Title of invention: control device, control method and program
Technical field
[0001]
　The present disclosure relates to control devices, control methods and programs.
Background technology
[0002]
　Conventionally, in Patent Document 1 below, it is determined whether or not slip occurs on the contact surface based on the amount of change in the center position of the pressure on the contact surface and the gripping force of the grip portion that grips the target object. It is stated that.
Prior art literature
Patent documents
[0003]
Patent Document 1: Japanese Unexamined Patent Publication No. 2008-55540
Outline of the invention
Problems to be solved by the invention
[0004]
　Detection of partial slippage is effective for grasping an object by a robot or the like. Partial slip is a phenomenon that occurs prior to total slip, in which the relative position with the object shifts and slides down, and a part of the contact surface starts to slip. At this time, in the state of partial slippage, the relative position with respect to the object does not deviate.
[0005]
　However, since the technique described in the above patent document is a method of detecting the total slip when the object starts to slide, the gripping force cannot be controlled unless the object starts to slide. Therefore, in the technique described in the above patent document, it is difficult to control the gripping force before the object starts to slide to perform stable gripping. Moreover, in the first place, the fact is that there is no effective technique for detecting partial slippage. When attempting to detect partial slip, shear deformation of the contact portion that occurs prior to partial slip is detected, making it difficult to determine the minimum gripping force based on partial slip. Furthermore, when the pressure distribution is flat, such as when the pressure distribution is flat, such as when the object is hard or the surface of the object is flat, the progress of partial slip becomes fast, and it is difficult to detect partial slip.
[0006]
　Therefore, it has been required to detect the slip of an object with high accuracy by detecting the partial slip.
Means to solve problems
[0007]
　According to the present disclosure, based on pressure information obtained from a plurality of regions having different sliding characteristics when a contacted object slides, a state of total slippage in which the object is sliding is detected in each of the plurality of the regions. A control device is provided that includes an overall slip detector.
　Further, according to the present disclosure, based on pressure information obtained from a plurality of regions having different sliding characteristics when a contacted object slides, a state of total slippage in which the object is sliding in each of the plurality of the regions A control method is provided that comprises detecting.
　Further, according to the present disclosure, based on pressure information obtained from a plurality of regions having different sliding characteristics when a contacted object slides, a state of total slippage in which the object is sliding is determined in each of the plurality of the regions. A program for operating a computer is provided as a means of detection.
The invention's effect
[0008]
　As described above, according to the present disclosure, it is possible to detect the slip of an object with high accuracy by detecting the partial slip.
　It should be noted that the above effects are not necessarily limited, and together with or in place of the above effects, any of the effects shown herein, or any other effect that can be grasped from this specification. May be played.
A brief description of the drawing
[0009]
FIG. 1 is a schematic diagram showing a configuration of a robot hand according to an embodiment of the present disclosure.
FIG. 2 is a schematic view showing a state in which an object is in contact with the first and second flexible layers.
FIG. 3A schematically shows how the contact state between the flexible layer and the object changes in time series from the state a to the state f in the model shown in FIG. 2 from the time when the object is gripped until the object starts to slide. It is a figure.
3B is a characteristic diagram showing how the pressure center position changes in each region of the first flexible layer and the second flexible layer in the states a to f of FIGS. 3A.
FIG. 4 is a characteristic diagram showing a case where the friction coefficients of the first flexible layer and the second flexible layer are the same for comparison with FIG. 3B.
FIG. 5A is a schematic view showing the division direction of the flexible layer.
[Fig. 5B] Fig. 5B is a schematic view showing the division direction of the flexible layer.
[Fig. 5C] It is a schematic diagram which shows the division direction of a flexible layer.
FIG. 6 is a schematic diagram showing a configuration example of a robot control system according to an embodiment of the present disclosure.
FIG. 7 is a schematic view showing a configuration of a gripping force calculation unit according to the first modification.
[Fig. 8A] Fig. 8A is a characteristic diagram showing an example of pressure distribution when a flexible layer comes into contact with an object.
[Fig. 8B] Fig. 8B is a characteristic diagram showing an example of pressure distribution when a flexible layer comes into contact with an object.
[Fig. 8C] Fig. 8C is a characteristic diagram showing an example of pressure distribution when a flexible layer comes into contact with an object.
FIG. 9 is a schematic view showing a configuration of a gripping force calculation unit according to the second modification.
FIG. 10 is a schematic diagram showing specific control of a hand.
FIG. 11 is a schematic diagram showing an example of division of a flexible layer.
[Fig. 12] Fig. 12 is a plan view showing an example of division that does not depend on the position of contact with an object.
FIG. 13 is a schematic diagram showing an example in which distributed pressure sensors are staggered and arranged in layers in order to pseudo-reduce the pitch width of the nodes of two distributed pressure sensors.
FIG. 14 is a schematic diagram showing an example in which flexible layers are arranged above and below a distributed pressure sensor.
[Fig. 15A] Fig. 15A is a schematic diagram showing the detection sensitivity of a distributed pressure sensor according to the difference in the thickness of the flexible layer.
15B is a characteristic diagram showing how the pressure center position changes in the examples (a) to (c) and the modified example 4 shown in FIG. 15A as in FIG. 3B.
FIG. 16A is a schematic view showing the reason why the movement direction of the pressure center position of the modified example 4 is opposite to that of the examples (a) to (c).
FIG. 16B is a schematic view showing the reason why the moving direction of the pressure center position of the modified example 4 is opposite to that of the examples (a) to (c).
FIG. 17 is a schematic view showing an example in which the surface area of ​​the flexible layer is changed.
FIG. 18 is a schematic diagram showing an example of changing the hardness of the flexible layer below the distributed pressure sensor in order to change the frictional force of the flexible layer above the distributed pressure sensor.
FIG. 19 is a schematic view showing a configuration example using a linear flexible layer according to the modified example 6.
FIG. 20 is a schematic view showing a configuration of a gripping force calculation unit using the linear flexible layer shown in FIG.
FIG. 21 is a schematic view showing an example in which the directions of the flexible layers are arranged in multiple directions in the configuration using the linear flexible layers shown in FIG.
Mode for carrying out the invention
[0010]
　Preferred embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. In the present specification and the drawings, components having substantially the same functional configuration are designated by the same reference numerals to omit duplicate description.
[0011]
　The explanations will be given in the following order.
　1. 1. Outline of the present disclosure
　2. Hand composition
　3. Sliding of an object against the flexible layer
　　3.1. "Whole slip" and "partial slip"
　　3.2. Changes in the contact state between the flexible layer and the object
　　3.3. Judgment of slip based on pressure center position
　　3.4. Parameters that make the timing of overall slippage different for each flexible layer
　　3.5. Dividing direction of flexible layer
　4. Configuration example of robot control system
　5. Modifications of this embodiment
　　5.1. Modification 1 (Example of adjusting the gripping force control gain according to the rigidity of the object)
　　5.2. Deformation example 2 (Example of controlling the position and posture of the finger in order to increase the difference in the timing of occurrence of overall slippage)
　　5.3. Modification 3 (Variation of arrangement of flexible layer and distributed pressure sensor)
　　5.4. Modification 4 (Example of arranging flexible layers above and below the distributed pressure sensor)
　　5.5. Modification 5 (Method of changing the coefficient of friction of the flexible layer)
　　5.6. Deformation example 6 (Example using a linear flexible layer)
[0012]
　1. 1. Outline of the present disclosure
　For example, when an object is gripped by a robot hand, it is desirable to grip the object with an appropriate force so as not to slip off the hand. As a result, the object is not destroyed by the gripping force, while the object can be reliably gripped. In particular, when gripping a flexible object, it is possible to suppress destruction and deformation of the object. The present disclosure relates to a technique for detecting a state of "partial slip" before a state of "total slip" in which an object starts to slide occurs when gripping an object, and gripping the object with an optimum gripping force.
[0013]
　2. 2. The configuration
　diagram 1 of the hand is a schematic diagram showing the configuration of the robot hand 500 according to the embodiment of the present disclosure. The hand 500 is provided at the tip of the arm 506 of the robot. As shown in FIG. 1, the hand 500 has a main body 501, links 512 and 514 that make up the first finger 502, and links 516 and 518 that make up the second finger 504. Further, actuators are provided at each joint 520, 522, 524, 526. The link 512 rotates with respect to the link 514 by the driving force of the actuator of the joint 520, and the link 514 rotates with respect to the main body 501 by the driving force of the joint 522 actuator. Similarly, the link 516 rotates with respect to the link 518 by the driving force of the actuator of the joint 524, and the link 518 rotates with respect to the main body 501 by the driving force of the actuator of the joint 526.
