Method for changing walking mode of walking robot having plurality of legs, and robot using same

Abstract

The present invention relates to a method for changing the walking mode of a walking robot having a plurality of legs, and a robot using same, the method comprising the steps of: in a first walking mode in which terrain is recognized on the basis of vision sensor information and a walking trajectory is determined accordingly, swinging a first leg toward a first touchdown position according to the trajectory determined on the basis of the vision sensor information; and a step of changing to a second walking mode in which a walking trajectory is determined on the basis of flat terrain without using the vision sensor information if a change in terrain, different from the terrain recognized on the basis of the vision sensor information, is detected during the swinging motion of the first leg, wherein, in the second walking mode, terrain information is constructed using the actual landing position of the first leg, and a trajectory for the swinging motion of a second leg following the first leg is determined on the basis of the constructed terrain information.

Description

A method for switching walking modes of a multi-legged walking robot and a robot using the same

[0001] The present invention relates to a method for a robot having multiple legs to switch walking modes depending on surrounding conditions such as terrain.

[0002] A robot is a multifunctional manipulator designed to perform specific movements and tasks through programmed, variable motions. Robots are widely used in various industries, including manufacturing, transportation, exploration, medicine, surveillance, and patrol.

[0003] Robots can be physically fixed in a specific location, such as an industrial robot arm, but they can also be configured to be mobile, with one or more legs or wheels. Mobile robots offer a wider range of applications and uses than fixed robots.

[0004] A representative example of a mobile robot is a multi-legged robot with one or more legs. Multi-legged robots must avoid stepping on or colliding with obstacles while moving, and walking control technology is required to maintain stable balance and speed while avoiding contact with obstacles.

[0005] The technical problem to be solved through embodiments of the present invention is to provide a method for switching the walking mode of a multi-legged walking robot, which enables the robot to walk stably by automatically switching to an appropriate walking mode according to surrounding conditions such as terrain, and a robot using the same.

[0006] The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art of the present invention from the description below.

[0007] In order to solve the above technical problem, a method for switching a walking mode of a multi-legged walking robot according to embodiments of the present invention includes: a step of swinging a first leg toward a first touchdown position according to a trajectory determined based on the vision sensor information in a first walking mode in which a terrain is recognized based on vision sensor information and a walking trajectory is determined; and a step of switching to a second walking mode in which a walking trajectory is determined based on a flat surface without using the vision sensor information when a change in terrain different from the terrain recognized based on the vision sensor information is detected during the swinging motion of the first leg.

[0008] The second walking mode configures terrain information using the actual landing position of the first leg, and determines a trajectory for a swing motion of the second leg following the first leg based on the configured terrain information.

[0009] At least some of the methods for switching walking modes of the above multi-legged walking robot can be implemented as a computer-readable recording medium recording a program for executing on a computer, and can be provided as the program itself.

[0010] Meanwhile, the method for switching the walking mode of the multi-legged walking robot can be performed by the multi-legged walking robot according to an embodiment of the present invention.

[0011] In addition, a multi-legged walking robot according to one embodiment of the present invention comprises: a processor; a memory for loading a computer program executed by the processor; and a storage for storing the computer program, wherein the computer program comprises: an operation for swinging a first leg toward a first touchdown position according to a trajectory determined based on the vision sensor information in a first walking mode for recognizing terrain and determining a walking trajectory based on vision sensor information; and

[0012] If a change in terrain that is different from the terrain recognized based on the vision sensor information is detected during the swing motion of the first leg, instructions may be included for executing an operation of switching to a second walking mode that determines a walking trajectory based on flat ground without using the vision sensor information.

[0013] According to an embodiment of the present invention, when a multi-legged walking robot walks blindly without using vision sensor information, if a change in terrain is detected while the robot is moving according to a trajectory determined based on flat ground, blind walking control is applied to a leading leg, and terrain information acquired according to the blind walking control is applied to a trailing leg to determine a swing trajectory, thereby preventing an environment in which unstable movement of the robot may occur from being repeated, and thereby enabling more stable blind walking.

[0014] According to another embodiment of the present invention, a first walking mode that recognizes terrain based on vision sensor information and determines a walking trajectory, and a second walking mode that determines a walking trajectory based on flat ground without using vision sensor information, are mutually switched according to changes in terrain, thereby enabling the walking of a multi-legged walking robot using a vision system to be implemented more stably.

[0015] The technical effects of the present invention are not limited to the technical effects mentioned above, and other technical effects not mentioned will be clearly understood by those skilled in the art from the description below.

[0016] FIG. 1 is a drawing showing exemplary forms of a multi-legged walking robot having multiple legs according to an embodiment of the present invention.

[0017] Figure 2 is a drawing for explaining an example of a leg swing motion for walking of a multi-legged walking robot.

[0018] FIG. 3 is a flowchart illustrating a blind walking control method of a multi-legged walking robot according to an embodiment of the present invention.

[0019] FIGS. 4 to 9 are drawings for explaining embodiments of a method for controlling blind walking of a multi-legged walking robot.

