Robot motion trajectory generation method, device, medium and robot

Abstract

The present disclosure relates to the technical field of robots, and provides a robot motion trajectory generation method, device, medium and robot. The method comprises: determining an initial motion trajectory of a robot in a case where an upper limb is fixed; and determining a target motion trajectory of the robot in a case where neither the upper limb nor the lower limb is fixed according to the initial motion trajectory, so that the robot moves according to the target motion trajectory. In this way, by first determining the initial motion trajectory of the robot in the case where the upper limb is fixed, only the degree of freedom factor of the lower limb of the robot is considered, and then the target motion trajectory of the robot in the case where both the upper limb and the lower limb are fixed is determined according to the initial motion trajectory, and the degree of freedom factors of the upper limb and the lower limb of the robot are considered, so that the target motion trajectory which makes the upper limb and the lower limb coordinate is quickly obtained.

Description

Technical Field

[0001] This disclosure relates to the field of robotics technology, and in particular to a method, apparatus, medium, and robot for generating robot motion trajectories. Background Technology

[0002] Robots are complex mechatronic systems integrating advancements in materials science, structural engineering, electrical engineering, communication bus technology, and high-performance chips. In recent years, with the development of high-energy-density frameless motor technology and industrial fieldbus technology, as well as breakthroughs in high-performance integrated chips, robots have achieved tremendous progress. More and more robots are participating in daily work and life, including in logistics, catering, service, education, and medical services. However, current robots are prone to exhibiting uncoordinated movements between their upper and lower limbs. Summary of the Invention

[0003] To overcome the problems existing in related technologies, this disclosure provides a method, apparatus, medium and robot for generating robot motion trajectories.

[0004] According to a first aspect of the present disclosure, a method for generating a robot motion trajectory is provided, the method comprising:

[0005] Determine the robot's initial motion trajectory with its upper limbs fixed;

[0006] Based on the initial motion trajectory, the target motion trajectory of the robot is determined when neither the upper nor lower limbs are fixed, so that the robot moves according to the target motion trajectory.

[0007] Optionally, determining the robot's initial motion trajectory with its upper limbs fixed includes:

[0008] The robot's upper limb degrees of freedom are set to a preset fixed angle, or the robot's upper limbs and torso are defined as a rigid body.

[0009] The initial motion trajectory of the robot with its upper limbs fixed is determined with the goal of minimizing the sum of squares of the joint torques.

[0010] Optionally, determining the initial motion trajectory of the robot with its upper limb fixed, with the objective of minimizing the sum of squared joint moments, includes:

[0011] Determine the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint condition, wherein the first preset constraint condition is used to constrain the movement of the robot with its upper limbs fixed;

[0012] Using the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint as the target constraint, and with the goal of minimizing the sum of squares of the joint torques of the robot, the first nonlinear optimization problem is solved to obtain the initial motion trajectory of the robot with its upper limbs fixed. The first nonlinear optimization problem is used to characterize the sum of squares of the joint torques of the robot within one motion cycle with its upper limbs fixed.

[0013] Optionally, the first preset constraint condition includes a first equality constraint condition and a first inequality constraint condition, wherein the first equality constraint condition includes robot gait periodicity constraint and / or robot walking speed constraint, and the first inequality constraint condition includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, and robot foot swing constraint.

[0014] Optionally, determining the walking dynamics model of the robot with its upper limbs fixed includes:

[0015] The continuous dynamics model and discrete dynamics model of the robot under the condition of upper limb fixation are determined. The continuous dynamics model is used to characterize the multi-rigid-body dynamics relationship and contact point acceleration relationship of the robot under the condition of upper limb fixation. The discrete dynamics model is used to characterize the position relationship, relative sliding relationship with the ground and velocity relationship of the robot before and after the occurrence of discrete events. The discrete events include foot landing events and foot lifting events.

[0016] Optionally, determining the discrete dynamic model of the robot with its upper limb fixed includes:

[0017] The discrete dynamic model of the robot with its upper limbs fixed is determined as follows:

[0018]

[0019] Wherein, the superscript + indicates the moment before the discrete event, the superscript - indicates the moment after the discrete event, the subscript c indicates the contact point between the robot and the ground, q indicates the generalized coordinate value corresponding to the robot, M indicates the inertia matrix corresponding to the robot, and F imp δ represents the instantaneous collision force corresponding to the robot, and δ represents the duration of the discrete event. c- J represents the contact Jacobian matrix corresponding to the robot's contact state with the ground before the discrete event. c+ The contact Jacobian matrix represents the contact state between the robot and the ground after the discrete event.

[0020] Optionally, determining the discrete dynamic model of the robot with its upper limb fixed includes:

[0021] The discrete dynamic model of the robot with its upper limbs fixed is determined according to the following rules:

[0022] When the landing event occurs, the instantaneous collision force of the robot is greater than or equal to the first threshold and less than or equal to the second threshold.

[0023] When the foot-lifting event occurs, the instantaneous collision force corresponding to the robot is zero.

[0024] Optionally, determining the target motion trajectory of the robot when neither its upper nor lower limbs are fixed, based on the initial motion trajectory, includes:

[0025] Constraints are constructed based on the initial motion trajectory, with the goal of minimizing the robot's energy consumption, to determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed.

[0026] Optionally, the constraint conditions constructed based on the initial motion trajectory, with the objective of minimizing the robot's energy consumption, determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, including:

[0027] A walking dynamics model and a second preset constraint condition are determined for the robot when its upper and lower limbs are not fixed. The second preset constraint condition is used to constrain the movement of the robot when its upper and lower limbs are not fixed.

