JP2018158439A - Object handling device, control device, and calibration method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object handling device that comprises, for example, a manipulator making accuracy of access by an effector higher.SOLUTION: An object handling device of an embodiment comprises, for example, a base, a manipulator, a first camera, and a calibration processing part. The manipulator, which is provided on a base, has a link, a movable part, and a motor for rotatively driving the link. The first camera is provided in the manipulator. The calibration processing part performs calibration processing in a position of the movable part on the basis of a load generated in the manipulator and based on the result of photographing in a predetermined position of the base by the first camera and the amount of change in angle of rotation between the link and the motor when the manipulator is brought into contact with the base.SELECTED DRAWING: Figure 18A

Description

  Embodiments relate to an object handling device, a control device, and a calibration method.

  Conventionally, an object handling apparatus including a manipulator such as a robot arm having an effector is known.

Special Table 2003-533162

  For example, it would be beneficial if an object handling device having a manipulator with higher access accuracy by an effector could be obtained.

  The object handling apparatus according to the embodiment includes, for example, a base, a manipulator, a first camera, and a calibration processing unit. The manipulator is provided on the base, and includes a link, a movable part, and a motor that rotationally drives the link. The first camera is provided in the manipulator. The calibration processing unit calibrates the position of the movable part based on the result of photographing the predetermined position of the base by the first camera and the load generated on the manipulator based on the amount of change in the rotation angle of the link and the motor when the manipulator contacts the base. Execute the process.

FIG. 1 is a schematic and exemplary perspective view of an object handling system including an object handling apparatus according to an embodiment. FIG. 2 is a schematic and exemplary block diagram of the controller of the embodiment. FIG. 3A is a schematic and exemplary side view (partially sectional view) showing the configuration of the hand of the embodiment, and is a view showing a state in which the suction pad protrudes. FIG. 3B is a schematic and exemplary side view (partially sectional view) showing the configuration of the hand of the embodiment, and is a view showing a state in which the suction pad protrudes and tilts. FIG. 3C is a schematic and exemplary side view (partially sectional view) showing the configuration of the hand of the embodiment, and is a view showing a state in which the suction pad is stored. FIG. 3D is a side view (partially sectional view) seen from a direction orthogonal to the line of sight of FIG. 3C schematically showing the configuration of the hand of the embodiment, and shows a state in which the suction pad is housed. FIG. Drawing 4 is a typical example perspective view of another hand of an embodiment, and is a figure showing the state where a finger closed. FIG. 5A is a diagram illustrating a state in which the finger of FIG. 4 is opened. FIG. 5B is a diagram illustrating a state in which the finger of FIG. 4 is opened, and is a diagram seen from a direction different from FIG. 5A. FIG. 6 is a block diagram of the hand control unit of the embodiment. 7A is a schematic and exemplary side view showing an example of an operation procedure of the hand of FIG. FIG. 7B is a schematic and exemplary side view showing an example of the operation procedure of the hand of FIG. 4, and shows the next stage of FIG. 7A. FIG. 7C is a schematic and exemplary side view showing an example of the operation procedure of the hand of FIG. 4, and shows the next stage of FIG. 7B. 7D is a schematic and exemplary side view showing an example of the operation procedure of the hand of FIG. 4, and shows the next stage of FIG. 7C. FIG. 7E is a schematic and exemplary side view showing an example of the operation procedure of the hand of FIG. 4, and shows the next stage of FIG. 7D. FIG. 8 is a schematic and exemplary side view of still another hand according to the embodiment. FIG. 9 is a flowchart illustrating an example of a basic operation of the system including the object handling apparatus according to the embodiment. FIG. 10 is a flowchart illustrating another example of the basic operation of the system including the object handling apparatus according to the embodiment. FIG. 11 is a schematic and exemplary perspective view of the information acquisition system according to the embodiment. FIG. 12 is a table showing object information stored in the object database of the embodiment. FIG. 13 is a table showing grip information stored in the grip database of the embodiment. FIG. 14 is an exemplary flowchart showing a procedure for acquiring object information by the information acquisition system. FIG. 15 is a schematic diagram of a gripping basic system model of the embodiment. FIG. 16 is a schematic diagram of basic data of a gripping method according to the embodiment. FIG. 17 is a schematic diagram illustrating a part of a procedure for automatically generating grip information by the information acquisition system of the embodiment. FIG. 18A is an exemplary flowchart illustrating an auto-calibration procedure of the object handling device according to the embodiment. FIG. 18B is an explanatory diagram of a plane error between the base coordinate system and the manipulator coordinate system of the embodiment. FIG. 18C is a schematic diagram illustrating an image of a manipulator provided with a calibration pattern photographed by the camera of the embodiment. FIG. 19A is an exemplary flowchart illustrating a handling procedure by a hand based on the object database and the gripping database according to the embodiment. FIG. 19B is an exemplary flowchart illustrating a handling procedure by a hand based on the object database and the gripping database of the embodiment, and is a diagram illustrating a continuation of FIG. 19A. FIG. 20A is a schematic and exemplary diagram illustrating the operation via point of the embodiment, and is a diagram illustrating an initial position, an object, and a conveyance target position. FIG. 20B is a schematic and exemplary diagram illustrating the operation via point according to the embodiment, and is a diagram illustrating the operation via point from the initial position to the grip target position. FIG. 20C is a schematic and exemplary diagram illustrating the operation via point according to the embodiment, and is a diagram illustrating the operation via point from the pull-up position to the grip release position. FIG. 21A is a schematic and exemplary block diagram of the manipulator control unit of the embodiment. FIG. 21B is a schematic and exemplary explanatory diagram of an image feature amount target value according to the embodiment. FIG. 21C is a schematic and exemplary block diagram of another example of the manipulator control unit according to the embodiment. FIG. 21D is a flowchart illustrating a procedure for changing parameters in the manipulator control unit of the embodiment. FIG. 22A is a flowchart showing a procedure of the state change operation of FIG. 19A. FIG. 22B is an exemplary flowchart showing a procedure different from FIG. 22A of the state change operation of FIG. 19A. FIG. 23 is an exemplary flowchart showing a procedure of the grip search operation of FIG. 19B. FIG. 24 is an exemplary flowchart showing the procedure of the retry operation of FIG. 19B. FIG. 25A is an exemplary flowchart illustrating a procedure for executing a learning operation and database generation according to the embodiment. FIG. 25B is an exemplary flowchart illustrating a procedure of gripping and transporting in the learning / database generation mode of the embodiment. FIG. 26 is an exemplary schematic configuration diagram of a learning system in which learning data is shared by a plurality of object handling devices according to the embodiment. FIG. 27 is a schematic and exemplary configuration diagram showing a system of another embodiment different from the object handling system of FIG. FIG. 28 is a schematic and exemplary cross-sectional view of the manipulator of the embodiment. FIG. 29 is a schematic diagram showing one link of the manipulator of FIG. FIG. 30 is a block diagram of a general speed control system. FIG. 31 is a diagram showing the effect of fine tuning of physical parameters. FIG. 32 is an exemplary block diagram of an observer of the manipulator control unit of the embodiment. FIG. 33 is a block diagram of a manipulator control unit having the observer of FIG. FIG. 34 is a block diagram illustrating another example of the manipulator control unit of the embodiment. FIG. 35 is a block diagram illustrating another example of the observer according to the embodiment. FIG. 36 is a block diagram illustrating still another example of the observer according to the embodiment.

  Hereinafter, exemplary embodiments of the present invention are disclosed. The configurations and controls (technical features) of the embodiments described below, and the operations and results (effects) brought about by the configurations and controls are examples. Moreover, the same component is contained in the some structure of embodiment illustrated below. In the following, similar constituent elements are given common reference numerals, and redundant description is omitted.

[System configuration]
FIG. 1 is a perspective view of an object handling system (robot handling system) including an object handling apparatus 1 (handling robot) according to an embodiment.

  The object handling apparatus 1 is an apparatus that conveys an object P. The object P is, for example, a product, an article, a part, and the like, and the specifications such as the size and the hardness are various. The object P can also be referred to as an object or a transported object.

  As shown in FIG. 1, the object handling apparatus 1 includes a controller 100, a manipulator 20, a base 30, a transport mechanism 40, a processing unit 50, a moving mechanism 6, and the like. The base 30 is configured to be freely movable on the floor or the ground by a moving mechanism 6 including a multi-directional moving wheel, a rotating wheel, a crawler, a drive motor, and the like, whereby a controller fixed to the base 30 is configured. 100, the manipulator 20, the transport mechanism 40, and the processing unit 50 can move freely on the floor and the ground together with the base 30.

  The transport mechanism 40 is a transport device for the object P provided separately from the manipulator 20. Specifically, it has a roller, a belt conveyor, etc., and the operation is controlled by the controller 100 so that the object P is conveyed between a predetermined position in the object handling apparatus 1 and a predetermined position outside the object handling apparatus 1. can do. For example, when the manipulator 20 carries the object P into the object handling apparatus 1, the carrying mechanism 40 can carry the object P out of the object handling apparatus 1 after the carry-in. On the other hand, when the manipulator 20 unloads the object P from the object handling apparatus 1, the transport mechanism 40 can load the object P into the object handling apparatus 1 before the unloading. Further, as shown in FIG. 1, the transport mechanism 40 can transport the object P while being accommodated in the case B or in the case B unit.

  The processing unit 50 is provided on the base 30 and is used for temporary holding, placement, storage, and the like of the object P. The processing unit 50 may be referred to as a work place, a work space, a work area, a work table, a buffer, or the like. The processing unit 50 is provided with a plurality of marks 51. The mark 51 becomes a target for photographing by the cameras 21 and 33, and is used for transporting the object P, calibration described later, and the like. The processing unit 50 may be provided with an RF-ID reader / writer, a barcode reader, various sensors, and the like (not shown). The camera 21 is an example of a first camera, and the camera 33 is an example of a second camera. The processing unit 50 is an example of a placement unit, and the mark 51 is an example of a first mark.

  In addition to the vacuum pump 1a and the battery 1b, the base 30 can be equipped with a cylinder, a compressor (not shown), a user interface, a safety mechanism, and the like. For example, when the manipulator 20 sucks the object P, the vacuum pump 1a generates a negative pressure of air or gas for sucking the object P. The battery 1b is a power source such as a motor when the vacuum pump 1a and the manipulator 20 are driven by a motor, for example. The cylinder and the compressor can be used, for example, for generating positive pressure of air or gas when the manipulator 20 is pneumatically driven, suction release processing, driving of an electromagnetic valve, or the like. Moreover, a cylinder or a compressor can be used for generating negative pressure by connecting to an ejector. The user interface is, for example, an operation button, a switch, a keyboard, or the like. The safety mechanism is, for example, a light curtain or a collision detector.

  The manipulator 20 has an arm 2 and a hand 200. The arm 2 is a so-called multi-joint or multi-axis robot arm, and is driven by, for example, a plurality of servo motors. Further, the arm 2 may be, for example, a vertical articulated type, a horizontal articulated type, a type of robot in which a plurality of linear motion robots are combined, or the like. The arm 2 is provided with various sensors (not shown) such as a force sensor. The hand 200 is a mechanism that grips the object P (target object) by a multi-finger mechanism such as suction, jamming, and pinching. The hand 200 can be referred to as a gripping mechanism, an effector, or an end effector. The hand 200 has, for example, a small actuator, an elastic mechanism 200e (elastic portion, FIG. 3A, etc.) for obtaining flexibility. The hand 200 is provided with a camera 21 and various sensors (not shown) such as a force sensor. The detailed configuration of the hand 200 will be described later. The elastic mechanism 200e can also be referred to as an elastic joint. The arm 2 and the hand 200 are an example of a movable part.

  The work table 70 is, for example, a rack, a shelf, a table for performing assembly work, and the like, and provides a place where the object P is carried out or carried in by the object handling apparatus 1. A mark 71 is provided on the work table 70. The mark 71 becomes a target for photographing by the cameras 21 and 33 and is used for transporting the object P, calibration described later, and the like. Further, the system may include a robot 8 or the like different from the object handling apparatus 1. The mark 71 is an example of a second mark.

[Controller configuration]
FIG. 2 is a block diagram of the controller 100. As shown in FIG. 2, the controller 100 includes an arithmetic processing unit 101, various databases, a main storage unit (not shown), an auxiliary storage unit (not shown), and the like. Examples of the database include an object database 102a (object DB), a grip database 102b (grip DB), a robot database 102c (robot DB), and an environment database 102d (environment DB). The main storage unit is, for example, a ROM or a RAM, and the auxiliary storage unit is, for example, an HDD, an SSD, or the like. The database may be a part of the auxiliary storage unit.

  The arithmetic processing unit 101 includes a manipulator control unit 110, an image processing unit 101a, a signal processing unit 101b, a plan generation unit 101c, a gripping / motion plan generation unit 101d, a state monitoring unit 101e, an error detection unit 101f, a learning control unit 101g, An object recognition unit 101h and the like are included. The manipulator control unit 110 includes an arm control unit 120, a hand control unit 140, and the like.

  The image processing unit 101a processes various image sensor information and generates information necessary for operation planning, operation control, error detection, and learning.

  The signal processing unit 101b performs signal processing such as signal amplification processing, analog-digital conversion processing, noise removal, feature information extraction, and state change determination processing on various sensor information.

