A foot-type robot

Abstract

The application discloses a kind of foot type robots, for enabling foot type robots to complete getting up with shorter action path, to improve the coherence of getting up action and getting up speed.The foot type robot provided by the application comprises a torso, two upper limbs, two lower limbs and corresponding limb ends connected with the torso respectively, at least one processor, and at least one storage medium, the at least one storage medium stores instructions, the instructions are executed by the at least one processor to enable the foot type robot: in response to getting up instruction, execute getting up action sequence, the getting up action sequence includes heterolateral hand-foot support stage and lower limb support stage, and the heterolateral hand-foot support stage is located before the lower limb support stage.

Description

Technical Field

[0001] This application relates to the field of robot control technology, and in particular to a legged robot. Background Technology

[0002] With the development of legged robot motion control technology, legged robots are gradually being applied to complex scenarios such as service, inspection, rescue, companionship, and industrial collaboration. In these applications, legged robots may fall due to external collisions, uneven ground, or other disturbances. To improve the autonomous operation capabilities of legged robots, they need to be able to autonomously get up and return to a standing position after a fall.

[0003] In existing standing-up solutions, robots typically need to gradually adjust their torso and limb postures according to pre-set action stages, switching support points multiple times during the process to gradually transition to a sitting, kneeling, semi-standing, or other intermediate postures before finally reaching a standing position. While these solutions achieve the standing-up function, the numerous switching of support points and the strong stage-dependent relationship between torso lifting, limb adjustment, and center of gravity transfer result in a long motion path, making the robot prone to pauses and leading to low overall standing-up efficiency. Summary of the Invention

[0004] This application provides a legged robot that can stand up with a shorter motion path, thereby improving the continuity and speed of the standing up action.

[0005] This application provides a legged robot, comprising:

[0006] The torso, including the hips; The head, two upper limbs and their corresponding ends, and two lower limbs and their corresponding ends are respectively connected to the torso, wherein the lower limbs are connected to the hip. At least one processor; and, At least one storage medium storing instructions that, when executed by the at least one processor, cause the legged robot to: In response to a standing command, a standing action sequence is executed, the standing action sequence including a contralateral hand and foot support phase and a lower limb support phase, the contralateral hand and foot support phase being preceding the lower limb support phase; The contralateral limb support phase includes: controlling the legged robot to enter an initial support posture, in which the ratio of the minimum distance between the head and the contact surface to the length of the torso is 0.5-1; the distal end of the first upper limb and the distal end of the first lower limb respectively establish support contact with the contact surface; the first upper limb and the first lower limb are located on different sides of the torso; the projection of the center of mass of the legged robot along the direction of gravity onto the contact surface falls within a first support polygon; wherein the first support polygon is formed by at least the distal end of the first upper limb and the distal end of the first lower limb together on the contact surface; the elbow joint angle of the first upper limb is 45° to 120°; and the knee joint angle of the first lower limb is 15° to 100°; after the legged robot enters the initial support posture, controlling the distal end of the first upper limb to output a thrust towards the contact surface, so that the legged robot enters the lower limb support phase; The lower limb support phase includes: controlling the upper limb end of the first upper limb to separate from the contact surface and move away from the contact surface, the centroid projection point continuously moving into the second support polygon, the second support polygon being formed by the lower limb end of at least one of the two lower limbs on the contact surface; The moment when the end of the first upper limb detaches from the contact surface is recorded as the first moment. At the first moment, the distance between the hip and the contact surface is 0mm-300mm. Before the first moment, the elbow joint angle of the first upper limb gradually increases. After the first moment, the knee joint angle of the first lower limb gradually increases.

[0007] As can be seen from the above technical solutions, this application has the following advantages: In response to the stand-up command, the legged robot is first controlled to enter an initial support posture in which at least the distal ends of the first upper limb and the distal ends of the first lower limb participate in the support. The center of mass projection point is then positioned within the corresponding first support polygon, allowing for a stable support configuration with the participation of both the contralateral limbs before the robot stands up. Simultaneously, by limiting the degree of head elevation relative to the contact surface, the elbow angle of the first upper limb, and the knee angle of the first lower limb, the robot can maintain a low-lying stand-up preparation state in the initial support posture while also allowing for subsequent upper limb push-off and lower limb extension movements.

[0008] Based on this, the upper limb of the first upper limb is controlled to output a thrust to the contact surface and release the support contact with the contact surface. The reaction force of the contact surface acting on the upper limb of the first upper limb is used to drive the center of mass projection point to move continuously into the second support polygon formed by at least one lower limb end. Furthermore, when the first upper limb leaves the contact surface, the elbow joint angle gradually increases before the first upper limb end leaves the contact surface, and the knee joint angle of the first lower limb increases after the first upper limb end leaves the contact surface, so that the upper limb pushing action and the lower limb extension action are continuously connected.

[0009] Based on the above method, the robot can establish stable support before standing up by using the first upper limb and the first lower limb located on different sides of the torso, and use the support reaction force of the contact surface acting on the first upper limb to continuously transfer the center of mass projection point from the support range corresponding to the initial support posture to the support range of the lower limb, thereby reducing the number of times the support state needs to be adjusted during the process of entering the standing posture from the low posture. This allows the legged robot to complete the standing up with a shorter action path, thereby improving the continuity and speed of the standing up action. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A schematic diagram of the overall structure of a humanoid robot provided as an example of a legged robot for this application; Figure 2 A schematic diagram of one embodiment of the standing motion sequence of the legged robot provided in this application; Figure 3 A schematic diagram of a support state of the legged robot provided in this application, in its initial support posture; Figure 4 A schematic diagram of a support state of the legged robot provided in this application as it transitions from an initial supporting posture to a standing posture; Figure 5 A schematic diagram of the first supporting polygon provided in this application in one supporting state; Figure 6 A schematic diagram of the second supporting polygon provided in this application in one supporting state; Figure 7 This is a schematic diagram illustrating a sub-stage division of the lower limb support stage provided in this application. Detailed Implementation

[0012] This application provides a legged robot that can stand up with a shorter motion path, thereby improving the continuity and speed of the standing up action.

[0013] The legged robot provided in this application includes a torso, two upper limbs, two lower limbs, at least one processor, and at least one storage medium. The two upper limbs are each connected to the torso, and the two lower limbs are each connected to the torso. For ease of description, the two upper limbs may be referred to as the first upper limb and the second upper limb, and the two lower limbs may be referred to as the first lower limb and the second lower limb. It should be noted that the terms "first upper limb," "second upper limb," "first lower limb," and "second lower limb" are only used to distinguish different limb structures and do not constitute a limitation on specific left and right positions.

[0014] Each upper limb includes an upper limb endpiece for establishing supporting contact with the contact surface, and each lower limb includes a lower limb endpiece for establishing supporting contact with the contact surface. It should be noted that the contact surface can be the ground, a platform surface, or other load-bearing surface capable of providing support for the legged robot; the upper limb endpiece can be a hand, a robotic hand, or other end structure capable of establishing supporting contact with the contact surface, and the lower limb endpiece can be a foot, a foot actuator, or other end structure capable of establishing supporting contact with the contact surface; this application does not limit the specific type of endpiece.

[0015] It should be noted that, in the embodiments of this application, "contact" refers to the state in which the limbs, torso, or other body parts of the legged robot are in physical contact with the contact surface; "supporting contact" refers to the state in which the above-mentioned contact provides support to the legged robot, enabling the corresponding part to bear part of the load from the body. Specifically, when a part of the legged robot only makes slight contact, rubbing, or non-load-bearing contact with the contact surface, that part can be understood as having ordinary contact with the contact surface; when there is a contact between that part and the contact surface that can bear the weight of the body or participate in posture stability, that part can be understood as establishing supporting contact with the contact surface. Whether supporting contact is established can be determined by force sensors, tactile sensors, joint torque, motor current, contact pressure distribution, or posture changes.

[0016] In some embodiments, the legged robot can be a humanoid robot, which is a specific implementation of legged robots. For ease of description, the following description mainly uses a humanoid robot as an example. Please refer to [link / reference]. Figure 1 , Figure 1This illustration shows an overall structural diagram of a humanoid robot, an example of a legged robot in this application. The humanoid robot includes a torso 3 and two upper limbs 1 and two lower limbs 2 connected to the torso 3. In some embodiments, the torso 3 may further include an upper torso 31 and a hip 32, wherein the upper torso 31 connects the two upper limbs 1 and the head 4, and the hip 32 connects the two lower limbs 2 and serves as a reference connection structure for lower limb movement. In some embodiments, the upper torso 31 and the hip 32 may be rotatably connected. In other embodiments, the legged robot may also adopt other limb configurations, as long as it can establish supporting contact with the contact surface through the limb ends or limb joints and complete the standing action according to the supporting state switching method of the embodiments of this application, it can be applied to the technical solution of this application.

