Signal processing device and robot device

Abstract

This signal processing device performs drive control of a proportional control valve, which is for adjusting the pressure of a fluid that is supplied to a hydraulic actuator, in a movable part drive unit having the hydraulic actuator and the proportional control valve, and comprises: a calculation unit that calculates a drive control value of the proportional control valve through feedback loop calculation using a detection value detected by a state sensor for detecting the state of the proportional control valve and / or the hydraulic actuator, a target value of the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the hydraulic actuator; and a selection processing unit that selects the valve characteristic information and the actuator characteristic information used in the feedback loop calculation from among a plurality of candidates.

Description

Signal processing device, robot device

[0001] The present technology relates to a signal processing device and a robot device including the signal processing device, and more particularly to a control technology for a fluid pressure actuator provided for driving a movable part of the robot device.

[0002] As an actuator for driving a movable part such as an arm part of a robot device, a fluid pressure actuator driven by the pressure of a fluid, such as a cylinder actuator or a rotary actuator, is widely known.

[0003] Since the fluid pressure actuator has non-linear characteristics, the difficulty of control tends to be higher than that of an electric actuator. However, it has advantages such as passive softness and a high mass-to-output ratio, and can generate a large force without a speed reducer.

[0004] For a fluid pressure actuator, a drive control method is adopted in which the pressure of the supplied fluid is adjusted by opening and closing control of a proportional control valve (see, for example, Patent Document 1 below). The proportional control valve means a solenoid valve that can adjust the opening degree continuously in proportion to an input signal (drive signal: voltage or current).

[0005] In the drive control of a fluid pressure actuator, it is important to apply a control algorithm considering valve characteristics and actuator characteristics in order to obtain appropriate actuator operation. In the conventional drive control of a fluid pressure actuator, valve control is performed by a dedicated driver that applies a control algorithm obtained specifically for a combination of a specific valve and actuator.

[0006] Japanese Patent Publication No. 2019-517059

[0007] However, according to the conventional method as described above, when there are a plurality of types of robot devices with different combinations of valves and actuators, it is necessary to develop a driver for each of those robot devices, resulting in a lack of convenience.

[0008] This invention has been made in view of the above circumstances, and aims to realize a versatile driver that can be used to control the drive of a fluid pressure actuator, and that can be used with multiple types of robotic devices that have different combinations of valves and actuators.

[0009] The signal processing device according to the present invention is a signal processing device for controlling the drive of a proportional control valve in a movable part drive unit having a fluid pressure actuator and a proportional control valve for adjusting the pressure of the fluid supplied to the fluid pressure actuator, and comprises a calculation unit that calculates a drive control value of the proportional control valve by a feedback loop calculation using a detected value from a state sensor that detects the state of at least one of the proportional control valve and the fluid pressure actuator, a target value of the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the fluid pressure actuator, and a selection processing unit that selects the valve characteristic information and actuator characteristic information to be used in the feedback loop calculation from among a plurality of candidates. This makes it possible to switch the combination of valve characteristic information and actuator characteristic information used to calculate the drive control value of the proportional control valve depending on the combination of proportional control valve and fluid pressure actuator used in the movable part drive unit.

[0010] According to the present invention, it is possible to realize a versatile driver that can be used to control the drive of a fluid pressure actuator, and that can be used with multiple types of robotic devices that have different combinations of valves and actuators.

[0011] This figure shows an example configuration of a robot device as an embodiment of the present invention. This figure explains an example configuration of a movable part drive unit in an embodiment. This figure illustrates the flow characteristics of a 3-port proportional control valve. This figure explains an example configuration of a driver unit in an embodiment. This is a functional block diagram explaining the functions for realizing a valve drive control value calculation method as an embodiment. This figure shows an example of a control model assumed in the calculation of drive control values. This figure shows an example of a characteristic information group. This is a flowchart of the process related to storing robot type identification information. This is a flowchart of the process related to selecting characteristic information. This is an explanatory diagram of an example configuration of a movable part drive unit when a proportional control valve with a 2-port valve is used. This figure illustrates the flow characteristics of a proportional control valve with a 2-port valve. This figure shows an example of a characteristic information group when a 2-port valve is used. This is an explanatory diagram of an example configuration of a movable part drive unit when a proportional control valve with a 5-port valve is used. This figure illustrates the flow characteristics of a proportional control valve with a 5-port valve. This figure illustrates a control model when a 5-port valve is used. This is an explanatory diagram of a robot device as another example of performing reaction force calculation of a gripping part. This is an explanatory diagram of a fluid pressure actuator using a rotary actuator.

[0012] The embodiments of the present invention will be described below in the following order with reference to the attached drawings. <1. Overview of the Robot Device Configuration> <2. Example of a Movable Part Drive Unit> <3. About the Driver Unit> [3-1. Example of Driver Unit Configuration] [3-2. Calculation Method for Valve Drive Control Values ​​as an Embodiment] [3-3. Processing Procedure] <4. Another Example of a Movable Part Drive Unit> <5. About Force Feedback> <6. Modified Examples> <7. Summary of Embodiments>

[0013] <1. Overview of the Robot Device Configuration> Figure 1 is a diagram showing an example of the configuration of a robot device 50 as an embodiment of the present invention. The robot device 50 of the embodiment is configured as a so-called robot arm, and its movable parts are driven based on drive control by a computer to perform operations such as handling work such as grasping and moving an object, and various processing work such as welding and paint application. In this example, the robot device 50 is configured as a device that performs the above-mentioned handling work.

[0014] As shown in the figure, the robot device 50 comprises a base portion 51, a plurality of link portions 52, and a plurality of joint portions 53. The base portion 51 is the part that is placed on the floor or the like in a room in a facility where handling work is performed, and it is the part that supports the arm portion composed of the link portions 52 and joint portions 53. Various installation methods for the robot device 50 are conceivable, and other configurations besides installation on the floor are also possible, such as being installed on a wall or suspended from the ceiling.

[0015] Figure 1 shows an example in which the robot device 50 is configured to have four joints 53. One of the four joints 53 is provided at the base portion of the base portion 51, which is the root portion of the arm (hereinafter referred to as the "root portion") (hereinafter referred to as the "first joint"). The end of the joint 53 as the first joint is provided with a second joint 53 (hereinafter referred to as the "second joint") via a first link portion 52 (hereinafter referred to as the "first link"). The end of the joint 53 as the second joint is provided with a third joint 53 (hereinafter referred to as the "third joint") via a second link portion 52 (hereinafter referred to as the "second link"). Furthermore, the end of the joint 53 as the third joint is provided with a fourth joint 53 (hereinafter referred to as the "fourth joint") via a third link portion 52 (hereinafter referred to as the "third link").

[0016] Each joint 53 from the first to the third joint functions as a joint for partially bending the arm portion.

[0017] On the other hand, the joint portion 53, which is the fourth joint located at the very tip of the arm portion, functions as a joint portion for driving the gripping portion 54 that grips an object. As shown in the figure, the gripping portion 54 has two gripping members 54a and 54b. These gripping members 54a and 54b extend in the direction toward the tip of the arm portion, and it is possible to grip an object between these gripping members 54a and 54b. Specifically, in the gripping portion 54 of this example, gripping member 54a is a movable member driven by the joint portion 53 as the fourth joint, and gripping member 54b is a fixed member that is fixed to, for example, the tip of the link portion 52 as the third link and is therefore immovable. The gripping member 54a is attached to the joint portion 53 as the fourth joint so as to be rotatable in the axial direction of the joint portion 53, thereby making it possible to grip an object between the gripping members 54a and 54b.

