A bipedal robot gait planning method and device for stepped terrain

Abstract

The present invention provides a method and device for gait planning of a bipedal robot for stepped terrain, belonging to the technical field of robot gait planning. The method comprises: establishing walking modes for ascending and descending stairs of the bipedal robot based on a linear inverted pendulum model; mapping the linear inverted pendulum model for stepped terrain into a segmented virtual height linear inverted pendulum model for planar terrain based on the walking modes; and utilizing model predictive control based on the segmented virtual height linear inverted pendulum model to implement gait planning for the bipedal robot for stepped terrain. The method and device for gait planning of a bipedal robot for stepped terrain provided by the present invention utilize a linear inverted pendulum model with a virtual center of mass height. The gait mode established based on terrain elevation information and foothold information of the stepped environment can adapt to irregular stepped terrain with variable heights and lengths, thus providing the gait with good terrain adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of robot gait planning, and in particular to a gait planning method and device for a bipedal robot for stepped terrain. Background Art

[0002] Because bipedal robots have multiple degrees of freedom and frequent mechanical interactions with the physical environment during locomotion, the mathematical models of their walking process are highly nonlinear and strongly coupled, making them difficult to control. By drawing on the characteristics of human walking, abstracting bipedal robots into simplified physical models and mathematically describing their walking patterns, we can effectively simplify the kinematic and dynamic models of bipedal robots, facilitate motion planning, and reduce control difficulties.

[0003] The Linear Inverted Pendulum Model (LIPM) concentrates the entire mass of a bipedal robot at a single center of mass, which interacts with the ground via an actively retractable massless pendulum. Its characteristic is that the center of mass remains constant, thus linearizing the center of mass dynamics during gait. The stability of a bipedal robot's gait can be measured by its zero moment point (ZMP). When this point lies within the robot's support domain, the robot's gait is considered stable. Based on the center of mass dynamics of the LIPM, a linear model predictive control (MPC) is introduced. Using real-time rolling optimization, the robot's zero moment point is constrained within the support domain, enabling online planning of the bipedal robot's gait.

[0004] Traditional predictive control methods based on the linear inverted pendulum model only implement gait planning for bipedal robots on flat terrain and cannot be further applied to stepped terrain in three-dimensional environments. However, real-world environments are often characterized by uneven steps, so traditional predictive control methods based on the linear inverted pendulum model limit the application of bipedal robots in real-world environments. Summary of the Invention

[0005] The present invention provides a gait planning method and device for a bipedal robot for stepped terrain, which are used to solve the defect that the traditional predictive control method based on a linear inverted pendulum model limits the application of bipedal robots in practical environments.

[0006] In the first aspect, the present invention provides a gait planning method for a bipedal robot on stepped terrain, comprising: establishing walking modes for ascending and descending stairs of the bipedal robot based on a linear inverted pendulum model; mapping the linear inverted pendulum model under the stepped terrain into a segmented virtual height linear inverted pendulum model under the planar terrain based on the walking mode; and utilizing model predictive control to realize gait planning of the bipedal robot under stepped terrain based on the segmented virtual height linear inverted pendulum model.

[0007] In a second aspect, the present invention further provides a bipedal robot gait planning device for stepped terrain, comprising:

[0008] The first processing module is used to establish the walking modes of the biped robot for ascending stairs and descending stairs respectively based on the linear inverted pendulum model;

[0009] The second processing module is configured to map the linear inverted pendulum model under the stepped terrain into a segmented virtual height linear inverted pendulum model under the flat terrain based on the walking pattern;

[0010] The third processing module is used to realize gait planning of the biped robot on stepped terrain by using model predictive control based on a segmented virtual height linear inverted pendulum model.

[0011] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the bipedal robot gait planning method for stepped terrain as described in any one of the above are implemented.

[0012] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-described bipedal robot gait planning methods for stepped terrain.

[0013] In a fifth aspect, the present invention also provides a computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of any of the above-mentioned bipedal robot gait planning methods for stepped terrain are implemented.

[0014] The biped robot gait planning method and device for stepped terrain provided by the present invention have the following advantages compared with the prior art:

[0015] (1) The present invention uses a linear inverted pendulum model with a virtual center of mass height. The gait pattern established based on the terrain elevation information and foothold information of the stair environment can adapt to irregular stair terrain with variable height and length, making the gait have better terrain adaptability.

