Robot control device and robot control method

Abstract

This robot control device comprises an external force estimation unit that acquires torque sensor information and tactile sensor information from, respectively, a torque sensor and a tactile sensor that are provided to a controlled portion of a robot and, on the basis of the acquired torque sensor information and tactile sensor information, separates intended external force applied to the robot by a task from the external force acting on the controlled portion and estimates unintended external force.

Description

Robot control device and robot control method

[0001] The present disclosure relates to a robot control device and a robot control method.

[0002] In robots, hands are often used to manipulate objects, making them prone to colliding with the target object or the environment. Highly rigid hand control provides high positional accuracy but reduces safety when coming into contact with the surrounding environment or objects. On the other hand, flexible hand control provides high safety but reduces positional accuracy, potentially leading to task failure. Patent Document 1 proposes a technology that provides high-precision and flexible gripping with the same mechanism by providing the fingers of a hand with two members of different rigidity.

[0003] JP 2009-101424 A

[0004] It is conceivable that achieving both high accuracy and high safety can be achieved not by the mechanical structure of the robot, but by adaptively changing hand control in response to the external force acting on the hand. In this case, it is desirable to estimate the external force with high accuracy.

[0005] It is desirable to provide a robot control device and a robot control method that are capable of estimating external forces with high accuracy.

[0006] A robot control device according to one embodiment of the present disclosure includes an external force estimation unit that acquires torque sensor information and tactile sensor information from a torque sensor and a tactile sensor, respectively, provided in a part of the robot to be controlled, and based on the acquired torque sensor information and tactile sensor information, separates intended external forces due to tasks assigned to the robot from external forces acting on the part to be controlled, and estimates unintended external forces.

[0007] A robot control method according to one embodiment of the present disclosure includes acquiring torque sensor information and tactile sensor information from a torque sensor and a tactile sensor, respectively, provided in a part of the robot to be controlled, and, based on the acquired torque sensor information and tactile sensor information, separating intended external forces due to a task assigned to the robot from external forces acting on the part to be controlled, and estimating unintended external forces.

[0008] In a robot control device or a robot control method according to one embodiment of the present disclosure, torque sensor information and tactile sensor information are acquired from a torque sensor and a tactile sensor, respectively, provided in a part of the robot to be controlled, and based on the acquired torque sensor information and tactile sensor information, intended external forces due to tasks assigned to the robot are separated from external forces acting on the part to be controlled, and unintended external forces are estimated.

[0009] 1 is an explanatory diagram showing an example of an environment in which an object that is a target of hand manipulation by a robot according to a comparative example is placed. FIG. 2 is an explanatory diagram showing an example of interference with a peripheral object accompanying hand manipulation by a robot according to a comparative example. FIG. 3 is an explanatory diagram showing an example of a decrease in position accuracy when flexible hand control is performed in a robot according to a comparative example. FIG. 4 is an external view schematically showing an example of a robot that is a target of control by a robot control device and a robot control method according to a first embodiment of the present disclosure. FIG. 5 is a configuration diagram schematically showing an example of a hand of a robot that is a target of control by a robot control device and a robot control method according to the first embodiment. FIG. 6 is a block diagram showing an example of a configuration of an external force estimating unit in the robot control device according to the first embodiment. FIG. 7 is an explanatory diagram showing an example of estimation of a specific task and an external force corresponding to the task. FIG. 8 is a block diagram showing an example of the overall configuration of a robot control device according to the first embodiment. FIG. 9 is an explanatory diagram showing an example of torque modulation by an external force adaptive control unit in the robot control device according to the first embodiment. FIG. 10 is an explanatory diagram showing an example of torque modulation and hand control corresponding to the task. FIG. 11 is a flowchart showing an example of control operation by the robot control device according to the first embodiment. FIG. 12 is an explanatory diagram showing an overview of hand control by a robot control device and a robot control method according to a second embodiment. FIG. 13 is a block diagram showing an example of the overall configuration of a robot control device according to a third embodiment. 10 is a block diagram showing an example of the overall configuration of a robot control device according to a fourth embodiment; and FIG. 11 is a flowchart showing an example of a control operation by the robot control device according to the fourth embodiment.

[0010] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. The description will be made in the following order: 0. Comparative Example (FIGS. 1 to 3) 1. First Embodiment (FIGS. 4 to 11) 1.1 Configuration of External Force Estimation Unit and Example of External Force Estimation Calculation 1.2 Configuration of Force Control Unit and Example of Hand Control 1.3 Operation 1.4 Effects 1.5 Modifications 2. Second Embodiment (Example of Individual Hand Control for Each Finger) (FIGS. 12 to 13) 3. Third Embodiment (Example of Task-Dependent Hand Control) (FIG. 14) 4. Fourth Embodiment (Example of Surrounding Environment-Dependent Hand Control) (FIGS. 15 to 16) 5. Other Embodiments

[0011] <0. Comparative Example> (Outline and Problems of a Robot Control Apparatus and a Robot Control Method According to a Comparative Example) FIG. 1 shows an example of an environment in which an object to be manipulated by a robot according to a comparative example is placed.

