JP2022157368A - Robot device - Google Patents

Abstract

To suppress elongation/contraction of an artificial muscle in an axial direction caused by minute vibration of a spool of a relief valve.SOLUTION: A robot device comprises: a robot body; at least one artificial muscle that receives fluid to drive the robot body; a relief valve that has a spool and an electromagnetic part for applying thrust to the spool and regulates and supplies fluid supplied from a fluid supply source to the artificial muscle on the basis of a current supplied to the electromagnetic part; and a control unit that controls the current supplied to the electromagnetic part on the basis of a control signal for setting a fluid pressure supplied from the relief valve to the artificial muscle to a target pressure required for driving the robot body. The control unit generates the control signal, on the basis of the target current based on the target pressure to the electromagnetic part, dither amplitude minutely vibrating the spool, and a dither command value having a dither period for preventing elongation/contraction of the artificial muscle in an axial direction following a change caused by minute vibration of the spool of the fluid pressure output from the relief valve.SELECTED DRAWING: Figure 8

Description

The present disclosure relates to robotic devices.

Conventionally, this type of technology includes an arm, first and second artificial muscles that are opposed to each other via a rope wound around a pulley that rotates integrally with the arm, and that are antagonistically driven; A driving device has been proposed that includes first and second pressure regulating valves that are respectively connected to second artificial muscles (see, for example, Patent Document 1). In this driving device, a position command signal and an in-phase disturbance signal having a frequency of ten times or more the bandwidth of the driving device are used to generate control signals for the first and second pressure regulating valves. Controls the pressure regulating valve. As a result, the hysteresis of the first and second artificial muscles is reduced by the dither effect.

JP-A-61-107405

In the drive device described above, at least one of the first and second artificial muscles may vibrate due to minute vibrations of the spools of the first and second pressure regulating valves, resulting in noise and reduced durability. There is

The robot device of the present disclosure is mainly intended to improve the control accuracy of the robot main body while suppressing the noise of the artificial muscle and the decrease in durability caused by the minute vibration of the spool of the pressure regulating valve.

The robotic device of the present disclosure employs the following means in order to achieve the above main object.

A robot apparatus of the present disclosure includes a robot body, at least one artificial muscle that receives a supply of fluid to drive the robot body, a spool, and an electromagnetic section that applies a thrust to the spool. a pressure regulating valve that regulates the pressure of the fluid based on the current supplied to the electromagnetic section and supplies it to the artificial muscle; and the fluid pressure that is supplied from the pressure regulating valve to the artificial muscle is required to drive the robot body. a controller for controlling current supplied to the electromagnetic unit based on a control signal for achieving a target pressure, wherein the controller controls the target pressure to the electromagnetic unit based on the target pressure. an electric current, a dither command value having a dither amplitude that causes the spool to vibrate slightly, and a dither cycle that does not cause the artificial muscle to expand or contract in the axial direction due to a change in the fluid pressure output from the pressure regulating valve caused by the slight vibration of the spool; , to generate the control signal.

In the robot apparatus of the present disclosure, the target current to the electromagnetic unit based on the target pressure required to drive the robot main body, the dither amplitude that micro-vibrates the spool, and the change in the fluid pressure output from the pressure regulating valve due to the micro-vibration of the spool. A control signal is generated based on a dither command value having a dither cycle that does not cause the artificial muscle to expand or contract in the axial direction. Based on this control signal, the current supplied to the electromagnetic portion of the pressure regulating valve is controlled. As a result, expansion and contraction of the artificial muscle in the axial direction due to minute vibrations of the spool of the pressure regulating valve can be suppressed. As a result, it is possible to improve the control accuracy of the robot main body while suppressing the noise of the artificial muscle and the decrease in durability caused by the slight vibration of the spool of the pressure regulating valve. For example, when the current position of the robot body is approaching the target position, the robot body can be made to reach the target position with high precision, and when the robot body is moved slowly, the robot body can be moved with high precision. can.

1 is a schematic configuration diagram showing a robot device of the present disclosure; FIG. It is an enlarged view showing a robot device. 1 is a schematic configuration diagram showing a fluid supply device of a robot device; FIG. FIG. 2 is a schematic configuration diagram showing first and second linear solenoid valves; (a), (b), and (c) are cross-sectional views for explaining the operation of the first and second linear solenoid valves of the present disclosure. FIG. 2 is a block diagram showing main parts of a control device of the robot device; FIG. 4 is an explanatory diagram illustrating a target pressure setting map; FIG. 2 is a block diagram showing main parts of a control device of the robot device; FIG. 4 is an explanatory diagram illustrating how current is supplied to the electromagnetic portion of the first linear solenoid valve; 9 is a flowchart illustrating an example of dither amplitude setting processing; 9 is a flowchart illustrating an example of dither amplitude setting processing; FIG. 2 is a block diagram showing the main part of the control device of the robot device of the present disclosure;

Next, embodiments for carrying out the invention of the present disclosure will be described with reference to the drawings.

FIG. 1 is a schematic configuration diagram showing a robot device 1 of the present disclosure, and FIG. 2 is an enlarged view showing the robot device 1. As shown in FIG. The robot device 1 shown in these drawings includes a robot arm (robot body) 2, a fluid supply device 10, and a control device 100 that controls the entire device. The robot arm 2 includes a plurality of (three in this embodiment) joints (pin joints) J1, J2, and J3, a plurality of (three in this embodiment) arms (links) 3, joints J1, A plurality of fluid actuators M as artificial muscles, for example, an even number (four in this embodiment) provided for each of J2 and J3, and a hand portion (robot hand) as a grip portion attached to the tip of the arm 3 on the tip side. ) 4 is an articulated arm. The hand unit 4 is controlled by the control device 100 so as to grip a target object (hereinafter referred to as a "gripping target"). Further, the fluid supply device 10 is controlled by the control device 100 to supply and discharge working oil (working fluid) as a fluid to each fluid actuator M. As shown in FIG. As a result, the robot arm 2 can be driven by hydraulic pressure (fluid pressure) to move the hand portion 4 to a desired position.

Each fluid actuator M of the robot arm 2 is, as shown in FIG. The tube T is formed in a cylindrical shape from an elastic material such as a rubber material having high oil resistance. A hydraulic oil inlet/outlet IO is formed in a sealing member C on the base end side of the tube T (on the fluid supply device 10 side, the lower end side in FIG. 2). The braided sleeve S is formed in a cylindrical shape by weaving a plurality of cords oriented in a predetermined direction so as to intersect each other, and is contractible in the axial and radial directions. As the cords forming the braided sleeve S, fiber cords, high-strength fibers, metal cords composed of ultrafine filaments, and the like can be used. Hydraulic oil is supplied from the inlet/outlet IO into the tube T of the fluid actuator M to increase the pressure of the hydraulic oil in the tube T, whereby the tube T expands radially and axially due to the action of the braided sleeve S. and generates a contraction force according to the pressure of the hydraulic oil inside.

As shown in FIGS. 1 and 2, among the plurality of arms 3, the arm 3 closest to the proximal end (the side closest to the fluid supply device 10) is rotatable by a support member 5 as a link via a joint J1. Supported. Also, the two arms 3 are rotatably connected to each other via a joint J2 or J3. Furthermore, a connecting member 6 is fixed to the tip of each arm 3 (end on the hand side). As illustrated, the support member 5 rotatably supports the sealing members C on the proximal side of the plurality (four) of the fluid actuators M corresponding to the joint J1 on the most proximal side. In addition, the connecting member 6 of each arm 3 rotatably supports sealing members C on the tip side (hand side) of a plurality (four) of fluid actuators M corresponding to the joint J1 or J2 located on the base end side. do. Further, each connecting member 6 rotatably supports sealing members C on the proximal side of a plurality (four) of fluid actuators M corresponding to joints J2 or J3 located on the distal side.

More specifically, the support member 5 rotatably supports the sealing member C on the base end side of the two fluid actuators M corresponding to the joint J1 via the first connecting shaft. Also, the connecting member 6 of the arm 3 on the most proximal side rotatably supports the sealing members C on the distal side of the two fluid actuators M corresponding to the joint J1 via the second connecting shaft. Further, the support member 5 rotates the sealing member C on the base end side of the remaining two fluid actuators M corresponding to the joint J1 via a third connecting shaft extending parallel to the first connecting shaft. To support. In addition, the connecting member 6 of the arm 3 on the most proximal side is a fourth connecting shaft extending parallel to the second connecting shaft through the sealing member C on the distal end side of the remaining two fluid actuators M corresponding to the joint J1. rotatably supported via Similarly, the connecting members 6 of the two arms 3 connected to each other via the joints J2 or J3 are also connected to a plurality of (four) fluids corresponding to the joints J2 or J3 via a plurality of connecting shafts as described above. A corresponding sealing member C of the actuator M is rotatably supported.