[0017] The aforementioned at least one storage medium stores instructions that, when executed by at least one processor, can control the joint motors of the legged robot to work together so that the legged robot responds to the stand-up command and executes a stand-up action sequence, which includes a contralateral arm and foot support phase and a lower limb support phase, with the contralateral arm and foot support phase preceding the lower limb support phase.

[0018] Please see Figure 2 , Figure 2 This is a schematic diagram of one embodiment of the standing action sequence provided in this application. The standing action sequence includes a contralateral hand and foot support stage 201 and a lower limb support stage 202. The contralateral hand and foot support stage 201 is located before the lower limb support stage 202. They will be described separately below.

[0019] Contralateral Hand and Foot Support Phase 201: Control the legged robot to enter the initial support posture. In the initial support posture, the ratio of the minimum distance between the head and the contact surface to the length of the torso is 0.5-1. The upper limb tip of the first upper limb and the lower limb tip of the first lower limb establish support contact with the contact surface respectively. The first upper limb and the first lower limb are located on different sides of the torso. The projection point of the legged robot's center of mass on the contact surface along the direction of gravity falls within the first support polygon. The first support polygon is formed by at least the upper limb tip of the first upper limb and the lower limb tip of the first lower limb on the contact surface. The elbow joint angle of the first upper limb is 45° to 120°, and the knee joint angle of the first lower limb is 15° to 100°. After the legged robot enters the initial support posture, control the upper limb tip of the first upper limb to output a thrust to the contact surface so that the legged robot enters the lower limb support phase 202.

[0020] In this embodiment, the stand-up command can be generated by the legged robot's own task planning module, or it can be issued by an external control terminal, voice interaction module, remote control device, or host computer. The stand-up command is used to control the legged robot to execute a sequence of stand-up actions. Before executing the sequence of stand-up actions, the legged robot can be in a lying posture, a side-lying posture, a diagonal posture, a semi-lying posture, or other low-position postures, which are not limited here. During the process of controlling the legged robot to enter the initial support posture, the processor can control the movement of at least some joints of the torso, the two upper limbs, and the two lower limbs, so that the torso of the legged robot gradually rises from a state close to the contact surface, and the upper limb tip of the first upper limb and the lower limb tip of the first lower limb move to the contact surface to establish support contact, thereby forming the initial support posture. The above-mentioned torso lifting, the movement of the first upper limb tip, and the movement of the first lower limb tip can be executed sequentially according to a preset order, or at least partially synchronously, which can be determined according to the initial posture, joint range of motion, and current support state of the legged robot, and this application does not impose any restrictions on this.

[0021] In some embodiments, the processor can determine whether the legged robot has reached the initial support posture based on the minimum distance between the head and the contact surface, the length of the torso, the contact state between the distal end of the first upper limb and the contact surface, the contact state between the distal end of the first lower limb and the contact surface, the elbow angle of the first upper limb, the knee angle of the first lower limb, and the position of the centroid projection point relative to the first support polygon. Here, "length of the torso" can be understood as the length of the torso along the vertical axis of the legged robot when it is standing, i.e., the total length of the upper torso and hips along the vertical axis.

[0022] When the ratio of the minimum distance between the head and the contact surface to the length of the torso falls between 0.5 and 1, and both the distal ends of the first upper limb and the distal ends of the first lower limb establish supporting contact with the contact surface, the included angle of the elbow joint of the first upper limb is between 45° and 120°, the included angle of the knee joint of the first lower limb is between 15° and 100°, and the centroid projection point falls within the first supporting polygon, it can be determined that the legged robot has reached the initial supporting posture. The following details the various contact states and posture determination parameters under the initial supporting posture: In the initial support posture, the ratio of the minimum distance between the head and the contact surface to the length of the torso is 0.5 to 1. Specifically, this ratio can be 0.5, 0.6, 0.7, 0.8, 0.9, or 1. By limiting the ratio between the minimum distance between the head and the contact surface and the length of the torso, the legged robot can maintain a posture in which the head and torso are raised, but the overall body remains in a low-lying, ready-to-stand position. This facilitates the subsequent transition to standing up using the first upper limb to push off the ground and the first lower limb to support the body.

[0023] In the initial support posture, the first upper limb and the first lower limb are located on opposite sides of the torso. These opposite sides of the torso can be determined with reference to the central axis, sagittal plane, or central plane extending along the length of the torso of the legged robot. For example, the first upper limb can be the right upper limb and the first lower limb can be the left lower limb; or, the first upper limb can be the left upper limb and the first lower limb can be the right lower limb. By positioning the first upper limb and the first lower limb on opposite sides of the torso, a contralateral hand-foot support configuration can be formed before standing up, improving stability in the initial support posture.

[0024] In the initial support posture, the elbow joint angle of the first upper limb is between 45° and 120°. This elbow joint angle is the internal angle between the upper arm link and the forearm link of the first upper limb; it increases when the elbow is extended and decreases when the elbow is flexed. Specifically, the elbow joint angle of the first upper limb can be 45°, 60°, 75°, 90°, 105°, or 120°. This angle range allows the first upper limb to maintain a certain degree of flexion in the initial support posture while retaining a margin for further extension, enabling the first upper limb to output thrust to the contact surface by increasing the elbow joint angle before entering the lower limb support phase.

[0025] In the initial support posture, the knee joint angle of the first lower limb is between 15° and 100°. The knee joint angle is the internal angle between the thigh link and the lower leg link of the first lower limb. The knee joint angle increases when the knee is extended and decreases when the knee is flexed. The knee joint angle of the first lower limb is between 15° and 100°, specifically 15°, 20°, 30°, 45°, 60°, 75°, 90°, or 100°. In this way, the first lower limb is in a flexed state, while retaining the margin for subsequent extension, which is beneficial for maintaining stability in the initial support posture and can reduce the difficulty of transitioning to a standing posture.

[0026] In the initial supported posture, the projection of the legged robot's center of mass onto the contact surface along the direction of gravity falls within the first supporting polygon. The center of mass can be calculated based on the mass of each link, the angle of each joint, and the robot's dynamics model. The first supporting polygon represents the contact range within which the legged robot can currently form a stable support. The projection of the center of mass onto the contact surface along the direction of gravity, located inside the first supporting polygon, helps maintain overall stability during the standing-up process.

[0027] Next, the meaning of the support polygon will be explained in detail. In one interpretation, the support polygon is determined by all the support areas of the legged robot and the contact surface. The smallest convex hull closed polygon that can enclose all support points is taken to obtain the support polygon corresponding to the current support state.

[0028] In another interpretation, at least three support points can be selected within the support area of ​​the legged robot's contact surface, and a support polygon can be constructed based on these selected support points. In this case, since each support area can contain multiple candidate support points, multiple candidate support polygons can be constructed based on different combinations of support points. In this interpretation, as long as the centroid projection point falls within any candidate support polygon constructed based on the selected support points, it can be understood that the centroid projection point falls within the support polygon corresponding to the current support state. For example, at least one support point can be selected within the support area corresponding to the upper limb end of the first upper limb, and at least two support points can be selected within the support area corresponding to the lower limb end of the first lower limb, with the selected support points forming the first support polygon; alternatively, at least two support points can be selected within the support area corresponding to the upper limb end of the first upper limb, and at least one support point can be selected within the support area corresponding to the lower limb end of the first lower limb, with the selected support points forming the first support polygon.

[0029] Understandably, in the initial support posture, the first support polygon is determined by at least the support regions corresponding to the distal ends of the first upper limb and the distal ends of the first lower limb. For example, the first support polygon can be determined by the support regions corresponding to the distal ends of the first upper limb and the distal ends of the first lower limb; it can also be determined by the support regions corresponding to the distal ends of the first upper limb, the distal ends of the first lower limb, and the support region corresponding to the hip; it can also be determined by the support regions corresponding to the distal ends of the first upper limb, the distal ends of the first lower limb, and the distal ends of the second lower limb; or it can be determined by the support regions corresponding to the distal ends of the first upper limb, the distal ends of the first lower limb, the distal ends of the second lower limb, and the support region corresponding to the hip.

[0030] After the legged robot enters the initial support posture, the processor controls the distal end of the first upper limb to output a thrust towards the contact surface. Simultaneously, the contact surface applies a reaction force to the distal end of the first upper limb. This reaction force can include a component along the normal direction of the contact surface and a component along the tangential direction of the contact surface. The component along the normal direction of the contact surface drives the legged robot's center of mass to rise, while the component along the tangential direction drives the projection point of the legged robot's center of mass to move towards the lower limb support area. This allows the legged robot to obtain the necessary center of mass shift during the upper limb push-off process to transition from the initial support posture to the standing posture, thereby enabling the legged robot to enter the lower limb support phase.

