Gripper device control device and gripper system

Abstract

Even when the backdrivability of a gripper device is low, the force applied to an object to be gripped is appropriately controlled without relying on sensing of a gripping mechanism. [Solution] The control device 20 of the gripper device 10 is equipped with a gripping mechanism 11 that grips an object to be gripped 2, a motor 12 that drives the gripping mechanism 11, and an angle sensor 13 for the motor 12, and is equipped with: a disturbance estimation unit 21 that estimates a disturbance torque acting on the gripper device 10 based on the output signal of the angle sensor 13 and the current signal output to the motor 12; a reaction force estimation unit 22 that determines the friction state inside the gripping mechanism 11 based on the estimated disturbance torque 72 of the disturbance estimation unit 21 and the motor angular velocity, and estimates a reaction force acting on the gripping mechanism 11 in accordance with the determined friction state; and a force controller 23 that acquires a target contact force 61 and outputs a control command value 63 that brings the reaction force 62 estimated by the reaction force estimation unit 22 closer to the target contact force 61.

Description

[Technical Field]

[0001] The present invention relates to a control device for a gripper device and a gripper system. [Background technology]

[0002] For example, Patent Document 1 discloses an electric hand mechanism including a plurality of L-shaped fingers with claws attached, nuts to which the base ends of the fingers are swingably joined via link pins, an opening / closing pin that swingably supports the bent portions of the fingers, and a motor having a rotating shaft formed with a threaded shaft onto which the nut screws. When the motor rotates the rotating shaft, the nut moves along the threaded shaft, and the fingers swing around the opening / closing pin in conjunction with the movement of the nut. The electric hand mechanism swings the plurality of fingers to open and close the claws of each finger, thereby gripping and releasing an object to be gripped. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2024-7159 Summary of the Invention [Problem to be solved by the invention]

[0004] When controlling the position of an actuator including a gripper device, a constant position command corresponding to the size of the object to be gripped is given to perform the gripping operation. However, objects to be gripped vary in size, and some objects are easily deformed. As a result, a constant position command may fail to grip the object. Increasing the gripping position command to increase gripping reliability may result in excessive force being applied to the object, potentially causing damage. For this reason, it is desirable to appropriately control the force (referred to as contact force in this specification) applied to the object to be gripped by the gripper device. While one possible way to control the contact force is to install a force sensor in the gripping mechanism of the gripper device, in applications requiring hygienic handling of the object to be gripped, such as in the food manufacturing industry, there is a need to perform processes such as sterilization on the gripping mechanism. Sterilization exposes the entire gripper device to a high-temperature environment, which may damage sensors and connected cables. Therefore, it is desirable to achieve contact force control using a sensorless structure that does not require force sensors or wiring in the gripping mechanism.

[0005] Gripper devices vary in backdrivability depending on the structure of their gripping mechanisms. Backdrivability refers to the ease with which external forces are transmitted from the output side to the motor side, i.e., backdrive. A gripping mechanism with high backdrivability easily transmits the reaction force from the object to be gripped to the motor side, making it easier to estimate the magnitude of the contact force acting on the object to be gripped. Therefore, a gripping mechanism with high backdrivability can easily achieve contact force control in a sensorless structure. On the other hand, a gripping mechanism with high backdrivability may be pushed back by the reaction force and lose grip when the motor side stops driving due to a power outage or other power loss. Therefore, the possibility of the object being gripped dropping when the power is lost can be a problem in applications requiring hygiene.

[0006] In contrast, in a gripping mechanism with low backdrivability, the reaction force from the object to be gripped is less likely to be transmitted to the motor, allowing the gripping mechanism to maintain its position and continue gripping even when the motor's drive is stopped. This has the advantage of preventing the object from being dropped when the power is lost, making it suitable for applications requiring hygiene. On the other hand, when backdrivability is low, the reaction force from the object to be gripped is less likely to be transmitted to the motor, making it difficult to grasp the magnitude of the contact force acting on the object to be gripped. Therefore, with a gripping mechanism with low backdrivability, it is necessary to measure the hand position of the gripping mechanism and the force acting on the gripping mechanism using a sensor. In other words, it is difficult to control the contact force with a sensorless structure in a gripper device using a gripping mechanism with low backdrivability.

[0007] The present disclosure has been made in consideration of the above-mentioned problems, and aims to provide a control device for a gripper device and a gripper system that can appropriately control the force applied to the object to be gripped without relying on sensing of the gripping mechanism, even when the backdrivability of the gripper device is low. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, a control device of a gripper device according to one aspect of the present disclosure is a control device of a gripper device that includes a gripping mechanism that grips an object to be gripped, a motor that drives the gripping mechanism, and an angle sensor of the motor, and includes: a disturbance estimation unit that estimates a disturbance torque acting on the gripper device based on an output signal of the angle sensor and a current signal output to the motor; a reaction force estimation unit that determines a friction state inside the gripping mechanism based on the disturbance torque estimated by the disturbance estimation unit and the motor angular velocity, and estimates a reaction force acting on the gripping mechanism in accordance with the determined friction state; and a force controller that obtains a target contact force and outputs a control command value that brings the reaction force estimated by the reaction force estimation unit closer to the target contact force.

[0009] In a preferred aspect of the control device of the above-mentioned gripper device, the reaction force estimation unit includes a state machine that determines the friction state and estimates a reaction torque acting on the motor in accordance with the determined friction state, and a conversion processing unit that converts the reaction torque estimated by the state machine into the estimated reaction force.

[0010] In a preferred aspect of the control device of the above-mentioned gripper device, the state machine determines which of a plurality of friction states the internal state of the gripping mechanism corresponds to, and calculates the estimated counter torque in the determined friction state based on a calculation formula for each friction state.

[0011] In a preferred aspect of the control device of the above-mentioned gripper device, the plurality of friction states include a dynamic friction state and a static friction state, and the state machine distinguishes between the dynamic friction state and the static friction state based on the motor angular velocity.

[0012] In a preferred embodiment of the control device of the above-mentioned gripper device, the static friction state includes a first static friction state in which a static friction force smaller than a maximum static friction force acts, and a second static friction state in which the maximum static friction force acts, and the state machine distinguishes between the first static friction state and the second static friction state based on the immediately preceding estimated counter torque and the motor torque calculated from the current signal.

[0013] In a preferred aspect of the control device for the gripper device, when the speed command value for the motor becomes equal to or less than a threshold value in the static friction state, the state machine maintains the immediately preceding estimated counter torque.

[0014] In a preferred aspect of the control device for the gripper device, the disturbance estimation unit includes a Kalman filter that outputs estimated values ​​of the estimated disturbance torque and the motor angular velocity based on the output signal of the angle sensor and the current signal.