As a result, two fluid actuators M are arranged parallel to the corresponding arm 3 on both sides of each arm 3 extending from the joint axis of the joints J1 to J3 to the hand side (hand part 4 side) in this embodiment. . Two fluid actuators M arranged on one side of each arm 3 constitute a first artificial muscle (one antagonistic muscle) AM1 (see FIG. 3) corresponding to one joint J1, J2 or J3. , two fluid actuators M arranged on the other side of each arm 3 are second artificial muscles (other antagonistic muscles) corresponding to one joint J1, J2 or J3 paired with the first artificial muscle AM1. Configure AM2 (see FIG. 3). However, the number of fluid actuators M forming the first artificial muscle AM1 may differ from the number of fluid actuators M forming the second artificial muscle AM2. Further, in this embodiment, the plurality (four) of fluid actuators M provided for one joint J1, J2 or J3 have the same specifications. However, the specifications of a plurality of fluid actuators M corresponding to one joint J1, J2 or J3 do not necessarily have to be the same. The specifications of the fluid actuator M that constitutes the second artificial muscle AM2 may be different. Furthermore, each arm 3 is hollow, and a plurality of hoses H (see broken lines in FIG. 2) as fluid supply pipes are arranged inside each arm 3 . Each hose H is connected to an inlet/outlet IO formed in a sealing member C on the base end side of the corresponding fluid actuator M. of working oil (hydraulic pressure) is supplied.

Therefore, by controlling the fluid supply device 10 with the control device 100, the hydraulic pressure in the tubes T of the two fluid actuators M constituting the first artificial muscle AM1 and the second hydraulic pressure paired with the first artificial muscle AM1 are controlled. The hydraulic pressures in the tubes T of the two fluid actuators M constituting the artificial muscle AM2 can be made different from each other. As a result, force (rotational torque) is transmitted from the four fluid actuators M, that is, a pair of first and second artificial muscles AM1 and AM2 to each arm 3 via the connecting member 6, and the support member It is possible to change the joint angles of the joints J1 to J3 by rotating each arm 3 with respect to the arm 3 on the proximal end side. In this embodiment, the two fluid actuators M constituting the first artificial muscle AM1 and the two fluid actuators M constituting the second artificial muscle AM2 paired with the first artificial muscle AM1 are the tubes T are axially contracted by a predetermined amount (for example, about 10% of the natural length), which is the initial state.

As shown in FIG. 1, the fluid supply device 10 of the robot device 1 includes a tank 11 defining a hydraulic oil reservoir (fluid reservoir), and a rotation shaft (a dashed line in FIG. 1) extending vertically through the tank 11. ) and a base portion 12 pivotably supported thereabout. The tank 11 is, for example, a cylindrical body with closed upper and lower ends, and can store hydraulic oil therein. In this embodiment, as shown in FIG. 2, the support member 5 of the robot arm 2 is fixed to the upper wall portion 11u of the tank 11 via bolts (not shown) or the like. That is, the robot arm 2 is supported by the tank 11 (upper wall portion 11 u) of the fluid supply device 10 .

The base portion 12 is fixed to an installation location of the robot device 1 so as to be positioned below the robot arm 2 and the tank 11, or is mounted (fixed) on an automatic guided vehicle (AGV) (not shown). The base portion 12 also supports a rotation unit (not shown) that rotates the tank 11 around the rotation shaft. Accordingly, by operating the rotating unit, the robot arm 2 and the tank 11 can be rotated integrally around the rotating shaft. The rotating unit may be a swing motor driven by hydraulic pressure supplied from the fluid supply device 10, or may include an electric motor or the like.

Further, as shown in FIG. 3, the fluid supply device 10 includes, in addition to the tank 11 and the base portion 12, a pump 13 as a fluid supply source, a valve body (not shown) arranged in the tank 11, and a source pressure generating unit. It includes a valve 14 and first and second linear solenoid valves 151 and 152 as a plurality of pressure regulating valves (pressure regulating devices). Both the pump 13 and the first and second linear solenoid valves 151 and 152 are controlled by the control device 100 . The first and second linear solenoid valves 151, 152 are provided one each for each of the joints J1-J3.

The pump 13 is, for example, an electric pump, and sucks hydraulic oil stored in the tank 11 and discharges it from a discharge port. The pump 13 includes a pump portion arranged inside the tank 11 and a drive portion having an electric motor and a reduction gear mechanism and arranged inside or outside the tank 11 . The original pressure generating valve 14 drains (regulates) part of the working oil discharged from the pump 13 according to the signal pressure from the signal pressure generating valve (not shown) to generate the original pressure, and the original pressure is applied to the valve body. is supplied to an oil passage (fluid passage) L0 formed in . As the signal pressure generating valve of the source pressure generating valve 14, for example, a linear solenoid valve whose energization is controlled by the control device 100 is used.

The first and second linear solenoid valves 151 and 152 are arranged in valve bodies and have the same structure as each other. As shown in FIGS. 20 and a valve portion 30 that is driven by the electromagnetic portion 20 to regulate the pressure of the hydraulic fluid. In this embodiment, the first and second linear solenoid valves 151 and 152 are normally closed linear solenoid valves that open when current is supplied to the electromagnetic section 20 .

The electromagnetic section 20 includes tubular first and second cores arranged side by side in the axial direction, tubular coils arranged so as to surround the first and second cores, and axial coils inside the second core. a plunger movably arranged in the first core, a rod 21 axially movable in conjunction with the plunger in the first core, and a yoke (case) accommodating these members (Fig. 4 shows the rod 21 only). When current flows through the coils of the electromagnetic portion 20, a magnetic flux circuit is formed that flows through the yoke, second core, plunger, and first core in that order. As a result, the plunger is attracted toward the first core, and in conjunction with the plunger, the rod 21 moves in a direction (to the right in FIG. 4) protruding from the first core. In this embodiment, the current supplied to the electromagnetic section 20 is controlled by a PWM signal generated based on the target current.

As shown in FIG. 4, the valve portion 30 includes a substantially cylindrical sleeve 40 incorporated in the above-described valve body, and a spool 50 arranged axially slidably (movably) inside the sleeve 40. have. One end of the sleeve 40 (left end in the drawing) is fixed to the electromagnetic section 20 (yoke), and an end (right end in the drawing) of the sleeve 40 opposite to the electromagnetic section 20 side has an end portion. A cap CP that closes the is fixed (screwed). A spring (elastic member) SP is arranged inside the sleeve 40 so as to be positioned between the spool 50 and the cap CP. The spring SP is a coil spring in this embodiment, and biases the spool 50 toward the electromagnetic section 20 (left side in FIG. 4).

As shown in FIG. 4, the sleeve 40 has an input port 41i, an output port 41o that can communicate with the input port 41i, a feedback port 41f that communicates with the output port 41o, and a drain port 41d that can communicate with the output port 41o. including. As shown in FIG. 3, the input ports 41i of the first and second linear solenoid valves 151, 152 communicate with the oil passage L0. The output port 41o of the first linear solenoid valve 151 is connected to the hydraulic fluid inlet/outlet IO of the two fluid actuators M (tubes T) constituting the first artificial muscle AM1 through the oil passages L11 and L12. communicate. The output port 41o of the first linear solenoid valve 151 is connected to the hydraulic fluid inlet/outlet IO of the two fluid actuators M (tube T) that constitute the second artificial muscle AM2 through the oil passages L21 and L22. communicate. The drain ports 41d of the first and second linear solenoid valves 151 and 152 communicate with the hydraulic oil reservoir in the tank 11 through the oil passages L3. In this embodiment, as shown in FIG. 4, the input port 41i, the output port 41o, the drain port 41d, and the feedback port 41f are spaced in this order from the electromagnetic portion 20 side toward the spring SP (cap CP) side. are formed in the sleeve 40 so as to line up in the axial direction. That is, the input port 41i is formed closer to the electromagnetic section 20 than the output port 41o, the drain port 41d is formed closer to the spring SP than the output port 41o, and the feedback port 41f is closer to the spring SP than the drain port 41d. formed in

Inside the sleeve 40, there are an input chamber 42i communicating with the input port 41i, an output chamber 42o communicating with the output port 41o, a drain chamber 42d communicating with the drain port 41d, and a feedback chamber 42f communicating with the feedback port 41f. are defined at intervals of . Further, inside the sleeve 40, there are a first communication chamber 45 opening at an input chamber 42i and an output chamber 42o, a second communication chamber 46 opening at an output chamber 42o and a drain chamber 42d, a drain chamber 42d and a feedback chamber 42f. A third communication chamber 47 that opens at . The input chamber 42i, the output chamber 42o, the drain chamber 42d, and the feedback chamber 42f are circular cross-sectional space portions having the same inner diameter (cross-sectional area). The first to third communication chambers 45, 46, and 47 are circular cross-sectional space portions having the same inner diameter (cross-sectional area) as each other and smaller than the inner diameter (cross-sectional area) of the input chamber 42i and the like. The input chamber 42i, the output chamber 42o, the drain chamber 42d, the feedback chamber 42f, and the first to third communication chambers 45, 46, 47 coaxially extend along the axis of the sleeve 40. As shown in FIG.