[0031] In some specific implementations, in the initial support posture, the distal end of the second upper limb of the two upper limbs is separated from the contact surface, the projection of the distal end of the second upper limb onto the contact surface along the direction of gravity does not fall into the first support polygon, the elbow joint angle of the first upper limb is 60° to 100°, and the knee joint angle of the first lower limb is 20° to 75°.

[0032] The separation of the distal end of the second upper limb from the contact surface indicates that, in the initial supporting posture, the second upper limb does not participate in constructing the first supporting polygon as a supporting component. During the subsequent transition from the initial supporting posture to the standing posture, the processor does not need to control the second upper limb to undergo a switching process from supporting contact to releasing support contact, thereby reducing changes in the supporting contact state and lowering the difficulty of transitioning from the initial supporting posture to the standing posture. The projection of the distal end of the second upper limb onto the contact surface along the direction of gravity does not fall within the first supporting polygon, indicating that the second upper limb has a certain abduction or extension distance relative to the first supporting polygon. Since the second upper limb itself has mass, changes in its position will affect the overall center of mass position of the humanoid robot. The processor can adjust the posture of the second upper limb to produce an adjustment effect to compensate for the tendency of the torso to tilt laterally, forward, or backward. In other words, the cooperation of the second upper limb helps maintain the balance of the robot.

[0033] The elbow angle of the first upper limb is between 60° and 100°, specifically 60°, 70°, 80°, 90°, or 100°. When the elbow angle is less than 60°, the first upper limb is in a large flexion state, requiring a large extension stroke during the subsequent push-off, which easily increases the time and joint output requirements of the push-off action. When the elbow angle is greater than 100°, the first upper limb is close to a more extended state, and the remaining range of motion for further extension of the elbow joint to output thrust to the contact surface is relatively reduced. By setting the elbow angle of the first upper limb to between 60° and 100°, the first upper limb can have both support stability and subsequent extension range of the push-off in the initial support posture, which is beneficial for the first upper limb to output thrust to the contact surface by increasing the elbow angle before entering the lower limb support phase.

[0034] The knee joint angle of the first lower limb is between 20° and 75°, specifically 20°, 30°, 45°, 60°, or 75°. Within this range, the first lower limb can maintain flexion support while preserving space for subsequent extension, allowing the robot to raise its center of mass by extending the first lower limb during the transition from the initial supported posture to a standing posture. Based on this posture, the robot's main support area is formed by the distal ends of the first upper limb and the distal ends of the first lower limb. The second upper limb can be used for swinging balance or avoiding the lower limb, thus simplifying the support relationship in the initial supported posture.

[0035] Lower limb support phase 202: Control the distal end of the first upper limb to separate from the contact surface and move away from the contact surface. The centroid projection point continuously moves into the second support polygon, which is formed by the distal ends of at least one of the two lower limbs on the contact surface. The moment when the distal end of the first upper limb leaves the contact surface is recorded as the first moment. At the first moment, the distance between the hip and the contact surface is 0mm-300mm. Before the first moment, the elbow joint angle of the first upper limb gradually increases. After the first moment, the knee joint angle of the first lower limb gradually increases.

[0036] After the legged robot enters the initial support posture, the processor controls the robot to transition from the contralateral hand and foot support stage to the lower limb support stage. During this transition, the first upper limb outputs thrust towards the contact surface, enabling the legged robot to obtain the necessary center of mass shift required to transition from the initial support posture to the standing posture, and then enters the lower limb support stage 202. After entering the lower limb support stage 202, the processor controls the distal end of the first upper limb to separate from the contact surface and move away from it. The moment when the distal end of the first upper limb separates from the contact surface is recorded as the first moment.

[0037] Before the first moment, the processor controls the elbow joint angle of the first upper limb to gradually increase. The gradual increase in the elbow joint angle can also cause the upper limb end to output thrust towards the contact surface to assist the center of mass to rise and move to the lower limb support area.

[0038] The first moment occurs when the distal end of the first upper limb detaches from the contact surface. At the first moment, the processor controls the distal end of the first upper limb to release its supporting contact with the contact surface and continues to control the distal end of the first upper limb to move away from the contact surface.

[0039] At the initial moment, the distance between the hip and the contact surface ranges from 0mm to 300mm, and this distance can be the minimum distance along the direction of gravity. When the distance is 0mm, it indicates that the hip is still in contact with the contact surface; when the distance is greater than 0mm and less than or equal to 300mm, it indicates that the hip is no longer in contact with the contact surface but is still in a low-lying, off-the-ground state. By ensuring that the hip is in contact or low-lying, off-the-ground state at the initial moment, the legged robot can switch from upper limb support to lower limb support in a lower posture, reducing the need to significantly raise the hip before switching support, thus shortening the standing motion path.

[0040] In some specific implementations, at the first moment, the distance between the head and the knee joint of the second lower limb is less than the distance between the head and the knee joint of the first lower limb. In other words, the torso is tilted, making the head closer to the knee joint of the second lower limb. This helps to maintain the balance of the aircraft after the first upper limb leaves the contact surface and achieves a smooth transfer of the center of gravity.

[0041] The second supporting polygon is formed by the distal end of at least one of the two lower limbs on the contact surface. The definition of the second supporting polygon can be understood with reference to the first supporting polygon. In one implementation, the second supporting polygon is formed by the distal end of the first lower limb on the contact surface (i.e., single-leg support). The second supporting polygon can be constructed based on the outer contour of the support area of ​​the distal end of the first lower limb, or at least three support points can be selected within the support area corresponding to the distal end of the first lower limb, and the second supporting polygon can be formed by these at least three support points. In another implementation, the second supporting polygon is jointly formed by the distal ends of both lower limbs on the contact surface (i.e., double-leg support). The second supporting polygon can be constructed based on the support areas corresponding to the distal ends of both lower limbs. For example, the smallest convex hull closed polygon including all support points can be taken as the second supporting polygon, or multiple support points can be selected within the support areas of each foot, and the second supporting polygon can be constructed based on the selected support points.

[0042] Following the initial moment, the processor controls the knee joint angle of the first lower limb to gradually increase. This gradual increase in the knee joint angle indicates a change from a flexed to an extended state, allowing the first lower limb to continue driving the robot's center of gravity upwards and to support the weight of the robot body after the first upper limb is de-supported. As the knee joint angle of the first lower limb gradually increases, the height of the robot's torso gradually increases, the center of gravity shifts further upwards, and the projection point of the center of gravity continuously moves into the second support polygon, causing the robot to gradually transition from a state supported by the opposite hand and foot to a standing posture supported by the lower limbs.

[0043] Please see Figure 3 , Figure 3This diagram illustrates a support state of a legged robot in its initial support posture. In the initial support posture of the corresponding embodiment, the head of the legged robot is raised relative to the contact surface, ensuring that the ratio of the minimum distance between the head and the contact surface to the length of the torso meets a preset range. The distal end of the first upper limb establishes support contact with the contact surface, and the distal end of the first lower limb establishes support contact with the contact surface. The first upper limb and the first lower limb are located on different sides of the torso, thus forming a contralateral hand-foot support configuration. The hip establishes support contact with the contact surface. The heel area of ​​the distal end of the second lower limb may or may not establish support contact with the contact surface. The following will specifically explain the first support polygon using the example where the distal end of the second lower limb does not establish support contact with the contact surface, i.e., the distal end of the second lower limb does not output an effective support force to the contact surface to bear the weight of the robot body. In this case, the first support polygon is jointly determined by the support area corresponding to the distal end of the first upper limb, the support area corresponding to the distal end of the first lower limb, and the support area corresponding to the hip.

[0044] It should be noted that, Figure 3 Only one implementation of the initial support posture is shown. Support can be formed by the left upper limb and right lower limb on opposite sides, or by the right upper limb and left lower limb on opposite sides. The hip can establish support contact with the contact surface or it can be separated from the contact surface; the lower limb tip of the second lower limb can have already established support contact with the contact surface in the initial support posture, or it can establish support contact with the contact surface during the subsequent transition from the initial support posture to the standing posture. As long as the upper limb tip of the first upper limb and the lower limb tip of the first lower limb can establish support contact with the contact surface, and the centroid projection point falls within the corresponding first support polygon, it can be used as the initial support posture in this embodiment.