[0014]
　Further, the arm 506 has multiple joints as an example, and a plurality of links are rotatably connected by each joint. Each link rotates with each other by the driving force of the actuator provided in each joint. As a result, an articulated arm 506 having a predetermined degree of freedom and capable of moving the hand 500 to a desired position is configured.
[0015]
　FIG. 1 shows a state in which the first finger 502 and the second finger 504 are gripping the object (grasping object) 600. Distributed pressure sensors 530 and 532 are provided inside the link 512 (object 600 side) of the first finger 502. A first flexible layer 540 is provided inside the distributed pressure sensor 530, and a second flexible layer 542 is provided further inside the distributed pressure sensor 532. Similarly, distributed pressure sensors 530 and 532 are provided inside the link 516 (object 600 side) of the second finger 504. A first flexible layer 540 is provided inside the distributed pressure sensor 530, and a second flexible layer 542 is provided inside the distributed pressure sensor 532. The first flexible layer 540 and the second flexible layer 542 are made of an elastic material having viscosity, elasticity, or both, and are made of a material that can be easily deformed by an external load, for example. It is composed of materials such as urethane gel and silicon gel. The first flexible layer 540 is made of a material having a lower coefficient of friction than the second flexible layer 542. The slip detection device according to the present embodiment includes first and second flexible layers 540 and 542 and distributed pressure sensors 530 and 532. The first and second flexible layers 540, 542 and the distributed pressure sensors 530, 532 may be directly attached to the arm 506. Further, even if the first and second flexible layers 540, 542 and the distributed pressure sensors 530, 532 are attached to the feet of the robot and detect the slip state between the feet and the ground (floor). good. In this way, the first and second flexible layers 540, 542 and the distributed pressure sensors 530, 532 can be attached to the action portion on which the robot acts on the object.
[0016]
　FIG. 2 is a schematic view showing how the first and second flexible layers 540, 542 of the first finger 502 are in contact with the object 600. Although the spherical object 600 is shown in FIG. 1, the rectangular parallelepiped object 600 is shown in FIG. The shape of the object 600 is not particularly limited. The direction of the x-axis shown in FIG. 2 is a direction in which the object 600 slides (or tends to slide) relative to the first and second flexible layers 540 and 542. As shown in FIG. 2, the two flexible layers 540 and 542 are arranged in the order of the second flexible layer 542 and the first flexible layer 540 with respect to the direction in which the object 600 slides. It is possible to increase the difference in the timing of occurrence of overall slippage between the flexible layer 540 of 1 and the flexible layer 542 of the second. In the example of FIG. 1, the x-axis direction corresponds to the gravity direction.
[0017]
　As shown in FIG. 2, a force Ft is applied to the object 600 in the x-axis direction. If the x-axis direction is the direction of gravity, the force Ft corresponds to gravity. Further, a force Fn is applied to the object 600 in the y direction orthogonal to the x-axis direction. The force Fn corresponds to the reaction force when the object 600 is gripped by the first finger 502 and the second finger 504.
[0018]
　3. 3. Sliding of an object against the flexible layer
　　3.1. "Whole slip" and "partial slip" As
　shown in FIG. 2, a force Ft is applied to the object 600 in the x-axis direction. When the x-axis direction is the direction of gravity, if the gripping force of the first finger 502 and the second finger 504 to grip the object 600 is weak, the object 600 slides down in the direction of gravity. The transition of the state from the stopped state to the start of sliding of the object 600 can be explained by the phenomena of "total slip" and "partial slip".
[0019]
　"Overall slip" is a state in which the relative positions of the object 600 and the flexible layer are displaced and the object is slipping down. "Partial slip" is a phenomenon that occurs prior to "total slip" and a part of the contact surface between the object 600 and the flexible layers 540 and 542 slips. In the present embodiment, when the object 600 is gripped, "partial slip" is detected in which the object 600 is gripped with the minimum force that does not cause the object 600 to slip off.
[0020]
　　3.2. Changes in the contact state between the flexible layer and the object
　FIG. 3A shows that in the model shown in FIG. 2, the contact state between the flexible layers 540, 542 and the object 600 changes between the time when the object 600 is gripped and the time when the object 600 starts to slide. It is a figure which shows typically how it changes in time series from a state a to a state f. In FIG. 3A, the upper surface of the flexible layers 540 and 542 is divided into a plurality of rectangular regions, and two types of dots attached to each rectangular region are used to determine the contact state with the object 600 by "fixing" and "sliding". It is classified into. In the "fixed" rectangular region, slip does not occur between the object 600 and the flexible layers 540 and 542, and both are fixed. On the other hand, in the rectangular region of "slip", slip occurs between the object 600 and the flexible layers 540 and 542. The state of each rectangular region can be obtained from, for example, analysis by simulation. In the following, based on FIG. 3A, a state in which "partial slip" and "total slip" occur between the time when the object 600 is gripped and the time when the object 600 starts to slide will be described.
[0021]
　The flexible layer provided on the distributed pressure sensors 530 and 532 is equally divided into two, a first flexible layer 540 and a second flexible layer 542, and by the division, the first flexible layer 540 and the first flexible layer 540 The second flexible layer 542 is arranged side by side in the sliding direction (x-axis direction) of the object 600. In the state a shown in FIG. 3A, all the rectangular regions on the upper surfaces of the first flexible layer 540 and the second flexible layer 542 are in the “fixed” state. Next, in the state b, the rectangular region on the upper surface of the first flexible layer 540, which is close to the second flexible layer 542, is changed to the “sliding” state. The other areas are in a "fixed" state.
[0022]
　Next, in the state c of FIG. 3A, the rectangular region in the “sliding” state is enlarged on the upper surface of the first flexible layer 540. In the next state d, all the rectangular areas on the upper surface of the first flexible layer 540 are in the "sliding" state, and a part of the rectangular area on the upper surface of the second flexible layer 542 is changed to the "sliding" state. ing. In the next state e, the rectangular region in the "sliding" state is enlarged on the upper surface of the second flexible layer 542. In the next state f, all the rectangular regions on the upper surface of the second flexible layer 542 are in the “sliding” state.
[0023]
　In each of the first flexible layer 540 and the second flexible layer 542, when all the rectangular regions are in the "sliding" state, the "overall sliding" state is obtained. The first flexible layer 540 is in the "overall slip" state in the state d, and is also in the "overall slip" state in the subsequent states e and f. On the other hand, the second flexible layer 542 is in a state of "overall slip" in the state f, lagging behind the first flexible layer 540.
[0024]
　As described above, in each of the first flexible layer 540 and the second flexible layer 542, the region of "slip" expands with the passage of time and becomes a state of "overall slip", but the friction coefficient is low. It can be seen that the timing of the first flexible layer 540 to be in the "whole slip" state is earlier. In other words, by making the friction coefficients of the two flexible layers 540 and 542 different, it is possible to make a difference in the timing at which the overall slip occurs.
[0025]
　In the state f in which the total slip occurs in the two regions of the first flexible layer 540 and the second flexible layer 542 at the same time, the total slip occurs in the entire region including the first flexible layer 540 and the second flexible layer 542. It has occurred. In this state, the object 600 is relatively moving with respect to both the first flexible layer 540 and the second flexible layer 542, and in FIG. 2, the object 600 is sliding in the x-axis direction. ..
[0026]
　Further, in the present embodiment, the states d and e in which the total slip occurs in one region of the first flexible layer 540 and the second flexible layer 542 and the total slip does not occur in the other region. , Is defined as a state in which "partial slip" occurs in the entire region including the first flexible layer 540 and the second flexible layer 542. In the states d and e where "partial slip" occurs, the object 600 does not move relative to the first flexible layer 540 and the second flexible layer 542. In FIG. 3A, the states d and e are surrounded by a thick line to indicate that the states d and e are in the “partial slip” state.
[0027]
　When focusing on one of the regions of the first flexible layer 540 and the second flexible layer 542, the region of the first flexible layer 540 in the states b and c, or the second region in the states d and e. In the region of the flexible layer 542, it can be considered that partial slippage has occurred. However, in the present embodiment, focusing on the entire region including the first flexible layer 540 and the second flexible layer 542, total slip occurs in one of the flexible layers 540 and 542, and the other. The states d and e in which total slippage does not occur in the entire region are defined as the states in which "partial slippage" occurs in the entire region.
[0028]
　Further, the states a to c in which the entire slip does not occur in any of the first flexible layer 540 and the second flexible layer 542 are the entire region including the first flexible layer 540 and the second flexible layer 542. Is in a state of being fixed to the object 600.
[0029]
　Therefore, the states d and e in which the "partial slip" occurs are the states immediately before the object 600 starts to slide, and the object 600 is relative to the first flexible layer 540 and the second flexible layer 542. Not moving. Therefore, by detecting the state of "partial slip" and gripping the object with a gripping force that causes partial slip, deformation and destruction of the object 600 are suppressed, and the object 600 is gripped with an optimum force that does not slip. be able to.
