Limb balance control method and system for humanoid robot

Abstract

The present application relates to the technical field of robot control, in particular to a limb balance control method and system for a humanoid robot, the method comprising: obtaining a forward path range of the humanoid robot, denoted as an alternative path set; emitting a first pre-set near-infrared light, receiving a reflected near-infrared light, denoted as a second near-infrared light, and capturing a scattered near-infrared light, denoted as a third near-infrared light; calculating reflectivity, scattering rate and ground smoothness in the alternative path set; searching for a path with the lowest ground smoothness, calculating an optimal center of gravity height of the humanoid robot, denoted as a first height; setting the forward path of the humanoid robot as the optimal path; and setting an average value of the center of gravity height during the forward movement of the humanoid robot according to the first height. The present application reduces the probability of the humanoid robot falling down on a smooth ground by searching for a path with the lowest ground smoothness and adjusting the center of gravity height according to the ground smoothness.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robot control, in particular to a limb balance control method and system for a humanoid robot. BACKGROUND

[0002] With the development of robot technology, humanoid robots will be more and more applied in industrial manufacturing and service industry. Especially in some environment not suitable for human work, humanoid robots have extremely broad application space. For example, in the cold chain logistics industry, if the ice layer is unevenly distributed on the ground of the cold storage, it is easy to slip and fall if the operation is carried out by manpower. Therefore, it is a feasible technical solution to use a humanoid robot to assist part of the operation. However, the humanoid robot operation still faces technical problems, that is, the limb balance problem of the humanoid robot on the smooth ground.

[0003] The principle of the humanoid robot advancing on the ground is similar to walking like a human being, mainly relying on the friction between the soles and the ground to move forward, so the humanoid robot is also easy to fall down when walking on the smooth ground. In order to prevent the humanoid robot from falling down on the smooth ground, the existing technical solution is to increase the suction cup on the soles of the humanoid robot to increase the friction, thereby preventing falling down, but this way consumes a lot of energy of the humanoid robot. Therefore, it is necessary to propose new technical means to solve this problem.

[0004] By learning from the way people walk on the smooth ground, there are two feasible technical solutions: first, since the ice layer on the ground of the cold storage is not evenly distributed, part of the ground is covered with ice layer, and there may also be areas without ice layer coverage, therefore, if the path with the lowest smoothness from the starting point to the ending point can be identified, and the path is followed, the probability of falling down will be greatly reduced; second, when walking on the smooth ground, the probability of falling down can be reduced by lowering the center of gravity. If the two solutions are combined, the probability of the humanoid robot falling down when walking on the smooth ground can be greatly reduced. SUMMARY

[0005] (1) Technical problem to be solved

[0006] The purpose of the present application is to provide a limb balance control method and system for a humanoid robot to reduce the probability of the humanoid robot falling down when walking on the smooth ground.

[0007] (2) Technical solution

[0008] In order to achieve the above-mentioned purpose, the present application provides a limb balance control method for a humanoid robot, which comprises the following steps:

[0009] S1, obtaining the advancing path range of the humanoid robot, denoted as the candidate path set.

[0010] S2, emitting a first pre-set near-infrared light to the set of candidate paths through the near-infrared light transceiver installed on the forehead support of the humanoid robot, receiving the reflected near-infrared light, denoted as second near-infrared light, capturing the scattered near-infrared light, denoted as third near-infrared light, through the near-infrared light scattering receiving device installed at the preset position.

[0011] S3, calculating the reflectivity and scattering rate according to the first, second and third near-infrared light, and calculating the ground smoothness in the set of candidate paths according to the reflectivity and scattering rate.

[0012] S4, searching for a path with the lowest ground smoothness in the set of candidate paths, denoted as the optimal path, recording the ground smoothness corresponding to the optimal path, denoted as the optimal path smoothness, calculating the optimal center of gravity height of the humanoid robot according to the optimal path smoothness, denoted as the first height, setting the advancing path of the humanoid robot as the optimal path, and if the first height is greater than the pre-set lower limit of the center of gravity height, controlling the walking posture of the limbs of the humanoid robot to make the average center of gravity height during the advancing of the humanoid robot be the first height, and then advancing along the optimal path, and if the first height is less than or equal to the lower limit of the center of gravity height, setting the humanoid robot to be static and issuing a ground smoothness warning.

[0013] Further, the method for obtaining the advancing path range of the humanoid robot, denoted as the set of candidate paths, comprises:

[0014] traversing all the non-backtracking paths to obtain a set of non-backtracking paths according to the pre-set advancing starting point, advancing ending point and pre-obtained minimum advancing step length and minimum side shift step length, wherein the non-backtracking path represents a path connecting the advancing starting point and the advancing ending point, and not repeating the same position and not retreating; searching for non-backtracking paths with path lengths less than the pre-set upper limit of path length in the set of non-backtracking paths to obtain the set of candidate paths by combination; and denoting the non-backtracking paths contained in the set of candidate paths as the first feasible path to the N feasible path, respectively. N The number of non-backtracking paths in the set of candidate paths is denoted as N.

[0015] Further, the method for emitting a first pre-set near-infrared light to the set of candidate paths through the near-infrared light transceiver installed on the forehead support of the humanoid robot, receiving the reflected near-infrared light, denoted as second near-infrared light, capturing the scattered near-infrared light, denoted as third near-infrared light, through the near-infrared light scattering receiving device installed at the preset position, comprises:

[0016] dividing the first feasible path to the N feasible path into to path segments; wherein, except for the path segment in the first feasible path to the last feasible path closest to the advancing terminal point, the length of each path segment is a pre-set standard path segment length N b N b N N

[0017] N N N N ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​The third near-infrared light is obtained by combining the values of the first near-infrared light and the second near-infrared light; and the near-infrared light scattering receiving device is in the shape of a concentric circle, and the near-infrared light transceiving device is located at the center of the concentric circle.

[0018] Further, the method for calculating the reflectivity and the scattering rate according to the first near-infrared light, the second near-infrared light and the third near-infrared light, and calculating the ground surface smoothness in the candidate path set according to the reflectivity and the scattering rate comprises:

[0019] The first near-infrared light intensity is obtained, denoted as , and the unit is watt per square meter.

[0020] The reflectivity is calculated according to the first near-infrared light and the second near-infrared light; the reflectivity comprises a first feasible path reflectivity to an nth feasible path reflectivity; and the calculation formula of the reflectivity is: N

[0021] ;

[0022] wherein, r i represents the ith feasible path reflectivity, dimensionless, and the value range is 0 to 1.

[0023] The scattering rate is calculated according to the first near-infrared light and the third near-infrared light; the scattering rate comprises a first feasible path scattering rate to an nth feasible path scattering rate; and the calculation formula of the scattering rate is: N

[0024] ;

[0025] wherein, s i represents the ith feasible path scattering rate, dimensionless, and the value range is 0 to 1.

[0026] The ground surface smoothness in the candidate path set is calculated according to the reflectivity and the scattering rate; the ground surface smoothness in the candidate path set comprises a first path ground surface smoothness to an nth path ground surface smoothness; and the calculation formula of the ground surface smoothness in the candidate path set is: N

[0027] ;

[0028] wherein, g i represents the ith path ground surface smoothness, dimensionless, and the value range is 0 to 1.

