Robot control method, robot control system, and program

Abstract

To improve walking stability when transferring a bipedal walking robot from a walking time to a both-leg standing-leg time when controlling the robot in accordance with an attitude of an operator.SOLUTION: In a robot control method, operator attitude information indicating an attitude of an operator is input and operation control of a bipedal walking robot which can switch a mode between a walking mode and a standing-leg mode is executed on the basis of the operator attitude information. When the mode is switched from the walking mode to the standing-leg mode (in the case of YES in step S4) in the operation control, reference attitude information indicating a reference attitude of the robot is calculated on the basis of initial attitude information indicating an initial attitude corresponding to an initial attitude of the operator and foot positional information indicating the positions of foot parts of the present robot (step S5). The attitude of the robot is controlled on the basis of the operator attitude information and the calculated reference attitude information. (step S6 to S8).SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a robot control method, a robot control system, and a program.

Background Art

[0002] Patent Document 1 describes a remote operation system aimed at enabling an object to be manipulated to perform flexible, stable, and high-speed operations that are difficult to achieve only by autonomous control for maintaining the standing posture of the object to be manipulated.

[0003] The remote operation system described in Patent Document 1 includes an object to be manipulated having a mechanism that enables a predetermined part to operate by the remote operation of an operator, an operation detection device that detects the motion state of the operator's body, a stimulus presentation device that operates to present a stimulus to the operator, and a control device that controls the operations of the object to be manipulated and the stimulus presentation device. Here, the object to be manipulated has a humanoid shape including parts corresponding to the arm, torso, and leg parts of the human body, and a foot that contacts a predetermined ground surface in a standing state is provided at the part corresponding to the leg part. Further, the stimulus presentation device is controlled to operate so as to apply a pressure stimulus as foot sole force sensation information corresponding to the magnitude of the ground contact pressure to the part of the operator's foot sole corresponding to the action site of the ground contact pressure that becomes the reaction force received by the foot due to the ground contact of the foot. Then, after the presentation of the pressure stimulus, the object to be manipulated is controlled to operate the parts corresponding to these arm and torso parts so as to follow the operations of the operator's arm and torso parts based on the detection results of the operation detection device.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Thus, in the technique described in Patent Document 1, a pressure stimulus is presented to the operator to avoid the operator from falling by himself / herself. However, in the technique described in Patent Document 1, although it seems that the real-time control of the waist position and walking can be compatible in order for the operator to always instruct the whole body posture of the object to be operated including the legs, in reality, it is difficult for the operator to walk while receiving a pressure stimulus as feedback, and the walking stability is lacking. Therefore, in a biped walking robot, it is desirable to improve the walking stability. In particular, in a biped walking robot, since it is easy to fall when shifting from the walking state to the double-leg standing state, control for improving the walking stability in this case is desirable.

[0006] The present disclosure has been made to solve such problems, and an object thereof is to provide a robot control method, a robot control system, and a program capable of improving the walking stability when shifting from the walking state to the double-leg standing state when controlling a biped walking robot according to the posture of an operator.

Means for Solving the Problems

[0007] A robot control method according to an aspect of the present disclosure inputs operator posture information indicating the posture of an operator, and executes operation control of a biped walking robot capable of switching modes between a walking mode and a standing mode based on the operator posture information. When the operation control is switched from the walking mode to the standing mode, based on initial posture information indicating an initial posture of the robot corresponding to the initial posture of the operator and foot position information indicating the current position of the feet of the robot, reference posture information indicating a reference posture of the robot is calculated, and includes standing transition control for controlling the posture of the robot based on the operator posture information and the calculated reference posture information.

[0008] A robot control system according to an aspect of the present disclosure includes an input unit that inputs operator posture information indicating the posture of an operator, and a control unit that executes operation control of a bipedal robot capable of switching between a walking mode and a standing mode based on the operator posture information. When switching from the walking mode to the standing mode, the control unit calculates reference posture information indicating a reference posture of the robot based on initial posture information indicating an initial posture of the robot corresponding to the initial posture of the operator and foot position information indicating the current position of the feet of the robot, and controls the posture of the robot based on the operator posture information and the calculated reference posture information.

