A method for standing up control of a humanoid robot and a humanoid robot

Abstract

The application discloses a method for getting-up control of a humanoid robot and the humanoid robot, and is used for enabling the humanoid robot to realize fast, stable and autonomous getting-up with basic fault-tolerant capability. The method comprises the following steps: if a getting-up triggering instruction of the humanoid robot is received, a posture adjustment control instruction is called to control the humanoid robot to be adjusted to a preset posture, the preset posture comprises the following conditions: a hip part is in contact with the ground, an end effector of a first lower limb and / or an end effector of a second lower limb is in contact with the ground, a first included angle is alpha, 45°< alpha < 180°, a second included angle is beta, 45°< beta < 180°; a getting-up control program is called to control the humanoid robot to execute a getting-up action sequence, so that the humanoid robot is transitioned from the preset posture to a double-foot standing posture, and the getting-up action sequence comprises multi-stage getting-up actions, and the getting-up action of each stage is used for controlling the center of mass and the plurality of end effectors of the humanoid robot to move according to a respective preset getting-up reference trajectory by a corresponding getting-up control subprogram.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of humanoid robot control, and particularly relates to a method for controlling a humanoid robot to stand up and a humanoid robot. BACKGROUND

[0002] With the development of service humanoid robots, special operation humanoid robots and companion humanoid robots, humanoid robots gradually become the focus of research and application because their appearance and action mode are close to human beings. In a complex environment, a humanoid robot inevitably falls due to external force impact, uneven ground or control failure. In order to ensure that the humanoid robot can continue to complete the task without human intervention, the humanoid robot needs to have the ability to stand up autonomously after falling.

[0003] In the prior art, there are two types of solutions for the problem of the humanoid robot standing up after falling: one type is a standing-up method based on a fixed trajectory, which drives the joints of the humanoid robot through a pre-designed action sequence to make the humanoid robot recover from a falling state to a standing state. This type of method has strong action repeatability, but often depends on specific initial posture and ground conditions, and once the actual environment deviates from the preset conditions, the standing-up will fail. The other type is a standing-up method based on optimization control, which solves the dynamics equation or the trajectory optimization problem online to generate a standing-up control instruction according to the current state. This type of method can adapt to different initial postures and environmental conditions in theory, but has high computational complexity, is difficult to meet the real-time requirements of the humanoid robot control system, and is strongly dependent on the accuracy of the humanoid robot dynamics model and contact parameters.

[0004] Therefore, the method of the prior art is mostly open-loop or quasi-open-loop, and for the situation that the supporting hand / foot slips due to changes in ground friction, uneven local terrain and other factors during execution, it is extremely easy to cause the standing-up action to be accidentally interrupted or completely failed in the middle of the way, and it is impossible to realize the fast and stable autonomous standing-up of the humanoid robot. SUMMARY

[0005] The present application provides a method for controlling a humanoid robot to stand up and a humanoid robot, which is used to enable the humanoid robot to realize fast, stable and basic fault-tolerant autonomous standing-up.

[0006] The first aspect of the present application provides a method for standing up control of a humanoid robot, the humanoid robot comprising a trunk, a first upper limb and an end effector thereof, a second upper limb and an end effector thereof, a first lower limb and an end effector thereof, and a second lower limb and an end effector thereof, the first upper limb and the second upper limb being connected to the trunk respectively, the trunk comprising a crotch, the first lower limb and the second lower limb being connected to the crotch respectively, the first lower limb and the second lower limb each comprising a thigh and a shank, the thigh and the shank being connected through a knee joint, a first included angle being formed between the thigh and the shank of the first lower limb, and a second included angle being formed between the thigh and the shank of the second lower limb;

[0007] The method comprises:

[0008] If a standing up trigger instruction of the humanoid robot is received, a posture adjustment control instruction is called to control the humanoid robot to adjust to a preset posture, the preset posture comprising: the crotch being in contact with the ground, the end effector of the first lower limb and / or the end effector of the second lower limb being in contact with the ground, the first included angle being α, 45° < α < 180°, and the second included angle being β, 45° < β < 180°.

[0009] A standing up control program is called to control the humanoid robot to execute a standing up action sequence, so that the humanoid robot transitions from the preset posture to a double-foot standing posture, the standing up action sequence comprising a plurality of stages of standing up actions, each stage of standing up action controlling the center of mass and the plurality of end effectors of the humanoid robot to move according to a respective preset standing up reference trajectory by a corresponding standing up control subprogram.

[0010] The second aspect of the present application provides a humanoid robot, the humanoid robot comprising:

[0011] a processor, a memory, an input / output unit, and a bus;

[0012] The processor is connected to the memory, the input / output unit, and the bus;

[0013] The memory stores a program, and the processor calls the program to execute the method for standing up control of the humanoid robot of the first aspect and any optional embodiment of the first aspect.

[0014] A third aspect of this application provides a system for controlling the standing up of a humanoid robot. The humanoid robot includes a torso, a first upper limb and its end effector, a second upper limb and its end effector, a first lower limb and its end effector, and a second lower limb and its end effector. The first upper limb and the second upper limb are respectively connected to the torso. The torso includes a hip. The first lower limb and the second lower limb are respectively connected to the hip. Both the first lower limb and the second lower limb include a thigh and a calf. The thigh and the calf are connected by a knee joint. A first angle is formed between the thigh and the calf of the first lower limb, and a second angle is formed between the thigh and the calf of the second lower limb. The system includes:

[0015] The first calling unit is used to call the posture adjustment control command if it receives a humanoid robot standing up trigger command, and control the humanoid robot to adjust to a preset posture. The preset posture includes: the hip touching the ground, the end effector of the first lower limb and / or the end effector of the second lower limb touching the ground, the first included angle is α, 45° < α < 180°, and the second included angle is β, 45° < β < 180°.

[0016] The second calling unit is used to call the standing control program to control the humanoid robot to execute a standing action sequence, so that the humanoid robot transitions from the preset posture to a standing posture. The standing action sequence includes multiple stages of standing action. Each stage of standing action is controlled by a corresponding standing control subroutine to move the center of mass of the humanoid robot and multiple end effectors according to their respective preset standing reference trajectories.

[0017] As can be seen from the above technical solutions, this application has the following advantages:

[0018] First, the posture adjustment control command is invoked. In the first control phase, the humanoid robot is adjusted from an arbitrary fallen state to a uniform preset posture. This eliminates the need for subsequent standing control to handle an infinite number of initial states, greatly improving the method's versatility and generalization ability to multiple initial postures. Then, the standing control program is invoked. In the second control phase, a standing action sequence containing multiple stages is introduced. This staged standing control effectively suppresses subsequent action errors caused by slipping or misstepping of the supporting hand / foot, thus avoiding unexpected interruptions or complete failure of the standing process. This application, through its staged control framework, enables the humanoid robot to achieve rapid, stable, and fundamentally fault-tolerant autonomous standing under varying initial postures and uncertain contact conditions. Attached Figure Description

[0019] In order to more clearly illustrate the technical solutions in the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0020] Figure 1 An embodiment flowchart of a method for standing up control of a humanoid robot provided by the present application is shown in the following figure.

[0021] Figure 2 A schematic diagram of adjusting a humanoid robot from a fallen state to a preset posture provided by the present application is shown in the following figure.

[0022] Figure 3 A schematic diagram of a support state of a humanoid robot in a standing up process provided by the present application is shown in the following figure.