  The plan generation unit 101c generates and manages a work plan for the object handling apparatus 1 based on the user input information, the system state, and various sensor information.

  The gripping / motion plan generating unit 101d generates an motion plan related to the gripping and transfer operations of the object P by the manipulator 20. As an example, when gripping the object P, a process of calculating a movement via point of the hand 200 from the current position of the hand 200 to a position where the object P can be gripped without interfering with the surrounding environment is performed. . In this case, the gripping / motion plan generating unit 101d may calculate a motion plan including position information, time information, speed information, acceleration information, and the like.

  The arm control unit 120 mainly controls the operation of the arm 2. The arm control unit 120 includes a hand trajectory generation unit 121, an axis angle generation unit 122, an attitude monitoring unit 123, a drive control unit 124, and the like.

  The hand trajectory generator 121 generates hand trajectory information for controlling the motion trajectory of the hand of the manipulator 20 (hand 200). The hand trajectory information is, for example, information on the hand trajectory that enables smooth movement without imposing a load on the arm 2 as much as possible, and may include position information, time information, speed information, acceleration information, and the like.

  The shaft angle generation unit 122 generates angle information of a motor (not shown) that changes the angle of each joint of the arm 2 for the hand of the manipulator 20 (hand 200) to move according to the hand trajectory information. The angle information can include, for example, angle information, time information, angular velocity information, angular acceleration information, and the like.

  The posture monitoring unit 123 monitors the positions of a plurality of shaft driving units (not shown) included in the manipulator 20 and calculates the posture of the manipulator 20.

  The drive control unit 124 controls the operation (position) of the motor of the manipulator 20 and eventually the operation (position, posture, etc.) of the manipulator 20 according to the angle information generated by the shaft angle generation unit 122.

  Based on the posture information acquired from the information obtained from the posture monitoring unit 123, the state monitoring unit 101e generates operation state information indicating the operation state of the manipulator 20 such as, for example, operating, stopped, or operation complete. To do.

  The error detection unit 101f measures the state of the object handling apparatus 1, the execution state of the work plan, the drive control state, the grip state of the object P, the transport state of the object P, and the like, and detects an error.

  The learning control unit 101g is used for robot model learning for improving operation accuracy such as vibration suppression of the manipulator 20, gripping control parameter learning for improving the gripping performance of the object P, gripping database learning, and work plan execution performance improvement. Controls learning functions such as error detection learning.

  The object recognition unit 101h recognizes the object P based on the image processing result including the image of the object P by the image processing unit 101a.

  The database information generation / update unit 101i (DB information generation / update unit) generates and updates object information stored in the object database 102a and grip information stored in the grip database 102b. The processing of the database information generation / update unit 101i will be described in detail later.

  The automatic calibration unit 101j performs auto calibration. Auto calibration will be described in detail later. The automatic calibration unit 101j is an example of a calibration processing unit.

  The object database 102a stores attribute information of the object P. Hereinafter, the attribute information of the object P is referred to as object information. In addition, for each object P, various information related to control for gripping the object is stored in the gripping database 102b. The object database 102a and the grip database 102b will be described in detail later.

  In the robot database 102c, the structure of the object handling apparatus 1, specifications of each part (for example, dimensions, weight, moment of inertia, etc.), and performance specifications (for example, operation range, speed, Torque performance etc.) are stored.

  The environment database 102d stores information on the work table 70 that the object handling apparatus 1 supports, and ambient environment information such as the operation range of the object handling apparatus 1 and surrounding interferences.

[Hand configuration example 1]
3A to 3D are side views (partially sectional views) showing a configuration example of the hand 200A (200). A hand 200A shown in FIGS. 3A to 3D has a plurality of gripping mechanisms such as suction pads 200a and fingers 200b. Further, the suction pad 200a may be configured by a jamming mechanism or the like. That is, the hand 200A can grip the object P by a plurality of different gripping methods. Further, the hand control unit 140 of the controller 100 can select and use a gripping mechanism suitable for the object P when gripping the object P. The finger 200b is a multi-finger finger or a multi-joint finger that moves in parallel.

  As shown in FIG. 3B, the hand 200A is configured to be able to change the posture (angle) of the suction pad 200a by controlling the posture variable mechanism 200d by the hand control unit 140 of the controller 100. Therefore, the hand 200A can hold the object P more reliably.

  The hand 200A has an elastic mechanism 200e and a jamming portion 200c. The force applied to the object P from the hand 200A is suppressed by the elastic mechanism 200e and the jamming part 200c. The hand 200A has a clutch 200f that switches between an elastically deformable state and an impossible state of the elastic mechanism 200e. Therefore, the vibration of the hand 200A due to the elasticity of the elastic mechanism 200e can be reduced by limiting the relative displacement of the elastic mechanism 200e in two places where the elastic mechanism 200e can be relatively displaced by the clutch 200f while the object P is being gripped and conveyed. Can be suppressed. For example, the jamming portion 200c is provided so as to cover the elastic expansion / contraction mechanism, and by fixing the state due to jamming, the relative displacement of the expansion / contraction portion is limited similarly to the clutch 200f, and the hand 200A caused by elasticity is limited. Vibration can be suppressed.

  The hand 200A is provided with a contact sensor 200g, a force sensor 200h, and a displacement sensor 200i (for example, a non-contact laser displacement meter). Furthermore, the camera 21 provided in the hand 200A is, for example, a camera (for example, an RGBD sensor) that can detect distance information. Therefore, the arm control unit 120 and the hand control unit 140 of the controller 100 control the control target such as a motor based on the detection results of these sensors and cameras, thereby further suppressing the force applied from the hand 200 to the object P. Can be done.

  Further, the suction pad 200a is configured to be movable between the protruding position pa (FIGS. 3A and 3B) and the storage position pb (FIGS. 3C and 3D), whereby the object P can be gripped by the suction pad 200a. Can be switched between a state (FIGS. 3A and 3B) and a state where the object P can be gripped by the finger 200b (FIGS. 3C and 3D).

[Hand configuration example 2]
4, 5A, and 5B are perspective views showing a configuration example of the hand 200B (200). 4 shows a state in which the finger 200b is closed, FIGS. 5A and 5B show a state in which the finger 200b is opened, and FIG. 5B is a view as seen from a direction different from FIG. 5A.

  The hand 200B has two fingers 200b provided on the base 200j so as to be movable in parallel through guides 200k. The guide 200k includes, for example, a rail and a slider, and can guide the finger 200b to be movable along a linear track and can be positioned at a plurality of positions on the track. With such a configuration, the plurality of fingers 200b can change the distance (interval) between each other. In addition, the guide direction of the finger 200b by the guide 200k is an intersecting direction (orthogonal direction) with the longitudinal direction of the distal end portion of the arm 2.

  The hand 200B has a rotation / linear motion conversion mechanism 200m. The rotation / linear motion conversion mechanism 200m is a mechanism that converts the rotation of a motor (not shown) accommodated in the base 200j to linear motion, and is, for example, a rack and pinion. With such a configuration, by controlling the motor by the hand control unit 140 of the controller 100, the position of the finger 200b, that is, the distance (interval) between the two fingers 200b can be controlled. In the present embodiment, the rotation / linear motion conversion mechanism 200m is provided corresponding to each of the fingers 200b. However, the rotation / linear motion conversion mechanism 200m is not limited to this, and various configurations and methods may be employed. As an example, the rotation / linear motion converting mechanism 200m has a pinion that rotates in conjunction with a motor and two parallel racks that mesh with the pinion, and each rack translates the fingers 200b in parallel along the guide 200k. As another example, the rotation / linear motion conversion mechanism 200m may be a screw mechanism having a female screw nut rotated by a motor and a male screw shaft that is prevented from rotating by the base 200j and linearly moves. Also good. Further, a reduction mechanism, a rotation direction conversion mechanism, or the like may be interposed between the motor and the rotation / linear motion conversion mechanism 200m.

  The finger 200b has a floating structure (elastic expansion / contraction structure) in which a distal end portion 200b2 is elastically extended and supported by a moving base 200b1 guided along a guide 200k. The tip portion 200b2 is supported by the moving base 200b1 via the elastic mechanism 200e so as to be movable in the direction in which the fingers 200b extend. The moving direction of the tip portion 200b2 with respect to the moving base 200b1 is the longitudinal direction of the tip portion of the arm 2. The displacement sensor 200i measures the relative displacement between the movement base 200b1 and the tip portion 200b2. The displacement sensor 200i may be an encoder, a super musical note sensor, a variable resistor, a capacitance sensor, a pulse coder, a fiber sensor, or the like. In the present embodiment, the finger 200b is tapered and plate-shaped, and the elastic mechanism 200e and the displacement sensor 200i are arranged apart from the tip of the finger 200b. With such a configuration, it is possible to obtain an effect that the finger 200b easily enters a narrow gap or the like.

  FIG. 6 is a block diagram of the hand control unit 140. The hand control unit 140 includes a command generation unit 140a, a target value generation unit 140b, a drive control unit 140c, a determination unit 140d, and a driver 140e.

  The command generation unit 140a operates an operation procedure necessary for each work process in accordance with an operation command input from the grasping / operation plan generation unit 101d (FIG. 2) via the plan generation unit 101c (FIG. 2). Generate as a command. The gripping / motion plan generating unit 101d, the plan generating unit 101c, and the command generating unit 140a can refer to the object database 102a and the gripping database 102b in the respective arithmetic processing. The use of the object database 102a and the grip database 102b will be described in detail later. The operation command is a command related to a series of operations of the arm 2 of the manipulator 20 and the hand 200, and is held in the form of a program, for example. The operation command may be given by the operator touching a command command displayed on a user interface (not shown) or may be given by a voice instruction of the worker.

  The target value generation unit 140b receives an operation command command for a drive unit (movable unit) such as a motor from the command generation unit 140a. The target value generation unit 140b calculates a target value for the drive unit, and generates a target command value for driving the drive unit.

  The drive control unit 140c receives the target command value of the drive unit from the target value generation unit 140b, and generates a drive instruction for driving the drive unit according to the target command value.

  The determination unit 140d receives detection results from various sensors from the signal processing unit 101b and also receives an operation command from the command generation unit 140a. The determination unit 140d determines whether or not the operation according to the operation command is obtained based on the detection results from the various sensors and the operation command, and returns to the command generation unit 140a according to the situation. A value command (correction command) can be given. The determination unit 140d can give a drive stop command to the drive control unit 140c according to the situation.

  The driver 140e receives a drive command for the drive unit from the drive control unit 140c, and generates a drive output for the drive unit. That is, the drive unit is driven by the driver 140e.

[Hand operation example]
7A to 7E are schematic diagrams illustrating an example of an operation procedure of the hand 200B. As illustrated in FIG. 7A, the hand 200 </ b> B provided at the distal end portion of the arm 2 is lowered from above the object P to a predetermined position in accordance with the operation of the arm 2 controlled by the arm control unit 120. At this time, the two fingers 200b of the hand 200B are closed.

  At this time, as shown in FIG. 7A, the arm control unit 120 drives the arm 2 based on the plan planned by the plan generation unit 101c, images taken by the camera 21, detection results of various sensors, and the like. The manipulator 20 can be stopped from descending at a predetermined position spaced above the object P.

  Next, as illustrated in FIG. 7B, the hand control unit 140 uses the two fingers 200b based on the plan planned by the plan generation unit 101c, images taken by the camera 21, detection results of various sensors, and the like. Is controlled to be slightly wider than the width of the object P.

  Next, as illustrated in FIG. 7C, the arm control unit 120 determines that the hand 200 </ b> B is an object based on the plan planned by the plan generation unit 101 c, images taken by the camera 21, detection results of various sensors, and the like. The drive unit of the arm 2 is controlled so as to descend to a predetermined position (gripping position) for gripping P.

  Next, as shown in FIG. 7D, the hand control unit 140 uses the two fingers 200b based on the plan planned by the plan generation unit 101c, the images taken by the camera 21, the detection results of various sensors, and the like. The driving unit of the hand 200B is controlled so that the distance between the two fingers 200b is decreased and the object P is gripped by the two fingers 200b.

  Next, as shown in FIG. 7E, the arm control unit 120 detects that the hand 200B is an object based on the plan planned by the plan generation unit 101c, images taken by the camera 21, detection results of various sensors, and the like. The drive unit of the arm 2 is controlled so as to rise to a predetermined position above the gripping position while gripping P.

  The arm control unit 120 and the hand control unit 140 can execute a series of operations of the arm 2 and the hand 200B in accordance with a plan planned in advance by the plan generation unit 101c. In addition, the arm control unit 120 and the hand control unit 140 can determine that the hand 200B is in contact with the object P or the processing unit 50 from the image taken by the camera 21, the detection result of the sensor, or the like. The arm 2 and the hand 200B may be appropriately controlled according to the results of the camera, sensor, and the like, such as stopping the descent of the arm 2 at the contact position and moving the arm 2 slightly upward. A series of operations of the descending, grasping, and ascending hands 200B can be executed regardless of the grasping method of the object P.