[0030]
　　3.3. Determining slip based on pressure center position In
　this embodiment, it is distributed that "partial slip" or "total slip" occurs in the entire region including the first flexible layer 540 and the second flexible layer 542. Judgment is made based on the pressure center position obtained from the mold pressure sensors 530 and 532.
[0031]
　FIG. 3B is a characteristic diagram showing how the pressure center position changes in each region of the first flexible layer 540 and the second flexible layer 542 in the states a to f of FIGS. 3A. The pressure center position X cop can be obtained from the following equation (1). Each of the distributed pressure sensors 530 and 532 is provided with a plurality of nodes for detecting pressure, which are formed in a matrix. In the formula (1), N is the number of sensor nodes distributed pressure sensor 530, 532, x i is the coordinate of the i th node, p (x i ) is the pressure at which the i-th node has detected. The pressure center position X cop is a value obtained by dividing the sum of the multiplication values ​​of the pressure and the coordinates by the sum of the pressures, and is a value indicating the center of the pressure in the distributed pressure sensors 530 and 532.
[0032]
[Number 1]

[0033]
　In FIG. 3B, in order from the left, the change in the pressure center position of the second flexible layer 542, the change in the pressure center position of the first flexible layer 540, and the pressure center position of the first and second flexible layers 540, 542. It shows each change. In the three characteristic diagrams, the horizontal axis indicates the number of steps in time, and the vertical axis indicates the pressure center position. The pressure center position on the vertical axis corresponds to the position in the x-axis direction in FIG. In the characteristics of the first flexible layer 540, the origin of the vertical axis corresponds to the origin of FIG. 2, and in the characteristics of the second flexible layer 542, the origin of the vertical axis corresponds to the coordinates of −L in FIG. It is assumed that the lengths of the first flexible layer 540 and the second flexible layer 542 in the x-axis direction are the same (= L).
[0034]
　Further, in FIG. 3B, the a to f attached to the characteristics of the pressure center position correspond to the states a to f in FIG. 3A, respectively.
[0035]
　As shown in FIG. 3B, with the passage of time, the pressure center positions of the first and second flexible layers 540 and 542 are both moved in the x-axis direction of FIG. At this time, the first flexible layer 540 having a smaller coefficient of friction moves faster at the pressure center position than the second flexible layer 542. In the first flexible layer 540, the movement of the pressure center position is stopped when the number of time steps exceeds 15, and the pressure center position is in a steady state of a constant value. On the other hand, in the second flexible layer 542, the movement of the pressure center position is stopped when the number of steps of time exceeds 25, and the pressure center position is in a steady state of a constant value.
[0036]
　As shown in FIG. 3B, in each of the first flexible layer 540 and the second flexible layer 542, the state in which the movement of the pressure center position is stopped corresponds to the state of "overall slip". On the other hand, in each of the first flexible layer 540 and the second flexible layer 542, the state in which the movement of the pressure center position is not stopped is the shear deformation and partial portion of the first flexible layer 540 or the second flexible layer 542. The pressure center position is moving due to the slippage. In the state where the movement of the pressure center position is not stopped, the relative movement between the object 600 and the flexible layer 540 of the first and the second flexible layer 542 does not occur. On the other hand, in a state where the movement of the pressure center position is not stopped, the absolute position of the object 600 may change due to the shear deformation of the first flexible layer 540 or the second flexible layer 542. As described above, by making the friction coefficients of the two flexible layers 540 and 542 different, it is possible to make a difference in the timing at which the overall slip occurs, and as is clear from FIG. 3B, the second flexible layer The timing at which the entire slip occurs in 542 is later than that in the first flexible layer 540.
[0037]
　The characteristic diagram shown on the right side of FIG. 3B is a superposition of the characteristic diagram shown on the left side of FIG. 3B and the characteristic diagram shown in the center. In the present embodiment, based on the change in the pressure center position, the first flexible layer 540 in which the total slip occurs first causes the total slip, and the second flexible layer 542 does not cause the total slip. It is determined that "partial slip" has occurred in the entire region including the first flexible layer 540 and the second flexible layer 542. Further, based on the change in the pressure center position, the state in which the entire slip occurs in both the first flexible layer 540 and the second flexible layer 542, the first flexible layer 540 and the second flexible layer 542 are changed. It is determined that "whole slip" has occurred in the entire area including the area. Further, based on the change in the pressure center position, a state in which overall slippage does not occur in both the first flexible layer 540 and the second flexible layer 542 is determined to be a fixed state.
[0038]
　From the above, by calculating the pressure center position in each region of the first flexible layer 540 and the second flexible layer 542, the total slip in each region can be detected. In the above example, since the number of regions is two, when total slip occurs in two regions at the same time, total slip occurs in the entire region including the first flexible layer 540 and the second flexible layer 542. It is determined that there is (state f). Further, when the total slip occurs in any of the regions, it is determined that the partial slip has occurred in the entire region (states d and e). If no total slip occurs in any of the regions, it is determined that the entire region is in the "fixed state" (states a, b, c).
[0039]
　Then, the ratio of the undetected region of "overall slip" to the entire region (contact region of the object 600) including the first flexible layer 540 and the second flexible layer 542 is defined as the fixation rate. In FIG. 3B, in the state f, both the region of the first flexible layer 540 and the region of the second flexible layer 542 are relative to the entire region including the first flexible layer 540 and the second flexible layer 542. The sticking rate is 0% because it is in a state of total slippage. Further, in the states d and e, since the region of the first flexible layer 540 is in a state of total slippage with respect to the entire region including the first flexible layer 540 and the second flexible layer 542, the adhesion rate is high. It is 50%. Further, in the states a, b, and c, the region of the first flexible layer 540 and the region of the second flexible layer 542 are relative to the entire region including the first flexible layer 540 and the second flexible layer 542. The sticking rate is 100% because none of them are in a state of total slippage.
[0040]
　In the present embodiment, the gripping force of the hand 500 is controlled according to the sticking rate. The larger the sticking rate, the smaller the gripping force by the hand 500, and the smaller the sticking rate, the larger the gripping force by the hand 500. As a result, the object 600 can be gripped with the minimum necessary force, and the destruction or deformation of the object 600 can be suppressed.
[0041]
　As the number of divisions of the flexible layer region increases, the resolution of the fixing rate increases, and the accuracy of gripping force control increases. Further, as the number of divisions of the flexible layer region increases, it becomes possible to detect the adhesion rate even for a smaller object or an object having irregularities. For example, when the number of divisions of the flexible layer is set to 3, the adhesion rate can be calculated in four stages of 0%, 33%, 66%, and 100% by obtaining the adhesion rate by the same method as described above.
[0042]
　For example, when the number of divisions of the flexible layer is three and the friction coefficient of each flexible layer is different, the adhesion rate is 0% if the entire flexible layer is in a state of total slippage. Become. Further, when the total slip occurs in the region of the flexible layer having the smallest friction coefficient and the region of the flexible layer having the next smallest friction coefficient, and the total slip does not occur in the region of the flexible layer having the largest friction coefficient, the adhesion rate Is 33%. Further, when the total slip occurs in the region of the flexible layer having the smallest friction coefficient, and then the total slip does not occur in the region of the flexible layer having the smallest friction coefficient and the region of the flexible layer having the largest friction coefficient. Is 66%. Further, when the whole slip occurs in all the regions of the flexible layer, the adhesion rate becomes 0%. From the same viewpoint, the more the number of divisions of the flexible layer region is increased, the more accurately the state of partial slip can be detected.
[0043]
　FIG. 4 is a characteristic diagram showing a case where the friction coefficients of the first flexible layer 540 and the second flexible layer 542 are the same for comparison with FIG. 3B. The conditions other than the coefficient of friction and the display method of the characteristic diagram are the same as those in FIG. 3B. As shown in FIG. 4, when the friction coefficients of the first flexible layer 540 and the second flexible layer 542 are the same, the total slippage occurs in each region of the first flexible layer 540 and the second flexible layer 542. Since there is no difference in the timing of occurrence, it is difficult to detect partial slip in the entire region including the first flexible layer 540 and the second flexible layer 542. Therefore, as in the present embodiment shown in FIG. 3B, “partial slip” occurs in the state where the adhesion rate is 50%, that is, in the entire region including the first flexible layer 540 and the second flexible layer 542. Cannot detect the state. According to the present embodiment, it is possible to detect a state in which "partial slip" occurs in the entire region including the first flexible layer 540 and the second flexible layer 542, which corresponds to the state of partial slip. It is possible to control the gripping force with high accuracy based on the sticking rate.
