5の機能と、
     を実行する、マニプレータ。
    A manipulator having a moving carriage part and an arm part mounted on the moving carriage part,
    The arm portion includes a first joint, a second joint, and a third joint in order from the movable carriage side,
    The arm portion is
    The third joint is driven by a movement control device that positions a reference point set at a part opposite to the moving carriage part of the third joint at a desired position,
    The movement control device includes:
    A first function that calculates the form of the arm portion that can be taken to position the reference point at a desired position as a plurality of candidate forms;
    A second function for calculating a reference straight line connecting the reference point and the first joint;
    A third function of calculating the form of the arm portion overlapping the reference straight line as a singular point form;
    Based on the comparison result with the singular point form, a fourth function of selecting a form of one arm part from the plurality of candidate forms and setting as a target form;
    A fifth function for determining movements of the first joint, the second joint, and the third joint to reach the target form;
    Run the manipulator.
  10.  前記第4の機能は、
     前記特異点形態に最も近い距離にある前記候補形態を目標形態として選択する、請求項9記載のマニプレータ。
    The fourth function is:
    The manipulator according to claim 9, wherein the candidate form that is closest to the singular point form is selected as a target form.
  11.  前記第4の機能は、
     前記特異点形態における各回転関節機構の角度と、前記複数の候補形態における各回転関節機構の角度の差の二乗を総和した値の平方根を用いて、前記距離を評価する、請求項9記載のマニプレータ。
    The fourth function is:
    The distance is evaluated using a square root of a value obtained by summing a square of a difference between an angle of each rotary joint mechanism in the singular point form and an angle of each rotary joint mechanism in the plurality of candidate forms. Manipulator.
  12.  前記アーム部において、前記第三関節の前記第二関節と逆側にカメラを備えており、
     前記基準点は前記カメラの視点である、請求項9記載のマニプレータ。
    In the arm portion, a camera is provided on the side opposite to the second joint of the third joint,
    The manipulator according to claim 9, wherein the reference point is a viewpoint of the camera.
PCT/JP2015/055256 2015-02-24 2015-02-24 Manipulator, motion planning method for manipulator, and control system for manipulator WO2016135861A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/JP2015/055256 WO2016135861A1 (en) 2015-02-24 2015-02-24 Manipulator, motion planning method for manipulator, and control system for manipulator
JP2017501607A JP6359756B2 (en) 2015-02-24 2015-02-24 Manipulator, manipulator operation planning method, and manipulator control system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2015/055256 WO2016135861A1 (en) 2015-02-24 2015-02-24 Manipulator, motion planning method for manipulator, and control system for manipulator

Publications (1)

Publication Number Publication Date
WO2016135861A1 true WO2016135861A1 (en) 2016-09-01

Family

ID=56788049

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2015/055256 WO2016135861A1 (en) 2015-02-24 2015-02-24 Manipulator, motion planning method for manipulator, and control system for manipulator

Country Status (2)

Country Link
JP (1) JP6359756B2 (en)
WO (1) WO2016135861A1 (en)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018020423A (en) * 2016-08-05 2018-02-08 株式会社日立製作所 Robot system and picking method
CN108051272A (en) * 2018-01-17 2018-05-18 吴泠霏 Method prepared by the automatic sand sample preparation facilities of robot and sand sample
CN112330878A (en) * 2019-08-01 2021-02-05 天津物网科技有限公司 Vending method and system of vending machine
CN113126614A (en) * 2020-01-10 2021-07-16 三菱重工业株式会社 Control system for vehicle, control method for vehicle, and program
EP3854534A1 (en) * 2020-01-23 2021-07-28 Seiko Epson Corporation Control method and control device for mobile robot, and robot system
EP3718931A4 (en) * 2018-02-26 2021-08-18 Kabushiki Kaisha Toshiba Control device, program, and system
CN113288477A (en) * 2021-07-01 2021-08-24 吕涛 Orthodontic self-ligating bracket system with open auxiliary groove
JP6957781B1 (en) * 2021-06-14 2021-11-02 Dmg森精機株式会社 Self-propelled device
WO2022030242A1 (en) * 2020-08-03 2022-02-10 ソニーグループ株式会社 Information processing device, information processing method, and program
CN114504385A (en) * 2022-01-18 2022-05-17 上海术航机器人有限公司 Surgical robot tip attitude adjustment method, system, device, and medium
JP7157853B1 (en) * 2021-06-30 2022-10-20 日本金銭機械株式会社 Article transport processing system
DE112021003148T5 (en) 2020-08-21 2023-05-04 Fanuc Corporation ROBOT CONTROL DEVICE
WO2023079809A1 (en) * 2021-11-02 2023-05-11 株式会社日立製作所 Trajectory planning device, trajectory planning method, and trajectory planning program
WO2023119350A1 (en) * 2021-12-20 2023-06-29 日本電気株式会社 Control device, control system, control method and storage medium

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109807892B (en) * 2019-02-19 2022-04-12 宁波凯德科技服务有限公司 Motion planning model of automobile welding robot

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0355179A (en) * 1989-07-25 1991-03-08 Canon Inc System of point-of-origin calibration and its method for arm of horizontal multi-joint type robot
JPH10128689A (en) * 1996-09-04 1998-05-19 Shinko Electric Co Ltd Visual correcting device of unmanned movable body
JP2003103481A (en) * 2001-09-28 2003-04-08 Honda Motor Co Ltd Attitude optimizing method and attitude optimizing device for articulated robot
JP2007203380A (en) * 2006-01-30 2007-08-16 Kawasaki Heavy Ind Ltd Instruction support device for robot
JP2009050936A (en) * 2007-08-24 2009-03-12 Nsk Ltd Interference determination device and leg wheel type robot
US20110093119A1 (en) * 2009-10-16 2011-04-21 Samsung Electronics Co., Ltd. Teaching and playback method based on control of redundancy resolution for robot and computer-readable medium controlling the same