[Example 3 of hand configuration]
FIG. 8 is a perspective view showing a configuration example of the hand 200C (200). As shown in FIG. 8, in the hand 200C, a suction pad 200a is provided at the tip of the finger 200b. In this case, the hand control unit 140 can select whether the object P is gripped by the suction pad 200a or the object 200 is gripped by the finger 200b according to the object P or according to other situations. In this case, it is preferable that the suction pad 200a is rotatably provided on the finger 200b via the posture changing mechanism 200d as in the hand 200A of FIG. 3B.

[Basic operation of the system 1 (supply of parts to the workbench)]
FIG. 9 is a flowchart showing an example of the basic operation of the system. Here, the transport mechanism 40 acquires (loads) a part as an object P accommodated in a part car as the case B into the object handling apparatus 1 together with the part car, and the manipulator 20 sends the part to the work table 70. The case of carrying out is illustrated.

  First, when the system is activated or reset in the controller 100 and an operation instruction is received from the host system (S101), the controller 100 first executes an initial calibration operation (S102). In this case, if there is no particular problem, the controller 100 notifies the upper system that there is no problem. As a result, the controller 100 instructs the controller 100 to supply the parts to the work table 70 or starts control based on the operation instruction (supply instruction) received at the beginning. The initial calibration operation will be described in detail later.

  Next, the controller 100 transports a component cage containing the component (part to be supplied) from a predetermined position outside the object handling device 1, for example, from a component shelf to a predetermined position in the object handling device 1. 40 is controlled (S103).

  Next, the controller 100 controls the camera 33 provided on the support column 32 of the base 30 (FIG. 1) so as to photograph the part cage (S104).

  Next, the controller 100 performs confirmation of the presence / absence of a supply target component in the component cage, and calculation of the position, orientation, and gripping position of the component in the component cage based on the image taken by the camera 33. Execute (S105).

  Next, the controller 100 controls the manipulator 20 to hold the supply target component in the component cage (S106).

  Next, the controller 100 confirms the gripping state of the parts to be supplied by the hand 200 based on images taken by the cameras 21 and 33, detection results of various sensors, and the like (S107).

  Next, the controller 100 confirms whether or not the work table 70 (conveyance destination) is in a state in which parts can be supplied based on images taken by the cameras 21 and 33 (S108).

  Next, the controller 100 controls the manipulator 20 so as to place the component gripped by the hand 200 at a specified position on the work table 70 in a predetermined posture (S109).

  When the series of work ends, the controller 100 controls the manipulator 20 to move to the standby position (S110).

[Basic operation of system 2 (part removal from work bench)]
FIG. 10 is a flowchart showing another example of the basic operation of the system. Here, a case where the manipulator 20 acquires (loads) a part as an object P from the work table 70 into the object handling apparatus 1 and the transport mechanism 40 carries the part out of the object handling apparatus 1 is exemplified.

  First, when the system is activated or reset in the controller 100 and an operation instruction is received from the host system (S201), the controller 100 first causes the hand 200 to move to a predetermined position near the work table 70. Then, the manipulator 20 is controlled (S202).

  Next, the controller 100 recognizes the position and orientation of the work table 70 based on the image of the mark 71 photographed by the camera 21 of the hand 200 and the virtual work space (calculation) based on the recognized result. The position data of the work table 70 in the space is corrected (S203).

  Next, the controller 100 controls the camera 21 of the hand 200 so as to photograph a part (part to be taken out) present on the work table 70, and recognizes the part by image processing or the like on the photographed image (S204). ).

  Next, the controller 100 controls the manipulator 20 so as to grip the part to be taken out present on the work table 70 (S205).

  Next, the controller 100 confirms the gripping state of the part to be picked up by the hand 200 based on images taken by the cameras 21 and 33, detection results of various sensors, and the like (S206).

  Next, the controller 100 confirms the state of the transport destination based on images taken by the cameras 21 and 33, detection results of various sensors, and the like (S207).

  Next, the controller 100 controls the manipulator 20 so as to place the component gripped by the hand 200 at a specified position of the object handling apparatus 1 in a predetermined posture (S208). The designated position in this case is on the transport mechanism 40, for example.

  Next, the controller 100 confirms the state of the part to be taken out, the barcode, and the like. Further, the object database 102a is updated (S209).

  Next, the controller 100 has a transport mechanism so that the component cage containing the component (part to be supplied) is transported from a predetermined position in the object handling device 1 to a predetermined position outside the object handling device 1, for example, a component shelf. 40 is controlled (S210).

[Configuration of information acquisition system]
FIG. 11 is a perspective view of the information acquisition system 300. The information acquisition system 300 shown in FIG. 11 includes a slide stage 301, a rotary stage 302, a position adjustment mechanism 303, a sensor arm 304, a slide mechanism 305, an RGB camera 306, a distance image camera 307, a laser sensor 308, an illumination arm 309, An illumination 310, a controller 320, and the like are provided.

  The controller 320 controls each unit so that the RGB camera 306 captures images from various directions of the object P, and the distance image camera 307 captures distance images from various directions of the object P. The object P is placed on the rotary stage 302. The position adjustment mechanism 303 changes the position of the rotary stage 302 in the height direction, and the slide stage 301 changes the position of the rotary stage 302 in the horizontal direction. The sensor arm 304 is rotatably supported by the position adjustment mechanism 303. In such a configuration, the controller 320 sets or adjusts the horizontal position of the slide stage 301, the angle of the rotary stage 302, the degree of expansion / contraction of the position adjustment mechanism 303, and the angle of the sensor arm 304, thereby adjusting the RGB camera 306. The shooting direction and shooting distance of the object P by the distance image camera 307 can be changed. Hereinafter, the slide stage 301, the rotary stage 302, the position adjustment mechanism 303, and the sensor arm 304 may be referred to as a movable mechanism. The distance measurement result by the laser sensor 308 can be used for calibration of the measurement distance by the distance image camera 307. In addition, the controller 320 can adjust the illumination condition by the illumination 310.

  The controller 320 generates object information of the object P based on the photographed image and the detection result by the sensor. The generated object information is stored in the object database 102a. Further, the controller 320 can generate grip information for gripping the object P based on the object information of the object P. The generated grip information is stored in the grip database 102b. In addition, the controller 320 determines whether the gripping information includes a gripping score indicating the gripping position and posture of the object P, ease of gripping, etc., a pushable amount at the time of gripping, gripping determination, based on input or received information. For example, a threshold value can be added and stored.

  The information acquisition system 300 also includes a user interface (not shown) for inputting an operation by the operator. The information acquisition system 300 can include a weight sensor and other sensors (not shown). In addition, the information acquisition system 300 can be configured so that a cover (not shown) covering the whole can be attached.

[Object database and gripping database]
FIG. 12 is a table showing object information stored in the object database 102a. Object information stored in the object database 102a includes, for example, an object ID number, object name information, item category information, multi-directional two-dimensional image information, three-dimensional image information, detailed CAD model information, simple CAD model information, bar Code information, identifier information for object determination, weight information, characteristic information at the time of gripping (for example, soft, fragile, shape change, etc.), imaging condition information, and the like can be included. Here, the discriminator information for object determination is information such as a classifier, a discriminator, a pattern matching dictionary, etc. used for discriminating objects. The shooting condition information is information indicating shooting date and time, environment, shooting equipment, and the like.

  FIG. 13 is a table showing grip information stored in the grip database 102b. The grip information stored in the grip database 102b includes, for example, an object ID number, object name information, grip shape category information, grip method information, grip position information, grip score information, search range information, push amount information, grip determination Information, transport posture information, allowable speed information, and the like can be included. In the gripping database 102b, gripping position information, gripping score information, search range information, push-in amount information, gripping determination information, transport posture information, allowable speed information, and the like are set for each gripping method. 12 and 13, the object information stored in the object database 102a and the grip information stored in the grip database 102b are associated with each other by the object ID number and the object name information.

[Obtaining Object Information by Information Acquisition System]
FIG. 14 is a flowchart illustrating a procedure for acquiring object information by the information acquisition system 300. When the cover is opened, the object P is set on the rotation stage 302 (S301), and the object name ID number or the object name information is input from the user interface or the like, the controller 320 transmits the object name to each part of the information acquisition system 300. An instruction to start execution of processing for acquiring information is output (S302).

  Next, adjustment of the position where the object P is installed is executed (S303). In S303, for example, the controller 320 adjusts the position of the object P by operating the movable mechanism based on the provisional captured images of the RGB camera 306 and the distance image camera 307 of the object P, and sets the assumed centroid position of the object P. Match the shooting center.

  Next, the controller 320 acquires the detection result of the weight of the object P by the weight sensor, and measures the weight of the object P based on the detection result (S304).

  Next, the controller 320 changes the position (angle) of the movable mechanism, for example, the rotary stage 302 or the sensor arm 304 in a stepwise manner, and the RGB camera 306 or the distance image camera so as to execute photographing at each position (angle). 307 is controlled (S305). At this time, the controller 320 obtains a distance measurement result by the laser sensor 308 in each step.

  Next, the controller 320 performs imaging similar to that in step S305, with the object P reversed, as necessary. The inversion of the object P may be performed by the operator, or may be executed by the controller 320 using the apparatus when the apparatus such as the manipulator 20 or another robot 8 can be used (S306). .

  When shooting is completed, the controller 320 automatically generates primary object information to be registered in the object database 102a (S307). Here, the primary object information is, for example, item category information, multi-directional two-dimensional image information, three-dimensional image information, detailed CAD model information, simple CAD model information, barcode information, identification for object determination of the object P. Vessel information. The automatic generation of primary object information will be described in detail later.

  Next, the controller 320 displays the automatically generated primary object information on the output unit (display or the like) of the user interface, and based on the operator's correction instruction and input data by the input unit (keyboard or the like) of the user interface. Next object information is corrected (S308).

  Next, the controller 320 automatically generates primary grip information based on the corrected primary object information (S309). Here, the primary gripping information refers to gripping shape category information, gripping method information, gripping position information, gripping score information, search range information, push amount information, gripping determination information, transport posture information, Permissible speed information. The automatic generation of primary grip information will be described in detail later.

  Next, the controller 320 displays the automatically generated primary grip information on the output unit (display or the like) of the user interface, and 1 based on the operator's correction instruction and input data by the input unit (keyboard or the like) of the user interface. The next grip information is corrected (S310).

  Next, the controller 320 updates learning-related processing such as article determination discriminator information (S311).

  Next, the controller 320 updates and registers the corrected primary object information in the object database 102a, and updates and registers the corrected primary grip information in the grip database 102b (S312).

[Automatic generation of object information in information acquisition system]
Among the primary object information, the detailed CAD model information includes, for example, the obtained multidirectional two-dimensional image information, three-dimensional image information obtained by reconstructing the multidirectional two-dimensional image information, distance image information, and the like. Can be obtained by converting three-dimensional point cloud information such as point cloud data obtained from the above into three-dimensional surface information or point cloud information such as STL or STEP.

  Furthermore, the centroid of the obtained detailed CAD model information and the reference coordinates of the centroid are defined. The reference coordinates may be based on, for example, the posture where the object P is placed on the information acquisition system 300. Specifically, for example, the reverse direction of the normal direction of the maximum surface of the circumscribed rectangle of the object P is the Z axis, the direction along the long side of the maximum surface is the Y axis, and the direction along the short side is the X axis. Rules can be applied.

  The simplified CAD model information is reduced in data size by reducing the vertex information of the obtained detailed CAD model or by applying the 3D point cloud information to a small number of basic shapes such as a cube or a cylinder. Information can be obtained.

  Further, the barcode information can be obtained by extracting a barcode area from the obtained multi-directional two-dimensional image information by image processing and analyzing the barcode information, thereby obtaining information such as the ID of the target item. In this case, the barcode information includes information indicating the position and direction of the barcode area in the object P, information indicating the position and direction of the barcode area in the detailed CAD model information or the simple CAD model information, The information is stored in the object database 102a in association with or including the information.

  In addition, item category information includes classifications that indicate the types of parts such as screws, nuts, fastening parts, and protective parts at the manufacturing site, and classifications of articles such as clothing, stationery, and electronic equipment at the distribution site of finished products. This is a classification to represent, and can be determined from image information and CAD model information using a category recognition technique.

  Further, the identification information for object determination, that is, information such as a classifier, a classifier, a pattern matching dictionary, etc., used for identifying an object is, for example, the obtained multidirectional two-dimensional image information, three-dimensional image information, object From name information, etc., using features such as SIFT, HOG, Haar-like feature quantities of 2D images, features using 3D shape data such as SHOT, deep learning, etc. Can be generated.

[Automatic generation of grip information in information acquisition system]
FIG. 15 is a schematic diagram of a gripping basic system model, and FIG. 16 is a schematic diagram of gripping method basic data. FIG. 17 is a schematic diagram illustrating a part of a procedure for automatically generating grip information by the information acquisition system 300.

  The controller 320 (database information generation unit) generates a gripping basic system model as illustrated in FIG. 15 that is a basic system of an object model in advance and a basic gripping method for the gripping basic system model in order to generate gripping information. And holding method basic system data as exemplified in FIG. The gripping basic system model and the gripping method basic system data may be described in a program, or may be registered as a database in an auxiliary storage unit or the like.

