Ground contact detection method and apparatus, and computer-readable storage medium and robot

Abstract

The present application belongs to the technical field of robots, and in particular relates to a ground contact detection method and apparatus, and a computer-readable storage medium and a robot. The method comprises: acquiring data information of leg joint encoders of a robot; on the basis of the data information of the leg joint encoders, determining a foot-end ground reaction force of the robot; and using a preset ground contact detection model to process the foot-end ground reaction force, in order to obtain a ground contact detection result of the robot, wherein the ground contact detection model is a logistic regression classifier model that is used for performing ground contact detection and obtained by means of training on the basis of a preset ground contact detection sample set, and each ground contact detection sample in the ground contact detection sample set comprises a sample foot-end ground reaction force and a corresponding ground contact detection result label. The present application eliminates the need to additionally configure sensors for robots, thereby reducing the usage cost of the robots, and avoiding the impact of sensor aging and damage, and thus maintaining relatively high accuracy.

Description

Ground contact detection methods, devices, computer-readable storage media, and robots

[0001] This application claims priority to Chinese Patent Application No. 202411855132.2, filed on December 14, 2024, entitled “Method, Apparatus, Computer-readable Storage Medium and Robot for Ground Contact Detection”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application belongs to the field of robotics technology, and in particular relates to a method, apparatus, computer-readable storage medium, and robot for detecting ground contact. Background Technology

[0003] In robot control, ground contact detection is frequently required, specifically checking whether the robot's feet are in contact with the ground. Existing ground contact detection methods primarily rely on force sensors on the feet, comparing the force sensor readings with a pre-set threshold to determine ground contact. Specifically, if the force sensor reading exceeds the threshold, it's considered to be in contact with the ground; otherwise, it's considered not in contact. However, this approach requires additional sensors for the robot, increasing operating costs. Furthermore, these sensors are prone to aging and damage, resulting in low accuracy in ground contact detection. Technical issues

[0004] In view of this, embodiments of this application provide a ground contact detection method, apparatus, computer-readable storage medium, and robot to solve the problems of high cost and low accuracy of existing ground contact detection methods. Technical solutions

[0005] A first aspect of this application provides a ground contact detection method, which may include:

[0006] Acquire data from the robot's leg joint encoders;

[0007] The foot support reaction force of the robot is determined based on the data information from the leg joint encoder;

[0008] The foot support reaction force is processed using a preset ground contact detection model to obtain the ground contact detection result of the robot;

[0009] The ground contact detection model is a logistic regression classifier model trained based on a preset ground contact detection sample set for ground contact detection. Each ground contact detection sample in the ground contact detection sample set includes the foot reaction force and the corresponding ground contact detection result label.

[0010] In one specific implementation of the first aspect, before processing the foot support reaction force using a preset ground contact detection model to obtain the ground contact detection result of the robot, the following may be included:

[0011] Obtain the ground contact detection sample set;

[0012] Determine the log-likelihood function corresponding to the ground contact detection sample set;

[0013] With the goal of maximizing the log-likelihood function, the model parameters of the ground contact detection model are estimated to obtain the model parameters of the ground contact detection model.

[0014] In one specific implementation of the first aspect, the step of estimating the model parameters of the ground contact detection model with the objective of maximizing the log-likelihood function to obtain the model parameters of the ground contact detection model may include:

[0015] The partial derivatives of the log-likelihood function with respect to the model parameters of the ground contact detection model are obtained;

[0016] Based on the partial derivative results and the preset learning rate, the adjustment amount of the model parameters of the ground contact detection model is determined;

[0017] The model parameters of the ground contact detection model are iteratively updated based on the adjustment amount to obtain the iteratively updated model parameters;

[0018] If the preset parameter iteration update termination condition is not met, return to the step of calculating the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model and its subsequent steps.

[0019] If the parameter iteration update termination condition is met, the iteratively updated model parameters are determined as the model parameters of the ground contact detection model.

[0020] In one specific implementation of the first aspect, processing the foot support reaction force using a preset ground contact detection model to obtain the robot's ground contact detection result may include:

[0021] The foot support reaction force is input into the ground contact detection model, and the ground contact probability of the robot output by the ground contact detection model is obtained;

[0022] The ground contact probability is determined as the ground contact detection result of the robot.

[0023] In one specific implementation of the first aspect, processing the foot support reaction force using a preset ground contact detection model to obtain the robot's ground contact detection result may include:

[0024] The foot support reaction force is input into the ground contact detection model, and the ground contact probability of the robot output by the ground contact detection model is obtained;

[0025] Based on the ground contact probability and a preset ground contact probability threshold, the robot's binary ground contact information is determined; wherein, the binary ground contact information includes whether the robot is in contact with the ground or not.

[0026] The binary ground contact information is determined as the ground contact detection result of the robot.

[0027] In one specific implementation of the first aspect, determining the robot's binary ground contact information based on the ground contact probability and a preset ground contact probability threshold may include:

[0028] If the ground contact probability is greater than or equal to the ground contact probability threshold, the binary ground contact information is determined to be ground contact.

[0029] If the ground contact probability is less than the ground contact probability threshold, the binary ground contact information is determined to be no ground contact.

[0030] In one specific implementation of the first aspect, the data information of the leg joint encoder may include the leg joint rotation angle, the leg joint rotation speed, and the leg joint motor current;

[0031] Determining the foot support reaction force of the robot based on the data information from the leg joint encoder may include:

[0032] Based on the leg joint rotation angle, the leg joint rotation speed, and the leg joint motor current, the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term are determined respectively.