[0044]
　Therefore, by making the friction coefficients of the first flexible layer 540 and the second flexible layer 542 different, the timing at which the entire slip occurs between the first flexible layer 540 and the second flexible layer 542 is different, and the first and the first are. Partial slip can be detected in the entire region including the second flexible layers 540 and 542. The greater the difference in timing at which the first flexible layer 540 and the second flexible layer 542 occur, the more partial slip occurs in the entire region including the first and second flexible layers 540 and 542. Since the time is increased, the gripping force can be easily controlled.
[0045]
　　3.4. Parameters that make the overall slip occurrence timing different for each flexible layer In the
　above description, the friction coefficients of the first flexible layer 540 and the second flexible layer 542 are different, and the first flexible layer 540 and the second flexible layer 540 and the second flexible layer are made different. The timing of occurrence of the overall slip of the layer 542 was changed. On the other hand, parameters other than the friction coefficient of the first flexible layer 540 and the second flexible layer 542 are made different so that the overall slip occurrence timing of the first flexible layer 540 and the second flexible layer 542 is different. You may. Parameters other than the coefficient of friction include Young's modulus, Poisson's ratio, thickness, radius of curvature, and the like.
[0046]
　In the case of the coefficient of friction, the smaller the coefficient of friction, the earlier the timing of occurrence of overall slip. In the case of Young's modulus, the larger the Young's modulus, the earlier the timing of occurrence of overall slippage. In the case of Poisson's ratio, the smaller the Poisson's ratio, the earlier the timing of overall slip occurrence. In the case of thickness, the smaller the thickness, the earlier the timing of occurrence of overall slippage. In the case of the radius of curvature, the larger the radius of curvature, the earlier the timing of occurrence of overall slip.
[0047]
　Further, in the above parameters, it is possible to further widen the difference in the timing at which the overall slip occurs by combining the conditions under which the overall slip occurs earlier or the conditions at which the overall slip occurs later. For example, by providing the first flexible layer having a small friction coefficient and a small thickness and the second flexible layer having a large friction coefficient and a large thickness, the entire slip of the first flexible layer and the second flexible layer can be prevented. It is possible to further expand the timing of occurrence.
[0048]
　　3.5. Dividing Direction of Flexible Layer
　FIGS. 5A to 5C are schematic views showing a dividing direction of the flexible layer. 5A and 5B show an example in which the first flexible layer 540 and the second flexible layer 542 are divided into two in the sliding direction, as in FIGS. 1 and 2. Further, FIG. 5C shows an example in which the first flexible layer 540 and the second flexible layer 542 are divided into two in a direction orthogonal to the sliding direction.
[0049]
　When the vertical axis (z-axis in FIGS. 5A to 5C) of the gripped object 600 is constrained with respect to the sliding direction of the object 600, each of the first flexible layer 540 and the second flexible layer 542 is complete. There is no difference in the timing of slippage. On the other hand, when the vertical axis is not constrained, a slight shaft rotation occurs due to the friction distribution, and when the vertical axis is divided perpendicular to the sliding direction, the difference in the timing at which complete slip occurs becomes larger. .. In a real environment, the axis of the object cannot be constrained, so it is more preferable that the object is divided in the sliding direction as shown in FIGS. 5A and 5B. Further, in FIG. 5A in which the second flexible layer 542 and the first flexible layer 540 are arranged in this order in the direction in which the object 600 slides, the first flexible layer 540 and the second in the direction in which the object 600 slides are arranged. It is possible to make a difference in the timing at which the entire slip occurs as compared with FIG. 5B in which the flexible layers 542 are arranged in this order. In other words, by arranging the second flexible layer 542 having a larger friction coefficient on the upstream side in the sliding direction, the difference in the occurrence timing of the total slip between the first flexible layer 540 and the second flexible layer 542 can be further increased. It can be made larger. In the case of the arrangement of FIG. 5B, since the sliding force of the first flexible layer 540 located on the upstream side in the sliding direction is blocked by the second flexible layer 542 on the downstream side, the entire sliding in the first flexible layer 540 Occurrence timing is relatively late. Therefore, in the case of the arrangement shown in FIG. 5B, the difference in the timing of occurrence of overall slip between the first flexible layer 540 and the second flexible layer 542 is relatively small. On the other hand, in the case of the arrangement of FIG. 5A, the sliding force of the second flexible layer 542 located on the upstream side in the sliding direction is not blocked by the first flexible layer 540 on the downstream side, and the first flexible layer on the downstream side is not blocked. Overall slip occurs in the flexible layer 540 at a relatively early timing. Therefore, in the case of the arrangement of FIG. 5A, the difference in the timing at which the overall slip occurs can be made larger.
[0050]
　Further, the plurality of flexible layers may not be provided adjacent to each other. For example, as shown in FIG. 10 described later, the first flexible layer 540 and the second flexible layer 542 may be separately provided on different fingers of the hand 500.
[0051]
　4. Configuration Example of Robot Control System
　FIG. 6 is a schematic diagram showing a configuration example of the robot control system (control device) 1000 according to the embodiment of the present disclosure. As shown in FIG. 6, the control system 1000 includes a recognition / planning unit 100, a gripping force calculation unit 200, and a control unit 300. The recognition / planning unit 100 includes a recognition unit 102, a command unit 104, a gripping position determination unit 106, and an operation planning unit 108. The gripping force calculation unit 200 includes a pressure acquisition unit 202, a contact detection unit 204, a pressure center position calculation unit 206, a pressure center movement amount calculation unit 208, an overall slip detection unit 210, a sticking rate calculation unit 212, and a gripping force control unit 214. have. The control unit 300 has an overall control unit 302 and a hand control unit 304.
[0052]
　The recognition / planning unit 100 recognizes the object 600 held by the robot and makes a plan for holding the object 600. The recognition unit 102 is composed of a camera, a ToF sensor, and the like, and recognizes the three-dimensional shape of the object 600. A command from the user is input to the command unit 104. The gripping position determination unit 106 uses the recognition result of the object by the recognition unit 102 and determines the position where the robot grips the object 600 based on the user's command input to the command unit 104. The motion planning unit 108 plans the operation of the robot arm 506 and the operation of the hand 500 provided at the tip of the arm 506 based on the gripping position determined by the gripping position determining unit 106.
[0053]
　The gripping force calculation unit 200 calculates the gripping force of the hand 500 that grips the object 600, and controls the gripping force. The pressure acquisition unit 202 acquires the pressure detected by the distributed pressure sensors 530 and 532. The contact detection unit 204 detects the contact between the first and second flexible layers 540 and 542 and the object 600 by using the distributed pressure value acquired by the pressure acquisition unit 202. For example, the contact detection unit 204 detects the contact between the first and second flexible layers 540, 542 and the object 600 when the distributed pressure value is equal to or higher than a predetermined value. The pressure center position calculation unit 206 uses the distributed pressure value acquired by the pressure acquisition unit 202 to obtain the pressure center position X from the above equation (1) in each region of the first and second flexible layers 540 and 542. Calculate COP .
[0054]
　The pressure center movement amount calculation unit 208 calculates the movement amount of the pressure center position in each region of the first and second flexible layers 540 and 542 by using the pressure center position calculated by the pressure center position calculation unit 206. To do. The pressure center movement amount calculation unit 208 calculates the movement amount ΔX COP of the pressure center position from the following equation (2) . The right side of the equation (2) shows the difference between the pressure center position X COP at time t + 1 and the pressure center position X COP at time t .
[0055]
[Number 2]

[0056]
　The overall slip detection unit 210 detects whether or not there is a change in the movement of the pressure center position in a preset time window by using the movement amount of the pressure center position calculated by the pressure center movement amount calculation unit 208. The time window is a predetermined time. If the pressure center position does not move during this predetermined time, the overall slip detection unit 210 detects that the pressure center position does not change and that the overall slip has occurred. The overall slip detection unit 210 detects that overall slip occurs in each region by monitoring the change in the pressure center position in each region of the two distributed pressure sensors 530 and 532.
[0057]
　The sticking rate calculation unit 212 calculates the ratio of the undetected region of the total slip to the entire region including the first flexible layer 540 and the second flexible layer 542, and sets the calculated ratio as the sticking ratio. As described above, when the flexible layer is divided into two, the adhesion rate is calculated as three kinds of values ​​of 0%, 50%, and 100%.
[0058]
　The gripping force control unit 214 determines the gripping force so that the sticking rate becomes a constant value. The gripping force control unit 214 controls the gripping force by feedback control so that the sticking rate calculated by the sticking rate calculation unit 212 becomes a predetermined value. As an example, the gripping force control unit 214 controls the gripping force so that the sticking rate is 50%.