  As shown in FIG. 15, the gripping basic system model is, for example, shape data of a rectangular parallelepiped Mo1, a flat plate Mo2, a square bar Mo3, a cylindrical Mo4, a round bar Mo5, and the like. The gripping method basic data defines gripping methods capable of gripping, such as suction Me11 to Me12, sandwiching Me22 to Me24, and gripping with respect to each surface and feature points of the gripping basic system model. In the example of FIG. 16, it is defined that the rectangular parallelepiped Mo1 can have the suction Me11 by the suction pad method with respect to each surface of the rectangular parallelepiped Mo1 and the sandwiching Me21 that sandwiches the opposing surfaces. 15 and 16 are merely examples, and it is of course possible to define basic gripping model other than that shown in FIG. 15 and basic gripping method data other than that shown in FIG.

  As shown in FIG. 17, for example, the gripping category information is classified by shape (for example, rectangular parallelepiped type, flat plate type, composite type, etc.) and structural material (for example, hardness, brittleness, flexibility, porosity, etc.). Is done. The shape of the gripping category can be determined by matching the CAD model information of the object information with the gripping basic system model and applying the closest gripping basic system model. The structural material information of the gripping category can be determined by category recognition from the RGBD image information of the object information or reclassification from the item category information.

  As shown in FIG. 17, the gripping method information can be determined by referring to the gripping method basic data from the shape information of the gripping category, for example. In addition, when the direction defined by the gripping method basic data is hidden by the combination of shapes, it is excluded from the gripping method.

  Like the grip method information, the grip position information is the grip position information for gripping with the grip method defined by the grip method basic data for the basic posture of the object defined by the shape information of the grip category. This can be determined by calculating the coordinate value and the gripping posture.

  The grip score information can be calculated according to a rule defined by a combination with the shape of the grip position for each grip method. For example, in the case of suction by a suction pad, the rule increases the score if the area of the flat part to be sucked is large, and lowers the score if the flat part of the CAD model has minute irregularities. If the structural material information is a material that does not allow air to pass, such as plastic, the score is increased. If the material is porous, such as air, the score is decreased. A gripping score can be determined through an identifier using these pieces of information as learning data. The gripping score can be considered as an index of the gripping success probability.

  The search range information is information used for re-searching the target when a gripping failure determination occurs during the gripping operation. The smaller the size of the object P determined by the shape of the gripping category, the smaller the search range is set. The larger the size, the larger the search range is set.

  The indentation amount information defines the indentation amount of the hand 200 at the time of gripping determination. From the structure material information of the grasping category, the indentation amount is set to be smaller for harder materials, and the indentation amount is set to be larger for softer materials. .

  The grip determination information is a threshold value when grip determination is performed. For example, in the case of suction, when the negative pressure (pressure) at the time of suction is used for grip determination, the grip position is determined from the shape information or the structural material information. When the leakage flow rate is expected to be small, a larger negative pressure is maintained, so the absolute value of the negative pressure that is a criterion for adsorption is increased. When it is expected that there is a leakage flow rate, the pressure is unlikely to decrease, so the absolute value of the negative pressure that is a criterion for adsorption is reduced. Even in this case, the object P can be gripped if the suction flow rate is sufficiently secured. In the case of clamping, for example, when the material is hard, the reaction force to the finger 200b at the time of clamping is set high, and when it is soft, the reaction force to the finger 200b is set low. Thus, by changing the determination value of the gripping state according to the characteristics of the object P, it is possible to stably grip more types of objects P.

  The above-described processing executed by the controller 320 of the information acquisition system 300 can be executed by the database information generation / update unit 101i of the object handling apparatus 1. That is, the object handling apparatus 1 is also an example of the information acquisition system 300. In this case, the object P placed on the processing unit 50 is photographed from a plurality of directions and a plurality of distances by the camera 21 of the hand 200 (manipulator 20). In this case, the posture of the object P in the processing unit 50 may be changed using the manipulator 20. Then, processing similar to the processing described above by the controller 320 is executed by the arithmetic processing unit 101.

[Characteristics of object handling device]
The object handling apparatus 1 has the following characteristic functions (1) to (5). Hereinafter, these functions will be described in order.
(1) Auto-calibration function to minimize manual calibration (2) Grasping function that realizes flexible handling of objects based on object database and gripping database with hands with elastic joints (3) Mechanism and sensing Model-based real-time visual feedback control function that corrects the trajectory of the target sequentially in response to errors and changes in the environment, and reliably captures the object (4) Monitors the work status at all times, selects the optimal work procedure, and executes error recovery Error recovery function (5) Learning control function that adapts to the usage environment by learning parameters during work

[(1) Auto calibration]
FIG. 18A is a flowchart illustrating a procedure for auto-calibration of the object handling apparatus 1. 18B is an explanatory diagram of a plane error between the base coordinate system S1 and the manipulator coordinate system S2. FIG. 18C shows an image of the manipulator 20 provided with the calibration pattern CP photographed by the camera 21. It is a schematic diagram.

  The auto calibration is executed by the automatic calibration unit 101j of the arithmetic processing unit 101. The automatic calibration unit 101j automatically grasps the deviation between the base coordinate system S1, the manipulator coordinate system S2, and the external device coordinate system S3 shown in FIG. 1, and obtains detection results and control results with less errors. Perform calibration. As a premise, the position of the mark 51 of the processing unit 50 is known as design information in the base coordinate system S1. Further, the position of the origin of each coordinate system S1, S2, S3 and the direction of the coordinate axis are not limited to those shown in FIG.

  First, the automatic calibration unit 101 j detects a plane error between the processing unit 50 and the manipulator 20 based on a photographed image of the camera 21 provided in the hand 200. Specifically, the automatic calibration unit 101j has a position and posture where the position of the mark 51 of the processing unit 50 in the design is assumed to overlap the center point of the field of view of the camera 21 of the hand 200 or a predetermined assumed point. The manipulator 20 is controlled. In this case, it is assumed that the optical axis of the camera 21 is orthogonal to the mark 51. Then, as shown in FIG. 18B, the deviation δx, δy (plane error) between the center point or the predetermined assumed point Pt and the XY plane (direction X and direction Y) of the mark 51 from the image obtained by capturing the mark 51 with the camera 21. ). The direction X and the direction Y are examples of the first direction. When the camera 21 is a distance image camera, not only the XY plane position of the mark 51 but also the Z direction position can be acquired. By acquiring the position in the Z direction, the search range of the depth position error by the manipulator 20 in the next step (S502) can be narrowed down. It should be noted that the positional accuracy of the distance image camera information varies depending on its direction. In general, since the accuracy in the depth direction is low, it is difficult to obtain sufficient accuracy only with the information of the distance image camera. The deviations δx and δy are an example of the first error (S501).

  Next, the automatic calibration unit 101j detects a depth position error between the processing unit 50 and the manipulator 20 based on detection results of the contact sensor 200g, the force sensor 200h, the displacement sensor 200i, and the like. Specifically, the automatic calibration unit 101j first detects the deviation δ1 obtained in step S501 on the XY plane from the position of the tip 51 of the hand 200 or the probe (not shown) provided on the hand 200 in the design of the mark 51. The manipulator 20 is controlled so as to move from a position moved in the decreasing direction and distance to a position shifted by a predetermined amount in a direction Z orthogonal to the XY plane. Next, the automatic calibration unit 101j controls the manipulator 20 so as to move to a position where the tip of the hand 200 or the probe contacts the mark 51. The contact with the mark 51 can be recognized from the detection results of the contact sensor 200g, the force sensor 200h, the displacement sensor 200i, and the like. The automatic calibration unit 101j stops the manipulator 20 at the position where the tip of the hand 200 or the probe contacts the mark 51, and determines the position (depth position, position in the Z direction) in the manipulator coordinate system S2 at the stopped position. get. For the plurality of marks 51, the above-described processing is executed to acquire the Z-direction position in the manipulator coordinate system S2, and further, a deviation δz (depth error) from the Z-direction position in the design of each mark 51 is calculated. The direction Z is an example of the second direction. The deviation δz is an example of a second error (S502).

  Next, the automatic calibration unit 101j calibrates the manipulator coordinate system S2 in the base coordinate system S1. Specifically, the automatic calibration unit 101j has a first correction matrix that minimizes the deviations δx, δy, and δz of the plurality of marks 51, that is, the position of the representative part of the manipulator 20 (hand 200) in the base coordinate system S1 and A first correction matrix for calculating an operation value for obtaining an actual position and posture of a representative part that is as close as possible to the command value is obtained for the command value of the posture. Here, obtaining the first correction matrix is an example of calibration of the manipulator coordinate system S2 in the base coordinate system S1. The first correction matrix is an example of an instruction calibration value (first instruction calibration value) (S503).

  Next, the automatic calibration unit 101j calibrates the position and orientation of the camera 33 of the base 30 in the base coordinate system S1. In this process, it is important that the manipulator coordinate system S2 is already calibrated with respect to the base coordinate system S1. Specifically, the automatic calibration unit 101j first controls the manipulator 20 so that the calibration pattern CP provided on the manipulator 20 (arm 2 or hand 200) as shown in FIG. To do. Next, the automatic calibration unit 101j controls the manipulator 20 so that the calibration pattern CP moves to a plurality of predetermined positions set in the field of view, and controls the camera 33 so as to capture the calibration pattern CP at each location. To do. Next, the automatic calibration unit 101j is based on the position and orientation obtained from the imaging result of the calibration pattern CP by the camera 33, and the position and orientation that should be obtained from the imaging result of the calibration pattern CP by the camera 33 by design. Then, the origin and orientation of the coordinate system of the internal optical system of the camera 33 in the base coordinate system S1 are calculated. The coordinate system of the internal optical system of the camera is called a camera coordinate system, and the origin and orientation are generally called external parameters, and the values can be calculated by a perspective projection matrix estimation method or the like. Here, the second correction matrix for obtaining the actual position and orientation of the camera 33 in the base coordinate system S1 is obtained by the known method. Thus, by calibrating the position and orientation of the camera 33 in the base coordinate system S1, it is possible to correctly calculate the position and orientation of the object to be imaged from the results obtained by the camera 33 thereafter. The second correction matrix is an example of an image calibration value (S504).

  Next, the automatic calibration unit 101j calibrates the position and orientation of the work table 70 (outside the apparatus) in the base coordinate system S1 through the following steps S505 and S506 (S507). Also in this process, it is important that the manipulator coordinate system S2 is already calibrated with respect to the base coordinate system S1. Specifically, first, the automatic calibration unit 101 j detects a plane error between the work table 70 and the manipulator 20 based on a photographed image of the camera 21 provided on the hand 200. Specifically, the automatic calibration unit 101j controls the manipulator 20 so that the mark 71 on the workbench 70 is within the field of view of the camera 21. Next, a rough position and orientation of the work table 70 are calculated from the mark 71 and the feature point information of the peripheral portion thereof by the camera 21. Next, the automatic calibration unit 101j controls the manipulator 20 so that the position of the mark 71 in the design is a center point of the field of view of the camera 21 of the hand 200 or a position and posture assumed to overlap a predetermined assumption point. To do. In this case, it is assumed that the optical axis of the camera 21 is orthogonal to the mark 71. Then, a deviation (plane error) between the center point or a predetermined assumed point and the XY plane of the mark 71 is measured from an image obtained by photographing the mark 71 with the camera 21 (S505). In the present embodiment, the camera 21 captures the image of the mark 71, but the camera 33 may capture the image of the mark 71. When using the camera 33, it is important that the position and orientation of the camera 33 are calibrated in the base coordinate system S1.

  Next, the automatic calibration unit 101j detects a depth position error between the processing unit 50 and the manipulator 20 based on detection results of the contact sensor 200g, the force sensor 200h, the displacement sensor 200i, and the like. Specifically, the automatic calibration unit 101j first shifts the tip of the hand 200 or a probe (not shown) provided on the hand 200 by a predetermined amount in the Z direction orthogonal to the XY plane from the position of the design of the mark 71. The manipulator 20 is controlled to move to the selected position. Next, the automatic calibration unit 101j controls the manipulator 20 so as to move to a position where the tip of the hand 200 or the probe contacts the mark 71. The contact with the mark 71 can be recognized from detection results of the contact sensor 200g, the force sensor 200h, the displacement sensor 200i, and the like. The automatic calibration unit 101j stops the manipulator 20 at the position where the tip of the hand 200 or the probe contacts the mark 71, and determines the position (depth position, position in the Z direction) in the base coordinate system S1 at the stopped position. get. For the plurality of marks 71, the above-described processing is executed to acquire the Z-direction position in the base coordinate system S1, and the deviation from the Z-direction position in the design of each mark 71 is calculated (S506).

  Next, the automatic calibration unit 101j calibrates the position and orientation of the work table 70 in the base coordinate system S1. Specifically, the automatic calibration unit 101j has a third correction matrix that minimizes the shift in the planar direction and the shift in the depth direction of the plurality of marks 71, that is, the position and orientation in the base coordinate system S1 of the representative part of the workbench 70. A third correction matrix for calculating an operation value for obtaining the actual position and posture of the representative part as close as possible to the command value is obtained. Here, obtaining the third correction matrix is an example of calibration of the position and orientation of the work table 70 in the base coordinate system S1. The third correction matrix is an example of an instruction calibration value (second instruction calibration value) (S507).