[0028] Using the initial motion trajectory, the walking dynamics model of the robot when neither the upper nor lower limbs are fixed, and the second preset constraint as the target constraint, the optimal solution determined based on the initial motion trajectory is used as the initial guess, and the goal is to minimize the robot's energy consumption to solve the second nonlinear optimization problem, thereby obtaining the target motion trajectory. The optimal solution is used to characterize time, the robot's generalized state, and the robot's generalized force. The second nonlinear optimization problem is used to characterize the energy consumption of the robot's motion within one motion cycle when neither the upper nor lower limbs are fixed.

[0029] Optionally, the second preset constraint includes a second equality constraint and a second inequality constraint, wherein the second equality constraint includes at least one of robot gait periodicity constraint, robot walking speed constraint and robot floating base trajectory constraint, and the second inequality constraint includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, robot foot swing constraint and robot angular velocity constraint.

[0030] According to a second aspect of the present disclosure, a robot motion trajectory generation apparatus is provided, the apparatus comprising:

[0031] The initial trajectory determination module is configured to determine the robot's initial motion trajectory with its upper limbs fixed.

[0032] The target trajectory determination module is configured to determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, based on the initial motion trajectory, so that the robot moves according to the target motion trajectory.

[0033] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a first processor, implement the steps of any robot motion trajectory generation method provided in the first aspect of the present disclosure.

[0034] According to a fourth aspect of the present disclosure, a robot is provided, comprising:

[0035] processor;

[0036] Memory used to store processor-executable instructions;

[0037] The processor is configured as follows:

[0038] Perform the steps of implementing any of the robot motion trajectory generation methods provided in the first aspect of this disclosure.

[0039] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects:

[0040] By first determining the robot's initial motion trajectory with its upper limbs fixed, and considering only the degree of freedom of the robot's lower limbs, the target motion trajectory of the robot with both upper and lower limbs fixed is determined based on this initial motion trajectory. By combining the degree of freedom of the robot's upper and lower limbs, the target motion trajectory that coordinates the upper and lower limbs can be obtained quickly.

[0041] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0043] Figure 1 This is a flowchart illustrating a method for generating a robot motion trajectory according to an exemplary embodiment;

[0044] Figure 2This is a schematic diagram illustrating the dynamics of robot walking according to an exemplary embodiment;

[0045] Figure 3 This is an implementation illustrated according to an exemplary embodiment. Figure 1 Flowchart of step S1;

[0046] Figure 4 This is an implementation illustrated according to an exemplary embodiment. Figure 2 Flowchart of step S12;

[0047] Figure 5 This is an implementation illustrated according to an exemplary embodiment. Figure 4 Flowchart of step S122;

[0048] Figure 6 This is an implementation illustrated according to an exemplary embodiment. Figure 1 Flowchart of step S2;

[0049] Figure 7 This is a block diagram illustrating a robot motion trajectory generation device according to an exemplary embodiment;

[0050] Figure 8 This is a block diagram of a robot for a robot motion trajectory generation method according to an exemplary embodiment. Detailed Implementation

[0051] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0052] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0053] Figure 1 This is a flowchart illustrating a robot trajectory generation method according to an exemplary embodiment, such as... Figure 1As shown, the robot motion trajectory generation method can be used in electronic devices and robots, and the electronic devices can be implemented in various forms. For example, the electronic devices described in this disclosure may include mobile phones, tablets, laptops, desktop computers, handheld computers, personal digital assistants (PDAs), portable media players (PMPs), navigation devices, wearable devices, smart bracelets, etc. If the robot motion trajectory generation method is applied to an electronic device, the electronic device can send the obtained target optimized trajectory to the robot. The robot can be a robot with both upper and lower limbs, such as a humanoid robot. The robot motion trajectory generation method includes:

[0054] Step S1: Determine the initial motion trajectory of the robot with its upper limbs fixed.

[0055] Please see Figure 2 , Figure 2 (a) shows a schematic diagram representing the robot's upper limbs being fixed.

[0056] The initial motion trajectory can be the motion trajectory of the lower limbs of the robot while the upper limbs are fixed.

[0057] Step S2: Based on the initial motion trajectory, determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, so that the robot moves according to the target motion trajectory.

[0058] Figure 2 (b) shows a schematic diagram representing a robot whose upper and lower limbs are not fixed.

[0059] The target motion trajectory can be the motion trajectory of the robot's upper and lower limbs when neither is fixed and both are free. In determining the target motion trajectory, the initial motion trajectory can be used as a constraint.

[0060] It should be noted that the target motion trajectory can be generated during the robot's movement or it can be generated offline. Offline generation can be done by the robot itself or by other electronic devices that generate the trajectory offline and then send it to the robot.

[0061] By first determining the robot's initial motion trajectory with its upper limbs fixed, and considering only the degree of freedom of the robot's lower limbs, the target motion trajectory of the robot with both upper and lower limbs fixed is determined based on this initial motion trajectory. By combining the degree of freedom of the robot's upper and lower limbs, the target motion trajectory that coordinates the upper and lower limbs can be obtained quickly.

[0062] By determining the initial motion trajectory when the upper limbs are fixed, we obtain the initial motion trajectory when the upper limbs have relatively fixed degrees of freedom. Then, based on this initial motion trajectory, we determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, which is the target motion trajectory when the upper and lower limbs have full degrees of freedom. Since the target motion trajectory is determined based on the initial motion trajectory, it can converge quickly compared to determining it directly from the case of full degrees of freedom of the upper and lower limbs, and thus quickly obtain the target motion trajectory.

