JP2018008019A - Walking support robot and walking support method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a walking support robot capable of performing walking support for a user, in a more comfortable manner.SOLUTION: A walking support robot for supporting walking of a user comprises: a body part 11; a handle part 12 which is provided on the body part, and can be held by a user; a detection part 13 for detecting a load applied to the handle part; a movement device 14 which has a rotating body 16, controls rotation of the rotating body according to the load detected by the detection part and moves the walking support robot; and a load tendency data generation part 15 for generating load tendency data indicating a tendency of a load applied to the handle part, on the basis of past load data applied to the handle part, which are acquired during movement of the walking support robot. The movement device comprises an actuator 22 for controlling a rotation amount of the rotating body, on the basis of the load detected by the detection part, and the load tendency data generated by the load tendency data generation part.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a walking support robot and a walking support method that support a user's walking.

  2. Description of the Related Art In recent years, a guidance mobile robot has been developed that calculates a moving speed based on user input and draws the user's hand to guide the user to a destination (see, for example, Patent Document 1).

  Patent Document 1 discloses a guidance robot that calculates a target speed of a base in accordance with an input to a gripper by a guided person and moves the base at the calculated target speed.

JP 2010-271911 A

  In the robot of Patent Document 1, there is still room for improvement from the viewpoint of providing more comfortable user walking support.

  This indication solves the above-mentioned subject, and provides a walking support robot and a walking support method which can perform walking support of a more comfortable user.

A walking support robot according to an aspect of the present disclosure is provided.
A walking support robot that supports user walking,
The main body,
A handle portion provided on the main body portion and grippable by the user;
A detection unit for detecting a load applied to the handle unit;
A moving device that has a rotating body and moves the walking support robot by controlling the rotation of the rotating body according to the load detected by the detection unit;
Based on past load data applied to the handle part acquired during the movement of the walking support robot, a load tendency data generation unit that generates load tendency data indicating a load tendency applied to the handle part;
With
The moving device includes an actuator that controls the amount of rotation of the rotating body based on the load detected by the detection unit and the load tendency data generated by the load tendency data generation unit.

A walking support method according to an aspect of the present disclosure includes:
A walking support method for supporting a user's walking using a walking support robot,
Detecting a load applied to a handle portion of the walking support robot by a detection unit;
Generating load trend data indicating a tendency of a load applied to the handle unit based on past load data applied to the handle unit acquired during the movement of the walking support robot;
Controlling the amount of rotation of the rotating body provided in the moving device of the walking support robot based on the load detected by the detection unit and the load tendency data;
including.

  As described above, according to the walking support robot and the walking support method of the present disclosure, more comfortable walking support for the user can be performed.

FIG. 1 is an external view of a walking assistance robot according to Embodiment 1 of the present disclosure. FIG. 2 is a diagram illustrating a state in which the user is walking with the walking support by the walking support robot according to the first embodiment of the present disclosure. FIG. 3 is a diagram illustrating a handle load detection direction detected by the detection unit according to the first embodiment of the present disclosure. FIG. 4 is a control block diagram illustrating a main control configuration in the walking assistance robot according to the first embodiment of the present disclosure. FIG. 5 is a control block diagram illustrating a control configuration of walking support of the walking support robot according to the first embodiment of the present disclosure. FIG. 6 is a diagram illustrating a load tendency map according to the first embodiment of the present disclosure. FIG. 7 is an exemplary flowchart of load trend data generation processing of the walking assistance robot according to the first embodiment of the present disclosure. FIG. 8 is an example of input waveform information of the handle load. FIG. 9A is a diagram illustrating an example of waveform information of load data in the Fz direction when the user is moving straight ahead. FIG. 9B is a diagram showing frequency components of the load data in the Fz direction shown in FIG. 9A. FIG. 10A is a diagram illustrating an example of waveform information of load data in the My direction when the user is moving straight ahead. FIG. 10B is a diagram showing frequency components of load data in the My direction shown in FIG. 10A. FIG. 11A is a diagram illustrating an example of waveform information of load data in the Fz direction when the user turns right. FIG. 11B is a diagram showing frequency components of load data in the Fz direction shown in FIG. 11A. FIG. 12 is an exemplary flowchart of the movement intention estimation process of the walking assistance robot according to the first embodiment of the present disclosure. FIG. 13A is a diagram illustrating an example of waveform information of load data in the Fz direction when the user is moving straight ahead. FIG. 13B is a diagram showing waveform information obtained by filtering fluctuation frequency components from the waveform information of the load data in the Fz direction shown in FIG. 13A. FIG. 14A is a diagram illustrating an example of waveform information of load data in the My direction when the user is moving straight ahead. 14B is a diagram illustrating waveform information obtained by filtering the fluctuation frequency component from the waveform information of the load data in the My direction illustrated in FIG. 14A. FIG. 15 is an exemplary flowchart of the driving force calculation process of the walking assistance robot according to the first embodiment of the present disclosure. FIG. 16 is a diagram illustrating a load tendency map according to the second embodiment of the present disclosure. FIG. 17 is an exemplary flowchart of load trend data generation processing of the walking assistance robot according to the second embodiment of the present disclosure. FIG. 18 is an exemplary flowchart of the movement intention estimation process of the walking assistance robot according to the second embodiment of the present disclosure. FIG. 19A is a diagram illustrating an example of waveform information of current load data in the Mz direction when the user is moving straight ahead. FIG. 19B is a diagram illustrating an average load value of past load data in the Mz direction. FIG. 19C is a diagram illustrating an example of waveform information of corrected load data in the second embodiment of the present disclosure. FIG. 20A is a diagram illustrating an example of waveform information of past load data in the Mz direction when the user is traveling straight ahead. 20B is a diagram showing an average load value of past load data in the Mz direction shown in FIG. 20A. FIG. 21A is a diagram illustrating an example of waveform information of current load data in the Mz direction when the user is traveling straight ahead. FIG. 21B is a diagram showing an average load value of the current load data in the Mz direction shown in FIG. 21A. FIG. 22 is a diagram illustrating an example of waveform information of corrected load data in the second embodiment of the present disclosure. FIG. 23 is a control block diagram illustrating a main control configuration in the walking assist robot according to the third embodiment of the present disclosure. FIG. 24 is a control block diagram illustrating a control configuration of walking support of the walking support robot according to the third embodiment of the present disclosure. FIG. 25 is an exemplary flowchart of the body information estimation process of the walking assistance robot according to the third embodiment of the present disclosure. FIG. 26A is an example of physical information stored in the physical information database of the walking assistance robot according to the third embodiment of the present disclosure. FIG. 26B is another example of the physical information stored in the physical information database of the walking assistance robot according to the third embodiment of the present disclosure. FIG. 27 is an exemplary flowchart of user movement intention estimation processing of the walking assistance robot according to the third embodiment of the present disclosure. FIG. 28 is another control block diagram showing a walking assistance control configuration of the walking assistance robot according to the third embodiment of the present disclosure.

(Background to the disclosure)
In recent years, with the aging population and declining birthrate in developed countries, the need for watching over elderly people and supporting their lives has increased. In particular, elderly people tend to have difficulty in maintaining QOL (Quality of Life) for home life due to a decline in physical ability associated with aging. In order to prevent elderly people such as sarcopenia and maintain physical ability, it is important to maintain muscle mass by continuing to exercise beyond a certain level. However, it is difficult to go out due to a decrease in physical ability, and it is difficult for an elderly person who tends to stay in the house to maintain a constant amount of exercise, resulting in a vicious circle in which muscle mass is further attenuated. In recent years, various devices for assisting the user's walking have been proposed under such a background.

  As described in the background art, Patent Document 1 discloses a guidance mobile robot that moves at a moving speed according to a user input. However, with the guidance mobile robot of Patent Document 1, when a user with low physical ability such as an elderly person uses the robot, it is difficult to perform comfortable walking support for the user.

  For example, an elderly person with low physical ability may lean forward with respect to the robot. In this case, in the robot of Patent Document 1, although the elderly person's walking speed is slow, the input value in the traveling direction increases, and thus the moving speed in the traveling direction increases. As a result, the elderly cannot go about the movement of the mobile robot. In addition, an elderly person may swing left and right without stabilizing the center of gravity of the body, even though he is moving straight ahead. In this case, the robot of Patent Document 1 may detect an input in the left-right direction and move in the left-right direction. As described above, in the robot of Patent Document 1, in the case of a user with low physical ability, the robot may move in a direction contrary to the user's intention to move although the user wants to perform a straight-ahead operation. . For this reason, in the robot of patent document 1, it is necessary for the user to walk while finely adjusting the traveling direction, and it is difficult to perform comfortable walking support for the user.

  Even if the user is walking staggered from side to side, the present intent makes the user's intention to walk by accumulating information on the movement of the walking support robot (for example, movement direction, movement speed, etc.). I found out that I can grasp it. Accordingly, the present inventors have reached the following invention in order to move the robot according to the user's movement intention based on the accumulated movement operation information even when the user is walking in a staggered manner from side to side. It was.

A walking support robot according to an aspect of the present disclosure is provided.
A walking support robot that supports user walking,
The main body,
A handle portion provided on the main body portion and grippable by the user;
A detection unit for detecting a load applied to the handle unit;
A moving device that has a rotating body and moves the walking support robot by controlling the rotation of the rotating body according to the load detected by the detection unit;
Based on past load data applied to the handle part acquired during the movement of the walking support robot, a load tendency data generation unit that generates load tendency data indicating a load tendency applied to the handle part;
With
The moving device includes an actuator that controls the amount of rotation of the rotating body based on the load detected by the detection unit and the load tendency data generated by the load tendency data generation unit.

  With such a configuration, walking support can be performed according to the physical ability of the user, so that more comfortable walking support for the user can be performed.

The moving device further includes a load correction unit that corrects a load value detected by the detection unit based on the load tendency data.
The actuator may control a rotation amount of the rotating body based on a load value corrected by the load correction unit.

  With such a configuration, the load value can be corrected based on the load tendency data, so that more comfortable walking support for the user can be performed.

The load tendency data generation unit generates load tendency data for each movement operation of the walking support robot,
The load correction unit may correct the value of the load based on the load tendency data corresponding to the movement operation of the walking support robot when the load is detected.

  With such a configuration, load tendency data can be generated for each movement operation, so that the user's load tendency can be grasped more accurately. Thereby, a more comfortable user's walking support can be performed.