[0059]
　The control unit 300 controls the operation of the robot. The overall control unit 302 controls the arm 506 of the robot based on the motion plan planned by the motion planning unit 108. The hand control unit 304 controls the hand 500 based on the control of the gripping force control unit 214. The gripping force control unit 214 and the hand control unit 304 may be integrally configured.
[0060]
　Each component of the recognition / planning unit 100, the gripping force calculation unit 200, and the control unit 300 of the control system 1000 shown in FIG. 6 functions as a circuit (hardware) or a central processing unit such as a CPU. It can be composed of a program (software) for. This program can be stored in a recording medium such as a memory included in the control system 1000 or a memory connected to the control system 1000 from the outside. The same applies to FIGS. 7, 9, and 20 described later.
[0061]
　5. Modifications
　of the present embodiment In the following, some modifications of the present embodiment will be described.
[0062]
　　5.1. Modification 1 (Example of adjusting the gripping force control gain according to the rigidity of the object) In the
　modification 1, the position of the hand 500 at the moment when the flexible layer comes into contact with the object 600 and the flexible layer is pushed into the object 600, or The physical information (rigidity) of the object 600 is calculated from the information on the contact area and the contact force between the flexible layer and the object 600. Then, the gripping force control gain is adjusted based on the physical information of the object 600. Here, the gripping force control gain is an increase rate when the gripping force is increased so that the sticking rate becomes a certain value.
[0063]
　FIG. 7 is a schematic view showing the configuration of the gripping force calculation unit 200 according to the first modification. As shown in FIG. 7, in addition to the configuration shown in FIG. 6, the gripping force calculation unit 200 according to the modification 1 includes a contact force calculation unit 216, a contact node number acquisition unit (contact radius calculation unit) 218, and physical information. It has a calculation unit 220.
[0064]
　The contact force calculation unit 216 calculates the contact force when the object 600 comes into contact with the first and second flexible layers 540 and 542. The contact force can be obtained by multiplying the number of contact nodes among all the nodes of the distributed pressure sensors 530 and 532 by the force (pressure) applied to each contact node. The contact node is a node of the distributed pressure sensors 530, 532 that are in contact with the object 600 via the first flexible layer 540 or the second flexible layer 542. In other words, the contact node is the node from which the pressure detection value is obtained (the detection value is not 0).
[0065]
　The contact node number acquisition unit 218 acquires the number of contact nodes based on the contact between the first and second flexible layers 540 and 542 detected by the contact detection unit 204 and the object 600. The number of contact nodes corresponds to the contact area. The physical information calculation unit 220 calculates the rigidity as physical information of the object 600 from the contact area information obtained from the contact node number acquisition unit 218 and the contact force information obtained from the contact force calculation unit 216.
[0066]
　Further, the rigidity can be calculated as physical information of the object 600 from the contact radius a when the object 600 is in contact with the flexible layer. In this case, the contact radius calculation unit is made to function instead of the contact node number acquisition unit 218. From Hertz's contact theory, the contact radius a between the robot finger (first finger 502, second finger 504) and the object can be expressed by the following equation (3).
[0067]
[Number 3]

[0068]
　In addition, r is the radius of the robot finger, and E * is the effective elastic modulus.
　Further, as shown in the following equation (4), the effective elastic modulus E * is given by the elastic moduli Ef and Eo of the robot finger and the object and the Poisson's ratios vf and vo, respectively.
[0069]
[Number 4]

[0070]
　Since the Poisson's ratio is a value of about 0.5 at most and is usually a smaller value , the Poisson's ratio is ignored as shown in the equation (5), assuming that the squared value does not significantly affect E *. can do.
[0071]
[Number 5]

[0072]
　Therefore, since the robot finger radius r and the Young's modulus Ef of the robot finger are known, the physics of the object 600 is calculated from the equation (3) based on the information of the contact radius a and the contact force Fn calculated by the contact radius calculation unit 218. Information (Young's modulus Eo) can be calculated.
[0073]
　The rigidity of the object 600 as physical information is sent to the gripping force control unit 214. The gripping force control unit 214 adjusts the gripping force control gain based on the physical information. As described above, the gripping force control gain is an increase rate when the gripping force is increased so that the sticking ratio becomes a certain value. When the rigidity of the object 600 is high, the gripping force control unit 214 has a relatively low probability of deforming or breaking the object 600. Therefore, when controlling the sticking rate to a target value, the gripping force control unit 214 increases the rate of increase in the gripping force. To do. On the other hand, when the rigidity of the object 600 is low, the gripping force control unit 214 has a relatively high probability that the object 600 will be deformed or broken. Therefore, when controlling the sticking rate to a target value, the gripping force control unit 214 increases the gripping force. To lower.
[0074]
　According to the first modification, by detecting the above-mentioned partial slip state, the gripping force can be controlled with the minimum necessary force so that the object 600 does not slip, and the gripping force is adjusted according to the hardness of the object 600. The rate of increase can be controlled. Therefore, it is possible to reliably suppress the deformation and destruction of the object 600 when gripping.
[0075]
　When determining the rigidity of the object 600, the rigidity is determined from the relationship between the position (pushing amount) of the hand 500 when the flexible layer is pushed into the object 600 and the contact force obtained from the distributed pressure sensors 530 and 532. You may ask for.
[0076]
　　5.2. Modification 2 (Example of controlling the position and posture of a finger in order to increase the difference in the timing of occurrence of overall slip) In
　Modification 2, the overall slip in each of the first flexible layer 540 and the second flexible layer 542. The position and posture of the fingers are controlled in order to increase the difference in the timing of occurrence of. Here, the steeper the pressure distribution when the flexible layers 540 and 542 come into contact with the object 600, the later the timing of occurrence of overall slippage. 8A to 8C are characteristic diagrams showing an example of the pressure distribution when the flexible layers 540 and 542 come into contact with the object 600. The pressure distribution becomes steeper in the order of FIGS. 8C, 8B, and 8A. Here, the steep pressure distribution means that the pressure gradient at the end of the region where the flexible layers 540 and 542 are in contact with the object 600 (region A1 shown in FIGS. 8A to 8C) is large. ..
[0077]
　In each of FIGS. 8A to 8C, the region where the pressure is high is the region where the flexible layers 540 and 542 and the object 600 are in contact with each other. At the end of the region where the flexible layers 540, 542 and the object 600 are in contact, there is a gradient in pressure. The larger the pressure gradient, the later the timing at which the overall slip occurs.
[0078]
　For example, in the region where the flexible layers 540 and 542 are in contact with the object 600, the shape of the object 600 is a convex surface, and the smaller the radius of curvature of the convex surface, the steeper the pressure gradient and the later the timing at which the entire slip occurs. Become.
[0079]
　In the second modification, the hand 500 is arranged so that the first flexible layer 540 and the second flexible layer 542 are arranged at positions where the pressure gradients are different in the first flexible layer 540 and the second flexible layer 542, respectively. To control.
[0080]
　FIG. 9 is a schematic view showing the configuration of the gripping force calculation unit 200 according to the modified example 2. As shown in FIG. 9, the gripping force calculation unit 200 according to the modified example 2 has a pressure gradient calculation unit 222 and an actuator control unit 224 in addition to the configuration shown in FIG.
[0081]
　When the contact detection unit 204 detects the contact between the first and second flexible layers 540, 542 and the object 600, the pressure gradient calculation unit 222 is based on the pressure detected by the distributed pressure sensors 530, 532. The characteristics shown in 8A to 8C are acquired. Then, the pressure gradient calculation unit 222 calculates the pressure gradient of the region A1 shown in FIGS. 8A to 8C. The pressure detected by the distributed pressure sensors 530 and 532 is acquired by the pressure acquisition unit 202 and sent to the pressure gradient calculation unit 222.
[0082]
　The actuator control unit 224 controls the actuator that controls the hand 500 or the arm 506 based on the pressure gradient calculated by the pressure gradient calculation unit 222. The actuator control unit 224 controls the actuator so as to grip the object 600 at a position where the difference in pressure gradient becomes larger at the contact portion between the first flexible layer 540 and the second flexible layer 542 and the object 600.
[0083]
　FIG. 10 is a schematic view showing specific control of the hand 500. In FIG. 10, the state before controlling the position and posture of the finger according to the pressure gradient is shown in the left figure, and the state after controlling the position and posture of the finger according to the pressure gradient is shown in the right figure. In FIG. 10, the first finger 502 is provided with the first flexible layer 540 and the distributed pressure sensor 530, and the second finger 504 is provided with the second flexible layer 542 and the distributed pressure sensor 532. Therefore, FIG. 10 shows an example in which the first flexible layer 540 and the second flexible layer 542 are not divided in the sliding direction of the object 600.