  According to the present embodiment, the automatic calibration unit 101j relatively easily or more accurately determines the instruction calibration value (the first instruction calibration value, the first instruction value) from the command value, the image captured by the camera 21, and the detection result by the sensor. (Two-indicating calibration value) can be acquired, so that an error with respect to the command value can be reduced. Furthermore, the automatic calibration unit 101j can acquire an image calibration value relatively easily or more accurately from a plurality of command values and a plurality of images taken by the camera 33 corresponding to each of the command values. Therefore, the position error obtained from the image can be reduced. In addition, it is preferable that the calibration processing by the automatic calibration unit 101j is automatically executed when the object handling apparatus 1 is started, for example, at a predetermined time or time interval. Such auto-calibration is useful in the object handling apparatus 1 in which the arm 2 or the hand 200 is provided with an elastic part such as the elastic mechanism 200e or an elastic joint. In the present embodiment, when the coordinate system is calibrated using a plurality of sensors such as a camera image and a force sensor 200h, information with high accuracy in each direction is more preferentially used. Thereby, a finer coordinate system can be calibrated. In the present embodiment, the case where the error calculation direction by the camera image and the error calculation direction of the force sensor 200h are orthogonal to each other has been described. However, these directions are completely orthogonal if there is an orthogonal component. Even if not, it is effective.

[(2) Flexible handling based on database]
FIG. 19A and FIG. 19B are flowcharts showing a flexible handling procedure by the hand 200 based on the object database and the gripping database. FIGS. 20A to 20C are diagrams illustrating operation waypoints. FIG. 20A is a diagram illustrating an initial position p0, an object (part, object), and a conveyance target position pe, and FIG. 20B is an initial position p0. FIG. 20C is a diagram showing operation transit points from the pull-up position p4 to the grip release position p6. Here, as an example, a procedure when the arithmetic processing unit 101 (FIG. 2) controls each unit to take out a component as the object P from a component cage as the case B is shown.

  The arithmetic processing unit 101 controls the camera 21 and the manipulator 20 so that the camera 21 captures the periphery of the component (object P). Here, images of a plurality of parts may be taken (S601).

  Next, the object recognition unit 101h compares the image of the part captured by the camera 21 with the object information in the object database 102a, and identifies the captured part. The object recognition unit 101h acquires the object ID number of the corresponding part from the object database 102a (S602).

  Next, the arithmetic processing unit 101 compares the object ID number of the part as the target instructed by the host with the object ID number of the part specified in S602, and whether the recognized part contains the target object. It is determined whether or not (S603).

  If there is no object in the recognized part (No in S603), the camera 30 of the base 30 or the camera of the hand 200 is changed by changing some shooting conditions such as the lighting conditions and the shooting posture of the camera 21 of the hand 200. The re-photographing is performed at 21 (S604), and the process returns to S602.

  If there is an object in the recognized part (Yes in S603), the object recognition unit 101h performs matching between the captured three-dimensional information and CAD model information in the object database 102a, and calculates the posture of the recognized article. To do. At this time, the calculation of the attitude of the parts cage, the attitude of the parts other than the object, and the like may be performed (S605).

  Next, the gripping / motion plan generating unit 101d refers to gripping information stored in the gripping database 102b and calculates a gripping method and gripping point candidates for the recognized component. Here, the gripping database 102b includes a plurality of gripping methods, a plurality of gripping points, and gripping scores for the parts, but the gripping / motion plan generating unit 101d does not recognize the recognized parts. Based on the position and orientation information and the position and orientation information of the component cage, recalculation of the grip score, selection of a grip point, and the like are executed. Specifically, among the gripping points arranged on each surface of the recognized component, the score of the gripping point facing the opening of the component cage is increased (S606).

  Next, the gripping / motion plan generating unit 101d calculates the motion via points shown in FIGS. 20A to 20C from the initial position p0 of the hand 200 to the conveyance target position pe. For example, the operation via point includes an insertion preparation position p1 for inserting the hand 200 into the component cage (case B), a position immediately before gripping p2 immediately before gripping the component (object P) with the hand 200, and a gripping operation of the component. A grip target position p3, a pull-up position p4 for lifting the gripped part from the part cage, a work table entry position p5 for inserting the hand 200 into the space of the work table 70, and an object at a predetermined position on the work table 70. For example, the grip release position p6 for placement. The insertion preparation position p1 is calculated from the recognition result of the position and orientation of the component cage in consideration of the position of the hand 200 where it is easy to confirm the component cage with the camera 21 of the hand 200 and the position where it is easy to move to the grip target position p3. The position immediately before gripping p2 is calculated in consideration of being able to reach the gripping target position p3 by a substantially linear motion, such as a position slightly away from the gripping target position p3 in the normal direction. The work table approach position p5 is calculated from the recognition result of the position and orientation of the work table 70 in consideration of the position of the hand 200 where the camera 70 of the hand 200 can easily confirm the work table 70 and the position where the work table 70 can easily move to the grip release position p6. To do. The grip release position p6 is calculated so that the component is placed in a predetermined position when the grip of the component is released (S607).

  Next, the hand trajectory generation unit 121 uses the information indicating the operation environment point, the shape of the object handling apparatus 1, the state of the external environment such as the part cage and the part, and the hand 200 performs the desired operation while avoiding interference. A hand trajectory that passes through the waypoint is generated. Specifically, it can be calculated using a technique such as RRT or PRM (S608).

  Here, when an effective trajectory cannot be calculated (No in S609), the arithmetic processing unit 101 checks whether or not there is a next gripping point candidate, and when there is no next gripping point candidate ( In S610, the operation shifts to a state change operation F22 (FIGS. 22A and 22B, which will be described in detail later). If there is a next gripping point candidate (Yes in S610), the arithmetic processing unit 101 sets the next gripping point candidate and executes the process of S607.

  When an effective trajectory can be calculated in S608 (Yes in S609), the manipulator control unit 110 moves the hand 200 from the initial position to the insertion preparation position p1. Here, the movement may be performed along a predetermined trajectory, or the trajectory is corrected by a visual feedback servo to be described later based on the position and orientation of the component cage by the camera 21 of the hand camera or the hand 200. However, it may be moved (S612).

  Next, the manipulator control unit 110 moves the hand 200 from the insertion preparation position p1 to the position immediately before gripping p2. At this time, based on the feature point information of the parts by the camera 21 of the hand 200, the camera 33 of the base 30, etc., it may be moved while correcting the trajectory by the visual feedback servo described later (S613).

  Next, when the position and orientation errors are calculated with respect to the grip target position p3, the arithmetic processing unit 101 calculates from the observation information of the hand 200 and the parts by the camera 21 of the hand 200, the camera 33 of the base 30, and the like. The grip target position p3 and the posture at that time are corrected (S614).

  Next, the manipulator control unit 110 moves the hand 200 from the position just before gripping p2 to the gripping target position p3 (S615).

  Next, the hand control unit 140 starts a gripping operation. Specifically, when gripping by the suction method, an electromagnetic valve or the like is scanned to generate a suction force. In the case of holding with two fingers or a multi-finger hand, an operation of closing the finger is started (S616).

  Next, the signal processing unit 101b performs processing of various sensor information in the gripping motion state, and the plan generation unit 101c monitors the gripping motion state information. Specifically, the gripping operation state information includes the vacuum pressure and flow rate during suction, the sensor value of the contact sensor, the displacement sensor value of the elastic joint, the torque of the gripping mechanism drive unit, the gripping operation processing time, and the like. The confirmation of the gripping operation state uses gripping determination information stored in the gripping database 102b. When the gripping operation is continued beyond the gripping target position p3, the pushing amount information of the gripping database 102b is used.

  Next, the state monitoring unit 101e determines whether or not a gripping confirmation state has been established based on the gripping operation state information. Specifically, the grip confirmation state is a state in which the vacuum pressure has reached a predetermined value during adsorption, a state in which contact has been confirmed by the contact sensor, and the displacement sensor and torque value have reached the desired values. It is in the state. The determination threshold here uses the grip determination information registered in the grip database 102b (S618).

  If the grip confirmation state is detected (Yes in S618), the state monitoring unit 101e determines the result of the grip state. Specifically, the determination can be made based on the vacuum pressure at the time of adsorption and the interval between two fingers if it is two fingers (S619).

  If it is determined in the determination of the gripping state result that the component has not been gripped (No in S620), the process proceeds to S621. In S621, it is confirmed whether or not the gripping point vicinity searching process has been performed at the current gripping point. If the search has been performed (Yes in S621), the next gripping point candidate is set (S622). If it has not been performed (No in S621), a grip search operation is performed.

  In the case of Yes in S620, the manipulator control unit 110 moves the hand 200 in the pulling direction (S623).

  Next, the arithmetic processing unit 101 confirms the gripping state again, and if the gripping state is unstable or cannot be gripped, that is, if it cannot be transported (No in S624), the retry operation F24 (FIG. 24). , Later described in detail). Note that the threshold value for confirming the gripping state differs depending on the object, the gripping method, and the like, and therefore grip determination information in the gripping database 102b is used.

  In the determination in S624, when the conveyance is possible (Yes in S624), the manipulator control unit 110 converts the posture of the hand 200 into the conveyance posture. At this time, since the posture suitable for conveyance and the gripping state differ depending on the parts, the posture is determined with reference to the conveyance posture information in the gripping database 102b. Specifically, in the case of suction, the posture specified in the gripping database 102b is a posture in which the suction pad 200a is difficult to apply a moment to the component and the suction pad 200a. In the case of pinching, the posture is such that a part of multiple fingers supports the lower part of the part. In any case, in order to suppress the vibration at the time of conveyance, it is possible to fix the elastic mechanism 200e by controlling the clutch 200f and make it in an inelastic state (S625).

  Next, the manipulator control unit 110 moves to the work table entry position p5 while maintaining the transport posture. Here, there is a risk of dropping the gripped component during movement during gripping. Therefore, using the allowable speed information in the gripping database 102b, the transport speed is kept below the allowable speed, thereby shortening the transport time to the maximum. However, stable conveyance can be realized. The conveyance speed here can be changed according to the gripping state. This can be realized by setting a plurality of levels in the grip determination information of the grip database 102b and setting a plurality of conveyance permissible speeds corresponding to the levels. For example, in the case of the adsorption method, the adsorption pressure can be used as the grip determination information. As an example, when a suction pressure is −40 kPa or less with respect to a certain part, since the suction grip is sufficiently performed, the conveyance allowable speed is set to 500 mm / sec. When the adsorption pressure is −40 kPa to −30 kPa, the conveyance speed is set to 200 mm / sec. When the adsorption pressure is −30 kPa to −15 kPa, the conveyance allowable speed is set to 200 mm / sec (S626). Note that the allowable value of the conveyance state does not need to be set according to only the speed, but may be set according to acceleration or jerk. On the other hand, if the suction pressure is greater than −15 kPa, a sufficient suction grip state has not been obtained, so it is determined that the conveyance is impossible in S624 (No in S624), and the process proceeds to a retry operation.

  The arithmetic processing unit 101 controls the camera 21 of the hand 200 to photograph the vicinity of the conveyance destination position of the work table 70 (S627). Next, the arithmetic processing unit 101 corrects the grip release position p6 of the component based on the instruction from the host and the photographing result (S628). Next, the manipulator control unit 110 moves the hand 200 to the grip release position p6, and the hand control unit 140 releases the grip of the component and places the component at the transport target position pe instructed by the host (S629). As a result, a series of processing ends.

[(3) Real-time visual feedback control]
FIG. 21A is a block diagram of the manipulator control unit 111 (110). FIG. 21B is an explanatory diagram of an image feature amount target value. The manipulator control unit 111 is characterized by having either or both of a feedback system for correcting the control amount of the motor 111e using the nonlinear dynamic model and a feature amount-based controller using the image feature amount. is there.

  In the manipulator control unit 111, when position information as a movement target of the hand 200 such as a grip position of a part (object P) is given from the plan generation unit 101c, the image feature amount generator 111a causes the object database 102a or the like to Using the information, an image feature amount target value for achieving movement to the grip target position p3 is generated. Specifically, as shown in FIG. 21B, the image feature amount target value is an arrangement of feature points of a component image obtained from a camera sensor such as the camera 21 of the hand 200 when the hand 200 reaches the grip target position p3. Information etc. In FIG. 21B, a feature amount image is shown for the sake of understanding, but in practice, numerical information such as the coordinates of feature points and the distance between feature points may be used. This is generated by extracting necessary surface information from the image and feature amount information of each surface of the part stored in the object database 102a, generating from CAD model information, or performing viewpoint conversion of the extracted information. be able to.

  The feature amount base controller 111b calculates the control amount of the motor 111e by using an error between the image feature amount target value and the image feature amount obtained from the camera sensor such as the camera 21 of the hand 200 as an input. The feature amount base controller 111b can be configured using an image Jacobian and using a general PI control system.

  The calculated control amount of the motor 111e is compensated for by the nonlinear dynamic model compensator 111d of the object handling apparatus 1. The details of the control amount compensation method using the nonlinear dynamic model will be described later. Here, in the nonlinear dynamic model, not only the object handling apparatus 1 but also information on the gripping state such as the gripping part model and the gripping position of the part such as weight is captured.