[0018] In the robot device 50, a driver unit 1 and a movable part drive unit 2 are provided for each joint 53. The movable part drive unit 2 is a unit that drives the target movable part, and specifically, each movable part drive unit 2 is provided as a unit that drives one target joint 53. As will be explained later, each movable part drive unit 2 is configured to have a fluid pressure actuator (fluid pressure actuator 21) that drives the movable part (joint 53 in this example) and a proportional control valve (proportional control valve 22) that adjusts the pressure of the fluid supplied to the fluid pressure actuator.

[0019] The driver unit 1 is a signal processing unit that controls the drive of the proportional control valve in the movable part drive unit 2. Each driver unit 1 controls the drive of the proportional control valve in the corresponding movable part drive unit 2.

[0020] Figure 1 shows an example in which a driver unit 1 and a movable part drive unit 2 corresponding to the joint portion 53 as the first joint is provided at the aforementioned root portion of the base portion 51, and the other driver unit 1 and movable part drive unit 2 pairs are provided at the link portion 52 located on the front side (root side) of the corresponding joint portion 53. Note that the installation configuration of the driver unit 1 and movable part drive unit 2 pairs is not limited to this, and other installation configurations can be adopted, for example, by installing each pair inside the case of the corresponding joint portion 53.

[0021] In the robot device 50, a main controller 55 is provided, which is a computer device for controlling the operation of each driver unit 1 (the operation of driving and controlling the proportional control valves). Figure 1 shows an example in which the main controller 55 is installed in the base unit 51, but the installation position of the main controller 55 in the robot device 50 is not particularly limited.

[0022] Here, the robot device 50 can be one of three types: a type that performs predetermined (programmed) movements (first type), a type that operates autonomously based on detection signals from sensors such as cameras (second type), or a type that operates in response to user operation (third type). In the first type, the main controller 55 controls each driver unit 1 based on predetermined program data that defines the movement of the arm (including the movement of the gripping unit 54), thereby realizing the predetermined movement of the arm. In the second type, the main controller 55 controls each driver unit 1 so that the movement of the arm calculated based on detection signals from sensors such as cameras is obtained. In the third type, the main controller 55 controls each driver unit 1 based on operation signals input from an operation device (not shown), thereby realizing the movement of the arm in response to user operation.

[0023] In this example, communication between the main controller 55 and each driver unit 1 is performed via network communication such as EtherCAT (registered trademark), FL-net, or PROFINET. In this example, the main controller 55 and each driver unit 1 are connected to a common bus wiring Bs (not shown in Figure 1) for network communication, enabling them to communicate data with each other.

[0024] <2. Example of a Movable Part Drive Unit> Figure 2 is a diagram illustrating an example of the configuration of a movable part drive unit 2. In Figure 2, along with the example of the configuration of the movable part drive unit 2, a driver unit 1, a main controller 55, and a power transmission unit 3 for transmitting power from the movable part drive unit 2 to the joint 53 are also shown, all of which are provided in conjunction with the movable part drive unit 2.

[0025] As shown in the figure, the movable part drive unit 2 has a fluid pressure actuator 21 and a proportional control valve 22, as well as a position sensor 23 and a pressure sensor 24.

[0026] The fluid pressure actuator 21 is an actuator driven by the pressure of a fluid. In this example, a pneumatic actuator driven by the pressure of air is used as the fluid pressure actuator 21. The fluid pressure actuator 21 in this example is configured as a cylinder actuator and has a cylinder portion 21a and a piston portion 21b as shown in the figure. In the fluid pressure actuator 21 as a cylinder actuator, the piston portion 21b is displaced in the axial direction of the cylinder portion 21a in response to changes in the pressure inside the cylinder portion 21a, thereby generating translational motion power.

[0027] In this example, the fluid pressure actuator 21 is a double-acting fluid pressure actuator. For clarification, a double-acting fluid pressure actuator is a fluid pressure actuator having a push-side port Pa and a pull-side port Pb. By supplying pressure to the push-side port Pa and releasing pressure from the pull-side port Pb, the movable part of the actuator (i.e., the piston part 21b in this example) is driven to push out, and conversely, by releasing pressure from the push-side port Pa and supplying pressure to the pull-side port Pb, the movable part is driven to retract. At this time, the force pushing out or retracting the movable part is determined by the product of the magnitude of the fluid pressure and the pressure-receiving area of ​​the movable part.

[0028] In this example, the movable part drive unit 2 is provided with two types of proportional control valves 22: one for supplying and exhausting air to the push-side port Pa of the fluid pressure actuator 21, and another for supplying and exhausting air to the pull-side port Pb. Here, a proportional control valve refers to a solenoid valve that can adjust its opening degree steplessly in proportion to the input signal (drive signal: voltage or current).

[0029] In this example, three-port valves are used as the two types of proportional control valves 22, with one proportional control valve 22 controlling the supply and exhaust of air to the push-side port Pa, and the other proportional control valve 22 controlling the supply and exhaust of air to the pull-side port Pb. As shown in the figure, each proportional control valve 22 has ports P1, P2, and P3, with port P3 of one proportional control valve 22 connected to the push-side port Pa, and port P3 of the other proportional control valve 22 connected to the pull-side port Pb. In each proportional control valve 22, port P1 is connected to the supply air passage from the air pressure source Sa, and port P2 is connected to the exhaust passage. In this example, it is assumed that the air pressure source Sa is located outside the robot device 50, but the air pressure source Sa may also be located inside the robot device 50.

[0030] One proportional control valve 22, with port P3 connected to the push-side port Pa, is configured to close port P1 (air supply port) and connect port P2 (exhaust port) to port P3 when de-energized, while closing port P2 and connecting port P1 to port P3 when energized. The other proportional control valve 22, with port P3 connected to the pull-side port Pb, is configured to connect port P1 (air supply port) to port P3 and close port P2 (exhaust port) when de-energized, while closing port P1 and connecting port P2 to port P3 when energized.

[0031] As a result, in the movable part drive unit 2 of this example, when one of the proportional control valves 22 is energized and the other proportional control valve 22 is de-energized, air is supplied to the push-side port Pa and exhausted from the pull-side port Pb, driving the piston portion 21b of the fluid pressure actuator 21 in the push-side direction. Conversely, when one of the proportional control valves 22 is de-energized and the other proportional control valve 22 is energized, air is supplied to the pull-side port Pb and exhausted from the push-side port Pa, driving the piston portion 21b in the pull-side direction.

[0032] As described above, the proportional control valve 22 is designed to allow adjustment of its opening degree in proportion to the drive signal (drive voltage in this example). Figure 3 shows the relationship between the valve voltage (drive voltage) and the flow rate at port P3 for one of the proportional control valves 22 described above. As shown in the figure, the flow rate at port P3 is at a negative minimum value (i.e., maximum exhaust volume) when the proportional control valve 22 is not energized, and gradually increases to a positive value (i.e., shifts to the intake side) as the valve voltage rises, before reaching a positive maximum value (i.e., maximum intake volume). Although not shown in the figure, the relationship between the valve voltage and the flow rate at port P3 for the other proportional control valve 22 is the opposite of that shown in Figure 3. When the valve is not energized, the flow rate is at a positive maximum value (i.e., maximum supply volume), and gradually decreases to a negative value (i.e., shifts to the exhaust side) as the valve voltage rises, before reaching a negative minimum value (i.e., maximum exhaust volume).