[0029] ​​​​​​​​​​Furthermore, the method for calculating the optimal center of gravity height of the humanoid robot based on the smoothness of the optimal path, denoted as the first height, includes:

[0030] Obtain a pre-prepared sample set; the sample set includes the critical center of gravity height of the humanoid robot under different surface smoothness conditions obtained from experiments; the value range of the critical center of gravity height is from the lower limit to the upper limit; the lower limit of the center of gravity height is obtained through pre-setting, representing the lowest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; the upper limit of the center of gravity height is obtained through pre-setting, representing the highest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; when the critical center of gravity height is the lower limit, it indicates that the humanoid robot... The humanoid robot will slip when the average center of gravity height is at the lower limit of the center of gravity height during forward movement. When the critical center of gravity height is the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is at the upper limit of the center of gravity height during forward movement. When the critical center of gravity height is greater than the lower limit of the center of gravity height but less than the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is less than or equal to the critical center of gravity height during forward movement, and the humanoid robot will slip when the average center of gravity height is higher than the critical center of gravity height during forward movement.

[0031] Based on the sample set, a nonlinear fitting algorithm is used to fit a functional relationship between ground smoothness and critical centroid height, denoted as the first function; the independent variable of the first function is ground smoothness, and the dependent variable is critical centroid height; the ground smoothness in the first function is set as the optimal path smoothness, and the value of the dependent variable is calculated, denoted as the second height; the first height is calculated based on the second height; the formula for calculating the first height is:

[0032] ;

[0033] in, This indicates the first altitude, expressed in millimeters. This indicates the second altitude, expressed in millimeters. This indicates the preset buoyancy height, in millimeters. This indicates the lower limit of the center of gravity height, in millimeters.

[0034] Based on the same inventive concept, in another aspect, the present invention also provides a limb balance control system for a humanoid robot, the system comprising:

[0035] The alternative path set acquisition module is used to obtain the range of forward paths for the humanoid robot, denoted as the alternative path set.

[0036] The near-infrared transceiver module, connected to the alternative path set acquisition module, is used to transmit a pre-set first near-infrared light to the alternative path set through a near-infrared transceiver device installed on the forehead support of the humanoid robot, and receive the reflected near-infrared light, denoted as the second near-infrared light. The scattered near-infrared light is captured by a near-infrared light scattering receiver installed at a preset position, denoted as the third near-infrared light.

[0037] The ground smoothness calculation module is connected to the near-infrared light transceiver module. It is used to calculate the reflectivity and scattering based on the first near-infrared light, the second near-infrared light, and the third near-infrared light, and to calculate the ground smoothness within the candidate path set based on the reflectivity and scattering.

[0038] The center of gravity height setting module, connected to the ground smoothness calculation module, is used to search for the path with the lowest ground smoothness within the set of candidate paths, and record it as the optimal path; record the ground smoothness corresponding to the optimal path, and record it as the optimal path smoothness; calculate the optimal center of gravity height of the humanoid robot based on the optimal path smoothness, and record it as the first height; set the humanoid robot's forward path as the optimal path; if the first height is greater than the preset lower limit of center of gravity height, the humanoid robot's limb walking posture is controlled to make the average center of gravity height during the humanoid robot's forward movement equal to the first height, and then the robot moves along the optimal path; if the first height is less than or equal to the lower limit of center of gravity height, the humanoid robot is set to stand still and a ground smoothness warning is issued.

[0039] Furthermore, the alternative path set acquisition module includes:

[0040] The path traversal module is used to traverse all non-backtracking paths based on a pre-set starting point, ending point, and pre-determined minimum forward step length and minimum lateral movement step length, obtaining a set of non-backtracking paths. A non-backtracking path is defined as a path connecting the starting point and ending point, without repeatedly traversing the same position or reversing. Based on a pre-set upper limit for path length, it searches the non-backtracking path set for paths with a length less than the upper limit, combining them to obtain a candidate path set. The non-backtracking paths included in the candidate path set are respectively denoted as the first feasible path to the next feasible path. N Feasible paths; among which, N This represents the number of paths without backtracking in the set of alternative paths.

[0041] Furthermore, the near-infrared transceiver module includes:

[0042] The segmentation module is used to connect the first feasible path to the first... N Feasible paths are divided into: to There are 10 path segments; among which, excluding the first feasible path to the 10th feasible path... NApart from the path segment closest to the destination in the feasible path, the length of each path segment is the pre-set standard path segment length. b The first feasible path to the first N The length of the path segment closest to the destination in the feasible path is less than or equal to b Move the first feasible path to the... N The path segments obtained from the feasible path partitioning are respectively denoted as the first path segment set to the second path segment set. N A set of path segments.

[0043] The transmit / receive module, connected to the segmentation module, is used to align the first path segment with the segmentation module via a near-infrared transceiver mounted on the forehead support of the humanoid robot. N Each path segment in the path segment set emits a pre-set first near-infrared light at its geometric center and receives the reflected near-infrared light, denoted as […]. Near-infrared light is reflected and captured by a near-infrared light scattering receiver installed at a preset position. These are denoted as follows: Scattered near-infrared light; the Reflected near-infrared light indicates the first The feasible path along the path direction from the starting point to the ending point. The near-infrared light reflected back from each path segment; measurement The intensity of reflected near-infrared light is denoted as The unit is watts per square meter; the aforementioned Scattered near-infrared light represents the first The feasible path along the path direction from the starting point to the ending point. The scattered near-infrared light corresponding to each path segment; measurement The intensity of scattered near-infrared light is denoted as The unit is watts per square meter; among which, The value is 1 to N Integers; The value is 1 to integers; for Reflecting near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the second near-infrared light; for Scattering near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the third near-infrared light; the shape of the near-infrared light scattering receiver is a concentric ring, and the near-infrared light transceiver is located at the center of the concentric ring.

[0044] Further, the ground smoothness calculation module comprises:

[0045] a near-infrared light intensity reading module, configured to obtain a first near-infrared light intensity, denoted as , in units of watts per square meter.

[0046] a reflectivity calculation module, connected to the near-infrared light intensity reading module, configured to calculate reflectivity according to the first near-infrared light and the second near-infrared light; the reflectivity comprises first feasible path reflectivity to N feasible path reflectivity; the calculation formula of the reflectivity is:

[0047] ;

[0048] wherein, represents the feasible path reflectivity, dimensionless, with a value range of 0 to 1.

[0049] a scattering rate calculation module, connected to the reflectivity calculation module, configured to calculate scattering rate according to the first near-infrared light and the third near-infrared light; the scattering rate comprises first feasible path scattering rate to N feasible path scattering rate; the calculation formula of the scattering rate is:

[0050] ;

[0051] wherein, represents the feasible path scattering rate, dimensionless, with a value range of 0 to 1.

[0052] a comprehensive calculation module, connected to the scattering rate calculation module, configured to calculate ground smoothness in the candidate path set according to the reflectivity and the scattering rate; the ground smoothness in the candidate path set comprises first path ground smoothness to N path ground smoothness; the calculation formula of the ground smoothness in the candidate path set is:

[0053] ;

[0054] wherein, represents the path ground smoothness, dimensionless, with a value range of 0 to 1. represents a pre-set weight coefficient, dimensionless, with a value range of 0 to 1.