[0023] Figure 4 A schematic diagram of a double-foot standing posture of a humanoid robot provided by the present application is shown in the following figure.

[0024] Figure 5 A schematic diagram of phase division of a standing up action sequence provided by the present application is shown in the following figure.

[0025] Figure 6 Another embodiment flowchart of a method for standing up control of a humanoid robot provided by the present application is shown in the following figure.

[0026] Figure 7 An embodiment structure schematic diagram of a system for standing up control of a humanoid robot provided by the present application is shown in the following figure.

[0027] Figure 8 An embodiment hardware structure schematic diagram of a humanoid robot provided by the present application is shown in the following figure. DETAILED DESCRIPTION

[0028] The present application provides a method for standing up control of a humanoid robot and a humanoid robot, which are used to make the humanoid robot realize fast, stable and self-standing up with basic fault tolerance.

[0029] It should be noted that the method for standing up control of the humanoid robot provided in the present application is applied to a humanoid robot and is designed for a blind standing up scene of the humanoid robot, that is, the humanoid robot realizes autonomous standing up without relying on a visual sensor. The humanoid robot comprises a torso, a first upper limb and an end effector thereof, a second upper limb and an end effector thereof, a first lower limb and an end effector thereof, and a second lower limb and an end effector thereof, the first upper limb and the second upper limb are connected with the torso respectively, the torso comprises a crotch, the first lower limb and the second lower limb are connected with the crotch respectively, the first lower limb and the second lower limb each comprise a thigh and a shank, the thigh and the shank are connected through a knee joint, a first included angle is formed between the thigh and the shank of the first lower limb, and a second included angle is formed between the thigh and the shank of the second lower limb. Each end effector of the humanoid robot is equipped with a force sensor or a tactile sensor for detecting a contact state and a force condition with the ground in real time. The humanoid robot further comprises a posture perception module, the posture perception module is used for acquiring posture data and a center of mass position of the humanoid robot in real time through an inertial measurement unit (IMU), an encoder and a joint angle sensor. For the convenience of description, the embodiments of the present application take a control system of the humanoid robot, referred to as a system, as an execution subject for description, but this is not a limitation on the protection scope of the present application.

[0030] The mechanical structure and part of the hardware of the humanoid robot provided in the present application are described above, and the method for standing up control of the humanoid robot provided in the present application is described below. Please refer to Figure 1 Figure 1 An embodiment of the method for standing up control of the humanoid robot provided in the present application comprises the following steps.

[0031] 101. If a standing up trigger instruction of the humanoid robot is received, a posture adjustment control instruction is called to control the humanoid robot to adjust to a preset posture, the preset posture comprising: the crotch touching the ground, the end effector of the first lower limb and / or the end effector of the second lower limb touching the ground, the first included angle being a, 45° < a < 180°, and the second included angle being b, 45° < b < 180°;

[0032] When the humanoid robot falls down due to external force impact, terrain disturbance or motion imbalance during the execution of a task, the posture state will deviate from the normal walking or standing posture, and the humanoid robot may fall down in various unpredictable postures. These arbitrary initial states have high uncertainty. If the standing up control program is directly called from such an arbitrary initial falling posture, the input state space of the control strategy will be too large to converge, and thus the standing up will fail. Therefore, in the present embodiment, a posture adjustment control instruction is introduced to smoothly adjust each joint of the humanoid robot, so that the humanoid robot adjusts from an arbitrary falling posture to a preset posture. Through this step, it can be ensured that the subsequent execution of the standing up control program has consistent and stable initial conditions, greatly improving the standing up stability. ​

[0033] In the embodiment, the posture adjustment control instruction can be specifically implemented by a PD position controller. The PD controller calculates an output joint torque instruction JointTorque according to a deviation between a preset target joint position q_des and a real-time detected current joint position q, and a deviation between a target joint speed qd_des and a current joint speed qd. Specifically, JointTorque = Kp * (q_des - q) + Kd * (qd_des - qd). Wherein, Kp is a proportional gain, and Kd is a differential gain. In the parameter design of the PD position controller, a moderate Kp value and a moderate Kd value are preferably adopted, with the purpose of reducing impact and chattering in the posture adjustment process in priority, rather than pursuing the fastest response speed. The moderate gain ensures that the limbs of the humanoid robot move smoothly in the process of transitioning to the preset posture, avoiding secondary rigid collision with the ground or high-frequency chattering.

[0034] Please refer to Figure 2 , Figure 2 Fig. 1 is a schematic diagram of a humanoid robot adjusting from a prone state to a preset posture. After posture adjustment control, the upper body of the humanoid robot is gradually lifted from lying down, the lower limbs are bent, and the hip is in contact with the ground, thereby forming a stable initial support structure, i.e., adjusting to a preset posture. It should be noted that the preset posture specifically refers to the stable posture of the humanoid robot before performing a standing-up action. The preset posture requires the hip to be in contact with the ground to form a reliable initial support surface, and at least one end effector of the lower limb should be in contact with the ground. The humanoid robot should have at least three support points in this preset posture, i.e., the contact point of the hip with the ground and other contact points formed by the end effectors of the limbs. The three support points can be formed by the end effectors of the two lower limbs and the ground, or by the end effector of one lower limb and the end effector of at least one upper limb. The body of the humanoid robot can be in a sitting or standing posture, or in a prone posture, which is not limited here. In this preset posture, the angles between the thighs and the shanks of the lower limbs are α and β, respectively, which need to satisfy the constraint range of 45° < α < 180° and 45° < β < 180°, to ensure that each joint has sufficient space and torque output capacity during the standing-up process.

[0035] 102, call the standing-up control program to control the humanoid robot to perform a standing-up action sequence, so that the humanoid robot transitions from the preset posture to a standing posture with both feet on the ground. The standing-up action sequence includes multiple stages of standing-up actions, and each stage of standing-up action controls the center of mass and multiple end effectors of the humanoid robot to move according to a respective preset standing-up reference trajectory by a corresponding standing-up control subprogram.

[0036] After the humanoid robot adjusts to the preset posture, the system calls a standing-up control program to control the humanoid robot to perform a standing-up action sequence, so that the humanoid robot transitions from the preset posture to a double-foot standing posture. The standing-up action sequence includes multiple stages of standing-up actions, i.e., the standing-up control program includes multiple sub-programs, and each stage of standing-up action is controlled by a corresponding standing-up control sub-program to move the center of mass and multiple end effectors of the humanoid robot according to a respective preset standing-up reference trajectory. The purpose of the preset standing-up reference trajectory is to guide the center of mass of the humanoid robot to gradually move from a starting position near the ground to a target position under the double-foot standing posture, and the height of the target position is higher than that of the starting position, so as to realize the dynamic transition from the prone state to the upright state. In this process, the preset standing-up reference trajectories of the center of mass and the multiple end effectors are optimized to ensure that the center of mass rising path is smooth, the energy consumption is reasonable, and the multiple end effectors maintain sufficient support stability at each stage.