[0020] Figures 10 to 13 are drawings for explaining embodiments of terrain information configured according to the movement of a multi-legged walking robot and the swing motion of the legs according to the terrain information.

[0021] Fig. 14 is a flowchart showing a method for switching walking modes of a multi-legged walking robot according to one embodiment of the present invention.

[0022] Figures 15 to 18 are drawings for explaining embodiments of a method for switching the walking mode of a multi-legged walking robot.

[0023] Fig. 19 is a flowchart showing a method for switching walking modes of a multi-legged walking robot according to another embodiment of the present invention.

[0024] FIG. 20 is a drawing for explaining embodiments of a method for restoring the walking mode of a multi-legged walking robot.

[0025] Fig. 21 is a block diagram showing the configuration of a multi-legged walking robot according to one embodiment of the present invention.

[0026] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. The advantages and features of the present invention, and methods for achieving them, will become clearer with reference to the embodiments described in detail below together with the attached drawings. However, the technical idea of ​​the present invention is not limited to the following embodiments, but can be implemented in various different forms. The following embodiments are provided only to complete the technical idea of ​​the present invention and to fully inform those skilled in the art of the present invention of the scope of the present invention, and the technical idea of ​​the present invention is defined only by the scope of the claims.

[0027] When assigning reference numerals to components in each drawing, it should be noted that identical components are assigned the same numerals whenever possible, even if they appear on different drawings. Furthermore, when describing the present invention, if a detailed description of a related known configuration or function is deemed likely to obscure the gist of the present invention, such detailed description will be omitted.

[0028] Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used in their common sense to those of ordinary skill in the art to which the present invention pertains. Furthermore, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise. The terminology used herein is for the purpose of describing embodiments and is not intended to limit the present invention. In this specification, singular forms also include plural forms, unless specifically stated otherwise.

[0029] Additionally, terms such as first, second, A, B, (a), (b), etc. may be used to describe components of the present invention. These terms are only intended to distinguish the components from other components, and the nature, order, or sequence of the components are not limited by the terms. When it is described that a component is "connected," "coupled," or "connected" to another component, it should be understood that the component may be directly connected or connected to the other component, but another component may also be "connected," "coupled," or "connected" between each component.

[0030] Hereinafter, some embodiments of the present invention will be described in detail with reference to the attached drawings.

[0031] FIG. 1 is a schematic diagram illustrating an exemplary form of a multi-legged walking robot having a plurality of legs according to one embodiment of the present invention.

[0032] Referring to FIG. 1, the robot (100) is a robot capable of walking movement, and may be a humanoid robot with two legs as shown in (a) of FIG. 1, or a quadruped walking robot with four legs as shown in (b) of FIG. 1.

[0033] However, the scope of the present invention is not limited to the shape of the robot illustrated in FIG. 1. For example, the robot (100) may have three, five, or more legs, and may have various shapes other than a humanoid robot or a quadruped robot.

[0034] Referring to (a) of FIG. 1, the robot (100) may include a main body (110) and two legs (120, 121), and each leg (120, 121) may be connected to the main body (110) and have an upper part and a lower part separated by a leg joint.

[0035] Meanwhile, referring to (b) of FIG. 1, the robot (100) may include a main body (115) and four legs (125, 126, 127, 128), and each leg (125, 126, 127, 128) may be connected to the main body (110) and have an upper part and a lower part separated by a leg joint.

[0036] For example, the robot (100) may further include one or more appendages, such as an articulated arm, which is arranged on the main body (110, 115) and configured to move relative to the main body (110, 115). One or more tools, such as a gripper for grasping / holding objects, may be provided at the end of the articulated arm.

[0037] The robot (100) may include a vision system having one imaging sensor or camera, for example, one or more cameras and one or more two-dimensional or three-dimensional LiDAR sensors may be mounted on the robot (100), but the present invention is not limited thereto.

[0038] Each sensor or camera mounted on the robot (100) collects image data or sensor data regarding the surrounding environment and terrain. The vision system can pan or tilt the camera to move the field of view of the robot (100) in any direction. The image data or sensor data collected by the cameras or sensors of the vision system can be provided to the central processing unit of the robot (100).

[0039] In addition, the robot (100) may be additionally equipped with sensors for measuring the dynamic state of the robot in addition to the vision sensors such as the above-mentioned camera and lidar sensor.

[0040] For example, the dynamic state of the robot (100) may include the inertia of the robot, the angle of each joint, force, posture, position or state, the velocity or acceleration of each component connected to the joint, the center of gravity of the robot, etc.

[0041] In order to measure the dynamic state of the robot (100) as described above, the robot (100) may include an IMU (Inertial Measurement Unit) sensor for measuring inertia, an angle sensor for measuring the angle of the joint, a current sensor for measuring the force of the joint, etc., but the present invention is not limited thereto.

[0042] The robot (100) walks by swinging each leg along a pre-planned trajectory. The leg swing cycle includes a lifting phase and a touchdown phase. The lifting phase refers to the period from when the leg touches the ground to when it reaches the highest point of the swing trajectory, and the touchdown phase refers to the period from when the leg descends from the highest point of the swing trajectory to when it lands (or touches down) on the ground.