[0012] In a robot, a hand is often used to manipulate an object, and therefore it is prone to colliding with the object or the environment. This is particularly noticeable in a cluttered environment where there are many surrounding objects 201 in addition to the object 200 to be manipulated, or in a case where the object 200 to be manipulated is placed near the environment, such as a shelf 300, as shown in FIG.

[0013] Fig. 2 shows an example of interference with a surrounding object 201 during hand manipulation by a robot according to a comparative example. In Fig. 2, (A) shows a state before the object 200 to be manipulated by the hand 101 is grasped, and (B) shows a state immediately before the object 200 to be manipulated is grasped by the hand 101. Fig. 3 shows an example of a decrease in positional accuracy when flexible hand control is performed in a robot according to a comparative example. As shown in Fig. 3, the hand 101 includes a first finger portion 111 and a second finger portion 112 as a grasping portion.

[0014] In the case of highly rigid hand control, as shown in FIG. 2B , although positional accuracy is high, safety is reduced due to contact (collision) with the surrounding environment or a surrounding object 201. On the other hand, flexible hand control provides high safety, but is sensitive to disturbances, noise, model errors, etc., resulting in low accuracy of the position and posture of the hand, which may lead to task failure. For example, as shown in FIG. 3 , when attempting to grasp an object 200 by aligning the object center 200C with the center position 101C of the hand 101, kinematic errors due to bending of the hand 101, model errors, etc. may cause the first finger 111 to collide with the object 200, resulting in task failure.

[0015] In general, it is difficult to achieve both highly rigid and highly accurate hand motion and flexible and highly safe hand motion using the same control law. Therefore, switching control laws depending on the task and environment is one option, but this poses the following challenges in terms of external force sensing and force control. One challenge with external force sensing is the difficulty of distinguishing between external forces caused by intentional contact, which is necessary for task execution, and external forces caused by unintentional collisions with the environment or obstacles. Another challenge with force control is the possibility of reduced stability, such as vibrations or discontinuities in motion, when switching motions.

[0016] Therefore, from the viewpoint of external force sensing, it is desirable to develop a technology that can estimate external forces with high accuracy. Also, from the viewpoint of force control, it is desirable to develop a technology that can perform adaptive force control according to external forces. As a result, it is desirable to develop a technology that can achieve both high accuracy and high safety in the control of robot hands, etc.

[0017] 1. First embodiment [1.1 Configuration of external force estimating unit and example of calculation for external force estimation] Fig. 4 schematically illustrates an example of a robot 100 that is an object to be controlled by a robot control device and a robot control method according to a first embodiment of the present disclosure. Fig. 5 schematically illustrates an example of a hand 101 of the robot 100.

[0018] The robot control device and robot control method according to the first embodiment are suitable for a robot 100 having, for example, an arm 102, as shown in Fig. 4. A hand 101 is provided at the tip of the arm 102 as a part to be controlled. However, the robot control device and robot control method according to the first embodiment can also use a part other than the hand 101 as a part to be controlled. For example, a foot (not shown) can also be used as a part to be controlled.

[0019] The hand 101 is capable of grasping an object 200 (see FIG. 1 , etc.) to be manipulated. The hand 101 has a gripping portion that grasps the object 200 to be manipulated, and a joint portion linked to the gripping portion. The configuration example in FIG. 5 shows an example in which the gripping portion has two fingers (a first finger 111 and a second finger 112). The first finger 111 is provided with a plurality of joint portions 401. The second finger 112 is provided with a plurality of joint portions 402. A tactile sensor 11 is provided at the tip of each of the first finger 111 and the second finger 112. A torque sensor 12 is provided at each of the plurality of joint portions 401 and the plurality of joint portions 402.

[0020] The gripping portion may have three or more fingers, and each finger may have only one joint, or may have three or more joints.

[0021] FIG. 6 shows an example of the configuration of the external force estimating unit in the robot control device according to the first embodiment.

[0022] In the following text, "(^)" indicates a character with a circumflex (^) above the character to its left. For example, "F(^)" indicates a character with a circumflex above F. Similarly, "(·), (··)" indicates a character with a dot (··) (differential symbol) above the character to its left. For example, "q(··)" indicates a character with a dot (··) above q.