[0055] Further, the gravity center height setting module comprises:

[0056] The sample set acquisition module is configured to acquire a pre-acquired sample set, the sample set comprising critical center of gravity heights of a trial human-shaped robot advancing on different ground smoothness levels, the critical center of gravity heights ranging from a lower limit of the center of gravity height to an upper limit of the center of gravity height, the lower limit of the center of gravity height being pre-set to represent an average value of the lowest center of gravity height maintained by the human-shaped robot in a walking posture during the advancing process, the upper limit of the center of gravity height being pre-set to represent an average value of the highest center of gravity height maintained by the human-shaped robot in the walking posture during the advancing process, the critical center of gravity height being equal to the lower limit of the center of gravity height, indicating that the human-shaped robot will fall down when the average value of the center of gravity height during the advancing process is equal to the lower limit of the center of gravity height, the critical center of gravity height being equal to the upper limit of the center of gravity height, indicating that the human-shaped robot will not fall down when the average value of the center of gravity height during the advancing process is equal to the upper limit of the center of gravity height, and the critical center of gravity height being greater than the lower limit of the center of gravity height and less than the upper limit of the center of gravity height, indicating that the human-shaped robot will not fall down when the average value of the center of gravity height during the advancing process is less than or equal to the critical center of gravity height, and the human-shaped robot will fall down when the average value of the center of gravity height during the advancing process is greater than the critical center of gravity height.

[0057] The first height calculation module is connected with the sample set acquisition module and is configured to acquire a function relationship between the ground smoothness level and the critical center of gravity height by using a nonlinear fitting algorithm to fit the sample set, denoted as a first function, the independent variable of the first function being the ground smoothness level, and the dependent variable of the first function being the critical center of gravity height, the ground smoothness level in the first function being set as the optimal path smoothness level, the dependent variable being calculated to obtain a value, denoted as a second height, the first height being calculated according to the second height, and the calculation formula of the first height being:

[0058] ;

[0059] wherein, the first height is denoted as mm; the second height is denoted as mm; the pre-set floating height is denoted as mm; the lower limit of the center of gravity height is denoted as mm.

[0060] (3) Beneficial effects

[0061] Compared with the prior art, the beneficial effects of the present application are:

[0062] According to the reflectivity and the scattering rate, the ground surface smoothness in the candidate path set is calculated, and the path with the lowest ground surface smoothness is searched, which is recorded as the optimal path, and the average value of the center of gravity height of the humanoid robot during the advancing is set according to the ground surface smoothness corresponding to the optimal path, so as to reduce the probability of the humanoid robot falling down on the smooth ground surface. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 A flow chart of the limb balance control method for the humanoid robot in the embodiment 1 of the present application is shown in the figure.

[0064] Figure 2 A module composition schematic diagram of the limb balance control system for the humanoid robot in the embodiment 2 of the present application is shown in the figure. DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the present application will be clearly and completely described in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0066] Before examples are given, the application scenario of the concept of the present application needs to be described. The present application is applied to reduce the probability of the humanoid robot falling down when walking on the ground surface of the cold storage with unevenly covered ice layer.

[0067] Embodiment 1: As shown in the figure, the present embodiment provides a limb balance control method for a humanoid robot, which comprises the following steps: Figure 1

[0068] S1, the advancing path range of the humanoid robot is obtained, which is recorded as a candidate path set.

[0069] S2, the pre-set first near-infrared light is emitted to the candidate path set through the near-infrared light transceiver device installed on the forehead support of the humanoid robot, and the reflected near-infrared light is received, which is recorded as the second near-infrared light, and the scattered near-infrared light is captured through the near-infrared light scattering receiving device installed at the preset position, which is recorded as the third near-infrared light.

[0070] S3, the reflectivity and the scattering rate are calculated according to the first near-infrared light, the second near-infrared light and the third near-infrared light, and the ground surface smoothness in the candidate path set is calculated according to the reflectivity and the scattering rate.

[0071] ​S4, searching for the path with the lowest ground smoothness from the ground smoothness of the set of candidate paths, and recording the path as the optimal path; recording the ground smoothness corresponding to the optimal path as the optimal path smoothness; calculating the optimal center of gravity height of the humanoid robot according to the optimal path smoothness, and recording the optimal center of gravity height as the first height; setting the advancing path of the humanoid robot as the optimal path; if the first height is greater than the preset lower limit of the center of gravity height, controlling the walking posture of the limbs of the humanoid robot to make the average height of the center of gravity of the humanoid robot during advancing equal to the first height, and then advancing along the optimal path; if the first height is less than or equal to the lower limit of the center of gravity height, setting the humanoid robot to be static and issuing a ground smoothness warning.

[0072] Exemplarily, the ground of the cold storage is converted into a grid map. The ground of the cold storage is unevenly covered with ice, and there is no obstacle on the ground. According to the preset advancing starting point and advancing ending point, combined with the kinematic constraints of the humanoid robot (including the minimum advancing step length and the minimum side shift step length), all the non-backtracking paths in the grid map are traversed. The non-backtracking path represents a path that connects the advancing starting point and the advancing ending point, and does not pass through the same position repeatedly or retreat. In the grid map, the non-backtracking path can be further represented as a path that connects the grid where the humanoid robot is during advancing and the grid where the advancing ending point is located, and does not pass through the same grid repeatedly or retreat. Then, the set of candidate paths is screened according to the upper limit of the path length. By using the grid map, the path search space can be limited, and the calculation amount can be reduced.

[0073] The set of candidate paths is divided into path segments, and then the near-infrared light emitting and receiving device installed on the forehead support of the humanoid robot emits near-infrared light to each path segment in the set of candidate paths. By adjusting the extension of the support, the angle between the line connecting the near-infrared light emitting and receiving device and the ground of the cold storage and 90 degrees is less than the preset upper limit of the deviation (20 degrees). The near-infrared light represents light waves with a wavelength between visible light and mid-infrared light. The wavelength of the emitted near-infrared light is 1540-1560 nanometers. The near-infrared light has the performance of resisting environmental light interference and resisting cold storage mist interference. The emitted near-infrared light is recorded as first near-infrared light. After the first near-infrared light is emitted to the ground of the cold storage, part of the first near-infrared light is reflected by the ground of the cold storage and then shot back to the near-infrared light emitting and receiving device, and another part of the first near-infrared light is scattered by the ground of the cold storage and then shot to the surrounding environment. The reflected near-infrared light received by the near-infrared light emitting and receiving device is recorded as second near-infrared light. At the same time, the scattered near-infrared light is captured by the near-infrared light scattering receiving device installed at a preset position, and recorded as third near-infrared light. The shape of the near-infrared light scattering receiving device is a concentric circular ring, and the near-infrared light emitting and receiving device is located at the center of the concentric circular ring.

[0074] The reflectivity and the scattering rate are calculated according to the first near-infrared light, the second near-infrared light and the third near-infrared light. The higher the reflectivity is, the smoother the ground of the cold storage is; the lower the scattering rate is, the smoother the ground of the cold storage is. The smoothness of the ground in the candidate path set is calculated according to the reflectivity and the scattering rate. The smoothness of the ground in the candidate path set considers both the reflectivity and the scattering rate which are two interrelated indexes, thereby improving the robustness of the calculation and reducing the influence caused by errors.