[0016] (2) The linear inverted pendulum model with virtual center of mass height proposed in this invention mainly realizes the gait planning of the biped robot walking forward in a straight line; combined with omnidirectional foothold planning and coordinate transformation matrix, the method proposed in this invention has the potential to realize omnidirectional walking on three-dimensional stepped terrain.

[0017] (3) In the proposed gait mode, the present invention adopts model predictive control to plan the gait trajectory of the biped robot in the forward and vertical directions through an optimization problem, avoiding the kinematic constraints caused by optimizing and solving the two directions separately, and can shorten the solution time, which is conducive to the real-time performance of gait planning.

[0018] (4) The step terrain gait planning method proposed in this invention constrains the robot's zero torque point to the support domain through secondary optimization solution according to the actual state of the robot, ensuring the stability of the bipedal robot's walking state and having a certain resistance to external force interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 is a schematic diagram of a walking pattern during a stair climbing process provided by the present invention;

[0021] Figure 2 is a schematic diagram of the walking mode during the stair descent process provided by the present invention;

[0022] Figure 3 is a schematic diagram of a virtual height linear inverted pendulum model provided by the present invention;

[0023] Figure 4 Schematic diagram of a segmented virtual height linear inverted pendulum walking mode provided by the present invention;

[0024] Figure 5 Schematic diagram of the structure of the biped robot gait planning device for stepped terrain provided by the present invention;

[0025] Figure 6 1 is a schematic diagram of a center of mass state curve on a sagittal plane when the biped robot of the present invention moves on a stepped terrain;

[0026] Figure 7 Schematic diagram of the center of mass state curve of the coronal plane of the bipedal robot provided by the present invention when moving on a stepped terrain;

[0027] Figure 8This is a simulation frame diagram of a biped robot walking in a stair environment using the gait planning method proposed in the present invention;

[0028] Figure 9 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0029] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0030] It should be noted that, in the description of the embodiments of the present invention, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "include a ..." do not exclude the presence of other identical elements in the process, method, article or device comprising the elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0031] The terms "first," "second," and the like in this application are used to distinguish similar objects, and are not used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, the objects distinguished by "first," "second," and the like generally refer to a class of objects and do not limit the number of objects. For example, the first object may be one or more.

[0032] The following combination Figures 1-9 The present invention describes a biped robot gait planning method and device for stepped terrain provided by an embodiment of the present invention.

[0033] The gait planning method for a bipedal robot on stepped terrain provided by the present invention includes but is not limited to the following steps:

[0034] Step 101: Based on the linear inverted pendulum model, establish the walking modes of the biped robot for ascending stairs and descending stairs respectively.

[0035] In the coronal plane, inspired by the human body's process of climbing and descending stairs, the present invention proposes a walking mode of a bipedal robot in a climbing and descending stair environment based on a linear inverted pendulum model.

[0036] When climbing stairs, the biped robot first uses the foot on the previous step as the supporting foot and the other foot as the swinging foot. After stepping onto the next step with the swinging foot, the foot on the next step becomes the supporting foot again, and the center of mass of the support rises to the height between the two steps. The supporting foot on the previous step becomes the swinging foot and swings to the next step, and the cycle repeats. Therefore, based on the linear inverted pendulum model, the walking pattern in the process of climbing stairs is proposed as follows: Figure 1 As shown, Figure 1 Schematic diagram of the walking mode of the stair climbing process provided by the present invention:

[0037] The dark circle represents the center of mass, the triangle represents the support foot, and H i is the height difference between the current support foot step and the previous step, Θ i is the angle between the line connecting the current foothold and the previous foothold and the horizontal plane, G is the total gravity of the robot, L i is the horizontal distance between the current foothold and the previous foothold, and h is the height of the center of mass on a flat terrain. In this walking mode, when the supporting foot is on the i-th step and the swing leg steps towards the i+1-th step, the center of mass rises by a height of H during this process. i , move forward L i , that is, the angle between the center of mass trajectory of the linear inverted pendulum and the horizontal plane in this stage is Θ i , ensuring that the height difference between the center of mass and the supporting foot is equal to h before each switching of the supporting foot.