  The load correction unit may correct the value of the load based on the load tendency data when the load tendency data corresponding to the movement operation of the walking support robot is equal to or greater than a predetermined threshold.

  With such a configuration, the load value can be corrected when the load tendency data corresponding to the movement operation of the walking support robot exceeds a predetermined threshold value, so that a more comfortable walking support for the user is performed. Can do.

The load tendency data is a fluctuation frequency calculated from the past load data,
The load correction unit may correct the value of the load by filtering the frequency component of the fluctuation from the load detected by the detection unit.

  With such a configuration, by using the fluctuation frequency as the load tendency data, it is possible to acquire the load tendency data of the user in a wide range from the fluctuation with small unevenness to the fluctuation with large unevenness, and correct the load value. Therefore, more comfortable walking support for the user can be performed.

The load tendency data is an average load value calculated from the past load data,
The load correction unit may correct the value of the load based on the average load value.

  With such a configuration, by using the average load value as the load trend data, it is possible to acquire the load applied constantly for each user as the load trend data and correct the load value. Walking support can be performed.

  The load correction unit may correct the load value by subtracting the average load value from the load detected by the detection unit.

  With such a configuration, by subtracting the average load value from the load detected by the detection unit, it is possible to reduce the constantly applied load for each user, and thus it is possible to perform more comfortable walking support for the user. .

In the walking support robot,
A body information estimation unit for estimating the user's body information;
The load correction unit may correct the load value based on the body information estimated by the body information estimation unit.

  With such a configuration, the intensity of correction can be adjusted based on physical information, and thus walking support corresponding to the physical ability of the user can be performed.

In the walking support robot,
You may provide the user notification part which notifies the said physical information to the said user.

  With such a configuration, the user himself / herself can grasp daily physical information, which leads to motivation to maintain and improve physical ability or to attract attention during walking.

In the walking support robot,
A user movement intention estimation unit that estimates the movement intention of the user based on the corrected load value;
The user notification unit may notify the user of the user's intention to move.

  With such a configuration, the user can grasp the control state of the walking support robot.

A storage unit for storing a control table indicating a correspondence relationship between a load applied to the handle unit and a rotation amount of the rotating body;
The actuator drives and controls the rotating body with a rotation amount corresponding to the load detected by the detection unit by the control table stored in the storage unit,
The control table may be updated by correcting the load value based on the load tendency data.

  With such a configuration, the correspondence relationship between the load applied to the handle and the rotation amount of the rotating body can be easily specified by the control table, so that walking support according to the user's physical ability can be performed more easily. it can.

The detection unit detects a plurality of axial loads applied to the handle unit,
The moving device may switch the movement operation of the walking support robot by controlling the rotation of the rotating body according to the magnitude of the load applied to each of the plurality of axial directions.

  With such a configuration, the user's intention to move can be detected more accurately by detecting loads applied in a plurality of axial directions. For this reason, the movement operation of the walking support robot can be switched according to the user's intention to move.

  The moving operation may include a straight movement operation, a backward movement operation, and a turning operation of the walking support robot.

  With such a configuration, since the movement operation of the walking support robot can be switched according to the user's intention to move, more comfortable walking support for the user can be performed.

  The actuator may change a turning radius in the turning operation based on the load tendency data.

  With such a configuration, the walking support robot performs a turning motion according to the physical ability of the user, so that it is possible to provide more comfortable walking support for the user.

A walking support method according to an aspect of the present disclosure includes:
A walking support method for supporting a user's walking using a walking support robot,
Detecting a load applied to a handle portion of the walking support robot by a detection unit;
Generating load trend data indicating a tendency of a load applied to the handle unit based on past load data applied to the handle unit acquired during the movement of the walking support robot;
Controlling the amount of rotation of the rotating body provided in the moving device of the walking support robot based on the load detected by the detection unit and the load tendency data;
including.

  With such a configuration, walking support can be performed according to the physical ability of the user, so that more comfortable walking support for the user can be performed.

In the walking support method,
Based on the load tendency data, including a step of correcting the value of the load detected by the detection unit,
The step of controlling the amount of rotation of the rotating body may control the amount of rotation of the rotating body based on the corrected load value.

  With such a configuration, the load value can be corrected based on the load tendency data, so that more comfortable walking support for the user can be performed.

In the walking support method,
Estimating the user's physical information,
The step of correcting the value of the load may correct the value of the load based on the physical information.

  With such a configuration, the intensity of correction can be adjusted based on physical information, and thus walking support corresponding to the physical ability of the user can be performed.

In the walking support method,
The method may include notifying the user of the physical information.

  With such a configuration, the user himself / herself can grasp daily physical information, which leads to motivation to maintain and improve physical ability or to attract attention during walking.

In the walking support method,
Estimating a user's intention to move based on the corrected load value,
The notifying step may notify the user of the user's intention to move.

  With such a configuration, the user can grasp the state of walking support.

  Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. In each drawing, each element is exaggerated for easy explanation.

(Embodiment 1)
[overall structure]
FIG. 1 is an external view of a walking support robot 1 (hereinafter referred to as “robot 1”) according to the first embodiment. FIG. 2 shows a state in which the user is walking with the walking support by the robot 1.

  As shown in FIGS. 1 and 2, the robot 1 moves the main body 11, a handle 12 that can be gripped by the user, a detection unit 13 that detects a handle load applied to the handle 12, and the main body 11. A moving device 14 and a load tendency data generation unit 15 are provided.

  The handle part 12 is provided on the upper part of the main body part 11, and is provided in a shape and a height position that can be easily grasped by both hands of a user who is walking.

  The detection unit 13 detects a load (handle load) applied by the user to the handle unit 12 when the user grips the handle unit 12. Specifically, when the user walks while holding the handle portion 12, the user applies a handle load to the handle portion 12. The detection unit 13 detects the direction and magnitude of the handle load applied to the handle unit 12 by the user.

FIG. 3 shows the detection direction of the handle load detected by the detection unit 13. As shown in FIG. 3, the detection unit 13 is a six-axis force sensor capable of detecting forces applied in three axial directions orthogonal to each other and moments around the three axes. The three axes orthogonal to each other are an x-axis extending in the left-right direction of the robot 1, a y-axis extending in the front-rear direction of the robot 1, and a z-axis extending in the height direction of the robot 1. The force applied in the triaxial direction is a force Fx applied in the x-axis direction, a force Fy applied in the y-axis direction, and a force Fz applied in the z-axis direction. In the first embodiment, the force applied to the right direction in the Fx and Fx ^+, a force applied to the left Fx ^- is set to. The force applied to front direction in the Fy and Fy ^+, the force applied to the rear direction Fy ^- is set to. The force applied to the vertically downward direction to the tread surface of the Fz direction and Fz ^+, Fz force exerted vertically upward with respect to the tread surface ^- is set to. The moments around the three axes are the moment Mx around the x-axis, the moment My around the y-axis, and the moment Mz around the z-axis.

  The moving device 14 moves the main body 11 based on the magnitude and direction of the handle load (force and moment) detected by the detector 13. In the first embodiment, the moving device 14 performs the following control. In this specification, Fx, Fy, Fz, Mx, My, and Mz may be referred to as loads.

<Forward movement>
When the detection unit 13 detects Fy ⁺ force, the moving device 14 moves the main body unit 11 in the forward direction. That is, when Fy ⁺ force is detected by the detection unit 13, the robot 1 performs forward movement. While the robot 1 is moving forward, if the Fy ⁺ force detected by the detector 13 increases, the moving device 14 increases the speed of the robot 1 moving in the forward direction. On the other hand, when the Fy ⁺ force detected by the detection unit 13 is reduced while the robot 1 is moving forward, the moving device 14 decreases the moving speed of the robot 1 in the forward direction.

<Backward movement>
When the detection unit 13 detects Fy ⁻ force, the moving device 14 moves the main body unit 11 in the backward direction. That is, when the detection unit 13 detects Fy ⁻ force, the robot 1 performs the backward movement. While the robot 1 is moving backward, when the Fy ⁻ force detected by the detection unit 13 increases, the moving device 14 increases the speed of the backward movement of the robot 1. On the other hand, when the Fy ⁻ force detected by the detection unit 13 is reduced while the robot 1 is moving backward, the moving device 14 decreases the moving speed of the robot 1 in the backward direction.

<Right turn motion>
When the detecting unit 13 detects the force of Fy ⁺ and the moment of Mz ⁺ , the moving device 14 turns the main body unit 11 in the right direction. That is, when the detection unit 13 detects a force of Fy ⁺ and a moment of Mz ⁺ , the robot 1 performs a right turn operation. While the robot 1 is performing the right turn operation, the turning radius of the robot 1 is reduced when the moment of Mz ⁺ detected by the detection unit 13 is increased. Moreover, when the force of Fy ⁺ detected by the detection unit 13 is increased while the robot 1 is performing a right turn operation, the turn speed is increased.

<Left turn motion>
When the detection unit 13 detects the force of Fy ⁺ and the moment of Mz ⁻ , the moving device 14 turns the main body unit 11 in the left direction. That is, when Fy ⁺ force and Mz ⁻ moment are detected by the detection unit 13, the robot 1 performs a left turn operation. While the robot 1 is performing the left turning motion, when the Mz ⁻ moment detected by the detection unit 13 is increased, the turning radius of the robot 1 is decreased. Further, when the force of Fy ⁺ detected by the detection unit 13 is increased while the robot 1 is performing the left turning operation, the turning speed is increased.

  Note that the control of the moving device 14 is not limited to the above-described example. For example, the moving device 14 may control the forward movement and the backward movement of the robot 1 based on the forces of Fy and Fz. Further, the moving device 14 may control the turning operation of the robot 1 based on, for example, the moment of Mx or My.

  In the first embodiment, the example in which the detection unit 13 is a six-axis force sensor has been described. However, the present invention is not limited to this. The detection unit 13 may use, for example, a triaxial sensor or a strain sensor.

  The moving device 14 includes a wheel 16 that is a rotating body provided at a lower portion of the main body 11 and a drive unit 17 that drives and controls the wheel 16.

  The wheel 16 supports the main body part 11 in a self-supporting state and is driven to rotate by the driving part 17 so that the main body part 11 is maintained in a self-supporting posture, for example, the direction of the arrow shown in FIG. Move forward (forward or backward). In the first embodiment, the case where the moving device 14 includes a moving mechanism using two wheels 16 has been described as an example. However, this is the case where a rotating body (travel belt, roller, etc.) other than wheels is used. May be.