[0075] The path with the lowest smoothness of the ground is searched according to the smoothness of the ground in the candidate path set, and is recorded as an optimal path. If the humanoid robot advances along the optimal path, the smoothness of the ground that the humanoid robot passes through is the lowest, that is, the roughness is the highest and the friction between the soles and the ground is the largest. The smoothness of the ground corresponding to the optimal path is recorded as an optimal path smoothness. The best center of gravity height of the humanoid robot is calculated according to the optimal path smoothness, and is recorded as a first height. When the humanoid robot walks, the higher the center of gravity height is, the more difficult it is to keep balance, and the lower the center of gravity height is, the easier it is to keep balance. However, if the center of gravity height is too low, the normal walking of the limbs cannot be maintained. Therefore, the best center of gravity height of the humanoid robot, that is, the first height, needs to be calculated according to the optimal path smoothness. Within a certain range, the higher the optimal path smoothness is, the lower the first height is, and the lower the optimal path smoothness is, the higher the first height is. The advancing path of the humanoid robot is set as the optimal path. If the first height is greater than a pre-set lower limit of the center of gravity height, the average value of the center of gravity height of the humanoid robot during the advancing period is controlled to be the first height by controlling the walking mode of the limbs of the humanoid robot, and then the humanoid robot advances along the optimal path. If the first height is less than or equal to the lower limit of the center of gravity height, it is indicated that the ground of the cold storage is too smooth, and even if the center of gravity of the humanoid robot is adjusted to the lowest, there is a high probability of falling down. At this time, the humanoid robot is set to be static and a ground smoothness warning is issued, and the humanoid robot starts again after the smoothness of the ground of the cold storage is reduced and the walking condition is met.

[0076] Further, the method for obtaining the advancing path range of the humanoid robot, recorded as a candidate path set, comprises:

[0077] According to the pre-set advancing starting point and advancing ending point and the pre-obtained minimum advancing step length and minimum side shift step length, all the non-backtracking paths are traversed to obtain a non-backtracking path set; the non-backtracking path represents a path connecting the advancing starting point and the advancing ending point and not repeatedly passing through the same position and not retreating; according to the pre-set upper limit of the path length, the non-backtracking paths with the path length less than the upper limit of the path length in the non-backtracking path set are searched to obtain the candidate path set by combination; the non-backtracking paths contained in the candidate path set are recorded as a first feasible path to an nth feasible path respectively; wherein, N n represents the number of non-backtracking paths in the candidate path set. N n represents the number of non-backtracking paths in the candidate path set.

[0078] For example, the starting point is set as shelf number 1, and the ending point is set as shelf number 3. Combining the kinematic constraints of the humanoid robot, the minimum forward stride length is set to 300 mm, which is the shortest stride distance for the humanoid robot, and the minimum lateral stride length is set to 150 mm, which is the minimum lateral stride distance of the humanoid robot's hip joint. The minimum forward stride length and minimum lateral stride length reflect the stride length of the humanoid robot moving on the cold storage floor. Based on these minimum forward stride lengths and minimum lateral stride lengths, all non-backtracking paths are traversed on the gridded map to obtain a set of non-backtracking paths, and finally, a set of candidate paths is obtained. This set of candidate paths includes the first feasible path to the 346th feasible path.

[0079] Further, the method of transmitting a pre-set first near-infrared light to a set of candidate paths via a near-infrared transceiver device installed on the forehead support of a humanoid robot, and receiving the reflected near-infrared light, denoted as the second near-infrared light, and capturing the scattered near-infrared light, denoted as the third near-infrared light, via a near-infrared light scattering receiver installed at a preset position includes:

[0080] Move the first feasible path to the... N Feasible paths are divided into: to There are 10 path segments; among which, excluding the first feasible path to the 10th feasible path... N Apart from the path segment closest to the destination in the feasible path, the length of each path segment is the pre-set standard path segment length. b The first feasible path to the first N The length of the path segment closest to the destination in the feasible path is less than or equal to b Move the first feasible path to the... N The path segments obtained from the feasible path partitioning are respectively denoted as the first path segment set to the second path segment set. N A set of path segments.

[0081] The near-infrared transceiver, mounted on the forehead support of the humanoid robot, is aligned with the first path segment and then converges to the second. N Each path segment in the path segment set emits a pre-set first near-infrared light at its geometric center and receives the reflected near-infrared light, denoted as […]. Near-infrared light is reflected and captured by a near-infrared light scattering receiver installed at a preset position. These are denoted as follows: Scattered near-infrared light; the Reflected near-infrared light indicates the first The feasible path along the path direction from the starting point to the ending point. The near-infrared light reflected back from each path segment; measurement The intensity of reflected near-infrared light is denoted as The unit is watts per square meter; the aforementioned Scattered near-infrared light represents the first The feasible path along the path direction from the starting point to the ending point. The scattered near-infrared light corresponding to each path segment; measurement The intensity of scattered near-infrared light is denoted as The unit is watts per square meter; among which, The value is 1 to N Integers; The value is 1 to integers; for Reflecting near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the second near-infrared light; for Scattering near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the third near-infrared light; the shape of the near-infrared light scattering receiver is a concentric ring, and the near-infrared light transceiver is located at the center of the concentric ring.

[0082] For example, the length of the first feasible path is 4.9 meters, or 4900 millimeters. This is a pre-defined standard path segment length. b The standard path segment length is 1000 mm. This length can be manually set. To achieve a more accurate surface smoothness, the standard path segment length should be set smaller, but this increases computational load and operational complexity. Conversely, to reduce computational load and operational complexity, the standard path segment length can be set larger, but this decreases the accuracy of the obtained surface smoothness. Due to the standard path segment length... b Given a length of 1000 mm and a first feasible path length of 4900 mm, the first feasible path can be divided into 5 path segments. The first to fourth path segments are each 1000 mm long, and the fifth path segment is 900 mm long. These 5 path segments constitute the first path segment set. Similarly, the first path segment set to the 346th path segment set can be obtained.

[0083] A near-infrared transceiver mounted on the forehead support of a humanoid robot emits a pre-set first near-infrared beam at the geometric center of each path segment in the first to the 346th path segment sets, and receives the reflected near-infrared beam. The geometric center is obtained using an image recognition algorithm. When the first near-infrared beam is emitted at the first path segment in the first path segment set, the received reflected near-infrared beam is denoted as... Reflecting near-infrared light. The scattered near-infrared light is captured by a near-infrared light scattering receiving device installed at a preset position, and when the first near-infrared light is emitted to align with the first path segment in the first path segment set, the received scattered near-infrared light is recorded as Scattering near-infrared light. By analogy, the reflected near-infrared light and the scattered near-infrared light of all path segments are obtained. The reflected near-infrared light of all path segments is combined and recorded as the second near-infrared light. The scattered near-infrared light of all path segments is combined and recorded as the third near-infrared light.