[0033] In Figure 2, the power transmission unit 3 transmits the power of the fluid pressure actuator 21, which is manifested as the translational motion of the piston portion 21b, to the joint portion 53. In this example, since the joint portion 53 performs rotational motion to bend a part of the arm portion, the power transmission unit 3 is configured to convert the translational motion of the fluid pressure actuator 21 into rotational motion and transmit it. Specifically, the power transmission unit 3 in this example has a rack gear 3a connected to the piston portion 21b of the fluid pressure actuator 21, and a pinion gear 3b that rotates due to the translational motion of the rack gear 3a, and the joint portion 53 performs rotational motion due to the rotational power of the pinion gear 3b. Note that there are various possible configurations for converting translational motion into rotational motion, and the above configuration is not the only one that can be considered.

[0034] The position sensor 23 detects the operating position of the piston portion 21b in the fluid pressure actuator 21. The operating position here refers to the position of the part of the fluid pressure actuator that is moved by the application of fluid pressure (hereinafter referred to as the "operating part"). In the case of the fluid pressure actuator 21 in this example, which is configured as a cylinder actuator, the piston portion 21b, which is the operating part, performs translational motion, so the "operating position" indicates the position in the translational direction. If the operating part rotates, as in the rotary actuator described later, it will indicate the position in the rotational direction. Here, a linear encoder can be used as the position sensor 23 that corresponds to translational motion.

[0035] The position information detected by the position sensor 23 is input to the driver unit 1.

[0036] The pressure sensor 24 detects the fluid pressure in the fluid pressure actuator 21. In this example, where a double-acting fluid pressure actuator 21 is used, the pressure sensor 24 detects both the fluid pressure in the pushing-side fluid chamber Ra (air pressure in this example) and the fluid pressure in the pulling-side fluid chamber Rb in the cylinder portion 21a. Hereinafter, the fluid pressure in the pushing-side fluid chamber Ra will be referred to as "pressure P_a" and the fluid pressure in the pulling-side fluid chamber Rb will be referred to as "pressure P_b". The pressures P_a and P_b detected by the pressure sensor 24 are input to the driver unit 1.

[0037] In the robot device 50 of this embodiment, the main controller 55 provides information on a target value (hereinafter referred to as "target position τref") for the operating position detected by the position sensor 23. The driver unit 1 controls the drive of each proportional control valve 22 according to the target position τref instructed in this manner. Specifically, a feedback loop calculation is performed using the operating position (hereinafter referred to as "position τ") detected by the position sensor 23 as a feedback input to calculate the drive control value for each proportional control valve 22 in order to match position τ with the target position τref, and the drive control of each proportional control valve 22 is performed according to the calculated drive control value. In this example, the drive control of each proportional control valve 22 is performed by PID (Proportional-Integral-Differential) control. In this example, the pressures P_a and P_b detected by the pressure sensor 24 are also used for the drive control of the proportional control valve 22, but this will be explained in more detail later.

[0038] <3. About the Driver Unit> [3-1. Example of Driver Unit Configuration] Figure 4 is a diagram illustrating an example of the configuration of the driver unit 1. In addition, Figure 4 shows the example of the configuration of the driver unit 1 along with the position sensor 23, pressure sensor 24, and bus wiring Bs. As shown in the figure, the driver unit 1 includes a CPU (Central Processing Unit) 10, ROM (Read Only Memory) 11, and RAM (Random Access Memory) 12, as well as an A / D (Analog to Digital) conversion unit 13, a digital input unit 14, a D / A (Digital to Analog) conversion unit 15, a first communication unit 16, and a second communication unit 17.

[0039] The CPU 10 executes various processes based on programs stored in the ROM 11 and programs loaded from the ROM 11 into the RAM 12. In particular, the CPU 10 performs various processes related to the drive control of the proportional control valve 22.

[0040] Here, the characteristic information group 11a is stored in ROM 11, but the details of this characteristic information group 11a will be described later.

[0041] As shown in the figure, the pressures P_a and P_b detected by the pressure sensor 24 are digitally sampled by the A / D conversion unit 13 and input to the CPU 10 as digital values.

[0042] Further, the position τ detected by the position sensor 23 is input to the CPU 10 as a digital value via the digital input unit 14. Here, although the explanation was made on the premise that the output of the pressure sensor 24 is an analog signal and the output of the position sensor 23 is a digital signal, the output of the pressure sensor 24 may be a digital signal, and the output of the position sensor 23 may also be an analog signal.

[0043] Further, the CPU 10 is connected to the first communication unit 16 that performs data communication with an external device via the bus wiring Bs. Thereby, the CPU 10 can exchange data with the main controller 55. Information on the target position τ_ref from the main controller 55 is input to the CPU 10 via the first communication unit 16.

[0044] The CPU 10 calculates drive control values for the proportional control valves 22 by feedback loop operation with the position τ as a feedback input based on the target position τ_ref and the pressures P_a and P_b. In this example, since the proportional control valve 22 corresponding to the push-side port Pa (one of the proportional control valves 22 described above) and the proportional control valve 22 corresponding to the pull-side port Pb (the other proportional control valve 22 described above) are provided, the CPU 10 calculates two drive control values for driving and controlling the individual proportional control valves 22. Hereinafter, the drive control value for the proportional control valve 22 corresponding to the push-side port Pa is denoted as "drive control value u_a", and the drive control value for the proportional control valve 22 corresponding to the pull-side port Pb is denoted as "drive control value u_b".

[0045] As shown in the figure, the drive control values u_a and u_b are each converted into an analog signal by the D / A conversion unit 15 and output as valve drive signals.

[0046] The CPU 10 is connected to the second communication unit 17. The second communication unit 17 performs data communication with an external device (computer device) connected to an external terminal Te provided in the driver unit 1. As a result, the CPU 10 can exchange data with the external device connected to the external terminal Te via the second communication unit 17. For example, the external terminal Te may be a USB (Universal Serial Bus) terminal, and the second communication unit 17 may be a communication device that performs data communication by the USB communication method. Note that the second communication unit 17 may also be a communication device that performs wireless data communication with an external device.

[0047] By providing the second communication unit 17, various settings of the driver unit 1 can be performed using an external device different from the main controller 55, such as a general-purpose computer device such as a personal computer or a smartphone. Thereby, it is possible to perform various settings of the driver unit 1 for the driver unit 1 before being attached to the robot device 50 or the driver unit 1 removed from the robot device 50.

[0048] [3-2. Calculation Method of Valve Drive Control Value as an Embodiment] Here, in the drive control of the fluid pressure actuator 21, in order to obtain an appropriate actuator operation, it is important to apply a control algorithm considering valve characteristics (characteristics of the proportional control valve 22) and actuator characteristics (characteristics of the fluid pressure actuator 21). In the conventional drive control of the fluid pressure actuator 21, valve control has been performed by a dedicated driver that applies a control algorithm obtained specifically for a combination of a specific proportional control valve 22 and the fluid pressure actuator 21.

[0049] However, according to such a conventional method, when there are a plurality of types of robot devices 50 with different combinations of the proportional control valve 22 and the fluid pressure actuator 21, it is necessary to develop a driver for each of those robot devices 50, which lacks convenience.

[0050] Therefore, in this embodiment, a method is adopted in which the actuator characteristics and bubble characteristics to be used are selected according to the type of robot device 50 when calculating the valve drive control value.

[0051] Figure 5 is a functional block diagram illustrating the functions of the CPU 10 in order to realize the valve drive control value calculation method as described above. As shown in the figure, the CPU 10 has an arithmetic unit F1 and a selection processing unit F2.