  The drive unit 17 includes a load correction unit 18, a user movement intention estimation unit 19, a driving force calculation unit 20, an actuator control unit 21, and an actuator 22.

  The load correction unit 18 corrects the handle load detected by the detection unit 13 based on the load tendency of the user. Specifically, the load correction unit 18 corrects the value of the handle load detected by the detection unit 13 based on the load tendency data generated by the load trend data generation unit 15. In the first embodiment, the fluctuation of the steering wheel load is corrected by calculating the fluctuation frequency from the past steering wheel load data when the user walks, and filtering the fluctuation frequency from the steering wheel load detected by the detection unit 13. Further, the load correction unit 18 may correct the value of the handle load based on the use place, use time, and physical condition of the user of the robot 1.

The user movement intention estimation unit 19 estimates the user's movement intention based on the handle load corrected by the load correction unit 18 (hereinafter referred to as “corrected handle load”). The user's intention to move includes a moving direction and a moving speed. In the first embodiment, the user movement intention estimation unit 19 estimates the user's movement intention from the value of the corrected handle load in each movement direction. For example, when the force of Fy ⁺ detected by the detection unit 13 is a value equal to or greater than a predetermined first threshold and the force of My ⁺ is a value less than the predetermined second threshold, the user movement intention estimation unit 19 It may be estimated that the user's intention to move is a straight-ahead operation. Further, the user movement intention estimation unit 19 may estimate the movement speed based on the value of the corrected handle load in the Fz direction. On the other hand, when the force of Fy ⁺ detected by the detection unit 13 is a value equal to or greater than a predetermined third threshold and the force of My ⁺ is equal to or greater than a predetermined second threshold, the user movement intention estimation unit 19 It may be estimated that the user's intention to move is a right turn operation. In addition, the user movement intention estimation unit 19 may estimate the turning speed based on the value of the corrected handle load in the Fz direction, and may estimate the turning radius based on the value of the corrected handle load in the My direction.

  The driving force calculation unit 20 calculates the driving force based on the handle load information corrected by the load correction unit 18. Specifically, the driving force calculation unit 20 calculates the driving force based on the user's movement intention estimated from the corrected handle load information, that is, the user's moving direction and moving speed. For example, when the user's intention to move is forward movement or backward movement, the driving force is calculated so that the rotation amounts of the two wheels 16 are equal. When the user's intention to move is a right turn operation, the driving force is calculated so that the rotation amount of the right wheel 16 of the two wheels 16 is larger than the rotation amount of the left wheel 16. Further, the magnitude of the driving force is calculated according to the moving speed of the user.

  The actuator control unit 21 performs drive control of the actuator 22 based on the driving force information calculated by the driving force calculation unit 20. Further, the actuator control unit 21 can acquire information on the rotation amount of the wheel 16 from the actuator 22, and can transmit information on the rotation amount of the wheel 16 to the driving force calculation unit 20 and the user load tendency extraction unit 23.

  The actuator 22 is, for example, a motor that drives the wheels 16 to rotate. The actuator 22 is connected to the wheel 16 via a gear mechanism or a pulley mechanism. The actuator 22 is driven and controlled by the actuator control unit 21 to rotationally drive the wheel 16.

The load tendency data generation unit 15 generates user load tendency data based on information on handle loads detected in the past. The load tendency data is data indicating the tendency of the user's handle load in a predetermined operation. The predetermined movement means, for example, a straight movement, a backward movement, and a turning movement. For example, a user with a bent waist leans on the robot 1 when holding the handle portion 12, and therefore the handle load in the vertically downward direction relative to the walking surface of the road on which the robot 1 moves, that is, Fz ⁺ There is a tendency to increase power. In addition, when the user who walks while swinging left and right grips the handle portion 12, the handle load in the left-right direction, that is, the moment of My, tends to increase despite the forward movement. As described above, the load tendency data generation unit 15 generates the load tendency of the user for each predetermined operation from the past load data.

[Control configuration of walking support robot]
In the walking support robot 1 having such a configuration, a control configuration for supporting the user's walking will be described. FIG. 4 is a control block diagram showing a main control configuration in the robot 1. The control block diagram of FIG. 4 also shows the relationship between each control configuration and information to be handled.

  As shown in FIG. 4, the detection unit 13 detects a handle load applied to the handle unit 12. Information on the handle load detected by the detection unit 13 is transmitted to the load correction unit 18. The load correction unit 18 corrects the value of the handle load detected by the detection unit 13 based on the load trend data generated by the load trend data generation unit 15. Information on the corrected handle load (corrected handle load) is transmitted to the user movement intention estimation unit 19. The user movement intention estimation unit 19 estimates the user's movement intention (movement direction and movement speed) based on the information on the corrected handle load. Information on the estimated movement intention of the user is transmitted to the driving force calculation unit 20. The driving force calculation unit 20 calculates the driving force based on the estimated information on the movement intention of the user. Information on the calculated driving force is transmitted to the actuator control unit 21. The actuator control unit 21 performs drive control of the actuator 22 based on the driving force information calculated by the driving force calculation unit 20. The actuator 22 is driven and controlled by the actuator control unit 21 to rotationally drive the wheel 16 and move the main body 11.

  Further, as shown in FIG. 4, the handle load information detected by the detection unit 13 is also transmitted to the load tendency data generation unit 15. The handle load information detected by the detection unit 13 is also used to generate and update load tendency data.

  Detailed control of walking support of the robot 1 will be described with reference to FIG. FIG. 5 is a control block diagram illustrating a detailed control configuration of the walking support of the robot 1.

  As shown in FIG. 5, the load trend data generation unit 15 includes a user load trend extraction unit 23 that extracts a user load trend corresponding to the user's moving direction, and a load trend map 24 that stores the user load trend data. .

  The user load tendency extracting unit 23 extracts a user load tendency corresponding to the moving direction of the user. Specifically, the user load tendency extracting unit 23 extracts the user load tendency data corresponding to the moving direction of the user from the load tendency map 24. For example, when the user is moving straight ahead (forward forward), the user load tendency extraction unit 23 extracts the load tendency of the user corresponding to the straight movement from the load tendency map 24. The user load trend extraction unit 23 transmits the load trend data extracted from the load trend map 24 to the load correction unit 18.

  Further, the user load tendency extraction unit 23 generates user load tendency data based on the handle load information detected by the detection unit 13 and the rotation amount information of the wheels 16 acquired by the actuator control unit 21. . The generated load trend data is transmitted to the load trend map 24. Thereby, the load tendency data of the load tendency map 24 is updated.

  The load tendency map 24 is a database that stores user load tendency data in each of the movement directions of the user. The load tendency map 24 stores user load tendency data for each movement direction. FIG. 6 shows a load trend map 24. As shown in FIG. 6, in the first embodiment, the load tendency map 24 stores the fluctuation frequency in the moving direction during walking and the fluctuation frequency in the bias direction of the center of gravity during walking as the user's load tendency data. . Further, the load tendency map 24 may store data of fluctuation frequencies calculated in the past.

  Although not shown in FIG. 6, the load tendency map 24 may store data such as the usage location, usage time, and physical condition of the user of the robot 1. These data may be used when the load correction unit 18 corrects the handle load.

[Generation of load trend data]
The generation of load tendency data will be described with reference to FIG. FIG. 7 shows an exemplary flowchart of a load trend data generation process.

  As shown in FIG. 7, in step ST <b> 1, it is determined whether the handle load is detected by the detection unit 13. In step ST1, it is determined whether or not the user is holding the handle portion 12. When the handle load is detected by the detection unit 13, the process proceeds to step ST2. When the detection unit 13 does not detect the handle load, step ST1 is repeated.

  In step ST <b> 2, the user load tendency extraction unit 23 estimates the moving direction of the user based on the information on the rotation amount of the wheel 16. Specifically, when a change in the handle load is detected in step ST1, the actuator control unit 21 acquires information on the rotation amount of the wheel 16. Information on the rotation amount acquired by the actuator control unit 21 is transmitted to the user load tendency extraction unit 23. The user load tendency extracting unit 23 estimates the moving direction of the user based on the information on the rotation amount of the wheel 16, that is, the rotation direction and the rotation number of the wheel. In the first embodiment, the user load tendency extracting unit 23 estimates the moving direction of the user based on the rotation amounts of the two wheels 16 arranged on the left and right. For example, when the rotation amount of the right wheel 16 is larger than the rotation amount of the left wheel 16, the user load tendency extraction unit 23 may estimate that the user is turning leftward. Further, the user load tendency extraction unit 23 may estimate that the robot 1 is performing a straight traveling operation when the rotation speeds of the left and right wheels 16 are the same and are rotating in the forward direction.

In step ST3, the user load tendency extraction unit 23 acquires the handle load waveform information in the estimated user movement direction. The handle load waveform information in the user movement direction is not particularly limited. For example, when the user movement direction is the Fy ⁺ direction, the handle load waveform information in the Fz direction or the moment waveform information in the My direction is used. There may be.

  In step ST4, the user load tendency extraction unit 23 adds the acquired handle load waveform information and the past handle load waveform information. For example, past waveform information is stored in the load tendency map 24. The user load trend extraction unit 23 reads past waveform information from the load trend map 24 and adds the acquired current waveform information to the past waveform information. FIG. 8 shows an example of input waveform information of the handle load. As shown in FIG. 8, the waveform information of the handle load detected so far is stored in the load tendency map 24.

  In step ST5, the user load tendency extraction unit 23 calculates a fluctuation frequency based on the combined waveform information. Specifically, the user load tendency extracting unit 23 calculates a fluctuation frequency by performing a frequency analysis of the handle load in the estimated moving direction of the user.

  As an example, calculation of the fluctuation frequency during a straight-ahead operation of a user with low physical ability will be described. FIG. 9A shows an example of waveform information of load data in the Fz direction when the user is moving straight ahead. FIG. 9B shows frequency components of the load data in the Fz direction shown in FIG. 9A. FIG. 10A shows an example of waveform information of load data in the My direction when the user is moving straight ahead. FIG. 10B shows frequency components of load data in the My direction shown in FIG. 10A. Although FIG. 9A shows the waveform of the load data for three steps, the frequency analysis is actually performed on the waveform of the load data for ten or more steps.