[0063] In one possible implementation, please refer to Figure 3 Step S1 may include:

[0064] Step S11: Determine the robot's upper limb degrees of freedom as a preset fixed angle or determine the robot's upper limbs and torso as a rigid body.

[0065] The robot's upper limbs can be fixed by setting the degrees of freedom of the robot's upper limbs to a preset fixed angle, or by fixing the robot's upper limbs and torso as a rigid body.

[0066] The robot's structure can include two upper limbs, a torso, and two lower limbs. Each upper limb has 7 joints (excluding the joints in the fingers), each lower limb has 6 joints (excluding the joints in the toes), and the waist has 0 to 3 joints. The floating base includes 6 degrees of freedom. For the robot, one joint corresponds to one degree of freedom. The 7 degrees of freedom of the upper limbs are equivalent to 3 degrees of freedom of the human shoulder joint, 1 degree of freedom of the human elbow joint, and 3 degrees of freedom of the human wrist joint; the 6 joints of the lower limbs are equivalent to 3 degrees of freedom of the human hip joint, 1 degree of freedom of the human knee joint, and 2 degrees of freedom of the human ankle joint.

[0067] The degrees of freedom corresponding to the joints of the robot's upper limbs are all set to preset fixed angles to fix the robot's upper limbs. The two upper limbs of the robot have 14 degrees of freedom, and these 14 degrees of freedom are set to preset fixed angles.

[0068] It should be understood that the preset fixed angle setting should aim to minimize the shift in the center of gravity of the torso and arms as a whole. For example, the angles of each upper limb joint can be when the two upper limbs are naturally downward, or when they are in a running position. The running position can be one upper limb forward and the other upper limb backward, or other positions. This embodiment does not limit the angles of each upper limb joint.

[0069] By defining the robot's upper limbs and torso as a single rigid body and fixing the robot's upper limbs, it can be understood that the robot's two upper limbs and torso are considered as a single unit, with the degrees of freedom of the upper limbs varying with the degrees of freedom of the torso, and the upper limbs being fixed relative to the torso.

[0070] Step S12: Determine the initial motion trajectory of the robot with its upper limbs fixed, with the goal of minimizing the sum of squares of the joint torques.

[0071] The formula for calculating the sum of squares of joint moments is:

[0072] ∫∑τ 2 (t)dt(1)

[0073] Where τ represents the joint torque and t represents time.

[0074] In one possible implementation, please refer to Figure 4 Step S12 may include:

[0075] Step S121: Determine the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint conditions. The first preset constraint conditions are used to constrain the robot's movement when its upper limbs are fixed.

[0076] Determining the walking dynamics model of a robot with its upper limbs fixed can include:

[0077] The continuous dynamics model and discrete dynamics model of the robot with its upper limbs fixed are determined. The continuous dynamics model is used to characterize the multi-rigid-body dynamics and contact point acceleration relationships of the robot with its upper limbs fixed. The discrete dynamics model is used to characterize the robot's positional relationship, relative sliding relationship with the ground, and velocity relationship before and after discrete events. Discrete events include foot landing events and foot lifting events.

[0078] Please continue reading. Figure 2 When a robot walks, the phases are divided into a single support phase (SSP) and a double support phase (DSP) based on the number of feet in contact with the ground. The walking dynamics of a robot is a highly free and highly nonlinear hybrid dynamics system. A hybrid dynamics system refers to a walking process that includes two continuous processes (SSP and DSP) and a discrete process connecting these two continuous processes. The landing of a swinging foot converts SSP into DSP, and the lifting of a new swinging foot switches DSP back to SSP. Here, the landing and lifting of the foot are discrete events.

[0079] The robot during SSP and DSP is modeled as a multi-rigid-body system in contact with the external environment (ground):

[0080]

[0081]

[0082] Where q represents the generalized coordinate value of the robot, τ represents the torque of the robot, M represents the inertia matrix of the robot, b represents the nonlinear terms of the robot (including the manifestation of Coriolis force, centrifugal force, and gravity in the generalized coordinate space), and f c This represents the contact force between the robot and the external environment, with the subscript 'c' indicating the contact point, and 'J' representing the contact force between the robot and the external environment. c Let be the contact Jacobian matrix at the contact point. Equation (1) represents the continuous multi-rigid-body dynamics relationship, and equation (2) represents the acceleration relationship at the contact point. That is, the acceleration in all directions at the contact point between the foot and the ground is zero, meaning that the foot will not slide relative to the ground after making contact.

[0083] It should be noted that for a robot in SSP (Special Spatial Response) mode, only one foot is in contact with the ground. The contact force exerted by the ground on the robot consists of forces in the x, y, and z directions and three torques about the xyz rotation axis. Therefore, f c It is a 6-dimensional vector. However, when the robot is in DSP mode, two feet are in contact with the ground, so the contact force exerted by the ground on both feet is included, f. c It is a 12-dimensional vector.

[0084] During walking, each foot cycles through a support phase and a swing phase, accompanied by the landing and take-off of the swinging foot. To describe this discrete event, the following assumptions are presented for the establishment and dissolution of contact relationships:

[0085] 1. The contact establishment process (foot landing) is assumed to be a fully plastic impact process, and the robot position is continuous before and after the contact is established, but the velocity has a sudden change.

[0086] 2. The contact release process (foot lifting) is a non-adhesive separation process, meaning that the robot's state does not change abruptly.