[0052] The calculation unit F1 calculates the drive control value of the proportional control valve 22 by performing a feedback loop calculation using the detected value from a state sensor that detects the state of at least one of the proportional control valve 22 and the fluid pressure actuator 21, the target value of the state, valve characteristic information indicating the characteristics of the proportional control valve 22, and actuator characteristic information indicating the characteristics of the fluid pressure actuator 21.

[0053] The calculation unit F1 in this example performs calculations using a cascade control method that employs a control model with a double loop as a feedback loop calculation, which includes a master loop that feeds back the detected value (position τ) from the position sensor 23 and a slave loop that feeds back the detected value from the valve output state sensor that performs state detection related to the output of the proportional control valve 22. The valve output state sensor broadly refers to a sensor that performs state detection related to the output of a proportional control valve, such as detecting the pressure inside the cylinder of the fluid pressure actuator 21 or detecting the flow rate of the fluid output from the proportional control valve 22 to the fluid pressure actuator 21.

[0054] In this example, the calculation unit F1 uses the values ​​detected by the pressure sensor 24 (pressures P_a and P_b mentioned above) as the values ​​detected by the valve output state sensor. Therefore, the calculation of the valve drive control values ​​(u_a and u_b) is performed based on a control model having a double loop consisting of a master loop that feeds back the position τ and a slave loop that feeds back the pressure detected by the pressure sensor 24.

[0055] Figure 6 shows an example of a control model assumed in the calculation of drive control values ​​by the calculation unit F1. In the figure, the PID block 31 performs PID processing on the difference value (error value) between the target position τref and position τ. That is, it performs proportional, integral, and differential calculations in PID control on the difference value between the target position τref and position τ.

[0056] The PID control value Fref obtained from the PID processing of PID block 31 is input to fact block 32. Based on actuator characteristic information that shows the characteristics of the fluid pressure actuator 21, fact block 32 calculates the target value P_aref, which is the target value of pressure P_a, and the target value P_bref, which is the target value of pressure P_b. The force that pushes out or pulls in the movable part of the fluid pressure actuator 21 is basically determined by the product of the magnitude of the fluid pressure and the pressure-receiving area of ​​the movable part, but strictly speaking, the driving resistance of the movable part (ease of movement of the movable part) should also be taken into consideration. From this point of view, fact block 32 in this example uses information on the pressure-receiving area and driving resistance of the movable part (piston part 21b in this example) of the fluid pressure actuator 21 as actuator characteristic information to calculate the target values ​​P_aref and P_bref from the control value Fref. Furthermore, various known methods can be used to calculate the target values ​​P_aref and P_bref from the control value Fref using various actuator characteristic information, and the method is not limited to any particular method.

[0057] The target value P_aref calculated in fact block 32 is input to the push-side slave loop that feeds back pressure P_a, and the target value P_bref is input to the pull-side slave loop that feeds back pressure P_b.

[0058] In the push-side slave loop, the difference (error value) between the target value P_aref and the pressure P_a is processed by PID in the PID block 33a to calculate the PID control value q_aref. This control value q_aref is then input to the fvalve_a block 34a to calculate the push-side drive control value u_a.

[0059] The fvalve_a block 34a calculates the drive control value u_a from the control value q_aref based on valve characteristic information that shows the characteristics of the push-side proportional control valve 22. If the type of proportional control valve 22 is different, the characteristics of the correspondence between valve voltage and output port flow rate (hereinafter referred to as "flow rate characteristics") as illustrated in Figure 3 above may be different. For this reason, in the fvalve_a block 34a of this example, the flow rate characteristics information of the push-side proportional control valve 22 is used as valve characteristic information to calculate the drive control value u_a from the control value q_aref. Note that various known methods can be used for calculating the drive control value u_a from the control value q_aref using valve characteristic information, and it is not limited to a specific method. The same applies to the calculation of the drive control value u_b, which will be explained below.

[0060] The drive control value u_a calculated by the fvalve_a block 34a is input to the Valve_a block 35a. The Valve_a block 35a corresponds to the push-side proportional control valve 22, and the pressure P_a is obtained as its output. As can be understood from the above explanation, in a real-world configuration, this pressure P_a is detected by the pressure sensor 24.

[0061] In the pull-side slave loop, the difference (error value) between the target value P_bref and the pressure P_a is processed by PID in PID block 33b to calculate the PID control value q_bref. This control value q_aref is then input to fvalve_b block 34b to calculate the pull-side drive control value u_b.

[0062] The fvalve_b block 34b calculates the drive control value u_b from the control value q_bref based on valve characteristic information (flow characteristics in this example) that shows the characteristics of the pull-side proportional control valve 22.

[0063] The drive control value u_b calculated by the fvalve_b block 34b is input to the Valve_b block 35b. The Valve_b block 35b corresponds to the proportional control valve 22 on the pull side, and the output is pressure P_b. In a real-world configuration, this pressure P_b is also detected by the pressure sensor 24.

[0064] The Actuator block 36 corresponds to the fluid pressure actuator 21. Pressures P_a and P_b are applied to the Actuator block 36, and as a result, the position τ is output. As shown in the figure, the position τ is fed back to the preceding PID block 31 and used to calculate the error with the target position τref.

[0065] By calculating the drive control values ​​u_a and u_b using the control model described above, it is possible to control the drive of the proportional control valve 22 so that the position τ matches the target position τref.

[0066] In Figure 5, the selection processing unit F2 selects valve characteristic information and actuator characteristic information to be used in the feedback loop calculation from among several candidates. In particular, the selection processing unit F2 in this example selects different valve characteristic information for the fluid pressure actuator 21 during air supply and exhaust. Furthermore, the selection processing unit F2 in this example selects different valve characteristic information for the fluid pressure actuator 21 during push-side control and pull-side control.

[0067] The selection processing unit F2 selects the characteristic information to be used for the feedback loop calculation from among the multiple actuator characteristic information and valve characteristic information stored as the characteristic information group 11a described above.

[0068] Figure 7 shows an example of characteristic information group 11a. The characteristic information group 11a is information that indicates the corresponding actuator characteristic information and valve characteristic information for each type of robot device 50 with different actuator characteristics and valve characteristics. In the figure, the characteristic information group 11a is shown as an example when the number of corresponding types of robot devices 50 is n (where n is a natural number of 2 or more). The actuator characteristic information is stored from the actuator characteristic information for the first type of robot device 50 corresponding to the first type of robot device 50 to the nth type of actuator characteristics corresponding to the nth type of robot device 50.

[0069] As for valve characteristic information, as described above, four types of characteristic information are stored as valve characteristic information groups for each type of robot device 50, so that corresponding characteristic information can be selected for supply and exhaust, and for push-side control and pull-side control (indicated in the figure from the valve characteristic information group for type 1 robot to the valve characteristic information group for type n robot). The discharge-side characteristic information of the pull-side valve indicates the valve characteristics of the pull-side proportional control valve 22 when it discharges fluid pressure from the pull-side port Pb of the fluid pressure actuator 21, and the supply-side characteristic information of the pull-side valve indicates the valve characteristics of the pull-side proportional control valve 22 when it supplies fluid pressure to the pull-side port Pb.

[0070] In this example, the selection processing unit F2 selects characteristic information to be used for calculating drive control values ​​from among the actuator characteristic information and valve characteristic information stored as characteristic information group 11a, based on robot type identification information, which is information that identifies the type of robot device 50. Specifically, it selects characteristic information from the first to the nth type that corresponds to the robot type indicated by the robot type identification information.