  Since the user with low physical ability walks while swinging from side to side, the steering wheel load is not stable even if the user moves straight at a constant speed. Therefore, as shown in FIG. 9A, fluctuations occur in the waveform information of the load data in the height direction of the robot 1, that is, in the Fz direction. Fluctuation means a component that waveform information fluctuates and is not stable, and specifically means fluctuation from an average value of load data.

  In this case, although the user intends to go straight forward, the robot 1 moves in the left-right direction, so the user walks while finely adjusting the direction of travel to the left and right. In the first embodiment, the user load tendency extraction unit 23 estimates that the user is walking while swinging left and right, and uses the fluctuation component of the load as load tendency data in order to correct the handle load. Hereinafter, an example of processing of the user load tendency extraction unit 23 will be described.

  The user load tendency extraction unit 23 performs frequency analysis on the waveform information of the load data in the Fz direction shown in FIG. 9A, and calculates the frequency component of the load data as shown in FIG. 9B. Thereby, the user load tendency extraction unit 23 can specify that there is a fluctuation frequency of 2 Hz in the Fz direction as shown in FIG.

  Further, as shown in FIG. 10A, fluctuations also occur in the waveform information of the load data in the My direction of the user with low physical ability. The user load tendency extraction unit 23 performs frequency analysis on the load data in the My direction shown in FIG. 10A, and calculates the frequency component of the load data as shown in FIG. 10B. Thereby, the user load tendency extraction unit 23 can specify that there is a fluctuation frequency of 2 Hz in the My direction as shown in FIG.

  As another example, calculation of the fluctuation frequency when the user with low physical ability turns rightward will be described. FIG. 11A shows an example of waveform information of load data in the Fz direction when the user turns right. FIG. 11B shows frequency components of the load data in the Fz direction shown in FIG. 11A.

  As shown in FIG. 11A, even when a user with low physical ability is turning in the right direction, the waveform information of the load data in the Fz direction fluctuates. The user load tendency extraction unit 23 performs frequency analysis on the load data in the Fz direction shown in FIG. 11A and calculates the frequency component of the load data as shown in FIG. 11B. The user load tendency extracting unit 23 can specify that there is a fluctuation frequency of 6 Hz in the Fz direction as shown in FIG. 11B when the user turns rightward.

  As described above, in step ST5, the user load tendency extracting unit 23 calculates the fluctuation frequency from the waveform information of the handle load added in the estimated moving direction of the user.

  Returning to FIG. 7, in step ST6, the user load tendency extraction unit 23 sets the fluctuation frequency calculated in step ST5 as load tendency data. Specifically, the user load tendency extracting unit 23 updates the user's load tendency data in the load tendency map 24 to the fluctuation frequency calculated in step ST5.

  Thus, in the first embodiment, by performing the processing of steps ST1 to ST6, the fluctuation frequency of the user's handle load can be calculated, and the fluctuation frequency can be used as the load tendency data. In the first embodiment, load tendency data can be generated for each moving operation.

  In step ST4 and ST5, the example in which the fluctuation frequency calculated based on the waveform information obtained by adding the acquired handle load waveform information and the past handle load waveform information is set as the load tendency data has been described. However, the present invention is not limited to this. After adding the fluctuation frequency calculated based on the acquired handle load waveform information to the fluctuation frequency calculated based on the past waveform information, average calculation is performed, and the average value of the calculated fluctuation frequency is applied to the load trend. It may be used as data. In addition, the median or mode of the fluctuation frequency is calculated based on the fluctuation frequency calculated from the acquired handle load waveform information and the past fluctuation frequency, and the median or mode of the fluctuation frequency is loaded. It may be used as data. Further, the average value, the median value, and the mode value of the fluctuation frequency may be combined and used as the load tendency data. Alternatively, the most recently calculated fluctuation frequency may be used as the load tendency data. The above-described load tendency data may be properly used depending on the situation and purpose. For example, for a user with a small amount of past fluctuation frequency data stored in the load tendency map 24, the latest fluctuation frequency calculated may be used as the load tendency data. On the other hand, for users with a large amount of past fluctuation frequency data stored in the load tendency map 24, the average value, median value, or mode value of the fluctuation frequency may be used as the load tendency data.

[Estimation of user's intention to move]
The estimation of the user's intention to move will be described with reference to FIG. FIG. 12 shows an exemplary flowchart of a process for estimating a user's movement intention.

  As shown in FIG. 12, in step ST <b> 11, the load correction unit 18 acquires information on the handle load detected by the detection unit 13.

  In step ST <b> 12, the user load trend extraction unit 23 acquires load trend data from the load trend map 24. Specifically, the user load tendency extraction unit 23 acquires the fluctuation frequency that is the target of the current movement direction of the user from the load tendency map 24. The user load tendency extraction unit 23 transmits the fluctuation frequency information to the load correction unit 18 as load tendency data. Note that the current movement direction of the user can be estimated by obtaining information on the rotation amount of the wheel 16 from the actuator control unit 21.

  In step ST13, the load correction unit 18 filters the fluctuation frequency component acquired in step ST12 from the handle load acquired in step ST11. As a result, the load correction unit 18 corrects the value of the handle load detected by the detection unit 13. Information on the corrected handle load obtained by the load correction unit 18 is transmitted to the user movement intention estimation unit 19.

  Moreover, the load correction | amendment part 18 may correct | amend the value of a handle | steering-wheel load based on the use place of the robot 1, use time, and a user's physical condition. In this case, the user load tendency extracting unit 23 extracts data on the usage place, use time, and user's physical condition of the robot 1 from the load tendency map 24, and transmits these data to the load correcting unit 18. For example, the load correction unit 18 reduces the handle load when the robot 1 is used in a living room or when the user is in poor health, compared to the handle load when the robot 1 is used in a hallway or when the user is in good health. The handle load may be corrected so that

  In step ST14, the user movement intention estimation unit 19 estimates the user's movement intention based on the corrected handle load acquired in step ST13. Specifically, the user movement intention estimation unit 19 estimates the movement direction and movement speed of the user based on the magnitudes of forces in the Fx, Fy, Fz, Mx, My, and Mz directions of the corrected handle load.

  As described above, in the first embodiment, by performing the processing of steps ST11 to ST14, the fluctuation frequency component is filtered from the waveform information of the user's handle load, and the user's movement is performed based on the obtained corrected handle load information. Estimating the intention.

  In the first embodiment, as filtering, correction may be performed so as to cut all frequency components corresponding to the fluctuation portion so as to remove all fluctuation components, and fluctuation data of the fluctuation components may be applied to load data during walking. You may perform correction | amendment which reduces a ratio.

  Further, the load correction unit 18 does not correct only with the user's load tendency data, but compares the user's load tendency data with the average load tendency data of a plurality of users, and corrects to reduce the difference portion. The ratio may be changed. As an average calculation method for multiple users, it may be created for each group classified by a combination of age, gender, location, walking ability (walking speed, walking rate, stride, standing posture, left and right shaking), etc. .

  As an example, a process for estimating a movement intention of a user with low physical ability will be described. FIG. 13A shows an example of waveform information of load data in the Fz direction when the user is moving straight ahead. FIG. 13B shows waveform information obtained by filtering fluctuation frequency components from the waveform information of the load data in the Fz direction shown in FIG. 13A. FIG. 14A shows an example of waveform information of load data in the My direction when the user is moving straight ahead. FIG. 14B shows waveform information obtained by filtering the fluctuation frequency component from the waveform information of the load data in the My direction shown in FIG. 14A. The waveform information shown in FIGS. 13A and 14A is the handle load waveform information acquired in step ST11. The waveform information shown in FIGS. 13B and 14B is the waveform information of the corrected handle load obtained by filtering the fluctuation frequency component in step ST13.

  As shown in FIG. 13A, since the user with low physical ability is not stable in walking, the waveform information of the load data in the Fz direction at the time of the straight traveling motion fluctuates. That is, the value of the handle load in the Fz direction detected by the detection unit 13 fluctuates when the user goes straight ahead. The load correction unit 18 filters the fluctuation frequency component from the waveform information of the handle load in the Fz direction acquired by the detection unit 13. As a result, as shown in FIG. 13B, fluctuations in the handle load waveform information in the Fz direction during the straight traveling operation can be cut off. Accordingly, the user movement intention estimation unit 19 can easily estimate that the user's movement intention is a straight movement operation based on the corrected handle load.

  Further, as shown in FIG. 14A, fluctuations also occur in the waveform information of the load data in the My direction when the user with low physical ability goes straight ahead. That is, the value of the handle load in the My direction detected by the detection unit 13 fluctuates when the user goes straight ahead. The load correction unit 18 filters the fluctuation frequency component from the waveform information of the handle load in the My direction acquired by the detection unit 13. Thereby, as shown to FIG. 14B, the fluctuation | variation of the waveform information of the handle | steering-wheel load in the My direction at the time of a rectilinear advance operation | movement can be cut. Thereby, the user movement intention estimation unit 19 can easily estimate that the user's movement intention is a straight-ahead operation based on the corrected handle load.

  Further, the user movement intention estimation unit 19 may estimate a turning radius at the time of turning. For example, for a user with weak legs, the robot 1 may turn gently by setting the turning radius larger than usual. On the other hand, for a user with strong legs and legs, the turning radius may be smaller than usual and the user may turn sharply. For example, the turning radius is estimated from the value of the corrected handle load.

  Further, the user movement intention estimation unit 19 obtains information on the rotation amount of the wheel 16 from the actuator control unit 21, and estimates the user's movement intention based on the rotation amount information and the corrected handle load information. Good.

[Calculation of driving force]
The calculation of the driving force will be described with reference to FIG. FIG. 15 shows an exemplary flowchart of a driving force calculation process.

  As shown in FIG. 15, in step ST <b> 21, the driving force calculation unit 20 acquires information on the user's movement intention from the user movement intention estimation unit 19.

  In step ST <b> 22, the driving force calculation unit 20 acquires information on the rotation amount of the wheel 16 from the actuator control unit 21.

  In step ST23, the driving force calculation unit 20 calculates the driving force based on the user's intention to move and information on the rotation amount of the wheels 16 acquired in step ST21. Specifically, the driving force calculation unit 20 calculates the difference between the current movement direction and movement speed calculated from the information on the rotation amount of the wheel 16 and the movement direction and movement speed estimated from the information on the user's movement intention. Based on this, the amount of rotation of the wheel 16 is calculated.