[0045] Please see Figure 4 , Figure 4 This diagram illustrates a support state for a legged robot transitioning from an initial supporting posture to a standing posture. In the corresponding embodiment, the legged robot can be based on... Figure 3 The initial support posture shown is used to perform the standing up motion. Figure 3 The state shown Figure 4 During the transition of the state shown, the processor controls and adjusts the position of the lower limb end of the second lower limb so that the foot area of ​​the lower limb end of the second lower limb establishes a support area with the contact surface, controls the hip to disengage from the contact surface, and controls the upper limb end of the first upper limb to output a thrust to the contact surface, so that the reaction force of the contact surface acting on the upper limb end of the first upper limb can drive the torso and hip to rise relative to the contact surface.

[0046] Furthermore, combined Figure 3 and Figure 4The construction methods of the first and second supporting polygons are explained in relation to a corresponding support state.

[0047] First supporting polygon: Taking the understanding that the supporting polygon is jointly determined by the legged robot and all supporting areas of the contact surface as an example, such as... Figure 5 As shown, the distal end of the first upper limb establishes supporting contact with the contact surface, forming a first upper limb supporting region 501; the distal end of the first lower limb establishes supporting contact with the contact surface, forming a first lower limb supporting region 502; and the hip establishes supporting contact with the contact surface, forming a hip supporting region 503. Convex hull calculations are performed on the outer boundaries of the hip supporting region 503, the first upper limb supporting region 501, and the first lower limb supporting region 502 on the contact surface. Taking the smallest convex hull closed polygon including all supporting points yields the first supporting polygon 500. At this point, the projection of the legged robot's center of mass onto the contact surface along the direction of gravity falls within the first supporting polygon, enabling the legged robot to assume its initial supporting posture.

[0048] Taking the understanding that a supporting polygon is constructed from supporting points selected within a supporting region as an example, at least one supporting point can be selected from the hip supporting region 503, the first upper limb supporting region 501, and the first lower limb supporting region 502, and a first supporting polygon can be constructed based on the selected at least three supporting points. For example, a first supporting point can be selected from the hip supporting region 503, a second supporting point from the first upper limb supporting region 501, and a third supporting point from the first lower limb supporting region 502, and the first, second, and third supporting points can be connected to form the first supporting polygon. Alternatively, two supporting points can be selected from the first lower limb supporting region 502, and at least one supporting point can be selected from at least one of the hip supporting region 503 and the first upper limb supporting region 501, and a first supporting polygon can be constructed based on the selected supporting points. Since the hip supporting region, the first upper limb supporting region, and the first lower limb supporting region can each include multiple supporting points, multiple supporting polygons can be formed based on different combinations of supporting points. When the centroid projection point is located within any candidate support polygon constructed from the selected support points, it can be understood that the centroid projection point falls within the first support polygon.

[0049] Second supporting polygon: Taking the understanding that the supporting polygon is jointly determined by the legged robot and all supporting areas of the contact surface as an example, such as... Figure 6As shown, during the transition of the legged robot from its initial supporting posture to its standing posture, the distal ends of both lower limbs can establish supporting contact with the contact surface. Specifically, the distal end of the first lower limb forms the first lower limb support region 601, and the distal end of the second lower limb forms the second lower limb support region 602. By calculating the convex hull of the projected outer boundaries of the first and second lower limb support regions 601 and 602 on the contact surface, and taking the smallest convex hull closed polygon including all support points, a second support polygon 600 formed by the distal ends of both lower limbs can be obtained. As the distal end of the first upper limb outputs a thrust towards the contact surface, the reaction force exerted by the contact surface on the distal end of the first upper limb drives the center of mass to rise and causes the projection point of the center of mass to move into the second support polygon 600. Once the projection point of the center of mass enters the second support polygon 600, the legged robot can transition to a standing state supported by both lower limbs.

[0050] When constructing the second supporting polygon using the support point selection method, at least three support points can be selected from the first lower limb support region 601 and the second lower limb support region 602. For example, two support points can be selected from the first lower limb support region 601 and at least one support point can be selected from the second lower limb support region 602, and the second supporting polygon can be constructed based on the selected support points; alternatively, at least one support point can be selected from the first lower limb support region 601 and two support points can be selected from the second lower limb support region 602, and the second supporting polygon can be constructed based on the selected support points. When the centroid projection point is located within any of the supporting polygons constructed from the selected support points, it can be understood that the centroid projection point falls within the second supporting polygon.

[0051] In this embodiment, in response to the stand-up command, the legged robot is first controlled to enter an initial support posture in which at least the distal ends of the first upper limb and the distal ends of the first lower limb participate in the support, and the centroid projection point falls within the corresponding first support polygon. This allows for the formation of a stable support configuration with the participation of both the opposite limbs before the robot stands up. Simultaneously, by limiting the degree of head elevation relative to the contact surface, the elbow angle of the first upper limb, and the knee angle of the first lower limb, the robot can maintain a low-lying stand-up preparation state in the initial support posture, while also allowing for subsequent upper limb push-off and lower limb extension movements.

[0052] Based on this, the upper limb of the first upper limb is controlled to output a thrust to the contact surface and release the support contact with the contact surface. The reaction force of the contact surface acting on the upper limb of the first upper limb is used to drive the center of mass projection point to move continuously into the second support polygon formed by at least one lower limb end. Furthermore, when the first upper limb leaves the contact surface, the elbow joint angle gradually increases before the first upper limb end leaves the contact surface, and the knee joint angle of the first lower limb increases after the first upper limb end leaves the contact surface, so that the upper limb pushing action and the lower limb extension action are continuously connected.

[0053] Based on the above method, the robot can establish stable support before standing up by using the first upper limb and the first lower limb located on different sides of the torso, and use the support reaction force of the contact surface acting on the first upper limb to continuously transfer the center of mass projection point from the support range corresponding to the initial support posture to the support range of the lower limb, thereby reducing the number of times the support state needs to be adjusted during the process of entering the standing posture from the low posture. This allows the legged robot to complete the standing up with a shorter action path, thereby improving the continuity and speed of the standing up action.

[0054] In some embodiments, the case of the hips landing in the initial support posture is further explained in detail: The torso includes hips that are connected to the two lower limbs respectively. In the initial support posture, the hips establish support contact with the contact surface. The first support polygon is formed by the hips, the upper limb end of the first upper limb, and the lower limb end of the first lower limb on the contact surface. At this time, there are at least three support areas between the legged robot and the contact surface. The first support polygon has a large area, which is beneficial to keep the center of mass projection point within the first support polygon and to provide stable initial conditions for the hips to detach from the contact surface and for the center of mass projection point to move to the lower limb support area.

[0055] The shortest distance between the hip and the end of the first lower limb is the first distance, and the shortest distance between the hip and the end of the first upper limb is the second distance. The ratio of the first distance to the second distance is 1-3.

[0056] In this embodiment, the first distance is the shortest distance between the support area corresponding to the hip and the support area corresponding to the lower limb end of the first lower limb; the second distance is the shortest distance between the support area corresponding to the hip and the support area corresponding to the upper limb end of the first upper limb. The ratio of the first distance to the second distance is from 1 to 3, specifically 1, 1.2, 1.5, 1.8, 2, 2.5, 2.8, or 3. For example, when the first distance is 750mm and the second distance is 300mm, the ratio of the first distance to the second distance is 2.5; when the first distance is 900mm and the second distance is 300mm, the ratio of the first distance to the second distance is 3. The above values ​​are only examples, and the specific values ​​of the first distance and the second distance can be determined according to the size of the legged robot, the length of the torso, the length of the upper limb, the length of the lower limb, and the actual support position under the initial support posture.

[0057] By setting the ratio of the first distance to the second distance to be between 1 and 3, a more balanced support span can be formed between the hip, the lower limb tip of the first lower limb, and the upper limb tip of the first upper limb. This ensures that the hip is positioned appropriately between the support areas of the first lower limb and the first upper limb in the initial support posture. When the ratio of the first distance to the second distance is less than 1, it indicates that the distance between the hip and the lower limb tip is less than the distance between the hip and the upper limb tip. In this case, the lower limb side support span may be too small, resulting in insufficient support lever arm provided by the first lower limb during subsequent extension. The available space when the center of mass projection point moves from near the hip to the lower limb support area is small, which can easily lead to the first upper limb having to bear a larger pushing force, increasing the risk of slippage or body tilting during the pushing process. When the ratio of the first distance to the second distance is greater than 3, it indicates that the distance between the hip and the end of the first lower limb is too large relative to the distance between the hip and the end of the first upper limb. In this case, the span between the support position of the first lower limb and the hip may be too large, which may cause the first lower limb to be in an excessively extended position. The displacement required for the centroid projection point to move from the first support polygon to the lower limb support area is large, which will increase the difficulty of control when transitioning from the initial support posture to the lower limb bearing state.