The spool 50 has four lands 51, 52, 53 and 54, a first shaft portion 55 between the lands 51 and 52, and a second shaft portion 56 between the lands 52 and 53, as shown in FIG. , and a third shank 57 between lands 53 and 54 . The lands 51, 52 and 53 are formed in a cylindrical shape having the same outer diameter (cross-sectional area), and the land 54 has an outer diameter (cross-sectional area) smaller than the outer diameter (cross-sectional area) of the lands 51-53. It is formed in a columnar shape. The outer diameters of the lands 52 and 53 are set to values slightly smaller than the inner diameters of the first to third communication chambers 45 , 46 , 47 of the sleeve 40 . Further, in this embodiment, the land 52 of the spool 50 has an axial length that is longer than the axial length of the output chamber 42o of the sleeve 40. As shown in FIG. The first to third shaft portions 55 - 57 are formed in a cylindrical shape having an outer diameter (cross-sectional area) smaller than at least the outer diameter (cross-sectional area) of the lands 52 and 53 . Lands 51 - 54 and first through third shaft portions 55 - 57 extend coaxially with each other along the axis of spool 50 .

A land 51 of the spool 50 is slidably disposed in a hole (circular hole) formed in the sleeve 40 so as to communicate with the input chamber 42i from the electromagnetic portion 20 side. A contact portion 51a and a stopper portion 51s are formed at the tip of the land 51 (the left end in FIG. 4) to contact the rod 21 of the electromagnetic portion 20. As shown in FIG. Further, the land 54 of the spool 50 is slidably disposed in a hole (circular hole) formed in the sleeve 40 so as to communicate with the feedback chamber 42f from the spring SP (cap CP) side. The spring SP described above is arranged between CP. As a result, the spool 50 is slidably disposed inside the sleeve 40 and biased toward the electromagnetic portion 20 by the spring SP. As the spool 50 moves, the land 52 of the spool 50 changes the state of communication between the input port 41i and the output port 41o of the sleeve 40 and the state of communication between the output port 41o and the drain port 41d.

In the first and second linear solenoid valves 151 and 152 configured as described above, the biasing force applied to the spool 50 by the spring SP toward the electromagnetic portion 20 and the biasing force applied to the electromagnetic portion 20 (coil) A thrust applied to the spool 50 from the electromagnetic unit 20 by the current applied to the spool 50 toward the side opposite to the electromagnetic unit 20 side and a hydraulic pressure supplied from the output port 41o to the feedback port 41f to the electromagnetic unit 20 side applied to the spool 50 By balancing with the thrust of , it is possible to adjust the pressure of the hydraulic oil supplied from the source pressure generating valve 14 (pump 13) side to the input port 41i and flowing out from the output port 41o to a desired pressure.

Specifically, in the first and second linear solenoid valves 151 and 152, when power is not supplied to the coil of the electromagnetic section 20, as shown in FIGS. 4 and 5(a), the spool 50 ( and rod 21) are pressed against the plunger of the electromagnetic section 20 (to the left in FIGS. 4 and 5) by the biasing force of the spring SP. As a result, as shown in FIG. 5A, the end surface 52i of the land 52 of the spool 50 on the side of the input chamber 42i is located in the first communication chamber 45, and the end surface 52d of the land 52 on the side of the drain chamber 42d is output. Located in chamber 42o. By positioning the end surface 52i of the land 52 on the input chamber 42i side within the first communication chamber 45, the input port 41i and the output port 41o form the sleeve 40 that defines the outer peripheral surface of the land 52 and the first communication chamber 45. communicates through a slight clearance with the inner peripheral surface of the Furthermore, by positioning the end face 52d of the land 52 on the drain chamber 42d side inside the output chamber 42o, the output chamber 42o and the drain chamber 42d communicate with each other through the second communication chamber 46 with a sufficient amount of communication. Hereinafter, such a state of the first and second linear solenoid valves 151, 152 will be referred to as "first state".

When a current is supplied to the electromagnetic portion 20 and the rod 21 moves together with the plunger toward the spring SP (to the right in FIGS. 4 and 5), the spool 50 is pressed by the rod 21 against the biasing force of the spring SP. Move to the SP (cap CP) side. In the first and second linear solenoid valves 151 and 152, as the spool 50 moves toward the spring SP, as shown in FIG. While positioned within the chamber 45 , the end surface 52 d of the land 52 on the drain chamber 42 d side crosses the boundary between the output chamber 42 o and the second communication chamber 46 and is positioned within the second communication chamber 46 . By positioning the end surface 52i of the land 52 on the input chamber 42i side within the first communication chamber 45, the input port 41i and the output port 41o form the sleeve 40 that defines the outer peripheral surface of the land 52 and the first communication chamber 45. communicates through a slight clearance with the inner peripheral surface of the Furthermore, the end surface 52d of the land 52 on the drain chamber 42d side is positioned within the second communication chamber 46, so that the output port 41o and the drain port 41d define the outer peripheral surface of the land 52 and the second communication chamber 46. It communicates with the inner peripheral surface of the sleeve 40 through a slight clearance. Hereinafter, such a state of the first and second linear solenoid valves 151, 152 will be referred to as "second state".

When the current supplied to the electromagnetic part 20 further increases and the spool 50 is further pressed by the rod 21 and moves toward the spring SP (cap CP), a land 52 is formed as shown in FIG. 5(c). While the end face 52d on the drain chamber 42d side of the land 52 is located in the second communication chamber 46, the end face 52i on the input chamber 42i side of the land 52 crosses the boundary between the first communication chamber 45 and the output chamber 42o and extends into the output chamber 42o. Located in By locating the end surface 52i of the land 52 in the output chamber 42o, the input chamber 42i and the output chamber 42o communicate with each other through the first communication chamber 45 with a sufficient amount of communication. Furthermore, the end surface 52d of the land 52 on the drain chamber 42d side is positioned within the second communication chamber 46, so that the output port 41o and the drain port 41d define the outer peripheral surface of the land 52 and the second communication chamber 46. It communicates with the inner peripheral surface of the sleeve 40 through a slight clearance. Hereinafter, such a state of the first and second linear solenoid valves 151, 152 will be referred to as "third state".

Hereinafter, the amount of movement of the spool 50 of the first and second linear solenoid valves 151, 152 from the initial position (the position in the first state) will be referred to as "stroke amounts Sc1, Sc2". Stroke amounts Sc1 and Sc2 when the end surface 52i of the land 52 of the spool 50 on the side of the input chamber 42i is positioned at the boundary between the first communication chamber 45 and the output chamber 42o are referred to as "predetermined stroke amounts Scref1 and Scref2." The amount of communication between the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of the sleeve 40 is controlled when the stroke amounts Sc1 and Sc2 are equal to or smaller than the predetermined stroke amounts Scref1 and Scref2 (first state and second state). state), the clearance becomes a predetermined amount (slight clearance between the outer peripheral surface of the land 52 and the inner peripheral surface of the sleeve 40 defining the first communication chamber 45), and the stroke amounts Sc1 and Sc2 reach the predetermined stroke amount Scref1. , Scref2 (in the third state), the stroke amounts Sc1 and Sc2 gradually increase from the predetermined amount as the stroke amounts Sc1 and Sc2 increase. Therefore, when the stroke amounts Sc1 and Sc2 are larger than the predetermined stroke amounts Scref1 and Scref2, the unit movement of the spool 50 of the valve portion 30 is smaller than when the stroke amounts Sc1 and Sc2 are equal to or less than the predetermined stroke amounts Scref1 and Scref2. The amount of change in the amount of communication between the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of the sleeve 40 per amount, and the output port 41o of the first and second linear solenoid valves 151 and 152 The amount of change in the output hydraulic pressure of becomes large.

Electric power is supplied to the pump 13, the signal pressure generating valve of the source pressure generating valve 14, the first and second linear solenoid valves 151 and 152, the control device 100, and the like from an auxiliary battery 60 as a power source. Auxiliary battery 60 is, for example, a lead-acid battery having a rated voltage of 12V.

A control device 100 of the robot device 1 includes a microcomputer including a CPU, a ROM, a RAM, an input/output interface, etc., and various logic ICs (all not shown). The control device 100 detects the hydraulic pressure (fluid pressure), which is the pressure of the working oil in the oil passage L0 downstream of the first and second linear solenoid valves 151 and 152, and the voltage of the auxiliary battery 60. A detection value of a voltage sensor (not shown) for detecting is inputted. The control device 100 duty-controls the pump 13 so that the oil pressure in the oil passage L0 detected by the source pressure sensor reaches a target value, and supplies the signal to the electromagnetic portion of the signal pressure generation valve of the source pressure generation valve 14. control the current. The control device 100 controls the first and second linear solenoid valves 151 and 152 so that hydraulic pressure corresponding to a request (target hydraulic pressure) is supplied from the first and second linear solenoid valves 151 and 152 to each fluid actuator M. A target current to the electromagnetic section 20 is set, and the current supplied to each electromagnetic section 20 is controlled based on the target current.