As an example, when the robot 1 is moving in the forward direction at a moving speed of 71 cm / s, the driving force calculation unit 20 when the user increases the force of Fy ⁺ to accelerate the moving speed to 77 cm / s. The operation of will be described. The driving force calculation unit 20 acquires information indicating that the amount of rotation of both the left and right wheels 16 is 2000 rpm in a state of moving in the forward direction at a speed of 71 cm / s. Next, the driving force calculation unit 20 calculates that the rotation amount of both the left and right wheels 16 needs 2500 rpm in order to set the moving speed of the robot 1 to 77 cm / s. The driving force calculation unit 20 calculates the driving force so that the rotation amount of the left and right wheels 16 is increased by 500 rpm.

  In the first embodiment, the driving force calculation unit 20 has been described with respect to an example in which the driving force is calculated based on information on the user's intention to move and information on the amount of rotation of the wheel 16 acquired from the actuator control unit 21. However, it is not limited to this. For example, the driving force calculation unit 20 may calculate the driving force only from information about the user's movement intention. That is, step ST22 may not be included in the driving force calculation process.

  Further, the driving force calculation unit 20 may calculate the driving force based on a control table indicating a correspondence relationship between the handle load and the rotation amount of the wheel 16. Specifically, the driving force calculation unit 20 may include a storage unit that stores a control table indicating a correspondence relationship between the handle load and the rotation amount of the wheel 16. The driving force calculation unit 20 may calculate the amount of rotation of the wheel 16 corresponding to the value of the handle load detected by the detection unit 13 using the control table stored in the storage unit. Further, the control table may be updated by correcting the value of the handle load in the control table based on the load tendency data extracted from the user load tendency extracting unit 23.

[effect]
According to the walking assist robot 1 according to the first embodiment, the following effects can be obtained.

  According to the walking assist robot 1 according to the first embodiment, the value of the handle load can be corrected based on the load tendency data of the user. With such a configuration, the robot 1 can correct the handle load value according to the user's tendency.

  For example, for a user who tends to walk left and right, the value of the handle load is corrected by cutting the frequency of fluctuation caused by the left and right swing from the handle load. Thus, since the value of the handle load can be corrected according to the physical ability of the user, the moving direction and moving speed of the robot 1 can be set for each user having different physical ability. Thereby, since the robot 1 can be moved according to a user's physical ability, a more comfortable user's walk assistance can be performed.

  In the first embodiment, the fluctuation frequency of the handle load is used as the load tendency data. By using the fluctuation frequency, the robot 1 can acquire the user's load tendency data in a wide range from the fluctuation with small unevenness shown in the waveform information of the handle load to the fluctuation with large unevenness, and can correct the handle load. Thereby, the robot 1 can perform walking support according to the user's physical ability.

  In the first embodiment, the load tendency data generation unit 15, the load correction unit 18, the user movement intention estimation unit 19, the driving force calculation unit 20, and the actuator control unit 21 have, for example, a program that causes these elements to function. A storage memory (not shown) and a processing circuit (not shown) corresponding to a processor such as a CPU (Central Processing Unit) may be provided, and the processor may function as these elements by executing a program. Alternatively, the load tendency data generation unit 15, the load correction unit 18, the user movement intention estimation unit 19, the driving force calculation unit 20, and the actuator control unit 21 may be configured using an integrated circuit that allows these elements to function. .

  In the first embodiment, the operation of the walking support robot 1 has been mainly described. However, these operations can also be executed as a walking support method.

  In Embodiment 1, although the example which sets a fluctuation frequency in a moving direction and a moment direction was demonstrated, respectively, it is not limited to this. For example, a common fluctuation frequency may be set in all directions. Thereby, the handle load can be easily corrected.

  In the first embodiment, the example in which the forward movement operation, the backward movement operation, the right turn operation, the left turn operation, and the like of the robot 1 are controlled by setting the rotation amounts of the two wheels 16 has been described. However, the present invention is not limited thereto. Not. For example, the operation of the robot 1 may be controlled by controlling the amount of rotation of the wheel 16 by a brake mechanism or the like.

In the first embodiment, the load correction unit 18 determines the value of the handle load detected by the detection unit 13 based on the load tendency data when the load tendency data corresponding to the movement operation of the robot 1 is equal to or greater than a predetermined threshold. May be corrected (filtered). For example, if the load tendency data (fluctuation frequency) in the Mz direction is equal to or higher than the threshold value of 0 Hz in the straight movement operation (forward movement in the Fy ⁺ direction) of the robot 1, load data corresponding to the straight movement operation of the robot 1 is set to the load tendency. You may correct | amend based on data. With such a configuration, it is possible to filter the fluctuation frequency in the Mz direction which is unnecessary in the straight movement operation of the robot 1. Note that the predetermined threshold value may be changed according to the physical ability of the user. For example, the predetermined threshold value may be changed to 1 Hz based on information that the fluctuation frequency occurs at 1 Hz in a healthy person. Further, the load tendency data corresponding to the movement operation of the robot 1 may be load tendency data in the same direction as the movement direction of the robot 1 or may be load tendency data in a direction different from the movement direction of the robot 1. Also good. For example, when another user's load tendency data is set to a predetermined threshold, the user's load tendency data and the other user's load tendency data are obtained in the same movement direction as the movement direction corresponding to the movement operation of the robot 1. You may compare.

(Embodiment 2)
A walking support robot according to the second embodiment of the present disclosure will be described. In the second embodiment, differences from the first embodiment will be mainly described. In the second embodiment, the same or equivalent components as those in the first embodiment will be described with the same reference numerals. In the second embodiment, descriptions overlapping with those in the first embodiment are omitted.

  The second embodiment is different from the first embodiment in that an average load value is used as load tendency data. The walking support robot according to the second embodiment has the same components as those of the walking support robot 1 according to the first embodiment, and is denoted by reference numeral “51” in FIG. 1, FIG. 2, and FIG. It is.

  FIG. 16 shows a load tendency map 24 in the second embodiment. As shown in FIG. 16, the load trend map 24 stores, as load trend data, the average load value in the moving direction during walking and the average load value in the bias direction of the center of gravity during walking for each moving direction of the user. ing.

[Generation of load trend data]
The generation of load tendency data will be described with reference to FIG. FIG. 17 shows an exemplary flowchart of a load trend data generation process of the walking support robot 51 (hereinafter referred to as “robot 51”).

  As shown in FIG. 17, in step ST31, it is determined whether or not the handle load is detected by the detection unit 13. In step ST31, it is determined whether or not the user is holding the handle portion 12. When the handle load is detected by the detection unit 13, the process proceeds to step ST32. When the detection unit 13 does not detect the handle load, step ST31 is repeated.

  In step ST <b> 32, the user load tendency extraction unit 23 estimates the current movement direction of the user based on information on the rotation amount of the wheel 16. Specifically, when a handle load is detected in step ST31, the actuator control unit 21 acquires information on the rotation amount of the wheel 16. Information on the rotation amount acquired by the actuator control unit 21 is transmitted to the user load tendency extraction unit 23. For example, the user load tendency extraction unit 23 estimates the moving direction of the user based on the rotation amounts of the two wheels 16 arranged on the left and right.

  In step ST33, the user load tendency extracting unit 23 adds the handle load detected in step ST31 to the past load data in the estimated movement direction of the user. Specifically, the user load tendency extraction unit 23 reads past load data stored in the load tendency map 24, and adds the handle load detected in step ST31 to the read past load data. The past load data means all load data detected so far.

  In step ST34, the user load tendency extraction unit 23 calculates an average load value in the moving direction and an average load value in the bias direction when the user walks.

  In step ST <b> 35, the user load tendency extracting unit 23 sets the calculated average load value in the moving direction and average load value in the bias direction as the load tendency data. Specifically, the user load tendency extraction unit 23 transmits information on the calculated average load value to the load tendency map 24, and the average load value in the movement direction and the average of the bias direction in the load tendency map 24 when the user walks. Update the load value.

[Estimation of user's intention to move]
The estimation of the user's movement intention will be described with reference to FIG. FIG. 18 shows an exemplary flowchart of a process of estimating a user's movement intention.

  As shown in FIG. 18, in step ST <b> 41, the load correction unit 18 acquires information on the current handle load detected by the detection unit 13.

  In step ST42, the user load tendency extracting unit 23 reads the user load tendency data. Specifically, the user load tendency extraction unit 23 reads the past average load value from the load trend map 24 and transmits the past average load value to the load correction unit 18.

  In step ST43, the load correction unit 18 subtracts the past average load value from the current load data. Thereby, the load correction unit 18 corrects the value of the handle load.

  In step ST44, the user movement intention estimation unit 19 estimates the user's movement intention based on the corrected handle load information.

  As an example, correction of the handle load in the second embodiment will be described. Here, correction of the handle load for a user who walks with the center of gravity biased to the right will be described.

[Correction of handle load using average load value]
FIG. 19A shows an example of waveform information of current load data in the Mz direction when the user is moving straight ahead. As shown in FIG. 19A, since the center of gravity of the user is biased to the right, the load (moment) in the Mz direction is detected by the detection unit 13 even during a straight traveling operation.

  FIG. 19B shows an average load value of past load data in the Mz direction. The user load tendency extracting unit 23 calculates an average load value of the past load data as shown in FIG. 19B by performing an average calculation on the waveform information of the past load data. In the case of FIG. 19B, the past average load value is 1.0 Nm in the Mz direction. In the second embodiment, the average load value shown in FIG. 19B is used as the load tendency data.

  Next, the load correction unit 18 corrects the current load data based on the load tendency data. Specifically, the load correction unit 18 subtracts the past average load value 1.0 Nm shown in FIG. 19B from the waveform information of the current load data shown in FIG. 19A in the Mz direction. FIG. 19C shows the waveform information of the current load data in the Mz direction corrected using the load tendency data. As shown in FIG. 19C, the load applied in the Mz direction is reduced as a whole by subtracting the past average load value from the current load data. Thereby, the load correction | amendment part 18 can correct | amend the bias | biasing of the load to a steady right direction.

  The user movement intention estimation unit 19 estimates the user's movement intention based on the corrected current handle load information. Thereby, since the robot 51 can accurately determine the movement intention of the user and move, the user does not need to finely adjust the traveling direction of the robot 51.

  In addition, although the example of the said correction demonstrated as an example the user who walks with the gravity center biased rightward, it is not limited to this. For example, a user whose hip is bent may have a downward load. In this case, the handle load value may be corrected using the average load value in the Fz direction.