[0015] A gripper system according to one aspect of the present disclosure includes a gripper device having a gripping mechanism for gripping an object to be gripped, a motor for driving the gripping mechanism, and an angle sensor for the motor, and a control device for controlling the gripper device. The control device includes: a disturbance estimation unit that estimates a disturbance torque acting on the gripper device based on an output signal of the angle sensor and a current signal output to the motor; a reaction force estimation unit that determines a friction state inside the gripping mechanism based on the disturbance torque estimated by the disturbance estimation unit and the motor angular velocity and estimates a reaction force acting on the gripping mechanism in accordance with the determined friction state; and a force controller that acquires a target contact force and outputs a control command value that brings the reaction force estimated by the reaction force estimation unit closer to the target contact force. [Effects of the Invention]

[0016] According to the present disclosure, even when the backdrivability of the gripper device is low, the force applied to the object to be gripped can be appropriately controlled without relying on sensing of the gripping mechanism. [Brief explanation of the drawings]

[0017] [Figure 1] FIG. 1 is a schematic diagram showing a schematic configuration of a gripper system according to an embodiment. [Figure 2] FIG. 2 is a schematic diagram showing an example of the configuration of a gripping mechanism of the gripper device. [Figure 3] FIG. 3 is a block diagram illustrating the control process performed by the control device according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating the friction states classified by the state machine. [Figure 5] FIG. 5 is a flowchart showing an example of the operation of the gripper system. DETAILED DESCRIPTION OF THE INVENTION

[0018] Preferred embodiments of a gripper device control device and a gripper system according to the present disclosure will be described in detail below with reference to the drawings. Note that the present invention is not limited to these embodiments, and when there are multiple embodiments, the present invention also includes configurations that combine the embodiments. Furthermore, the components in the embodiments include those that can be easily imagined by a person skilled in the art, those that are substantially the same, and those that are within the so-called equivalent range.

[0019] (Gripper System) FIG. 1 is a schematic diagram showing a schematic configuration of a gripper system according to an embodiment.

[0020] The gripper system 1 of the embodiment includes a gripper device 10, a control device 20 for the gripper device 10, and a movement mechanism 30.

[0021] The gripper device 10 has a movable part that can be opened and closed, and can grip and release the gripping object 2 by opening and closing the movable part. The gripper device 10 includes a gripping mechanism 11 that grips the gripping object 2, a motor 12 that drives the gripping mechanism 11, and an angle sensor 13 for the motor 12. The motor 12 is an electric rotary motor. The angle sensor 13 is a sensor that detects the rotation angle of the rotation shaft of the motor 12 and is configured, for example, by a rotary encoder. The motor 12 drives the gripping mechanism 11 by rotating the rotation shaft and inputting torque to the gripping mechanism 11. The gripping mechanism 11 is a power conversion mechanism that converts the torque input by the motor 12 into opening and closing motion of the movable part. The gripping mechanism 11 includes an output member 54 as a movable part, and converts the rotation motion of the rotation shaft of the motor 12 into opening and closing motion (linear motion) of the output member 54. A claw member 55 shaped according to the gripping object 2 is attached to the output member 54. The claw members 55 are the gripping parts of the gripper device 10 and come into contact with the object to be gripped 2. The gripper device 10 grips the object to be gripped 2 by the gripping mechanism 11 operating the output member 54 and applying a contact force to the object to be gripped 2 by the claw members 55.

[0022] 1, the gripping mechanism 11 includes a pair of output members 54 that are movable toward and away from each other, and grips the object 2 to be gripped by clamping it with claw members 55 attached to the pair of output members 54. Thus, while FIG. 1 illustrates a two-claw type gripping mechanism 11 provided with a pair of output members 54 (claw members 55), the number of output members 54 may be three or more.

[0023] The object 2 to be gripped by the gripper device 10 is not particularly limited. The gripper device 10 can grip any object 2 to be gripped by attaching appropriate claw members 55 to the output member 54 depending on the shape of the object 2 to be gripped. FIG. 1 shows an example in which linear, plate-shaped claw members 55 are attached in parallel to the gripping mechanism 11, but the shape and attachment angle of the claw members 55 are not limited to this, and any shape and attachment angle suitable for gripping the object 2 to be gripped may be used.

[0024] The control device 20 controls the gripper device 10. The control device 20 receives information acquired from a sensor or the like as input, calculates a command value (a corrected current command 67 described later) for the gripper device 10, and outputs the command value. The control device 20 is realized by a processing circuit such as a system LSI (Large Scale Integration), or a processor such as a CPU (Central Processing Unit) that executes a program stored in a memory or the like.

[0025] The control device 20 is electrically connected to the angle sensor 13 and the motor driver 41 of the gripper device 10. The control device 20 controls the gripper device 10 via the motor driver 41. The control device 20 acquires an output signal (motor angle 71) from the angle sensor 13, and acquires a current signal to be output to the motor 12 from the motor driver 41. The control device 20 calculates a command value based on the output signal of the angle sensor 13 and the current signal to be output to the motor 12, and outputs the calculated command value to the motor driver 41. The motor driver 41 controls a current 43 to be supplied to the motor 12 of the gripper device 10 based on the command value, thereby controlling the operation of the gripper device 10.

[0026] The movement mechanism 30 supports and moves the gripper device 10. The movement mechanism 30 can move the gripper device 10 within a predetermined movable range using an actuator such as a motor. The movement mechanism 30 moves the gripper device 10, for example, between a position where the object to be gripped 2 is placed and a position to which the object to be gripped 2 is to be transferred. The movement mechanism 30 can transport the object to be gripped 2 gripped by the gripper device 10 to a predetermined transfer destination.

[0027] The configuration and degree of freedom of the moving mechanism 30 are not particularly limited and are set according to the application of the gripper system 1. The moving mechanism 30 may be, for example, a linear motion mechanism, a rotating arm mechanism, or a traveling mechanism that moves the gripper device 10 along a trajectory of an arbitrary path. The moving mechanism 30 may be, for example, a mechanism that can move the gripper device 10 to an arbitrary position within a movable range, such as an articulated robot, an orthogonal robot, or a parallel link robot. The actuator that serves as the driving source of the moving mechanism 30 may be a motor (electric rotary motor), a linear motor, a fluid actuator that uses a working fluid such as air, or the like. In one example, the moving mechanism 30 includes an electric rotary motor as an actuator and a position sensor such as a linear encoder.

[0028] In the example of FIG. 1, the control device 20 controls the moving mechanism 30 via a motor driver 42. The control device 20 outputs a control command value 32 (for example, a current command value) to the motor driver 42 based on an output signal 31 (position measurement value) from a position sensor of the moving mechanism 30. The motor driver 42 controls the operation of the moving mechanism 30 by controlling a current 44 supplied to the motor of the moving mechanism 30 based on the control command value 32. Note that the control device 20 of the gripper device 10 does not necessarily have to control the moving mechanism 30, and a control device for the moving mechanism 30 may be provided separately from the control device 20 of the gripper device 10, and the respective control devices may perform cooperative control through two-way communication.