FIG. 6 is a block diagram showing control units for the first and second linear solenoid valves 151 and 152 in the control device 100. As shown in FIG. As shown in the figure, the control device 100 includes a target position setting unit 101 constructed by at least one of hardware such as a computer CPU, ROM, and RAM, and software such as a control program installed in the computer. , a current position derivation unit 102, a target torque setting unit 105 including a torque calculation unit 103 and a gravity compensation unit 104, a target stiffness setting unit 106, a contraction rate setting unit 107, a contraction force calculation unit 108, and a target pressure derivation unit It has a functional block including a target pressure setting section 110 including 109 , a target current setting section 111 , and first and second valve drive sections 112 and 113 .

The target position setting unit 101 sets the target position, which is the final target position of the hand unit 4, based on the position of the object to be gripped by the hand unit 4 and the target velocity and target acceleration during movement of the hand unit 4 given by the user. A position (three-dimensional coordinates) and a target trajectory, which is a trajectory from the initial position of the hand unit 4 to a target arrival position and includes a plurality of target positions, that is, intermediate positions (three-dimensional coordinates), are set.

Based on the joint angles θ1, θ2, θ3 of the joints J1-J3 of the robot arm 2 and the specifications of the robot arm 2 (robot apparatus 1) (such as the dimensions of the arm 3), the current position derivation unit 102 calculates the position of the hand unit 4. A current position (three-dimensional coordinates) of (predetermined reference point) is derived. The joint angles θ1-θ3 of the joints J1-J3 are detected by corresponding one of the joint angle sensors 7 provided on the robot arm 2 . Hereinafter, "i" is a joint number (in this embodiment, i=1, 2, 3), the i-th joint is referred to as "joint Ji", and the joint angle of joint Ji is "θi ”.

The torque calculation unit 103 of the target torque setting unit 105 calculates two arms 3 (the arm 3 and the support arm 3) connected via the joint Ji so that the hand unit 4 moves from the current position to the target position for each joint J1-J3. A joint torque Tj(i) for relatively rotating the member 5) is calculated. The gravity compensation unit 104 of the target torque setting unit 105 adjusts the robot arm 2 based on the joint angles θ1-θ3 and the specifications of the robot arm 2 (robot device 1) (such as the dimensions of the arm 3) for each of the joints J1-J3. Gravity compensation torque Tc(i) required to maintain the attitude of . Then, the target torque setting unit 105 sets the sum of the joint torque Tj(i) and the gravity compensation torque Tc(i) as A target torque Ttag(i), which is a target value (target driving force) of the joint torque, is set.

Based on at least the target position of the robot device 1, that is, the hand unit 4, the target stiffness setting unit 106 determines the stiffness that the joints Ji should have, that is, the two arms 3 and the like connected via the joints Ji, for each of the joints J1 to J3. The force (torque) required to relatively rotate the (link) by a unit angle, and indicates the difficulty of movement of the joint Ji against an external force that relatively rotates the two arms 3, etc. A target stiffness R(i) is set.

The contraction rate setting unit 107 of the target pressure setting unit 110 sets the first artificial muscle corresponding to the joint Ji based on the joint angle θi of the joint Ji corresponding to the current position of the hand unit 4 for each of the joints J1 to J3. The contraction rate Cr1(i) of the two fluid actuators M forming AM1 and the contraction rate Cr2(i) of the two fluid actuators M forming the second artificial muscle AM2 corresponding to the joint Ji are set. The contraction ratios Cr1(i) and Cr2(i) indicate the ratio of the axial length of the contracted tube T to the natural length of the tube T of the corresponding fluid actuator M in the axial direction.

The contractile force calculator 108 of the target pressure setting unit 110 calculates the target torque Ttag(i) set by the target torque setting unit 105 and the target stiffness R ( i), when the two arms 3 or the like connected via the joint Ji are relatively rotated by the target torque Ttag(i), a plurality (a pair of) fluid actuators M corresponding to the joint Ji are provided with Calculate the required contractile forces Fc1(i) and Fc2(i). The contractile force Fc1(i) is the force to be generated by contracting the tubes T of the two fluid actuators M constituting the first artificial muscle AM1 corresponding to each joint Ji, and the contractile force Fc2(i) This is the force to be generated by the contraction of the tubes T of the two fluid actuators M forming the second artificial muscle AM2 corresponding to the joint Ji.

The target pressure derivation unit 109 of the target pressure setting unit 110 calculates the contraction rate Cr1(i) set by the contraction rate setting unit 107 and the contraction force from the target pressure setting map illustrated in FIG. 7 for each of the joints J1 to J3. A pressure corresponding to the contractile force Fc1(i) calculated by the unit 108 is derived and set as the target pressure Ptag1(i) of the two fluid actuators M constituting the first artificial muscle AM1. The target pressure derivation unit 109 calculates the contraction rate Cr2(i) set by the contraction rate setting unit 107 and the contraction force calculation unit 108 from the target pressure setting map illustrated in FIG. 7 for each of the joints J1 to J3. The target pressure Ptag2(i) of the two fluid actuators M constituting the second artificial muscle AM2 is set by deriving the pressure corresponding to the contraction force Fc2(i).

The target pressure setting map in FIG. 7 shows the static characteristics of the fluid actuator M as an artificial muscle. It was created through experiments and analyzes in advance so as to define the relationship of In this way, pressures corresponding to contraction rates Cr1(i), Cr2(i) and contraction forces Fc1(i), Fc2(i) of the tube T are set as target pressures Ptag1(i), Ptag2(i). , the target pressures Ptag1(i) and Ptag2(i) can be accurately set according to the requirements for the robot arm 2. FIG.

The target current setting unit 111 applies the target pressures Ptag1(i) and Ptag2(i) set by the target pressure setting unit 110 to the target currents to the electromagnetic units 20 of the first and second linear solenoid valves 151 and 152 of the joint Ji. Convert to Itag1(i) and Itag2(i). In this embodiment, since the first and second linear solenoid valves 151 and 152 are normally closed linear solenoid valves, the target current setting unit 111 increases as the target pressures Ptag1(i) and Ptag2(i) increase. The target currents Itag1(i) and Itag2(i) are set as follows.

The first and second valve driving units 112, 113 are provided corresponding to the first and second linear solenoid valves 151, 152 of the joints J1-J3, respectively. The first and second valve driving units 112 and 113 operate the first and second linear solenoid valves 151 and 152 of the joint Ji based on the target currents Itag1(i) and Itag2(i) set by the target current setting unit 111. controls the current supplied to each electromagnetic part 20 of the . In this manner, the first and second linear solenoid valves 151, 152 of the joint Ji are controlled to generate hydraulic pressures corresponding to the target pressures Ptag1(i), Ptag2(i).

Next, details of the first and second valve drive units 112, 113 corresponding to the first and second linear solenoid valves 151, 152 of the joint Ji will be described. FIG. 8 is a block diagram showing details of the first and second valve drive units 112, 113 corresponding to the first and second linear solenoid valves 151, 152 of the joint Ji. As shown in the figure, the first and second valve drive units 112 and 113 of the joint Ji respectively include a target voltage setting unit 115, a dither command value setting unit 116, a voltage superimposition unit 117, and a PWM signal generation unit. 118 , a drive unit 119 , a current detection unit 120 and a filter processing unit 121 .

Each target voltage setting unit 115 sets the target current Itag1(i), Itag2(i) of the first and second linear solenoid valves 151, 152 of the joint Ji set by the target current setting unit 111, and the corresponding current Filter processing by the filter processing unit 121 corresponding to the currents Is1(i) and Is2(i) supplied to the first and second linear solenoid valves 151 and 152 of the joint Ji detected by the detection unit 120. The first and second linear solenoid valves 151, 152 of the joint Ji are controlled by PI control (proportional-integral control) or PID control (proportional-integral-derivative control) based on the difference between the post-currents Isf1(i), Isf2(i). Target voltages Vtag1(i) and Vtag2(i) are calculated. That is, each target voltage setting unit 115 sets the sum of a proportional term and an integral term based on the difference between the target currents Itag1(i), Itag2(i) and the post-processing currents Isf1(i), Isf2(i) to the target voltage. Calculate as Vtag1(i) and Vtag2(i).

Each dither command value setting unit 116 converts the voltage command values Vcom1(i) and Vcom2(i) generated by the corresponding voltage superimposing unit 117 to the target voltage Vtag1(i) set by the corresponding target voltage setting unit 115 ( i), dither command values Vdiz1(i) for varying the dither cycles Tdiz1(i), Tdiz2(i) and the dither amplitudes Adiz1(i), Adiz2(i) for Vtag2(i), for example, sinusoidally, Generate Vdiz2(i).

Here, the dither cycles Tdiz1(i) and Tdiz2(i) are the cycles (PWM cycles) Tpwm1(i) and Tpwm2 of the PWM signals Spwm1(i) and Spwm2(i) generated by the PWM signal generator 118, respectively. It is set to a time longer than (i). The dither amplitudes Adiz1(i), Adiz2(i) are the PWM periods Tpwm1(i), The amount of variation (amplitude) in the dither cycles Tdiz1(i) and Tdiz2(i) is set to be larger than the amount of variation (amplitude) in Tpwm2(i). Details of the dither periods Tdiz1(i), Tdiz2(i) and the dither amplitudes Adiz1(i), Adiz2(i) will be described later.