  Further, when the forward movement of the robot 51 is performed based on the handle load values in the Fy direction and the Fz direction, average load values in the Fy direction and the Fz direction may be used as the load tendency data. That is, during the forward movement of the robot 51, the handle load value may be corrected using the average load values in the Fy direction and the Fz direction. When the turning operation of the robot 51 is performed based on a load (moment) in the Mz direction, an average load value in the Mz direction may be used as the load tendency data. That is, during the turning operation of the robot 51, the handle load value may be corrected using the average load value in the Mz direction. Alternatively, the average load value in all directions of Fx, Fy, Fz, Mx, My, and Mz may be calculated, and the handle load value may be corrected using the average load value in all directions. In this way, by correcting the handle load using the average load values in a plurality of directions, the user's load tendency can be grasped more accurately, so that the operation of the robot 51 suitable for the user's physical ability can be further improved. It becomes possible. In the correction of the handle load, at least one average load value in the Fx, Fy, Fz, Mx, My, and Mz directions is calculated according to the movement control of the robot 51, and the handle is calculated using the calculated average load value. What is necessary is just to correct | amend a load.

[effect]
According to the walking assist robot 51 according to the second embodiment, the following effects can be obtained.

  According to the walking assist robot 51 according to the second embodiment, the average load value of the handle load is used as the load tendency data of the user. With such a configuration, it is possible to acquire a load that is constantly applied for each user as load tendency data and correct the value of the load. Therefore, it is possible to perform walking support more suitable for the physical ability of the user. . Further, by using the average load value of the handle load as the user's load tendency data, it is possible to reduce errors in extracting the user's load tendency.

  In the second embodiment, an example is described in which all handle loads detected in the past are used as past load data when calculating load trend data. However, the present invention is not limited to this. The past load data used when calculating the load tendency data may be, for example, load data within a predetermined period. For example, the past load data used when calculating the load tendency data may be past load data detected within a predetermined period (for example, one year). Thus, it becomes easy to extract the user's current load tendency by using only relatively new load data.

  In the second embodiment, the load tendency map 24 may store load tendency data during stable walking. The user load trend extraction unit 23 may acquire load trend data from the load trend map 24 and transmit the load trend data during stable walking to the load correction unit 18. The load correction unit 18 may compare the load tendency data during stable walking with the current user's load data, and correct the handle load value when the two data are different. For example, in the past load tendency, when the load in the Fz direction at the time of stable walking of the straight movement of the user is 10N, when the user walks in a forward leaning posture and the handle load in the Fz direction becomes 20N, the load correction unit 18 May correct the handle load in the Fz direction to the value of the handle load in the Fz direction during stable walking. That is, the load correction unit 18 may correct the load of 20N in the Fz direction to ½.

  In Embodiment 2, although the load correction | amendment part 18 demonstrated the example which correct | amends the present load data by subtracting the past average load value from the present load data, it is not limited to this. For example, the load correction unit 18 may consider another parameter so that the handle load can be corrected in accordance with the use place of the robot 51, the use time, the physical condition of the user, and the like.

  Further, when the robot 1 performs movement control with the sum of the Fz value and the Fy value, the handle load may be corrected by changing the ratio when the Fz and Fy values are summed. For example, the control performed at Fz: Fy = 8: 2 may be changed to Fz: Fy = 6: 4. Also, instead of making corrections using only the user's load trend data, the user's load trend data may be compared with the average load trend data of multiple users, and the correction ratio may be changed to reduce the difference. Good. As an average calculation method for multiple users, it may be created for each group classified by a combination of age, gender, location, walking ability (walking speed, walking rate, stride, standing posture, left and right shaking), etc. .

  The load correction unit 18 may correct the handle load by multiplying the current load data by a correction coefficient calculated from past load tendency data. Hereinafter, an example of correction of the handle load using the correction coefficient will be described.

[Correction of handle load using correction coefficient]
FIG. 20A shows an example of waveform information of past load data in the Mz direction when the user is moving straight ahead. FIG. 20B shows an average load value of past load data in the Mz direction shown in FIG. 20A. The user load tendency extraction unit 23 performs an average calculation on the waveform information of the past load data shown in FIG. 20A. Thereby, the average load value of the past load data as shown in FIG. 20B is calculated as the load tendency data. In the case of FIG. 20B, the past average load value is −1.0 Nm in the Mz direction.

  Next, an average load value is calculated from the current load data. FIG. 21A shows an example of the waveform information of the current load data in the Mz direction when the user is moving straight ahead. FIG. 21B shows the average load value of the current load data in the Mz direction shown in FIG. 21A.

  The load correction unit 18 performs an average calculation on the waveform information of the current load data shown in FIG. 21A. Thereby, the average load value of the current load data as shown in FIG. 21B is calculated. In the case of FIG. 21B, the current average load value is −2.0 Nm in the Mz direction.

  The load correction unit 18 calculates a correction coefficient by dividing the past average load value by the current average load value. In this case, the correction coefficient is (−1.0 Nm / −2.0 Nm) = 0.5. The load correction unit 18 corrects the handle load by multiplying the waveform information of the current load data by this correction coefficient. That is, the value of the handle load in the Mz direction detected by the detection unit 13 is corrected by multiplying the waveform information of the current load data shown in FIG. 21A by the correction coefficient 0.5.

  FIG. 22 shows an example of the waveform information of the corrected load data. As shown in FIG. 22, the handle load (see the waveform information in FIG. 21A) detected by the detector 13 is corrected by multiplication of a correction coefficient. As described above, the load correction unit 18 may correct the current handle load by multiplying the current load data by the correction coefficient calculated based on the past load tendency data.

(Embodiment 3)
A walking assistance robot according to the third embodiment of the present disclosure will be described. In the third embodiment, differences from the first and second embodiments will be mainly described. In the third embodiment, the same or equivalent components as those in the first and second embodiments will be described with the same reference numerals. In the third embodiment, the description overlapping with the first and second embodiments is omitted.

  The third embodiment is different from the first and second embodiments in that the load is corrected based on the body information.

  FIG. 23 is a control block diagram illustrating a main control configuration in the walking assist robot 61 (hereinafter referred to as “robot 61”) according to the third embodiment. Further, the control block diagram of FIG. 23 also shows the relationship between each control configuration and information to be handled. FIG. 24 is a control block diagram showing a detailed control configuration for walking support of the robot 61.

  As shown in FIGS. 23 and 24, the robot 61 of the third embodiment is different from the robots 1 and 51 of the first and second embodiments in that the robot 61 includes a body information estimation unit 25 and a body information database 26. In the third embodiment, the physical information database 26 is not an essential configuration.

  The body information estimation unit 25 estimates the user's body information. In this specification, the body information is body information related to walking, and includes, for example, walking speed, walking rate, body tilt, body shake, stride, and muscle strength. The physical information is not limited to these. For example, the body information may include an average load in the moving direction, an average load in the biased direction of the center of gravity, a fluctuation frequency in the moving direction, a fluctuation frequency in the left-right direction, and the like.

  The body information estimation unit 25, for example, information on the handle load detected by the detection unit 13, information on the amount of rotation of the rotating body 16 acquired by the actuator control unit 21, and the driving calculated by the driving force calculation unit 20. The body information is estimated based on the force information.

  The physical information database 26 stores physical information for each user. The physical information database 26 stores and updates the physical information estimated by the physical information estimation unit 25 for each user.

[Body information estimation]
The estimation of physical information will be described with reference to FIG. FIG. 25 shows an exemplary flowchart of the body information estimation process of the walking support robot 61 according to the third embodiment.

  As shown in FIG. 25, in step ST51, the detection unit 13 detects a change in handle load. When the detection unit 13 detects a change in the handle load, the process proceeds to step ST52. If the detection unit 13 does not detect a change in handle load, step ST51 is repeated.

  In step ST52, the body information estimation unit 25 estimates the moving direction of the user based on the information on the rotation amount of the wheel 16. Specifically, when a change in the handle load is detected in step ST51, the actuator control unit 21 acquires information on the rotation amount of the wheel 16. Information on the rotation amount acquired by the actuator control unit 21 is transmitted to the body information estimation unit 25. The body information estimation unit 25 estimates the moving direction of the user based on the information on the rotation amount of the wheel 16, that is, the rotation direction and the rotation speed of the wheel. In Embodiment 3, the body information estimation part 25 estimates a user's moving direction based on the rotation amount of the two wheels 16 arrange | positioned at right and left. For example, when the rotation amount of the right wheel 16 is larger than the rotation amount of the left wheel 16, the body information estimation unit 25 may estimate that the user is turning leftward. Alternatively, the body information estimation unit 25 may estimate that the robot 61 is performing a straight-ahead operation when the rotation speeds of the left and right wheels 16 are the same and are rotating in the forward direction.

In step ST53, the body information estimation unit 25 acquires handle load waveform information in the estimated movement direction of the user. The handle load waveform information in the user's moving direction is not particularly limited. For example, when the user's moving direction is the Fy ⁺ direction, the handle load waveform information in the Fz direction or the moment waveform information in the My direction may be used. May be.

  In step ST54, the body information estimation unit 25 adds the acquired handle load waveform information and the past handle load waveform information. For example, past waveform information is stored in the body information database 26. The body information estimation unit 25 reads past waveform information from the body information database 26 and adds the acquired current waveform information to the past waveform information. The handle load input waveform information includes, for example, handle load waveform information shown in FIG.

  In step ST55, the body information estimation unit 25 acquires driving force information. Specifically, the body information estimation unit 25 acquires driving force information from the driving force calculation unit 20.

  In step ST56, the body information estimation unit 25 estimates the body information based on the waveform information added in step ST54 and the driving force information acquired in step ST55.

  In Embodiment 3, the body information estimation unit 25 estimates walking speed, walking rate, body tilt, body shake, stride, and muscle strength as body information.

  The walking speed is calculated by, for example, calculating the moving distance based on the driving force information and dividing by the moving time.

The walking rate means the number of steps per unit time. The walking rate is calculated by dividing the number of steps by the travel time. The number of steps is calculated based on the change information of the handle load. For example, when the user is moving straight, the user walks with the right foot and the left foot alternately forward. The waveform information of the handle load of the user who is moving straight ahead changes in conjunction with the walking cycle. For this reason, the handle load waveform information has a peak point that is convex in the Fz ⁺ direction when the toe of the right foot or left foot is separated from the ground, that is, when the toe is away from the ground. Therefore, the number of steps can be calculated by counting one step from the peak point to the next peak point. The method for calculating the convex peak point may be calculated based on, for example, a point at which the amount of change in the handle load changes from increase to decrease, or a quadratic curve is estimated by the least square method, and the estimated quadratic curve is calculated. It may be calculated based on the maximum value of.