[0087] 3. The establishment and contact of contact are completed instantaneously, and are therefore considered discrete events.

[0088] In one possible implementation, determining the discrete dynamics model of the robot with its upper limb fixed includes:

[0089] The discrete dynamics model of the robot with its upper limbs fixed is determined as follows:

[0090]

[0091] Where, the superscript + indicates the moment before the discrete event, the superscript - indicates the moment after the discrete event, the subscript c indicates the contact point between the robot and the ground, q indicates the generalized coordinate value of the robot, M indicates the inertia matrix of the robot, and F... impJ represents the instantaneous collision force corresponding to the robot, δ represents the duration of the discrete event, and δ is generally set to a minimum value. c- J represents the contact Jacobian matrix corresponding to the robot's contact state with the ground before the discrete event. c+ This represents the contact Jacobian matrix corresponding to the robot's contact state with the ground after a discrete event.

[0092] In one possible implementation, determining the discrete dynamics model of the robot with its upper limb fixed includes:

[0093] The discrete dynamics model of the robot with its upper limbs fixed is determined according to the following rules:

[0094] When a landing event occurs, the instantaneous collision force of the robot is greater than or equal to the first threshold and less than or equal to the second threshold.

[0095] When the foot lifts, the robot experiences zero instantaneous collision force.

[0096]

[0097] in, Indicates the first threshold. F represents the second threshold. imp This represents the instantaneous collision force on the robot.

[0098] Taking the discrete event of "landing" as an example: the contact between the swinging foot and the ground is established instantaneously; the generalized coordinate values ​​q are equal before and after landing. + =q - The relationship between the speeds before and after landing satisfies The contact state before landing is that only one foot is in contact with the ground, and the contact Jacobian is J. c- And there is no relative sliding in contact. The contact state after landing is that both feet are in contact with the ground, and the contact Jacobian is J. c+ And there is no relative sliding during contact.

[0099] Based on the above equations (2)(3)(4)(5) related to the hybrid dynamics system, the dynamic model of the robot walking in the multi-contact mode is determined, including the continuous dynamic model and the discrete dynamic model.

[0100] Because robot motion is complex and time-varying, involving continuous processes, discrete transitions, and rigid-flexible coupling, among other factors, this paper proposes hypotheses and extracts key features to prevent the trajectory optimization problem from becoming overly complex. Specifically, the periodic walking process of a robot at a specified speed is defined as including two continuous dynamic processes and two discrete transition events. Reasonable collision assumptions are given, and a hybrid dynamic system for the entire process is established, which better reflects the actual situation of robots.

[0101] In one possible implementation, the first preset constraints include a first equality constraint and a first inequality constraint. The first equality constraint includes robot gait periodicity constraints and / or robot walking speed constraints. The first inequality constraint includes at least one of optimization variable range constraints, robot-ground contact constraints, robot center-of-mass height constraints, and robot foot-swing constraints. Gait periodicity refers to the periodic movement of the robot's generalized state and generalized forces. Optimization variables include time, generalized state, and generalized force. Robot-ground contact constraints include the foot contact force satisfying a friction cone constraint and the constraint that the foot will not flip during contact with the ground. Robot foot-swing constraints include the height constraint of the swinging foot at the midpoint of the swing.

[0102] Step S122: Using the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint as the target constraint, and with the goal of minimizing the sum of squares of the robot's joint torques, the first nonlinear optimization problem is solved to obtain the initial motion trajectory of the robot with its upper limbs fixed. The first nonlinear optimization problem is used to characterize the sum of squares of the joint torques of the robot in one motion cycle with its upper limbs fixed.

[0103] Movement can include jumping, running, walking, dancing, etc. A movement cycle can refer to the time required to complete the smallest cyclical movement in that movement. For example, for walking, the cycle is: single-leg support phase → foot landing → double-leg support phase → foot lifting → single-leg support phase, and so on. The time from one single-leg support phase to the next single-leg support phase is one movement cycle.

[0104] The expression for the first nonlinear optimization problem is:

[0105]

[0106]

[0107] Where t represents time, t∈[t0,t1] f [] represents continuous time, from t0 to t fFor one motion cycle, τ represents the joint torque. The optimization variables are time t, and the trajectory of the robot's generalized state x and generalized force u as they change over time. The objective constraints include two dimensions: a) the robot's walking dynamics model with its upper limbs fixed, i.e., the system dynamic equations. b) The first pre-set constraint condition, including the first equality constraint condition, i.e., the equality constraint C other than the dynamic equation. eq (x(t),u(t),t)=0, and the first inequality constraint, i.e., inequality constraint C. iq (x(t),u(t),t)≤0.

[0108] For the robot's walking process, the generalized state and generalized force are respectively u=[τ,f c ] T The system dynamics equations during SSP and DSP are as follows:

[0109]

[0110] There are also similarities for placing and lifting the foot. For the sake of brevity, the time-related function variable (t) is omitted.

[0111] For the first nonlinear optimization problem, in order to quickly find the optimal solution, all degrees of freedom of the robot's upper limbs can be fixed. For the robot, since each leg has 6 joints, each arm has 7 joints, and the floating basis has 6 degrees of freedom, and because the upper limbs are fixed, the state in the optimization variables does not include the state of the upper limb joints. Therefore, at each time step, the dimension of the state x(t) ∈ R(t). 36×1 Here, 36 includes the velocities of 12 joints in the two lower limbs, the positions of the 12 joints in the two lower limbs, the velocities of 6 floating bases, and the positions of 6 floating bases. The dimension of the joint torque τ(t)∈R(t) 12×1 Among them, 12 includes the joint torques of the 12 joints of the 2 lower limbs.