Each voltage superimposing unit 117 applies the dither command value Vdiz1 ( i) and Vdiz2(i) are superimposed to generate voltage command values Vcom1(i) and Vcom2(i) for the first and second linear solenoid valves 151 and 152 of the joint Ji. That is, each voltage superimposing unit 117 generates voltage commands obtained by varying target voltages Vtag1(i) and Vtag2(i) with dither amplitudes Adiz1(i) and Adiz2(i) and dither cycles Tdiz1(i) and Tdiz2(i). Generate the values Vcom1(i) and Vcom2(i). Generating the voltage command values Vcom1(i) and Vcom2(i) by superimposing the dither command values Vdiz1(i) and Vdiz2(i) on the target voltages Vtag1(i) and Vtag2(i) is hereinafter referred to as “dither control. ”.

Each PWM signal generation unit 118 applies the voltage command values Vcom1(i) and Vcom2(i) generated by the corresponding voltage superimposition unit 117 to the joint Ji in each PWM period Tpwm1(i) and Tpwm2(i), respectively. are converted into pulse width modulation (PWM) signals Spwm1(i) and Spwm2(i) of the first and second linear solenoid valves 151 and 152, and output to the corresponding drive units 119. FIG. As described above, the PWM cycles Tpwm1(i), Tpwm2(i) are set shorter than the dither cycles Tdiz1(i), Tdiz2(i).

Each driving unit 119 has a switching element SW, which is a MOSFET. The drain of the switching element SW is connected to the positive electrode of the auxiliary battery 60, and the source of the switching element SW is connected to the electromagnetic section 20 of the first and second linear solenoid valves 151 and 152 of the joint Ji and the current detection section 120 (not shown). The gate of the switching element SW is connected to the corresponding PWM signal generator 118 . Note that the switching element SW may be a bipolar transistor or an IGBT instead of the MOSFET.

In each drive unit 119, when the switching element SW is turned on by the PWM signals Spwm1(i) and Spwm2(i) from the corresponding PWM signal generation unit 118, the voltage Vbat of the auxiliary battery 60 changes to the corresponding electromagnetic unit. 20 (coil), and an electromotive current flows through the electromagnetic section 20 . Further, in each driving unit 119, when the switching element SW is turned off by the PWM signals Spwm1(i) and Spwm2(i) from the corresponding PWM signal generating unit 118, the corresponding electromagnetic unit 20 (coil) is grounded. be done. Hereinafter, ON/OFF control of the switching element SW using the PWM signals Spwm1(i) and Spwm2(i) is referred to as "PWM control".

Each current detection unit 120 is a shunt having one end connected to the electromagnetic unit 20 of the corresponding one of the first and second linear solenoid valves 151 and 152 of the joint Ji and the other end grounded. An operational amplifier that detects the voltage between the resistor, the corresponding electromagnetic part 20 side of the shunt resistor, and the ground side, and the output of the operational amplifier (analog signal of voltage) is converted into an analog signal of current value, and the current value as a digital signal. and an A/D converter for converting to Is1(i) and Is2(i) (both not shown).

Each filter processing unit 121 filters the current values Is1(i) and Is2(i) detected by the corresponding current detection unit 120 so that the above dither cycles Tdiz1(i) and Tdiz2(i) are obtained. The processed currents Isf1(i) and Isf2(i) from which the frequency components have been removed are output to the corresponding target voltage setting units 115 . Each filter processing unit 121 can remove the frequency components of the dither cycles Tdiz1(i) and Tdiz2(i) as filter processing. For example, a band stop filter (notch filter) or the like is used.

By means of such functional blocks of the control device 100, in particular, the current supplied to the electromagnetic parts 20 of the first and second linear solenoid valves 151, 152 of the joint Ji by the first and second valve drive parts 112, 113 of the joint Ji Is1(i), Is2(i) can be varied based on dither amplitudes Adiz1(i), Adiz2(i) and dither periods Tdiz1(i), Tdiz2(i). FIG. 9 is an explanatory diagram illustrating how the current Is1(1) supplied to the electromagnetic portion 20 of the first linear solenoid valve 151 of the joint J1. As shown in the figure, the dither cycle Tdiz1(1) is longer than the PWM cycle Tpwm1(1), and the variation (amplitude) of the current Is1 caused by the dither control is equal to the current Is1(1) caused by the PWM control. ) is larger than the variation (amplitude) of Currents Is2(1), Is1(2), Is2(2), Is2(2), The states of Is1(3) and Is2(3) are the same as those of the current Is1(i). Therefore, based on the variation of currents Is1(1), Is2(1), Is1(2), Is2(2), Is1(3), and Is2(3) mainly due to dither control, The plungers of the first and second linear solenoid valves 151 and 152, the rod 21, and the spool 50 can be finely vibrated. As a result, the hysteresis of the output pressure due to the sliding resistance of the plunger, the rod 21, and the spool 50 can be more appropriately reduced, and the responsiveness of the first and second linear solenoid valves 151 and 152 for each joint J1-J3 can be reduced. Variation and output pressure variation can be suppressed more appropriately.

In this embodiment, the dither amplitudes Adiz1(i) and Adiz2(i) are set as amplitudes capable of micro-vibrating the spools 50 of the first and second linear solenoid valves 151 and 152 of the joint Ji. Tdiz1(i) and Tdiz2(i) correspond to changes in hydraulic pressure (fluid pressure) output from the first and second linear solenoid valves 151 and 152 of the joint Ji due to minute vibrations of the spool 50. and a frequency that does not extend or contract the tubes T of the two fluid actuators M constituting the second artificial muscles AM1 and AM2 in the axial direction. By setting the dither cycles Tdiz1(i), Tdiz2(i) and the dither amplitudes Adiz1(i), Adiz2(i) in this way, the first and second linear solenoid valves 151, 152 for each joint J1-J3 It is possible to suppress expansion and contraction in the axial direction of the tube T of the fluid actuator M of the first and second artificial muscles AM1 and AM2 caused by minute vibrations of the spool 50, thereby suppressing the generation of vibration and noise and the deterioration of durability. can. At this time, since the hysteresis of the output pressures of the first and second linear solenoid valves 151 and 152 is sufficiently reduced due to minute vibrations of the spool 50, the target position of the hand unit 4 set by the target position setting unit 101 is When the amount of deviation between the position (the target arrival position or the intermediate position on the trajectory from the initial position to the target arrival position) and the current position of the hand unit 4 derived by the current position derivation unit 102 is less than or equal to the predetermined deviation amount, i.e. , the hand portion 4 can be made to reach the target position with high accuracy when the hand portion 4 is approaching the target position. Further, when the target speed during movement of the hand portion 4 given by the user is equal to or less than a predetermined speed, that is, when the hand portion 4 is operated slowly, the hand portion 4 can be operated with higher accuracy. When the hand unit 4 is slowly moved, for example, when the hand unit 4 is approaching the target position, when the hand unit 4 grips a fragile object to be gripped, and when the hand unit 4 grips a fragile object to be gripped. is grasped by the hand portion 4 and moved.

In this embodiment, the dither cycles Tdiz1(1), Tdiz2(1), Tdiz1(2), Tdiz2(2), Tdiz1(2), and Tdiz2(2) are all the same for ease of design. The dither amplitudes Adiz1(1), Adiz2(1), Adiz1(1), Adiz2(2), Adiz1(1), Adiz2(2) are all set to be the same. Also, the PWM cycles Tpwm1(1), Tpwm2(1), Tpwm1(2), Tpwm2(2), Tpwm1(3), Tpwm2(3) are all set to be the same. However, it is not limited to this. For example, at least part of the dither cycles Tdiz1(1), Tdiz2(1), Tdiz1(2), Tdiz2(2), Tdiz1(2), Tdiz2(2) may be different from each other. At least some of the dither amplitudes Adiz1(1), Adiz2(1), Adiz1(1), Adiz2(2), Adiz1(1), and Adiz2(2) may be different from each other. Furthermore, at least part of the PWM cycles Tpwm1(1), Tpwm2(1), Tpwm1(2), Tpwm2(2), Tpwm1(3), and Tpwm2(3) may be different from each other.

As described above, in the robot apparatus 1, the target voltages Vtag1(i) and Vtag2(i) based on the target currents Itag1(i) and Itag2(i), the dither amplitudes Adiz1(i) and Adiz2(i), and the dither PWM signals Spwm1(i), Spwm2(i) based on dither command values Vdiz1(i), Vdiz2(i) having periods Tdiz1(i), Tdiz2(i) are used to determine the first and second It controls the current supplied to the linear solenoid valves 151 and 152 . In this case, the dither amplitudes Adiz1(i) and Adiz2(i) are set as amplitudes capable of micro-vibrating the spools 50 of the first and second linear solenoid valves 151 and 152 of the joint Ji. In addition, the dither cycles Tdiz1(i) and Tdiz2(i) are set according to changes in the hydraulic pressure (fluid pressure) output from the first and second linear solenoid valves 151 and 152 of the joint Ji due to minute vibrations of the spool 50. A frequency is set at which the tubes T of the two fluid actuators M constituting the first and second artificial muscles AM1 and AM2 of Ji are not stretched or contracted in the axial direction. As a result, the axial direction of the tube T of the fluid actuator M of the first and second artificial muscles AM1 and AM2 caused by the slight vibration of the spools 50 of the first and second linear solenoid valves 151 and 152 for each joint J1-J3 It is possible to suppress the expansion and contraction at, and suppress the generation of vibration and noise, and the deterioration of durability. Further, at this time, since the hysteresis of the output pressures of the first and second linear solenoid valves 151 and 152 is sufficiently reduced due to minute vibrations of the spool 50, for example, when the hand unit 4 is approaching the target position, Moreover, the hand portion 4 can be caused to reach the target position with high accuracy, and the hand portion 4 can be moved more accurately when the hand portion 4 is moved slowly.