The tilt of the body is calculated based on the handle load information. The inclination of the body is calculated based on the load bias caused by the inclination of the user's center of gravity. For example, for a user who walks with the center of gravity biased to the right, the load in the Fx ⁺ direction is calculated as the body inclination.

  The body shake is calculated by calculating the fluctuation frequency based on the combined waveform information. Specifically, the body information estimation unit 25 calculates a fluctuation frequency by performing a frequency analysis of the handle load in the estimated movement direction of the user.

The stride is calculated based on the moving distance between the toes. For example, the movement distance between the toes is calculated based on the handle load waveform information and the driving force information. As described above, the handle load waveform information and the walking cycle change in conjunction with each other. For example, in the handle load waveform information, the body information estimation unit 25 counts one peak from the peak point convex in the Fz ⁺ direction to the next peak point, and calculates the load in the Fx direction and / or the moment in the Mz direction. The right foot or the left foot is specified based on this. Next, the body information estimation unit 25 calculates the movement distance for each step based on the driving force information.

  The muscular strength is calculated from the bias of the load value for each foot position, the difference between the left and right stride, the difference in the movement amount, and the like. For example, the muscular strength is expressed by six-level evaluation (levels 0 to 5) for each leg muscle (for example, anterior tibialis aneurysm, peroneus muscle) used for each user's walking motion. In addition, a level shows that muscular strength is so strong that a number becomes large.

  In Embodiment 3, the above-described physical information data is calculated based on information for 10 steps. Specifically, an average value of data for 10 steps is calculated as physical information. The physical information is not limited to the average value of data for 10 steps. For example, the physical information is not limited to data for 10 steps, but is based on data of 1 step or more and less than 10 steps, data on the number of steps greater than 10 steps, or (data for 10 steps) × (multiple data). May be calculated. Further, the physical information may be calculated by a median value in addition to the average value of the data for 10 steps.

  The physical information data calculated in this way is stored in the physical information database 26. Further, the physical information stored in the physical information database 26 is updated to new information when the physical information is estimated.

  FIG. 26A is an example of physical information stored in the physical information database 26 of the robot 61. As shown in FIG. 26A, as the body information of the user A, walking speed, walking rate, body tilt, body shake, stride, and leg muscle strength are used.

  FIG. 26B is another example of the physical information stored in the physical information database 26 of the robot 61. As shown in FIG. 26B, the body information of the straight movement of the user A includes walking speed, walking rate, average load in the moving direction, average load in the biased direction of the center of gravity, fluctuation frequency in the moving direction, fluctuation frequency in the left-right direction, step length , Using the muscle strength of the foot. In addition, the body information shown in FIG. 26B indicates that the input waveform of the handle load is the sum of “No. 1” and “No. 3”.

[Estimation of user's intention to move]
In the third embodiment, the handle load is corrected based on the body information, and the user's intention to move is estimated based on the corrected handle load information. FIG. 27 shows an exemplary flowchart of a process of estimating the movement intention of the user of the robot 61.

  As illustrated in FIG. 27, in step ST61, the detection unit 13 acquires handle load information.

  In step ST62, the body information estimation unit 25 acquires body information corresponding to the moving direction of the user based on the handle load information. Specifically, the body information estimation unit 25 estimates the moving direction of the user based on the handle load information. Next, the body information estimation unit 25 acquires body information corresponding to the estimated movement direction of the user from the body information database 26. The body information estimation unit 25 transmits the acquired body information to the load correction unit 18.

  In step ST63, the load correction unit 18 sets a threshold value as to whether or not to correct the handle load information based on the body information. In step ST63, the load correction unit 18 adjusts the intensity of the correction by setting a threshold value for correcting the handle load information in accordance with the user's physical ability.

  In the third embodiment, the threshold value indicating whether or not to apply correction means a threshold value indicating whether or not to correct body tilt (load applied in the tilt direction), body shake (fluctuation frequency), and the like.

  In this specification, the intensity of correction means the ease with which correction is performed. A high correction intensity means that correction is easily performed, and a low correction intensity means that correction is difficult to apply.

  For example, a user with high physical ability can walk as intended by the user without correcting the handle load. In this case, since the user wants to move quickly, the handle load may be corrected to be small even if the load is increased in the straight direction. In order to avoid this, the load correction unit 18 sets the threshold value high and reduces the correction strength. On the other hand, a user with low physical ability has a large tilt or shake of the body, and it is difficult to walk as intended by the user unless the handle load is corrected. In this case, the load correction unit 18 sets the threshold value low and increases the correction strength.

  In the present specification, a user having high physical ability means, for example, a user having physical ability equal to or greater than the average physical ability of people of the same age as the user. In addition, a user having low physical ability means, for example, a user having physical ability lower than the average physical ability of people of the same age as the user.

  In Embodiment 3, the load correction | amendment part 18 determines the height of a user's physical ability on the basis of the average body information of the person of the same age as a user. Specifically, the load correction unit 18 uses the average value of walking speeds of people of the same age as the user (hereinafter referred to as “average walking speed of the same age”) as the physical information serving as a reference, and uses the physical ability of the user. Judging the height. For example, when the user's walking speed is equal to or higher than the average walking speed of the same age, the load correction unit 18 determines that the physical ability is high, and when the user's walking speed is slower than the average walking speed of the same age, the physical ability is low. judge.

  The average physical information of people of the same age as the user is stored in the physical information database 26, for example. Note that the average physical information of people of the same age as the user means, for example, average physical information of people aged 63 when the user is 63 years old. The average physical information of people of the same age as the user is stored in the physical information database 26, for example. In Embodiment 3, when the body information estimation part 25 transmits a user's body information to the load correction | amendment part 18, you may transmit the average body information of the person of the same age as a user.

  When it is determined that the user's physical ability is high, that is, the user's walking speed is equal to or higher than the average walking speed of the same age, the load correction unit 18 sets a high threshold value for whether or not to apply correction. On the other hand, when it is determined that the user's physical ability is low, that is, the user's walking speed is slower than the average walking speed of the same age, the load correction unit 18 sets a low threshold for whether or not to apply correction. As described above, in the third embodiment, the walking speed is used as a reference for determining a threshold value for determining whether or not to apply correction.

  As an example, a case where the load correction unit 18 determines that the user's walking speed is equal to or higher than the average walking speed of the same age will be described. In this example, when the average value of the load in the straight line direction of people of the same age as the user is 20N, the load correction unit 18 sets the correction threshold of the body inclination to 20N. Moreover, when the average value of the straight-direction swing of people of the same age as the user is 1.0 Hz, the load correction unit 18 sets the correction threshold of the body swing to 1.0 Hz. Thus, for users with high physical ability, the threshold value can be set high to make correction difficult.

  Moreover, the case where the load correction | amendment part 18 determines with a user's walking speed being slower than the average walking speed of the same age is demonstrated. In this example, in the case of a user with low physical ability, correction according to the load tendency may be always performed. In addition, the form which sets a threshold value lower than the threshold value of a user with high physical ability may be sufficient so that correction may be easily applied to a user with low physical ability.

  In step ST64, the load correction unit 18 determines whether or not the handle load information is greater than or equal to the threshold set in step ST63. If the handle load information is greater than or equal to the threshold value, the process proceeds to step ST65 to correct the handle load information. When the handle load information is less than the threshold value, the process proceeds to step ST67.

  As an example, when it is determined that the user's walking speed is equal to or higher than the average walking speed of the same age, the load correction unit 18 is applied when a load of 20 N or more is applied in the straight traveling direction, or the swing in the straight traveling direction is 1. When it becomes 0 Hz or more, the process proceeds to step ST65. On the other hand, when a load of less than 20N is applied in the straight direction and the fluctuation in the straight direction is less than 1.0 Hz, the load correction unit 18 proceeds to step ST67.

  When it is determined that the user's walking speed is slower than the average walking speed of the same age, the load correction unit 18 proceeds to step ST65 when the straight-direction swing is 0 Hz or more. In the third embodiment, since the threshold value for correcting the straight direction swing is set to 0 Hz, when the handle load is applied, the correction is substantially applied.

  In step ST65, the load correction unit 18 corrects the handle load information based on the load tendency data (fluctuation frequency and the like). Since step ST65 is the same as the load correction process of the first and second embodiments, the description thereof is omitted.

  In step ST66, the user movement intention estimation unit 19 estimates the user's movement intention (movement direction, movement speed) based on the corrected handle load information. Since step ST66 is the same as the user movement intention estimation process in the first and second embodiments, the description thereof is omitted.

  In step ST67, the load correction unit 18 estimates the user's intention to move without correcting the handle load information. About step ST67, since it is the same as the process of user movement intention estimation of Embodiment 1 and Embodiment 2, description is abbreviate | omitted.

  As described above, the robot 61 performs walking support according to the physical ability of the user by setting a threshold value for determining whether or not to correct the load based on the physical information.

[effect]
According to the walking assist robot 61 according to Embodiment 3, the following effects can be achieved.

  According to the walking assist robot 61 according to the third embodiment, the threshold value for determining whether or not to correct the load is set based on the body information. With such a configuration, the intensity of correction can be adjusted, so that walking support according to the user's physical ability can be performed.

  For example, a user with high physical ability can walk as intended by the user without correction. In this case, the robot 61 can suppress the correction of the handle load from being excessively applied by increasing the threshold value for determining whether or not to apply the correction. For example, it is possible to suppress control contrary to the user's intention such that if the user wants to go straight ahead quickly, the speed is suppressed by the correction even though the handle load is applied.

  In addition, a user with low physical ability may receive a load on the handle that is not intended by the user. In this case, the robot 61 makes it easier to correct the handle load by lowering the threshold value for determining whether or not to apply the correction. Thereby, even a user with low physical ability can walk as the user intends.

  In the third embodiment, the example in which the walking speed is used as a reference for determining the threshold value for whether or not to apply correction is described. However, the present invention is not limited to this. The reference for determining the threshold may be set based on physical information, and for example, walking speed, walking rate, stride, muscle strength, load bias, load swing, or a combination thereof may be used. By using these items as criteria for determining the threshold value, the intensity of correction can be set more finely.