[0009] A program according to an aspect of the present disclosure causes a computer to input operator posture information indicating the posture of an operator, and execute robot control for controlling the operation of a bipedal robot capable of switching between a walking mode and a standing mode based on the operator posture information. The operation control includes, when switching from the walking mode to the standing mode, calculating reference posture information indicating a reference posture of the robot based on initial posture information indicating an initial posture of the robot corresponding to the initial posture of the operator and foot position information indicating the current position of the feet of the robot, and controlling the posture of the robot based on the operator posture information and the calculated reference posture information. The operation control includes control at the time of standing transition for controlling the posture of the robot.

[0010] According to any one of the above robot control method, the above robot control system, and the above program, when controlling a bipedal robot in accordance with the posture of an operator, it is possible to improve the walking stability when transitioning from the walking state to the two-foot standing state.

Advantages of the Invention

[0011] According to the present disclosure, it is possible to provide a robot control method, a robot control system, and a program capable of improving the walking stability when transitioning from the walking state to the two-foot standing state when controlling a bipedal robot in accordance with the posture of an operator.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Modes for Carrying Out the Invention

[0013] Hereinafter, the present invention will be described through embodiments of the invention, but the invention according to the claims is not limited to the following embodiments. Also, not all of the configurations described in the embodiments are necessarily essential as means for solving the problems.

[0014] (Embodiment) FIG. 1 is a block diagram showing a configuration example of a robot control system according to the present embodiment, and FIG. 2 is a schematic diagram showing the appearance of the robot control system of FIG. 1. As shown in FIGS. 1 and 2, the robot control system 1 according to the present embodiment can include a biped walking robot (hereinafter simply referred to as a robot) 10 and a control device 20.

[0015] Robot 10 is a robot that can perform bipedal walking and can switch modes between a walking mode and a standing mode, and it can be any robot that can be controlled to reproduce the actions of the operator U from the control device 20. For example, as shown in FIG. 2, robot 10 can be a humanoid robot with a waist on its legs and a configuration similar to that of a human body such as arms and a head, but it is not limited to this. For example, it is not necessary to have all of the arms and head. The standing mode refers to a mode in which the robot stands on both feet, that is, the two-legged standing mode. The standing mode and the walking mode can be, for example, modes in which the mainly targeted center-of-gravity posture and waist angle are made different from each other.

[0016] Robot 10 can include a control unit 11, a sensor group 12, an input unit 13, and a drive unit 14. The control device 20 is a device for the operator U to control (operate) the actions such as the waist posture of robot 10, and can include a control unit 21, a sensor group 22, and an output unit 23.

[0017] The control unit 11 is a part that controls the operation of robot 10, and can also be a part that controls the entire robot 10. The control unit 11 can be realized, for example, by an integrated circuit (Integrated Circuit), and can be realized, for example, by a processor such as an MPU (Micro Processor Unit) or a CPU (Central Processing Unit), a working memory, and a non-volatile storage device. By storing a control program executed by the processor in this storage device and having the processor read out and execute the program in the working memory, the functions of robot 10 can be achieved.

[0018] The sensor group 12 is composed of a plurality of sensors that detect the position, angle, etc. at various locations of the robot 10 and transmit them to the control unit 11. Regardless of the type of each sensor, as long as the posture of the robot 10 can be directly or computationally detected by the sensor group 12. Although the illustration of the sensors included in the sensor group 12 is omitted in FIG. 2, they can be arranged at positions corresponding one-to-one with the sensors of the sensor group 22. However, only some of the sensors included in the sensor group 12 can be arranged at the corresponding positions, or none of them can be arranged at the corresponding positions, as long as various information for operations as described later can be obtained computationally.