[0037] It should be noted that each stage in the standing-up action sequence has a clear time boundary, and each stage corresponds to an independent starting time point and an ending time point. The main basis for stage division is the change in the contact state of the end effectors of the humanoid robot, i.e., when it is detected that the contact mode of the hand or foot end effector with the ground changes relative to the previous stage, the system determines that the current stage ends and enters the next stage. In addition to the contact state, the center of mass position or the upper body posture can also be introduced as a supplementary basis for stage division in auxiliary judgment. In some complex scenarios, the humanoid robot may have ambiguous contact signals due to uneven terrain or end sensor delay, at which point the system can determine whether the stage transition has been completed based on the trajectory of the center of mass. If it is detected that the center of mass reaches a predetermined threshold in the vertical or horizontal direction (e.g., the rising height exceeds a certain proportion), it can be inferred that the current stage action has been completed, thereby triggering the execution of the next stage. In addition, the upper body posture angle (such as the pitch angle and roll angle) can also be further introduced, and the change in the posture of the upper body is measured by an inertial measurement unit (IMU). When it is detected that the rotation angle of the body reaches a preset range, for example, from the supine state to the state of bending or lifting the torso by more than a certain angle, the system can also determine the stage switching based on this.

[0038] Further, to enable the humanoid robot to adapt to different terrain conditions, the system can further introduce a smoothing compensation mechanism for the reference trajectory when executing the standing up control program, that is, by adding a smoothing compensation amount to the original center of mass reference trajectory and the whole body end effector trajectory to realize real-time adaptation to the terrain height and slope. For example, when detecting that the foot end effector is stepping on a platform higher than the surrounding ground, the system will superimpose a smooth height compensation amount on the original center of mass reference trajectory, for example, move the center of mass reference trajectory upward by about 10 cm as a whole, to ensure that the body center of gravity and the support surface maintain a reasonable geometric relationship in subsequent actions, and the reference trajectories of all end effectors of the humanoid robot will also be adjusted according to the same pose change rule to maintain posture coordination and balance. If the humanoid robot is on a non-horizontal terrain (such as a slope or stairs), the system will further superimpose a slope correction on the basis of height compensation. In this way, the humanoid robot can not only successfully complete the standing up action on flat ground, but also achieve smooth and natural standing up in the case of inclined ground or height difference.

[0039] In this embodiment, first, the posture adjustment control instruction is called, and in the first control stage, the humanoid robot is adjusted from an arbitrary fallen state to a unified preset posture, so that the subsequent standing up control does not need to deal with infinite initial states, greatly improving the universality and generalization ability to multiple initial postures of the method. Then the standing up control program is called, and in the second control stage, a standing up action sequence including multiple stages is introduced. This staged standing up control can effectively suppress the subsequent action errors caused by slipping or missing of the supporting hands / feet, thereby avoiding accidental interruption or complete failure of the standing up process. Through the staged control framework, the humanoid robot can achieve fast, stable and basic fault-tolerant self-standing up under variable initial postures and uncertain contact conditions.

[0040] Please refer to Figures 3-4 , Figure 3 for a standing up process of the humanoid robot, Figure 4 for a schematic diagram of a double-foot standing posture. During the entire standing up process, the humanoid robot gradually lifts the hip off the ground through the coordinated movement of the upper limbs and lower limbs, and finally makes the end effectors of the double lower limbs stably contact the ground, and the torso remains vertical, thereby realizing the posture transition from falling to complete standing.

[0041] Please refer to Figure 5 , Figure 5 for a schematic diagram of stage division of the standing up action sequence. In some specific embodiments, in order to realize continuous and stable standing up, the multiple stages of standing up actions at least include: a single-hand supporting stage, a single-hand single-foot supporting stage, a double-hand single-foot supporting stage, a double-hand double-foot supporting stage, and a double-foot supporting stage. The stages are described in detail as follows:

[0042] 1. Single hand support stage: This stage is to establish the initial support from the preset posture (i.e. the standing-up preparation posture). The system controls the first upper limb of the humanoid robot to move, so that its end effector establishes stable contact with the ground, thereby forming the first support point on the ground. In this stage, the system also synchronously controls the torso of the humanoid robot to turn sideways towards the first support point, with the purpose of actively shifting the center of gravity of the humanoid robot towards the first support point that has been established, so as to create space for the subsequent movement of the limbs on the non-supporting side and ensure posture stability.

[0043] 2. Single hand and single leg support stage: This stage is to increase the support of the lower limbs to raise the center of gravity on the basis of the first support point that has been established. While maintaining the stable support of the first upper limb (the first support point), the system controls the first lower limb (i.e. the lower limb on the same side as the first upper limb) of the humanoid robot to move. Specifically, the end effector of the first lower limb moves from an initial first position to a second position, which is closer to the crotch of the humanoid robot than the first position, i.e. performs a similar action of “leg folding”. When the end effector of the first lower limb establishes stable contact with the ground at the second position, the second support point is established. It should be noted that during the movement in this stage, in order to cooperate with the “leg folding” and raise the body, the knee joint of the first lower limb moves towards the direction away from the ground (i.e. the knee is bent), while the knee joint of the second lower limb on the other side moves towards the direction close to the ground. This set of coordinated movements makes the first angle between the thigh and the lower leg of the first lower limb smaller than the second angle between the thigh and the lower leg of the second lower limb (i.e. the first lower limb is bent more), thereby creating a favorable geometric configuration for the subsequent raising of the center of gravity and rotation of the body.

[0044] 3. Double hand and single leg support stage: This stage is to establish a three-point support and prepare for the landing of both feet. The system first controls the second upper limb (i.e. the upper limb on the non-supporting side) of the humanoid robot to move, so that its end effector contacts the ground, thereby establishing the third support point and forming a stable three-point support structure of double hands and single leg. On this basis, the system further controls the second lower limb (i.e. the leg on the non-supporting side) of the humanoid robot to move, so that its end effector moves from the current third position to a fourth position, which is also closer to the crotch of the humanoid robot than the third position. During the movement of the second lower limb, the angle between its thigh and lower leg gradually decreases (i.e. the leg is gradually bent and folded), and the purpose of this action is to cooperate with the upper body, so that the overall center of gravity of the humanoid robot is inclined forward, thereby preparing for the subsequent transfer of the center of gravity to both feet and the raising of the crotch.

[0045] 4. Both hands and both feet support stage: This stage is the key stage for the humanoid robot to transition from the sitting posture to the four-point support. The system performs two coordinated actions while maintaining the stability of the first, second and third support points that have been established: one is to control the crotch of the humanoid robot to move upward, so that it moves from the fifth position (lower height) to the sixth position, and the height of the sixth position from the ground is greater than that of the fifth position, that is, to perform an action similar to "lifting the hips". The second is to control the end effector (foot) of the second lower limb that has been folded in the previous stage to contact the ground, thereby establishing the fourth support point. After this stage is completed, the humanoid robot forms a stable two-handed and two-foot four-point support state.

[0046] 5. Both feet support stage: This stage is the final stage of the getting-up action, and the goal is to transition to an upright standing posture. On the basis of the "both hands and both feet" four-point support, the system controls the center of mass position of the humanoid robot to move from the current seventh position (lower squat center of gravity) to the eighth position (standing center of gravity), and the height of the eighth position from the ground is greater than that of the seventh position. During this process of lifting and moving forward of the center of mass, as the center of gravity gradually shifts and stabilizes within the two-foot support surface, the system will control the end effector of the first upper limb and the end effector of the second upper limb to separate from the ground. Finally, the humanoid robot completes the entire transition from falling to a two-foot standing posture, relying only on two-foot support.