[0112] Since the sum of squares of joint torques is continuous and smooth, the first nonlinear optimization problem can be solved quickly, thereby quickly obtaining the initial motion trajectory and improving the efficiency of generating the target motion trajectory.

[0113] In one possible implementation, please refer to Figure 5 Step S122 may include:

[0114] Step S1221: Using the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint as the target constraint, and with the goal of minimizing the sum of squared joint torques of the robot, the first nonlinear optimization problem is solved to obtain the first optimal solution, wherein the first optimal solution is used to characterize time, robot generalized state and robot generalized force.

[0115] Step S1222: Fit the first optimal solution using the least squares method to obtain the initial motion trajectory of the robot with its upper limbs fixed.

[0116] Directly solving nonlinear optimization problems is quite time-consuming and difficult. It is usually necessary to convert them into parametric optimization problems through discretization and approximation, and then solve them using mature parametric optimization solvers. The conversion methods can include the direct placement method or the multiple target method, utilizing mature parametric optimization problem solvers, such as the Interior Point OPTimizer (Ipopt), the Sparse Nonlinear OPTimizer (NOPT), and FMINCON Matlab. Solving the first nonlinear optimization problem yields a set of discrete first optimal solutions, i.e., the robot's generalized state x. 1d Robot generalized force u 1d and time t 1d Then, the least squares method is used for fitting, and the optimal trajectory x of the first nonlinear optimization problem is obtained by fitting these discrete points. 1(t) u 1(t) t1 and t1 represent the initial trajectory, where x is the initial trajectory. 1(t) The functional relationship between the robot's generalized state and time, u 1(t) This represents the functional relationship between the robot's generalized force and time, and t1 represents the range of values ​​for time.

[0117] In one possible implementation, step S2 may include:

[0118] Constraints are constructed based on the initial motion trajectory, and the target motion trajectory of the robot is determined when neither the upper nor lower limbs are fixed, with the goal of minimizing the robot's energy consumption.

[0119] The energy consumption of a robot can be considered as the sum of the energy consumed by its joints. For example, the formula for calculating the total joint energy is:

[0120]

[0121] Where τ represents the joint torque, t represents time, and q represents the generalized coordinate value of the robot.

[0122] By constructing constraints based on the initial motion trajectory, the desired robot behavior can be customized, and the feasible region can be restricted to a relatively optimal area. Furthermore, mandatory range constraints can be imposed on key features, such as changes in the height of the center of mass and changes in angular momentum, to ensure that the target motion trajectory does not deviate from the desired optimality.

[0123] In one possible implementation, please refer to Figure 6 Using the initial motion trajectory as a constraint, and with the goal of minimizing the robot's energy consumption, the target motion trajectory of the robot is determined when neither the upper nor lower limbs are fixed. Specifically, this may include:

[0124] Step S21: Determine the walking dynamics model and the second preset constraint condition of the robot when neither the upper nor lower limbs are fixed. The second preset constraint condition is used to constrain the movement of the robot when neither the upper nor lower limbs are fixed.

[0125] Determining the walking dynamics model of a robot when neither its upper nor lower limbs are fixed can include:

[0126] The continuous dynamics model and discrete dynamics model of the robot are determined when neither the upper nor lower limbs are fixed. The continuous dynamics model is used to characterize the multi-rigid-body dynamics and contact point acceleration relationships of the robot when neither the upper nor lower limbs are fixed. The discrete dynamics model is used to characterize the robot's positional relationship, relative sliding relationship with the ground, and velocity relationship before and after discrete events. Discrete events include foot landing events and foot lifting events.

[0127] It should be understood that the steps for determining the walking dynamics model of a robot when neither the upper nor lower limbs are fixed can refer to the steps for determining the walking dynamics model of a robot when the upper limbs are fixed, the difference being that the degrees of freedom of the upper limbs are completely freed.

[0128] In one possible implementation, the second preset constraints include a second equality constraint and a second inequality constraint. The second equality constraint includes at least one of robot gait periodicity constraint, robot walking speed constraint, and robot floating base trajectory constraint. The second inequality constraint includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, robot foot swing constraint, and robot angular velocity constraint. Gait periodicity refers to the periodic movement of the robot's generalized state and generalized forces. The robot floating base trajectory constraint means that the trajectory of the floating base is equal to the trajectory of the floating base in the initial optimized trajectory. Optimization variables include time, generalized state, and generalized force. The optimization variable range constraint also includes the range constraints of the aforementioned optimization variables corresponding to the elbow joint of the upper limb. The robot-ground contact constraint includes the foot contact force satisfying the friction cone constraint and the constraint that the foot will not flip during contact with the ground. The robot foot swing constraint includes the height constraint of the swinging foot at the midpoint of the swing. The robot angular velocity constraint refers to the average angular velocity constraint of the shoulder forward swing joint and the elbow joint.

[0129] By using floating base trajectory constraints, the optimization direction of the robot's lower limb joints will be correctly guided, preventing it from diverging too much. By introducing optimization variable range constraints for the upper limb joints, as well as average angular velocity constraints for the shoulder forward swing joint and elbow joint, the arm can be guided to swing more periodically. Furthermore, the swinging patterns of human upper limbs can be incorporated into this second inequality constraint.