[0084] The shape of the near-infrared light scattering receiving device is a concentric circle, and the near-infrared light transceiver device is located at the center of the concentric circle. The outer diameter of the concentric circle is 200 mm, and the inner diameter is 50 mm, so as to cover the typical scattering angle.

[0085] Further, the method for calculating the reflectivity and the scattering rate according to the first near-infrared light, the second near-infrared light and the third near-infrared light, and calculating the ground smoothness in the alternative path set according to the reflectivity and the scattering rate comprises:

[0086] Obtaining the intensity of the first near-infrared light, recorded as , unit: watt per square meter.

[0087] Calculating the reflectivity according to the first near-infrared light and the second near-infrared light; the reflectivity includes the first feasible path reflectivity to the N feasible path reflectivity; the calculation formula of the reflectivity is:

[0088] ;

[0089] Among them, represents the feasible path reflectivity, dimensionless, the value range is 0 to 1.

[0090] Calculating the scattering rate according to the first near-infrared light and the third near-infrared light; the scattering rate includes the first feasible path scattering rate to the N feasible path scattering rate; the calculation formula of the scattering rate is:

[0091] ;

[0092] Among them, represents the feasible path scattering rate, dimensionless, the value range is 0 to 1.

[0093] Calculating the ground smoothness in the alternative path set according to the reflectivity and the scattering rate; the ground smoothness in the alternative path set includes the first path ground smoothness to the NThe path ground smoothness degree; the calculation formula of the ground smoothness degree in the alternative path set is:

[0094] ;

[0095] Wherein, The path ground smoothness degree, dimensionless, the value range is 0 to 1; The path ground smoothness degree, dimensionless, the value range is 0 to 1; The preset weight coefficient, dimensionless, the value range is 0 to 1.

[0096] Exemplarily, the first feasible path reflectivity reflects the smoothness degree of the first feasible path to a certain extent, and the higher the first feasible path reflectivity is, the more smooth the first feasible path is. The first feasible path scattering rate also reflects the smoothness degree of the first feasible path to a certain extent, and the lower the first feasible path scattering rate is, the more smooth the first feasible path is. From the physical principle, the reflectivity and the scattering rate should be strictly negatively correlated. However, the first feasible path reflectivity and the first feasible path scattering rate obtained by the embodiment are measured, and there are inevitably errors. Therefore, in order to improve the robustness of the calculation, the first feasible path reflectivity and the first feasible path scattering rate are comprehensively used to obtain the path ground smoothness degree. The calculation formula of the path ground smoothness degree is: The preset weight coefficient reflects the judgment of the measurement accuracy of the reflectivity and the scattering rate. Since the near-infrared light scattering receiving device can only receive part of the scattered near-infrared light, the measurement error of the scattering rate is greater than that of the reflectivity, The value should be less than 0.5. In the embodiment, the value is 0.2. Further, the method for calculating the best center of gravity height of the humanoid robot according to the optimal path smoothness degree, recorded as the first height, comprises: Further, the method for calculating the best center of gravity height of the humanoid robot according to the optimal path smoothness degree, recorded as the first height, comprises: The calculation formula of the path ground smoothness degree is: The preset weight coefficient reflects the judgment of the measurement accuracy of the reflectivity and the scattering rate. Since the near-infrared light scattering receiving device can only receive part of the scattered near-infrared light, the measurement error of the scattering rate is greater than that of the reflectivity, The value should be less than 0.5.

[0097] Further, the method for calculating the best center of gravity height of the humanoid robot according to the optimal path smoothness degree, recorded as the first height, comprises:

[0098] ​​​Obtain a pre-prepared sample set; the sample set includes the critical center of gravity height of the humanoid robot under different surface smoothness conditions obtained from experiments; the value range of the critical center of gravity height is from the lower limit to the upper limit; the lower limit of the center of gravity height is obtained through pre-setting, representing the lowest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; the upper limit of the center of gravity height is obtained through pre-setting, representing the highest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; when the critical center of gravity height is the lower limit, it indicates that the humanoid robot... The humanoid robot will slip when the average center of gravity height is at the lower limit of the center of gravity height during forward movement. When the critical center of gravity height is the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is at the upper limit of the center of gravity height during forward movement. When the critical center of gravity height is greater than the lower limit of the center of gravity height but less than the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is less than or equal to the critical center of gravity height during forward movement, and the humanoid robot will slip when the average center of gravity height is higher than the critical center of gravity height during forward movement.

[0099] Based on the sample set, a nonlinear fitting algorithm is used to fit a functional relationship between ground smoothness and critical centroid height, denoted as the first function; the independent variable of the first function is ground smoothness, and the dependent variable is critical centroid height; the ground smoothness in the first function is set as the optimal path smoothness, and the value of the dependent variable is calculated, denoted as the second height; the first height is calculated based on the second height; the formula for calculating the first height is:

[0100] ;

[0101] in, This indicates the first altitude, expressed in millimeters. This indicates the second altitude, expressed in millimeters. This indicates the preset buoyancy height, in millimeters. This indicates the lower limit of the center of gravity height, in millimeters.

[0102] Exemplarily, the smoother the ground is, the more the humanoid robot needs to lower the height of the center of gravity to prevent falling down. The critical height of the center of gravity of the humanoid robot advancing on the ground with different degrees of smoothness is obtained through experiments, and then the optimal height of the center of gravity of the humanoid robot when walking is determined according to the calculated degree of smoothness of the optimal path. Since the height of the center of gravity of the humanoid robot fluctuates up and down during walking, it is not constant, so in this embodiment, the first height is taken as the average value of the height of the center of gravity during the advancement of the humanoid robot. The control of the average value of the height of the center of gravity during the advancement of the humanoid robot is realized by controlling the bending angle of the knee joint of the humanoid robot. The preset lowering height in the embodiment is based on the consideration of safety margin, thereby reducing the probability of the humanoid robot falling down.

[0103] Embodiment 2: Based on the same inventive concept, as Figure 2 shown, the embodiment also provides a limb balance control system for a humanoid robot, which comprises:

[0104] An alternative path set acquisition module is configured to acquire an advancement path range of the humanoid robot, denoted as an alternative path set.

[0105] A near-infrared light transceiver module is connected to the alternative path set acquisition module, configured to emit a preset first near-infrared light to the alternative path set through a near-infrared light transceiver device installed on a forehead support of the humanoid robot, and receive reflected near-infrared light, denoted as second near-infrared light, and capture scattered near-infrared light, denoted as third near-infrared light, through a near-infrared light scattering receiving device installed at a preset position.

[0106] A ground smoothness calculation module is connected to the near-infrared light transceiver module, configured to calculate reflectivity and scattering rate according to the first near-infrared light, the second near-infrared light and the third near-infrared light, and calculate the degree of smoothness of the ground in the alternative path set according to the reflectivity and the scattering rate.

[0107] A center of gravity height setting module is connected to the ground smoothness calculation module, configured to search for a path with the lowest degree of smoothness of the ground in the alternative path set, denoted as an optimal path; record the degree of smoothness of the ground corresponding to the optimal path, denoted as the degree of smoothness of the optimal path; calculate the optimal height of the center of gravity of the humanoid robot according to the degree of smoothness of the optimal path, denoted as the first height; set the advancement path of the humanoid robot as the optimal path; if the first height is greater than a preset lower limit of the height of the center of gravity, control the walking posture of the limbs of the humanoid robot so that the average value of the height of the center of gravity during the advancement of the humanoid robot is the first height, and then advance along the optimal path; if the first height is less than or equal to the lower limit of the height of the center of gravity, set the humanoid robot to be static and issue a ground smoothness warning.