[0029] (Example of gripper device configuration) Although the configuration of the gripper device 10 is not particularly limited, in the embodiment, the gripper device 10 is provided with a gripping mechanism 11 of a type with low backdrivability. That is, the gripping mechanism 11 according to the embodiment has a structure with low backdrivability, in which a reaction force from the object to be gripped 2 is not easily transmitted to the motor 12 side. As a result, even if the driving (torque input) on the motor 12 side is stopped, the gripping mechanism 11 can maintain the position of the claw members 55 and continue to grip the object to be gripped 2. Therefore, even if the power source for driving the motor 12 is lost, for example, in the event of a power outage or a malfunction, the gripper device 10 according to the embodiment can maintain the position of the claw members 55 attached to the gripping mechanism 11 (i.e., the position of the output member 54) without being driven by the motor 12, and can continue to grip the object to be gripped 2.

[0030] Fig. 2 is a schematic diagram showing an example of the configuration of the gripping mechanism of the gripper device. The gripping mechanism 11 converts the rotational motion of the motor 12 into the opening and closing motion of the output member 54. The claw members 55 shown in Fig. 1 are fixed to this output member 54, and the claw members 55 release and grip the object 2 to be gripped in accordance with the opening and closing motion of the output member 54.

[0031] The gripping mechanism 11 includes a screw member 51 that rotates integrally with the rotating shaft 12A of the motor 12, a movable nut 52 that meshes with the screw member 51, a link member 53 that is connected to the movable nut 52, and an output member 54 that is connected to the movable nut 52 via the link member 53. As described above, the number of output members 54 and link members 53 may be three or more.

[0032] The movable nuts 52 are connected to the respective link members 53 by connecting pins 56. The movable nuts 52 are constrained from rotating about the central axis AX of the rotating shaft 12A, and move linearly in a direction along the central axis AX due to rotation of the screw members 51. The link members 53 are generally L-shaped. One end of the link members 53 is swingably connected to the movable nuts 52 by the connecting pins 56. The pair of link members 53 are connected to the movable nuts 52 in a positional relationship that is symmetrical to each other with respect to the central axis AX. The other end of the link members 53 is swingably connected to the output member 54 by connecting pins 57. The link members 53 are swingably supported by support pins 58 at a bent portion between both ends. Each output member 54 is supported by a guide member 59 so as to be able to move linearly in a direction perpendicular to the central axis AX.

[0033] When the screw member 51 is rotated in one direction by the motor 12, the movable nut 52 moves toward the motor 12, and as the movable nut 52 moves, the connecting pin 56 moves toward the motor 12, causing each link member 53 to rotate around the support pin 58. As each link member 53 rotates, the pair of output members 54 move toward each other along the guide members 59, thereby closing the gripper device 10 (jaw members 55). The direction in which the pair of output members 54 move toward each other is the closing direction.

[0034] When the screw member 51 is rotated in the other direction by the motor 12, the movable nut 52 moves away from the motor 12, and as the movable nut 52 moves, the connecting pin 56 moves away from the motor 12, causing each link member 53 to rotate around the support pin 58. As each link member 53 rotates, the pair of output members 54 move in directions away from each other along the guide members 59, thereby opening the gripper device 10 (jaw members 55). The direction in which the pair of output members 54 move away from each other is the opening direction.

[0035] As described above, in the configuration example of FIG. 2, the gripping mechanism 11 converts the rotational motion of the motor 12 into linear axial motion along the central axis AX using the screw member 51 and the movable nut 52, and further converts the linear axial motion into opening and closing motion of the output member 54 using the link member 53 and the guide member 59.

[0036] In the configuration example of FIG. 2, the screw member 51 and the movable nut 52 directly mesh with each other, with the male thread portion formed on the screw member 51 and the female thread portion formed on the movable nut 52 forming a sliding screw (feed screw). A sliding screw has a large frictional resistance compared to a low-friction linear motion mechanism such as a ball screw. In the configuration example of FIG. 2, the backdrivability of the gripping mechanism 11 can be considered to be substantially determined by the frictional resistance of the sliding screw. In the embodiment, the parameters of the sliding screw are selected so as to obtain a frictional resistance sufficient to maintain the gripped state of the gripped object 2 without being driven by the motor 12, thereby realizing a gripping mechanism 11 with low backdrivability.

[0037] On the other hand, because the gripping state of the gripping object 2 can be maintained without being driven by the motor 12, even if a reaction force generated during a gripping operation acts on the claw members 55, the reaction force is hardly transmitted to the motor 12 via the gripping mechanism 11. Therefore, it is difficult to estimate the reaction force input from the gripping object 2 from the current signal on the motor 12 side or the output signal of the angle sensor 13. Therefore, when a gripping mechanism 11 with low backdrivability is used, it is usually necessary to detect the claw members 55 and the output member 54, on which the reaction force actually acts, using dedicated sensors to determine whether the gripping object 2 is being gripped and how much force is being applied to the gripping object 2. In contrast, in the embodiment, even when a gripping mechanism 11 with low backdrivability is used, the control device 20 is able to appropriately estimate the reaction force acting on the gripping mechanism 11 from the current signal on the motor 12 side and the output signal of the angle sensor 13. Therefore, in the embodiment, it is not necessary to directly sense the external force acting on the claw members 55 and the output member 54 or the displacement of these members in order to obtain the reaction force.

[0038] (Control processing of the control device) FIG. 3 is a block diagram illustrating the control process performed by the control device according to the embodiment.

[0039] The control device 20 includes a disturbance estimation unit 21, a reaction force estimation unit 22, and a force controller 23. In the example of Fig. 3, the control device 20 further includes a disturbance observer 24. The disturbance estimation unit 21, the reaction force estimation unit 22, the force controller 23, and the disturbance observer 24 represent a set of arithmetic processing executed by the control device 20, and may be realized as independent hardware (processors), or may be realized by the same hardware using a program that defines the content of the arithmetic processing.

[0040] In the control device 20 according to the embodiment, the disturbance estimation unit 21 estimates the disturbance torque using the output signal of the angle sensor 13 and the current signal to the motor 12, and the reaction force estimation unit 22 determines the state of the gripping mechanism 11 involved in the force (torque) transmission of the motor 12, and estimates the reaction force acting on the gripping mechanism 11 according to the determined state.