[0130] Step S22: Using the initial motion trajectory, the walking dynamics model of the robot when neither the upper nor lower limbs are fixed, and the second preset constraint as the target constraint, the optimal solution determined based on the initial motion trajectory is used as the initial guess, and the goal is to minimize the robot's energy consumption to solve the second nonlinear optimization problem, thereby obtaining the target motion trajectory. The optimal solution is used to characterize time, the robot's generalized state, and the robot's generalized force. The second nonlinear optimization problem is used to characterize the energy consumption of the robot's motion within one motion cycle when neither the upper nor lower limbs are fixed.

[0131] This can be understood as follows: taking the initial motion trajectory, the walking dynamics model of the robot when neither the upper nor lower limbs are fixed, and the second preset constraint as the target constraint, the second nonlinear optimization problem is solved based on the optimal solution determined by the initial motion trajectory, i.e., the first optimal solution, as the initial guess, and minimizing the robot's energy consumption as the objective, to obtain the second optimal solution. The second optimal solution is then fitted using the least squares method to obtain the target motion trajectory of the robot when neither the upper nor lower limbs are fixed.

[0132] The expression for the second nonlinear optimization problem is:

[0133]

[0134] Where t represents time, t∈[t0,t1] f [] represents continuous time, from t0 to t f For one motion cycle, τ represents the joint torque. The optimization variables are time t, the trajectory of the robot's generalized state x, and the robot's generalized force u as a function of time. The target constraints include three dimensions: a) constraints constructed based on the initial motion trajectory; b) the walking dynamics model of the robot when neither the upper nor lower limbs are fixed; and c) a second set of pre-defined constraints.

[0135] For the second nonlinear optimization problem, since all degrees of freedom of the robot's upper limbs are no longer fixed, the state in the optimization variables includes the state of the upper limb joints. Therefore, at each time step, the dimension of the state x(t) ∈ R(t). 64×1 Among them, 64 includes the velocities of 12 joints in 2 lower limbs, the positions of 12 joints in 2 lower limbs, the velocities of 6 floating bases, the positions of 6 floating bases, the velocities of 14 joints in 2 upper limbs, and the positions of 14 joints in the upper limbs. The dimension of the joint torque τ(t)∈R(t) 26×1 Among them, 26 represents the joint torques of 12 joints in the two lower limbs and 14 joints in the two upper limbs.

[0136] Directly solving nonlinear optimization problems is quite time-consuming and difficult. It is usually necessary to convert them into parametric optimization problems through discretization and approximation, and then solve them using mature parametric optimization solvers. The conversion methods can be the direct placement method or the multiple target method, utilizing mature parametric optimization problem solvers, such as the Interior Point OPTimizer (Ipopt), the Sparse Nonlinear OPTimizer (NOPT), and FMINCON Matlab. Solving the second nonlinear optimization problem yields a set of discrete second optimal solutions, i.e., the robot's generalized state x. 2d Robot generalized force u 2d and time t 2d Then, the least squares method is used for fitting, and the optimal trajectory x of the second nonlinear optimization problem is obtained by fitting these discrete points. 2(t) u 2(t) t2 and t2 represent the target trajectory, where x is the target trajectory. 2(t) The functional relationship between the robot's generalized state and time, u 2(t) This represents the functional relationship between the robot's generalized force and time, and t2 represents the range of values ​​for time.

[0137] It should be understood that other aspects of the second nonlinear optimization problem can be referred to the description of the first nonlinear optimization problem, and will not be repeated here in this embodiment.

[0138] Since the joint energy and formula are absolute value functions and are non-smooth, they are difficult to converge. Furthermore, since the upper and lower limbs are not fixed and have full degrees of freedom, conventional trajectory optimization methods are difficult to solve. Therefore, the first optimal solution is used as the initial guess to promote the convergence of the optimization process of the second nonlinear optimization problem and improve the efficiency of generating the target motion trajectory.

[0139] Figure 7 This is a block diagram illustrating a robot motion trajectory generation device 400 according to an exemplary embodiment. The robot motion trajectory generation device 400 includes an initial trajectory determination module 401 and a target trajectory determination module 402.

[0140] The initial trajectory determination module 401 is configured to determine the initial motion trajectory of the robot when its upper limbs are fixed.

[0141] The target trajectory determination module 402 is configured to determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, based on the initial motion trajectory, so that the robot moves according to the target motion trajectory.

[0142] Optionally, the initial trajectory determination module 401 includes an upper limb fixation module and a first determination module:

[0143] The upper limb fixation module is configured to determine the robot's upper limb degrees of freedom to a preset fixed angle or to define the robot's upper limb and torso as a rigid body.

[0144] The first determining module is configured to determine the initial motion trajectory of the robot with its upper limbs fixed, with the objective of minimizing the sum of squares of the robot's joint torques.

[0145] Optionally, the first determining module includes a first constraint determining module and an initial trajectory solving module:

[0146] The first constraint determination module is configured to determine the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint conditions. The first preset constraint conditions are used to constrain the robot's movement when its upper limbs are fixed.

[0147] The initial trajectory solving module is configured to use the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint as the target constraint, and to solve the first nonlinear optimization problem with the goal of minimizing the sum of squares of the robot's joint torques, so as to obtain the initial motion trajectory of the robot with its upper limbs fixed. The first nonlinear optimization problem is used to characterize the sum of squares of the joint torques of the robot in one motion cycle with its upper limbs fixed.

[0148] Optionally, the first preset constraint condition includes a first equality constraint condition and a first inequality constraint condition, wherein the first equality constraint condition includes robot gait periodicity constraint and / or robot walking speed constraint, and the first inequality constraint condition includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, and robot foot swing constraint.