[0019] The input unit 13 inputs information related to the operation from the operating device 20 and transmits it to the control unit 11. The information input by the input unit 13 includes operator posture information indicating the posture of the operator U. The operator posture information can be information indicating the posture of the operator U itself, or can also be information indicating displacements at various locations such as the feet and waist of the operator U. Based on the operator posture information, the control unit 11 can execute the motion control of the robot 10 while also referring to the information related to the posture of the robot 10 detected by the sensor group 12. This motion control includes the control of the posture of the robot 10. Note that the positions of the control unit 11 and the input unit 13 shown in FIG. 2 are merely illustrative.

[0020] The drive unit 14 is a part that drives the movable parts arranged at various locations of the robot 10 and can include motors and the like provided at various locations. The drive unit 14 can move the movable parts arranged at various locations according to the control from the control unit 11 to change the posture of the robot 10. The control unit 21 is a part that controls the entire operating device 20. For example, it can adopt a configuration similar to that of the control unit 11, but the control program is for fulfilling the functions of the operating device 20.

[0021] The sensor group 22 is composed of a plurality of sensors that detect the position, angle, etc. of the operator U at various locations and transmit them to the control unit 21. Regardless of the type of each sensor, as long as the posture of the operator U can be directly or computationally detected by the sensor group 22 and the movement of the operator U can be tracked. The positions where each sensor of the sensor group 22 can be attached can include the waist and the soles of the feet of the operator U, but can also be arranged at more joints. For example, as shown in FIG. 2, the operating device 20 can include sensors 22a to 22e worn by the operator U at various locations such as the soles of the feet, knees, and waist as the sensor group 22. The positions where each sensor is attached will vary depending on the range and accuracy in which the movement of the robot 10 is to be reflected.

[0022] Also, although detailed description is omitted, some or all of the sensors included in the sensor group 22 can be those not worn by the operator U. For example, it can be composed of a camera that captures the operator U as a moving image or a still image at a predetermined interval and an analysis device that analyzes the captured image data to perform motion tracking. In this way, the operating device 20 can include a motion capture device that can acquire the positions and angles of various parts of the operator U. The motion capture device is not limited to the method in which the operator U wears the sensor group 22 as exemplified here, and any method that can acquire the movement of the operator U, such as a method of estimating from camera images, is acceptable. "Various parts of the operator U" can refer to each joint of the operator U. Also, only the positions of each joint can be acquired, and the angles can be acquired by inverse kinematics calculation.

[0023] The control unit 21 obtains, for example, by generating operator posture information, which is information to be transmitted to the robot 10 side, from the information regarding the posture of the operator U detected by the sensor group 22, and passes it to the output unit 23. The output unit 23 outputs the operator posture information received from the control unit 21 to the robot 10, that is, to the input unit 13, thereby giving a walking command. In FIG. 2, an example is given in which the control unit 21 and the output unit 23 are provided on a belt with the waist sensor 22a attached, but these arrangements are not limited to the exemplified ones.

[0024] In this way, the control device 20 can be a controller for outputting a walking command indicated by the operator's posture information to the robot 10, and can be, for example, a shoe-shaped controller, a controller operated by a foot pedal or by hand. The walking command can include commands such as the foot to step out during walking, the stepping width, and the direction.

[0025] Note that both the input unit 13 and the output unit 23 can be wired or wireless communication units. Further, the control device 20 can be configured such that the control unit 21 is omitted and the output unit 23 directly outputs the detection result of the sensor group 22 to the robot 10.

[0026] The robot control system 1 according to the present embodiment has the following features. That is, when the control unit 11 is switched from the walking mode to the standing mode, based on the initial posture information of the robot 10 and the current foot position information of the robot 10, the control unit 11 calculates reference posture information indicating the reference posture of the robot 10. The above-mentioned initial posture information of the robot 10 is information indicating the initial posture of the robot 10 corresponding to the initial posture of the operator U. The above-mentioned current foot position information of the robot 10 is information indicating the position of the foot of the current robot 10, and can be acquired by the sensor group 12, but can also be obtained from the immediately preceding control value.

[0027] Then, the control unit 11 controls the posture of the robot 10 based on the calculated reference posture information and the input operator posture information. Of course, the operator posture information used here can be current information, that is, the latest information. Note that the information regarding the position, angle, etc. of the feet, waist, etc. of the robot 10 or the operator U can be information indicating the relative position, angle, etc. from a certain reference. Further, the control when switching from the walking mode to the standing mode can be referred to as standing transition control.