[0023] The robot control device according to the first embodiment includes an external force estimating unit 20. The external force estimating unit 20 receives torque sensor information τ from the torque sensor 12 and the tactile sensor 11 provided in the hand 101, which is a part to be controlled of the robot 100. ｅ and tactile sensor information p. The external force estimation unit 20 acquires the acquired torque sensor information τ ｅ Based on the tactile sensor information p, the intended external forces due to the task given to the robot 100 are separated from the external forces acting on the controlled part, and unintended external forces are estimated.

[0024] The external force estimation unit 20 includes a pressure-load conversion unit 21 , an inverse dynamics calculation unit 22 , a subtractor 23 , a torque-force moment conversion unit 24 , and a subtractor 25 .

[0025] The external force estimation unit 20 estimates the torque sensor information τ ｅ The first external force (external force F(^)) calculated based on ｔｒｑ ) to obtain the second external force (external force F(^) calculated based on the tactile sensor information p. ｐｒｅ ) to calculate the unintended external force (estimated external force F(^) ｅｘ ) is estimated.

[0026] The external force estimation unit 20 estimates the torque sensor information τ ｅ and a first external force (external force F(^)) based on an estimated torque (estimated joint torque, dynamics component τ(^)) calculated using inverse dynamics calculation based on a target motion of the controlled part. ｔｒｑ ) is calculated.

[0027] An example of calculation of external force estimation by the external force estimating unit 20 when the controlled part is the hand 101 will be described in more detail below.

[0028] The tactile sensor 11 receives an external operation force (external force F(^)) via the object 200 grasped by the hand 101. ｐｒｅ The pressure-load converter 21 converts the pressure distribution information (tactile sensor information p) from the tactile sensor 11 into, for example, six-dimensional force / moment information (external force F(^) ｐｒｅ On the other hand, the joint torque detected by the torque sensor 12 (torque sensor information τ ｅ) includes, as a dynamics component τ(^), a force (torque) generated by the dynamics of the robot 100 itself, in addition to external forces from the object 200 and the environment. The dynamics component τ(^) can be calculated by the inverse dynamics calculation using the following equation (1) by the inverse dynamics calculation unit 22. In equation (1), q is the joint angle vector, M is the inertia matrix (inertia force term), c is the centrifugal force / Coriolis force term, and g is the gravity term.

[0029]

[0030] Torque sensor information τ from the torque sensor 12 ｅ The dynamics component τ(^) corresponding to the current state and the target motion is subtracted by the subtractor 23 from the torque information τ(^), thereby eliminating the dynamics component τ(^) of the robot 100 itself and subtracting the components from the object 200 and the environment. ｅ The extracted torque information τ(^) ｅ From this, the torque-force moment conversion unit 24 converts the hand information (external force F(^)) into ｔｒｑ ) to obtain information (external force F(^) ｐｒｅ ) and the dimension is the same. ｔｒｑ can be calculated by the following equation (2) using the Jacobian matrix J.

[0031]

[0032] Finally, the subtractor 25 calculates the external operation force (external force F(^)) via the object 200. ｐｒｅ ) and external force F(^) from object 200 and the environment ｔｒｑ By comparing the external force with the external force, it is possible to determine whether the external force is an intended external force due to task operation or an unexpected and unintended external force generated by contact with the environment (estimated external force F(^) ｅｘ ) can be separated (Equation (3)).

[0033]

[0034] FIG. 7 shows an example of a specific task and an estimation of an external force corresponding to the task.

[0035] Specific examples of tasks include approaching an object 200 in free space (Example 1) and interacting with the environment via a gripping tool (Example 2). Examples of tasks in which an unintended external force occurs include an unexpected collision with the environment or object 200 (Example 3) and an unexpected collision with the environment or object 200 while interacting with the environment via a gripping tool (Example 4).

[0036] If there is no unexpected collision, no unintended external force occurs, or the unintended external force is small, the estimated external force F(^) ｅｘ is 0 or a value close to 0. When an unexpected collision occurs and an unintended external force occurs, the estimated external force F(^) ｅｘ is a value that corresponds to an unintended external force.

[0037] [1.2 Configuration of Force Control Unit and Example of Hand Control] In the robot control device and robot control method according to the first embodiment, the estimated external force F(^) ｅｘ Based on this, force control of the control target portion of the robot 100 is performed. Fig. 8 shows an example of the overall configuration of a robot control device according to the first embodiment, which is equipped with a force control unit 30 that performs such force control.

[0038] The robot control device according to the first embodiment includes a force control unit 30 and a joint control unit 41 .

[0039] The force control unit 30 includes a pressure-load conversion unit 31, a Cartesian space control unit 32, a torque command value calculation unit 33, an adder , and an external force adaptive control unit .