[0084]
　Before controlling the position and posture of the finger, the contact surface of 600 of the object is a curved surface with respect to both the first flexible layer 540 and the second flexible layer 542. On the other hand, after controlling the position and posture of the finger, the contact surface of the object 600 with respect to the second flexible layer 542 is a curved surface, but the contact surface of the object 600 with respect to the first flexible layer 540. The surface is flat.
[0085]
　As described above, since the friction coefficient of the first flexible layer 540 is smaller than the friction coefficient of the second flexible layer 542, the timing of occurrence of total slippage is later in the second flexible layer 542. In addition to this, after controlling the position and posture of the finger, the contact surface of the object 600 with respect to the first flexible layer 540 becomes a flat surface, and the contact surface of the object 600 with respect to the second flexible layer 542. The surface is a curved surface. Therefore, the pressure distribution in the second flexible layer 542 becomes steeper than the pressure distribution in the first flexible layer 540, and the timing of occurrence of total slip in the second flexible layer 542 becomes later. Therefore, the difference in the timing of occurrence of overall slippage can be made larger between the first flexible layer 540 and the second flexible layer 542.
[0086]
　In the above-mentioned example, the position and posture of the fingers of the hand 500 are controlled according to the pressure gradient, but the three-dimensional information of the object 600 obtained by observing the shape of the object 600 by the recognition unit 102 can be used. Based on this, the position and posture of the fingers of the hand 500 may be controlled. Also in this case, based on the three-dimensional information, it is possible to bring the first flexible layer 540 into contact with a portion having a small pressure gradient and to bring the second flexible layer 542 into contact with a portion having a large pressure gradient.
[0087]
　　5.3. Modification 3 (Variation of arrangement of flexible layer and distributed pressure sensor)
　As described above, it is more preferable that the first flexible layer 540 and the second flexible layer 542 are divided in the direction in which the object 600 slides. Is. In the third modification, when a plurality of different sliding directions of the object 600 are assumed according to the postures of the hand 500 and the arm 506, the division is performed independently of the sliding directions.
[0088]
　FIG. 11 is a schematic view showing an example of division of the flexible layer. In the example shown in FIG. 11, the Young's modulus is different for each of the plurality of flexible layers 544, 546, 548, 550. Each of the flexible layers 544, 546, 548, 550 is divided by a circular boundary, and the Young's modulus is smaller toward the center. According to such a division method, it is possible to deal with sliding in multiple directions as shown by a plurality of arrows in FIG. 11, and the flexible layer is divided in the sliding direction in any of the sliding directions. It will be.
[0089]
　Further, as described above, regarding the parameters of the flexible layer that cause the overall slip occurrence timing to be different, the difference in the timing at which the overall slip occurs by combining the conditions where the overall slip occurs earlier or the conditions where the overall slip occurs later. Can be further expanded.
[0090]
　Therefore, in the example shown in FIG. 11, the conditions of the thickness and the Young's modulus are combined so that the flexible layer closer to the center has a larger thickness and a smaller Young's modulus. The thickness and Young's modulus were set. As a result, the larger the Young's modulus, the earlier the overall slip occurrence timing, and the smaller the thickness, the earlier the overall slip occurrence timing. Therefore, the peripheral flexible layer shown in FIG. 11 has an earlier overall slip occurrence timing. Therefore, the closer the flexible layer is to the center, the later the timing of occurrence of overall slippage.
[0091]
　Further, regarding the parameters that cause the overall slip occurrence timing to be different, by combining the thickness and Young's modulus, the flexible layer 550 having a large Young's modulus in the periphery is used for the heavy object 600 that requires a large gripping force, but the large gripping force Only the flexible layer 544 with a small Young's modulus in the center is used for gripping the light and fragile object 600 that does not require the above. Therefore, it is possible to use an appropriate flexible layer depending on the gripped object.
[0092]
　Further, FIG. 12 is a plan view showing an example of division that does not depend on the position of contact with the object 600. In the example shown in FIG. 12, a flexible layer 552 having a small Young's modulus and a flexible layer 554 having a large Young's modulus are used. As shown in the plan view, the flexible layers 552 and the flexible layers 554 are alternately arranged in a staggered pattern. Although two flexible layers with different Young's moduluss 552 and 554 are shown in FIG. 12, three or more flexible layers having different Young's moduluss may be arranged. In this case, by arranging the flexible layers adjacent to each other in the row direction or the column direction so as to have different Young's modulus in the figure, even when three or more flexible layers having different Young's modulus are arranged, the arrangement is staggered. It can be performed. Further, when two flexible layers 552 and 554 having different Young's moduluss are arranged, they are arranged so that the Young's modulus of the flexible layers in the same column or the same row is the same in FIG. Young's modulus may be different.
[0093]
　Further, as described above, as the number of divisions of the region increases, the resolution of the fixing rate increases and the accuracy of gripping force control increases. Further, as the number of divisions of the region increases, the sticking rate can be detected even for a smaller object or an object having unevenness. However, the number of divisions depends on the pitch width of the nodes of the distributed pressure sensor, and when an attempt is made to increase the number of divisions, there is a hardware limit. Therefore, as shown in FIG. 13, in order to pseudo-reduce the pitch width of the nodes of the two distributed pressure sensors 560 and 562, the distributed pressure sensors 560 and 562 are staggered and arranged in layers.
[0094]
　In the example shown in FIG. 13, when two distributed pressure sensors 560 and 562 are stacked by shifting them in the node arrangement direction (x direction) in order from the top (example (a)), the node arrangement direction (x direction, Three methods are shown: a case where the pressure sensors 562 are shifted in the y direction) and stacked (example (b)), and a case where one of the distributed pressure sensors 562 is rotated by 45 ° and stacked (example (c)).
[0095]
　In the example (a) shown in FIG. 13, the distributed pressure sensors 560 and 562 are arranged so as to be offset from each other in the x direction by 1/2 of the width of the node. In the example (b) shown in FIG. 13, the distributed pressure sensors 560 and 562 are arranged so as to be offset from each other in the x direction and the y direction by 1/2 of the width of the node. The arrangement and stacking method of the distributed pressure sensors are not limited to the example of FIG.
[0096]
　As described above, by shifting the distributed pressure sensors 560 and 562 and arranging them in an overlapping manner, the pitch width of the nodes can be pseudo-reduced, and the number of divisions of the region can be increased.
[0097]
　　5.4. Modification 4 (Example of arranging flexible layers above and below the distributed pressure sensor)
　As described above, as a method of delaying the occurrence timing of overall slip, there is a method of increasing the thickness of the flexible layer. On the other hand, if the thickness of the flexible layer is increased, the sensitivity of the distributed pressure sensor is lowered.
[0098]
　In the modified example 4, as shown in FIG. 14, a part of the thickness of the flexible layer 570 above the distributed pressure sensor 564 is arranged below the distributed pressure sensor 564, so that the distributed pressure sensor 564 is placed above and below the distributed pressure sensor 564. The flexible layers 527 and 574 are arranged. The movement of the pressure center position depends on the total thickness of the flexible layers 527 and 574, while the detection sensitivity depends on the flexible layer 574 above the distributed pressure sensor 564, which reduces the detection sensitivity of the distributed pressure sensor 564. It is possible to delay the occurrence timing of the entire slip without doing so.
[0099]
　FIG. 15A is a schematic view showing the detection sensitivity by the distributed pressure sensor 564 according to the difference in the thickness of the flexible layer. In FIG. 15A, the detection sensitivity when three types of flexible layers 570 having different thicknesses are arranged on the distributed pressure sensor 564 (examples (a) to (c)) and the flexibility up and down of the distributed pressure sensor 564. The detection sensitivity when the layers 527 and 574 are arranged (modification example 4) is shown.
[0100]
　The example (a) shown in FIG. 15A shows an example in which a flexible layer 570 having a thickness of 1 mm is arranged on the distributed pressure sensor 564. Further, the example (b) shown in FIG. 15A shows an example in which the flexible layer 570 having a thickness of 3 mm is arranged on the distributed pressure sensor 564, and the example (c) shows an example in which the flexible layer 570 having a thickness of 3 mm is arranged on the distributed pressure sensor 564. An example in which a flexible layer 570 having a thickness of 5 mm is arranged is shown.
[0101]
　Further, in the modified example 4 shown in FIG. 15A, a flexible layer 574 having a thickness of 1 mm is arranged on the distributed pressure sensor 564, and a flexible layer 572 having a thickness of 4 mm is arranged under the distributed pressure sensor 564. Shown.
[0102]
　Further, FIG. 15A shows the pressure detected by the distributed pressure sensor 564 and the standard deviation thereof for Examples (a) to (c) and Modified Example 4. As shown in Examples (a) to (c), the larger the thickness of the flexible layer 570 on the distributed pressure sensor 564, the larger the standard deviation of the pressure detection value, and the larger the detection sensitivity of the distributed pressure sensor 564. It can be seen that is decreasing.