[0041] Specifically, in the embodiment, the equation of motion of the gripper device 10 is expressed by equation (1).

number

[0042] From equation (1), the following equation (2) is obtained.

number

[0043] From equation (2), the disturbance torque T d is estimated, and the disturbance torque T d By subtracting the sum of the frictional forces f from ex In the embodiment, a disturbance estimation unit 21 estimates the disturbance torque using the output signal of the angle sensor 13 and the current signal to the motor 12. A reaction force estimation unit 22 estimates the friction force of the gripping mechanism 11 and calculates the reaction force acting on the gripping mechanism 11 (i.e., the external force F ex ) is calculated. Then, the force controller 23 performs force control using the estimated reaction force. As a result, in the embodiment, even when a gripping mechanism 11 with low backdrivability is used, it is possible to estimate the reaction force and control the contact force from information observable from the motor 12 side (the output signal of the angle sensor 13 and the current signal to the motor 12) without sensing the reaction force from the gripping target 2.

[0044] The control process of the control device 20 will be described in detail below.

[0045] The force controller 23 acquires a target contact force 61 and outputs a control command value 63 that brings the reaction force 62 estimated by the reaction force estimator 22 closer to the target contact force 61. The estimated reaction force 62 is an estimated value of the force (reaction force) acting on the claw member 55 from the grasped object 2. The force controller 23 receives the target contact force 61 as a force command value and the estimated reaction force 62 from the reaction force estimator 22. The force controller 23 calculates an angular acceleration command for the motor 12 so that the estimated reaction force 62 approaches the target contact force 61, and outputs the calculated angular acceleration command as the control command value 63. The target contact force 61 is set in advance according to the grasped object 2. The force controller 23 acquires the target contact force 61, for example, by reading out a set value of the target contact force 61 from a memory.

[0046] The control device 20 calculates the nominal moment of inertia J of the motor 12. an and the nominal torque constant K of the motor 12tn The control command value 63 (i.e., angular acceleration command) of the force controller 23 is converted into a current command value 65 by a current conversion coefficient 64 including the nominal moment of inertia J an and nominal torque constant K tn is a constant value that can be measured in advance, but a catalog value for the motor 12 may also be used.

[0047] The disturbance observer 24 estimates the disturbance torque acting on the motor 12 from the motor angular velocity 66 and the corrected current command 67, and outputs a disturbance correction current 68 so as to cancel out the estimated disturbance torque. The disturbance observer 24 has an internal low-pass filter to prevent noise amplification. The low-pass filter may be, for example, a first-order lag low-pass filter with a predetermined cutoff frequency, but other low-pass filters such as a second-order lag low-pass filter may also be used.

[0048] The control device 20 outputs a corrected current command 67, which is obtained by subtracting a disturbance correction current 68 from a current command value 65, to the motor driver 41 (see FIG. 1) as a motor current command value. The motor 12 is operated by this corrected current command 67, and a motor torque 70 corresponding to an actual torque constant 69 of the motor 12 is generated. As shown in equation (1), the motor torque 70 (T M ) and the actual disturbance torque (T d ) and the actual moment of inertia J of the gripper device 10 is applied to the motor 12. a The angular acceleration corresponding to the moment of inertia J in Fig. 3 is a is the moment of inertia J in Eq. (1) ALL The motor angle 71, which corresponds to the value obtained by integrating the angular acceleration of the motor 12 twice, is detected by the angle sensor 13. The motor angular velocity 66 calculated by differentiating the angle θ output from the angle sensor 13 is input to the disturbance observer 24. Note that the symbol s in the block diagram of FIG. 3 is the Laplace operator.

[0049] The disturbance estimation unit 21 estimates a disturbance torque acting on the gripper device 10 based on the output signal of the angle sensor 13 and the current signal output to the motor 12. In this embodiment, the disturbance estimation unit 21 includes a Kalman filter 21A that outputs an estimated disturbance torque 72 and an estimated motor angular velocity 73 that is an estimated value of the motor angular velocity based on the output signal of the angle sensor 13 and the current signal. Here, the Kalman filter 21A is a linear Kalman filter.

[0050] The control device 20 calculates a nominal motor torque 75 based on the corrected current command 67 and a nominal torque constant 74 of the motor 12. The Kalman filter 21A receives the motor angle 71, the motor angular velocity 66, and the nominal motor torque 75 as inputs, and outputs an estimated disturbance torque 72 and an estimated motor angular velocity 73 for the motor 12. Note that the hat symbols above each variable in FIG. 3 indicate estimated values.

[0051] The Kalman filter 21A is composed of a state equation for the system shown in FIG. 3 that includes an estimated disturbance torque 72 and an estimated motor angular velocity 73 as state variables, and an observation equation that includes a motor angle 71, a motor angular velocity 66, and a nominal motor torque 75 as observables. For each sampling period, the disturbance estimation unit 21 performs the following steps for the Kalman filter 21A: prior estimation, updating of a prior error covariance matrix, calculation of a Kalman gain, posterior estimation, and updating of a posterior error covariance matrix. The prior estimation is a process of calculating an estimate of the state in the current sample (a prior state estimate) from a posterior estimate based on the previous sample. The updating of the prior error covariance matrix is ​​a process of calculating an error covariance matrix in the current sample (a prior error covariance matrix) from the posterior error covariance matrix based on the previous sample. The calculation of the Kalman gain is a process of calculating the Kalman gain from the calculated prior error covariance matrix in the current sample. The Kalman gain is the sensitivity of the Kalman filter. Posterior estimation is a process of calculating a posterior state estimate for the current sample from the prior state estimate for the current sample, the Kalman gain, and the observations acquired as the current sample (motor angle, motor angular velocity, and nominal motor torque). Updating the posterior error covariance matrix is ​​a process of calculating a posterior error covariance matrix to be used in the next sampling period from the prior error covariance matrix and Kalman gain for the current sample.

[0052] As a result of the above-described series of processes, the Kalman filter 21A calculates an estimated disturbance torque 72 and an estimated motor angular velocity 73, which are estimated as ex-post state estimates, for each sampling period.

[0053] The reaction force estimator 22 determines the internal friction state of the gripping mechanism 11 based on the estimated disturbance torque 72 and motor angular velocity (estimated motor angular velocity 73) of the disturbance estimator 21, and estimates the reaction force acting on the gripping mechanism 11 according to the determined friction state. The internal friction state of the gripping mechanism 11 is a state that determines the friction force generated between components of the power transmission path of the motor 12, from the motor 12 to the contact portion of the claw member 55 that contacts the gripping target 2. The friction state indicates, for example, whether two contacting objects are in a state of static friction or a state of kinetic friction. In the configuration example of the gripping mechanism 11 shown in FIG. 2, the friction between the screw member 51 and the movable nut 52 that constitute a sliding screw is the largest among the friction forces of each part in the power transmission path. Therefore, in determining the internal friction state of the gripping mechanism 11, the friction state between the screw member 51 and the movable nut 52 is the most influential and dominant factor.