[0043] Referring to Figure 2, the leg swing cycle of the robot when walking is illustrated. The leg swing cycle begins at time t0, when the leg is in contact with the ground, passes through time t1, when the leg swings to its highest point, and ends at time t2, when the leg lands on the ground again.

[0044] At this time, the section from t0 to t1, when the leg is lifted from the ground and reaches the highest point, is the lifting phase, and the section from t1 to t2, when the leg is lowered from the highest point and until the leg lands on the ground again, is the touchdown phase.

[0045] A robot (100) according to one embodiment of the present invention may be capable of blind walking based only on the dynamic state of the robot (100) without using vision sensors such as a camera and a lidar sensor.

[0046] Blind walking can be performed when the robot (100) is not equipped with a vision system such as a camera or a lidar sensor, but the present invention is not limited thereto, and even when the robot (100) is equipped with a vision system, it can be performed when walking using vision sensors is difficult depending on the surrounding circumstances.

[0047] According to an embodiment of the present invention, when a multi-legged walking robot (100) walks blindly without using vision sensor information, if a change in terrain is detected while the robot (100) moves according to a trajectory determined based on flat ground, blind walking control is applied to a preceding leg, and terrain information acquired according to the blind walking control is applied to a following leg to determine a swing trajectory, thereby preventing an environment in which unstable movement of the robot may occur from being repeated, and thereby enabling more stable blind walking.

[0048]

[0049] Hereinafter, with reference to FIGS. 3 to 9, embodiments of a blind walking control method of a multi-legged walking robot according to the present invention will be described in more detail.

[0050] FIG. 3 is a flowchart illustrating a blind walking control method of a multi-legged walking robot according to an embodiment of the present invention.

[0051] Referring to FIG. 3, the robot (100) swings the first leg toward the first touchdown position according to a trajectory determined based on the flat ground (step S300).

[0052] When a robot (100) walks blindly, assuming that the terrain is flat, the walking trajectory can be controlled based on the flat terrain.

[0053] For example, as illustrated in FIG. 4, the first touchdown position (A), which is the position where the first leg (125) of the robot (100) lands, can be determined under the assumption that the terrain of the ground (400) on which the robot (100) walks is flat.

[0054] Meanwhile, after the first leg (125) of the robot (100) lands at the first touchdown position (A), the second leg (127) that follows can swing and land, and the touchdown position where the second leg (127) lands can also be determined assuming that the terrain of the ground (400) is flat.

[0055] Here, the third leg (126) can be swung together with the second leg (127), and the touchdown position where the third leg (126) lands can also be determined based on the flat ground.

[0056] And the fourth leg (128) can be swung together with the first leg (125), and the touchdown position where the fourth leg (128) lands can also be determined based on the flat ground.

[0057] The swing motion and trajectory of the robot (100) described with reference to FIG. 4 are examples for explaining blind walking, and the present invention is not limited thereto, and the swing order of each leg (125, 126, 127, 128) may be changed.

[0058] If a change in terrain is detected during the swing motion of the first leg (step S310), the swing motion of the first leg is controlled so that the first leg lands at a second touchdown position different from the first touchdown position (step S320).

[0059] Meanwhile, if no change in terrain is detected at step S310, the remaining trailing legs also swing according to the landing position determined based on flat ground, so that the robot (100) can walk.

[0060] The change in terrain detected at step S310 may include at least one of a change in the height and a change in the slope of the ground on which the robot (100) walks.

[0061] For example, as illustrated in FIG. 5, a second ground (410) higher than a first ground (400), which is flat, may exist on the terrain on which the robot (100) walks, and a slope (530) may be formed between the first ground (400) and the second ground (410).

[0062] In this case, due to the height difference (H) between the first ground (400) and the second ground (410), the first touchdown position (A), which is the landing position of the first leg (125) determined based on the flat ground, is located below the ground, and accordingly, the first leg (125) of the robot (100) may collide with the second ground (410) before landing at the first touchdown position (A).

[0063] In this way, if an impact due to a collision with the second ground (410) is detected before the first leg (125) lands at the first touchdown position (A), it can be determined that the terrain of the ground on which the robot (100) walks has changed.

[0064] For example, in step S310, an impact applied to the first leg (125) can be detected based on the dynamic state of the robot (100) measured in the touchdown phase of the first leg (125), and the dynamic state of the robot (100) can include at least one of inertial information about the robot, angle information of the joint, and force information of the joint.

[0065] Specifically, while the robot (100) is walking, the inertia of the robot can be measured by the IMU sensor, and the force of each joint can be measured through the current sensor along with the angle of each joint, and based on the inertia, angle of the joint, and force measured as described above during the process of the first leg (125) landing at the first touchdown position (A), the impact applied when the first leg (125) comes into contact with the second ground (410) can be detected.

[0066] In the process of controlling the walking trajectory of the robot (100) based on flat ground, if an impact due to a change in terrain as described above is detected, a blind walking control algorithm can be applied to overcome a certain range of height changes or slope changes and allow the leg to land at a new touchdown position.