[0038] The process of descending stairs is exactly the opposite of the process of ascending stairs. The biped robot first lowers its center of mass to the height difference between the two steps, then uses the foot on the first step as the supporting foot and the other foot as the swinging foot. After stepping down the next step with the swinging foot, the foot on the next step is used as the supporting foot again, and the cycle repeats. Therefore, based on the linear inverted pendulum model, the walking pattern of the stair descent process is proposed as follows: Figure 2 As shown, Figure 2 Schematic diagram of the walking mode during the stair descent process provided by the present invention:

[0039] In this walking mode, when the supporting foot is on the i-th step, the center of mass drops by a height H every time the swing leg steps up the i+1-th step. i+1 , move forward L i+1 , that is, the angle between the center of mass trajectory of the linear inverted pendulum and the horizontal plane in this stage is Θ i+1 , ensuring that the height difference between the center of mass and the supporting foot is equal to h at the moment after the supporting foot is switched.

[0040] In this walking pattern of going up and down stairs, the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage form a parallelogram, which only depends on the adjacent step height H and the stride length L.

[0041] Step 102: Based on the walking pattern, the linear inverted pendulum model under the stepped terrain is mapped into a segmented virtual height linear inverted pendulum model under the flat terrain.

[0042] Based on the aforementioned walking pattern, model predictive control is required to generate a smooth center of mass trajectory to ensure the stability of the bipedal robot's motion. On stepped terrain, a linear inverted pendulum has center of mass displacements along the horizontal (x-axis) and vertical (y-axis) axes (denoted by x and y, respectively). The center of mass height is constantly changing, and the mathematical model of the linear inverted pendulum is also constantly changing, making it difficult to directly generate a smooth center of mass trajectory using such a model.

[0043] Taking the above steps as an example, the present invention proposes to rotate the parallelogram formed by the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage by an angle θ i , that is, to perform a coordinate rotation transformation on the linear inverted pendulum model at this stage, and transform the gravity component GcosΘ i Considered as equivalent gravity, GsinΘ i Considered as an external force that pushes the center of mass of the linear inverted pendulum in the horizontal opposite direction, Figure 3 As shown, Figure 3 Schematic diagram of a virtual height linear inverted pendulum model provided by the present invention.

[0044] At this time, the linear inverted pendulum can be regarded as a linear inverted pendulum with a constant center of mass height. The center of mass height is called the virtual height. :

[0045] ;

[0046] in, It is the horizontal distance between the center of mass and the supporting foot just before the supporting foot is switched.

[0047] In the horizontal external force GsinΘ i The dynamic equation of the linear inverted pendulum model under push is:

[0048] ;

[0049] in, is the equivalent gravitational acceleration after coordinate rotation transformation, m is the mass center of the linear inverted pendulum, τ is the torque at the support foot, xis the displacement of the center of mass in the horizontal direction. At this time, the zero moment point position of the linear inverted pendulum model for:

[0050] ;

[0051] Similarly, when going down the stairs, rotate the parallelogram formed by the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage by an angle Θ i+1 , then the virtual height and zero moment point position of the linear inverted pendulum model are:

[0052] ;

[0053] ;

[0054] By performing such a coordinate rotation transformation in each step phase, the variable-height linear inverted pendulum model under the stepped terrain can be mapped into a segmented virtual-height linear inverted pendulum model under the flat terrain. Figure 4 Schematic diagram of the segmented virtual height linear inverted pendulum walking mode provided by the present invention, such as Figure 4 Since the virtual height of the center of mass of each segment remains unchanged, model predictive control can be used to plan the center of mass trajectory.

[0055] Step 103: Based on the segmented virtual height linear inverted pendulum model, model predictive control is used to realize gait planning of the biped robot on stepped terrain.

[0056] The present invention proposes a walking mode under stepped terrain, which Figure 4 Equivalent segmented virtual height linear inverted pendulum model. Next, based on this model, model predictive control is used to realize the robot's gait planning.

[0057] Assuming the minimum time interval is δ, the robot's center of mass state at the kth moment (center of mass displacement, velocity, acceleration, and acceleration derivative in the horizontal axis direction) is recorded as:

[0058] 、 ;

[0059] The position of the zero moment point is recorded as , the discretized state transfer equation is:

[0060] ;

[0061] ;

[0062] in, 、 They represent the robot's position at the kth moment. iThe virtual height and equivalent gravitational acceleration when taking the first step. The prediction window length of the model predictive control rolling optimization is N (i.e., the prediction window length for the next N steps is N). δ Time is used for predictive control), and the recursive relationship of the state transfer equation can be used to obtain the relationship between the ZMP position and the center of mass state in the prediction window:

[0063] ;

[0064] in, , , , , is a Toeplitz matrix with N rows and N columns.