[0058] In some embodiments, in the initial support posture, the distance between the knee joint of the second lower limb and the contact surface is less than or equal to the distance between the knee joint of the first lower limb and the contact surface, and the knee joint angle of the second lower limb is greater than or equal to the knee joint angle of the first lower limb, wherein the knee joint angle of the second lower limb is 20°-150°.

[0059] It should be noted that the distance between the knee joint of the second lower limb and the contact surface can be considered as the minimum distance between the knee joint of the second lower limb and the contact surface along the direction of gravity; similarly, the distance between the knee joint of the first lower limb and the contact surface can be considered as the minimum distance between the knee joint of the first lower limb and the contact surface along the direction of gravity. When the contact surface is a horizontal surface, the above distances can be understood as the height of the corresponding knee joint relative to the horizontal surface.

[0060] Because the first upper limb and the first lower limb are located on opposite sides of the trunk, the second lower limb is usually located on the same side of the trunk as the first upper limb. In the initial support posture, the distal end of the first upper limb participates in the support, and the trunk tends to tilt towards the side where the first upper limb is located. When the hip establishes support contact with the contact surface, the space between the second lower limb, which is on the same side as the first upper limb, and the trunk is relatively limited. If the second lower limb is excessively flexed, it is easy for excessive folding to occur between the second lower limb and the trunk, increasing the risk of joint injury and increasing the difficulty of subsequent adjustment of the second lower limb posture and establishment of lower limb support contact.

[0061] In the initial support posture corresponding to this embodiment, the distance between the knee joint of the second lower limb and the contact surface is less than or equal to the distance between the knee joint of the first lower limb and the contact surface, and the knee joint angle of the second lower limb is greater than or equal to the knee joint angle of the first lower limb, indicating that the degree of flexion of the second lower limb is less than or equal to the degree of flexion of the first lower limb. Therefore, the possibility of excessive folding between the second lower limb and the torso can be reduced.

[0062] The knee angle of the second lower limb should be between 20° and 150°, specifically 20°, 30°, 45°, 60°, 75°, 90°, 120°, 135°, or 150°. When the knee angle is less than 20°, the second lower limb is in an over-flexed state, and there is a greater possibility of excessive folding between the second lower limb and the trunk, making it difficult to stand up. When the knee angle is greater than 150°, the second lower limb is close to an extended state, and the lower limb tip is difficult to retract to the target support area near the hip in a short time. Therefore, by setting the knee angle of the second lower limb to 20° to 150°, the possibility of excessive folding between the second lower limb and the trunk or hip can be reduced, and the lower limb tip can be adjusted to the target support area and establish support contact more quickly during the subsequent transition.

[0063] The following is a detailed explanation of the process of transitioning from an initial hip-ground support posture to a standing posture: In some embodiments, in the initial support posture, the foot area at the end of the second lower limb does not make contact with the contact surface, the distance between the knee joint of the second lower limb and the contact surface is less than the distance between the knee joint of the first lower limb and the contact surface, and the knee joint angle of the second lower limb is greater than the knee joint angle of the first lower limb, with the knee joint angle of the second lower limb being 90° to 150°.

[0064] The foot area at the end of the second lower limb does not make contact with the contact surface. For example, the heel area at the end of the second lower limb can make contact with the contact surface, or the end of the second lower limb can be separated from the contact surface. The distance between the knee joint of the second lower limb and the contact surface is less than the distance between the knee joint of the first lower limb and the contact surface. The knee joint angle of the second lower limb is greater than the knee joint angle of the first lower limb. The knee joint angle of the second lower limb is 90° to 150°. For example, this angle can be 90°, 100°, 110°, 120°, 130°, 135°, 140° or 150°. In this way, the second lower limb can maintain a relatively extended and unsupported state in the initial support posture, further reducing the possibility of excessive folding between the second lower limb and the torso or hip.

[0065] The lower limb support phase includes: Step A: Controlling the lower limb tip of the second lower limb to move towards the hip, and controlling the hip to disengage from the contact surface, so that the foot area of ​​the lower limb tip of the second lower limb establishes support contact with the contact surface; Step B: Controlling the upper limb tip of the first upper limb to output thrust towards the contact surface and then disengage from the contact surface, so that the support reaction force of the contact surface acting on the upper limb tip of the first upper limb is used to drive the center of mass of the legged robot to rise, and drive the center of mass projection point to move into the second support polygon formed by the lower limb tips of the two lower limbs.

[0066] It should be noted that although step B is described in form after step A, there is no necessary sequential relationship between step A and step B. Step A can be performed before step B, step A can be performed after step B, or step A and step B can be performed simultaneously.

[0067] In some embodiments, the foot region of the lower limb of the second lower limb establishes supporting contact with the contact surface, which may occur before the upper limb of the first upper limb detaches from the contact surface. (See [link to relevant documentation]). Figure 7 , Figure 7 This is a diagram showing the division of a sub-stage in the lower limb support stage. The lower limb support stage may include a single-hand and double-foot support sub-stage 701 and a double-foot support sub-stage 702. The single-hand and double-foot support sub-stage 701 precedes the double-foot support sub-stage 702. In the single-hand, two-foot support sub-stage 701, the lower limb is controlled to move from the initial position to the first position, and the hip is controlled to disengage from the contact surface. After the hip disengages from the contact surface, the lower limb is controlled to move from the first position to the second position, and the foot area of ​​the lower limb is controlled to establish supporting contact with the contact surface. The distance between the initial position, the first position, the second position, and the hip gradually decreases; that is, the second lower limb gradually moves closer to the hip and establishes supporting contact with the contact surface.

[0068] It should be noted that during the initial position, the first position, and the movement from the initial position to the first position, the distal end of the second lower limb may remain in contact with the contact surface. For example, the heel area of ​​the second lower limb may remain in contact with the contact surface, while the sole area may not establish contact with the contact surface. The distal end of the second lower limb may also detach from the contact surface and move from the initial position to the first position by swinging.

[0069] In this embodiment, before the hip detaches from the contact surface, the distal end of the second lower limb moves a certain distance from its initial position towards a first position closer to the hip, allowing the distal end of the second lower limb to approach the second position ahead of time. Since the hip can still participate in support together with the distal ends of the first upper limb and the distal ends of the first lower limb before detaching from the contact surface, the legged robot has a relatively stable support foundation in this sub-stage. Therefore, while maintaining the stability of the initial support posture, a portion of the position adjustment of the second lower limb can be completed in advance. After the hip detaches from the contact surface, the number of support positions of the legged robot decreases, and the support relationship is not yet stable. At this time, the distal end of the second lower limb only needs to continue moving a short distance to reach the second position and establish support contact between the foot area and the contact surface, thereby reducing the movement distance of the distal end of the second lower limb in the unstable state after the hip leaves the ground. Meanwhile, since the foot area at the end of the second lower limb establishes support contact after the hip leaves the contact surface, it avoids the second lower limb from folding into a foot support posture in advance when the hip is on the ground and the torso is tilted towards the side of the first upper limb. This helps to reduce the possibility of excessive folding between the second lower limb and the torso or hip, and reduces the risk of the second lower limb joint approaching the movement limit position.

[0070] After the foot region at the end of the second lower limb establishes supporting contact with the contact surface, the supporting parts of the legged robot change from the hip, the end of the first upper limb, and the end of the first lower limb to the end of the first upper limb, the end of the first lower limb, and the end of the second lower limb. At this time, the end of the first upper limb, the end of the first lower limb, and the end of the second lower limb together form a third supporting polygon on the contact surface, and the centroid projection point falls within the third supporting polygon. The understanding of the third supporting polygon is similar to that of the first and second supporting polygons, and will not be repeated here.

[0071] In the bipedal support sub-stage 702, the distal end of the first upper limb is controlled to output a thrust towards the contact surface, and the support contact between the distal end of the first upper limb and the contact surface is released. When the distal end of the first upper limb outputs a thrust towards the contact surface, the reaction force exerted by the contact surface on the distal end of the first upper limb can drive the center of mass of the legged robot to rise, and drive the projection point of the center of mass to move within the second support polygon formed by the distal ends of the two lower limbs on the contact surface. This second support polygon can be jointly determined by the first lower limb support area corresponding to the distal end of the first lower limb and the second lower limb support area corresponding to the distal end of the second lower limb.

[0072] In the bipedal support sub-stage 702, as the distal end of the first upper limb releases its support contact with the contact surface, the current support polygon is updated from the third support polygon to the second support polygon. Subsequently, the processor controls the knee joints of the first and second lower limbs to continue extending, further raising the center of mass of the legged robot and gradually transitioning it to a standing posture. Since the foot region at the distal end of the second lower limb has already established support contact before the first upper limb leaves the contact surface, after the first upper limb leaves the ground, the legged robot can have its weight supported by both lower limbs, reducing postural instability caused by the reduction in support points and lowering the risk of tipping over during support switching, thereby improving the stability of the transition from the single-arm bipedal support sub-stage 701 to the bipedal support sub-stage 702.