Also, the dither cycles Tdiz1(i) and Tdiz2(i) are set longer than the cycles (PWM cycles) Tpwm1(i) and Tpwm2(i) of the PWM signals Spwm1(i) and Spwm2(i). As a result, the plungers, rods 21, and spools 50 of the first and second linear solenoid valves 151 and 152 can be slightly vibrated based on fluctuations in the currents Is1(i) and Is2(i) mainly due to dither control. can. As a result, it is possible to more appropriately reduce the sliding resistance between the spool 50 and the sleeve 40, and to more appropriately suppress the decrease in responsiveness of the first and second linear solenoid valves 151 and 152 and the variation in response. .

In the above embodiment, the dither command values Vdiz1(1), Vdiz2(1), Vdiz1(2), Vdiz2(2), Vdiz1(2) of the first and second linear solenoid valves 151, 152 for each joint J1-J3 , Vdiz2(2) are omitted. However, the phases of the dither command values Vdiz1(1), Vdiz2(1), Vdiz1(2), Vdiz2(2), Vdiz1(2), and Vdiz2(2) may all be the same. At least part of them may be different from each other. In the description of the example in this case, for ease of description, the dither periods Tdiz1(1), Tdiz2(1), Tdiz1(2), Tdiz2(2), Tdiz1(3), Tdiz2(3) are the same. (hereinafter referred to as “Tdiz”).

A case in which the phases of the dither command values Vdiz1(1), Vdiz2(1), Vdiz1(2), Vdiz2(2), Vdiz1(2), and Vdiz2(2) are made different from each other, for example, by 60° will be described. In this case, the voltage command values Vcom1(1), Vcom2(1), Vcom1(2), Vcom2(2), Vcom1(3), Vcom2 of the first and second linear solenoid valves 151, 152 for each joint J1-J3 Currents Is1(1), Is2(1), Is1(2), and Is2(2) supplied to the electromagnetic units 20 with different phases (hereinafter referred to as “dither phases”) in the dither period Tdiz of (3) , Is1(3) and Is2(3) are different from each other. When the dither phases of the currents Is1(1), Is2(1), Is1(2), Is2(2), Is1(3), and Is2(3) supplied to the electromagnetic units 20 overlap, the auxiliary battery 60 increases, the voltage drop of the auxiliary battery 60 increases, and there is a possibility that sufficient current will not be supplied to each electromagnetic section 20 . In this case, sufficient hydraulic oil (hydraulic pressure) is not supplied to at least one fluid actuator M, causing the behavior of the robot arm 2 to become unstable, or the hand unit 4 to move along the target trajectory to the target arrival position. It may become impossible to move. On the other hand, the dither phases of the currents Is1(1), Is2(1), Is1(2), Is2(2), Is1(3), and Is2(3) supplied to the electromagnetic units 20 are made different. As a result, the peak of the current from the auxiliary battery 90 can be dispersed, the voltage drop of the auxiliary battery 90 can be suppressed, and a sufficient current can be supplied to each electromagnetic part 20 . As a result, sufficient hydraulic oil (hydraulic pressure) can be supplied to each fluid actuator M, and the robot arm 2 can be operated appropriately. Furthermore, by dispersing the peaks of the current from auxiliary battery 60, noise can be reduced, and adverse effects on control device 100 and devices electrically connected thereto can be suppressed.

A case has been described where the phases of the dither command values Vdiz1(1), Vdiz2(1), Vdiz1(2), Vdiz2(2), Vdiz1(2), and Vdiz2(2) are made different from each other. However, the dither command values Vdiz1(1) and Vdiz2(1) have the same phase, the dither command values Vdiz1(2) and Vdiz2(2) have the same phase, and the dither command values Vdiz1(3) and Vdiz2(3) may be the same, and the phases of the dither command values Vdiz1(1), Vdiz1(2), and Vdiz1(3) may be different from each other, for example, by 120 degrees. In this case, the voltage command values Vcom1(1) and Vcom2(1) have the same dither phase, the voltage command values Vcom1(2) and Vcom2(2) have the same dither phase, and the voltage command values Vcom1(3) and Vcom2 The dither phases of (3) are the same, and the dither phases of the voltage command values Vcom1(1), Vcom1(2), and Vcom1(3) are different from each other. Then, the currents Is1(1) and Is2(1) supplied to the electromagnetic section 20 have the same dither phase, the currents Is1(2) and Is2(2) have the same dither phase, and the currents Is1(3) and Is2 The dither phases of (3) are the same, and the dither phases of the currents Is1(1), Is(2) and Is(3) are different from each other. As a result, while suppressing the voltage drop of the auxiliary battery 90 by dispersing the peak of the current from the auxiliary battery 90 to some extent, the two fluid actuators M constituting the first artificial muscle AM1 corresponding to the joint Ji and The two fluid actuators M constituting the second artificial muscle AM2 can be satisfactorily driven in opposition.

In the above embodiment, each voltage superimposing unit 117 superimposes the dither command values Vdiz1(i) and Vdiz2(i) on the target voltages Vtag1(i) and Vtag2(i) to obtain the voltage command values Vcom1(i) and Vcom2(i). i). That is, the dither control is always executed. However, at least when the amount of deviation between the target position of the hand unit 4 (the target position or intermediate position on the trajectory from the initial position to the target position) and the current position of the hand unit 4 is less than or equal to the predetermined amount of deviation, and Alternatively, dither control may be executed when the target speed during movement of the hand unit 4 is equal to or lower than a predetermined speed. Therefore, each voltage superimposing unit 117 is set to the target position when the amount of deviation between the target position and the current position of the hand unit 4 is larger than the predetermined amount of deviation or when the target speed during movement of the hand unit 4 is larger than the predetermined speed. Voltages Vtag1(i) and Vtag2(i) may be set to voltage command values Vcom1(i) and Vcom2(i), that is, dither control may not be executed.

In the above embodiment, the dither amplitudes Adiz1(i), Adiz2(i) and the dither cycles Tdiz1(i), Tdiz2(i) of the dither command values Vdiz1(i), Vdiz2(i) are set to constant values, respectively. However, at least one of these may be variable. A case where the dither amplitudes Adiz1(i) and Adiz2(i) are variable will be described. FIG. 10 shows dither amplitude setting processing that is repeatedly executed for setting the dither amplitude Adiz1(1) by the dither command value setting unit 116 of the first driving unit 112 corresponding to the first linear solenoid valve 151 of the joint J1. It is a flow chart which shows an example. Note that the dither amplitudes Adiz2(1), Adiz1(2), Adiz2(2), Adiz1(3), and Adiz2(3) are set in the same manner as the dither amplitude Adiz1(1).

In the dither amplitude setting process of FIG. 10, the dither command value setting unit 116 inputs the target current Itag1(1) set by the corresponding target current setting unit 111 (step S100), and the input target current Itag1(1) (step S110). Here, the target current increase rate Iup1(1) is obtained, for example, by subtracting the previous target current (previous Itag1(1)) from the current target current Itag1(i) and dividing it by the input period Δt of the target current Itag1. calculated as

Subsequently, the target current increase rate Iup(1) is compared with the threshold value Iupref(1) (step S120). Here, the threshold value Iupref(1) is set from the side where the stroke amount Sc1(1) of the first linear solenoid valve 151 of the joint J1 is smaller than the vicinity of the predetermined stroke amount Scref1(1) to the vicinity of the predetermined stroke amount Scref1(1) or less. is a threshold value used to determine whether there is a possibility that the stroke amount is greater than or is already near or greater than the predetermined stroke amount Scref1 (1) (hereinafter referred to as "first possibility"), It is determined in advance by experiments and analyses. When the first linear solenoid valve 151 of the joint J1 is a normally closed linear solenoid valve, the stroke amount Sc1(1) increases as the current supplied to the electromagnetic section 20 increases, and the stroke amount Sc1(1) becomes When the stroke amount Scref1(1) is greater than the predetermined stroke amount Scref1(1), the input chamber 41i of the sleeve 40 per unit movement of the spool 50 is greater than when the stroke amount Sc1(1) is equal to or less than the predetermined stroke amount Scref1(1). (input port 41i) and the output chamber 42o (output port 41o), the amount of change in the amount of communication, and by extension, the amount of change in the output hydraulic pressure from the output port 41o of the first linear solenoid valve 151 of the joint J1 increases. The processing of step S120 is performed in consideration of this. Note that when the target current increase rate Iup(1) is equal to or greater than the threshold Iupref(1), the target pressure Ptag1(i) of the two fluid actuators M constituting the first artificial muscle AM1 corresponding to the joint J1 is When there is a rapid increase, for example, when the hand unit 4 rapidly moves the grasped object can be mentioned.