  In addition, as a criterion for determining the threshold, an average value, a median value, or the like of physical information (walking speed, walking rate, etc.) of people of the same age as the user may be used.

  In the third embodiment, when it is determined that the user's physical ability is low, that is, the user's walking speed is slower than the average walking speed of the same age, an example of setting the threshold for correcting the body shake in the straight direction to 0 Hz Although described, it is not limited to this. When the user's physical ability is low, for example, the body shake correction threshold value may be set to a different value such as 0.5 Hz, or the body tilt correction threshold value may be set to 10 N in the straight direction. .

  In Embodiment 3, when it is determined that the user's physical ability is high, that is, the user's walking speed is equal to or higher than the average walking speed of the same age, an average value 20N of the load of the body inclination of people of the same age as the user, Although the example which sets the average value 1.0Hz of a body shake as a threshold value was demonstrated, it is not limited to this. When the user's physical ability is high, for example, the threshold value may be set to a different value or may be set not to be corrected.

  In Embodiment 3, the load correction | amendment part 18 demonstrated the example which sets the threshold value according to a user's physical ability by determining the height of a user's physical ability based on the physical information of people of the same age as a user. However, it is not limited to this.

  As an example, the load correction part 18 demonstrates the example which sets a threshold value according to the physical ability according to age. In this example, a threshold is set for each age. For example, in the ages of 50 to 60 years old, the ages of 60 to 70 years of age, and the ages of 70 to 80 years of age, the load threshold in the traveling direction is set to the average value of each age 15N, 10N, 5N. The threshold value of the shaking in the traveling direction may be set to an average value of each age of 1.0 Hz, 1.2 Hz, and 1.4 Hz.

  In this example, the load correction unit 18 may determine which age the body has a physical ability based on the user's physical information, and may set a threshold corresponding to the age. For example, when the load correction unit 18 determines that the user's physical ability corresponds to the physical ability of the age of 60 years old or older and less than 70 based on the user's physical information (walking speed, walking rate, etc.), the load in the traveling direction May be set to 10 N, and the threshold value of the shaking in the traveling direction may be set to 1.2 Hz.

  In the third embodiment, the example of setting the threshold when the physical ability of the user is high and low is described, but the present invention is not limited to this. For example, when it is determined that the user's physical ability is low, as described above, a threshold set according to the physical ability by age group may be set. For example, when it is determined that the user's physical ability is low, the threshold value may be changed stepwise according to the physical ability of each age group.

  In Embodiment 3, although the body information estimation part 25 demonstrated the example which estimates body information based on the waveform information added by step ST54, and the information of the driving force acquired by step ST55, it is limited to this. Not. For example, the body information estimation unit 25 may estimate the body information based on the waveform information added in step ST54 and the rotation amount of the rotating body 16 measured by the actuator control unit 21.

[User Notification Section]
FIG. 28 shows another control block diagram illustrating a control configuration of walking support of the robot 61. As shown in FIG. 28, the robot 61 may include a user notification unit 27.

  The user notification unit 27 notifies the user of at least one of physical information and user movement intention. Specifically, the user notification unit 27 acquires physical information estimated from the physical information estimation unit 25. In addition, the user notification unit 27 acquires information on the user's movement intention from the user movement intention estimation unit 19.

  The user notification unit 27 includes, for example, an LED, a display, or a speaker. In addition, the user notification part 27 may be comprised by LED, a display, a speaker, or these combination.

  A case where the user notification unit 27 includes an LED will be described. For example, when the user notification unit 27 acquires physical information, the user notification unit 27 may correct the load based on the physical information and turn on the LED when the user's intention to move is estimated. In addition, you may identify the information to present according to the lighting pattern of LED.

  A case where the user notification unit 27 has a display will be described. When the user notification unit 27 acquires physical information, for example, a message such as “Your walking speed is OO”, “Walking rate is OO”, or “Right leg muscle strength is weak” is displayed on the display. It may be displayed. In addition, when the user notification unit 27 corrects the load based on the body information and estimates the user's intention to move, for example, “support to you” on the display, “change the control to suit you” ”,“ Strengthen brake ”,“ Suppress shaking ”,“ Stabilize ”, etc. may be displayed. Note that the message displayed on the display is not limited to these.

  A case where the user notification unit 27 includes a speaker will be described. When the user notification unit 27 acquires physical information, for example, the speaker outputs sounds such as “Your walking speed is XX”, “Walking rate is XX”, “Right leg muscle strength is weak”, etc. May be. In addition, when the user notification unit 27 corrects the load based on the physical information and estimates the user's intention to move, the user notification unit 27, for example, uses a speaker to “support to you” or “change control to suit you” ”,“ Strengthen brake ”,“ Suppress shaking ”,“ Stabilize ”, etc. may be output. Note that the sound output by the speaker is not limited to these.

  Thus, by providing the user notification unit 27, the user can acquire body information or robot walking support information visually and / or auditoryly.

  By the user notification unit 27 being notified, the user himself / herself grasps daily physical information, which leads to motivation to maintain and improve physical ability, or alertness at the time of walking.

  In addition, the user notification unit 27 notifies the user of the control state of the robot 61, and can adapt to a large change in operational feeling such that the brake is strengthened.

  Although this disclosure has been described in some embodiments with some detail, the disclosure of these embodiments should vary in configuration details. In addition, combinations of elements and changes in the order in each embodiment can be realized without departing from the scope and spirit of the present disclosure.

  The correction of the handle load based on the load tendency data described in Embodiment 1-3 is an example, and the present invention is not limited to these. Various known correction methods may be employed as the correction of the handle load based on the load tendency data. As a correction method, for example, with respect to fluctuation in the center of gravity direction, a method of performing smoothing with a moving average according to the degree of fluctuation, a method of removing fluctuations by performing smoothing with a median filter, and performing frequency analysis You may employ | adopt the method of cutting and reducing a specific frequency.

  The present disclosure can be applied to a walking support robot and a walking support method that perform walking support of a more comfortable user.

DESCRIPTION OF SYMBOLS 1, 51, 61 Walking assistance robot 11 Main part 12 Handle part 13 Detection part 14 Moving device 15 Load tendency data generation part 16 Rotating body 17 Drive part 18 Load correction part 19 User movement intention estimation part 20 Driving force calculation part 21 Actuator control Unit 22 Actuator 23 User load tendency extraction unit 24 Load trend map 25 Physical information estimation unit 26 Physical information database 27 User notification unit

Claims (19)

A walking support robot that supports user walking,
The main body,
A handle portion provided on the main body portion and grippable by the user;
A detection unit for detecting a load applied to the handle unit;
A moving device that has a rotating body and moves the walking support robot by controlling the rotation of the rotating body according to the load detected by the detection unit;
Based on past load data applied to the handle part acquired during the movement of the walking support robot, a load tendency data generation unit that generates load tendency data indicating a load tendency applied to the handle part;
With
The moving device includes an actuator that controls the amount of rotation of the rotating body based on the load detected by the detection unit and the load tendency data generated by the load tendency data generation unit.
Walking support robot. The moving device further includes a load correction unit that corrects a load value detected by the detection unit based on the load tendency data.
The actuator controls the amount of rotation of the rotating body based on the load value corrected by the load correction unit.
The walking support robot according to claim 1. The load tendency data generation unit generates load tendency data for each movement operation of the walking support robot,
The load correction unit corrects the value of the load based on the load tendency data corresponding to the movement operation of the walking support robot when the load is detected.
The walking support robot according to claim 2. The load correction unit corrects the value of the load based on the load tendency data when the load tendency data corresponding to the movement operation of the walking support robot is equal to or greater than a predetermined threshold.
The walking support robot according to claim 3. The load tendency data is a fluctuation frequency calculated from the past load data,
The load correction unit corrects the value of the load by filtering the frequency component of the fluctuation from the load detected by the detection unit,
The walking support robot according to any one of claims 2 to 4. The load tendency data is an average load value calculated from the past load data,
The load correction unit corrects the load value based on the average load value.
The walking support robot according to any one of claims 2 to 4. The load correction unit corrects the value of the load by subtracting the average load value from the load detected by the detection unit.
The walking support robot according to claim 6. Furthermore,
A body information estimation unit for estimating the user's body information;
The load correction unit corrects the value of the load based on the body information estimated by the body information estimation unit.
The walking support robot according to any one of claims 2 to 7. Furthermore,
The walking support robot according to claim 8, further comprising a user notification unit that notifies the user of the physical information. Furthermore,
A user movement intention estimation unit that estimates the movement intention of the user based on the corrected load value;
The user notification unit notifies the user of the user's intention to move.
The walking support robot according to claim 9. Furthermore,
A storage unit for storing a control table indicating a correspondence relationship between a load applied to the handle unit and a rotation amount of the rotating body;
The actuator drives and controls the rotating body with a rotation amount corresponding to the load detected by the detection unit by the control table stored in the storage unit,
The control table is updated by correcting the load value based on the load trend data.
The walking support robot according to claim 1. The detection unit detects a plurality of axial loads applied to the handle unit,
The moving device switches the movement operation of the walking support robot by controlling the rotation of the rotating body according to the load applied to each of the plurality of axial directions.
The walking support robot according to any one of claims 1 to 11.   The walking support robot according to claim 12, wherein the moving operation includes a straight movement operation, a backward movement operation, and a turning operation of the walking support robot.   The walking assist robot according to claim 13, wherein the actuator changes a turning radius in the turning motion based on the load tendency data. A walking support method for supporting a user's walking using a walking support robot,
Detecting a load applied to a handle portion of the walking support robot by a detection unit;
Generating load trend data indicating a tendency of a load applied to the handle unit based on past load data applied to the handle unit acquired during the movement of the walking support robot;
Controlling the amount of rotation of the rotating body provided in the moving device of the walking support robot based on the load detected by the detection unit and the load tendency data;
A walking support method. Furthermore,
Based on the load tendency data, including a step of correcting the value of the load detected by the detection unit,
The step of controlling the amount of rotation of the rotator controls the amount of rotation of the rotator based on the corrected load value.
The walking support method according to claim 15. Furthermore,
Estimating the user's physical information,
The step of correcting the value of the load corrects the value of the load based on the physical information.
The walking support method according to claim 16. Furthermore,
The walking support method according to claim 17, comprising notifying the user of the physical information. Furthermore,
Estimating a user's intention to move based on the corrected load value,
The notifying step notifies the user of the user's intention to move;
The walking support method according to claim 18.

Images