[0054] The reaction force estimation unit 22 includes a state machine 22A that determines the friction state and estimates the reaction torque acting on the motor 12 according to the determined friction state, and a conversion processing unit 22B that converts an estimated reaction torque 77 by the state machine 22A into an estimated reaction force 62. The estimated reaction torque 77 is an estimated value of the reaction torque (reaction torque), which is a torque acting on the output shaft (rotation shaft 12A) of the motor 12 due to the reaction force from the grasp target 2.

[0055] The state machine 22A determines the friction state based on information calculated from the output signal of the angle sensor 13 and the current signal output to the motor 12. Specifically, the state machine 22A receives as input an estimated disturbance torque 72 and an estimated motor angular velocity 73 by the Kalman filter 21A, a nominal motor torque 75, and a speed command value 76 of the motor 12. The control device 20 calculates the speed command value 76 by integrating the control command value 63 (angular acceleration command) output from the force controller 23. The state machine 22A determines the friction state inside the gripping mechanism 11 based on the estimated motor angular velocity 73 and the speed command value 76 of the motor 12, which are among the input information.

[0056] In the embodiment, the state machine 22A determines which of a plurality of friction states the internal state of the gripping mechanism 11 corresponds to, and calculates the estimated counter torque 77 in the determined friction state based on a calculation formula for each friction state.

[0057] FIG. 4 is a diagram illustrating friction states classified by the state machine. The multiple friction states include a kinetic friction state 81 and a static friction state 82. The static friction state 82 includes a first static friction state 82A in which a static friction force smaller than the maximum static friction force acts, and a second static friction state 82B in which the maximum static friction force acts. The state machine 22A determines whether the internal state of the gripping mechanism 11 is the kinetic friction state 81, the first static friction state 82A, or the second static friction state 82B, based on predetermined conditions. The state machine 22A calculates the estimated reaction torque 77 using a calculation formula corresponding to each friction state.

[0058] Furthermore, when the state machine 22A determines based on a predetermined condition that the motor 12 is in a static friction state 82 but is in a low-speed state where the motor 12 has almost stopped rotating, the state machine 22A maintains the previous estimated counter torque 77. In this specification, this low-speed state of the motor 12 is distinguished from the dynamic friction state 81 (first state), the first static friction state 82A (second state), and the second static friction state 82B (third state), and is referred to as an estimated stopped state 83 (fourth state). The estimated stopped state 83 (fourth state) is a state in which the state machine 22A stops estimating the counter torque (thereby maintaining the value calculated in the previous sampling period).

[0059] 4, the state machine 22A distinguishes between a dynamic friction state 81 and a static friction state 82 based on the motor angular velocity (estimated motor angular velocity 73). In the static friction state 82, the state machine 22A distinguishes between a first static friction state 82A and a second static friction state 82B based on the immediately preceding estimated counter torque 77 and the motor torque calculated from the current signal (nominal motor torque 75).

[0060] (1) State of kinetic friction When the estimated motor angular velocity 73 calculated by the Kalman filter 21A is greater than a predetermined angular velocity threshold, the state machine 22A determines that the internal state of the gripping mechanism 11 is in a kinetic friction state 81. That is, when the state discrimination formula shown in equation (3) is satisfied, the state machine 22A determines that the internal state of the gripping mechanism 11 is in a kinetic friction state 81.

number

[0061] When the state discrimination formula shown in formula (3) is satisfied, the state machine 22A calculates the estimated counter torque using the calculation formula for the state of dynamic friction shown in formula (4).

number

[0062] In this way, when the internal state of the gripping mechanism 11 is a state 81 of kinetic friction, the state machine 22A estimates the counter torque acting on the motor 12 based on the estimated counter torque 77 from the previous time (previous sampling period), a function of the kinetic friction force and the speed-dependent viscous resistance.

[0063] (2) First static friction state The state machine 22A determines that the internal state of the gripping mechanism 11 is in the first static friction state 82A when the estimated motor angular velocity 73 calculated by the Kalman filter 21A is equal to or less than a predetermined angular velocity threshold value and the sum of the immediately preceding estimated counter torque 77 and the motor torque (nominal motor torque 75) is equal to or less than the maximum static friction force. That is, the state machine 22A determines that the internal state of the gripping mechanism 11 is in the first static friction state 82A when the state discrimination formula shown in Equation (5) is satisfied.

number

[0064] When the state discrimination formula shown in formula (5) is satisfied, the state machine 22A calculates the estimated counter torque using the calculation formula for the first static friction state shown in formula (6).

number

[0065] In this way, when the internal state of the gripping mechanism 11 is the first static friction state 82A, the state machine 22A estimates the counter torque acting on the motor 12 based on the immediately preceding (previous sampling period) estimated counter torque 77 and a function of the viscous resistance. In the first static friction state 82A, the friction force is any magnitude between zero and the maximum static friction force, but the static friction force can be ignored in consideration of the balance between the friction force and the acting force.

[0066] (3) Second static friction state The state machine 22A determines that the internal state of the gripping mechanism 11 is in the second static friction state 82B when the estimated motor angular velocity 73 calculated by the Kalman filter 21A is equal to or less than a predetermined angular velocity threshold value and the sum of the immediately preceding estimated counter torque 77 and the motor torque (nominal motor torque 75) is greater than the maximum static friction force. That is, the state machine 22A determines that the internal state of the gripping mechanism 11 is in the second static friction state 82B when the state discrimination formula shown in equation (7) is satisfied.

number

[0067] When the state discrimination formula shown in formula (7) is satisfied, the state machine 22A calculates the estimated counter torque using the calculation formula for the second static friction state shown in formula (8).

number

[0068] In this way, when the internal state of the gripping mechanism 11 is the second static friction state 82B, the state machine 22A estimates the counter torque acting on the motor 12 using a function of the immediately preceding estimated counter torque 77, the maximum static friction force, and the viscous resistance. In the second static friction state 82B, the friction force is a force with a magnitude (constant value) of the maximum static friction force, in a direction corresponding to the direction of the resultant force of the immediately preceding estimated counter torque 77 and the motor torque (nominal motor torque 75).