[0073] In some embodiments, the foot region of the second lower limb establishes supporting contact with the contact surface after the upper limb of the first upper limb detaches from the contact surface. In some embodiments, the establishment of supporting contact between the foot region of the second lower limb and the contact surface and the detachment of the upper limb of the first upper limb from the contact surface can occur simultaneously.

[0074] Because the second lower limb and the first upper limb are located on the same side of the torso, in the initial support posture, after the distal end of the first upper limb participates in the support, the torso tends to tilt towards the side where the first upper limb is located. Simultaneously, the hip remains in support contact with the contact surface, restricting the space for movement between the second lower limb and the torso and hip. If the foot area of ​​the second lower limb prematurely establishes support contact with the contact surface in this state, it increases the likelihood of excessive folding between the second lower limb and the torso or hip, increasing the risk of the joints of the second lower limb approaching their movement limits. By configuring the above embodiment so that the foot area of ​​the distal end of the second lower limb does not immediately establish contact with the contact surface in the initial support posture, a larger space for posture adjustment can be reserved for the second lower limb. The second lower limb can maintain contact through the heel area or other localized areas, or it can be completely separated from the contact surface. When transitioning from the initial supported posture to the standing posture, the distal end of the second lower limb first moves towards the hip, and after the hip disengages from the contact surface, the foot area establishes supporting contact with the contact surface. The distal end of the first upper limb outputs a force to the contact surface, driving the center of mass upward and causing the projection point of the center of mass to move into the second supporting polygon. Thus, the legged robot can reduce the risk of excessive folding between the second lower limb and the torso or hip, while also reducing the likelihood of the second lower limb joint approaching its movement limit position, and improving the smoothness of the transition from the initial supported posture with hip support to the standing posture.

[0075] The following section will elaborate on another process for transitioning from an initial hip-ground support posture to a standing posture: In some specific embodiments, the lower limb support phase includes: controlling the upper limb end of the first upper limb to output a thrust to the contact surface and releasing the support contact with the contact surface, the hip disengaging from the contact surface, and the support reaction force of the contact surface acting on the upper limb end of the first upper limb to drive the center of mass to rise and drive the center of mass projection point to move into the second support polygon formed by the lower limb end of the first lower limb.

[0076] In this embodiment, the hip is disengaged from the contact surface by pushing the ground directly with the first upper limb, and the center of mass projection point is sent into the second support polygon formed by the lower limb tip of the first lower limb. When the second support polygon is composed of a single support area, that support area can be understood as the corresponding second support polygon.

[0077] During the lower limb support phase, the distal end of the first upper limb can output thrust towards the contact surface within a short ground-pushing time window. The processor controls the elbow joint of the first upper limb to gradually extend, enabling the distal end of the first upper limb to output thrust that raises the center of gravity and moves the center of gravity projection point toward the second support polygon. When the distal end of the first upper limb outputs thrust towards the contact surface, the reaction force exerted by the contact surface on the distal end of the first upper limb lifts the hip off the contact surface and moves the center of gravity projection point from the area of ​​the first support polygon to the area of ​​the second support polygon. At this time, the distal end of the first lower limb gradually bears more of the fuselage weight. The first lower limb can use knee joint extension, hip joint adjustment, and / or ankle joint adjustment to move the center of gravity projection point into or toward the second support polygon. After the distal end of the first upper limb releases support contact with the contact surface, the knee joint angle of the first lower limb can be further increased, causing the first lower limb to change from a flexed support state to an extended support state, thereby continuing to drive the center of gravity upward. The second lower limb can establish supporting contact with the contact surface after the first lower limb has completed the support, so that the legged robot can gradually transition from a single lower limb support state to a bipedal standing posture.

[0078] With the above-described embodiments, the legged robot can directly lift its hips off the contact surface using the ground-pushing reaction force of its first upper limbs, from an initial support posture with the hips in contact, and directly project its center of mass into the second support polygon formed by the ends of its first lower limbs. This method reduces the requirement for the second lower limbs to establish support contact first, but places higher demands on the pushing power of the first upper limbs and the instantaneous load-bearing capacity of the first lower limbs. The shorter motion path allows the robot to transition from the initial support posture to the lower limb load-bearing state with a shorter support switching path, further improving its standing speed.

[0079] In some embodiments, the situation where the hips are off the ground in the initial support posture is further explained in detail: In the initial support posture, the hip separates from the contact surface. The upper limb tip of the first upper limb, the lower limb tip of the first lower limb, and the lower limb tip of the second lower limb together form the first support polygon on the contact surface. At this time, there are three support areas between the legged robot and the contact surface. The first support polygon has a large area, which is beneficial to keep the center of mass projection point within the first support polygon and to provide stable initial conditions for the subsequent hip to separate from the contact surface and for the center of mass projection point to move to the lower limb support area.

[0080] Among them, the shortest distance between the lower limb tip of the first lower limb and the lower limb tip of the second lower limb is the third distance, and the shortest distance between the lower limb tip of the first lower limb and the upper limb tip of the first upper limb is the fourth distance. The ratio of the fourth distance to the third distance is 1-3.

[0081] The understanding of the first supporting polygon is similar to that in the previous embodiment, and will not be repeated here. The third distance can be the shortest distance between the supporting area corresponding to the lower limb end of the first lower limb and the supporting area corresponding to the lower limb end of the second lower limb. The fourth distance can be the shortest distance between the supporting area corresponding to the lower limb end of the first lower limb and the supporting area corresponding to the upper limb end of the first upper limb.

[0082] The ratio of the fourth distance to the third distance is between 1 and 3, specifically 1, 1.2, 1.5, 1.8, 2, 2.5, 2.8, or 3. For example, when the fourth distance is 450mm and the third distance is 300mm, the ratio is 1.5; when the fourth distance is 750mm and the third distance is 300mm, the ratio is 2.5; and when the fourth distance is 900mm and the third distance is 300mm, the ratio is 3. These values ​​are merely examples; the specific values ​​for the third and fourth distances can be determined based on the legged robot's foot size, upper limb length, lower limb length, torso height, and current initial support posture.

[0083] By setting the ratio of the fourth distance to the third distance to 1 to 3, the upper limb tip of the first upper limb can form an appropriate hand and foot support span relative to the lower limb tips of the two lower limbs. On the one hand, the first upper limb can form an effective lever arm during the subsequent pushing process, so that the reaction force of the contact surface acting on the upper limb tip of the first upper limb assists the center of mass to rise and drives the center of mass projection point to move to the lower limb support area. On the other hand, the distance between the upper limb tip of the first upper limb and the lower limb support area is kept within a controllable range, which can reduce the possibility of excessive deflection of the torso posture during the pushing process, thereby improving the stability and continuity of the legged robot's transition from the initial support posture to the standing posture.

[0084] When the ratio of the fourth distance to the third distance is less than 1, it indicates that the distance between the tip of the first upper limb and the tip of the first lower limb is less than the distance between the tips of the two lower limbs. In this case, the hand-foot support span is too small, and the reaction force exerted by the contact surface on the tip of the first upper limb has a weak driving effect on the rise of the center of mass and the movement of the center of mass projection point towards the foot support area, increasing the difficulty of getting up. When the ratio of the fourth distance to the third distance is greater than 3, it indicates that the distance between the tip of the first upper limb and the tip of the first lower limb is too large relative to the distance between the tips of the two lower limbs. This makes it easy for the trunk to have a large tendency to lean forward, lean to the side, or twist during the push-off process, increasing the difficulty of posture control when transitioning from the initial support posture to the standing posture.

[0085] With the above-described embodiments, in the initial support posture where the hips are separated from the contact surface, the legged robot can form a first support polygon together from the tips of the first upper limb, the tips of the first lower limb, and the tips of the second lower limb, with the center of mass projection point falling within this first support polygon. In this state, the hips are detached from the contact surface, and the two lower limbs can participate in bearing loads in advance. The first upper limb can continue to assist in the upward movement and transfer of the center of mass through the ground-pushing reaction force during the subsequent transition process, thereby shortening the motion path from the initial support posture to the standing posture.

[0086] In other embodiments, in the initial support posture, the hips are separated from the contact surface, and the lower limb of the second lower limb has not yet established support contact with the contact surface. The first support polygon is jointly formed by the upper limb of the first upper limb and the lower limb of the first lower limb. Subsequently, by pushing the ground briefly with the first upper limb, the center of mass projection point enters the second support polygon formed by the lower limb of the first lower limb, and the second lower limb then lands and enters a standing posture.