[0047] By dividing the getting-up action sequence into single-hand support, single-hand and single-foot support, double-hand and single-foot support, double-hand and double-foot support, and double-foot support stages, the stability and controllability of the getting-up process can be significantly improved. By dividing the complex whole-body movement into a series of sub-actions with stable intermediate postures and controlling them through corresponding getting-up control sub-programs, it can be ensured that the humanoid robot always maintains a reliable support polygon during the transition from the preset posture to the final two-foot standing posture. This gradual action design achieves a smooth and gradual lifting of the center of mass, for example, the system uses the support points established by the upper limbs to assist in lifting the crotch, rather than relying solely on the lower limbs or the waist and abdomen for high-torque, high-risk explosive actions. This design avoids sudden changes in posture and instability caused by rapid movement of the center of mass, ensuring that the humanoid robot can safely transfer the center of mass from near the ground to within the two-foot support surface in a predictable, low-impact and efficient manner, thereby greatly improving the robustness and success rate of the getting-up action.

[0048] In some embodiments, a corresponding contact state reference sequence can be defined for the getting-up action sequence. The contact state reference sequence is used to define the timing of the contact or separation of the end effectors such as the hands and feet of the humanoid robot with the ground during the entire getting-up action. In the contact state reference sequence, the contact state at each time point is composed of multiple bits, each representing the contact state of the left leg, the right leg, the left hand, and the right hand. For example, in some embodiments, the contact state has 4 bits, representing the left leg / right leg / left hand / right hand, and when the identifier is "1", it indicates that the corresponding end effector is in contact with the ground, and when it is "0", it indicates that it is separated. The contact states of multiple stages are as follows:

[0049]

[0050] The definition of the contact reference timing enables the humanoid robot to sequentially establish or release the support points at different stages. By pre-establishing the contact template during the training stage, the actual contact state can be compared and monitored during the execution process, and the corresponding mechanism can be triggered when there is a deviation to ensure stable execution of the action. The contact state reference sequence predefines the touchdown time point and the contact duration of each end effector, both of which are determined according to the planning of the multi-stage getting-up action sequence and are fixed timing indicators. During execution, the system will determine the establishment and release timing of each support point according to these time parameters, thereby coordinating the force and posture changes of each end effector, making the getting-up action rhythm natural, continuous, and predictable.

[0051] The method for controlling the getting-up of the humanoid robot provided in the present application will be described in detail below. Please refer to Figure 6 , Figure 6 Another embodiment of the method for controlling the getting-up of the humanoid robot provided in the present application comprises:

[0052] 601, if a getting-up trigger instruction of the humanoid robot is received, a posture adjustment control instruction is called to control the humanoid robot to adjust to a preset posture, the preset posture comprising: the hip touching the ground, the end effector of the first lower limb and / or the end effector of the second lower limb touching the ground, the first included angle being a, 45° < a < 180°, and the second included angle being b, 45° < b < 180°;

[0053] 602, a getting-up control program is called to control the humanoid robot to execute a getting-up action sequence to make the humanoid robot transition from the preset posture to a double-foot standing posture, the getting-up action sequence comprising multiple stages of getting-up actions, each stage of getting-up action being controlled by a corresponding getting-up control subprogram to move the center of mass and multiple end effectors of the humanoid robot according to respective preset getting-up reference trajectories;

[0054] In this embodiment, steps 601-602 are similar to the aforementioned embodiment steps 101-102, which will not be repeated here.

[0055] 603. Record a plurality of three-dimensional spatial positions of the plurality of end effectors in actual contact with the ground;

[0056] 604. Calculate the terrain of the current position of the humanoid robot based on the plurality of three-dimensional spatial positions;

[0057] 605. If the terrain of the current position of the humanoid robot is a non-horizontal terrain, adjust the preset reference trajectory of the center of mass of the humanoid robot and the plurality of end effectors according to the terrain;

[0058] In steps 603-605, when the humanoid robot is in a fallen state and performs a standing-up action sequence, the system can obtain the three-dimensional spatial position coordinates (X, Y, Z) of any end effector in the humanoid robot base coordinate system at the moment when the system detects that the end effector establishes effective contact with the ground. Specifically, at the moment when the end effector comes into contact with the ground, the real-time joint angles of all related joints on the limb kinematic chain can be collected through the joint encoder, and the three-dimensional spatial position coordinates of the end effector in the humanoid robot base coordinate system can be calculated in real time in combination with the known humanoid robot link parameters. With the execution of the multi-stage standing-up action, the system can continuously store these three-dimensional spatial position points in actual contact. According to these three-dimensional spatial position points, the system can calculate a mathematical plane that best fits these spatial points using a geometric plane fitting algorithm, and then obtain the normal vector and height information of the current support surface. If the fitting result shows that the normal vector of the plane deviates from the vertical direction by more than a preset threshold, or there is a significant difference in the height of the contact points, it is determined that the current terrain is a non-horizontal terrain, such as an inclined slope, a stair step, or a local protruding platform. If no trajectory adjustment is performed on such terrain and the original reference trajectory is forcibly executed, the center of mass of the humanoid robot will deviate from the center of the actual support surface, or the end effector will exert a support force at the wrong height and angle, which will easily cause slipping or falling.

[0059] Therefore, when the system determines that the terrain where the humanoid robot is located is a non-horizontal plane terrain, in order to prevent instability of support caused by centroid deviation, the controller will adaptively adjust the reference trajectory and target position of the centroid according to the terrain parameters. Specifically, the system will superimpose a compensation amount on the basis of the original centroid reference trajectory according to the height and normal vector direction of the fitting plane. For example, when it is detected that the feet of the humanoid robot fall on a platform 10 cm higher than the reference plane, the system will raise the centroid trajectory of all subsequent stages by 10 cm as a whole, and simultaneously increase the target height of the final standing posture; if the terrain has a slope, in addition to height compensation, the system will also correct the spatial inclination angle of the centroid trajectory based on the normal vector direction to ensure that the centroid is always perpendicular to the actual support surface. Further, the adjustment is not limited to the reference trajectory of the centroid. When the centroid reference trajectory is corrected, the reference trajectories of other parts of the humanoid robot, such as the reference trajectories of all end effectors, can also be adjusted synchronously to ensure the coordination and physical feasibility of the overall motion. If new contact point data is generated during the process of getting up, the system can also dynamically update the terrain model and correct the trajectory in real time to further improve the adaptive ability of the humanoid robot in complex environments.

[0060] 606、In the process of the humanoid robot executing the getting-up motion sequence, the contact state signals of the end effectors of the humanoid robot with the ground are monitored;

[0061] In the process of the humanoid robot executing the getting-up motion sequence, the contact state signals of the end effectors of the humanoid robot with the ground are monitored;

[0062] During the whole rising process, the system can compare the real-time monitored contact state signal with the pre-stored contact state reference sequence to determine whether the current action is consistent with the planned trajectory. If it is detected that the contact timing of the end effector is consistent with the reference sequence, it is considered that the rising action is executed normally; if there is deviation of early contact or delayed contact, different control strategies are executed according to the deviation type, that is, steps 607 or 608 are executed respectively.

[0063] 607, if it is monitored that at least one end effector establishes unexpected contact with the ground before the target time point, it is determined that at least one end effector completes the rising action sequence of the current stage it is in, and other end effectors except at least one end effector continue to execute the rising action sequence of the current stage they are in;

[0064] Step 607 is used to deal with the early contact phenomenon of the humanoid robot in the actual rising process.