[0069] (4) Estimated stop state 83 When the speed command value 76 to the motor 12 becomes equal to or less than the threshold value in the static friction state 82, the state machine 22A maintains the immediately preceding estimated counter torque 77. In other words, when the motor angular velocity (estimated motor angular velocity 73) is a low value equal to or less than the threshold value and the speed command value 76 to the motor 12 also becomes a low value equal to or less than the threshold value, the state machine 22A determines that the internal state of the gripping mechanism 11 is in an estimated stopped state 83. In other words, when the state discrimination formula shown in equation (9) is satisfied, the state machine 22A determines that the internal state of the gripping mechanism 11 is in the estimated stopped state 83.

number

[0070] In equation (9), α is set to a value that satisfies equation (9) when the rotation of the motor 12 is in a low-speed state that can be considered equivalent to a stopped state in terms of both the angular velocity and the speed command value. When the state discrimination formula shown in equation (9) is satisfied, the state machine 22A calculates the estimated counter torque using the calculation formula for the estimated stopped state 83 shown in equation (10).

number

[0071] In this way, when the internal state of the gripping mechanism 11 is the estimated stopped state 83, the state machine 22A sets the estimated counter torque at the current sample (time k) to the value of the estimated counter torque at the previous sample (time k-1).

[0072] As described above, the state machine 22A determines whether the internal state of the gripping mechanism 11 is one of the four states, and calculates the estimated reaction torque 77 based on a calculation formula corresponding to the determined state. The state machine 22A outputs the calculated estimated reaction torque 77 to the conversion processing unit 22B.

[0073] The conversion processing unit 22B performs a calculation to convert the estimated reaction torque 77 output from the state machine 22A into an estimated reaction force acting on the claw members 55 of the gripping mechanism 11. The conversion processing unit 22B calculates the estimated reaction force using equation (11).

number

[0074] In addition, the coefficient P Finger corresponds to the reduction ratio R in equation (1). In this way, the reaction force estimating unit 22 calculates the estimated reaction force 62 acting on the gripping mechanism 11. The reaction force estimating unit 22 outputs the calculated estimated reaction force 62 to the force controller 23. The force controller 23 calculates a control command value 63 (angular acceleration command value) so as to make the deviation between the target contact force 61 and the estimated reaction force 62 zero, and as a result, the gripper device 10 is force-controlled so as to maintain the magnitude of the contact force applied to the gripping object 2 at the value of the target contact force 61.

[0075] (Gripper system operation) Next, the flow of operations of the gripper system 1 according to the embodiment will be described. Fig. 5 is a flowchart showing an example of operations of the gripper system.

[0076] Before the gripper system 1 starts operating, the control device 20 undergoes parameter setting and tuning. Specifically, first, the parameters of the gripper device 10 are acquired. For example, a nominal torque constant and a nominal moment of inertia are acquired from the specification information of the motor 12. The parameters of the disturbance observer 24 are also set. The cutoff frequency is set to, for example, the target speed response frequency of the actuator. Although not particularly limited, a sufficient response frequency can be ensured at around 100 Hz. The parameters of the maximum static friction force, kinetic friction force, and viscous resistance coefficient used in the processing of the state machine 22A are also identified. These parameters can be acquired, for example, experimentally. The maximum static friction force is determined by, for example, gradually increasing the current flowing through the motor 12 from a stopped state and multiplying the current value at the moment the motor starts moving by the nominal torque constant to obtain the maximum static friction force (torque). The kinetic friction force and the viscous resistance coefficient can be obtained, for example, by conducting a test in which the gripper device 10 is opened and closed at a constant speed at multiple speeds. When the relationship between the speed and the estimated disturbance torque is plotted, the intercept is obtained as the dynamic friction force (torque) and the slope is obtained as the viscous resistance coefficient. The identified parameters are set in the reaction force estimation unit 22. Furthermore, parameters of the force controller 23 are set. The force controller 23 is not particularly limited, but may be, for example, a PID controller. The parameters of the PID controller are tuned so that the output (angular acceleration command) does not diverge during force control and the response speed falls within an allowable range. This completes the advance preparation. After that, the gripper system 1 starts operating.

[0077] 5, the control device 20 controls the movement mechanism 30 via the motor driver 42 to move the gripper device 10 to the gripping position (step S10). That is, the control device 20 controls the movement mechanism 30 to move the gripper device 10 so that the pair of claw members 55 of the gripper device 10 are positioned on both sides of a predetermined gripping position where the object to be gripped 2 is located. The gripping mechanism 11 is stopped at an open position where the distance between the output members 54 is larger than the dimension of the object to be gripped 2.

[0078] The control device 20 controls the motor 12 of the gripper device 10 via the motor driver 41 to operate the gripping mechanism 11 from the open position toward the closing direction (step S11). As a result, the pair of claw members 55 moves toward the object to be gripped 2. The control device 20 controls the motor 12 so that the estimated reaction force 62 obtained from the estimated reaction torque 77 approaches the target contact force 61.

[0079] In the initial state after the motor 12 starts to drive, the state machine 22A determines that the friction state is an estimated stopped state 83 because the motor angular velocity is equal to or less than the angular velocity threshold and the motor velocity command is also equal to or less than the threshold, thereby satisfying equation (9). The state machine 22A calculates the estimated counter torque 77 using equation (11). In this case, a preset initial value is calculated for the estimated counter torque 77. Thereafter, when the motor angular velocity becomes greater than the angular velocity threshold, equation (3) is satisfied, and the state machine 22A determines that the friction state is a kinetic friction state. Therefore, the state machine 22A calculates the estimated counter torque 77 using equation (4).

[0080] As the motor 12 continues to drive the gripping mechanism 11 in the closing direction, the motor angular velocity decreases, and the contact force with respect to the gripping object 2 (i.e., the reaction force from the gripping object 2) increases. For example, when the pair of claw members 55 contacts the gripping object 2, the contact force increases and the motor angular velocity decreases (step S12). The change in the motor angular velocity also depends on the degree of elastic deformation of the gripping object 2. Accordingly, equation (5) or equation (7) is satisfied. If equation (5) is satisfied, the state machine 22A determines that the friction state is the first static friction state 82A and calculates the estimated reaction torque 77 using equation (6). If equation (7) is satisfied, the state machine 22A determines that the friction state is the second static friction state 82B and calculates the estimated reaction torque using equation (8).

[0081] Whether the friction state becomes the first static friction state 82A or the second static friction state 82B depends on various factors, such as the physical properties of the grasped object 2 and the structure and physical properties of the contact portions of the claw members 55, and is not specified as either. Also, as shown in FIG. 4 , depending on the time changes in the motor angular velocity, nominal motor torque 75, and estimated counter torque, the friction state may transition from the kinetic friction state 81 to the first static friction state 82A, from the kinetic friction state 81 to the second static friction state 82B, or between the first static friction state 82A and the second static friction state 82B after transitioning from the kinetic friction state 81 to one of the static friction states. Note that this example is a typical example, and in reality, the kinetic friction state 81 may also transition to the estimated stop state 83. Even if the contact position between the claw member 55 and the object to be grasped 2 varies due to dimensional variations in the object to be grasped 2, the result of the friction state determination changes depending on changes in the motor angular velocity, nominal motor torque 75, and estimated reaction torque values ​​associated with the contact, and the estimated reaction force 62 is calculated depending on the determination result, so the grasping operation is not affected by dimensional variations in the object to be grasped 2.