[0071] In this example, the instruction regarding the robot type to the driver unit 1 is given from an external device via the second communication unit 17 mentioned above before attachment to the robot device 50. The CPU 10 in this example stores the robot type information instructed from the external device as robot type identification information in a predetermined storage device such as R0M11 (which is a writable memory in this example).

[0072] The selection processing unit F2 selects the appropriate valve characteristic information depending on whether push-side control or pull-side control is performed. Specifically, when push-side control is performed, the selection processing unit F2 selects the valve characteristic information corresponding to push-side control from the group of valve characteristic information corresponding to the robot type indicated by the robot type identification information, as the valve characteristic information used to calculate the drive control values ​​u_a and u_b. In this example, when push-side control is performed to drive the piston section 21b in the push direction, the push-side proportional control valve 22 is the supply side and the pull-side proportional control valve 22 is the discharge side. Therefore, the valve characteristic information selected for push-side control is the push-side valve supply-side characteristic information and the pull-side valve discharge-side characteristic information. For clarification, the selected push-side valve supply-side characteristic information is used in the calculation in the fvalve_a block 34a explained in Figure 6, and the pull-side valve discharge-side characteristic information is used in the calculation in the fvalve_b block 34b.

[0073] On the other hand, when pull-side control is performed, the selection processing unit F2 selects the valve characteristic information corresponding to the pull-side control from the group of valve characteristic information corresponding to the robot type indicated by the robot type identification information, specifically the push-side valve discharge-side characteristic information and the pull-side valve supply-side characteristic information. Similar to the push-side control, the selected push-side valve characteristic information is used in the calculation in the fvalve_a block 34a, and the pull-side valve characteristic information is used in the calculation in the fvalve_b block 34b.

[0074] Furthermore, the determination of whether it is push-side control or pull-side control can be made, for example, based on the relative magnitudes of pressures P_a and P_b. That is, if P_a > P_b, it can be determined to be push-side control; otherwise, it can be determined to be pull-side control.

[0075] Here, the characteristic information group 11a may be stored in the driver unit 1 from an external device via the second communication unit 17. By enabling storage from an external device in this way, it becomes possible to additionally store new characteristic information in the driver unit 1 when increasing the number of compatible models. In this case, the ROM 11 is, for example, a memory capable of data writing, such as an EEP (Electrically Erasable Programmable) ROM.

[0076] [3-3. Processing Procedure] Referring to the flowcharts in Figures 8 and 9, specific processing procedures to be performed to realize the drive control value calculation method as described above will be explained. The processes shown in Figures 8 and 9 are executed by the CPU 10 based on a program stored in a predetermined storage device such as ROM 11.

[0077] Figure 8 is a flowchart of the process related to the storage of robot type identification information. It is assumed that the CPU 10 is in a state where it can communicate with an external device via the second communication unit 17 when the process shown in this figure begins.

[0078] In step S101, the CPU 10 waits for instructions regarding the type of robot to be used. That is, it waits for instructions regarding the type of robot to be used from an external device that has been made available for communication via the second communication unit 17.

[0079] When an instruction is given for the type of robot to be used, the CPU 10 proceeds to step S102 and stores information indicating the instructed robot type as robot type identification information in, for example, RAM 12, and completes the series of processes shown in Figure 8.

[0080] Figure 9 is a flowchart of the process related to the selection of characteristic information. First, in step S201, the CPU 10 waits for the calculation to start. That is, it waits until the start conditions for the feedback loop calculation (calculation for calculating the drive control value) described in Figure 6 are met. Possible conditions for starting the calculation include, for example, an instruction to start the calculation from the main controller 55.

[0081] If it is determined in step S201 that the calculation should begin, the CPU 10 proceeds to step S202 and selects the actuator characteristic information identified from the stored robot type identification information as the characteristic information to be used. That is, as the actuator characteristic information to be used in the calculation of the fact block 32 shown in Figure 6 (calculation to find the target pressure values ​​P_aref and P_bref from the PID control value Fref), the CPU 10 selects the actuator characteristic information identified by the robot type identification information from among the actuator characteristic information in the characteristic information group 11a.

[0082] In step S203 following step S203, the CPU 10 determines whether it is push-side control or not (either push-side control or pull-side control). If it is push-side control, the CPU 10 proceeds to step S204 and selects the valve characteristic information corresponding to push-side control from the group of valve characteristic information identified from the stored robot type identification information as the usage characteristic information. Specifically, from the group of valve characteristic information identified from the robot type identification information, the CPU 10 selects the push-side valve supply-side characteristic information and the pull-side valve discharge-side characteristic information.

[0083] On the other hand, if step S203 determines that it is not push-side control (i.e., pull-side control), the CPU 10 proceeds to step S205 and selects the valve characteristic information corresponding to pull-side control from the group of valve characteristic information identified from the stored robot type identification information as the usable characteristic information. That is, from the group of valve characteristic information identified from the robot type identification information, the CPU 10 selects the push-side valve discharge-side characteristic information and the pull-side valve supply-side characteristic information.

[0084] Depending on whether the CPU 10 has performed the selection process in either step S204 or S205 described above, it proceeds to step S206. In step S206, the CPU 10 determines whether the calculation is complete or not, that is, whether the termination condition of the feedback loop calculation has been met. Possible termination conditions for the calculation in step S206 include, for example, a calculation completion instruction from the main controller 55 or the cessation of power supply to the driver unit 1.

[0085] If it is determined in step 2206 that the calculation is not yet complete, the CPU 10 returns to step S203. This ensures that during the feedback loop calculation, appropriate switching of valve characteristic information is performed according to whether it is push-side or pull-side control.

[0086] If the CPU 10 determines in step S206 that the calculation is complete, it finishes the series of processes shown in Figure 9.

[0087] <4. Another example of a movable part drive unit> Here, an example was given above in which a 3-port type valve is used as the proportional control valve 22, but the proportional control valve 22 is not limited to a 3-port type valve.

[0088] Figure 10 is an explanatory diagram illustrating an example configuration of the movable part drive unit 2 when a proportional control valve 22 using a two-port valve is used. In this case as well, the movable part drive unit 2 is equipped with a position sensor 23 and a pressure sensor 24, but these are omitted from the illustration.

[0089] In the case of a two-port valve, it is not possible to switch between supply and exhaust air as in a three-port valve, so as shown in the figure, a total of four proportional control valves 22 are provided: two on the pushing side and two on the pulling side. The two proportional control valves 22 on the pushing side have port P2 connected to the pushing port Pa of the fluid pressure actuator 21, and one of the proportional control valves 22 has port P1 connected to the supply air passage from the air pressure source Sa, functioning as a valve for supplying air to the pushing port Pa (referred to as the first valve). The other proportional control valve 22 on the pushing side has port P1 connected to the exhaust passage, functioning as a valve for exhausting air from the pushing port Pa (referred to as the second valve). The two proportional control valves 22 on the pulling side have port P2 connected to the pulling port Pb of the fluid pressure actuator 21, and one of the proportional control valves 22 has port P1 connected to the exhaust passage, functioning as a valve for exhausting air from the pulling port Pb (referred to as the third valve). Furthermore, the other proportional control valve 22 on the pull side has its port P1 connected to the air supply passage from the air pressure source Sa, and functions as an air supply valve for the pull side port Pb (referred to as the fourth valve).