[0067] For example, if an impact is detected before the first leg (125) of the robot (100) lands at the first touchdown position (A), the location where the impact is detected and the dynamic state of the robot (100) can be considered to determine the location as the new touchdown position.

[0068] Referring to FIG. 6, when the robot (100) is walking on a first ground (400) which is a flat ground, the first touchdown position (A), which is the next landing position of the first leg (125), is set based on the flat ground, and the first touchdown position (A) can be located below a second ground (410) which is higher than the first ground (400).

[0069] Referring to FIG. 7, when the first leg (125) swings toward the first touchdown position (A) and lands, the first leg (125) collides with the second ground (410) due to the height difference (H), and accordingly, an impact can be detected based on the dynamic state of the robot (100) as described above.

[0070] At this time, if the height (H) of the second ground (410) is below a certain height, the height difference (H) is overcome according to the blind walking control algorithm, and the landing position of the first leg (125) can be changed from the previously set first touchdown position (A) to the second touchdown position (B) on the second ground (410).

[0071] When the first leg (125) lands at the second touchdown position in step S320, the robot (100) configures terrain information including the second touchdown position (step S330).

[0072] In step S330, the terrain information may include three-dimensional coordinate information of the second touchdown position, and the three-dimensional coordinate information may indicate the second touchdown position on the robot's internal coordinate system.

[0073] Meanwhile, terrain information is stored in memory in the form of a virtual map on the robot's internal coordinate system, including information on the locations where the robot's (100) legs land while walking, and can be used thereafter to control the walking trajectory of the robot (100).

[0074] To this end, while the robot (100) walks, 3D coordinate information corresponding to each landing position of the first leg (125) is collected, and coordinate information for the terrain on which the robot (100) walks is generated by interpolating between adjacent landing positions of the first leg (125) based on the collected 3D coordinate information, and the generated coordinate information for the terrain can be stored in memory.

[0075] That is, by interpolating and connecting adjacent landing positions that are sequentially collected as described above, information about the terrain on which the robot (100) walks can be created as a virtual map.

[0076] Thereafter, the robot (100) determines a trajectory for the swing motion of the second leg following the first leg based on the terrain information configured in step S330 (step S340).

[0077] For example, the landing position of the second leg may be determined as the third touchdown position based on flat ground, and if a change in terrain is detected during the touchdown phase of the first leg and the landing position of the first leg is changed, the landing position of the second leg may be changed from the existing third touchdown position to the fourth touchdown position.

[0078] At this time, in the three-dimensional coordinates of the robot's internal coordinate system, the third touchdown position of the second leg determined based on the flat ground and the first touchdown position of the first leg can be determined to have the same height, and the fourth touchdown position, which is the new landing position of the second leg, can be determined to have the same height as the changed second touchdown position of the first leg.

[0079] Referring to FIG. 8, when the first leg (125) of the robot (100) lands on the second ground (410), the landing position of the second leg (127) following it is set to the third touchdown position (C) assuming a flat ground, and the third touchdown position (C) can be located below the second ground (410).

[0080] In this case, based on the terrain information (e.g., 3D coordinate information for the second touchdown position (B) on the second ground (410)) identified from the swing motion of the first leg (125) in advance, the landing position of the second leg (127) can be changed to the fourth touchdown position (D) on the second ground (410).

[0081] Referring to FIG. 9, the landing position of the second leg (127) is preset to change from the third touchdown position (C) to the fourth touchdown position (D) on the second ground (410), so that the second leg (127) can swing toward the fourth touchdown position (D) according to the changed trajectory.

[0082] As described above, the second leg (127) following the trailing leg starts the swing motion with the landing position changed to the fourth touchdown position (D) on the second ground (410), so that the impact due to the collision with the second ground (410) when the first leg (125) lands does not occur repeatedly, and the blind walking control algorithm for overcoming the height difference (H) of the ground may not need to be repeatedly performed.

[0083]

[0084] Hereinafter, with reference to FIGS. 10 to 13, examples of terrain information configured according to the movement of a multi-legged walking robot and swing motions of the legs according to the terrain information will be described in more detail.

[0085] Referring to FIG. 10, a virtual map (1000) having three-dimensional coordinates on the robot's internal coordinate system representing terrain information can be created using the landing positions of the first leg (125) collected while the robot (100) is walking.

[0086] Meanwhile, the landing locations of the second leg (127) following the first leg (125) can be determined based on terrain information according to the virtual map (1000) generated as above.

[0087] Specifically, in the case of (a) of FIG. 10, the first leg (125) and the second leg (127) of the robot (100) can walk with landing positions set according to a trajectory determined based on the flat ground.

[0088] When a ground as high as a height (h1) within a certain range exists as shown in (b) of FIG. 10 while the robot (100) is walking, a collision with the high ground occurs during the flat-ground touchdown phase of the first leg (125), and accordingly, the landing position of the first leg (125) may be changed to a second touchdown position (B) on the high ground.