[0040] The joint 40 includes joints 401 and 402 in the example configuration of the hand 101 shown in FIG.

[0041] The external force adaptive control unit 35 detects the unintended external force (estimated external force F(^)) estimated by the external force estimation unit 20. ｅｘ ) based on the magnitude of the torque sensor information τ ｅ For example, as the unintended external force increases, the external force adaptive control unit 35 modulates the torque sensor information τ ｅ The torque sensor information τ ｅ Modulation is performed.

[0042] The adder 34 calculates a torque command value τ for the control target part calculated based on the exercise purpose. ｄ and the torque sensor information τ after modulation by the external force adaptive control unit 35. ａｄｊ The adder 34 calculates a final torque command value for the controlled part based on the above. The adder 34 corresponds to a specific example of a "torque control unit" in the technology of the present disclosure.

[0043] Hereinafter, an example of hand control of the robot 100 will be described as an example of force control by the force control unit 30.

[0044] In general, task-based hand 101 movement goal x ｄ is indicated in Cartesian space coordinates. The Cartesian space control unit 32 includes various controllers, and controls the motion target x ｄ Acceleration target x (...) ｄ The controller may be, for example, a PID (Proportional Integral Differential) controller that realizes a position target at the tip of the hand 101, or a force information (external force F(^) ｐｒｅ In addition, a linear controller or a non-linear controller may be used to further improve the positional accuracy of the tip of the hand 101.

[0045] Generally, the torque command value calculation unit 33 calculates the acceleration target x(...) ｄ Using dynamic calculations, the torque command value τ ｄ and instructs the joint control unit 41 to obtain the motion target x ｄ However, with only this type of control, the external force and external torque components (torque sensor information τ ｅ , τ ａｄｊ ) is not taken into consideration, the joints 40 of the hand 101 are insensitive to external forces, making it difficult to achieve any other movements than those with high rigidity.

[0046] On the other hand, by using the torque sensor 12, the external torque acting on the joint 40 of the hand 101 from the environment or the like is calculated as torque sensor information τ ｅIn the force control method according to the first embodiment, the external torque (torque sensor information τ ｅ ) is converted into an estimated external force value (estimated external force F(^) ｅｘ ) based on 0 to τ ｅ Then, in the adder 34, the modulated torque sensor information τ ａｄｊ The torque command value τ ｄ By adding this to the control, natural switching between high rigidity control and flexible control is realized.

[0047] The external force estimated value (estimated external force F(^)) by the external force adaptive control unit 35 ｅｘ ) using torque sensor information τ ｅ The modulation of is expressed as in equation (4).

[0048] FIG. 9 shows an example of torque modulation by the external force adaptive control unit 35. For example, in the case of linear modulation as shown in FIG. 9, the following modulation results are obtained: ｅｘ ) is smaller than the threshold value th1, the modulated torque sensor information τ ａｄｊ is τ ａｄｊ = 0.0. Estimated external force (estimated external force F(^) ｅｘ ) is greater than the threshold value th2, the modulated torque sensor information τ ａｄｊ is τ ａｄｊ = τ ｅ ・External force estimated value (estimated external force F(^) ｅｘ ) is between the thresholds th1 and th2, the modulated torque sensor information τ ａｄｊ is transformed into τ by a linear transformation such as equation (5). ａｄｊ = 0 to τ ｅ This becomes:

[0049]

[0050] The torque modulation is not limited to the linear modulation example described above, but may be modulation using a non-linear function.

[0051] FIG. 10 shows an example of a specific task and torque modulation and hand control corresponding to the task.

[0052] In the case of approaching the object 200 in free space in Example 1, the motion objective is to control the trajectory tracking of the tip position and tip attitude of the hand 101, and the result of external force estimation by the external force estimating unit 20 is F(^). ｅｘ In this case, the external force adaptive control unit 35 modulates the torque to obtain τ ａｄｊ = 0. This weakens the ability to adapt to external forces, and prioritizes position control and attitude control of the tip of the hand 101 (resulting in highly rigid and highly accurate hand operation).

[0053] In the case of acting on the environment via the gripping tool in Example 2, the objective of the movement is to control the impedance of the force via the object 200, and the result of the external force estimation by the external force estimating unit 20 is F(^). ｅｘ In this case, the external force adaptive control unit 35 modulates the torque to obtain τ ａｄｊ = 0. This makes it possible to realize a desired operation (such as a highly accurate impedance operation of the hand 101) via the object 200 or the tool.