[0090] Figure 11 illustrates the flow rate characteristics of a proportional control valve 22 using a two-port valve, specifically the flow rate change characteristics at port P2 in relation to the valve voltage. As shown in the figure, the proportional control valve 22 using a two-port valve has the characteristic that the flow rate at port P2 increases from 0 to the maximum value in proportion to the valve voltage. In other words, for the first and fourth valves for air supply, the amount of air supplied from port P2 to the fluid pressure actuator 21 side increases from 0 to the maximum value in proportion to the increase in valve voltage, and for the second and third valves on the exhaust side, the amount of exhaust from the fluid pressure actuator 21 to port P2 increases from 0 to the maximum value in proportion to the increase in valve voltage.

[0091] When using a two-port valve, the four proportional control valves 22, which are the first to fourth valves as described above, will be individually driven and controlled. Therefore, as shown in Figure 12, the characteristic information group 11a stores the characteristic information for each of the first to fourth valves as a valve characteristic information group for each robot type.

[0092] When using a two-port valve, the push-side control involves supplying air to the push-side port Pa with the first valve while keeping the second valve closed, exhausting air from the pull-side port Pb with the third valve, and keeping the fourth valve closed. Conversely, the pull-side control involves keeping the first valve closed while exhausting air from the push-side port Pa with the second valve, supplying air to the pull-side port Pb with the third valve, and keeping the fourth valve closed.

[0093] As can be understood from this point, in this case, the selection processing unit F2, when performing push-side control, selects the first valve characteristic information (air supply to the push-side port Pa) and the third valve characteristic information (exhaust from the pull-side port Pb) from the group of valve characteristic information corresponding to the robot type identified by the robot type identification information. On the other hand, when performing pull-side control, it selects the second valve characteristic information (exhaust from the push-side port Pa) and the fourth valve characteristic information (air supply to the pull-side port Pb) from the group of valve characteristic information corresponding to the robot type identified by the robot type identification information.

[0094] In this case, when push-side control, the CPU 10 drives the proportional control valve 22 as the first valve using a drive control value u_a calculated using the first valve characteristic information, and drives the proportional control valve 22 as the third valve using a drive control value u_b calculated using the third valve characteristic information. When push-side control, the proportional control valves 22 as the second and fourth valves are de-energized (closed). Also, when pull-side control, the CPU 10 drives the proportional control valve 22 as the second valve using a drive control value u_a calculated using the second valve characteristic information, and drives the proportional control valve 22 as the fourth valve using a drive control value u_b calculated using the fourth valve characteristic information. When pull-side control, the proportional control valves 22 as the first and third valves are de-energized (closed).

[0095] Furthermore, a 5-port valve can also be used as the proportional control valve 22. Figure 13 is an explanatory diagram of an example configuration of the movable part drive unit 2 when a 5-port valve is used as the proportional control valve 22. In this case as well, the movable part drive unit 2 is equipped with a position sensor 23 and a pressure sensor 24, but these are omitted from the illustration here.

[0096] In the case of a 5-port valve, a single proportional control valve 22 can supply and exhaust air to the push-side port Pa and to the pull-side port Pb. As shown in the figure, in the movable part drive unit 2 in this case, port P4 of the proportional control valve 22 of the 5-port valve is connected to the push-side port Pa, port P2 is connected to the pull-side port Pb, port P1 is connected to the air supply passage from the air pressure source Sa, and ports P3 and P5 are connected to the exhaust passage.

[0097] Figure 14 illustrates the flow rate characteristics of a proportional control valve 22 using a 5-port valve, specifically the flow rate change characteristics on the supply side of ports P4 and P2 in response to the valve voltage. As shown in the figure, in the case of a 5-port valve, the flow rate on the supply side of port P4 (i.e., the flow rate of air supplied to the push-side port Pa) and the flow rate on the supply side of port P2 (the flow rate of air supplied to the pull-side port Pb) are both set to 0 when the valve voltage is at a predetermined value. The flow rate on the supply side of port P4 gradually increases towards its maximum value as the valve voltage decreases from the predetermined value. In the region where the valve voltage exceeds the predetermined value, the flow rate on the supply side of port P4 remains at 0. Similarly, the flow rate on the supply side of port P2 (the flow rate of air supplied to the pull-side port Pb) gradually increases towards its maximum value as the valve voltage increases from the predetermined value. In the region where the valve voltage falls below the predetermined value, the flow rate on the supply side of port P2 remains at 0. Although not shown in the diagram, the exhaust flow rate characteristics from port P5 are similar to those of port P2 described above. They remain at 0 until the valve voltage reaches the predetermined value, and then gradually increase towards the maximum value as the valve voltage rises from the predetermined value. Similarly, the exhaust flow rate characteristics from port P3 are similar to those of port P4 described above. They gradually increase towards the maximum value as the valve voltage decreases from the predetermined value, and remain at 0 in the region where the valve voltage is above the predetermined value.

[0098] When using a 5-port valve, the valve drive control value is calculated based on the ratio of the pushing pressure P_a to the pulling pressure P_b (hereinafter referred to as "pressure ratio P_r"). Figure 15 illustrates a control model when using a 5-port valve. As shown in the figure, the slave loop in this case is a single loop and has a PID block 33 and an fvalve block 34.

[0099] In this case, the fact block 32 takes the control value Fref obtained from the PID block 31 as input and calculates the target value P_rref of the pressure ratio P_r based on the actuator characteristic information. Then, the PID block 33 in the slave loop performs PID processing on the difference value (error value) between the target value P_rref and the pressure ratio P_r, and takes the control value qref obtained from the PID processing as input, and the fvale block 34 calculates the drive control value u of the proportional control valve 22 using a 5-port valve based on the valve characteristic information.

[0100] Even when using a 5-port valve, the valve characteristics to be considered in calculating the drive control value u are the four characteristics: the characteristics related to the supply of air to the push-side port Pa and the characteristics related to the exhaust from the pull-side port Pb during push-side control, and the characteristics related to the supply of air to the pull-side port Pb and the characteristics related to the exhaust from the push-side port Pa during pull-side control.

[0101] Therefore, even when using a 5-port valve, the characteristic information group 11a is the same as that shown in Figure 7 above. That is, as a valve characteristic information group for each robot type, a characteristic information group 11a is used that has four types of valve characteristic information, each showing the characteristics related to the supply of air to the push-side port Pa during push-side control (i.e., the flow rate characteristics of port P4 during push-side control) and the exhaust from the pull-side port Pb (i.e., the flow rate characteristics of port P3 during push-side control), and the characteristics related to the supply of air to the pull-side port Pb during pull-side control (i.e., the flow rate characteristics of port P5 during pull-side control) and the exhaust from the push-side port Pa (i.e., the flow rate characteristics of port P1 during pull-side control).

[0102] In this case, the selection processing unit F2 makes the following selections regarding the valve characteristic information used to calculate the drive control value u. Specifically, during push-side control, it selects from the group of valve characteristic information identified by the robot type identification information valve characteristic information indicating the characteristics related to air supply to the push-side port Pa during push-side control, and valve characteristic information indicating the characteristics related to exhaust from the pull-side port Pb during push-side control. Similarly, during pull-side control, it selects from the group of valve characteristic information identified by the robot type identification information valve characteristic information indicating the characteristics related to air supply to the pull-side port Pb during pull-side control, and valve characteristic information indicating the characteristics related to exhaust from the push-side port Pa during pull-side control.

[0103] <5. Force Feedback> When the fluid pressure actuator 21 is used to drive the gripping part 54, force feedback can be provided by calculating the reaction force of gripping the object based on the change in pressure in the fluid pressure actuator 21 and feeding this reaction force back to the operator.