[0103]
　On the other hand, as shown in FIG. 15A, in the modified example 4, the total thickness of the flexible layer 572 and the flexible layer 574 is the same as that of the example (c), but the thickness of the flexible layer 574 on the distributed pressure sensor 564 is the same. Since the diameter is 1 mm, the standard deviation of the detected pressure value is suppressed. Therefore, according to the modified example 4, at least the same detection sensitivity as in the example (a) can be obtained.
[0104]
　FIG. 15B is a characteristic diagram showing how the pressure center position changes in the examples (a) to (c) and the modified example 4 shown in FIG. 15A as in FIG. 3B. As shown in Examples (a) to (c) of FIG. 15B, the larger the thickness of the flexible layer 570 on the distributed pressure sensor 564, the later the timing of occurrence of overall slippage.
[0105]
　Further, as shown in FIG. 15B, in the modified example 4, the overall slip occurrence timing is the same as that in the example (c). Therefore, according to the modified example 4, the thickness of the flexible layer 574 on the distributed pressure sensor 564 is made equal to that of the flexible layer 570 of the example (a), thereby ensuring the same detection sensitivity as the example (a). it can. Further, according to the modified example 4, the total thickness of the upper and lower flexible layers 527 and 574 of the distributed pressure sensor 564 is made equal to that of the flexible layer 570 of the example (c), so that the overall slip occurrence timing is an example. It can be made equivalent to (c).
[0106]
　In the characteristic of the modified example 4 shown in FIG. 15B, the moving direction of the pressure center position is opposite to that of the examples (a) to (c). This is due to the influence of the vertical pressure applied to the distributed pressure sensor 564.
[0107]
　16A and 16B are schematic views showing the reason why the moving direction of the pressure center position of the modified example 4 is opposite to that of the examples (a) to (c). 16A and 16B each show a state in which the object 600 is not gripped (left figure) and a state in which the object 600 is gripped (right figure), and each flexible layer is formed by gripping the object 600. Shows how it deforms. Further, in FIGS. 16A and 16B, the arrow A11 indicates the direction in which the object 600 tends to slide. For convenience of explanation, the illustration of the object 600 is omitted in FIGS. 16A and 16B.
[0108]
　FIG. 16A shows an example (a), showing how the flexible layer 570 is deformed in the direction of arrow A11 when the object 600 tries to slide in the direction of arrow A11. At this time, as for the pressure detection value of the distributed pressure sensor 564, since the flexible layer 570 does not exist in the region A3 shown in FIG. 16A and the pressure becomes small, the pressure becomes small at the left end of the distributed pressure sensor 564 and becomes small at the right end. The pressure increases. Therefore, as shown in FIG. 15B, the pressure center position moves in the positive direction of the x-axis.
[0109]
　On the other hand, FIG. 16B shows a modification 4, showing that the flexible layer 572 and the flexible layer 574 are deformed in the direction of the arrow A11 when the object 600 tries to slide in the direction of the arrow A11. At this time, as for the pressure detection value of the distributed pressure sensor 564, since the flexible layer 572 does not exist in the region A4 shown in FIG. 16B and the pressure becomes small, the pressure becomes small at the right end of the distributed pressure sensor 564 and becomes small at the left end. The pressure increases. Therefore, as shown in FIG. 15B, the pressure center position moves in the negative direction of the x-axis. In FIG. 15B, the moving direction of the pressure center position is different from the examples (a) to (c) of FIGS. 3B and 15A, but a method of determining that overall slippage occurs when the movement of the pressure center position is stopped. It is the same as the examples (a) to (c) of FIGS. 3B and 15A.
[0110]
　　5.5. Deformation example 5 (Method of changing the frictional force of the
　flexible layer ) As a method of changing the friction coefficient of the flexible layer, in addition to the method of changing the material, fine processing is applied to the surface of the flexible layer, and the surface of the flexible layer is coated. There is a way to do it. This makes it possible to generate various friction coefficient distributions even when a plurality of flexible layers are made of the same material.
[0111]
　Further, in order to change the frictional force of the flexible layer, a method of changing the surface area of ​​the region of each flexible layer can be mentioned. The larger the surface area of ​​the flexible layer, the greater the frictional force. FIG. 17 is a schematic view showing an example in which the surface area of ​​the flexible layer is changed. Since the area of ​​the central flexible layer 580 is larger than the area of ​​the surrounding flexible layer 582, the frictional force of the central flexible layer 580 is larger than the frictional force of the surrounding flexible layer 582. Therefore, the timing of occurrence of overall slippage can be different between the flexible layer 580 and the flexible layer 582.
[0112]
　Further, FIG. 18 shows an example in which the flexible layers 527 and 574 are arranged above and below the distributed pressure sensor 564 as in the modified example 4 of FIG. 14, and the flexible layer 574 above the distributed pressure sensor 564 is shown. In order to change the frictional force, a method of changing the hardness of the flexible layer 572 under the distributed pressure sensor 564 is shown. In FIG. 18, the hardness of the flexible layer 572 below the distributed pressure sensor 564 is set to three types: low (flexible layer 572a), medium (flexible layer 572b), and high (flexible layer 572c).
[0113]
　When the displacement amount of the flexible layers 572a to 572c is constant, the reaction force Fn generated in each flexible layer 572a to 572c is different, and the distribution of the frictional force Ft (= Fn × μ · Fn) generated in each flexible layer 574a to 574c. Can be generated. This also facilitates the replacement of the flexible layer on the surface.
[0114]
　　5.6. Modification 6 (Example using a linear flexible layer) In the
　modification 6, instead of dividing the flexible layer, a linear flexible layer is arranged. FIG. 19 is a schematic view showing a configuration example using the linear flexible layers 590 and 592 according to the modified example 6. As shown in FIG. 19, linear flexible layers 590 and 592 are provided on the distributed pressure sensors 594 and 596. The flexible layer 590 is provided on the distributed pressure sensor 594, and the flexible layer 592 is provided on the distributed pressure sensor 596. The flexible layer 592 is made of a material having a higher coefficient of friction than the flexible layer 590.
[0115]
　FIG. 19 shows a state in which the object 600 is gripped by using the flexible layers 590 and 592. Therefore, the object 600 is in contact with the flexible layers 590 and 592.
[0116]
　Further, FIG. 19 shows time-series movements at time t1, time t2, and time t3. Time t1 indicates a fixed state. In this state, the tips of the flexible layers 590 and 592 point uniformly to the left.
[0117]
　Next, the time t2 indicates a state in which the object 600 is partially slipped. In this state, the tip of the flexible layer 590 having a small coefficient of friction faces to the right. On the other hand, the tip of the flexible layer 592 having a large coefficient of friction is maintained in a state of facing left. When a partial slip occurs, when the orientation of the flexible layer 590, which is a slip region portion, changes, the pressure in that region decreases, and the change can be detected by the distributed pressure sensor 594.
[0118]
　Next, the time t3 indicates a state in which the object 600 is totally slipped. In this state, the object 600 slides to the right, and the tips of the flexible layers 590 and 592 point uniformly to the right.
[0119]
　FIG. 20 is a schematic view showing the configuration of the gripping force calculation unit 200 using the linear flexible layers 590 and 592 shown in FIG. The pressure acquisition unit 202 acquires the pressure detected by the distributed pressure sensor 594,596. The overall slip detection unit 210 monitors the change in pressure in the linear flexible layers 590 and 592, and detects slip in each flexible layer 590 and 592. As described above, at time t2 in FIG. 19, only the pressure detection value of the distributed pressure sensor 594 corresponding to the flexible layer 590 decreases, so that the state of partial slip can be detected. The sticking rate calculation unit 212 calculates the ratio of the undetected region of partial slippage in the total region. The gripping force control unit 214 determines the gripping force so that the sticking rate calculated by the sticking rate calculation unit 212 becomes a constant value.
[0120]
　When the pressure detection value decreases, the entire slip can be detected. The pressure change in each of the linear flexible layers 590 and 592 acquired by the pressure acquisition unit 202 is monitored, and when the pressure exceeds a certain threshold value, the total slip is detected. The sticking rate calculation unit 212 calculates the ratio of the entire non-slip detection region to the total region. In the example of FIG. 19, the entire slip occurs only in the flexible layer 590 at time t2, and the flexible layer 592 is in a fixed state. Since the total number of the flexible layers 590 and 592 is 7, and the number of the flexible layers 592 is 4, the adhesion rate is 57% (= (4/7) × 100).