[0065] Establish a quadratic optimization problem, and the cost function is:

[0066] ;

[0067] ;

[0068] Among them, Q and R are constants used to make a trade-off between ZMP tracking error and the oscillation of the center of mass acceleration derivative. is the lower limit of the support domain, is the upper limit of the support domain.

[0069] When the height of the terrain and the position sequence of the foothold Known, the reference position of ZMP Known, we can find (in 、 represents the horizontal distance between the current foothold and the previous foothold when the robot is on the i-th step at the k-th moment, as well as the angle between the line connecting the two points and the horizontal plane). Take this sequence Substitute the first element in the discretized state transfer equation to obtain the optimal center of mass state at the next moment The center of mass state is Figure 4 The center of mass state of the segmented virtual height linear inverted pendulum in the coordinate system is essentially the vector sum of the center of mass states in the x and y directions under the stepped terrain. Then decompose it along the x and y axes of the world coordinate system of the stepped terrain to obtain Figure 1 The center of mass state of the stepped terrain in the world coordinate system is shown:

[0070] ;

[0071] Thus, this invention has completed the center of mass trajectory planning (x- and y-axis directions) in the sagittal plane based on the virtual height linear inverted pendulum model predictive control. In the coronal plane, the center of mass trajectory in the z-axis direction can be directly calculated using traditional model predictive control, which will not be repeated here.

[0072] After completing the center of mass trajectory planning, in order to achieve complete gait planning, the third-order and fifth-order Bezier curves are used in the robot's local coordinate system to plan the swing leg's trajectory in the x-axis and y-axis directions, respectively. Specifically:

[0073] ;

[0074] in, t is the normalized time coefficient, n is the order of the Bezier curve, A is the control point of the Bezier curve. To avoid impact with the ground, the position and velocity of the foot at the initial and final moments are constrained. The preset Bezier curve constraint parameters are shown in Table 1:

[0075] Table 1 Bezier curve constraint parameter table

[0076]

[0077] To better understand the control method of the embodiment of the present invention, a specific biped robot gait planning example is provided below for illustration.

[0078] Using the model and gait planning method provided by the present invention, gait planning is performed for a bipedal robot with a total mass of 90 kg, a thigh length and mass of 0.39 m and 6.7 kg, a calf length and mass of 0.34 m and 5.8 kg, a trunk mass of 65 kg, and a foot length, width, and thickness of 0.3 m, 0.2 m, and 0.1 m, respectively. The center of mass height when standing still on a flat terrain is , minimum time interval ,A simulation experiment is conducted in Webots to obtain a stable gait under stepped terrain.

[0079] Figure 5 This figure shows the center of mass height curve of the bipedal robot in a stairway environment. The blue solid line represents the height curve of the stairway, the black dashed line represents the theoretically optimal center of mass height curve obtained through optimization, and the red solid line represents the actual center of mass height curve of the robot in a simulation experiment. It can be seen that in an environment with varying stair length and height, the robot's center of mass height can be rationally planned based on the terrain height, demonstrating good stair-crossing capability.

[0080] Figure 6is a schematic diagram of the center of mass state curve on the sagittal plane when the biped robot of the present invention moves on a stepped terrain, wherein Z max , Z min , Z real , Z ref The upper and lower limits of the support region, the robot's actual ZMP position, and the reference ZMP position are shown, respectively. Between ascending and descending stairs, the robot takes a 20-second step to switch from the ascending gait to the descending gait. It can be seen that during the stair crossing process, the robot's actual ZMP position fluctuates slightly between the ZMP reference positions and remains within the support region. The forward trajectory of the center of mass is smooth, demonstrating the excellent stability of the gait planning method proposed in this invention in the sagittal plane.

[0081] Figure 7 is a schematic diagram of the center of mass state curve of the coronal plane when the biped robot provided by the present invention moves on a stepped terrain, wherein Z max , Z min , Z real , Z ref The upper and lower limits of the support region, the robot's actual ZMP position, and the reference ZMP position are shown in Figure 1. It can be seen that while the robot's actual ZMP position fluctuates somewhat during the stair crossing process, it remains near the ZMP reference position and never exceeds the support region. The lateral trajectory of the center of mass is smooth, demonstrating that the proposed gait planning method has excellent stability in the sagittal plane.

[0082] Figure 8 This is a simulation frame diagram of a biped robot walking in a stair environment using the gait planning method proposed in the present invention.