[0104] Conventionally, force feedback was implemented by providing a torque sensor on the gripping section 54 and calculating the reaction force from the detection signal of the torque sensor. However, providing a dedicated sensor for reaction force calculation in this way leads to an increase in the number of parts and costs of the robot device 50, which is undesirable.

[0105] Therefore, it is conceivable to adopt a configuration in which the reaction force is calculated based on the detection signal of the pressure sensor 24.

[0106] Figure 16 is an explanatory diagram of a robot device 50 as an alternative example in which the reaction force is calculated based on the detection signal of the pressure sensor 24. In Figure 16, only the gripping part 54, the joint part 53 as the fourth joint corresponding to the gripping part 54, and the set of the movable part drive unit 2 and driver part 1 provided in conjunction with it are shown. The other components of the robot device 50 are the same as those shown in Figure 1 and are therefore omitted from the illustration. In addition to the configuration shown, the movable part drive unit 2 in the figure also has one or more proportional control valves 22, similar to the movable part drive unit 2 described above.

[0107] As shown in the figure, in the robot device 50 as an alternative example, a driver unit 1A is provided as a driver unit corresponding to the joint 53 as the fourth joint. This driver unit 1A differs from driver unit 1 in that a CPU 10A is provided instead of CPU 10. In Figure 16, only the CPU 10A, ROM 11, A / D conversion unit 13, and digital input unit 14 are shown as the configuration of driver unit 1A, but driver unit 1A also has the other configurations that driver unit 1 has (RAM 12, D / A conversion unit 15 to second communication unit 17).

[0108] CPU 10A differs from CPU 10 in that it has a calculation unit F1A instead of the calculation unit F1. The calculation unit F1A has the function of the calculation unit F1 as described above, as well as the function of a reaction force calculation unit F11.

[0109] The reaction force calculation unit F11 calculates the reaction force of the gripping part for gripping an object based on the value detected by the pressure sensor 24. In this example, the reaction force calculation unit F11 uses the actuator characteristic information in the characteristic information group 11a to calculate the reaction force. As mentioned above, the actuator characteristics may differ depending on the type of robot device 50, so the accuracy of the reaction force calculation is improved by using actuator characteristic information (for example, the pressure-receiving area and driving resistance of the movable part) that corresponds to the type of robot device 50.

[0110] To this end, the reaction force calculation unit F11 selects actuator characteristic information to be used for calculating the reaction force from among several candidates. Specifically, it selects actuator characteristic information corresponding to the robot type identified by the robot type identification information mentioned above from among the actuator characteristic information stored in the characteristic information group 11a.

[0111] Regarding the method for calculating the reaction force of the gripping portion 54 based on the pressure of the fluid pressure actuator 21 (in this example, at least one of pressure P_a and pressure P_b), various known methods have been proposed, and the method is not limited to a specific one. For example, a method using position τ information detected by the position sensor 23 along with the pressure can be considered. When the reaction force changes, the amount of change in position τ changes in relation to the amount of change in pressure, so based on this principle, the reaction force can be calculated based on pressure and position τ (see, for example, "Japanese Patent No. 6650153").

[0112] <6. Modifications> The embodiments are not limited to the specific examples described above, and various modified configurations can be adopted. For example, although an example using a fluid pressure actuator 21 with a cylinder actuator was given above, a fluid pressure actuator 21A with a rotary actuator, as illustrated in Figure 17, can also be used instead of the cylinder actuator. As shown in the figure, in the fluid pressure actuator 21A as a rotary actuator, a movable part 25 is provided which operates in response to the application of fluid pressure, and which has a shaft portion 25b and a pressure receiving portion 25a extending radially from the shaft portion 25b. Power is generated when the movable part 25 rotates around the shaft portion 25b by adjusting the fluid pressure applied to the pressure receiving portion 25a.

[0113] In Figure 17, a double-acting actuator having a push-side port Pa and a pull-side port Pb is shown as an example of the fluid pressure actuator 21A, but a single-acting actuator can also be used. The same applies to the fluid pressure actuator 21 using a cylinder actuator.

[0114] Furthermore, while the above example illustrates the case where a fluid pressure actuator is a pneumatic actuator, it is also conceivable to use a liquid pressure actuator that operates using hydraulics, for example, as the fluid pressure actuator.

[0115] Furthermore, while the above example illustrates the use of cascaded control (double-loop control) of position and pressure for the feedback loop calculation to determine the valve drive control value, it is also possible to use other combinations of double-loop control besides position and pressure, such as double-loop control of position and flow rate.

[0116] Furthermore, the control is not limited to double-loop control; single-loop control, such as a position loop only, is also conceivable. The single-loop control envisioned here involves performing a calculation using actuator characteristic information and valve characteristic information as a feedback loop calculation of the drive control value.

[0117] Furthermore, while the position sensor 23 was given as an example of a state sensor according to the present invention above, other sensors that sense states other than position, such as the angle, torque, and speed of a movable part, can also be used as state sensors. In other words, the target value of the feedback loop calculation may also be an element other than position, such as angle, torque, or speed.

[0118] Furthermore, when using fluid pressure actuators or proportional control valves with significantly different characteristics, it may be necessary to select not only actuator characteristic information and valve characteristic information, but also the control model itself according to the type of robot.

[0119] Furthermore, it is conceivable that temperature and atmospheric pressure information could be used in the feedback loop calculations for determining drive control values. In that case, the detection information from temperature and atmospheric pressure sensors would be used in the feedback loop calculations.

[0120] Furthermore, while the above example illustrates a case where the robot device 50 performs handling tasks, various applications for the robot device 50 are conceivable. For example, in addition to industrial applications for performing the various processing tasks mentioned above, applications such as assistance robots, hobby robots, and surgical support robots that assist in surgery in medical facilities such as hospitals can be cited. In the case of a surgical support robot, the movable part drive unit 2 drives movable parts related to the operation of surgical instruments such as scopes (endoscopy) and forceps. The present invention can be broadly and suitably applied to robot devices equipped with movable parts.

[0121] <7. Summary of Embodiments> As described above, the signal processing device (1, 1A) as an embodiment is a signal processing device that controls the drive of a proportional control valve in a movable part drive unit (2) having a fluid pressure actuator (21, 21A) and a proportional control valve (22) that adjusts the pressure of the fluid supplied to the fluid pressure actuator, and comprises a calculation unit (F1, F1A) that calculates the drive control value of the proportional control valve by a feedback loop calculation using a detected value from a state sensor (position sensor 23) that detects the state of at least one of the proportional control valve and the fluid pressure actuator, a target value of the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the fluid pressure actuator, and a selection processing unit (F2) that selects the valve characteristic information and actuator characteristic information to be used in the feedback loop calculation from among a plurality of candidates. Therefore, as a driver for controlling the drive of a fluid pressure actuator, it is possible to realize a versatile driver that can be used with multiple types of robotic devices that have different combinations of valves and actuators.

[0122] Furthermore, in the signal processing device as an embodiment, the calculation unit performs calculations using a cascade control method that employs a control model with a double loop as a feedback loop calculation, which includes a master loop that feeds back detected values ​​from a position sensor that detects the operating position of a fluid pressure actuator, and a slave loop that feeds back detected values ​​from a valve output state sensor that detects the state related to the output of a proportional control valve. By adopting a cascade control method, the system becomes less susceptible to disturbances, and the accuracy and stability of the control can be improved. If it were a single position loop, even if the pressure inside the fluid pressure actuator changes due to the influence of a disturbance, corrective control would not be applied unless that influence manifested as a position error. With cascade control using both position and pressure, the degree of valve opening and closing is controlled immediately in response to pressure changes, so no position error occurs, and the influence on the master loop is suppressed. Therefore, the system becomes less susceptible to disturbances, and the accuracy and stability of the control can be improved.