[0065] During the execution of the rising action of multiple stages, the end effectors of the humanoid robot establish or release support according to the timing in the contact state reference sequence. However, due to uneven ground, environmental interference or slight deviation of the posture of the humanoid robot itself, it is possible that a certain end effector contacts the ground before the target time point. This early contact belongs to an unexpected contact event, and if not handled, it can lead to misjudgment of the completion of the current stage or cause a sudden change in posture. For example, when the end effector contacts the ground early and still continues to apply driving force or maintains the original downward trajectory, it will cause the humanoid robot to be "lifted up" locally, thereby affecting the stability of the whole rising process.

[0066] Therefore, the embodiment needs to continuously calculate the target time point deviation of each end effector. When it is detected that a certain end effector contacts the ground stably before the planned contact time, the system will not interrupt the action execution of other end effectors in the current stage, but only mark the current stage action of the end effector as "completed", that is, it is considered that the end effector has established a stable support point, and the end effector will no longer continue to execute the remaining motion instructions of the stage. But other end effectors except the end effector still continue to execute their action sequence according to the reference trajectory of the current stage. For example, if it is detected that the right foot of the humanoid robot contacts the ground early before the target time point, the system keeps the current position of the right foot and adjusts the force control output to make it stable support, while the left hand and the left foot still continue to complete the action of the current stage according to the original trajectory. This independent control logic of different ends makes the humanoid robot still maintain the continuity and balance of the whole action in the case of early establishment of partial support.

[0067] Further, in order to prevent attitude mutation or force mutation caused by early contact, the embodiment can introduce a smooth transition mechanism in the control logic. When a certain end effector touches the ground early, the system will not immediately cut off its original trajectory command, but will gradually reduce the end effector speed and torque output to zero through an interpolation algorithm, so that the contact process is smoothly transitioned to the support state. At the same time, the center of mass reference trajectory and the actions of other limbs will be fine-tuned according to the new support conditions to compensate for the center of gravity deviation caused by early contact.

[0068] 608、If it is monitored that at least one end effector does not establish target contact with the ground at the target time point, contact compensation control is triggered; after the contact compensation control ends, the remaining action sequence of the multiple stages of the rising action sequence corresponding to the at least one end effector is executed until the humanoid robot reaches a double-foot standing posture.

[0069] Step 608 is used to cope with the delayed contact phenomenon of the humanoid robot in the actual rising process.

[0070] When the humanoid robot performs the multiple stages of rising actions, the contact state reference sequence can be used to define the explicit contact time point and holding time for each end effector. When a certain end effector does not detect an effective contact signal with the ground at the predetermined target time point during the system operation, it indicates that the end effector has a delayed contact. The delayed contact phenomenon can be caused by various reasons, such as uneven terrain, soft or slippery support surface, blocked end effector action, or sensor delay, etc.

[0071] In order to ensure the continuity and stability of the rising action, the embodiment needs to perform contact compensation control after detecting the delayed contact. The contact compensation control is an adaptive adjustment strategy adopted by the system when it finds that the end effector does not touch the ground as planned. The main purpose is to quickly establish effective support with the ground by readjusting the action of the end effector and related limbs, so as to avoid the instability of the posture or the deviation of the center of gravity of the humanoid robot due to insufficient support. After the contact compensation control ends, the system will continue to execute the remaining action sequence of the multiple stages of the rising action corresponding to the at least one end effector, and continuously monitor the contact state between the end effector of the humanoid robot and the ground. With the continuous execution of the subsequent actions, the humanoid robot gradually enters the final double-foot support stage. In the final stage, the system detects that the double-foot end effector is in stable contact with the ground and the center of mass height reaches the target value of the reference trajectory, and determines that the rising process is completed, and the humanoid robot successfully recovers to a double-foot standing posture. At this time, the hip, knee and ankle angles are in a balanced state, the center of mass is located in the center area of the double-foot support surface, and the humanoid robot has a stable foundation to continue to perform walking or standing tasks.

[0072] In this embodiment, first, the posture adjustment control instruction is called, and the humanoid robot is adjusted from an arbitrary fallen state to a unified preset posture in the first control stage, so that the subsequent standing up control does not need to deal with infinite initial states, greatly improving the universality and generalization ability to multiple initial postures of the method. Then, the standing up control program is called, and in the second control stage, a standing up action sequence including multiple stages is introduced, and a closed-loop feedback mechanism for continuously monitoring whether the end effector successfully establishes target contact is combined. By actively monitoring whether the real contact between the end effector of the humanoid robot and the physical world is successfully established in each standing up stage, the method can assist in judging whether the key node of the standing up action is reached, that is, whether the action sequence of each stage is completed. After the previous stage is completed, the next stage will be executed. This mechanism can effectively suppress the subsequent action errors caused by slipping or missing of the supporting hand / foot, thereby avoiding accidental interruption or complete failure of the standing up process.

[0073] Through the phased control framework and the closed-loop execution mechanism based on physical contact confirmation, the embodiment can cover various fallen postures that may occur in actual applications, and enable the humanoid robot to realize fast, stable and basically fault-tolerant autonomous standing up under uncertain contact conditions.

[0074] In some specific embodiments, considering that there are many reasons for delayed contact, which may be caused by slight ground depression, large terrain drop, sensor measurement delay or posture cumulative error of the humanoid robot body. A single compensation strategy may not cover all abnormal situations, or may cause action failure or long-time deadlock in some scenarios. To cope with such complexity, the embodiment proposes multiple types of contact compensation control strategies. The difference between these strategies mainly lies in how the system re-plans or adjusts the logical path of the action when the end effector fails to contact the ground on time. The system can select the most suitable compensation logic according to task requirements or computing resources, such as prioritizing time or prioritizing success rate. The following describes each strategy of the contact compensation control in step 608:

[0075] Strategy one, continue to probe + time termination:

[0076] If any end effector fails to establish target contact with the ground at the target time point, the control end effector continues to probe; if during the first preset time period, the end effector fails to monitor the establishment of effective contact signal with the ground during the continuous probing process, the end effector of the current stage of target contact is ended, and the contact compensation control is ended. The first preset time period is extended after the target time point.

[0077] In the first strategy, when the end effector fails to establish the contact signal at the target time point, the system first performs the operation of continuing to probe, i.e., extending the movement of the end effector in the original reference trajectory direction, to try to find the actual contact surface within a shorter distance. The first strategy is suitable for scenarios where the contact surface is slightly lower than the planned position or the terrain surface has flexible materials causing sensor contact delay. If the system still fails to detect the contact signal within the first preset time period, it is determined that the end effector cannot establish support in the current stage, and the system automatically ends the target contact of the stage and exits the compensation control to prevent the end effector from continuing to probe uselessly, causing mechanical interference or unstable posture. The first strategy has simple logic and fast response, and is suitable for flat environments with high detection noise, which can effectively avoid unnecessary repeated attempts.

[0078] It should be noted that the total execution time of the entire standing-up process is a fixed value, for example, about 5 seconds, and this total execution time will not be extended due to contact delay or compensation action of a certain end effector. That is, even if a certain end effector fails to complete the action of the current stage as planned, the system still needs to maintain the continuous execution of the overall timing to ensure that other end effectors complete their respective trajectory planning in pace.