  The motor 111e is driven according to the compensated control amount, and performs tasks such as moving the arm 2 and moving the hand 200 to a predetermined position. The image feature amount calculator 111f calculates an image feature amount for a target using an output image from a camera sensor such as the camera 21 of the hand 200.

  In the manipulator control unit 111 in FIG. 21A, the manipulator 20 operates so that the image observed by the camera 21 of the hand 200 matches the grip target position p3. Even if there is an error, the error is finally suppressed, and the hand 200 can be moved to an accurate position. Further, even if the component is moving, the hand 200 can be followed in accordance with the movement of the component. Further, the manipulator control unit 111 includes a nonlinear dynamic model compensator 111d. This compensates for the deflection and vibration of the arm 2 and enables the hand 200 to follow more accurately. In the control system using the camera image, when the arm 2 vibrates, the image information greatly fluctuates, so it is difficult to realize stable control. However, in this system, vibration suppression by the nonlinear dynamic model compensator 111d is difficult. It is possible to realize stable control.

  Furthermore, in the manipulator control unit 111 of FIG. 21A, the nonlinear dynamic model compensator 111d can be switched by the nonlinear dynamic model parameter management unit 111g depending on the gripping state of the part (object P). Normally, in a model-based control system using a non-linear dynamic model, when the load of the arm 2 changes due to the hand 200 gripping or releasing an object, the difference between the model and the reality increases. Becomes unstable. However, the manipulator control unit 111 can realize stable control in any process of the gripping and transporting work by switching the nonlinear dynamic model from the recognized object information and gripping state information. Here, the feature point of the component is used as information indicating the grip target position p3. However, the feature point of the work table 70, the mark 51 of the processing unit 50, or the component cage (case B) can also be used. is there.

  FIG. 21C is a block diagram illustrating another example of the manipulator control unit 111A (110). When a target image or feature point cannot be obtained, a PI controller 111h of a normal PI control system as shown in FIG. 21C is used instead of the feature amount base controller 111b in FIG. 21A. Even a control system such as the manipulator control unit 111A can perform a relatively stable operation by sufficient calibration, and more accurate control can be performed by the nonlinear dynamic model compensator 111d.

  FIG. 21D is a flowchart illustrating a procedure for changing parameters in the manipulator control unit 111. Here, a procedure for gripping and transporting a component as an object P from the component cage to the work table 70 is shown.

  The arithmetic processing unit 101 controls the camera 33 of the base 30 so as to photograph the periphery of the component (S701).

  Next, the object recognition unit 101h specifies the object ID information of the photographed part using the image photographed by the camera 33 of the base 30 and the object information in the object database 102a (S702).

  Next, the object recognition unit 101h performs matching between the captured three-dimensional information and the CAD model information in the object database 102a, and calculates the posture of the recognized component. At this time, the calculation of the attitude of the part cage and the attitude of an object other than the part is also executed (S703).

  Next, the gripping / motion plan generation unit 101d refers to the gripping database 102b and calculates a gripping method and a gripping position for the recognized component (S704).

  Next, the manipulator control unit 111 (110) performs a state conversion operation of the hand 200 in accordance with the calculated gripping method. Specifically, when the suction method is calculated, the suction pad 200a is protruded, and when the pinching method using the finger 200b is calculated, the suction pad 200a is stored (S705). .

  Next, the nonlinear dynamic model parameter management unit 111g sets the model parameters of the dynamic model used in the nonlinear dynamic model compensator 111d in accordance with the state of the hand 200 set in S705. As model parameters to be set, those determined by prior learning are used. Further, when the prior learning is not performed, parameter learning may be performed by applying a random signal or the like to the control system at this timing (S706).

  Next, the manipulator control unit 111 controls the manipulator 20 so that the hand 200 moves to the insertion preparation position p1 using the set model parameter. Here, since an optimal model is set, stable operation is possible even if the manipulator 20 is moved at the highest speed. Furthermore, by performing the feature-based control utilizing the feature points of the component cage, the component cage can be more accurately moved to the insertion preparation position p1 (S707).

  Next, the manipulator control unit 111 controls the manipulator 20 so that the hand 200 moves from the insertion preparation position p1 to the position just before gripping p2. By performing the feature amount-based control utilizing the feature point information of the part, it is possible to move to the position p2 immediately before gripping more accurately (S708).

  Next, the manipulator control unit 111 controls the manipulator 20 so that the hand 200 moves from the position immediately before gripping p2 to the gripping target position p3 at a medium speed. At this time, even if the hand 200 and the target object P or the surrounding object P come into contact with each other due to the influence of misrecognition of parts, the hand 200 has an elastic part, so the elastic part absorbs the impact and There is no hurt. Further, while the elastic part absorbs an impact or a pushing force, the displacement sensor 200i or the contact sensor 200g determines the contact and stops the operation of the manipulator 20 or moves it in a direction to release the force, thereby moving the object P, The destruction of the hand 200 can be avoided (S709).

  Next, the hand control unit 140 controls the hand 200 to hold the component. Here, since the contact state can be accurately detected, the component can be stably held (S710).

  Next, the nonlinear dynamic model parameter management unit 111g sets a dynamic model in accordance with the gripping method, the gripping position, and the component characteristics. The gripping method, the gripping position, and the characteristics of the object to be gripped can be calculated from the object database 102a and the gripping database 102b, and the accurate gripping position may be corrected from the sensor information that actually performs the gripping operation. . The nonlinear dynamic model parameter management unit 111g has model parameters learned in advance under conditions such as a plurality of tip load weights and gravity center offsets. The designated gripping method, gripping position, and object to be gripped The optimum parameter is determined by selecting the optimum model parameter from the above characteristics or by interpolation using a plurality of model parameters having relatively close conditions (S711).

  Next, the manipulator control unit 111 controls the manipulator 20 to pull out the gripped component from the component cage (S712).

  Next, the manipulator control unit 111 refers to the gripping database 102b and controls the manipulator 20 so as to transport the component at the transport allowable speed based on the transport allowable value information. At this time, since the more accurate feature point amount can be obtained by using the mark 71 provided on the work table 70 of the transport destination as the feature point, stable feature amount-based control can be performed (S713). ).

  Next, the manipulator control unit 111 places the component at the conveyance target position pe of the work table 70 (S714).

  Next, the hand control unit 140 converts the state of the hand 200 so as to be in the initial state (S715).

  Next, the nonlinear dynamic model parameter management unit 111g sets model parameters in accordance with the state of the hand 200 (S716).

  Next, the manipulator control unit 111 controls the manipulator 20 so that the hand 200 moves to the initial position p0 at the maximum speed (S717).

[(4) Error recovery]
FIG. 22A is a flowchart showing the procedure of the state change operation F22 of FIG. 19A. Regarding the state change operation F22 of FIG. 19A, the arithmetic processing unit 101 first controls the manipulator 20 and the transport mechanism 40 so that the position and posture of the component cage (case B) change (S801). Next, the process returns to S601 in FIG. 19A (S802). By such a state change recovery, the positional relationship between the object handling device 1 and the component (object P) to be gripped changes, and therefore the probability that a path to grip the component can be set, that is, the probability that the component can be gripped. Easy to increase.

  FIG. 22B is a flowchart showing a procedure different from FIG. 22A of the state change operation F22 of FIG. 19A. With regard to the state change operation F22 in FIG. 19A, in this example, the arithmetic processing unit 101 first determines the periphery of the component, for example, the center on the image of the component or the component cage or the position away from the assumed center of gravity calculated from the image. The contactable position where the hand 200 can contact is calculated. This can be obtained by calculating the outer position of the part (object P) from the information of the distance image camera such as a three-dimensional point cloud and calculating the route to the position (S901). Next, the manipulator control unit 110 controls the manipulator 20 so that a predetermined force is applied to a position where the tip of the hand 200 can be contacted. At this time, the hand 200 has an elastic mechanism 200e and a force sensor h, and the manipulator control unit 110 can control the magnitude of the force to be applied, acceleration, and the like. A predetermined force can be applied (S902). Next, the manipulator control unit 110 returns the hand 200 to the initial position p0 (S903). And it returns to S601 of FIG. 19A (S904). In this procedure, the arrangement state of the article is changed by applying a force to the part or its periphery in S902. The recovery of such a state change also changes the positional relationship between the object handling device 1 and the article to be gripped, and thus increases the possibility of creating a route to the article gripping position.

  FIG. 23 is a flowchart showing the procedure of the grip search operation F23 of FIG. 19B. After S621 in FIG. 19B, the manipulator control unit 110 controls the manipulator 20 so that the hand 200 temporarily moves to a position away from the grip target position p3 by a predetermined distance (S1001). The arithmetic processing unit 101 captures the periphery of the grip target position p3 (S1002). If the arithmetic processing unit 101 determines from the photographing result that an image of the part (object P) to be grasped is included and that effective photographing satisfying a predetermined condition has been performed (Yes in S1003), The grip target position p3 is calculated (S1005). On the other hand, when it is not determined in S1003 that the effective photographing satisfying the predetermined condition has been made (No in S1003), the arithmetic processing unit 101 uses the search range information included in the gripping database 102b. Then, a position within the range specified by the search range information and shifted by a predetermined distance from the already calculated grip target position p3 is set as a new grip target position p3 (S1004). After S1004 or S1005, the process returns to S607 in FIG. 19A. In S607, an operation via point is calculated for the new gripping target position p3 set in S1004 or S1005 (S1006). In the search range information, an efficient search range is designated according to the part (object P). Specifically, for example, as the part (object P) is smaller, the search range is set narrower, and the offset distance (step distance) from the grip target position p3 of the search range is also set smaller, in other words, the part is larger. As the search range is set wider, the offset distance (step distance) from the grip target position p3 of the search range is also set larger. Such grip search recovery tends to increase the probability that a route until a processing component is gripped, that is, the probability that a component can be gripped.

  FIG. 24 is a flowchart showing the procedure of the retry operation F24 of FIG. 19B. If it is determined in S624 in FIG. 19B that the conveyance is impossible, the manipulator control unit 110 moves the hand 200 to the component cage (case B) by a predetermined distance, and grips the component (object P) on the component cage. The manipulator 20 is controlled so as to release. As a result, the component can be returned to the component cage (S1101). And it returns to S601 of FIG. 19A (S1102).

[(5) Parameter learning control during work]
FIG. 25A is a flowchart illustrating a procedure of executing a learning operation and database generation. The object handling apparatus 1 utilizes learning data created in advance, such as a database such as the object database 102a and the gripping database 102b, and an object recognition unit 101h that uses a learning result. However, when handling various objects, an object with insufficient or no learning data may be handled. In such a case, the procedure shown in FIG. 25A is effective.

  First, the arithmetic processing unit 101 receives a grip / convey command for a specific object from a host system (S1201). Next, the learning control unit 101g (FIG. 2) acquires object information related to the object from the object database 102a based on the object ID number of the object, and acquires grip information related to the object from the grip database 102b (S1202). ).

  Next, the learning control unit 101g determines whether the object information or the grip information about the object is sufficient (S1203). If it is determined that the information is sufficient (Yes in S1203), the manipulator control is performed. The unit 110 performs gripping and conveyance of the object in the normal mode as shown in FIGS. 19A and 19B (S1205). On the other hand, if it is determined in S1203 that the information is insufficient (No in S1203), the arithmetic processing unit 101 controls each unit to execute gripping and conveyance in the learning / database generation mode. The gripping and conveyance in the learning / database generation mode will be described in detail later (S1204).

  When the gripping and transporting process is completed (S1206), the learning control unit 101g learns both when the gripping and transporting is executed in the learning / database generation mode and when the gripping and transporting is executed in the normal mode. Data and object database 102a and gripping database 102b are updated. For example, even when gripping and transporting are executed in the normal mode, the update process here uses the results such as the success or failure of gripping, the success or failure of waypoint generation as reward information, and the selected gripping position or For example, the score information associated with the gripping method is updated (S1207).

  FIG. 25B is a flowchart illustrating a procedure of gripping and transporting in the learning / database generation mode. First, the image processing unit 101a acquires an object ID number and item category information from an image obtained by capturing a part cage with the camera 33 of the base 30, and the object recognition unit 101h stores the object ID number and item category information in the object database 102a. It is checked whether or not a matching item is included (S1301). If there is a match in the object database 102a, the object is subject to gripping and transport processing. Next, the arithmetic processing unit 101 performs matching between the 3D image information around the object stored in the object database 102a and the gripping basic system model, and selects the gripping basic system model having the highest matching rate ( S1302). Then, the gripping / motion plan generating unit 101d refers to the gripping method basic data from the gripping basic system model, and calculates the gripping method and the gripping position in the base coordinate system S1 (S1303). Next, the manipulator control unit 110 controls the manipulator 20 to execute gripping and conveyance at a lower speed than in the normal gripping mode (S1304).

  The learning control unit 101g performs detailed learning when a processing in the learning / database creation mode is instructed from the host device or when there is sufficient time until the processing completion time designated by the host device. The process proceeds to S1307 (Yes in S1305).

  In the case of No in S1305, gripping and transporting to a predetermined position are executed while maintaining a low speed state (S1306).