[0123] Furthermore, in the signal processing device as an embodiment, the selection processing unit selects different valve characteristic information for use in feedback loop calculations depending on whether the fluid pressure actuator is supplying or exhausting air. This makes it possible to appropriately switch the valve characteristic information used in calculating the drive control value in response to cases where the valve characteristics when supplying air to the fluid pressure actuator and the valve characteristics when exhausting air from the fluid pressure actuator differ for a proportional control valve. Consequently, the calculation accuracy of the drive control value of the proportional control valve can be improved, and the control accuracy of the fluid pressure actuator can be improved.

[0124] Furthermore, in the signal processing device as an embodiment, the fluid pressure actuator is a double-acting fluid pressure actuator, and the movable part drive unit has one or more proportional control valves for adjusting the fluid pressure applied to the pushing-side fluid chamber and the pulling-side fluid chamber of the fluid pressure actuator, respectively. The selection processing unit selects different valve characteristic information for use in feedback loop calculations depending on whether the fluid pressure actuator is being pushed or pulled. This makes it possible to appropriately switch the valve characteristic information used in calculating the drive control value in response to cases where the valve characteristics when performing pushing-side control and when performing pulling-side control of the fluid pressure actuator are different for the proportional control valve. Therefore, the calculation accuracy of the drive control value of the proportional control valve can be improved, and the control accuracy of the fluid pressure actuator can be improved.

[0125] Furthermore, in the signal processing device as an embodiment, the fluid pressure actuator is an actuator that drives a movable part that acts as a gripping part for gripping an object, and the calculation unit has a reaction force calculation unit (F11) that calculates the reaction force of object gripping at the gripping part based on the detected value of a pressure sensor that detects the fluid chamber pressure of the fluid pressure actuator and actuator characteristic information, and the reaction force calculation unit selects the actuator characteristic information to be used for calculating the reaction force from among several candidates. As a result, the calculation of the reaction force used for force feedback is performed accurately, taking into account the characteristics of the fluid pressure actuator in the target robot device. Therefore, the accuracy of force feedback can be improved. In addition, since it is not necessary to provide a force sensor on the gripping part when calculating the reaction force, the number of parts of the robot device can be reduced and costs can be reduced.

[0126] Furthermore, in the signal processing device as an embodiment, the drive control of a proportional control valve in a movable part drive unit that drives a movable part related to surgical instrument operation in a surgical support robot is performed. High precision control is required for surgical instrument operation. With the above configuration, it is possible to calculate a highly precise drive control value that takes into account the characteristics of the actuator and valve, which is suitable for the drive control of the movable part that requires such high precision control.

[0127] The robot device (50) as an embodiment includes a movable part (joint part 53), a movable part drive unit having a fluid pressure actuator for driving the movable part and a proportional control valve for adjusting the pressure of the fluid supplied to the fluid pressure actuator, and a signal processing unit (driver unit 1, 1A) that controls the drive of the proportional control valve in the movable part drive unit, which includes a calculation unit that calculates a drive control value for the proportional control valve by a feedback loop calculation using a detected value from a state sensor that detects the state of at least one of the proportional control valve and the fluid pressure actuator, a target value for the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the fluid pressure actuator, and a selection processing unit that selects the valve characteristic information and actuator characteristic information to be used in the feedback loop calculation from among a plurality of candidates. This makes it possible to realize an inexpensive robot device in which a general-purpose driver is used as the driver for controlling the movement of the movable part, rather than a specially developed driver.

[0128] 1, 1A Driver unit (signal processing unit) 2 Movable part drive unit 3 Power transmission unit 3a Rack gear 3b Pinion gear 10, 10A CPU 11a Characteristic information group Te External terminal 21 Fluid pressure actuator 21a Cylinder part 21b Piston part Pg Push-side port Pb Pull-side port Ra Push-side fluid chamber Rb Pull-side fluid chamber 22 Proportional control valve 23 Position sensor 24 Pressure sensor Bs Bus wiring Sa Air pressure supply source F1, F1A Calculation unit F2 Selection processing unit 50 Robot device 51 Base part 52 Link part 53 Joint part 54 Gripping part 54a, 54b Gripping member 55 Main controller F11 Reaction force calculation unit 21A Fluid pressure actuator 25 Movable part 25a Pressure receiving part 25b Shaft part

Claims

1. A signal processing device for controlling the drive of a proportional control valve in a movable part drive unit having a fluid pressure actuator and a proportional control valve for adjusting the pressure of the fluid supplied to the fluid pressure actuator, comprising: a calculation unit that calculates a drive control value for the proportional control valve by a feedback loop calculation using a detected value from a state sensor that detects the state of at least one of the proportional control valve and the fluid pressure actuator, a target value of the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the fluid pressure actuator; and a selection processing unit that selects the valve characteristic information and the actuator characteristic information to be used in the feedback loop calculation from a plurality of candidates.

2. The signal processing device according to claim 1, wherein the calculation unit performs calculations using a cascade control method that employs a control model having a double loop as the feedback loop calculation, the master loop which feeds back a value detected by a position sensor that detects the operating position of the fluid pressure actuator, and the slave loop which feeds back a value detected by a valve output state sensor that detects the state related to the output of the proportional control valve.

3. The signal processing device according to claim 1 or 2, wherein the selection processing unit selects different characteristic information for the valve characteristic information used in the feedback loop calculation when the fluid pressure actuator is supplying air and when it is exhausting air.

4. The signal processing device according to any one of claims 1 to 3, wherein the fluid pressure actuator is a double-acting fluid pressure actuator, the movable part drive unit has one or more proportional control valves for adjusting the fluid pressure applied to the pushing fluid chamber and the pulling fluid chamber of the fluid pressure actuator, respectively, and the selection processing unit selects different valve characteristic information for use in the feedback loop calculation depending on whether the fluid pressure actuator is being pushed or pulled.

5. The signal processing device according to any one of claims 1 to 4, wherein the fluid pressure actuator is an actuator that drives a movable part as a gripping part for gripping an object, the calculation unit has a reaction force calculation unit that calculates the reaction force of gripping an object at the gripping part based on the detected value of a pressure sensor that detects the fluid chamber pressure of the fluid pressure actuator and the actuator characteristic information, and the reaction force calculation unit selects the actuator characteristic information to be used in calculating the reaction force from among a plurality of candidates.

6. A signal processing device according to any one of claims 1 to 5, which controls the drive of a proportional control valve in a movable part drive unit that drives a movable part related to the operation of a surgical instrument in a surgical support robot.

7. A robotic device comprising: a movable part; a movable part drive unit having a movable part, a fluid pressure actuator for driving the movable part, and a proportional control valve for adjusting the pressure of the fluid supplied to the fluid pressure actuator; a signal processing unit for controlling the drive of the proportional control valve in the movable part drive unit, the calculation unit calculating a drive control value for the proportional control valve by a feedback loop calculation using a detected value from a state sensor that detects the state of at least one of the proportional control valve and the fluid pressure actuator, a target value for the state, valve characteristic information indicating the characteristics of the proportional control valve, and actuator characteristic information indicating the characteristics of the fluid pressure actuator; and a selection processing unit that selects the valve characteristic information and actuator characteristic information to be used in the feedback loop calculation from among a plurality of candidates.

Images