[0083] On the other hand, the present invention also provides a bipedal robot gait planning device for stepped terrain, the device comprising:

[0084] The first processing module is used to establish the walking modes of the biped robot for ascending stairs and descending stairs respectively based on the linear inverted pendulum model;

[0085] The second processing module is configured to map the linear inverted pendulum model under the stepped terrain into a segmented virtual height linear inverted pendulum model under the flat terrain based on the walking pattern;

[0086] The third processing module is used to realize gait planning of the biped robot on stepped terrain by using model predictive control based on a segmented virtual height linear inverted pendulum model.

[0087] It should be noted that the bipedal robot gait planning device for stepped terrain provided in an embodiment of the present invention can execute the bipedal robot gait planning method for stepped terrain described in any of the above embodiments during specific operation, which will not be elaborated in this embodiment.

[0088] In summary, the bipedal robot gait planning method and device for stepped terrain provided by the present invention have the following advantages compared with the prior art:

[0089] (1) The present invention uses a linear inverted pendulum model with a virtual center of mass height. The gait pattern established based on the terrain elevation information and foothold information of the stair environment can adapt to irregular stair terrain with variable height and length, making the gait have better terrain adaptability.

[0090] (2) The linear inverted pendulum model with virtual center of mass height proposed in this invention mainly realizes the gait planning of the biped robot walking forward in a straight line; combined with omnidirectional foothold planning and coordinate transformation matrix, the method proposed in this invention has the potential to realize omnidirectional walking on three-dimensional stepped terrain.

[0091] (3) In the proposed gait mode, the present invention adopts model predictive control to plan the gait trajectory of the biped robot in the forward and vertical directions through an optimization problem, avoiding the kinematic constraints caused by optimizing and solving the two directions separately, and can shorten the solution time, which is conducive to the real-time performance of gait planning.

[0092] (4) The step terrain gait planning method proposed in this invention constrains the robot's zero torque point to the support domain through secondary optimization solution according to the actual state of the robot, ensuring the stability of the bipedal robot's walking state and having a certain resistance to external force interference.

[0093] Figure 9 Schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 9 As shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940, wherein the processor 910, the communication interface 920, and the memory 930 communicate with each other via the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the bipedal robot gait planning method for stepped terrain.

[0094] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the bipedal robot gait planning method for stepped terrain provided in the above-mentioned embodiments.

[0095] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the bipedal robot gait planning method for stepped terrain provided in the above-mentioned embodiments.

[0096] Through the description of the above embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A gait planning method for a bipedal robot on stepped terrain, characterized in that: include: Based on the linear inverted pendulum model, the walking modes of the biped robot for ascending stairs and descending stairs are established respectively; Based on the walking pattern, the linear inverted pendulum model under the stepped terrain is mapped into a segmented virtual height linear inverted pendulum model under the flat terrain; Based on a segmented virtual height linear inverted pendulum model, model predictive control is used to realize gait planning of a bipedal robot on stepped terrain. Based on the linear inverted pendulum model, the walking modes of the biped robot for climbing and descending stairs are established respectively, including: The walking pattern for stair climbing is as follows: During the stair climbing process, the bipedal robot first uses the foot on the previous step as the supporting foot, and the other foot as the swinging foot. After stepping onto the next step with the swinging foot, the foot on the next step becomes the supporting foot again, and the center of mass of the support rises to the height between the two steps. The supporting foot on the previous step becomes the swinging foot, and swings to the next step, and the cycle repeats. The walking pattern for stair descent is as follows: During the stair descent, the bipedal robot first lowers its center of mass to the height difference between the two steps. Then, it uses the foot on the first step as its support foot and the other foot as its swing foot. After stepping down the next step with its swing foot, it uses the foot on the next step as its support foot again, and repeats this cycle. Also includes: Let H i is the height difference between the current support foot step and the previous step, Θ i is the angle between the line connecting the current foothold and the previous foothold and the horizontal plane, G is the total gravity of the robot, L i is the horizontal distance between the current foothold and the previous foothold, and h is the center of mass height under flat terrain; In the stair climbing walking mode, when the supporting foot is on the i-th step and the swing leg steps towards the i+1-th step, the center of mass rises by a height H during this process. i , move forward L i , the angle between the center of mass trajectory of the linear inverted pendulum model and the horizontal plane is Θ i , ensuring that the height difference between the center of mass and the supporting foot is equal to h before each switching of the supporting foot; In the stair-down walking mode, when the supporting foot is on the i-th step and the swing leg steps onto the i+1-th step, the center of mass drops by a height H during this process. i+1 , move forward L i+1 , the angle between the center of mass trajectory of the linear inverted pendulum model and the horizontal plane is Θ i+1 , ensuring that the height difference between the center of mass and the supporting foot is equal to h at the moment after the supporting foot is switched; In the walking mode of going up and down stairs, the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage form a parallelogram; Based on the walking pattern, the linear inverted pendulum model under the stepped terrain is mapped to a segmented virtual height linear inverted pendulum model under the flat terrain, including: In the walking mode of climbing stairs, the parallelogram formed by the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage is rotated by an angle θ i , the gravity component GcosΘ i Considered as equivalent gravity, GsinΘ i Considered as an external force that pushes the center of mass of the linear inverted pendulum in the horizontal reverse direction, the linear inverted pendulum can be regarded as a linear inverted pendulum with a constant center of mass height. The center of mass height is called the virtual height. : ； in, is the horizontal distance between the center of mass and the supporting foot just before the supporting foot is switched; In the horizontal external force GsinΘ i The dynamic equation of the linear inverted pendulum model under push is: ； in, is the equivalent gravitational acceleration after coordinate rotation transformation, m is the mass center of the linear inverted pendulum, τ is the torque at the support foot, x is the displacement of the center of mass in the horizontal direction; the zero moment point position of the linear inverted pendulum model for: ； In the walking mode of going down stairs, the parallelogram formed by the line connecting the front and rear support legs of each step and the displacement trajectory of the center of mass at the current stage is rotated by an angle θ i+1 , then the virtual height and zero moment point position of the linear inverted pendulum model are: ； ； The linear inverted pendulum model of each step stage is subjected to coordinate rotation transformation, so that the variable height linear inverted pendulum model under the stepped terrain is mapped into a segmented virtual height linear inverted pendulum model under the flat terrain.