[0028] Next, an example of the robot control method in the robot control system 1 will be described with reference to FIGS. 3 to 6. FIG. 3 is a flowchart for explaining an example of the robot control method in the robot control system 1, and FIG. 4 is a flowchart following FIG. 3. FIG. 5 is a schematic diagram for explaining the posture information that can be used in the robot control methods of FIGS. 3 and 4, and FIG. 6 is a schematic diagram for explaining an example of the interpolation process in the robot control method.

[0029] In the robot control system 1, first, the initial posture of the robot 10 is determined (step S1). This determination can be made, for example, as follows. At the start of operation or before the start of operation, the operating device 20 measures the position (X dw , Y dw , Z dw ) and angles (Α dw , Β dw , Γ dw ) of the waist of the operator U. At this time, for setting the initial posture of the robot 10 described later, the operator U bends the knees so that the operator can move the feet with a margin without being tied to the waist position and measures the position (X dw , Y dw , Z dw ) and angles (Α dw , Β dw , Γ dw ) of the waist. The operating device 20 outputs the measured result to the robot 10 as operator initial posture information indicating the initial posture.

[0030] The robot 10 inputs this operator initial posture information and sets initial posture information indicating the initial posture of the robot 10 corresponding to this operator initial posture information. The initial posture information to be set is, as exemplified by INI in FIG. 5, the initial foot sole position (the position (x dl , y dl , z dl ) of the left foot sole SL, the position (x dr , y dr , z dr ) of the right foot sole SR) and the height z dwThis can be done. In step S1, in this way, the initial posture of the robot 10 can be determined in advance before motion control. Note that the information (parameters) included in the operator initial posture information and the information (parameters) included in the initial posture information of the robot 10 corresponding to the operator posture information are information indicating the posture of the waist and information indicating the position and orientation of the center of gravity, respectively. Although they are not information at the same position, since the relationship between the waist and the center of gravity is known, it can be said that they are corresponding information to each other.

[0031] The initial posture information will be described with reference to INI in FIG. 5. First, the reference positions of the right foot sole SR and the left foot sole SL are determined in advance, and the xy coordinates of the reference position of the zero moment point (zmp) are defined to be at the midpoint on the xy plane. The height Z of the center of gravity G in the initial posture dw is the distance l from the right foot sole SR to the center of gravity G f and the distance from the left foot sole SL to the center of gravity G (since it is symmetric left and right in the placement posture, the distance l f ) are set to be shorter than the original leg length. Taking the right foot sole SR as an example, the length of the leg of the robot 10 itself can be defined as the distance from the right foot sole SR to the waist when the leg is extended. Being shorter than the original leg length means that the distance from the right foot sole SR to the waist is shorter than this distance, that is, the knee or the part corresponding to the knee in the leg is in a slightly bent state. The same applies to the left foot sole SL.

[0032] In this way, the initial posture information can be set. As described above, since the height Z of the center of gravity G is set in a state where the robot 10 bends the knee or the part corresponding to the knee, an initial posture with a margin that allows the robot 10 to perform follow-up control by the movement of the waist even when an external force acts on the robot 10 can be set. dw

[0033] Next to step S1, the control unit 11 performs a mode switching determination. There are several examples of switching determination. As an example, when automatically switching the mode, it can be determined based on whether the switching has been made. Also, in the standing mode, when the operator U gives a new walking instruction, the mode can be shifted to the walking mode. In this case, it can be determined that the mode has switched from the standing mode to the walking mode by receiving the walking instruction. Further, in the walking mode, when a condition such as no next walking instruction within a certain time after the end of walking is satisfied, the mode can be shifted to the standing mode. In this case, it can be determined that the mode has switched from the walking mode to the standing mode when the above condition is satisfied. In addition to this, the switching between the walking mode and the standing mode may be performed based on an operation such as a manual switch provided on the control device 20.