[0054] In the case of the environment of Example 3 or an unexpected collision with the object 200, the motion objective is to control the trajectory tracking of the tip position and tip posture of the hand 101, and the result of the external force estimation by the external force estimating unit 20 is F(^). ｅｘ In this case, the external force adaptive control unit 35 modulates the torque to ａｄｊ = τ ｅ This will strengthen the ability to adapt to external forces and prioritize highly flexible and safe hand movements.

[0055] In the case of an unexpected collision with the environment or the object 200 during an action on the environment via the gripping tool in Example 4, the motion objective is to control the impedance of the force via the object 200, and the external force estimation result by the external force estimating unit 20 is F(^). ｅｘ In this case, the external force adaptive control unit 35 modulates the torque by τ ａｄｊ = 0 to τ ｅ This allows the hand to perform a flexible hand movement while maintaining the desired operation via the object 200 or tool as much as possible. How flexible and safe the movement should be depends on modulation by the external force adaptive control unit 35.

[0056] [1.3 Operation] FIG. 11 is a flowchart showing an example of a control operation by the robot control device according to the first embodiment.

[0057] First, the force control unit 30 calculates the external operation force (external force F(^) ｐｒｅ ), and torque sensor information τ ｅ (Step S101). Next, the external force estimating unit 20 calculates an estimated joint torque (dynamics component τ(^)) using inverse dynamics calculation based on the target motion (Step S102). Next, the external force estimating unit 20 calculates an estimated joint torque (dynamics component τ(^)) using inverse dynamics calculation based on the target motion (Step S102). Next, the external force estimating unit 20 calculates an estimated joint torque (dynamics component τ(^)) using inverse dynamics calculation based on the target motion (Step S103). ｅ , and the estimated external force (external force F(^) ｔｒｑ Next, the external force estimating unit 20 calculates the estimated external force (external force F(^) ｔｒｑ ) and external operating force (external force F(^) ｐｒｅ ) and the force difference amount (estimated external force F(^) ｅｘ ) is calculated (step S104).

[0058] Next, the external force adaptive control unit 35 calculates the modulated torque sensor information τ ａｄｊ Next, the adder 34 calculates the torque command value τ ｄ and the modulated torque sensor information τ ａｄｊ From this, the final joint command torque is calculated (step S106). Next, the joint control unit 41 controls the target joint (joint unit 40) based on the final joint command torque (step S107).

[0059] [1.4 Effects] As described above, according to the robot control device and the robot control method according to the first embodiment, torque sensor information τ ｅ and tactile sensor information p is acquired, and the acquired torque sensor information τ ｅ Based on the tactile sensor information p, the intended external force due to the task given to the robot is separated from the external force acting on the controlled part, and the unintended external force (estimated external force F(^) ｅｘThis makes it possible to estimate the external force with high accuracy.

[0060] According to the robot control device and the robot control method of the first embodiment, the estimated external force F(^) ｅｘ This makes it possible to achieve both high precision and high safety in the control of the hand 101 of the robot 100. For example, it is possible to switch between high rigidity and high precision hand control and high flexibility and high safety hand control under the same control law.

[0061] The robot control device and robot control method according to the first embodiment make it possible to adaptively realize high-rigidity, high-precision hand control and high-flexibility, high-safety hand control according to the task and environment. Automatic hand control can be performed based on estimated external forces without the need for manual setting. For example, the following hand control can be automatically performed:

[0062] For example, when approaching the object 200 in free space, if there is no unexpected external force (no unexpected external force on the tactile sensor 11, no unexpected external force on the torque sensor 12), priority can be given to position control and attitude control of the tip of the hand 101. This allows for highly rigid and highly accurate hand operation, aiming for a high task success rate.

[0063] Furthermore, for example, when acting on the environment via the grasped object 200 or the grasping tool, if there is an external force (an external force on the tactile sensor 11, but no unexpected external force on the torque sensor 12), priority can be given to the force (impedance) control of the tip of the hand 101. This allows for highly accurate impedance operation of the hand 101, aiming to improve the success rate of contact tasks.

[0064] Furthermore, for example, in the case of an unintentional collision with the environment or an obstacle, an external force is detected (no external force detected by the tactile sensor 11, but an external force detected by the torque sensor 12), and priority can be given to adaptive control of the external force acting on the joints 40 of the hand 101. This allows for highly flexible and safe hand operation, thereby reducing the possibility of damage to the environment or the object 200.

[0065] Furthermore, for example, if the fingers of the hand 101 unintentionally collide with the environment or an obstacle while acting on the environment via the grasped object 200 or tool, an external force is applied (an external force is applied to the tactile sensor 11 and an external force is applied to the torque sensor 12), and adaptive control of the external force acting on the joints 40 of the hand 101 can also be taken into consideration. This makes it possible to realize highly accurate impedance operation of the hand 101 while also taking into account highly flexible and safe hand operation. How highly flexible and safe the operation should be depends on modulation by the external force adaptive control unit 35.