[0089] In this case, the location and height (h1) of the ground are reflected in the 3D virtual map (1000) according to the changed landing position of the first leg (125), and the landing position of the second leg (127) following can be changed and set based on the virtual map (1000) having the terrain information.

[0090] Referring to (c) of Fig. 10, the landing position of the second leg (127) is changed to the fourth touchdown position (D) on the high ground according to the terrain information of the virtual map (1000), and the second leg (127) can swing with the fourth touchdown position (D) as the landing position.

[0091] In the case of (a) of Fig. 11, the first leg (125) and the second leg (127) of the robot (100) can walk with landing positions set according to a trajectory determined based on flat ground.

[0092] When a ground that is as low as a depth (h2) within a certain range exists as shown in (b) of FIG. 11 while the robot (100) is walking, a collision with the low ground occurs after passing the first touchdown position (A) set based on flat ground during the touchdown phase of the first leg (125), and accordingly, the landing position of the first leg (125) may be changed to the second touchdown position (B) on the low ground.

[0093] In this case, the location and depth (h2) of the ground are reflected in the 3D virtual map (1000) according to the changed landing position of the first leg (125), and the landing position of the second leg (127) following can be changed and set based on the virtual map (1000) having the terrain information.

[0094] Referring to (c) of Fig. 11, the landing position of the second leg (127) is changed to the fourth touchdown position (D) on the low ground according to the terrain information of the virtual map (1000), and the second leg (127) can swing with the fourth touchdown position (D) as the landing position.

[0095] Meanwhile, the blind walking control method of a multi-legged walking robot according to the present invention can be applied not only to the above-described stair-shaped terrain but also to the slope-shaped terrain.

[0096] In the case of (a) of Fig. 12, the first leg (125) and the second leg (127) of the robot (100) can walk with landing positions set according to a trajectory determined based on flat ground.

[0097] When there is an inclined surface that rises by an angle (θ1) within a certain range as shown in (b) of FIG. 12 while the robot (100) is walking, a collision with the inclined surface occurs during the flat-ground touchdown phase of the first leg (125), and accordingly, the landing position of the first leg (125) may be changed to a second touchdown position (B) on the inclined surface.

[0098] Here, the stair-shaped terrain and the slope-shaped terrain described with reference to FIGS. 10 and 11 can be distinguished from each other depending on the dynamic state of the robot measured in the touchdown phase of the first leg (125), such as the inertia of the robot, the angle of the joint, or the direction of the force.

[0099] In this case, the location and slope angle (θ1) of the ground are reflected in the 3D virtual map (1000) according to the changed landing position of the first leg (125), and the landing position of the second leg (127) following behind can be changed and set based on the virtual map (1000) having the terrain information.

[0100] Referring to (c) of Fig. 12, the landing position of the second leg (127) is changed to the fourth touchdown position (D) on the sloped ground according to the terrain information of the virtual map (1000), and the second leg (127) can swing with the fourth touchdown position (D) as the landing position.

[0101] In the case of (a) of Fig. 13, the first leg (125) and the second leg (127) of the robot (100) can walk with landing positions set according to a trajectory determined based on flat ground.

[0102] When there is a slope that descends by an angle (θ2) within a certain range as shown in (b) of FIG. 13 during the walking of the robot (100), a collision with a low, sloping ground occurs after passing the first touchdown position (A) set as a flat ground base during the touchdown phase of the first leg (125), and accordingly, the landing position of the first leg (125) may be changed to the second touchdown position (B) on the descending, sloping ground.

[0103] In this case, the location and inclination angle (θ2) of the ground are reflected in the 3D virtual map (1000) according to the changed landing position of the first leg (125), and the landing position of the second leg (127) following behind can be changed and set based on the virtual map (1000) having the terrain information.

[0104] Referring to (c) of Fig. 13, the landing position of the second leg (127) is changed to the fourth touchdown position (D) on the downwardly inclined ground according to the terrain information of the virtual map (1000), and the second leg (127) can swing with the fourth touchdown position (D) as the landing position.

[0105] Meanwhile, terrain information (e.g., virtual map) configured as described above can be stored in memory and used to control the walking trajectory for blind walking of the robot (100).

[0106] Specifically, upon return of the robot (100), a trajectory for the swing motion of the preceding first leg (125) can also be determined based on terrain information (e.g., a virtual map) stored in the memory, and trajectories for the swing motion of the remaining legs following can be determined as described above.

[0107] In the above, referring to FIGS. 10 to 13, the blind walking control method of the multi-legged walking robot according to the present invention has been described for a stepped terrain with a certain range of height difference and a sloped terrain with a certain range of angles. However, the present invention is not limited thereto, and can be applied to blind walking in various rough terrains such as a protruding or sunken terrain, a bushy terrain, a sandy terrain, or a snow-covered terrain.

[0108]

[0109] Meanwhile, the blind walking control method as described above can also be applied to a case where a robot (100) is equipped with a vision system such as a camera and a lidar sensor and walks using vision sensor information.