[0087] In some specific embodiments, the torso also includes at least one sensor unit. When the instructions are executed by at least one processor, in response to a rising command while the robot is in a lying posture, the robot performs a sequence of rising actions. During the contralateral arm and leg support phase, the legged robot controls the output torque of the joint motors in the hip area, causing the torso to rotate relative to the lower limbs under the inertial constraint of the lower limbs, thereby driving the torso to rise. During the torso rising process, the robot acquires attitude tilt angle data and angular velocity data output by the sensor unit. Based on the attitude tilt angle data and angular velocity data, the robot adjusts the output of the joint motors in a closed loop to control the torso's rising height.

[0088] In this embodiment, the sensor unit may include at least one of an inertial measurement unit, a gyroscope, an accelerometer, and an attitude sensor, used to acquire attitude tilt angle data and angular velocity data of the legged robot during the standing-up process. During the contralateral hand-foot support phase, the output torque of the joint motors included in the hip joint can be controlled. Since at least part of the structure of the two lower limbs can contact the contact surface or maintain a large inertial constraint, after the hip joint motors output torque, the torso can rotate relative to the lower limbs, thereby gradually raising the torso from a state close to the contact surface. During the torso raising process, the processor acquires the attitude tilt angle data and angular velocity data output by the sensor unit, and determines the current raising state of the torso based on the attitude tilt angle data and angular velocity data. For example, the attitude tilt angle data can be used to determine the degree of tilt of the torso relative to the contact surface, and the angular velocity data can be used to determine the rotational speed during the torso raising process. The processor can perform closed-loop adjustment of the output torque of the hip joint motors according to the current attitude tilt angle, angular velocity, and target raising height of the torso. When the torso lifting speed is below the preset speed range, the output of the hip joint motor can be increased; when the torso lifting speed is above the preset speed range or the torso posture is close to the target posture, the output of the hip joint motor can be decreased. This allows the torso to be smoothly lifted to a height suitable for establishing the initial support posture.

[0089] In some specific embodiments, when the instructions are executed by at least one processor, when controlling the output of the pushing torque at the end of the first upper limb, the legged robot: calculates the desired pushing force that the end of the first upper limb needs to output based on the target position where the driving center of mass moves continuously; acquires actual feedback data collected by sensors in the first upper limb, including joint torque data and / or current data; calculates the actual pushing reaction force generated by the end of the first upper limb based on the actual feedback data; and, based on the deviation between the desired pushing force and the pushing reaction force, performs closed-loop adjustment and issues joint control instructions.

[0090] In this embodiment, during the process of controlling the distal end of the first upper limb to output a pushing force towards the contact surface, the processor can calculate the desired pushing force that the distal end of the first upper limb needs to output based on the target position where the center of mass moves continuously. The target position can be the position where the projection point of the center of mass enters the second supporting polygon, or it can be the intermediate position where the projection point of the center of mass moves towards the second supporting polygon. The processor can determine the desired pushing force that the distal end of the first upper limb needs to output based on the current center of mass position of the legged robot, the target position, the body mass, the current posture of the first upper limb, and a preset pushing time window. During the pushing process, the processor acquires actual feedback data collected by sensors in the first upper limb, including joint torque data and / or current data. For example, joint torque data can be acquired by torque sensors installed at the joints of the first upper limb, or the output torque of the corresponding joint can be calculated based on the current data of the joint motors. The processor can convert the joint torque data and / or the joint output torque calculated from the current data into the actual pushing reaction force generated by the distal end of the first upper limb based on the kinematic model of the first upper limb.

[0091] Furthermore, the processor can compare the desired pushing force with the pushing reaction force and adjust the joint control commands of the first upper limb in a closed loop based on the deviation between the two. For example, when the pushing reaction force is less than the desired pushing force, the processor can increase the output torque of at least one joint in the first upper limb; when the pushing reaction force is greater than the desired pushing force, the processor can decrease the output torque of the corresponding joint; when there is a deviation between the direction of the pushing reaction force and the direction of the desired pushing force, the processor can adjust the target angle or target angular velocity of the shoulder joint, elbow joint, and / or wrist joint so that the distal end of the first upper limb outputs a pushing force to the contact surface in the target direction.

[0092] In this way, the processor can determine the desired pushing force based on the target center of mass movement requirements during the first upper limb pushing the ground, and correct the pushing output by combining the actual feedback data of the first upper limb, so that the support reaction force of the contact surface acting on the upper limb end of the first upper limb can drive the center of mass to rise and drive the center of mass projection point to move into the second support polygon.

[0093] In some specific embodiments, the legged robot further includes at least one visual perception module. When the instructions are executed by at least one processor, the legged robot: before executing the standing action sequence, acquires visual data of the surrounding environment based on the visual perception module, and extracts information on the flatness of the contact surface and the distribution of obstacles in the left and right regions of the legged robot based on the visual data; performs a support stability assessment on the left and right regions based on the contact surface flatness information and the obstacle distribution information; and determines the first upper limb and the first lower limb based on the results of the support stability assessment.

[0094] In this embodiment, the visual perception module may include an RGB camera, a depth camera, a binocular camera, a structured light camera, a LiDAR, or other perception modules capable of acquiring three-dimensional environmental information, used to acquire visual data of the environment surrounding the legged robot. The visual data may include environmental images, depth images, environmental point cloud data, or an environmental map obtained by fusing the above data. Before executing the standing-up action sequence, the processor may acquire visual data of the surrounding environment based on the visual perception module and determine the environmental state of the left and right sides of the legged robot based on the visual data. The left and right sides may be contact surface areas located on the left and right sides of the legged robot's torso, respectively, and within the reach of the upper or lower limbs.

[0095] The processor can extract contact surface flatness and obstacle distribution information for the left and right regions based on visual data. Contact surface flatness can be determined based on the height difference of the contact surface point cloud, local slope, plane fitting error, or degree of unevenness. Obstacle distribution information can include the location, height, volume, distance from the expected support location, and whether it is located within the movement path of the upper or lower limb. Based on the contact surface flatness and obstacle distribution information, the processor performs support stability assessments for the left and right regions separately. For example, when a region has a flatter contact surface, fewer obstacles, and a larger area available for contact with the upper or lower limb, the corresponding support stability score can be higher; when a region has protrusions, pits, steps, loose objects, or obstacles interfering with the limb movement path, the corresponding support stability score can be lower. Subsequently, the processor determines the first upper limb and the first lower limb based on the support stability assessment results. For example, when the right side of a legged robot is more suitable for establishing support contact at the end of the upper limb, and the left side is more suitable for establishing support contact at the end of the lower limb, the right upper limb can be designated as the first upper limb, and the left lower limb as the first lower limb. Conversely, when the left side of a legged robot is more suitable for establishing support contact at the end of the upper limb, and the right side is more suitable for establishing support contact at the end of the lower limb, the left upper limb can be designated as the first upper limb, and the right lower limb as the first lower limb. This allows the first upper limb and first lower limb to be dynamically selected based on the surrounding environment before entering the initial support posture, improving the reliability of subsequent support contact establishment.

[0096] In some specific embodiments, the legged robot also includes at least one visual perception module. When the instructions are executed by at least one processor, the legged robot also: before executing the standing action sequence, calls the visual perception module to obtain environmental point cloud data of the current surrounding environment of the legged robot; if it is determined based on the environmental point cloud data that the current surrounding environment is not suitable for standing up, then generates a translation path in combination with the environmental point cloud data; after executing the translation path, controls the legged robot to execute the standing action sequence.

[0097] In this embodiment, the visual perception module may include an RGB camera, a depth camera, a binocular camera, a structured light camera, a LiDAR, or other perception modules capable of acquiring three-dimensional information about the environment. The processor can identify the available standing space, contact surface state, and obstacle distribution around the legged robot based on environmental point cloud data, and determine whether the current surrounding environment is suitable for standing up. In some implementations, situations where the current surrounding environment is unsuitable for standing up may include: obstacles above the legged robot's torso or along its lifting path, posing a collision risk to the torso, head, or upper limbs during lifting; protrusions, pits, steps, gaps, loose objects, or sharp objects in the intended support area of ​​the legged robot's first upper limb or first lower limb, making it difficult for the limb's end to establish stable support contact; the slope of the contact surface being greater than a preset slope threshold, or the local height difference of the contact surface being greater than a preset height difference threshold, making it difficult for the centroid projection point to stably fall into the support polygon; the effective area of ​​the contact surface available for support of the upper or lower limb end being less than a preset area threshold; walls, furniture, steps, equipment casings, or other obstacles around the legged robot, restricting the unfolding path, swinging path, or support landing point of the upper or lower limbs; and the support area that can be formed in the current posture of the legged robot cannot meet the support stability conditions required for the initial support posture.