[0066] The effects described in this specification are merely examples and are not limiting, and other effects may also be achieved. The same applies to the effects of other embodiments described below.

[0067] [1.5 Modifications] In the above description, an example has been given in which torque modulation by the external force adaptive control unit 35 is performed using a predetermined function, but torque modulation may also be performed using modulation parameters obtained by machine learning, for example, rather than a predetermined function. This makes it possible to acquire control parameters that are robust and suitable for hand operation in a learning environment.

[0068] 2. Second Embodiment Next, a robot control device and a robot control method according to a second embodiment of the present disclosure will be described. Note that, hereinafter, parts that are substantially the same as the components of the robot control device and the robot control method according to the first embodiment described above will be assigned the same reference numerals, and descriptions thereof will be omitted as appropriate.

[0069] FIG. 12 shows an outline of hand control by a robot control device and a robot control method according to the second embodiment.

[0070] For example, individual hand control may be performed according to the positions of the fingers (first finger 111, second finger 112) of the hand 101 as a gripping unit. For example, as shown in Fig. 12, when the first finger 111 and the second finger 112 are in different positional relationships with respect to the direction of gravity, hand control may be performed such that the finger on the floor side (e.g., the second finger 112) is operated with high rigidity to prioritize accuracy, and the opposite finger (e.g., the first finger 111) is operated with low rigidity and flexibility to hold the object 200.

[0071] Furthermore, for example, if the hand 101 has a finger portion corresponding to the thumb and other finger portions, the hand may be controlled so that the finger portion corresponding to the thumb operates with high rigidity and the other finger portions operate with low rigidity and flexibility.

[0072] FIG. 13 shows an example of the overall configuration of a robot control device according to the second embodiment, which realizes the above-described hand control.

[0073] In the robot control device according to the second embodiment, the force control unit 30 has a plurality of adders 341 and 342 as the adder 34. The adders 341 and 342 correspond to a specific example of a "torque control unit" in the technology of the present disclosure.

[0074] The external force estimating unit 20 receives the joint angle vectors q1 and q2 from the plurality of joints 401 and 402 in the first finger 111 and the second finger 112, respectively.

[0075] The external force adaptive control unit 35 calculates torque sensor information τ ｅ is modulated individually according to the positions of the joints 401 and 402, and the modulated torque sensor information τ ａｄｊ 1, τ ａｄｊ 2 is output to the adders 341 and 342.

[0076] The adders 341 and 342 calculate torque command values ​​τ for the respective joints 401 and 402 based on the motion objective. ｄ 1, τ ｄ 2, and modulated torque sensor information τ at each of the plurality of joints 401, 402 individually modulated by the external force adaptive control unit 35.ａｄｊ 1, τ ａｄｊ 2, the final torque command values ​​for each of the joints 401 and 402 are calculated.

[0077] The joint control units 411 and 412 control the joint units 401 and 402 based on the final torque command values ​​for the joint units 401 and 402, respectively.

[0078] Other configurations, operations, and effects may be substantially the same as those of the robot control device and robot control method according to the first embodiment.

[0079] 3. Third Embodiment Next, a robot control device and a robot control method according to a third embodiment of the present disclosure will be described. Note that, hereinafter, parts that are substantially the same as the components of the robot control device and the robot control method according to the first or second embodiment will be assigned the same reference numerals, and descriptions thereof will be omitted as appropriate.

[0080] FIG. 14 shows an example of the overall configuration of a robot control device according to the third embodiment.

[0081] In the robot control device according to the third embodiment, the force control unit 30 performs force control in accordance with the task specified by the task setting unit 50. The external force adaptive control unit 35 controls the force in accordance with the unintended external force (estimated external force F(^)) estimated by the external force estimating unit 20. ｅｘ ) and the task designated by the task setting unit 50, the torque sensor information τ ｅ Modulation is performed.

[0082] According to the robot control device and robot control method of the third embodiment, for tasks that do not require precision, such as opening a door, it is possible to perform hand control that prioritizes safe and flexible hand movements.

[0083] Furthermore, for example, in a task where positional accuracy is important, even if the hand 101 comes into contact with the environment, it is possible to perform hand control that prioritizes high rigidity and high precision operation.

[0084] Furthermore, for example, hand control can be performed to reset priorities in response to replanning of a task in the event of a failure.

[0085] Other configurations, operations, and effects may be substantially the same as those of the robot control device and robot control method according to the first embodiment.