[0034] In this way, the mode switching can be executed according to a switching instruction that is at least one of the walking or standing instruction indicated by the operator posture information and the instruction based on the switching operation by the operator U. And the switching from the walking mode to the standing mode can also be executed when there is no switching instruction indicating walking for a predetermined period after the end of walking in the walking mode. The switching from the standing mode to the walking mode can be executed based on at least one of the standing instruction indicated by the operator posture information and the switching operation by the operator U in the standing mode.

[0035] In the examples of FIGS. 3 and 4, next to step S1, the control unit 11 first performs a mode determination (step S2). When the determination result is the standing mode (YES in step S3), it is determined whether the mode has changed from the previous time (step S4). On the other hand, even when the determination result is the walking mode (NO in step S3), it is determined whether the mode has changed from the previous time (step S10). Whether the mode has changed from the previous time can be determined by storing the result of the previous mode determination and referring to it.

[0036] When in the walking mode and when the mode has changed from the previous time (YES in step S10), command values for the center-of-gravity position and the waist angle, which are target values for stably achieving the instructed walking, are generated in real time on the walking calculator side (step S11), and a command value for interpolating between the current posture of the robot 10 and the target value is generated (step S12). When in the walking mode and when the mode has not changed from the previous time (NO in step S10), command values for the center-of-gravity position and the waist angle, which are target values for stably achieving the instructed walking, are generated in real time on the walking calculator side (step S13). The processes of steps S11 and S13 can use existing technologies. The process of step S12 is the same as the process described later for step S7.

[0037] When in the stance mode and when the mode has changed from the previous time (YES in step S4), the control unit 11 calculates the reference posture of the robot 10 (step S5). As can be understood from the determination in step S4, the process of step S5 is executed only for the first time after the transition to the stance mode. In step S5, the reference posture is the reference zmp (x nz , y nz , z nz ) of the robot, the center-of-gravity height z nw including the center-of-gravity position (x nw ,y nw ,z nw ) and the waist angle ([[]] α nw , β nw , γ nw ), which can be determined by calculating from the current sole posture (left sole position (x nl , y nl , z nl ), left foot yaw angle θ nl , right sole position (x nr , y nr , z nr ), right foot yaw angle θ nr ). These values of the reference posture are obtained by the following formulas (1) to (5). Note that α nw ,β nwcan all be set to 0, and as can be seen from FIG. 5, the center-of-gravity position x nw ,y nw , can directly use the values of x nz ,y nz .

[0038]

Number

[0039] Referring to REF in FIG. 5, the setting of the reference positions represented by the above formulas (1) to (5) will be described. The xy coordinates of the target ZMP are set to the midpoint of the xy coordinates of both feet, and the height (z coordinate) of the center of gravity G is set to a height where l f does not change, and the angle of the waist W changes to the middle of the directions of both feet. For example, when the left foot flat SL is stepped out in the lower right direction of FIG. 5, as illustrated by REF in FIG. 5, the xy coordinates of the target ZMP move in the lower right direction of FIG. 5 as the feet open. The xy coordinates of the center of gravity G are calculated from the target ZMP and the whole body posture, and the z coordinate of the center of gravity G becomes a value that maintains the posture of bending the legs by dropping the waist by the amount the foot width has opened, and the angle of the waist W is set to face the middle direction because the directions of both feet are open.

[0040] Next, the control unit 11 adds to the value of the reference posture calculated in step S5 the displacement of the current operator U's waist posture (position ( X nw , Y nw , Z nw ), angle (Α nw , Β nw , Γ nw )) from the Those reference value are operated onA value obtained by multiplying the conversion constant D considering the physical difference between the operator U and the robot 10 is applied (added) to generate a command value (step S6). The above reference value refers to the initial waist posture of the operator U explained in step S1. Even when in the standing leg mode and there has been no mode change since the previous time (NO in step S4), the same processing as in step S6 is performed (step S9). In step S9, the value calculated when shifting to the standing leg mode can be obtained by referring to it if it is stored as the value of the reference posture. Thus, the position (x z , y z , 0) of the target ZMP of the robot 10, the height z of the center of gravity G, and the waist angles (α, β, γ) can be calculated by the following equations. x z = x nw + D(X nw - X dw ), α = α nw + D(Α nw - Α dw ), y z = y nw + D(Y nw - Y dw ), β = β nw + D(Β nw - Β dw ), z = z nw + D(Z nw - Z dw ), γ = γ nw + D(Γ nw - Γ dw )