[0079] Strategy two, probe + retreat + re-planning:

[0080] If it is monitored that any end effector fails to establish the target contact with the ground at the target time point, the end effector is controlled to continue to probe; if, within the first preset time period, the end effector fails to monitor the establishment of the effective contact signal with the ground during the process of continuing to probe, the end effector is controlled to retreat to the initial position corresponding to the standing-up action sequence of the current stage, and the reference trajectory of the end effector in the current stage is re-planned, and the re-planned reference trajectory is executed from the initial position corresponding to the standing-up action sequence of the current stage. If the end effector establishes contact with the ground at the adjusted target time point, the remaining action sequence of the standing-up action of the end effector in the multiple stages is continued to be executed, and the first preset time period is postponed after the target time point. If the end effector fails to monitor the establishment of the effective contact signal with the ground at the adjusted target time point, the target contact of the end effector in the current stage is ended to end the contact compensation control.

[0081] In strategy two, when the delayed contact occurs, the system first performs the continue down action. If no valid contact signal is detected within the first preset time period, the control end effector retreats to the initial position of the current stage of the get-up action sequence. Then, the system re-plans the reference trajectory based on the current position of the end effector and the current terrain estimation, so that it has a new target contact time and contact position. When the end effector successfully contacts the ground at the adjusted target time point, the system considers that the compensation is successful, and continues to perform the subsequent action sequence; if no contact signal is detected, the current stage target contact is ended and the compensation process is exited.

[0082] Strategy two is more robust than strategy one, and through the closed-loop mechanism of retreat-replan-reexecute, the success rate of establishing support in an environment with large terrain differences is improved.

[0083] Strategy three, continue down + re-plan:

[0084] If any end effector is detected to not establish a target contact with the ground at the target time point, the control end effector continues to descend; if within the first preset time period, the end effector does not monitor an effective contact signal with the ground during the continue down process, the reference trajectory of the end effector in the current stage is re-planned based on the current position of the end effector, and the re-planned reference trajectory is continued to be executed; if the end effector establishes contact with the ground at the adjusted target time point, the remaining action sequence of the get-up action of the multiple stages of the end effector is continued to be executed, and the first preset time period is extended after the target time point; if no effective contact signal between the end effector and the ground is monitored at the adjusted target time point, the target contact of the end effector in the current stage is ended to end the contact compensation control.

[0085] This strategy three does not retreat after the continue down fails, but re-plans the trajectory based on the current position as the new reference state. Unlike the aforementioned strategy two, this scheme does not perform a retreat operation, but directly re-plans the reference trajectory of the current stage at the current position of the end effector, trying some new contact points around the original position. Moreover, this re-planning is only performed for the end effector that has an abnormality, and other end effectors are not affected, while the total execution time of the entire get-up action remains unchanged. When re-planning, the system generates a new trajectory based on the current pose of the end effector and dynamic constraints, thereby reducing the time delay and energy consumption caused by the retreat action. If an effective contact is established at the adjusted target time point, the system continues to execute the subsequent stage; if no effective signal is detected, the target contact of the current stage is ended and the contact compensation control is ended. Strategy three reduces the energy consumption and delay caused by the retreat action compared to strategy two, and balances response speed and adaptability.

[0086] Strategy four, retreat + re-plan:

[0087] If any of the end effectors fails to establish the target contact with the ground at the target time point, the control system controls the end effectors to retreat to the initial position corresponding to the current stage of the rising motion sequence, and re-plans the reference trajectory of the end effectors at the current stage, and continues to execute the re-planned reference trajectory from the initial position corresponding to the current stage of the rising motion sequence; if the end effectors establish contact with the ground at the adjusted target time point, the control system continues to execute the remaining action sequence of the rising motion of the multiple stages of the end effectors; if the end effectors fail to establish the effective contact signal with the ground at the adjusted target time point, the control system ends the target contact of the current stage of the end effectors, and ends the contact compensation control.

[0088] Strategy four does not perform the downward motion after detecting the delayed contact, but directly controls the end effectors to retreat to the initial position of the current stage of the rising motion sequence. The system then re-plans the reference trajectory of the end effectors based on the new environment data, and re-executes from the initial position. If the effective contact is detected at the adjusted target time point, the subsequent rising motion is continued; if the effective contact is still not detected, the target contact of the stage is ended. The core of strategy four is the retreat priority, which is mainly applicable when the posture of the humanoid robot has deviated from the reference trajectory or the end effectors are constrained, for example, the hands or knees are limited by obstacles, and the continued downward motion will cause collision risk. By retreating, the posture state can be reset to provide reliable initial conditions for re-planning, and ensure the safety of the motion.

[0089] Five, instant re-planning:

[0090] If any of the end effectors fails to establish the target contact with the ground at the target time point, the control system controls the end effectors to retreat to the initial position corresponding to the current stage of the rising motion sequence, and re-plans the reference trajectory of the end effectors at the current stage, and continues to execute the re-planned reference trajectory from the initial position corresponding to the current stage of the rising motion sequence; if the end effectors establish contact with the ground at the adjusted target time point, the control system continues to execute the remaining action sequence of the rising motion of the multiple stages of the end effectors; if the end effectors fail to establish the effective contact signal with the ground at the adjusted target time point, the control system ends the target contact of the current stage of the end effectors, and ends the contact compensation control.

[0091] Strategy five also does not perform the dive-down and does not perform the fallback, but directly replans the reference trajectory of the stage based on the current position of the end effector, which is suitable for scenarios where the humanoid robot needs to perform a fast sit-up or perform a task in a dynamic environment. When a delayed contact is detected, the system does not perform any dive-down or fallback operation, but directly replans the reference trajectory of the current stage based on the current pose and position of the end effector as a new starting point, and delays the execution pace of the original target time point in time. The goal of this strategy is to maintain the time continuity of the overall motion, and its replanning algorithm prioritizes fast convergence and whole-body coordination, which is suitable for high-response scenarios or humanoid robot platforms with high degrees of freedom control. If a contact is established at the adjusted target time point, the system continues to perform the subsequent motion; if no valid contact signal is detected, it is determined that the contact of the current stage is not achievable and the stage target contact is ended to maintain the stability of the overall sit-up process.

[0092] It should be noted that in the strategies involving fallback operations (strategy two and strategy four described above), when the end effector with contact anomaly occurs during the fallback, the other end effectors can have three different motion conditions:

[0093] In one embodiment, as a preferred solution, in some scenarios, the system can only perform a fallback operation on the end effector with contact anomaly, while keeping the other end effectors without anomaly continue to track the corresponding reference trajectory. This way can reduce the time consumption brought by the fallback and ensure the sit-up speed, which is suitable for the case where only partial support points deviate while the overall pose is still stable. For example, when one of the upper limbs of the humanoid robot does not touch the ground at the expected time, while the other upper limb and both feet have established stable support, the system can only control the end effector of the upper limb to fallback and replan its contact trajectory, while the other support points do not fallback, thereby speeding up the compensation execution speed and maintaining the overall pose continuity.

[0094] In another embodiment, all end effectors can be controlled to fallback synchronously to the initial position corresponding to the current stage of the sit-up motion sequence. The advantage of this way is that each limb re-establishes a unified motion starting point, which facilitates the system to recalculate the dynamic balance constraint conditions of the whole body, and can maintain the whole-body coordination and center of mass balance in the new sit-up stage. This way is particularly suitable for the case where the humanoid robot is in a complex terrain or has a large pose deviation, which can significantly improve the stability of the sit-up after recovery.