[0108] Further, the alternative path set acquisition module comprises:

[0109] The path traversal module is used to traverse all non-backtracking paths based on a pre-set starting point, ending point, and pre-determined minimum forward step length and minimum lateral movement step length, obtaining a set of non-backtracking paths. A non-backtracking path is defined as a path connecting the starting point and ending point, without repeatedly traversing the same position or reversing. Based on a pre-set upper limit for path length, it searches the non-backtracking path set for paths with a length less than the upper limit, combining them to obtain a candidate path set. The non-backtracking paths included in the candidate path set are respectively denoted as the first feasible path to the next feasible path. N Feasible paths; among which, N This represents the number of paths without backtracking in the set of alternative paths.

[0110] Furthermore, the near-infrared transceiver module includes:

[0111] The segmentation module is used to connect the first feasible path to the first... N Feasible paths are divided into: to There are 10 path segments; among which, excluding the first feasible path to the 10th feasible path... N Apart from the path segment closest to the destination in the feasible path, the length of each path segment is the pre-set standard path segment length. b The first feasible path to the first N The length of the path segment closest to the destination in the feasible path is less than or equal to b Move the first feasible path to the... N The path segments obtained from the feasible path partitioning are respectively denoted as the first path segment set to the second path segment set. N A set of path segments.

[0112] The transmit / receive module, connected to the segmentation module, is used to align the first path segment with the segmentation module via a near-infrared transceiver mounted on the forehead support of the humanoid robot. N Each path segment in the path segment set emits a pre-set first near-infrared light at its geometric center and receives the reflected near-infrared light, denoted as […]. Near-infrared light is reflected and captured by a near-infrared light scattering receiver installed at a preset position. These are denoted as follows: Scattered near-infrared light; the Reflected near-infrared light indicates the first The feasible path along the path direction from the starting point to the ending point. The near-infrared light reflected back from each path segment; measurement The intensity of reflected near-infrared light is denoted as The unit is watts per square meter; the aforementioned Scattered near-infrared light represents the first The feasible path along the path direction from the starting point to the ending point. The scattered near-infrared light corresponding to each path segment; measurement The intensity of scattered near-infrared light is denoted as The unit is watts per square meter; among which, The value is 1 to N Integers; The value is 1 to integers; for Reflecting near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the second near-infrared light; for Scattering near-infrared light, Traverse 1 to N The value of will Traverse 1 to The values ​​of are combined to obtain the third near-infrared light; the shape of the near-infrared light scattering receiver is a concentric ring, and the near-infrared light transceiver is located at the center of the concentric ring.

[0113] Furthermore, the ground smoothness calculation module includes:

[0114] The near-infrared light intensity reading module is used to acquire the first near-infrared light intensity, denoted as... The unit is watts per square meter.

[0115] The reflectance calculation module, connected to the near-infrared light intensity reading module, is used to calculate the reflectance based on the first near-infrared light and the second near-infrared light; the reflectance includes the reflectance of the first feasible path to the second... N Feasible path reflectivity; the formula for calculating the reflectivity is:

[0116] ;

[0117] in, Indicates the first The reflectivity of the feasible path is dimensionless and ranges from 0 to 1.

[0118] The scattering rate calculation module, connected to the reflectance calculation module, is used to calculate the scattering rate based on the first near-infrared light and the third near-infrared light; the scattering rate includes the scattering rate of the first feasible path to the scattering rate of the third feasible path. N Feasible path scattering rate; the formula for calculating the scattering rate is:

[0119] ;

[0120] in, Indicates the first The feasible path scattering rate is dimensionless and ranges from 0 to 1.

[0121] The integrated calculation module, connected to the scattering rate calculation module, is used to calculate the ground smoothness within the candidate path set based on reflectivity and scattering rate; the ground smoothness within the candidate path set includes the ground smoothness of the first path to the... N The smoothness of the path surface; the formula for calculating the smoothness of the surface within the set of candidate paths is:

[0122] ;

[0123] in, Indicates the first The smoothness of the path surface is dimensionless and ranges from 0 to 1. This represents a pre-defined weighting coefficient, which is dimensionless and ranges from 0 to 1.

[0124] Furthermore, the center of gravity height setting module includes:

[0125] A sample set acquisition module is used to acquire a pre-obtained sample set. The sample set includes the critical center of gravity height for a humanoid robot to move forward under different surface smoothness conditions, obtained from experiments. The critical center of gravity height ranges from the lower limit to the upper limit. The lower limit is pre-set and represents the lowest average center of gravity height required to maintain a limb walking posture during the robot's movement. The upper limit is also pre-set and represents the highest average center of gravity height required to maintain a limb walking posture during the robot's movement. When the critical center of gravity height is the lower limit, it indicates... When a humanoid robot moves forward, it will slip if the average height of its center of gravity is at the lower limit of its center of gravity. When the critical center of gravity height is the upper limit of its center of gravity, the humanoid robot will not slip if the average height of its center of gravity is at the upper limit of its center of gravity. When the critical center of gravity height is greater than the lower limit of its center of gravity but less than the upper limit of its center of gravity, the humanoid robot will not slip if the average height of its center of gravity is less than or equal to the critical center of gravity height, and it will slip if the average height of its center of gravity is greater than the critical center of gravity height.

[0126] The first height calculation module is connected with the sample set acquisition module, and is configured to obtain a function relationship between the ground smoothness and the critical barycentric height by using a nonlinear fitting algorithm according to the sample set, and the function relationship is recorded as a first function; the independent variable of the first function is the ground smoothness, and the dependent variable of the first function is the critical barycentric height; the ground smoothness in the first function is set as the optimal path smoothness, the dependent variable value is calculated, and the dependent variable value is recorded as a second height; the first height is calculated according to the second height; and a calculation formula of the first height is:

[0127] ;

[0128] wherein, represents the first height, and the unit is millimeter; represents the second height, and the unit is millimeter; represents a preset downward floating height, and the unit is millimeter; represents a barycentric height lower limit, and the unit is millimeter.

[0129] It should be noted that, as to the system in the above-mentioned embodiments, the specific manner in which each module performs an operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

[0130] Finally, it should be noted that: although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or equivalently replace part of the technical features, any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for limb balance control in humanoid robots, characterized in that, The method includes the following steps: S1, obtain the range of forward paths for the humanoid robot, denoted as the set of candidate paths; S2, a pre-set first near-infrared light is emitted to the set of alternative paths through a near-infrared light transceiver device installed on the forehead support of the humanoid robot, and the reflected near-infrared light is received and recorded as the second near-infrared light. The scattered near-infrared light is captured by a near-infrared light scattering receiver installed at a preset position and recorded as the third near-infrared light. S3, the reflectivity and scattering rate are calculated based on the first near-infrared light, the second near-infrared light, and the third near-infrared light, and the smoothness of the ground within the candidate path set is calculated based on the reflectivity and scattering rate. S4. Search for the path with the lowest ground smoothness in the candidate path set and record it as the optimal path; record the ground smoothness corresponding to the optimal path and record it as the optimal path smoothness; calculate the optimal center of gravity height of the humanoid robot based on the optimal path smoothness and record it as the first height; set the forward path of the humanoid robot as the optimal path; if the first height is greater than the preset lower limit of the center of gravity height, control the walking posture of the humanoid robot to make the average center of gravity height of the humanoid robot during forward movement equal to the first height, and then move forward along the optimal path; if the first height is less than or equal to the lower limit of the center of gravity height, set the humanoid robot to stand still and issue a ground smoothness warning.