  The detailed learning mode is performed after S1307. First, in S1307, the manipulator control unit 110 conveys the object P to the processing unit 50 at a lower speed than usual and places it on the processing unit 50 (S1307). Next, the arithmetic processing unit 101 acquires object information for the object P placed on the processing unit 50 by the same procedure as shown in FIG. 14 (S1308). Next, the arithmetic processing unit 101 acquires an RGB image and a distance image of the object P with the camera 21 provided in the hand 200 (S1309). Next, the manipulator control unit 110 controls the manipulator 20 so that the posture of the object is reversed (S1310). Next, the arithmetic processing unit 101 controls each unit so that the camera 21 provided in the hand 200 captures the object P from multiple directions and obtains an RGB image and a distance image (S1311). Next, the database information generation / updating unit 101i generates object information such as 3D image information, detailed CAD model information, and simple CAD model information from the captured image information (S1312). In addition, the database information generation / update unit 101i generates grip information from the simple CAD model information and the captured multidirectional image (S1313).

  Next, the arithmetic processing unit 101 performs a grip test. Specifically, for example, a plurality of gripping methods are tried, the gripping success or failure is confirmed, and the data of the failed gripping method is deleted. Further, the grip determination information is generated by using the vacuum pressure at the time of successful gripping (S1314).

  Next, the database information generation update unit 101i updates the gripping database 102b by writing the gripping information generated in S1314 to the gripping database 102b (S1315).

  Next, the manipulator control unit 110 controls the hand 200 to grip the object P based on the updated gripping information in the gripping database 102b, and the learning control unit 101g holds the object P in a state where the hand 200 grips the object P. Thus, a process of supplying a test signal for nonlinear dynamic model learning to the manipulator control unit 110 is executed. Then, based on the execution result, the object database 102a and the gripping database 102b are updated (S1316). Then, the manipulator control unit 110 controls the manipulator 20 so as to grip and convey the object P to a predetermined position (S1317).

  FIG. 26 is a schematic configuration diagram of a learning system 400 that shares learning data among a plurality of object handling apparatuses 1. The learning operation described above learns not only the data for object recognition but also the data for gripping and the data for conveyance, so there are many combination conditions and more learning data is required. In this regard, as shown in FIG. 26, by sharing learning data among a plurality of object handling apparatuses 1, higher speed learning is possible. Further, in this learning system 400, learning data for operation of the manipulator 20 (robot) is also generated, so that not only when a new object becomes a target of gripping / conveying but also when a new robot is introduced. In addition, the learning data can be utilized to realize quicker and more stable gripping and conveyance.

  FIG. 27 is a block diagram showing a system different from the object handling system of FIG. In the system of FIG. 27, the object handling apparatus 1A is fixedly installed without the moving mechanism 6, and the work table 70A is provided with the moving mechanism 6A. Even in such a system, the object handling apparatus 1A can execute the same auto calibration as described above, and can also execute other processes in the same manner.

[Learning of nonlinear dynamic model and vibration suppression]
FIG. 28 is a cross-sectional view of the manipulator 20A. The manipulator 20A includes a gantry 501, a first link 503, a first motor 504, a first speed reducer 505, a first encoder 506, a second link 508, a second motor 509, a second speed reducer 510, and a second encoder 511. Have. One end of the first link 503 is attached to the top of the gantry 501, and the second link 508 is attached to the other end of the first link 503. A load 512 is added to the tip of the second link 508. Here, the load 512 is a combination of the hand 200 of the object handling apparatus 1 or the hand 200 and the object being grasped.

  The controller 100A horizontally turns the first link 503 with respect to the gantry 501 around the first shaft 502 by the combination of the first motor 504 and the first encoder 506 and the first speed reducer 505 having a spring characteristic. Further, the controller 100A moves the second link 508 with respect to the first link 503 around the second shaft 507 by combining the second motor 509 and the second encoder 511 with the second speed reducer 510 having spring characteristics. Turn horizontally.

FIG. 29 is a schematic diagram showing one link of this manipulator. This is called a two-inertia system. With reference to FIG. 28 and FIG. 29, this manipulator can be modeled as a nonlinear dynamics as a serial two-link arm having an elastic joint (elastic mechanism, elastic part). In FIG. 29, the first axis is shown as a representative of the physical parameters, moment of inertia, friction coefficient, spring coefficient, and the like necessary for describing the nonlinear dynamic model for one link. In FIG. 29, the link 530 is driven and controlled by the motor 520 via the speed reducer 525. As physical parameters regarding the first axis, the torque input of the motor 520 is represented by u ₁ , the moment of inertia of the motor 520 is represented by m _M1 , and the rotation angle of the motor 520, that is, the output detected by the encoder is θ _M1 . Where the viscous friction coefficient of the motor 520 is represented by d _M1 , the Coulomb dynamic friction coefficient of the motor 520 is represented by f _M1 , the damping coefficient of the speed reducer 525 is represented by d _G1 , and the spring coefficient of the speed reducer 525 is k It is represented by _G1 , the moment of inertia of the link 530 is represented by m _L1 , the viscous friction coefficient of the link 530 is represented by d _L1 , and the rotation angle of the link 530 is represented by θ _L1 .

In the nonlinear dynamic model of a serial two-link arm having an elastic joint, the following motor side is given by equation (1) and the link side is given by equation (2).
However,
θ _M = [θ _M1 , θ _M2 ] ^T : Motor rotation speed (suffixes 1 and 2 are axis numbers)
_{θ L = [θ L1, θ} L2] T: Rotation speed _{α = [α 1, α 2} ] Link ^T: Link translational acceleration ^{_{M L (θ L) ∈R 2}} × 2: link inertia matrix c _L (d [theta] _L / dt, θ _L ) R ^{2 × 1} : Centrifugal force and Coriolis force vector M _M = diag (m _M1 , m _M2 ): Motor and reduction gear high-speed inertia D _M = diag (d _M1 , d _M2 ): Motor shaft viscous friction coefficient D _L = diag (d _L1 , d _L2 ): Link shaft viscous friction coefficient K _G = diag (k _G1 , k _G2 ): Reducer spring coefficient D _G = diag (d _G1 , d _G2 ): Reduction gear damping coefficient N _G = diag (n _G1 , n _G2 ): reduction ratio (n _G1 , n _G2 ≦ 1)
f _M = [f _M1 , f _M2 ] ^T : Coulomb dynamic friction torque of motor shaft E = diag (e ₁ , e ₂ ): Torque / voltage (current command value) constant u = [u ₁ , u ₂ ] ^T : Input Voltage (command value to the motor current control system)
It is. Here, sgn (a) in equation (1) represents a sign function, and sgn (a) takes values of 1, -1, and 0, respectively, depending on whether a is positive, negative, or zero.
Further, diag (a, b) represents a diagonal matrix having a and b as diagonal elements.

Here, if α, β, and γ are basis parameters composed of the link length, the center of gravity position, the mass, and the inertia, the inertia matrix of the link is expressed as Equation (3).
The details of the basis parameters are expressed as the following formula (4).

α = m ₁ l _g1 ² + I _z1 + m ₂ l ₁ ²
β = m ₂ l _g2 ² + l _z2 (4)
γ = m ₂ l ₁ l _{g 1}
However,
l _i : length of each link (i is an axis number)
m _i : Mass of each link l _gi : _Center of gravity of each link (links are symmetrical in the longitudinal direction)
l _zi : Moment of inertia around the center of gravity of each link. When the tip load 512 changes, the mass m _{2 of the} link to which the tip load 512 is added changes, and the base parameters α, β, and γ all change.
About a centrifugal force and a Coriolis force vector, it becomes like a formula (5).

The motor angular velocity control system has a basic configuration based on Formula (6) in which PI (proportional and integral) control is FF-IP (feedforward-integral-proportional) control and two-degree-of-freedom PI control is performed.
However,
dθ _MRi / dt: Motor angular velocity target value (i is the axis number)
dθ _Mi / dt: motor angular velocity k _FVi : motor angular velocity target value feedforward control gain k _IVi : motor angular velocity deviation integral feedback control gain k _PVi : motor angular velocity proportional feedback control gain u _i : input voltage (motor current) Command value to control system (torque input)
It is.

  FIG. 30 is a block diagram of a general speed control system. The two-degree-of-freedom PI speed control system is configured as a control system cascade-connected inside the position control system (P control) as shown in FIG. FIG. 30 shows a two-degree-of-freedom PI speed control system related to the first axis.

  In the following, focusing on the angular velocity control system of the motor excluding the position control system, it is assumed that the control cycle is sufficiently short, and the explanation will be made assuming that the angular control system of the motor is a continuous system.

  First, the load 512 at the tip of the two link arm shown in FIG. 28 is set to 5 kg, for example, and the physical parameters appearing in the equation (1) (hereinafter, including the base parameters α, β, and γ) are used using a known identification method. To estimate. As this identification method, for example, an identification method described in Patent Document 3 considering the joint stiffness of the robot model is used.

  Next, these estimated physical parameters are substituted into equation (1) representing the dynamic model and equation (6) representing the speed feedback control law together with the motor speed control gain. As a result, a time response simulator of the speed control system of the serial 2-link arm is constructed, and it is confirmed whether or not it matches well with the time response of the actual robot.

  If the physical parameters are accurately estimated by the above identification method, these time responses should match well. If there is a slight deviation, the physical parameters are fine-tuned using optimization calculation by a non-linear least-squares method (such as Levenberg-Marquardt method) using the time response simulator. FIG. 31 is a diagram showing the effect of fine tuning of physical parameters. For example, the angular velocity step response before optimization is shown on the upper side of FIG. In FIG. 31, the horizontal axis indicates time, the vertical axis indicates angular velocity, the solid line indicates the response of the actual robot, and the broken line indicates the response result by the simulator. As can be seen from the lower graph after optimization, by optimizing the physical parameters, the angular velocity step response of the simulator and the actual robot can be matched well.

  Therefore, in this embodiment, an observer is configured to estimate the angular velocity of the link from the angular velocity of the motor measured by the encoder, and not only the measured value of the angular velocity of the motor but also the estimated value of the angular velocity of the link is fed back. Operates accurately while suppressing vibration at the tip. In the following embodiment, focusing on the fact that the actual robot whose speed is controlled and the time response of the simulation match well, an observer is constructed that takes advantage of the good accuracy of estimation of physical parameters. Hereinafter, the observer according to each embodiment will be described.

(Manipulator control unit with nonlinear dynamic model (1))
FIG. 32 is a block diagram of the observer 600 of the manipulator control unit 112. The observer 600 of this embodiment includes first and second axis PI controllers 601, 602, a two-link arm nonlinear dynamic model 603, and integrators 604a, 604b that integrate the outputs of the nonlinear dynamic model 603, 604c, 604d, 605a, 605b, 605c, 605d. The PI controller 601 performs PI control based on the deviation between the speed dθ _M1 / dt of the motor driving the first axis and the estimated value of the rotation speed of the motor driving the first axis. The PI controller 602 performs PI control based on the deviation between the speed dθ _M2 / dt of the motor driving the second axis and the estimated value of the rotation speed of the motor driving the second axis. The two-link arm nonlinear dynamic model 603 includes a first input τ _{1 that} is the sum of the output of the PI controller 601 and the operation amount u ₁ of the first axis, and the output of the PI controller 602 and the operation amount of the second axis. Based on the second input τ ₂ that is the sum of u ₂ , the angular acceleration of the first and second links is estimated based on the nonlinear dynamic model 603 and the angles of the motors that drive the first and second links, respectively. Acceleration is estimated and the estimated angular acceleration is output.

  The integrator 604a integrates the estimated value of the angular acceleration of the first link output from the nonlinear dynamic model 603, and outputs the estimated value of the angular velocity. The integrator 604b integrates the output of the integrator 604a and outputs an estimated value of the rotation angle of the first link. The integrator 604c integrates the estimated value of the angular acceleration of the motor driving the first link output from the nonlinear dynamic model 603, and outputs the estimated value of the angular velocity of the motor. The integrator 604d integrates the output of the integrator 604c, and outputs an estimated value of the rotation angle of the motor that drives the first link.

  The integrator 605a integrates the estimated value of the angular acceleration of the second link output from the nonlinear dynamic model 603, and outputs the estimated value of the angular velocity. The integrator 605b integrates the output of the integrator 605a and outputs an estimated value of the rotation angle of the second link. The integrator 605c integrates the estimated value of the angular acceleration of the motor driving the second link output from the nonlinear dynamic model 603, and outputs the estimated value of the angular velocity of the motor. The integrator 605d integrates the output of the integrator 605c, and outputs an estimated value of the rotation angle of the motor that drives the second link.

  The estimated values of the angular acceleration of the first and second links, the estimated values of the angular velocities of the first and second links, the estimated values of the angular velocities of the motors driving the first and second links, and the estimated rotation angles of the motors. Each value is output from the observer 600.

Further, in this observer 600, the estimated value of the angular velocity of the first link, the reduction ratio n _G1 of the reduction gear provided between the first link and the motor driving the first link, and the motor driving the first link Based on the estimated value of the angular velocity, the estimated value of the torsional angular velocity of the first link is calculated, and this estimated value is output. Similarly, the estimated value of the angular speed of the second link, the reduction ratio n _G2 of the speed reducer provided between the second link and the motor driving the second link, and the estimated angular speed of the motor driving the second link Based on the value, an estimated value of the torsional angular velocity of the second link is calculated, and this estimated value is output.