[0095] In another embodiment, only the end effector with contact anomaly can be controlled to perform a fallback operation, while the other end effectors without anomaly remain stationary on the original trajectory, but this way will prolong the total execution time of the entire sit-up motion.

[0096] In some specific embodiments, the standing-up control program is trained according to an imitation learning reference trajectory by a reinforcement learning strategy, and a training process of the standing-up control program adopts a reward function, the reward function comprising: a collision penalty function and a slipping penalty function; the collision penalty function is used to impose a penalty when the torso of the humanoid robot or its own limbs have an undesired contact with the surrounding environment; the slipping penalty function is used to impose a penalty when tangential slipping occurs when any end effector as a support point is in contact with the ground.

[0097] This embodiment takes into account that the multi-stage standing-up motion sequence of the humanoid robot is a complex and multi-point contact switching dynamic process. In order to achieve smooth and robust control effect, the reinforcement learning RL method can be used to train the standing-up control program, and the imitation learning reference trajectory based on human standing-up motion capture data is introduced in the training process. Only relying on imitation learning is not enough to solve the key physical stability problem in the standing-up process, especially slipping and collision. Therefore, the embodiment particularly includes a targeted penalty function in the reward function to optimize the robustness of the strategy. Specifically, the collision penalty function aims to reduce undesired collisions in the standing-up process, including self-collisions between limbs and undesired environmental collisions of key parts with the ground (such as the torso and head touching the ground). This function can be configured to impose a penalty when the torso of the humanoid robot or its own limbs have an undesired contact with the surrounding environment. Through this penalty term, the system guides the control strategy to learn to avoid high-risk actions and choose a safer standing-up path. The purpose of the slipping penalty function is to suppress the tangential sliding of the support point on the ground and prevent imbalance and falling due to unstable support. This function can be configured to impose a penalty if it is monitored that the tangential velocity of any end effector as a support point in contact with the ground exceeds a threshold. This penalty term encourages the control strategy to learn to generate sufficient normal force and reduce tangential force to ensure the stability of the hand or foot support point during the weight transfer process.

[0098] In some specific embodiments, the training process of the standing-up control strategy adopts a curriculum learning mechanism and a domain randomization mechanism; the curriculum learning mechanism is used to gradually increase the difficulty of the standing-up task on one or more difficulty axes and automatically adjust the difficulty axes according to the evaluation results of the standing-up success rate, collision rate and time consumption, the difficulty axes including friction coefficient, terrain slope, initial posture complexity, effector constraint, sensor noise and external disturbance; the domain randomization mechanism is used to randomize the ground friction coefficient, recovery coefficient, terrain parameter, humanoid robot dynamics parameter, execution delay, sensor noise and external disturbance in different training rounds.

[0099] The embodiment is considering that the standing-up control strategy trained in only a single idealized simulation environment is not robust enough to cope with complex physical conditions in the real world, especially variable ground friction, uneven terrain, or sensor noise, etc. In order to systematically improve the generalization ability of the standing-up control strategy, the embodiment introduces a curriculum learning and domain randomization mechanism in the training process.

[0100] The function of the curriculum learning mechanism is to guide the control strategy to gradually increase the difficulty of the standing-up task on one or more difficulty axes. The system starts training from a simple task (for example, high friction, flat ground), and automatically adjusts the difficulty axis according to the evaluation results (for example, standing-up success rate, collision rate and time consumption) in the training process. If the success rate is high, the difficulty is increased (such as reducing the friction); if the success rate is low, the difficulty is reduced to maintain the stability and scientificity of the training. The difficulty axis is configured to include but not limited to: friction coefficient (for example, from high to low), terrain slope or unevenness (for example, step height gradually increasing), initial posture complexity (for example, from standard supine to lateral lying, prone), actuator constraint (for example, gradually tightening the torque upper limit), sensor noise (for example, from low to high), and external disturbance (for example, light push disturbance).

[0101] The function of the domain randomization mechanism is to introduce diversity in the simulation environment, so that the standing-up control strategy can foresee various extreme or non-ideal working conditions during the training phase. The mechanism is configured to randomize a series of simulation parameters in different training rounds. These randomized parameters include: ground friction coefficient, recovery coefficient, terrain parameters (such as height field, step distribution), humanoid robot dynamics parameters (such as mass, moment of inertia ratio), execution delay, sensor noise (such as IMU noise or drift), and external disturbance (such as random pulse thrust). In this way, the trained standing-up control strategy is forced to learn a robust control law that is not sensitive to parameter changes, thereby significantly improving the adaptability and reliability of the standing-up control strategy in the real physical environment.

[0102] The system for controlling the standing-up of a humanoid robot provided in the present application will be described in detail below. Please refer to Figure 7 , Figure 7 An embodiment of the system for controlling the standing-up of a humanoid robot provided in the present application, the system comprises:

[0103] The first calling unit 701 is configured to call a posture adjustment control instruction to control the humanoid robot to adjust to a preset posture if a standing-up trigger instruction of the humanoid robot is received, the preset posture comprising: a hip part touching the ground, an end effector of a first lower limb and / or an end effector of a second lower limb touching the ground, a first included angle being α, 45° < α < 180°, and a second included angle being β, 45° < β < 180°.

[0104] The second calling unit 702 is configured to call a standing-up control program to control the humanoid robot to perform a standing-up action sequence, so that the humanoid robot transits from a preset posture to a posture with both feet standing, and the standing-up action sequence includes a plurality of stages of standing-up actions, and each stage of standing-up action controls the center of mass and the plurality of end effectors of the humanoid robot to move according to a respective preset standing-up reference trajectory.

[0105] In the system of the embodiment, the functions of the units correspond to the steps in the method embodiments shown in the foregoing Figure 1 or Figure 2 The steps in the method embodiments shown in the foregoing

[0106] The application also provides a humanoid robot, please refer to Figure 8 , Figure 8 An embodiment of the humanoid robot provided by the application comprises:

[0107] The processor 801, the memory 802, the input and output unit 803, and the bus 804;

[0108] The processor 801 is connected with the memory 802, the input and output unit 803, and the bus 804;

[0109] The memory 802 stores a program, and the processor 801 calls the program to perform the method of standing-up control of the humanoid robot.

[0110] The application also relates to a computer readable storage medium, and the computer readable storage medium stores a program, and when the program runs on a computer, the computer performs the method of standing-up control of the humanoid robot.

[0111] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the system, the device and the unit described above can refer to the corresponding processes in the foregoing method embodiments, and will not be described here.

[0112] In the several embodiments provided by the application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. There can be another division manner for actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0113] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0114] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0115] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical scheme of the present application essentially or the part of the prior art that contributes to the technical scheme or the whole or part of the technical scheme can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method of each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, read-only memory), a random access memory (RAM, random access memory), a magnetic disk or an optical disk, and various program code storage media.