[0086] 4. Fourth Embodiment Next, a robot control device and a robot control method according to a fourth embodiment of the present disclosure will be described. Note that, hereinafter, parts that are substantially the same as the components of the robot control device and the robot control method according to any one of the first to third embodiments will be assigned the same reference numerals, and descriptions thereof will be omitted as appropriate.

[0087] The external force adaptive control unit 35 calculates torque sensor information τ based on the magnitude of the unintended external force estimated by the external force estimating unit 20 and environmental information around the robot 100. ｅ The environmental information may include distance information about the object 200 in the vicinity of the robot 100.

[0088] FIG. 15 shows an example of the overall configuration of a robot control device according to a fourth embodiment, which realizes hand control using the distance information described above.

[0089] In the robot control device according to the second embodiment, the force control unit 30 further includes a distance calculation unit 14. The distance calculation unit 14 calculates the distance d between the robot 100 and the surrounding environment or objects based on sensor information from the distance sensor 13. The distance sensor 13 is configured, for example, with a camera (image sensor) or LiDAR (Light Detection and Ranging).

[0090] The robot control device according to the fourth embodiment enables hand control according to the surrounding environment. For example, in an environment where there are people or objects around the robot 100, the hand can be controlled to perform safe and flexible movements.

[0091] Fig. 16 is a flowchart showing an example of a control operation by the robot control device according to the fourth embodiment. In the flowchart shown in Fig. 16, the processing steps other than step S101A and step S105A are the same as those in the flowchart shown in Fig. 11.

[0092] First, the force control unit 30 calculates the external operation force (external force F(^) ｐｒｅ ), torque sensor information τ ｅ , and the distance d are acquired (step S101A). Next, the external force estimating unit 20 calculates the estimated joint torque (dynamics component τ(^)) using inverse dynamics calculation based on the target motion (step S102). Next, the external force estimating unit 20 calculates the estimated joint torque (dynamics component τ(^)), torque sensor information τ ｅ , and the estimated external force (external force F(^) ｔｒｑ Next, the external force estimating unit 20 calculates the estimated external force (external force F(^) ｔｒｑ ) and external operating force (external force F(^) ｐｒｅ ) and the force difference amount (estimated external force F(^) ｅｘ ) is calculated (step S104).

[0093] Next, the external force adaptive control unit 35 sets a threshold value for torque modulation based on the distance d (step S105A). Next, the external force adaptive control unit 35 calculates the torque sensor information τ ａｄｊ Next, the adder 34 calculates the torque command value τ ｄ and the modulated torque sensor information τ ａｄｊ From this, the final joint command torque is calculated (step S106). Next, the joint control unit 41 controls the target joint (joint unit 40) based on the final joint command torque (step S107).

[0094] Other configurations, operations, and effects may be substantially the same as those of the robot control device and robot control method according to the first embodiment.

[0095] 5. Other Embodiments The technology according to the present disclosure is not limited to the above-described embodiments, and various modifications are possible.

[0096] For example, the present technology can be configured as follows. According to the present technology configured as follows, torque sensor information and tactile sensor information are acquired from torque sensors and tactile sensors provided in the controlled parts of the robot, respectively, and based on the acquired torque sensor information and tactile sensor information, intended external forces due to tasks assigned to the robot are separated from external forces acting on the controlled parts, and unintended external forces are estimated. This makes it possible to estimate external forces with high accuracy.