[0041] Then, based on the position of this target ZMP, with the upper body posture and the height z of the center of gravity G as constraint conditions, the x and y coordinates of the position of the center of gravity G can be obtained. Also, the waist position can be automatically determined from the information including the position (x, y, z) of this center of gravity G.

[0042] Also, immediately after the transition to the stance mode, there are differences between the target ZMP and the waist posture of the robot 10 in the previous mode and the previous values of the robot 10. Therefore, if the displacement is made all at once during this period, there is a possibility of falling. Accordingly, after the process of step S6, an interpolation value is generated for an arbitrarily determined transition time between the current posture and the target posture, and the command value is gradually transitioned (step S7). Immediately after the transition to the walking mode, there is also a possibility of falling for the same reason. Therefore, an interpolation command value is similarly generated, and the command value is gradually transitioned (step S12).

[0043] Regarding the generation of the command value for controlling the posture including the transition timing at the time of mode switching, it will be described with reference to FIG. 6. In FIG. 6, an interval including the double-support phase operated in the stance mode, the walking phase (for one step) operated in the walking mode, and the double-support phase operated in the stance mode is extracted and illustrated. Further, in FIG. 6, the waist posture Y nw of the operator U is shown by the upper broken line, and the y nz of the target ZMP of the robot 10 generated by the walking calculation is shown by the middle solid line, and the y z of the target ZMP of the robot 10 used for control is shown in the lower part. Note that in FIG. 6, only the y-direction of the ZMP is shown, but the same applies to the x-direction of the ZMP and the waist postures α, β, γ.

[0044] As shown in the lower part of FIG. 6, during the double-support phase, it is controlled so that the waist posture of the operator U is scaled to the size of the robot 11 and reflected in the posture. During the subsequent walking phase, it is controlled so that the stable waist posture trajectory indicated by the target ZMP generated by the walking calculation is obtained. During the subsequent double-support phase, it is controlled in the same manner as the first double-support phase. However, immediately after the mode switching, since the target ZMP cannot maintain a stable posture, between points 51 and 52 and between points 53 and 54 in FIG. 6, control is performed using the interpolated command value shown by the dashed-dotted line. For such interpolation, in the case of the transition to the stance mode, the latest waist posture of the operator U is always acquired as the target posture during the transition and continuously updated. Thereby, a stable and visually discomfort-free state transition can be achieved.

[0045] After the processes of steps S7, S9, S12, and S13, in each case, the control unit 11 reflects the command value on the robot 11, controls the whole-body posture so as to achieve the command value (step S8), and ends the process. In step S8, the posture of the upper body can be fixed or controlled based on the joint angles of the operator U acquired by a motion tracker or the like.

[0046] As described above, in this embodiment, when switching from the walking mode to the standing mode (when YES in step S4), based on the initial posture information indicating the initial posture of the robot 10 corresponding to the initial posture of the operator U and the foot position information indicating the current position of the feet (equal feet) of the robot 10, the reference posture information indicating the reference posture of the robot 10 is calculated (step S5). And in this embodiment, based on the operator posture information and the reference posture information, the posture of the robot 10 is controlled (steps S6 to S8). Of course, the operator posture information used here can be the current information, that is, the latest information. Also, as described above, the operator posture information can include the position and angle of the waist of the operator U. In this case, when controlling the posture of the robot 10, control is performed based on the current waist posture (position and angle) of the operator U indicated by the operator posture information. Also, as described above, the posture information of the robot 10 (information indicating the posture, such as the reference posture information, information indicating the posture of the robot 10 during the standing transition control) can include the position of the zmp of the robot 10, the height of the center of gravity G, and the angle of the waist.