[0149] Optionally, the first constraint determination module can be specifically configured as follows:

[0150] The continuous dynamics model and discrete dynamics model of the robot with its upper limbs fixed are determined. The continuous dynamics model is used to characterize the multi-rigid-body dynamics and contact point acceleration relationships of the robot with its upper limbs fixed. The discrete dynamics model is used to characterize the robot's positional relationship, relative sliding relationship with the ground, and velocity relationship before and after discrete events. Discrete events include foot landing events and foot lifting events.

[0151] Optionally, the first constraint determination module can be configured as follows:

[0152] The discrete dynamics model of the robot with its upper limbs fixed is determined as follows:

[0153]

[0154] Where, the superscript + indicates the moment before the discrete event, the superscript - indicates the moment after the discrete event, the subscript c indicates the contact point between the robot and the ground, q indicates the generalized coordinate value of the robot, M indicates the inertia matrix of the robot, and F... imp J represents the instantaneous collision force corresponding to the robot, δ represents the duration of the discrete event, and J represents the instantaneous collision force corresponding to the robot. c- J represents the contact Jacobian matrix corresponding to the robot's contact state with the ground before the discrete event. c+ This represents the contact Jacobian matrix corresponding to the robot's contact state with the ground after a discrete event.

[0155] Optionally, the first constraint determination module can be configured as follows:

[0156] The discrete dynamics model of the robot with its upper limbs fixed is determined according to the following rules:

[0157] When a landing event occurs, the instantaneous collision force of the robot is greater than or equal to the first threshold and less than or equal to the second threshold.

[0158] When the foot lifts, the robot experiences zero instantaneous collision force.

[0159] Optionally, the target trajectory determination module 402 is specifically configured as follows:

[0160] Constraints are constructed based on the initial motion trajectory, and the target motion trajectory of the robot is determined when neither the upper nor lower limbs are fixed, with the goal of minimizing the robot's energy consumption.

[0161] Optionally, the target trajectory determination module 402 includes a second constraint determination module and a target trajectory solving module:

[0162] The second constraint determination module is configured to determine the walking dynamics model of the robot when neither the upper limb nor the lower limb is fixed and the second preset constraint condition, wherein the second preset constraint condition is used to constrain the movement of the robot when neither the upper limb nor the lower limb is fixed.

[0163] The target trajectory solving module is configured to use the initial motion trajectory, the walking dynamics model of the robot when neither the upper nor lower limbs are fixed, and the second preset constraint as the target constraint. Based on the optimal solution determined by the initial motion trajectory as the initial guess, and minimizing the robot's energy consumption as the objective, the module solves the second nonlinear optimization problem to obtain the target motion trajectory. The optimal solution is used to characterize time, the robot's generalized state, and the robot's generalized force. The second nonlinear optimization problem is used to characterize the energy consumption of the robot's motion within one motion cycle when neither the upper nor lower limbs are fixed.

[0164] Optionally, the second preset constraint includes a second equality constraint and a second inequality constraint, wherein the second equality constraint includes at least one of robot gait periodicity constraint, robot walking speed constraint and robot floating base trajectory constraint, and the second inequality constraint includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, robot foot swing constraint and robot angular velocity constraint.

[0165] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0166] Figure 8 This is a block diagram of a robot 800 for a robot motion trajectory generation method according to an exemplary embodiment.

[0167] Reference Figure 8 The robot 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input / output interface 812, a sensor component 814, and a communication component 816.

[0168] Processing component 802 typically controls the overall operation of robot 800, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the robot motion trajectory generation method described above. Furthermore, processing component 802 may include one or more modules to facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.

[0169] Memory 804 is configured to store various types of data to support the operation of robot 800. Examples of this data include instructions for any application or method used to operate on robot 800, contact data, phone book data, messages, pictures, videos, etc. Memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0170] The power supply assembly 806 provides power to the various components of the robot 800. The power supply assembly 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the robot 800.

[0171] The multimedia component 808 includes a screen that provides an output interface between the robot 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component 808 includes a front-facing camera and / or a rear-facing camera. When the robot 800 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0172] Audio component 810 is configured to output and / or input audio signals. For example, audio component 810 includes a microphone (MIC) configured to receive external audio signals when robot 800 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 804 or transmitted via communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

[0173] Input / output interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0174] Sensor assembly 814 includes one or more sensors for providing status assessments of various aspects of robot 800. For example, sensor assembly 814 can detect the on / off state of robot 800, the relative positioning of components such as the robot 800's display and keypad, changes in the position of robot 800 or one of its components, the presence or absence of user contact with robot 800, robot 800 orientation or acceleration / deceleration, and temperature changes of robot 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 814 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0175] Communication component 816 is configured to facilitate wired or wireless communication between robot 800 and other devices. Robot 800 can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 816 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 816 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0176] In an exemplary embodiment, robot 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the robot motion trajectory generation method described above.

[0177] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, which can be executed by the processor 820 of the robot 800 to complete the robot motion trajectory generation method described above. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0178] The aforementioned device can be a standalone robot or a part of a standalone robot. For example, in one embodiment, the robot can be an integrated circuit (IC) or a chip, where the integrated circuit can be a single IC or a collection of multiple ICs. The chip can include, but is not limited to, the following types: GPU (Graphics Processing Unit), CPU (Central Processing Unit), FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), and SoC (System on Chip). The aforementioned integrated circuit or chip can be used to execute executable instructions (or code) to implement the robot motion trajectory generation method described above. These executable instructions can be stored in the integrated circuit or chip or obtained from other devices or equipment. For example, the integrated circuit or chip may include a processor, memory, and an interface for communicating with other devices. The executable instructions can be stored in the memory, and when the executable instructions are executed by the processor, the above-mentioned robot motion trajectory generation method is implemented; or, the integrated circuit or chip can receive the executable instructions through the interface and transmit them to the processor for execution to implement the above-mentioned robot motion trajectory generation method.