[0110] According to another embodiment of the present invention, a first walking mode that recognizes terrain based on vision sensor information and determines a walking trajectory, and a second walking mode that determines a walking trajectory based on flat ground without using vision sensor information, are mutually switched according to changes in terrain, thereby enabling the walking of a multi-legged walking robot using a vision system to be implemented more stably.

[0111] For example, the first walking mode may be a mode for walking using vision sensor information as a default walking mode of the robot (100), and the second walking mode may be a mode for performing blind walking as described with reference to FIGS. 3 to 13, but the present invention is not limited thereto.

[0112] FIG. 14 is a flowchart illustrating a method for switching walking modes of a multi-legged walking robot according to an embodiment of the present invention. Among the illustrated methods, descriptions of those identical to those described with reference to FIGS. 1 to 13 will be omitted.

[0113] Referring to FIG. 14, first, the robot (100) operates in a first walking mode in which it recognizes terrain based on vision sensor information and determines a walking trajectory, and swings its first leg toward a first touchdown position according to the trajectory determined based on the vision sensor information (step S1400).

[0114] In the first walking mode, the surrounding terrain is recognized using vision sensor information acquired through a vision system such as a camera and lidar sensor equipped in the robot (100), and the touchdown positions of each leg are determined based on the recognized surrounding terrain information and the moving direction of the robot, so that walking can be performed.

[0115] Since the walking method of a multi-legged walking robot using vision sensor information can be applied to various conventional walking algorithms, a detailed description thereof will be omitted.

[0116] While the robot (100) is operating in the first walking mode, if a change in terrain that is different from the terrain recognized based on vision sensor information is detected during the swing motion of the first leg (step S1410), the robot switches to the second walking mode in which the walking trajectory is determined based on flat ground without using the vision sensor information (step S1420).

[0117] At step S1410, the change in terrain may include at least one of a change in height and a change in slope, for example, a case where the height or slope of the terrain recognized based on vision sensor information is different from the height or slope of the actual terrain.

[0118] Specifically, when walking in the first walking mode, if the position at which the first leg touches the ground during the touchdown phase is different from the first touchdown position determined based on vision sensor information, a change in terrain can be detected at step S1410.

[0119] For example, as illustrated in FIG. 15, when a robot (100) walks on a first ground (1500) according to terrain recognized based on vision sensor information in the first walking mode, a first touchdown position (A), which is a landing position of the first leg (125), can be determined on a second ground (1501) recognized based on vision sensor information.

[0120] However, the actual terrain may be a third ground (1510) with a certain depth difference (h3), which may occur when there is an object such as a bush that cannot support the legs of the robot (100) or when there is an error in the vision system.

[0121] In this way, when the first touchdown position (A) of the first leg (125) set based on the vision sensor information in the first walking mode does not match the height of the actual ground (1510), a change in the terrain is detected, and the walking mode of the robot (100) can be switched from the first walking mode to the second walking mode.

[0122] Here, the second walking mode may be a blind walking mode according to the method described with reference to FIGS. 3 to 13. Accordingly, in the second walking mode, terrain information may be constructed using the actual landing position of the first leg, and a trajectory for the swing motion of the second leg following the first leg may be determined based on the constructed terrain information.

[0123] At step S1420, the robot (100) can control the swing motion of the first leg so that the first leg lands at a second touchdown position that is different from the first touchdown position set based on vision sensor information in the first walking mode.

[0124] For example, referring to FIG. 16, if the first leg (125) of the robot (100) does not land at the first touchdown position (A) set in the first walking mode, the walking mode of the robot (100) may be switched to the second walking mode.

[0125] At this time, the first leg (125) collides with the actual third ground (1510) lower than the first touchdown position (A), and the impact applied to the first leg (125) can be detected based on the dynamic state of the robot (100).

[0126] In the second walking mode, the first leg (125) lands on the second touchdown position (B) on the third ground (151), and terrain information including three-dimensional coordinate information of the second touchdown position (B) can be configured.

[0127] Thereafter, in the second walking mode, the touchdown position of the second leg following can be set based on terrain information constructed using the actual landing position of the first leg (125) (e.g., the second touchdown position (B)).

[0128] The method of constructing terrain information (e.g., a virtual map on 3D coordinates) using the actual landing position of the first leg (125) in the second walking mode and setting the touchdown position of the following leg based thereon may be as described with reference to FIGS. 3 to 13, and therefore a detailed description thereof will be omitted.

[0129] Referring to (a) of FIG. 17, while the robot (100) is walking on the first ground (1500) in the first walking mode, the third ground (1510) that is lower by a certain range of depth (h3) than the first ground (1500) is not recognized by the vision sensor information, and thus the first leg (125) may collide with the third ground (1510) and land at the second touchdown position (B).

[0130] In this case, the walking mode of the robot (100) is switched to the second walking mode, and the landing position of the second leg (127) following can be set to the third touchdown position (C) based on flat ground, as shown in (b) of FIG. 17.

[0131] After that, as shown in (c) of Fig. 17, the robot (100) walks on the third ground (1510) in the second walking mode, and the fourth touchdown position (D) of the first leg (125) is determined based on the flat ground, and a virtual map on three-dimensional coordinates representing terrain information can be generated based on the landing positions of the first leg (125).