[0098] If the processor determines, based on environmental point cloud data, that the current surrounding environment is unsuitable for getting up, it can generate a translation path using the environmental point cloud data. The translation path is the movement path of the legged robot in a fallen posture, used to move the legged robot from its current position to a more suitable position for getting up. A more suitable position for getting up is one with more overhead space, a flatter contact surface, less obstacle interference, and stable support contact that can be established at the ends of the upper and lower limbs. During the execution of the translation path, the processor can control the legged robot to move along the translation path in a fallen or low-lying posture using upper limb push-off, lower limb extension, trunk rolling, hip twisting, or a combination of these actions. After the movement is complete, the processor can again call the visual perception module to obtain updated environmental point cloud data, and when it confirms that the updated position meets the preset conditions for getting up, it controls the legged robot to execute a sequence of getting-up actions.

[0099] In this way, the legged robot can determine whether the current environment is safe for standing up before executing the standing action sequence. If there are obstacles above, uneven support areas, insufficient support area, or limited limb movement paths, the robot can first move to a more suitable position for standing up by translating its path before executing the standing action sequence, thereby reducing the risk of collisions, loss of support, or limb interference during the standing process.

[0100] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0102] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0103] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0104] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A legged robot, characterized in that, include: The torso, including the hips; The head, two upper limbs and their corresponding ends, and two lower limbs and their corresponding ends are respectively connected to the torso, wherein the lower limbs are connected to the hip. At least one processor; and, At least one storage medium storing instructions that, when executed by the at least one processor, cause the legged robot to: In response to a standing command, a standing action sequence is executed, the standing action sequence including a contralateral hand and foot support phase and a lower limb support phase, the contralateral hand and foot support phase being preceding the lower limb support phase; The contralateral hand and foot support phase includes: controlling the legged robot to enter an initial support posture, in which the ratio of the minimum distance between the head and the contact surface to the length of the torso is 0.5-1; the distal end of the first upper limb and the distal end of the first lower limb respectively establish support contact with the contact surface; the first upper limb and the first lower limb are located on different sides of the torso; the projection of the center of mass of the legged robot along the direction of gravity onto the contact surface falls within a first support polygon; wherein the first support polygon is formed by at least the distal end of the first upper limb and the distal end of the first lower limb together on the contact surface; the elbow joint angle of the first upper limb is 45° to 120°; and the knee joint angle of the first lower limb is 15° to 100°; after the legged robot enters the initial support posture, controlling the distal end of the first upper limb to output a thrust towards the contact surface, so that the legged robot enters the lower limb support phase; The lower limb support phase includes: controlling the upper limb end of the first upper limb to separate from the contact surface and move away from the contact surface, the centroid projection point continuously moving into the second support polygon, the second support polygon being formed by the lower limb end of at least one of the two lower limbs on the contact surface; The moment when the end of the first upper limb detaches from the contact surface is recorded as the first moment. At the first moment, the distance between the hip and the contact surface is 0mm-300mm. Before the first moment, the elbow joint angle of the first upper limb gradually increases. After the first moment, the knee joint angle of the first lower limb gradually increases.

2. The legged robot according to claim 1, characterized in that, In the initial support posture, the end of the second upper limb of the two upper limbs is separated from the contact surface, and the projection of the end of the second upper limb on the contact surface along the direction of gravity does not fall into the first support polygon. The elbow joint angle of the first upper limb is 60° to 100°, and the knee joint angle of the first lower limb is 20° to 75°. At the first moment, the distance between the head and the knee joint of the second lower limb is less than the distance between the head and the knee joint of the first lower limb.

3. The legged robot according to claim 1, characterized in that, In the initial support posture, the hip establishes support contact with the contact surface, and the first support polygon is formed by at least the hip, the upper limb end of the first upper limb, and the lower limb end of the first lower limb on the contact surface; The shortest distance between the hip and the end of the first lower limb is the first distance, and the shortest distance between the hip and the end of the first upper limb is the second distance. The ratio of the first distance to the second distance is 1-3.

4. The legged robot according to claim 3, characterized in that, In the initial support posture, the distance between the knee joint of the second lower limb and the contact surface is less than or equal to the distance between the knee joint of the first lower limb and the contact surface, and the knee joint angle of the second lower limb is greater than or equal to the knee joint angle of the first lower limb, with the knee joint angle of the second lower limb being 20°-150°.

5. The legged robot according to claim 4, characterized in that, In the initial support posture, the foot area at the end of the second lower limb does not make contact with the contact surface. The distance between the knee joint of the second lower limb and the contact surface is less than the distance between the knee joint of the first lower limb and the contact surface. The knee joint angle of the second lower limb is greater than the knee joint angle of the first lower limb. The knee joint angle of the second lower limb is 90°-150°.

6. The legged robot according to claim 5, characterized in that, The lower limb support phase includes: controlling the lower limb tip of the second lower limb to move towards the hip, and controlling the hip to disengage from the contact surface, with the foot area of ​​the lower limb tip of the second lower limb establishing support contact with the contact surface; controlling the upper limb tip of the first upper limb to output a thrust towards the contact surface and then disengage from the contact surface, with the support reaction force of the contact surface acting on the upper limb tip of the first upper limb used to drive the center of mass of the legged robot to rise, and driving the center of mass projection point to move into the second support polygon formed by the lower limb tips of the two lower limbs.

7. The legged robot according to claim 4, characterized in that, The lower limb support phase includes: The upper limb of the first upper limb is controlled to output a thrust to the contact surface and release the support contact with the contact surface. The hip part is disengaged from the contact surface. The support reaction force of the contact surface acting on the upper limb of the first upper limb is used to drive the center of mass to rise and drive the projection point of the center of mass to move into the second support polygon formed by the lower limb ends of the first lower limb.

8. The legged robot according to claim 1, characterized in that, In the initial support posture, the hip is separated from the contact surface, and the upper limb end of the first upper limb, the lower limb end of the first lower limb, and the lower limb end of the second of the two lower limbs together form the first support polygon on the contact surface; The shortest distance between the lower limb tip of the first lower limb and the lower limb tip of the second lower limb is the third distance, and the shortest distance between the lower limb tip of the first lower limb and the upper limb tip of the first upper limb is the fourth distance. The ratio of the fourth distance to the third distance is 1-3.

9. The legged robot according to claim 1, characterized in that, The torso also includes at least one sensor unit, and when the command is executed by the at least one processor, in response to the rise command while the robot is in a lying posture, the command also causes the legged robot to: During the contralateral hand and foot support phase, the joint motors in the hip area are controlled to output torque, causing the torso to rotate relative to the lower limbs under the inertial constraint of the lower limbs, thereby driving the torso to lift. During the process of torso lifting, the attitude tilt angle data and angular velocity data output by the sensor unit are acquired; The output of the joint motor is adjusted in a closed loop based on the attitude tilt angle data and angular velocity data to control the lifting height of the torso.

10. The legged robot according to claim 1, characterized in that, When the instruction is executed by the at least one processor, when the control of the first upper limb's distal end to output a thrust towards the contact surface causes the legged robot to: Based on the target position of the driving center of mass moving continuously, calculate the expected pushing force that the distal end of the first upper limb needs to output. Acquire actual feedback data collected by sensors in the first upper limb, the actual feedback data including joint torque data and / or current data; Calculate the actual ground-pushing reaction force generated at the distal end of the first upper limb based on the actual feedback data; Based on the deviation between the desired pushing force and the pushing reaction force, closed-loop adjustment is performed and joint control commands are issued.

11. The legged robot according to claim 1, characterized in that, The legged robot further includes at least one visual perception module, and the instructions, when executed by the at least one processor, cause the legged robot to: Before executing the standing action sequence, visual data of the surrounding environment is acquired based on the visual perception module, and the flatness of the contact surface and the distribution of obstacles in the left and right regions of the legged robot are extracted based on the visual data. Based on the contact surface flatness information and the obstacle distribution information, the support stability of the left and right regions is evaluated respectively. The first upper limb and the first lower limb are determined based on the results of the support stability assessment.

12. The legged robot according to claim 1, characterized in that, The legged robot further includes at least one visual perception module, and the instructions, when executed by the at least one processor, also cause the legged robot to: Before executing the standing action sequence, the visual perception module is invoked to obtain environmental point cloud data of the current surrounding environment of the legged robot. If it is determined based on the environmental point cloud data that the current surrounding environment is not suitable for getting up, then a translation path is generated in conjunction with the environmental point cloud data; After executing the translation path, the legged robot is controlled to execute the standing action sequence.

Images