2. The limb balance control method for humanoid robots as described in claim 1, characterized in that, The method for obtaining the range of the humanoid robot's forward path, denoted as the candidate path set, includes: Based on the pre-set starting point, ending point, and pre-obtained minimum forward step length and minimum lateral movement step length, all non-backtracking paths are traversed to obtain a set of non-backtracking paths. A non-backtracking path is defined as a path that connects the starting point and ending point, does not repeatedly pass through the same position, and does not retreat. Based on the pre-set upper limit of path length, non-backtracking paths with a path length less than the upper limit are searched in the set of non-backtracking paths and combined to obtain a set of candidate paths. The non-backtracking paths included in the set of candidate paths are denoted as the first feasible path to the Nth feasible path, where N represents the number of non-backtracking paths in the set of candidate paths.

3. The limb balance control method for humanoid robots as described in claim 2, characterized in that, The method of transmitting a pre-set first near-infrared light to a set of candidate paths via a near-infrared transceiver device installed on the forehead support of a humanoid robot, receiving the reflected near-infrared light (referred to as the second near-infrared light), and capturing the scattered near-infrared light (referred to as the third near-infrared light) via a near-infrared light scattering receiver installed at a preset position includes: Divide the first feasible path to the Nth feasible path into three parts. to There are 1,2,200 path segments; wherein, except for the path segment closest to the destination in the first feasible path to the Nth feasible path, the length of each path segment is the pre-set standard path segment length b; the length of the path segment closest to the destination in the first feasible path to the Nth feasible path is less than or equal to b; the path segments obtained by dividing the first feasible path to the Nth feasible path are respectively denoted as the first path segment set to the Nth path segment set. By using a near-infrared transceiver mounted on the forehead support of a humanoid robot, pre-set first near-infrared light is emitted towards the geometric center of each path segment in the first to Nth path segment sets, and the reflected near-infrared light is received, denoted as follows: Near-infrared light is reflected and captured by a near-infrared light scattering receiver installed at a preset position. These are denoted as follows: Scattered near-infrared light; the Reflected near-infrared light indicates the first The feasible path along the path direction from the starting point to the ending point. The near-infrared light reflected back from each path segment; measurement The intensity of reflected near-infrared light is denoted as The unit is watts per square meter; the aforementioned Scattered near-infrared light represents the first The feasible path along the path direction from the starting point to the ending point. The scattered near-infrared light corresponding to each path segment; measurement The intensity of scattered near-infrared light is denoted as The unit is watts per square meter; among which, The integers are values ​​from 1 to N; The value is 1 to integers; for Reflecting near-infrared light, Iterate through the values ​​from 1 to N, and... Traverse 1 to The values ​​of are combined to obtain the second near-infrared light; for Scattering near-infrared light, Iterate through the values ​​from 1 to N, and... Traverse 1 to The values ​​of are combined to obtain the third near-infrared light; the shape of the near-infrared light scattering receiver is a concentric ring, and the near-infrared light transceiver is located at the center of the concentric ring.

4. The limb balance control method for humanoid robots as described in claim 3, characterized in that, The method for calculating reflectance and scattering based on first near-infrared light, second near-infrared light, and third near-infrared light, and for calculating the ground smoothness within the candidate path set based on reflectance and scattering, includes: Obtain the first near-infrared light intensity, denoted as The unit is watts per square meter; The reflectance is calculated based on the first near-infrared light and the second near-infrared light; the reflectance includes the reflectance of the first feasible path to the reflectance of the Nth feasible path; the formula for calculating the reflectance is: ； in, Indicates the first Reflectivity of feasible path, dimensionless, ranging from 0 to 1; The scattering rate is calculated based on the first near-infrared light and the third near-infrared light; the scattering rate includes the scattering rate of the first feasible path to the scattering rate of the Nth feasible path; the formula for calculating the scattering rate is: ； in, Indicates the first Feasible path scattering rate, dimensionless, ranging from 0 to 1; The surface smoothness within the candidate path set is calculated based on reflectivity and scattering rate; the surface smoothness within the candidate path set includes the surface smoothness of the first path to the Nth path; the calculation formula for the surface smoothness within the candidate path set is: ； in, Indicates the first The smoothness of the path surface is dimensionless and ranges from 0 to 1. This represents a pre-defined weighting coefficient, which is dimensionless and ranges from 0 to 1.

5. The limb balance control method for humanoid robots as described in claim 4, characterized in that, The method for calculating the optimal center of gravity height of the humanoid robot based on the smoothness of the optimal path, and denoting it as the first height, includes: Obtain a pre-prepared sample set; the sample set includes the critical center of gravity height of the humanoid robot under different surface smoothness conditions obtained from experiments; the value range of the critical center of gravity height is from the lower limit to the upper limit; the lower limit of the center of gravity height is obtained through pre-setting, representing the lowest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; the upper limit of the center of gravity height is obtained through pre-setting, representing the highest average center of gravity height for maintaining the limb walking posture during the humanoid robot's movement; when the critical center of gravity height is the lower limit, it indicates that the humanoid robot... The humanoid robot will slip when the average center of gravity height is at the lower limit of the center of gravity height during forward movement; when the critical center of gravity height is the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is at the upper limit of the center of gravity height during forward movement; when the critical center of gravity height is greater than the lower limit of the center of gravity height but less than the upper limit of the center of gravity height, it means that the humanoid robot will not slip when the average center of gravity height is less than or equal to the critical center of gravity height during forward movement, and the humanoid robot will slip when the average center of gravity height is greater than the critical center of gravity height during forward movement. Based on the sample set, a nonlinear fitting algorithm is used to fit a functional relationship between ground smoothness and critical centroid height, denoted as the first function; the independent variable of the first function is ground smoothness, and the dependent variable is critical centroid height; the ground smoothness in the first function is set as the optimal path smoothness, and the value of the dependent variable is calculated, denoted as the second height; the first height is calculated based on the second height; the formula for calculating the first height is: ； in, This indicates the first altitude, expressed in millimeters. This indicates the second altitude, expressed in millimeters. This indicates the preset buoyancy height, in millimeters. This indicates the lower limit of the center of gravity height, in millimeters.