[0132] Referring to (a) of FIG. 18, while the robot (100) is walking on the first ground (1500) in the first walking mode, the third ground (1510) that is lowered by a certain range of inclination (θ3) than the first ground (1500) is not recognized by the vision sensor information, and thus the first leg (125) may collide with the third ground (1510), which is a descending incline, and land at the second touchdown position (B).

[0133] In this case, the walking mode of the robot (100) is switched to the second walking mode, and the second touchdown position (B), which is the actual landing position of the first leg (125), is reflected on the virtual map in three-dimensional coordinates, and as shown in (b) of FIG. 18, the landing position of the second leg (127) following based on the virtual map can be set to the third touchdown position (C) on the third ground surface (1510), which is a downward slope.

[0134] After that, as shown in (c) of Fig. 18, the robot (100) can walk on the third ground (1510) in the second walking mode, so that the first leg (125) can land on the fourth touchdown position (D) on the third ground (1510).

[0135]

[0136] Meanwhile, when the terrain recognized based on the vision sensor information matches the actual terrain while operating in the second walking mode, the walking mode of the robot (100) can be returned to the first walking mode.

[0137] FIG. 19 is a flowchart illustrating a method for switching walking modes of a multi-legged walking robot according to another embodiment of the present invention. Among the illustrated methods, descriptions of those identical to those described with reference to FIGS. 1 to 18 will be omitted.

[0138] Referring to Fig. 19, the robot (100), while operating in the second walking mode, recognizes the surrounding terrain using vision sensor information and determines whether the terrain recognized based on the vision sensor information during the swing motion of the first leg matches the actual terrain (step S1900).

[0139] For example, when the robot (100) passes through terrain where there are objects that cannot support the legs, such as bushes, or when the vision system operates normally, terrain information recognized using the vision sensor information can match the actual terrain.

[0140] Specifically, when walking in the second walking mode, if the position at which the first leg touches the ground during the touchdown phase is the same as the touchdown position of the first leg according to the trajectory determined based on the vision sensor information, it can be determined that the terrain matches at step S1900.

[0141] As a result of the judgment, if the terrain matches, the walking mode of the robot (100) is switched from the second walking mode to the first walking mode (step S1910).

[0142] Referring to (a) and (b) of FIG. 20, when the robot (100) is walking on the ground (1510) in the second walking mode, if the ground (1501) recognized using the vision sensor information is identical to the terrain of the actual ground (1510), the robot (100) can return the walking mode to the first walking mode and set the landing position of the first leg to the fifth touchdown position (E) based on the vision sensor information.

[0143]

[0144] The blind walking control method of a multi-legged walking robot as described with reference to FIGS. 3 to 13 and the walking mode switching method of a multi-legged walking robot as described with reference to FIGS. 14 to 20 can be performed by a robot according to an embodiment of the present invention.

[0145] Meanwhile, the embodiments of the present invention have been described above by taking as an example the case where the robot (100) is a quadruped walking robot having four legs, but the present invention is not limited thereto, and can be applied to a humanoid robot having two legs as shown in (a) of FIG. 1 or a multi-legged walking robot having three or five or more legs.

[0146]

[0147] Hereinafter, with reference to FIG. 21, an exemplary computing device (500) in which the methods described in various embodiments of the present invention are implemented will be described. For example, the computing device (500) of FIG. 21 may be implemented as at least some components of a robot according to an embodiment of the present invention.

[0148] Referring to FIG. 21, a computing device (500) may include one or more processors (510), a bus (550), a communication interface (570), a memory (530) for loading a computer program (591) executed by the processor (510), and a storage (590) for storing the computer program (591). However, only components related to an embodiment of the present invention are illustrated in FIG. 21. Therefore, a person skilled in the art to which the present invention pertains will understand that other general components may be included in addition to the components illustrated in FIG. 21.

[0149] The processor (510) controls the overall operation of each component of the computing device (500). The processor (510) may be configured to include at least one of a CPU (Central Processing Unit), an MPU (Micro Processor Unit), an MCU (Micro Controller Unit), a GPU (Graphics Processing Unit), or any other type of processor well known in the art of the present invention. In addition, the processor (510) may perform operations for at least one application or program for executing methods / operations according to various embodiments of the present invention. The computing device (500) may include one or more processors.

[0150] The memory (530) stores various data, commands, and / or information. The memory (530) can load one or more programs (591) from the storage (590) to execute methods / operations according to various embodiments of the present invention. An example of the memory (530) may be, but is not limited to, RAM.

[0151] The bus (550) provides communication between components of the computing device (500). The bus (550) may be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

[0152] The communication interface (570) supports wired and wireless Internet communication of the computing device (500). The communication interface (570) may also support various communication methods other than Internet communication. To this end, the communication interface (570) may be configured to include a communication module well known in the technical field of the present invention.

[0153] Storage (590) can non-temporarily store one or more computer programs (591). Storage (590) can be configured to include a volatile memory such as a Read Only Memory (ROM), an Erasable Programmable ROM (EPROM), an Electrically Erasable Programmable ROM (EEPROM), a flash memory, a hard disk, a removable disk, or any form of computer-readable recording medium well known in the art to which the present invention pertains.