[0097] (1) A robot control device comprising: an external force estimating unit that acquires torque sensor information and tactile sensor information from a torque sensor and a tactile sensor, respectively, provided in a portion of the robot to be controlled, and that separates an intended external force due to a task assigned to the robot from external forces acting on the portion to be controlled based on the acquired torque sensor information and tactile sensor information, and estimates an unintended external force. (2) The robot control device described in (1) above, in which the external force estimating unit estimates the unintended external force by subtracting a second external force calculated based on the tactile sensor information from a first external force calculated based on the torque sensor information. (3) The robot control device described in (2) above, in which the external force estimating unit calculates the first external force based on the torque sensor information and an estimated torque calculated using inverse dynamics calculation based on a desired motion of the portion to be controlled. (4) The robot control device described in any one of (1) to (3) above, further comprising: an external force adaptive control unit that modulates the torque sensor information based on the magnitude of the unintended external force estimated by the external force estimating unit. (5) The robot control device according to any one of (4) above, further comprising a torque control unit that calculates a final torque command value for the control-target part based on the torque command value for the control-target part calculated based on a motion purpose and the torque sensor information modulated by the external force adaptive control unit. (6) The robot control device according to (4) or (5) above, wherein the external force adaptive control unit modulates the torque sensor information so that the torque value indicated by the torque sensor information increases as the unintended external force increases.(7) The robot control device according to (5) above, wherein the portion to be controlled has a plurality of joints, the torque sensor is provided at each of the plurality of joints, the external force adaptive control unit individually modulates the torque sensor information at each of the plurality of joints in accordance with positions at which the plurality of joints are provided, and the torque control unit calculates a final torque command value for each of the plurality of joints based on the torque command value for each of the plurality of joints calculated based on the movement purpose and the torque sensor information at each of the plurality of joints individually modulated by the external force adaptive control unit. (8) The robot control device according to (4) or (5) above, wherein the external force adaptive control unit modulates the torque sensor information based on the magnitude of the unintended external force estimated by the external force estimator and a specified task. (9) The robot control device according to (4) above, wherein the external force adaptive control unit modulates the torque sensor information based on the magnitude of the unintended external force estimated by the external force estimator and environmental information around the robot. (10) The robot control device according to (9) above, wherein the environmental information includes distance information for objects in the vicinity of the robot. (11) The robot control device according to any one of (1) to (10) above, wherein the robot includes, as the controlled portion, a hand capable of grasping an object, the hand having a gripping portion that grips the object and a joint that moves the gripping portion, the torque sensor being provided at the joint of the hand, and the tactile sensor being provided at the gripping portion of the hand. (12) A robot control method comprising: acquiring torque sensor information and tactile sensor information from a torque sensor and a tactile sensor provided at the controlled portion of the robot, respectively; and separating intended external forces due to a task assigned to the robot from external forces acting on the controlled portion based on the acquired torque sensor information and tactile sensor information, and estimating unintended external forces.

[0098] This application claims priority based on Japanese Patent Application No. 2021-111776, filed on July 5, 2021, in the Japan Patent Office, the entire contents of which are incorporated herein by reference.

[0099] Those skilled in the art will recognize that various modifications, combinations, subcombinations, and variations may occur depending on design requirements and other factors, and are intended to be within the scope of the appended claims and their equivalents.

Claims

1. A robot control device comprising an external force estimation unit that acquires torque sensor information and tactile sensor information from a torque sensor and a tactile sensor provided at a controlled part of a robot, respectively, and separates an intended external force due to a task given to the robot from an external force acting on the controlled part based on the acquired torque sensor information and tactile sensor information, and estimates an unintended external force.

2. The robot control device according to claim 1, wherein the external force estimation unit estimates the unintended external force by subtracting a second external force calculated based on the tactile sensor information from a first external force calculated based on the torque sensor information.

3. The robot control device according to claim 2, wherein the external force estimation unit calculates the first external force based on the torque sensor information and an estimated torque calculated using an inverse dynamics operation based on a target motion of the controlled part.

4. The robot control device according to claim 1, further comprising an external force adaptation control unit that modulates the torque sensor information based on a magnitude of the unintended external force estimated by the external force estimation unit.

5. The robot control device according to claim 4, further comprising a torque control unit that calculates a final torque command value for the controlled part based on a torque command value for the controlled part calculated based on a motion purpose and the torque sensor information modulated by the external force adaptation control unit.

6. The robot control device according to claim 4, wherein the external force adaptation control unit modulates the torque sensor information so that a torque value indicated by the torque sensor information increases as the unintended external force increases.

7. The controlled part has a plurality of joint parts, the torque sensors are provided at each of the plurality of joint parts, the external force adaptation control part performs individual modulation on the torque sensor information at each of the plurality of joint parts according to the positions where the plurality of joint parts are provided, and the torque control part calculates a final torque command value for each of the plurality of joint parts based on the torque command value for each of the plurality of joint parts calculated based on the motion purpose and the torque sensor information at each of the plurality of joint parts individually modulated by the external force adaptation control part. The robot control device according to claim 5.

8. The external force adaptation control part performs modulation of the torque sensor information based on the magnitude of the unintended external force estimated by the external force estimation part and the specified task. The robot control device according to claim 4.

9. The external force adaptation control part performs modulation of the torque sensor information based on the magnitude of the unintended external force estimated by the external force estimation part and the environmental information around the robot. The robot control device according to claim 4.

10. The environmental information includes distance information with respect to an object around the robot. The robot control device according to claim 9.

11. The robot includes, as the controlled part, a hand capable of gripping an object, the hand has a gripping part for gripping the object and a joint part, the torque sensor is provided at the joint part of the hand, and the tactile sensor is provided at the gripping part of the hand. The robot control device according to claim 1.

12. A robot control method including: acquiring torque sensor information and tactile sensor information from a torque sensor and a tactile sensor provided at a controlled part of a robot respectively; and separating an intended external force due to a task given to the robot from an external force acting on the controlled part based on the acquired torque sensor information and tactile sensor information, and estimating an unintended external force.