6. A limb balance control system for humanoid robots, characterized in that, The system includes: The alternative path set acquisition module is used to obtain the range of forward paths for the humanoid robot, which is denoted as the alternative path set. The near-infrared transceiver module is connected to the alternative path set acquisition module. It is used to transmit a pre-set first near-infrared light to the alternative path set through the near-infrared transceiver device installed on the forehead support of the humanoid robot, and receive the reflected near-infrared light, which is denoted as the second near-infrared light. The scattered near-infrared light is captured by the near-infrared light scattering receiver installed at a preset position, which is denoted as the third near-infrared light. The ground smoothness calculation module is connected to the near-infrared light transceiver module. It is used to calculate the reflectivity and scattering based on the first near-infrared light, the second near-infrared light, and the third near-infrared light, and to calculate the ground smoothness within the candidate path set based on the reflectivity and scattering. The center of gravity height setting module, connected to the ground smoothness calculation module, is used to search for the path with the lowest ground smoothness within the set of candidate paths, and record it as the optimal path; record the ground smoothness corresponding to the optimal path, and record it as the optimal path smoothness; calculate the optimal center of gravity height of the humanoid robot based on the optimal path smoothness, and record it as the first height; set the humanoid robot's forward path as the optimal path; if the first height is greater than the preset lower limit of center of gravity height, the humanoid robot's limb walking posture is controlled to make the average center of gravity height during the humanoid robot's forward movement equal to the first height, and then the robot moves along the optimal path; if the first height is less than or equal to the lower limit of center of gravity height, the humanoid robot is set to stand still and a ground smoothness warning is issued.

7. The limb balance control system for a humanoid robot as described in claim 6, characterized in that, The alternative path set acquisition module includes: The path traversal module is used to traverse all non-backtracking paths based on a pre-set starting point, a pre-set ending point, and a pre-determined minimum forward step length and minimum lateral movement step length, to obtain a set of non-backtracking paths. A non-backtracking path is defined as a path that connects the starting point and the ending point, does not repeatedly pass through the same position, and does not backtrack. Based on a pre-set upper limit for path length, the module searches for non-backtracking paths in the set whose path length is less than the upper limit, and combines them to obtain a set of candidate paths. The non-backtracking paths included in the candidate path set are denoted as the first feasible path to the Nth feasible path, where N represents the number of non-backtracking paths in the candidate path set.

8. The limb balance control system for a humanoid robot as described in claim 7, characterized in that, The near-infrared transceiver module includes: The segmentation module is used to divide the first feasible path to the Nth feasible path into segments. to There are 1,2,200 path segments; wherein, except for the path segment closest to the destination in the first feasible path to the Nth feasible path, the length of each path segment is the pre-set standard path segment length b; the length of the path segment closest to the destination in the first feasible path to the Nth feasible path is less than or equal to b; the path segments obtained by dividing the first feasible path to the Nth feasible path are respectively denoted as the first path segment set to the Nth path segment set. The transmitter-receiver module, connected to the segmentation module, is used to transmit pre-set first near-infrared light through a near-infrared transceiver device mounted on the forehead support of the humanoid robot, aiming at the geometric center of each path segment from the first path segment set to the Nth path segment set, and to receive the reflected near-infrared light, denoted as . Near-infrared light is reflected and captured by a near-infrared light scattering receiver installed at a preset position. These are denoted as follows: Scattered near-infrared light; the Reflected near-infrared light indicates the first The feasible path along the path direction from the starting point to the ending point. The near-infrared light reflected back from each path segment; measurement The intensity of reflected near-infrared light is denoted as The unit is watts per square meter; the aforementioned Scattered near-infrared light represents the first The feasible path along the path direction from the starting point to the ending point. The scattered near-infrared light corresponding to each path segment; measurement The intensity of scattered near-infrared light is denoted as The unit is watts per square meter; among which, The integers are values ​​from 1 to N; The value is 1 to integers; for Reflecting near-infrared light, Iterate through the values ​​from 1 to N, and... Traverse 1 to The values ​​of are combined to obtain the second near-infrared light; for Scattering near-infrared light, Iterate through the values ​​from 1 to N, and... Traverse 1 to The values ​​of are combined to obtain the third near-infrared light; the shape of the near-infrared light scattering receiver is a concentric ring, and the near-infrared light transceiver is located at the center of the concentric ring.

9. The limb balance control system for a humanoid robot as described in claim 8, characterized in that, The ground smoothness calculation module includes: The near-infrared light intensity reading module is used to acquire the first near-infrared light intensity, denoted as... The unit is watts per square meter; A reflectance calculation module, connected to a near-infrared light intensity reading module, is used to calculate the reflectance based on the first near-infrared light and the second near-infrared light; the reflectance includes the reflectance of the first feasible path to the reflectance of the Nth feasible path; the formula for calculating the reflectance is: ； in, Indicates the first Reflectivity of feasible path, dimensionless, ranging from 0 to 1; A scattering rate calculation module, connected to the reflectance calculation module, is used to calculate the scattering rate based on the first near-infrared light and the third near-infrared light; the scattering rate includes the scattering rate of the first feasible path to the scattering rate of the Nth feasible path; the calculation formula for the scattering rate is: ； in, Indicates the first Feasible path scattering rate, dimensionless, ranging from 0 to 1; The integrated calculation module, connected to the scattering rate calculation module, is used to calculate the surface smoothness within the candidate path set based on reflectivity and scattering rate. The surface smoothness within the candidate path set includes the surface smoothness of the first path to the Nth path. The calculation formula for the surface smoothness within the candidate path set is as follows: ； in, Indicates the first The smoothness of the path surface is dimensionless and ranges from 0 to 1. This represents a pre-defined weighting coefficient, which is dimensionless and ranges from 0 to 1.

10. The limb balance control system for a humanoid robot as described in claim 9, characterized in that, The center of gravity height setting module includes: A sample set acquisition module is used to acquire a pre-obtained sample set. The sample set includes the critical center of gravity height for a humanoid robot to move forward under different surface smoothness conditions, obtained from experiments. The critical center of gravity height ranges from the lower limit to the upper limit. The lower limit is pre-set and represents the lowest average center of gravity height required to maintain a limb walking posture during the robot's movement. The upper limit is also pre-set and represents the highest average center of gravity height required to maintain a limb walking posture during the robot's movement. When the critical center of gravity height is the lower limit, it indicates... When a humanoid robot moves forward, it will slip if the average height of its center of gravity is at the lower limit of its center of gravity. When the critical center of gravity height is the upper limit of its center of gravity, it will not slip if the average height of its center of gravity is at the upper limit of its center of gravity. When the critical center of gravity height is greater than the lower limit of its center of gravity and less than the upper limit of its center of gravity, it will not slip if the average height of its center of gravity is less than or equal to the critical center of gravity height, and it will slip if the average height of its center of gravity is greater than the critical center of gravity height. The first height calculation module, connected to the sample set acquisition module, is used to fit a functional relationship between ground smoothness and critical centroid height using a nonlinear fitting algorithm based on the sample set, denoted as the first function; the independent variable of the first function is ground smoothness, and the dependent variable of the first function is critical centroid height; the ground smoothness in the first function is set as the optimal path smoothness, and the value of the dependent variable is calculated, denoted as the second height; the first height is calculated based on the second height; the formula for calculating the first height is: ； in, This indicates the first altitude, expressed in millimeters. This indicates the second altitude, expressed in millimeters. This indicates the preset buoyancy height, in millimeters. This indicates the lower limit of the center of gravity height, in millimeters.

Images