2. The gait planning method for a bipedal robot on stepped terrain according to claim 1, characterized in that: Based on a segmented virtual height linear inverted pendulum model, model predictive control is used to implement gait planning for a bipedal robot on stepped terrain, including: Assuming the minimum time interval is δ, the center of mass state of the robot at the kth moment is recorded as 、 ; The position of the zero moment point is recorded as , the discretized state transfer equation is: ； ； in, 、 They represent the robot's position at the kth moment. i The virtual height and equivalent gravitational acceleration when taking the steps; the prediction window length of the model predictive control rolling optimization is N. The recursive relationship of the state transfer equation is used to obtain the relationship between the ZMP position and the center of mass state in the prediction window: in, , , , , is a Toeplitz matrix with N rows and N columns; Establish a quadratic optimization problem, and the cost function is: ； ； Where Q and R are constants, is the lower limit of the support domain, is the upper limit of the support domain; When the height of the terrain and the position sequence of the foothold Known, the reference position of ZMP Known, sought ;in 、 It represents the horizontal distance between the current foothold and the previous foothold when the robot is on the i-th step at the k-th moment, as well as the angle between the line connecting the two points and the horizontal plane; take the sequence Substitute the first element in the discretized state transfer equation to obtain the optimal center of mass state at the next moment ; Center of mass state Decompose the horizontal and vertical axes to obtain the center of mass state of the stepped terrain in the world coordinate system; After completing the planning of the center of mass trajectory, Bezier curves are used to plan the trajectories of the swing leg in the horizontal and vertical directions in the local coordinate system of the robot.

3. The gait planning method for a bipedal robot on stepped terrain according to claim 1, characterized in that: The Bezier curve is: ； in, t is the normalized time coefficient, n is the order of the Bezier curve, and A is the control point of the Bezier curve; The preset Serre curve constraint parameters are used to constrain the position and speed at the initial and final moments of the foot trajectory.

4. A bipedal robot gait planning device for stepped terrain, implementing the bipedal robot gait planning method according to any one of claims 1 to 3, characterized in that: include: The first processing module is used to establish the walking modes of the biped robot for ascending stairs and descending stairs respectively based on the linear inverted pendulum model; The second processing module is configured to map the linear inverted pendulum model under the stepped terrain into a segmented virtual height linear inverted pendulum model under the flat terrain based on the walking pattern; The third processing module is used to realize gait planning of the biped robot on stepped terrain by using model predictive control based on a segmented virtual height linear inverted pendulum model.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the biped robot gait planning method for stepped terrain according to any one of claims 1 to 3 are implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the biped robot gait planning method for stepped terrain according to any one of claims 1 to 3 are implemented.

7. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the biped robot gait planning method for stepped terrain according to any one of claims 1 to 3 are implemented.

Images