[0179] In another exemplary embodiment, a computer program product is also provided, the computer program product comprising a computer program executable by a programmable device, the computer program having a code portion for performing the robot motion trajectory generation method described above when executed by the programmable device.

[0180] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of this disclosure. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0181] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for generating robot motion trajectories, characterized in that, The method includes: Determine the robot's initial motion trajectory with its upper limbs fixed; Constraints are constructed based on the initial motion trajectory. The optimal solution determined based on the initial motion trajectory is used as an initial guess. With the goal of minimizing the robot's energy consumption, the target motion trajectory of the robot is determined when neither the upper nor lower limbs are fixed, so that the robot moves according to the target motion trajectory. The optimal solution is used to characterize time, robot generalized state, and robot generalized force.

2. The method according to claim 1, characterized in that, Determining the robot's initial motion trajectory with its upper limbs fixed includes: The robot's upper limb degrees of freedom are set to a preset fixed angle, or the robot's upper limbs and torso are defined as a rigid body. The initial motion trajectory of the robot with its upper limbs fixed is determined with the goal of minimizing the sum of squares of the joint torques.

3. The method according to claim 2, characterized in that, Determining the initial motion trajectory of the robot with its upper limb fixed, with the objective of minimizing the sum of squared joint moments, includes: Determine the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint condition, wherein the first preset constraint condition is used to constrain the movement of the robot with its upper limbs fixed; Using the walking dynamics model of the robot with its upper limbs fixed and the first preset constraint as the target constraint, and with the goal of minimizing the sum of squares of the joint torques of the robot, the first nonlinear optimization problem is solved to obtain the initial motion trajectory of the robot with its upper limbs fixed. The first nonlinear optimization problem is used to characterize the sum of squares of the joint torques of the robot within one motion cycle with its upper limbs fixed.

4. The method according to claim 3, characterized in that, The first preset constraint conditions include a first equality constraint condition and a first inequality constraint condition. The first equality constraint condition includes robot gait periodicity constraint and / or robot walking speed constraint. The first inequality constraint condition includes at least one of optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, and robot foot swing constraint.

5. The method according to claim 3, characterized in that, Determining the walking dynamics model of the robot with its upper limbs fixed includes: Determine the discrete and continuous dynamic models of the robot when its upper limbs are fixed. The continuous dynamic model is used to characterize the multi-rigid-body dynamics and contact-point acceleration relationships of the robot when its upper limbs are fixed. The discrete dynamic model is used to characterize the robot's positional relationship, relative sliding relationship with the ground, and velocity relationship before and after a discrete event. The discrete events include foot-landing events and foot-lifting events.

6. The method according to claim 5, characterized in that, Determining the discrete dynamic model of the robot with its upper limb fixed includes: The discrete dynamic model of the robot with its upper limbs fixed is determined according to the following rules: When the landing event occurs, the instantaneous collision force of the robot is greater than or equal to the first threshold and less than or equal to the second threshold. When the foot-lifting event occurs, the instantaneous collision force corresponding to the robot is zero.

7. The method according to claim 1, characterized in that, The process of constructing constraints based on the initial motion trajectory, using the optimal solution determined by the initial motion trajectory as an initial guess, and aiming to minimize the robot's energy consumption, determines the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, including: A walking dynamics model and a second preset constraint condition are determined for the robot when its upper and lower limbs are not fixed. The second preset constraint condition is used to constrain the movement of the robot when its upper and lower limbs are not fixed. Using the initial motion trajectory, the walking dynamics model of the robot when neither the upper nor lower limbs are fixed, and the second preset constraint as the target constraint, the optimal solution determined based on the initial motion trajectory is used as the initial guess, and the goal is to minimize the energy consumption of the robot, the second nonlinear optimization problem is solved to obtain the target motion trajectory. The second nonlinear optimization problem is used to characterize the energy consumption of the robot's motion within one motion cycle when neither the upper nor lower limbs are fixed.

8. The method according to claim 7, characterized in that, The second preset constraint includes a second equality constraint and a second inequality constraint. The second equality constraint includes at least one of the following: robot gait periodicity constraint, robot walking speed constraint, and robot floating base trajectory constraint. The second inequality constraint includes at least one of the following: optimization variable range constraint, robot-ground contact constraint, robot center of mass height constraint, robot foot swing constraint, and robot angular velocity constraint.

9. A robot motion trajectory generation device, characterized in that, The device includes: The initial trajectory determination module is configured to determine the robot's initial motion trajectory with its upper limbs fixed. The target trajectory determination module is configured to construct constraints based on the initial motion trajectory, use the optimal solution determined by the initial motion trajectory as an initial guess, and determine the target motion trajectory of the robot when neither the upper nor lower limbs are fixed, with the goal of minimizing the robot's energy consumption, so that the robot moves according to the target motion trajectory. The optimal solution is used to characterize time, robot generalized state, and robot generalized force.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When executed by the first processor, the program instructions implement the steps of the method according to any one of claims 1 to 8.

11. A robot, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured as follows: The steps of performing the method according to any one of claims 1 to 8.

Images