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS62199382A (en) * 1986-02-26 1987-09-03 株式会社東芝 Method of setting position of origin of horizontal turning type robot
JP2011093015A (en) * 2009-10-27 2011-05-12 Ihi Corp Control device of hand-eye bin picking robot

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0355179A (en) * 1989-07-25 1991-03-08 Canon Inc System of point-of-origin calibration and its method for arm of horizontal multi-joint type robot
JPH10128689A (en) * 1996-09-04 1998-05-19 Shinko Electric Co Ltd Visual correcting device of unmanned movable body
JP2003103481A (en) * 2001-09-28 2003-04-08 Honda Motor Co Ltd Attitude optimizing method and attitude optimizing device for articulated robot
JP2007203380A (en) * 2006-01-30 2007-08-16 Kawasaki Heavy Ind Ltd Instruction support device for robot
JP2009050936A (en) * 2007-08-24 2009-03-12 Nsk Ltd Interference determination device and leg wheel type robot
US20110093119A1 (en) * 2009-10-16 2011-04-21 Samsung Electronics Co., Ltd. Teaching and playback method based on control of redundancy resolution for robot and computer-readable medium controlling the same

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018020423A (en) * 2016-08-05 2018-02-08 株式会社日立製作所 Robot system and picking method
CN108051272A (en) * 2018-01-17 2018-05-18 吴泠霏 Method prepared by the automatic sand sample preparation facilities of robot and sand sample
EP3718931A4 (en) * 2018-02-26 2021-08-18 Kabushiki Kaisha Toshiba Control device, program, and system
CN112330878A (en) * 2019-08-01 2021-02-05 天津物网科技有限公司 Vending method and system of vending machine
CN112330878B (en) * 2019-08-01 2022-06-17 天津物网科技有限公司 Vending method and system of vending machine
CN113126614A (en) * 2020-01-10 2021-07-16 三菱重工业株式会社 Control system for vehicle, control method for vehicle, and program
EP3854534A1 (en) * 2020-01-23 2021-07-28 Seiko Epson Corporation Control method and control device for mobile robot, and robot system
WO2022030242A1 (en) * 2020-08-03 2022-02-10 ソニーグループ株式会社 Information processing device, information processing method, and program
DE112021003148T5 (en) 2020-08-21 2023-05-04 Fanuc Corporation ROBOT CONTROL DEVICE
JP6957781B1 (en) * 2021-06-14 2021-11-02 Dmg森精機株式会社 Self-propelled device
JP2022190478A (en) * 2021-06-14 2022-12-26 Dmg森精機株式会社 Self-propelled device
JP7157853B1 (en) * 2021-06-30 2022-10-20 日本金銭機械株式会社 Article transport processing system
WO2023276236A1 (en) * 2021-06-30 2023-01-05 日本金銭機械株式会社 Article conveyance processing system
CN113288477A (en) * 2021-07-01 2021-08-24 吕涛 Orthodontic self-ligating bracket system with open auxiliary groove
WO2023079809A1 (en) * 2021-11-02 2023-05-11 株式会社日立製作所 Trajectory planning device, trajectory planning method, and trajectory planning program
WO2023119350A1 (en) * 2021-12-20 2023-06-29 日本電気株式会社 Control device, control system, control method and storage medium
CN114504385A (en) * 2022-01-18 2022-05-17 上海术航机器人有限公司 Surgical robot tip attitude adjustment method, system, device, and medium

Also Published As

Publication number Publication date
JP6359756B2 (en) 2018-07-18
JPWO2016135861A1 (en) 2017-06-22

Similar Documents

Publication Publication Date Title
JP6359756B2 (en) Manipulator, manipulator operation planning method, and manipulator control system
US11607804B2 (en) Robot configuration with three-dimensional lidar
US10759051B2 (en) Architecture and methods for robotic mobile manipulation system
JP5620445B2 (en) Article takeout device for determining holding position and posture of robot based on selection condition
US11584004B2 (en) Autonomous object learning by robots triggered by remote operators
US11407111B2 (en) Method and system to generate a 3D model for a robot scene
Bostelman et al. Survey of research for performance measurement of mobile manipulators
US10675759B2 (en) Interference region setting apparatus for mobile robot
JP6855492B2 (en) Robot system, robot system control device, and robot system control method
JP2019214084A (en) Route planning device, route planning method, and route planning program
JP6950638B2 (en) Manipulator controller, manipulator control method, and manipulator control program
US11597104B2 (en) Mobile robot sensor configuration
US20220296754A1 (en) Folding UV Array
CN111319038A (en) Track generation system and track generation method
JP6328796B2 (en) Manipulator control method, system, and manipulator
US20230330764A1 (en) Autonomous assembly robots
JP2008264901A (en) Mobile robot, its automatic coupling method, and drive control method for parallel link mechanism
WO2022264672A1 (en) Self-propelled device
WO2022014133A1 (en) Mobile manipulator, method for controlling same, program
US11407117B1 (en) Robot centered augmented reality system
Tan et al. An integrated robotic system for transporting surgical tools in hospitals
JP7475663B2 (en) Mobile manipulator and control method and program thereof
JP2023123358A (en) Robot and system
Cho Pose Selection of a Mobile Manipulator for a Pick and Place Task

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15883158

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2017501607

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15883158

Country of ref document: EP

Kind code of ref document: A1