[0047] As described above, according to this embodiment, when controlling the bipedal walking robot 10 in accordance with the posture of the operator U, the switching from the walking mode to the standing mode can be performed seamlessly, and even if the posture of both feet changes during walking, the waist posture of the operator U can be reflected without the risk of falling. Therefore, according to this embodiment, the walking stability when transitioning from walking to standing on both feet can be improved.

[0048] Supplement on such effects. Not limited to the present embodiment, as a method of controlling a bipedal robot, there is a method in which an operator wears a motion tracker or the like to reflect the movement of the whole body. At this time, regarding the movement of the legs, it is common to give a walking instruction by a foot pedal or a shoe-shaped sensor, and the robot calculates and follows a leg movement and a waist posture that achieves walking without falling using the command value. On the other hand, when standing on both feet, a method of performing an operation with high expressive power similar to human movement by reflecting the waist posture of the operator is also conceivable. In order to achieve both of the above, it is necessary to switch in real time between a waist posture with high stability during walking and a waist posture with high expressive power when standing on both feet. In the present embodiment, the walking mode and the standing mode are seamlessly switched, and particularly during the transition to standing, even if the posture of both feet changes due to walking, the waist posture of the operator U can be reflected without the risk of falling. For example, when the foot width is widened, if the waist height before opening is set as the reference posture, the feet may stretch out and the movement of the waist in the front, rear, left, and right directions may not be reflected. However, in the present embodiment, such a problem can be solved, and the mode can be switched dynamically and safely.

[0049] Also, the control unit 11 in the present embodiment can also prevent the target zmp displacement value exceeding a predetermined maximum value for the robot 10 from being reflected in the control of the robot 10. Thereby, it is possible to prevent the movement of the waist position of the operator U from exceeding the stable range of the robot 11. Also, a mode for determining whether or not to execute such processing may be provided, and it may be determined whether or not to execute such processing according to the mode selected by the operator U.

[0050] Also, in step S7 described above, an example in which a transition time is set for the waist posture displacement at the time of mode change was given. As an alternative example, the transition may be executed based on a speed criterion. In this case, a target speed and a target acceleration are set, and the transition is made to the command value of the transition destination while satisfying the values. Thereby, when the current waist posture and the target waist posture are close, the transition can be made in a shorter time.

[0051] In addition, when the control unit 11 in the present embodiment indicates that the operator's posture information has a deviation from the initial posture by a predetermined value or more, the control unit 11 can also prevent the operator's posture information from being reflected in the control of the robot 10. Alternatively, the control unit 21 can prevent the operator's posture information that satisfies such conditions from being output from the output unit 23 to the robot 10.

[0052] For example, due to measurement errors of the sensor group 22 or the like, abnormal values may be input to the input of the waist posture. If this is adopted as a command, the robot 10 is likely to fall. Therefore, when a command value with a deviation from the initial posture by a certain amount (outside the predetermined range) is input, it is detected, and the robot waits until the value returns within the certain range. In addition, a mode for determining whether to execute such processing may be provided, and it may be determined whether to execute such processing according to the mode selected by the operator U. The modes to be selected can include the following fixed mode and transition mode. In the fixed mode, when a value outside the predetermined range is input, the robot 10 detects it, discards the command value, and fixes it to the command value immediately before going outside the predetermined range. In the transition mode, when it returns within the specified range, it is detected, the target posture is calculated from the command value, an interpolation value at an arbitrary time is created between the current posture and the target posture, and the robot is gently transitioned. Thereby, even when the command value after the return deviates from the command value immediately before going outside the predetermined range, it can be restored stably and without a sense of discomfort.

[0053] In addition, although the description has been made on the premise that the control unit 11 on the robot 10 side calculates the control value (command value) for controlling the posture of the robot 10, this is not the only case. For example, the control device 20 worn by the operator U may calculate the above control value, or alternatively, a relay control device may be provided separately, and that control device may calculate the above control value. That is, the calculation of the reference posture information among the controls described as the features of the control unit 11, or all of the controls, can also be executed by the control unit 21 on the control device 20 side via the output unit 23 and the input unit 13, or by a relay control device.