[0082] After the pair of jaw members 55 contacts the gripping object 2, when the motor 12 continues to drive the gripping mechanism 11 in the closing direction, the motor angular velocity becomes equal to or less than the angular velocity threshold, and the motor speed command also becomes equal to or less than the threshold, thereby satisfying equation (9). Therefore, the friction state is determined to be an estimated stopped state 83. For example, when the estimated reaction force 62 becomes approximately equal to the target contact force 61 and the motor speed command becomes approximately zero, the movement of the jaw members 55 essentially stops (step S13). This realizes a state in which the gripper device 10 grips the gripping object 2 with a contact force approximately equal to the target contact force 61 (i.e., a gripping state). The state machine 22A calculates the estimated reaction torque 77 using equation (11) corresponding to the estimated stopped state 83. In other words, since the estimated reaction torque 77 maintains the value immediately before (the previous sample), the estimated reaction force 62 estimated by the reaction force estimator 22 maintains the value immediately before (the previous sample). As a result, unless the gripping state changes, the contact force applied to the object to be gripped 2 is maintained almost constant (target contact force 61), preventing excessive contact force from acting on the object to be gripped 2 or the contact force acting being too small.

[0083] For example, when the deviation between the estimated reaction force 62 and the target contact force 61 falls within an allowable range, the control device 20 controls the movement mechanism 30 via the motor driver 42 to move the gripper device 10 to a target position (step S14). The target position is a position to which the object to be gripped 2 is to be transferred. The control device 20 continues to control the force of the gripper device 10 even during movement to the target position.

[0084] When the gripper device 10 reaches the target position, the control device 20 controls the motor 12 of the gripper device 10 via the motor driver 41 to operate the gripping mechanism 11 in the opening direction (step S15). As a result, the pair of claw members 55 move in a direction away from the object to be gripped 2, and the grip of the object to be gripped 2 is released. As a result, the object to be gripped 2 is placed at the target position.

[0085] This completes the example of the operation of transferring the object to be gripped 2 by the gripper system 1.

[0086] According to the above-described embodiment, the control device 20 of the gripper device 10 is a control device 20 of the gripper device 10 that includes a gripping mechanism 11 that grips the object to be gripped 2, a motor 12 that drives the gripping mechanism 11, and an angle sensor 13 of the motor 12, and includes: a disturbance estimation unit 21 that estimates a disturbance torque acting on the gripper device 10 based on the output signal of the angle sensor 13 and the current signal output to the motor 12; a reaction force estimation unit 22 that determines the friction state inside the gripping mechanism 11 based on the estimated disturbance torque 72 and motor angular velocity (estimated motor angular velocity 73) of the disturbance estimation unit 21 and estimates a reaction force acting on the gripping mechanism 11 in accordance with the determined friction state; and a force controller 23 that acquires a target contact force 61 and outputs a control command value 63 that brings the reaction force 62 estimated by the reaction force estimation unit 22 closer to the target contact force 61.

[0087] According to the control device 20 of the embodiment, the reaction force acting on the gripping mechanism 11 is estimated from the output signal of the angle sensor 13 (the motor angle 71 and the motor angular velocity (estimated motor angular velocity 73) calculated from the time derivative of the angle) and the current signal output to the motor 12. Therefore, by controlling the obtained estimated reaction force 62 to approach the target contact force 61, the object to be gripped 2 can be gripped with a constant contact force without being affected by variations in size or deformation. Dropping or damage to the object to be gripped 2 can be prevented. Furthermore, even when the backdrivability of the gripping mechanism 11 is low (i.e., when the frictional resistance in the power transmission path is high), the magnitude of the frictional resistance according to the frictional state can be estimated by determining the frictional state inside the gripping mechanism 11, and as a result, an appropriate estimated reaction force 62 can be obtained. As a result, even when the backdrivability of the gripper device 10 is low, the force applied to the object to be gripped 2 can be appropriately controlled without sensing the gripping mechanism 11. As a result, in the embodiment, it is possible to construct a gripper system 1 suitable for applications requiring hygienic handling of the gripped object 2, such as in the food manufacturing field. Specifically, since contact force control can be achieved using a sensorless structure that does not require a force sensor or wiring in the gripping mechanism 11, sterilization and other processes can be performed on the gripping mechanism 11. For example, when sterilizing and disinfecting the gripping mechanism 11 using high-temperature steam, there is no need to protect the force sensor and wiring from high temperatures or steam, which offers significant advantages in terms of reducing the number of parts in the gripping mechanism 11, reducing assembly time, and improving maintenance workability. Furthermore, since force control using a sensorless structure can be achieved with the gripping mechanism 11, which has low backdrivability, dropping of the gripped object 2 can be prevented even in the event of a power loss. The gripper system 1 and control device 20 according to the embodiment are applicable to any field of use, not just the food manufacturing field, and may be applied to any field, such as the medical field, pharmaceutical and chemical manufacturing field, or any other field.

[0088] The reaction force estimation unit 22 also includes a state machine 22A that determines the friction state and estimates the reaction torque acting on the motor 12 in accordance with the determined friction state, and a conversion processing unit 22B that converts the reaction torque 77 estimated by the state machine 22A into an estimated reaction force 62. As a result, by determining the reaction torque (reaction torque) in the motor 12 and determining in advance a relationship (conversion formula) related to torque transmission from the motor 12 to the gripped object 2, the reaction force acting on the contact portion of the gripper device 10 can be accurately estimated.

[0089] Furthermore, the state machine 22A determines which of a plurality of friction states the internal state of the gripping mechanism 11 corresponds to, and calculates the estimated counter torque 77 for the determined friction state based on a calculation formula for each friction state. In this way, by using a calculation formula for the counter torque for each friction state, even when backdrivability is low, it is possible to appropriately estimate the frictional resistance that varies depending on the friction state, and therefore it is possible to accurately calculate the estimated counter torque 77.

[0090] The plurality of friction states include a dynamic friction state 81 and a static friction state 82, and the state machine 22A distinguishes between the dynamic friction state 81 and the static friction state 82 based on the motor angular velocity (estimated motor angular velocity 73). This makes it possible to accurately estimate the magnitude of frictional resistance, which differs greatly between the dynamic friction state 81 and the static friction state 82. As a result, it is possible to accurately estimate the reaction force acting on the contact portion of the gripper device 10.