  The observer 600 of this embodiment incorporates a whole nonlinear dynamic model 603 of the manipulator, and the observer gain of the PI controllers 601 and 602 uses the PI control gain of the existing speed control system of each axis of the manipulator as it is. The simulator itself. That is, the observer gain of the PI controllers 601 and 602 is an observer gain equivalent to the speed control. In addition, since integration control is included, a function to make the steady deviation of the estimated output value zero is also built-in. There is no engineering cost for approximation and gain adjustment in observer implementation.

The PI control type observer based on the nonlinear dynamic model 603 is represented by a second-order differential form shown in Expression (7) obtained by modifying Expression (1). In Expression (7), the symbol “^” represents an estimated value.
However,
dθ _M / dt = [dθ _M1 / dt, dθ _M2 / dt] ^T : Input motor angular velocity to observer u = [u ₁ , u ₂ ] ^T : Input to observer (motor current command value)
K _PV = diag (k _PV1 , k _PV2 ): Speed deviation proportional control gain K _IV = diag (k _IV1 , k _IV2 ): Speed deviation integral control gain τ = [τ ₁ , τ ₂ ] ^T : Input to observer ( Motor current command value)
It is.

  Here, the observer gains of the PI controllers 601 and 602 are selected to be the same as the gains of P and I in the FF-IP control (two-degree-of-freedom PI control) of the speed loop of the actual robot. In this case, since only one degree of freedom of the follow-up performance of the observer needs to be considered, FF = P, that is, PI control with the feedforward system as a proportional system is set. This system is a kind of nonlinear observer with a constant gain. In addition, since integral control is included, the steady-state deviation of the output estimation value becomes zero, and a function as a disturbance removal observer is also incorporated.

  The state observer represented by the equation (7) is mounted on the controller 100A, and the state can be estimated by integrating twice in real time using the integrator shown in FIG. Actual integration is performed by numerical integration.

  In the two-inertia system angular velocity control, if the three state quantities of the motor angular velocity, the angular velocity of the link, and the axial torsional angular velocity of the motor and the link are fed back, any pole can be arranged, and the speed response (control system) The speed of following the target value) and the attenuation rate can be freely set. However, it is difficult to migrate because it involves readjustment of the existing PI control system. Therefore, as an input to the PI control system, first, only the estimated value of the axial torsional angular velocity (the difference between the link angular velocity and the motor angular velocity) output from the observer 600 is fed back. The feedback of the axial torsional angular velocity is expected to increase only the attenuation rate without changing the gain crossover frequency of PI control. For this reason, hand adjustment at the site is easy, so it is particularly easy to introduce it into a control device for an industrial robot.

FIG. 33 is a block diagram of the manipulator control unit 112 (110) having the observer 600 of FIG. The observer 600 receives the motor angular velocity (encoder difference) and the input voltage (current command value) to the motor driver. The state feedback units 701, 702, 703, and 704 of the estimated values of the axial torsional angular velocities for the first and second links by the PI control type observer 600 are two-degree-of-freedom PI control (FF-I) shown in Expression (7). -P control) can be realized by plugging in, and the control law shown in equation (8) is obtained.

The state feedback control gain k _TVi (i = 1, 2) of the estimated value of the axial torsional angular velocity can be easily adjusted manually while viewing the time response waveform. In addition, since an observer based on an accurate nonlinear dynamic model of a manipulator having an elastic joint is used, in the vibration suppression control of a manipulator configured with a serial link, a sufficient effect can be obtained with only the state feedback of only the first axis. is there. FIG. 33 shows a link angular acceleration feedback constant calculation unit 705 for increasing the inertia. The function of the feedback constant calculation unit 705 will be described later.

In the present embodiment, high inertia control for pretending to show a large inertia on the link side when the hand load is light will be described. The control law is obtained by adding negative feedback of the estimated value of the link angular acceleration to Expression (8) (state feedback units 703 and 704 in FIG. 33), and is as shown in Expression (9).

Here, the state feedback control gain k _AVi of the estimated value of the angular acceleration has the dimension of the inertia moment of the link, and it is considered that the inertia variation Δm of the tip load of the manipulator can be compensated. For example, when the inertia matrix of Equation (3) receives a change (decrease) in Δm, the base parameters α, β, and γ change to α ′, β ′, and γ ′ as shown in Equation (10), respectively.
α '= α-Δm
β ′ = β−Δm (10)
γ ′ = γ−Δm
Here, α ′, β ′, γ ′ after inertia change of the tip load of the manipulator shown in Expression (10) is substituted into the (1, 1) component of the inertia matrix of Expression (3), and the reduction ratio n _G1 Is taken into account, the equation (11) is obtained.
k _AV1 = (2 + 2 cos (θ _L2 )) Δm × n _G1 (11)
From this equation (11), it is possible to calculate the state feedback control gain of the estimated value of the angular acceleration of the first axis. Expression (11) is a function of the link angular acceleration feedback constant calculation unit 705 for increasing inertia shown in FIG.

(Manipulator control unit based on nonlinear dynamic model (2))
FIG. 34 is a block diagram of the manipulator control unit 113 (110). The manipulator control unit 113 is provided with link angular acceleration feedback control by high inertia and constant inertia and a gain calculation function thereof.

  The above-described decrease in inertia occurs not only when the load mass at the arm tip is light, but also when the manipulator changes in posture. In the two-link manipulator, for example, when the angle of the second axis increases, the inertia around the first axis decreases. Therefore, it is considered to compensate for a decrease in inertia due to a change in posture by feedback of link angular acceleration.

In the case of a two-link manipulator, when the amount of inertia reduction from the maximum value of the (1, 1) component of the inertia matrix shown in Equation (3) is taken into account, Equation (12) is obtained.
k _AV1 = 2γ (1-cos (θ _L2 )) n _G1 (12)
By this equation (12), the feedback control gain of the link angular acceleration of the first axis can be calculated. By this feedback, the inertia is kept constant regardless of the posture of the arm, that is, constant inertia control is realized, and the vibration suppression effect equivalent to the above-described high inertia control is obtained.

This constant inertial control and the high inertial control of the embodiment can be used in combination, and when Formula (11) and Formula (12) are merged, Formula (13) is obtained.
k _AV1 = [(2 + 2 cos (θ _2L )) Δm + 2γ (1−cos (θ _L2 ))] n _G1 (13)
From this equation (13), it is possible to calculate the feedback control gain of the link angular acceleration by the high inertia and constant inertia control of the first axis. This is the function of the link angular acceleration feedback constant calculation unit 705A by the high inertia and constant inertia control shown in FIG.

Further, paying attention to the non-diagonal elements of the inertia matrix shown in the equation (3), the feedback control of the link angular acceleration by the high inertia and constant inertia control is also performed for the (1, 2) component and the (2, 1) component. The gains k _BV1 and k _BV2 can be calculated as shown in equations (14) and (15), respectively.
k _BV1 = [(1 + cos (θ _2L )) Δm + γ (1−cos (θ _L2 ))] n _G1 (14)
k _BV2 = [(1 + cos (θ _2L )) Δm + γ (1−cos (θ _L2 ))] n _G1 (15)

Using these k _{BV1 and} k _BV2 , when the feedback of the angular acceleration of the second axis is added to the first axis and the feedback of the angular acceleration of the first axis is added to the second axis, the equation (16) is obtained. It becomes like this.

  In this way, a vibration suppression control system to which generalized high inertia and constant inertia control is added is obtained. FIG. 34 is a block diagram of the entire vibration suppression control system to which high inertia and constant inertia control is added.

  FIG. 35 is a block diagram of another observer 600A. The observer 600A of this embodiment has a configuration in which a physical parameter switching unit 606 is newly provided in the observer 600 of FIG. In the observer 600A of this embodiment, the physical parameter set switching (gain) of the nonlinear dynamics model 603 is performed by the physical parameter switching unit 606, as shown in FIG. Scheduling) is performed so that the vibration suppression performance is robust regardless of fluctuations in the robot tip load and friction force. In order to switch the physical parameter set explicitly, for example, a robot language such as “Payload (5 kg)” may be implemented and used at the timing when the tip load changes. Examples of the physical parameter set include mass, moment of inertia, friction coefficient, spring coefficient, and the like.

  In the example illustrated in FIG. 35, the physical parameter switching unit 606 requires five physical parameter sets with a tip load of 5, 4, 3, 2, 1, 0 kg. However, as long as the physical parameter set of 5 kg is accurately estimated, the physical parameter set of the tip load of 4 kg can be generated by the optimization method based on the nonlinear least square method described above. In the same manner, a physical parameter set of 4 kg to 3 kg, 3 kg to 2 kg, 2 kg to 1 kg, and 1 kg to 0 kg can be generated.

  FIG. 36 is a block diagram of still another observer 600B. The observer 600B of this embodiment has a configuration in which a table storage unit 607, a step response data comparison unit 608, and a physical parameter optimization unit 609 are newly provided in the observer 600 of the embodiment shown in FIG. Yes. In the observer 600B of the fifth embodiment, the tip load may deviate from numerical values such as 5 kg, 4 kg, 3 kg, 2 kg, 1 kg, and 0 kg. In FIG. 36, a table storage unit 607 for storing speed step response data for each mass (5 kg, 4 kg, 3 kg, 2 kg, 1 kg, 0 kg) acquired in advance and a physical parameter set at that time is prepared. ing.

  First, speed step response data of an actual machine having an unknown tip load is acquired. Next, the step response data comparison unit 608 compares the acquired speed step response data with the speed step response data for each tip load stored in the table storage unit 607. Based on the comparison result, the tip load corresponding to the speed step response data of the actual machine is the closest from the tip loads stored in the table storage unit 607, for example, 5 kg, 4 kg, 3 kg, 2 kg, 1 kg, 0 kg. A tip load is selected, and a physical parameter set corresponding to the selected tip load is selected as an initial value. The selection of the tip load closest to the tip load corresponding to the speed step response data of the actual machine is, for example, the difference between the speed step response data of the actual machine and the speed step response data of each tip load stored in the table storage unit 607. Select the tip load that minimizes the sum of the squares.

  Further, optimization calculation is performed by the physical parameter optimization unit 609 using the physical parameter set selected as the initial value. This optimization calculation is exactly the same as the above-described method based on the nonlinear least square method. As described above, since an accurate physical parameter set can be obtained even for an unknown tip load, an accurate observer can be constructed and effective vibration suppression control can be provided.

  DESCRIPTION OF SYMBOLS 1,1A ... Object handling apparatus, 2 ... Arm (movable part), 20 ... Manipulator, 21 ... Camera (first camera), 30 ... Base, 33 ... Camera (second camera), 50 ... Processing part (mounting part) ), 51 ... Mark (first mark), 71 ... Mark (second mark), 110 ... Manipulator control unit, 101 j ... Automatic calibration unit (calibration processing unit), 200 ... Hand (effector, movable unit), 200 g ... Contact Sensor (sensor), 200h ... Force sensor (sensor), 200i ... Displacement sensor (sensor), X ... direction (first direction), Y ... direction (first direction), Z ... direction (second direction), δx, δy: deviation (first error), δz: deviation (second error).

Claims (8)

Base and
A manipulator provided on the base and having a link, a movable part, and a motor for rotationally driving the link;
A first camera provided in the manipulator;
Based on a result of photographing the predetermined position of the base by the first camera and a load generated in the manipulator based on a change amount of a rotation angle of the link and the motor when the manipulator contacts the base. A calibration processing unit for executing a position calibration process;
An object handling device comprising: A second camera provided on the base;
The calibration processing unit performs a calibration process by acquiring an image calibration value for a position obtained from an image of the second camera based on a plurality of images captured by the second camera of the movable unit. The object handling apparatus according to 1. A first mark provided on the base;
The object handling apparatus according to claim 1, wherein the calibration processing unit performs a calibration process based on an image obtained by photographing the first mark by the first camera. A placement unit for placing an object;
The object handling device according to claim 3, wherein the first mark is provided at or adjacent to the mounting portion.   5. The object handling device according to claim 1, wherein the calibration processing unit executes a calibration process based on an image obtained by photographing the predetermined position outside the object handling device by the first camera.   The said calibration process part performs a calibration process based on the image which the said 1st camera image | photographed the 2nd mark provided in the access target by the effector located outside the said object handling apparatus. Object handling device. An object handling device, comprising: a base; a manipulator provided on the base; a link; a movable part; a motor that rotationally drives the link; and a first camera provided on the manipulator. A control device that executes a calibration process,
A first acquisition unit for acquiring a photographing result at a predetermined position of the base by the first camera;
A second acquisition unit that acquires a load generated in the manipulator based on a change amount of a rotation angle of the motor and the link when the manipulator contacts the base;
A calibration processing unit that executes a calibration process of the position of the movable unit based on the imaging result obtained by the first obtaining unit and the load obtained by the second obtaining unit;
A control device comprising: A base, a manipulator provided on the base, a link, a movable part, and a motor that rotationally drives the link, a first camera provided on the manipulator, and a calibration process for the position of the movable part A calibration method executed by an object handling device comprising:
The first camera captures a predetermined position of the base;
The position of the movable part is based on a load generated on the manipulator based on a result of photographing by the first camera and a change amount of a rotation angle of the link and the motor when the manipulator contacts the base. Perform the calibration process of
Calibration method.