[0054] Also, each of the above-described robot 10 and control device 20 can have a hardware configuration including, for example, a processor, a memory, and a communication interface. These devices are realized by the processor reading and executing a program stored in the memory. Such a program includes a set of instructions (or software code) for causing a computer to perform one or more functions described in the embodiments when loaded into the computer. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of non-limiting example, the computer-readable medium or tangible storage medium includes random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD), or other memory technologies, CD-ROM, digital versatile disc (DVD), Blu-ray (registered trademark) disc, or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage, or other magnetic storage devices. The program may be transmitted on a transitory computer-readable medium or a communication medium. By way of non-limiting example, the transitory computer-readable medium or communication medium includes electrical, optical, acoustic, or other forms of propagated signals.

[0055] Note that the present invention is not limited to the above-described embodiments and can be appropriately modified without departing from the spirit. For example, the robot control system is not limited to the configuration of FIG. 1 and can be constructed as a system in which functions are appropriately distributed or aggregated. Also, the bipedal robot 10, the sensor group of the control device 20, etc. are not limited to the appearance illustrated in FIG. 2. For example, the bipedal robot may be a robot composed only of a lower body without a head or an upper body.

Explanation of Reference Numerals

[0056] U... operator, 1... robot control system, 10... bipedal robot (robot), 11... control unit, 12... sensor group, 13... input unit, 14... drive unit, 20... control device, 21... control unit, 22... sensor group, 23... output unit.

Claims

1. Inputting operator posture information indicating the posture of an operator, executing operation control of a bipedal robot capable of switching modes between a walking mode corresponding to a swing phase and a standing mode corresponding to a stance phase based on the current operator posture information, wherein the operation control includes, when switching from the walking mode to the standing mode, calculating reference posture information indicating the current reference posture of the robot based on initial posture information indicating the initial posture of the robot corresponding to the initial posture at the start of operation of the operator and foot position information indicating the current position of the feet of the robot, and controlling the posture of the robot based on the current operator posture information and the calculated reference posture information, including stance transition control; wherein the operator posture information includes the position and angle of the waist of the operator, wherein the reference posture information and the information indicating the posture of the robot in the stance transition control include the position of the zero moment point, the height of the center of gravity, and the angle of the waist of the robot, A robot control method.

2. An input unit for inputting operator posture information indicating the posture of an operator, a control unit for executing operation control of a bipedal robot capable of switching modes between a walking mode corresponding to a swing phase and a standing mode corresponding to a stance phase based on the current operator posture information, comprising, wherein when the control unit switches from the walking mode to the standing mode, it calculates reference posture information indicating the current reference posture of the robot based on initial posture information indicating the initial posture of the robot corresponding to the initial posture at the start of operation of the operator and foot position information indicating the current position of the feet of the robot, and executes stance transition control for controlling the posture of the robot based on the current operator posture information and the calculated reference posture information, wherein the operator posture information includes the position and angle of the waist of the operator, wherein the reference posture information and the information indicating the posture of the robot in the stance transition control include the position of the zero moment point, the height of the center of gravity, and the angle of the waist of the robot, A robot control system.

3. On a computer, inputting operator posture information indicating the posture of an operator, a robot control for executing operation control of a bipedal robot capable of switching modes between a walking mode corresponding to a swing phase and a standing mode corresponding to a stance phase based on the current operator posture information, The operation control includes, when switching from the walking mode to the standing mode, calculating reference posture information indicating the current reference posture of the robot based on initial posture information indicating the initial posture of the robot corresponding to the initial posture at the start of operation of the operator and foot position information indicating the current position of the feet of the robot, and controlling the posture of the robot based on the current operator posture information and the calculated reference posture information, which is control during standing transition. A program for executing robot control, wherein the operator posture information includes the position and angle of the waist of the operator, the reference posture information and the information indicating the posture of the robot in the control during standing transition include the position of the zero moment point, the height of the center of gravity, and the angle of the waist of the robot, program.

Images