[0091] The static friction state 82 includes a first static friction state 82A in which a maximum static friction force, which is a static friction force smaller than the maximum static friction force, acts, and a second static friction state 82B. The state machine 22A distinguishes between the first static friction state 82A and the second static friction state 82B based on the immediately preceding estimated counter torque 77 and the motor torque (nominal motor torque 75) calculated from the current signal. By classifying and distinguishing the static friction state 82 into the first static friction state 82A and the second static friction state 82B in this way, the friction state of the gripping mechanism 11 when the gripping mechanism 11 contacts the gripping object 2 and applies force can be grasped in detail, and the magnitude of the frictional resistance can be estimated more precisely. As a result, even when the gripping object 2 varies in size or is deformed, the contact force can be more accurately brought closer to a constant target value.

[0092] Furthermore, when the speed command value 76 to the motor 12 becomes equal to or less than the threshold value in the static friction state 82, the state machine 22A maintains the immediately preceding estimated counter torque 77. This makes it possible to maintain the immediately preceding estimated counter torque 77 and keep the contact force constant, for example, in a state where the gripping mechanism 11 has gripped the gripping object 2 and is barely moving.

[0093] The disturbance estimation unit 21 also includes a Kalman filter 21A that outputs estimated values ​​of the estimated disturbance torque 72 and the motor angular velocity (estimated motor angular velocity 73) based on the output signal and current signal of the angle sensor 13. This makes it possible to accurately acquire the estimated disturbance torque 72 and the estimated values ​​of the motor angular velocity (estimated motor angular velocity 73). As a result, it is possible to accurately determine the friction state of the gripping mechanism 11.

[0094] Moreover, the gripper system 1 of the embodiment includes a gripper device 10 having a gripping mechanism 11 for gripping an object to be gripped 2, a motor 12 for driving the gripping mechanism 11, and an angle sensor 13 for the motor 12, and a control device 20 for controlling the gripper device 10. The control device 20 includes a disturbance estimation unit 21 that estimates a disturbance torque acting on the gripper device 10 based on the output signal of the angle sensor 13 and the current signal output to the motor 12, a reaction force estimation unit 22 that determines the friction state inside the gripping mechanism 11 based on the estimated disturbance torque 72 and motor angular velocity (estimated motor angular velocity 73) of the disturbance estimation unit 21 and estimates a reaction force acting on the gripping mechanism 11 in accordance with the determined friction state, and a force controller 23 that acquires a target contact force 61 and outputs a control command value 63 that brings the reaction force 62 estimated by the reaction force estimation unit 22 closer to the target contact force 61.

[0095] According to the gripper system 1 of the embodiment, the reaction force acting on the gripping mechanism 11 is estimated from the output signal of the angle sensor 13 (the motor angle 71 and the motor angular velocity (estimated motor angular velocity 73) calculated from the time derivative of the angle) and the current signal output to the motor 12. Therefore, by controlling the estimated reaction force 62 to approach the target contact force 61, the gripping target 2 can be gripped with a constant contact force without being affected by variations in size or deformation. Dropping or damage to the gripping target 2 can be prevented. Even when the backdrivability of the gripping mechanism 11 is low (i.e., when the frictional resistance in the power transmission path is high), the magnitude of the frictional resistance according to the frictional state can be estimated by determining the internal frictional state of the gripping mechanism 11. As a result, an appropriate estimated reaction force 62 can be calculated. As a result, even when the backdrivability of the gripper device 10 is low, the force applied to the gripping target 2 can be appropriately controlled without sensing the gripping mechanism 11.

[0096] [Variations] It should be noted that within the scope of the present invention, the embodiments may be freely combined, or any of the components of the embodiments may be modified, or any of the components of the embodiments may be omitted. [Explanation of symbols]

[0097] 1 Gripper System 2 Grasped object 10 Gripper device 11 Gripping mechanism 12 motors 13 Angle Sensor 20 Control device 21 Disturbance estimation section 21A Kalman filter 22 Reaction force estimation unit 22A State Machine 22B Conversion processing section 23 Force Controller 63 Control command value 61 Target contact force 62 Estimated reaction force 66 Motor angular velocity 71 Motor Angle 72 Estimated disturbance torque 73 Estimated motor angular velocity 76 Speed ​​command value 77 Estimated counter torque 81 State of kinetic friction 82 State of static friction 82A First static friction state 82B Second static friction state 83 Estimated stop state

Claims

1. A control device for a gripper device including a gripping mechanism that grips an object to be gripped, a motor that drives the gripping mechanism, and an angle sensor for the motor, a disturbance estimation unit that estimates a disturbance torque acting on the gripper device based on an output signal of the angle sensor and a current signal output to the motor; a reaction force estimating unit that determines a friction state inside the gripping mechanism based on the estimated disturbance torque and motor angular velocity of the disturbance estimating unit, and estimates a reaction force acting on the gripping mechanism in accordance with the determined friction state; a force controller that acquires a target contact force and outputs a control command value that causes the reaction force estimated by the reaction force estimator to approach the target contact force. A control device for a gripper device.

2. the reaction force estimating unit includes a state machine that determines the friction state and estimates a reaction torque acting on the motor in accordance with the determined friction state, and a conversion processing unit that converts the reaction torque estimated by the state machine into the estimated reaction force. The control device for a gripper device according to claim 1 .

3. The state machine determining whether the internal state of the gripping mechanism corresponds to one of a plurality of friction states; calculating the estimated counter torque in the determined friction state based on a calculation formula for each friction state; The control device for a gripper device according to claim 2 .

4. the plurality of friction states include a dynamic friction state and a static friction state, the state machine distinguishes between the dynamic friction state and the static friction state based on the motor angular velocity; The control device for a gripper device according to claim 3.

5. the static friction state includes a first static friction state in which a static friction force smaller than a maximum static friction force acts, and a second static friction state in which the maximum static friction force acts, the state machine distinguishes between the first static friction state and the second static friction state based on the immediately preceding estimated counter torque and the motor torque calculated from the current signal; The control device for a gripper device according to claim 4.

6. the state machine maintains the immediately preceding estimated counter torque when a speed command value to the motor becomes equal to or less than a threshold value in the static friction state. The control device for a gripper device according to claim 4.

7. the disturbance estimation unit includes a Kalman filter that outputs the estimated disturbance torque and the estimated motor angular velocity based on the output signal of the angle sensor and the current signal. The control device for a gripper device according to claim 1 .

8. a gripper device including a gripping mechanism for gripping an object to be gripped, a motor for driving the gripping mechanism, and an angle sensor for the motor; a control device for controlling the gripper device, The control device a disturbance estimation unit that estimates a disturbance torque acting on the gripper device based on an output signal of the angle sensor and a current signal output to the motor; a reaction force estimating unit that determines a friction state inside the gripping mechanism based on the estimated disturbance torque and motor angular velocity of the disturbance estimating unit, and estimates a reaction force acting on the gripping mechanism in accordance with the determined friction state; a force controller that acquires a target contact force and outputs a control command value that causes the reaction force estimated by the reaction force estimator to approach the target contact force. Gripper system.

Images