[0121]
　FIG. 21 is a schematic view showing an example in which the directions of the flexible layers 590 and 592 are arranged in multiple directions in the configuration using the linear flexible layers 590 and 592 shown in FIG. As shown in the state at time t1 in FIG. 19, the tips of the flexible layers 590 and 592 are oriented in the same direction in the fixed state. As shown in FIG. 21, by arranging the directions of the flexible layers 590 and 592 in the fixed state in multiple directions, it is possible to configure the structure so that it can correspond to a large number of sliding directions.
[0122]
　As described above, according to the present embodiment, it is possible to detect partial slippage of an object based only on a simple configuration and arithmetic processing, and it is possible to optimally control the gripping force of the object. In addition, by making the timing of occurrence of total slip in a plurality of flexible layers different, partial slip can be detected accurately even under various conditions such as a hard object, an object having a flat surface, and a flat pressure distribution. It becomes possible.
[0123]
　Although the preferred embodiments of the present disclosure have been described in detail with reference to the accompanying drawings, the technical scope of the present disclosure is not limited to such examples. It is clear that anyone with ordinary knowledge in the technical field of the present disclosure may come up with various modifications or modifications within the scope of the technical ideas set forth in the claims. Of course, it is understood that it belongs to the technical scope of the present disclosure.
[0124]
　For example, in the above embodiment, an example in which the flexible layer and the distributed pressure sensor are arranged on the hand 500 that grips the object 600 is shown, but the present technology is not limited to such an example. For example, a flexible layer and a distributed pressure sensor may be placed on the ground contact surface of the toes of a walking robot to detect slippage of the toes. As described above, the present embodiment can be widely applied in detecting slippage.
[0125]
　In addition, the effects described herein are merely explanatory or exemplary and are not limited. That is, the techniques according to the present disclosure may exhibit other effects apparent to those skilled in the art from the description herein, in addition to or in place of the above effects.
[0126]
　The following configurations also belong to the technical scope of the present disclosure.
(1) Overall slip detection that detects the state of overall slip in which the object is sliding in each of the plurality of said regions based on pressure information obtained from a plurality of regions having different sliding characteristics when the contacted object slides. A control device including a unit.
(2) The overall slip detection unit detects the overall slip state when the pressure center position becomes a constant value based on the change in the pressure center position in each of the plurality of the regions. The control device according to 1).
(3) The control device according to (2) above, wherein the overall slip detection unit detects the state of the overall slip when the pressure center position reaches the constant value within a predetermined time.
(4) The control device according to any one of (1) to (3) above, wherein the plurality of said regions have different slip characteristics, so that the timing at which the overall slip occurs is different.
(5) When the state of total slip is detected in at least one of the regions and the state of total slip is not detected in the other regions, the object partially slides with respect to the plurality of regions. The control device according to any one of (1) to (4) above, which determines that the state of partial slippage is occurring.
(6) The plurality of regions are determined to be in the partial slip state based on the difference in the timing at which the total slip occurs due to the difference in the slip characteristics. The control device according to 5).
(7) When the state of the overall slip is not detected in any of the plurality of the regions, it is determined that the object and the plurality of the regions are in a fixed state. The control device according to any one of 6).
(8) The above (1), which comprises a sticking rate calculation unit which calculates the sticking rate between the object and the area based on the ratio of the area where the state of the whole slip is not detected to all the areas. The control device according to any one of (7).
(9) The above (8), which includes a gripping force control unit that controls a gripping force when gripping the object by using a gripping unit that grips the object provided with a plurality of the regions, based on the sticking rate. ).
(10) The control device according to (9), wherein the gripping force control unit controls the gripping force so that the sticking rate becomes a predetermined value.
(11)
　The control device according to (9), further comprising an acquisition unit that acquires the rigidity of the object, and the gripping force control unit controls the gripping force based on the rigidity.
(12) The control device according to (11) above, wherein the gripping force control unit reduces the rate of increase when the gripping force is increased as the rigidity of the object is lower.
(13)
　A control unit including a pressure gradient calculation unit that calculates a pressure gradient when the object comes into contact with a plurality of the objects, and controls positions where the plurality of the regions come into contact with the object based on the pressure gradient. The control device according to any one of (1) to (12) above.
(14) The control unit brings the region of 1 into contact with the first portion of the object having a low pressure gradient, and at a second location where the pressure gradient of the object is higher than that of the first portion. The control device according to (13), wherein the regions of the above are brought into contact with each other.
(15) The control device according to any one of (1) to (14) above, wherein the plurality of the regions are arranged side by side in the sliding direction of the object.
(16) The control according to any one of (1) to (15) above, wherein the plurality of regions differ in at least one of a friction coefficient, Young's modulus, Poisson's ratio, thickness, and curvature as the slip characteristics. apparatus.
(17) Based on pressure information obtained from a plurality of regions having different sliding characteristics when a contacted object slides, it is provided to detect a state of total slippage in which the object is sliding in each of the plurality of the regions. , Control method.
(18) As a means for detecting the state of overall slippage in which the object is sliding in each of the plurality of said regions based on pressure information obtained from a plurality of regions having different sliding characteristics when the contacted object slides. A program that makes a computer work.
Description of the sign
[0127]
　210 Overall slip detection unit
　212 Sticking rate calculation unit
　214 Gripping force control unit
　222 Pressure gradient calculation unit
　224 Actuator control unit
The scope of the claims
[Claim 1]
　Based on pressure information obtained from a plurality of regions having different sliding characteristics when a contacted object slides, an overall slip detecting unit for detecting a state of overall slip in which the object is sliding is provided in each of the plurality of said regions. ,Control device.
[Claim 2]
　The first aspect of the present invention, wherein the overall slip detection unit detects the overall slip state when the pressure center position becomes a constant value based on a change in the pressure center position in each of the plurality of the regions. Control device.
[Claim 3]
　The control device according to claim 2, wherein the overall slip detection unit detects the state of the overall slip when the pressure center position reaches the constant value within a predetermined time.
[Claim 4]
　The control device according to claim 1, wherein the plurality of said regions have different slip characteristics, so that the timing at which the overall slip occurs is different.
[Claim 5]
　A portion where the object is partially sliding with respect to a plurality of the regions when the overall slip condition is detected in at least one of the regions and the overall slip condition is not detected in the other regions. The control device according to claim 1, wherein the control device is determined to be in a slipping state.
[Claim 6]
　The fifth aspect of the present invention, wherein the plurality of said regions have different slip characteristics, so that a difference occurs in the timing at which the entire slip occurs, and the partial slip state is determined based on the difference in the timing. Control device.
[Claim 7]
　The control device according to claim 1, wherein when the overall slip state is not detected in any of the plurality of the regions, it is determined that the object and the plurality of the regions are in a fixed state.
[Claim 8]
　The control according to claim 1, further comprising a sticking rate calculation unit that calculates a sticking rate between the object and the area based on the ratio of the area in which the state of total slippage is not detected to all the areas. apparatus.
[Claim 9]
　The eighth aspect of the present invention, wherein the gripping force control unit for controlling the gripping force when gripping the object by using the gripping portion for gripping the object provided with the plurality of regions is provided based on the sticking rate. Control device.
[Claim 10]
　The control device according to claim 9, wherein the gripping force control unit controls the gripping force so that the sticking rate becomes a predetermined value.
[Claim 11]

　The control device according to claim 9 　, further comprising an acquisition unit for acquiring the rigidity of the object, wherein the gripping force control unit controls the gripping force based on the rigidity.
[Claim 12]
　The control device according to claim 11, wherein the gripping force control unit reduces the rate of increase when the gripping force is increased as the rigidity of the object is lower.
[Claim 13]
　A pressure gradient calculation unit for calculating a pressure gradient when the
　object contacts a plurality of the objects is provided, and a control unit for controlling a position where the plurality of the regions come into contact with the object based on the pressure gradient is provided. The control device according to claim 1.
[Claim 14]
　The control unit brings the region of 1 into contact with the first location where the pressure gradient of the object is low, and the other region is located at a second location where the pressure gradient of the object is higher than the first location. 13. The control device according to claim 13.
[Claim 15]
　The control device according to claim 1, wherein the plurality of regions are arranged side by side in the sliding direction of the object.
[Claim 16]
　The control device according to claim 1, wherein the plurality of regions differ in at least one of a friction coefficient, Young's modulus, Poisson's ratio, thickness, and curvature as the slip characteristics.
[Claim 17]
　A control method comprising detecting a state of total slippage in which the object is sliding in each of the plurality of said regions based on pressure information obtained from a plurality of regions having different sliding characteristics when the contacted object slides. ..
[Claim 18]
　The computer functions as a means for detecting the state of total slippage of the object in each of the plurality of regions based on the pressure information obtained from the plurality of regions having different sliding characteristics when the contacted object slides. Program to make you.