[0154] The computer program (591) may include one or more instructions implementing methods / operations according to embodiments of the present invention as described with reference to FIGS. 3 to 20.

[0155] For example, the computer program (591) may include instructions for executing an operation of swinging a first leg toward a first touchdown position according to a trajectory determined based on flat ground, an operation of controlling the swing operation of the first leg so that the first leg lands at the first touchdown position and the second touchdown position when a change in terrain is detected during the swing operation of the first leg, an operation of constructing terrain information including the second touchdown position, and an operation of determining a trajectory for the swing operation of the second leg following the first leg based on the constructed terrain information.

[0156] In addition, the computer program (591) may further include instructions for executing an operation of swinging the first leg toward the first touchdown position according to the trajectory determined based on the vision sensor information in a first walking mode that recognizes the terrain based on the vision sensor information and determines the walking trajectory, and an operation of switching to a second walking mode that determines the walking trajectory based on a flat surface without using the vision sensor information when a change in the terrain different from the terrain recognized based on the vision sensor information is detected during the swinging motion of the first leg.

[0157] When the computer program (591) is loaded into the memory (530), the processor (510) can perform methods / operations according to various embodiments of the present invention by executing one or more of the instructions.

[0158]

[0159] The methods according to the present invention described above can be produced as a program to be executed on a computer and stored in a computer-readable recording medium. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0160] Computer-readable recording media can be distributed across network-connected computer systems, allowing computer-readable code to be stored and executed in a distributed manner. Furthermore, functional programs, codes, and code segments for implementing the above method can be readily inferred by programmers skilled in the art to which the present invention pertains.

[0161] Although the preferred embodiments of the present invention have been illustrated and described above, the present invention is not limited to the specific embodiments described above, and various modifications can be made by those skilled in the art without departing from the gist of the present invention as claimed in the claims. Furthermore, such modifications should not be understood individually from the technical idea or prospect of the present invention.

Claims

1. A method for switching the walking mode of a multi-legged walking robot, In a first walking mode for recognizing terrain and determining a walking trajectory based on vision sensor information, a step for swinging the first leg toward a first touchdown position according to the trajectory determined based on the vision sensor information; and Including a step of switching to a second walking mode in which a walking trajectory is determined based on a flat surface without using the vision sensor information when a change in terrain different from the terrain recognized based on the vision sensor information is detected during the swing motion of the first leg; The second walking mode is A method for switching walking modes of a multi-legged walking robot, characterized in that terrain information is configured using the actual landing position of the first leg, and a trajectory for a swing motion of a second leg following the first leg is determined based on the configured terrain information.

2. In paragraph 1, the change in the terrain A method for switching walking modes of a multi-legged walking robot, characterized by including at least one of a change in height and a change in inclination.

3. In paragraph 1, A method for switching walking modes of a multi-legged walking robot, characterized in that when the position at which the first leg touches the ground during the touchdown phase while walking in the first walking mode is different from the first touchdown position, a change in the terrain is detected.

4. In the first paragraph, the step of switching to the second walking mode is A method for switching walking modes of a multi-legged walking robot, characterized by comprising: a step of controlling a swing motion of the first leg so that the first leg lands at a second touchdown position different from the first touchdown position.

5. In paragraph 4, A step of configuring terrain information including the second touchdown location; and A method for switching walking modes of a multi-legged walking robot, characterized in that it further includes a step of setting a touchdown position of the second leg based on the configured terrain information.

6. In paragraph 4, the terrain information is A method for switching walking modes of a multi-legged walking robot, characterized in that it includes three-dimensional coordinate information of the second touchdown position.

7. In paragraph 1, A method for switching walking modes of a multi-legged walking robot, characterized in that it further includes a step of switching the walking mode of the robot from the second walking mode to the first walking mode when it is determined that the terrain recognized based on the vision sensor information during the swing motion of the first leg matches the actual terrain.

8. In paragraph 7, A method for switching walking modes of a multi-legged walking robot, characterized in that when the position at which the first leg touches the ground during the touchdown phase while walking in the second walking mode is identical to the touchdown position of the first leg according to the trajectory determined based on the vision sensor information, the terrain is determined to match.

9. A computer program stored on a computer-readable recording medium for executing any one of the methods of claims 1 to 8 in combination with hardware.

10. A multi-legged walking robot performing any one of the methods of clauses 1 to 8.

11. Processor; A memory that loads a computer program to be executed by said processor; and Including storage for storing the above computer program, The above computer program, In a first walking mode in which a walking trajectory is determined by recognizing the terrain based on vision sensor information, an action of swinging the first leg toward the first touchdown position according to the trajectory determined based on the vision sensor information; and A multi-legged walking robot characterized by including instructions for executing an operation of switching to a second walking mode in which a walking trajectory is determined based on flat ground without using vision sensor information when a change in terrain different from the terrain recognized based on the vision sensor information is detected during the swing motion of the first leg.

Images