When it is determined in step S120 that the target current increase rate Iup1(1) is less than the threshold value Iupref(1), it is determined that there is no first possibility, and a relatively large first amplitude A1 is added to the dither amplitude Adiz1(1). Amplitude normal processing to be set is executed (step S130), and this processing ends. By executing the normal amplitude processing, the voltage command value Vcom1(1) generated by the corresponding voltage superimposing unit 117 fluctuates with a relatively large amplitude, and the corresponding drive circuit 119 causes the first linear solenoid valve 151 of the joint J1 to change. The current supplied to the electromagnetic portion 20 of the rotor fluctuates with a relatively large amplitude, causing the spool 50 to oscillate with a relatively large amplitude. As a result, the hysteresis of the output pressure due to the influence of the sliding resistance between the spool 50 and the sleeve 40 of the first linear solenoid valve 151 of the joint J1 is sufficiently reduced, and the responsiveness of the linear solenoid valve 151 of the joint J1 is reduced. Variation and deterioration of output pressure control accuracy can be sufficiently suppressed.

When it is determined in step S120 that the target current increase rate Iup1(1) is equal to or greater than the threshold value Iupref(1), it is determined that there is the first possibility, and the dither amplitude Adiz1(1) is equal to or greater than 0 and the first amplitude. Amplitude reduction processing is executed to set a second amplitude A2 smaller than A1 (step S130), and this processing ends. By executing the amplitude reduction process, the amplitude of the fluctuation of the voltage command value Vcom1(1) generated by the corresponding voltage superimposition section 117 becomes smaller than when the normal amplitude process is executed, and the corresponding drive circuit 119 reduces the amplitude of the fluctuation of the current supplied to the electromagnetic part 20 of the first linear solenoid valve 151 of the joint J1, and reduces the amplitude of the oscillation of the spool 50 . As a result, when the stroke amount Sc1(1) fluctuates near or larger than the predetermined stroke amount Scref1(1), the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of the sleeve 40 It is possible to suppress large fluctuations in the amount of communication with the joint J1, and to suppress large fluctuations in the output oil pressure from the output port 41o of the first linear solenoid valve 151 of the joint J1.

In the above embodiment, the first and second linear solenoid valves 151, 152 corresponding to the joints J1-J3 are both normally closed linear solenoid valves. However, at least one of the first and second linear solenoid valves 151, 152 corresponding to the joints J1-J3 may be a normally open linear solenoid valve. Here, the case where the first and second linear solenoid valves 151 corresponding to the joints J1-J3 are both normally open linear solenoid valves will be described. As a hardware configuration in this case, a sleeve 40 and a spool 50 similar to those of the linear solenoid valves 151 and 152 in FIG. 4 are provided. A configuration for moving the spool 50 to the left in FIG. 4 against the urging force of the spring SP when is supplied will be described as an example. For ease of explanation, the same reference numerals as those of the first and second linear solenoid valves 151 and 152 in FIG. 4 are used.

Hereinafter, the amount of movement of the spool 50 from the initial position when the first and second linear solenoid valves 151 are normally closed linear solenoid valves will be referred to as "stroke amounts So1, So2". Stroke amounts So1 and So2 when the end surface 52i of the land 52 of the spool 50 on the side of the input chamber 42i is positioned at the boundary between the first communication chamber 45 and the output chamber 42o are referred to as "predetermined stroke amounts Soref1 and Soref2." The amount of communication between the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of the sleeve 40 is such that when the stroke amounts So1 and So2 are smaller than the predetermined stroke amounts Soref1 and Soref2, the stroke amounts So1 and So2 are As it becomes larger, it gradually decreases from a second predetermined amount, which is larger than a predetermined amount (a slight clearance between the outer peripheral surface of the land 52 and the inner peripheral surface of the sleeve 40 defining the first communication chamber 41), toward the predetermined amount. , and when the stroke amounts So1 and So2 are greater than or equal to the predetermined stroke amounts Soref1 and Soref2, they are the predetermined amounts. Therefore, when the stroke amounts So1 and So2 are smaller than the predetermined stroke amounts Soref1 and Soref2, compared to when the stroke amounts So1 and So2 are equal to or larger than the predetermined stroke amounts Soref1 and Soref2, the sleeve per unit movement amount of the spool 50 is reduced. The amount of change in the amount of communication between the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of 40, and the output hydraulic pressure from the output port 41o of the first and second linear solenoid valves 151 and 152 the amount of change increases.

When the first and second linear solenoid valves 151 and 152 are normally open linear solenoid valves, the control device 100 can be configured to operate in the same way as when the linear solenoid valves 151 and 152 are normally closed linear solenoid valves. It has functional blocks as shown in FIG. However, the target current setting units 111 of the first and second linear solenoid valves 151 and 152 of the joint Ji set the target currents Itag1(i) and Set Itag2(i).

Further, each dither command value setting unit 116 sets the dither amplitudes Adiz1(i), Adiz2(i) and the dither cycles Tdiz1(i), Tdiz2(i) of the dither command values Vdiz1(i), Vdiz2(i) to constant values. It may be a value, or at least one of them may be variable. A case where the dither amplitudes Adiz1(i) and Adiz2(i) are variable will be described. FIG. 11 shows dither amplitude setting processing that is repeatedly executed for setting the dither amplitude Adiz1(1) by the dither command value setting unit 116 of the first driving unit 112 corresponding to the first linear solenoid valve 151 of the joint J1. It is a flow chart which shows an example. Note that the dither amplitudes Adiz2(1), Adiz1(2), Adiz2(2), Adiz1(3), and Adiz2(3) are set in the same manner as the dither amplitude Adiz1(1).

In the dither amplitude setting process of FIG. 11, the dither command value setting unit 116 inputs the target current Itag1(1) set by the corresponding target current setting unit 111 (step S200), and the input target current Itag1(1) , the target current decrease rate Idn1(1), which is the amount of decrease in the target current Itag1(i) per unit time, is calculated (step S210). Here, the target current decrease rate Idn1(1) is obtained by, for example, dividing a value obtained by subtracting the current target current Itag1(i) from the previous target current (previous Itag1(1)) by the input period Δt of the target current Itag1. calculated as

Subsequently, the target current decrease rate Idn(1) is compared with the threshold value Idnref(1) (step S220). Here, the threshold value Idnref(1) is set from the side where the stroke amount So1(1) of the first linear solenoid valve 151 of the joint J1 is greater than the predetermined stroke amount Soref1(1) to the predetermined stroke amount Soref1(1) or so. is a threshold value used to determine whether there is a possibility that the stroke amount has reached the side smaller than or is already close to or smaller than the predetermined stroke amount Soref1 (1) (hereinafter referred to as "second possibility"), It is determined in advance by experiments and analyses. When the first linear solenoid valve 151 of the joint J1 is a normally open linear solenoid valve, the stroke amount So1(1) decreases as the current supplied to the electromagnetic section 20 decreases, and the stroke amount So1(1) When the stroke amount Sc1(1) is smaller than the predetermined stroke amount Soref1(1), the input chamber 41i of the sleeve 40 per unit movement amount of the spool 50 is greater than when the stroke amount Sc1(1) is equal to or greater than the predetermined stroke amount Scref1(1). (input port 41i) and the output chamber 42o (output port 41o), the amount of change in the amount of communication, and by extension, the amount of change in the output hydraulic pressure from the output port 41o of the first linear solenoid valve 151 of the joint J1 increases. The process of step S220 is performed in consideration of this. Note that when the target current decrease rate Idn(1) is equal to or greater than the threshold Idnref(1), the target pressure Ptag1(i) of the two fluid actuators M constituting the first artificial muscle AM1 corresponding to the joint J1 is When there is a rapid increase, for example, when the hand unit 4 rapidly moves the grasped object can be mentioned.

When it is determined in step S220 that the target current decrease rate Idn1(1) is less than the threshold Idnref(1), it is determined that there is no second possibility, and a relatively large first amplitude A1 is added to the dither amplitude Adiz1(1). Amplitude normal processing to be set is executed (step S230), and this processing ends. By executing the normal amplitude processing, the voltage command value Vcom1(1) generated by the corresponding voltage superimposing unit 117 fluctuates with a relatively large amplitude, and the corresponding drive circuit 119 causes the first linear solenoid valve 151 of the joint J1 to change. The current supplied to the electromagnetic portion 20 of the rotor fluctuates with a relatively large amplitude, causing the spool 50 to oscillate with a relatively large amplitude. As a result, the hysteresis of the output pressure due to the influence of the sliding resistance between the spool 50 and the sleeve 40 of the first linear solenoid valve 151 of the joint J1 is sufficiently reduced, and the responsiveness of the linear solenoid valve 151 of the joint J1 is reduced. Variation and deterioration of output pressure control accuracy can be sufficiently suppressed.