Claims

1. A method for controlling the standing up of a humanoid robot, characterized in that, The humanoid robot includes a torso, a first upper limb and its end effector, a second upper limb and its end effector, a first lower limb and its end effector, and a second lower limb and its end effector. The first upper limb and the second upper limb are respectively connected to the torso. The torso includes a hip. The first lower limb and the second lower limb are respectively connected to the hip. The first lower limb and the second lower limb each include a thigh and a calf. The thigh and the calf are connected by a knee joint. A first angle is formed between the thigh and the calf of the first lower limb, and a second angle is formed between the thigh and the calf of the second lower limb. The method includes: If a humanoid robot is given a command to stand up, the posture adjustment control command is invoked to control the humanoid robot to adjust to a preset posture. The preset posture includes: the hips touching the ground, the end effector of the first lower limb and / or the end effector of the second lower limb touching the ground, the first included angle being α, 45° < α < 180°, and the second included angle being β, 45° < β < 180°. The stand-up control program is invoked to control the humanoid robot to execute a stand-up action sequence, so that the humanoid robot transitions from the preset posture to a standing posture. The stand-up action sequence includes multiple stages of stand-up actions. Each stage of stand-up action is controlled by a corresponding stand-up control subroutine to move the center of mass of the humanoid robot and multiple end effectors according to their respective preset stand-up reference trajectories.

2. The method according to claim 1, characterized in that, The multiple phases of the standing up movement include at least: a single-hand support phase, a single-hand and single-leg support phase, a double-hand and single-leg support phase, a double-hand and double-leg support phase, and a double-leg support phase; The single-handed support phase includes moving the first upper limb of the humanoid robot so that its end effector contacts the ground, thereby establishing a first support point on the ground; The single-hand, single-leg support phase includes, while maintaining support at the first support point, moving the first lower limb of the humanoid robot, causing the end effector of the first lower limb to move from a first position to a second position. The second position is closer to the humanoid robot's hip than the first position, and the end effector of the first lower limb contacts the ground at the second position, thereby establishing a second support point. During the single-hand, single-leg support phase, the knee joint of the first lower limb moves away from the ground, forming a first angle between the thigh and lower leg of the first lower limb, and a second angle between the thigh and lower leg of the second lower limb, the second angle being greater than the first angle. The two-handed, one-legged support phase includes moving the humanoid robot's second upper limb so that its end effector contacts the ground, thereby establishing a third support point; and moving the humanoid robot's second lower limb so that the end effector of the second lower limb moves from a third position to a fourth position, the fourth position being closer to the humanoid robot's hip than the third position; during the process of the second lower limb moving from the third position to the fourth position, the angle between the thigh and the lower leg of the second lower limb gradually decreases; The two-handed and two-footed support phase includes, while maintaining support at the first, second, and third support points, moving the hips of the humanoid robot from the fifth position to the sixth position, where the sixth position is higher than the fifth position. At the same time, the second lower limb of the humanoid robot is moved so that its end effector contacts the ground, thereby establishing the fourth support point. The bipedal support phase includes controlling the humanoid robot's center of gravity to shift from a seventh position to an eighth position. The height of the eighth position above the ground is greater than that of the seventh position. During the shift of the humanoid robot's center of gravity from the seventh position to the eighth position, the end effectors of the first and second upper limbs separate from the ground.

3. The method according to claim 1, characterized in that, The method further includes: Record multiple three-dimensional spatial positions where the multiple end effectors actually come into contact with the ground; The terrain at the current location of the humanoid robot is calculated based on the multiple three-dimensional spatial positions; If the terrain at the current location of the humanoid robot is a non-horizontal surface, then the center of mass of the humanoid robot and the preset standing reference trajectories of multiple end effectors are adjusted according to the terrain.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: During the process of the humanoid robot performing the standing action sequence, the contact status signal between the humanoid robot's end effector and the ground is monitored; If it is detected that at least one of the end effectors establishes unexpected contact with the ground before the target time point, it is determined that the at least one end effector has completed the standing action sequence of its current stage, and the other end effectors besides the at least one end effector continue to execute the standing action sequence of their current stage. or, If it is detected that at least one of the end effectors fails to establish target contact with the ground at the target time point, contact compensation control is triggered; After the contact compensation control is completed, the remaining action sequence of the standing action sequence corresponding to the at least one end effector is executed until the humanoid robot reaches a bipedal standing posture.

5. The method according to claim 4, characterized in that, The contact compensation control includes: If it is detected that any of the end effectors fails to establish target contact with the ground at the target time point, the end effector is controlled to continue descending; If, during the first preset time period, the end effector fails to detect a valid contact signal with the ground during its continued descent, the target contact of the end effector in the current stage is terminated to end the contact compensation control, and the first preset time period is extended after the target time point.

6. The method according to claim 4, characterized in that, The contact compensation control includes: If it is detected that any of the end effectors fails to establish target contact with the ground at the target time point, the end effector is controlled to continue descending; If, during the first preset time period, the end effector fails to detect a valid contact signal with the ground while continuing to descend, the end effector is controlled to retreat to the initial position corresponding to the current stage's standing action sequence, and the reference trajectory of the end effector in the current stage is replanned. The replanned reference trajectory is then executed from the initial position corresponding to the current stage's standing action sequence. If the end effector establishes contact with the ground at the adjusted target time point, the remaining action sequence of the multiple stages of the end effector's standing action will continue to be executed, and the first preset time period will be postponed after the target time point; If no effective contact signal is detected between the end effector and the ground at the adjusted target time point, the target contact of the end effector in the current stage will be terminated to end the contact compensation control.

7. The method according to claim 4, characterized in that, The contact compensation control includes: If it is detected that any of the end effectors fails to establish target contact with the ground at the target time point, the end effector is controlled to continue descending; If, within the first preset time period, the end effector fails to detect a valid contact signal with the ground during its continued descent, the reference trajectory of the end effector at the current stage is replanned based on its current position, and the replanned reference trajectory is then executed. If the end effector establishes contact with the ground at the adjusted target time point, the remaining action sequence of the multiple stages of the end effector's standing action will continue to be executed, and the first preset time period will be postponed after the target time point; If no effective contact signal is detected between the end effector and the ground at the adjusted target time point, the target contact of the end effector in the current stage will be terminated to end the contact compensation control.

8. The method according to claim 4, characterized in that, The contact compensation control includes: If it is detected that any of the end effectors fails to establish target contact with the ground at the target time point, the end effector is controlled to revert to the initial position corresponding to the current stage's standing action sequence, and the reference trajectory of the end effector in the current stage is replanned. The replanned reference trajectory is then executed from the initial position corresponding to the current stage's standing action sequence. If the end effector establishes contact with the ground at the adjusted target time point, the remaining action sequence of the multiple stages of the end effector's standing action will continue to be executed; if no effective contact signal is detected between the end effector and the ground at the adjusted target time point, the target contact of the end effector in the current stage will be terminated to end the contact compensation control.

9. The method according to claim 4, characterized in that, The contact compensation control includes: If it is detected that any of the end effectors fails to establish target contact with the ground at the target time point, the reference trajectory of the end effector at the current stage is replanned based on the current position of the end effector, and the replanned reference trajectory is executed. If the end effector makes contact with the ground at the adjusted target time point, the remaining action sequence of the multiple stages of the end effector's standing action will continue to be executed; If no effective contact signal is detected between the end effector and the ground at the adjusted target time point, the target contact of the end effector in the current stage will be terminated to end the contact compensation control.

10. A humanoid robot, characterized in that, The humanoid robot includes: Processor, memory, input / output units, and bus; The processor is connected to the memory, the input / output unit, and the bus; The memory stores a program, which the processor invokes to perform the method as described in any one of claims 1 to 9.

Images