When it is determined in step S220 that the target current decrease rate Idn1(1) is equal to or greater than the threshold value Idnref(1), it is determined that there is the second possibility, and the dither amplitude Adiz1(1) is equal to or greater than 0 and the first amplitude. Amplitude reduction processing is executed to set a second amplitude A2 smaller than A1 (step S230), and this processing ends. By executing the amplitude reduction process, the amplitude of the fluctuation of the voltage command value Vcom1(1) generated by the corresponding voltage superimposition section 117 becomes smaller than when the normal amplitude process is executed, and the corresponding drive circuit 119 reduces the amplitude of the fluctuation of the current supplied to the electromagnetic part 20 of the first linear solenoid valve 151 of the joint J1, and reduces the amplitude of the oscillation of the spool 50 . As a result, when the stroke amount So1(1) fluctuates near or smaller than the predetermined stroke amount Soref1(1), the input chamber 42i (input port 41i) and the output chamber 42o (output port 41o) of the sleeve 40 It is possible to suppress large fluctuations in the amount of communication with the joint J1, and to suppress large fluctuations in the output oil pressure from the output port 41o of the first linear solenoid valve 151 of the joint J1.

In the above embodiment, the first and second valve drive units 112 and 113 corresponding to the first and second linear solenoid valves 151 and 152 of the joint Ji are respectively the target voltage setting unit 115 and the dither command value setting unit 116. , a voltage superimposition unit 117 , a PWM signal generation unit 118 , a drive unit 119 , a current detection unit 120 and a filter processing unit 121 . However, as shown in FIG. 12, the first and second valve drive sections 112, 113 may be replaced with first and second valve drive sections 112B, 113B. The first and second valve driving sections 112B and 113B in FIG. 12 are the same as the target voltage setting section 115, the dither command value setting section 116, the voltage superimposing section 117 and the PWM of the first and second valve driving sections 112 and 113 in FIG. The first and second valve drive sections 112, 113 is different. Therefore, of the first and second valve drive sections 112B and 113B in FIG. 12, the same parts as those of the first and second valve drive sections 112 and 113 in FIG. do.

Each current command value setting unit 115B corresponds to the target currents Itag1(i) and Itag2(i) of the first and second linear solenoid valves 151 and 152 of the joint Ji set by the target current setting unit 111. The currents Is1(i) and Is2(i) supplied to the first and second linear solenoid valves 151 and 152 of the joint Ji detected by the current detection unit 120 are filtered by the corresponding filter processing unit 121. Current command values Icom1(i), Icom2 ( Calculate i). That is, each target current setting unit 115B sets the sum of a proportional term, an integral term, etc. based on the difference between the target currents Itag1(i), Itag2(i) and the post-processing currents Isf1(i), Isf2(i). Calculated as values Icom1(i) and Icom2(i).

Each dither command value setting unit 116B converts the superimposed currents Icom1′(i) and Icom2′(i) generated by the corresponding current superimposing unit 117B to the currents set by the corresponding current command value setting unit 115B. For varying the dither cycles Tdiz1′(i), Tdiz2′(i) and the dither amplitudes Adiz1′(i), Adiz2′(i) with respect to the command values Icom1(i), Icom2(i), for example, sinusoidally. Dither command values Idiz1(i) and Idiz2(i) are generated. Here, the dither periods Tdiz1'(i), Tdiz2'(i) and the dither amplitudes Adiz1'(i), Adiz2'(i) are the dither periods Tdiz1(i), Tdiz2(i) and the dither amplitudes Adiz1(i). i), set in the same manner as Adiz2(i).

Each current superimposing unit 117B adds the current command values Icom1(i) and Icom2(i) set by the corresponding current command value setting unit 115B to the dither command values set by the corresponding dither command value setting unit 116. Idiz1(i) and Idiz2(i) are superimposed to generate superimposed currents Icom1'(i) and Icom2'(i). Each PWM signal generation unit 118B generates the superimposed currents Icom1′(i) and Icom2′(i) generated by the corresponding current superimposing unit 117B, or the superimposed currents Icom1′(i) and Icom2′(i). is converted into PWM signals Spwm1(i), Spwm2(i) of the first and second linear solenoid valves 151, 152 of the joint Ji, and output to the corresponding drive unit 119.

Even when the first and second valve drive sections 112 and 113 are replaced with the first and second valve drive sections 112B and 113B, the same effects as when the first and second valve drive sections 112 and 113 are provided can be obtained. can be done.

In the above embodiment, at least one of the first and second linear solenoid valves 151, 152 for each joint J1-J3 does not have a dedicated feedback port and outputs pressure inside the sleeve containing the spool. may be configured to act on the spool as feedback pressure (see, for example, Japanese Unexamined Patent Application Publication No. 2020-41687). At least one of the first and second linear solenoid valves 151 and 152 for each of the joints J1-J3 includes a linear solenoid valve that outputs a signal pressure corresponding to the current supplied to the electromagnetic section, and a It may be replaced with a control valve that adjusts hydraulic fluid pressure accordingly. Furthermore, at least one of the first and second linear solenoid valves 151 and 152 for each of the joints J1-J3 is controlled so that the hydraulic pressure (fluid pressure) supplied to the corresponding fluid actuator M becomes the target pressure. It may be replaced by a flow control valve that Additionally, the original pressure generating valve 14 may be omitted from the fluid supply device 10 . Further, the fluid supply device 10 may be provided with an accumulator (pressure accumulator) that stores the hydraulic pressure generated by the pump 13 . Further, the fluid supply device 10 may be configured to supply the fluid actuator M with a liquid other than hydraulic oil such as water or a gas such as air.

In the above-described embodiment, the fluid actuator M as an artificial muscle includes a tube T that is supplied with hydraulic oil and radially expands and axially contracts in response to an increase in internal hydraulic pressure. and a braided sleeve S that covers a McKibben-type artificial muscle. However, the configuration of the fluid actuator M in the robot device 1 is not limited to this. That is, the fluid actuator M may include a tube that expands in the radial direction and contracts in the axial direction when supplied with fluid. An axial fiber reinforced fluid comprising an outer tubular member formed and coaxially disposed outside the inner tubular member, and a fiber layer disposed between the inner tubular member and the outer tubular member. It may be an actuator (see, for example, JP-A-2011-137516). Further, fluid actuator M may be a fluid cylinder including a cylinder and a piston.

The robot device 1 is not limited to including the robot arm 2 having at least one fluid actuator M and the hand portion 4, and includes at least one fluid actuator M, a tool such as a drill bit, a switch, or the like. It may also include a robot arm to which elements other than the hand unit 4, such as a pressing member, are attached to the hand. Also, the robot apparatus 1 may be a walking robot, a wearable robot, or the like. Furthermore, the robot apparatus 1 may include only one joint, or may include only one fluid actuator M as an artificial muscle.

Although the embodiments for carrying out the present disclosure have been described above, the present disclosure is not limited to such embodiments, and can be implemented in various forms without departing from the gist of the present disclosure. Of course.

INDUSTRIAL APPLICABILITY The present disclosure is applicable to the robot device manufacturing industry and the like.

1 robot device, 2 robot arm (robot body), 10 fluid supply device, 13 pump (fluid supply source), 20 electromagnetic section, 40 spool, 100 control device, 151 first linear solenoid valve (pressure regulating valve), 152 second second Linear solenoid valve (pressure regulating valve), AM1 first artificial muscle, AM2 second artificial muscle, M fluid actuator (artificial muscle).

Claims (5)

the robot body and
at least one artificial muscle supplied with fluid to drive the robot body;
a pressure regulating valve having a spool and an electromagnetic section that applies a thrust force to the spool, the pressure regulating valve that regulates the pressure of the fluid from a fluid supply source based on the current supplied to the electromagnetic section and supplies the artificial muscle with the fluid;
a control device for controlling the current supplied to the electromagnetic unit based on a control signal for setting the fluid pressure supplied from the pressure regulating valve to the artificial muscle to a target pressure required for driving the robot main body;
A robotic device comprising
The control device controls a target current to the electromagnetic unit based on the target pressure, a dither amplitude for micro-vibrating the spool, and a change in the fluid pressure output from the pressure regulating valve due to the micro-vibration of the spool. generating the control signal based on a dither command value having a dither cycle that does not axially expand or contract the artificial muscle;
robotic device. The robot device according to claim 1,
The control device generates the control signal based on the target current and the dither command value at least when the amount of deviation between the current position of the robot body and the target position is equal to or less than a predetermined amount.
robotic device. The robot device according to claim 1 or 2,
The control device generates the control signal based on the target current and the dither command value at least when the target speed of the robot body is equal to or less than a predetermined speed.
robotic device. The robot device according to any one of claims 1 to 3,
the control signal is a PWM signal;
the dither period is longer than the period of the PWM signal;
robotic device. The robot device according to any one of claims 1 to 4,
The pressure regulating valve is a normally closed pressure regulating valve,
When the target current increase rate, which is the amount of increase in the target current per unit time, is equal to or greater than a predetermined increase rate, the control device reduces the dither amplitude compared to when the target current increase rate is less than the threshold value. to be smaller,
robotic device.

Images