[0033] The foot reaction force of the robot is determined based on the Jacobian matrix of the robot's legs, the torque vector of the leg joints, the centrifugal force term of the legs, the Coriolis force term, and the gravity term.

[0034] A second aspect of this application provides a ground contact detection device, which may include:

[0035] The data acquisition module is used to acquire data information from the robot's leg joint encoders;

[0036] The foot support reaction force determination module is used to determine the foot support reaction force of the robot based on the data information of the leg joint encoder;

[0037] The ground contact detection module is used to process the foot support reaction force using a preset ground contact detection model to obtain the ground contact detection result of the robot.

[0038] The ground contact detection model is a logistic regression classifier model trained based on a preset ground contact detection sample set for ground contact detection. Each ground contact detection sample in the ground contact detection sample set includes the foot reaction force and the corresponding ground contact detection result label.

[0039] In one specific implementation of the second aspect, the ground contact detection device may further include:

[0040] A ground contact detection sample set acquisition module is used to acquire the ground contact detection sample set;

[0041] The log-likelihood function determination module is used to determine the log-likelihood function corresponding to the ground contact detection sample set;

[0042] The model parameter estimation module is used to estimate the model parameters of the ground contact detection model with the goal of maximizing the log-likelihood function, so as to obtain the model parameters of the ground contact detection model.

[0043] In one specific implementation of the second aspect, the model parameter estimation module can be specifically used to: calculate the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model; determine the adjustment amount of the model parameters of the ground contact detection model based on the partial derivative and a preset learning rate; iteratively update the model parameters of the ground contact detection model based on the adjustment amount to obtain the iteratively updated model parameters; if the preset parameter iteration update termination condition is not met, return to the step of calculating the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model and its subsequent steps; if the parameter iteration update termination condition is met, determine the iteratively updated model parameters as the model parameters of the ground contact detection model.

[0044] In one specific implementation of the second aspect, the ground contact detection module may include:

[0045] The model interaction unit is used to input the foot support reaction force into the ground contact detection model and obtain the ground contact probability of the robot output by the ground contact detection model.

[0046] The first result determination unit is used to determine the ground contact probability as the ground contact detection result of the robot.

[0047] In one specific implementation of the second aspect, the ground contact detection module may include:

[0048] The model interaction unit is used to input the foot support reaction force into the ground contact detection model and obtain the ground contact probability of the robot output by the ground contact detection model.

[0049] The binary ground contact information determination unit is used to determine the robot's binary ground contact information based on the ground contact probability and a preset ground contact probability threshold; wherein, the binary ground contact information includes whether the robot is in contact with the ground or not.

[0050] The second result determination unit is used to determine the binary ground contact information as the ground contact detection result of the robot.

[0051] In one specific implementation of the second aspect, the binary ground contact information determination unit may be specifically used to: determine the binary ground contact information as ground contact when the ground contact probability is greater than or equal to the ground contact probability threshold; and determine the binary ground contact information as no ground contact when the ground contact probability is less than the ground contact probability threshold.

[0052] In one specific implementation of the second aspect, the data information of the leg joint encoder includes the leg joint rotation angle, the leg joint rotation speed, and the leg joint motor current;

[0053] The foot support reaction force determination module can be specifically used to: determine the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term based on the leg joint rotation angle, leg joint rotation speed, and leg joint motor current, respectively; and determine the robot's foot support reaction force based on the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term.

[0054] A third aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the above-described ground contact detection methods.

[0055] A fourth aspect of this application provides a robot including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above-described ground contact detection methods.

[0056] A fifth aspect of this application provides a computer program product that, when run on a robot, causes the robot to perform the steps of any of the above-described ground contact detection methods. Beneficial effects

[0057] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment acquires data information from the robot's leg joint encoder; determines the robot's foot support reaction force based on the data information from the leg joint encoder; processes the foot support reaction force using a preset ground contact detection model to obtain the robot's ground contact detection result; wherein, the ground contact detection model is a logistic regression classifier model for ground contact detection trained based on a preset ground contact detection sample set, and each ground contact detection sample in the ground contact detection sample set includes the sample foot support reaction force and the corresponding ground contact detection result label. Through this application embodiment, there is no need to configure additional sensors for the robot, reducing the robot's operating cost, and it is not affected by sensor aging and damage, maintaining a high accuracy rate. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 is a flowchart of an embodiment of a ground contact detection method according to this application;

[0060] Figure 2 is a schematic flowchart of the training process of the ground contact detection model;

[0061] Figure 3 is a schematic diagram of the training process of the ground contact detection model;

[0062] Figure 4 is a schematic diagram of the trained ground contact detection model;

[0063] Figure 5 is a comparative diagram of the ground contact test results;

[0064] Figure 6 is a structural diagram of an embodiment of a ground contact detection device according to this application;

[0065] Figure 7 is a schematic block diagram of a robot according to an embodiment of this application. Embodiments of the present invention

[0066] To make the inventive objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0067] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0068] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0069] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0070] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if [the described condition or event] is detected" may be interpreted, depending on the context, as "once determined," "in response to determination," "once [the described condition or event] is detected," or "in response to detection of [the described condition or event]."

[0071] Furthermore, in the description of this application, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0072] In robot control, ground contact detection is frequently required, specifically checking whether the robot's feet are in contact with the ground. Existing ground contact detection methods primarily rely on force sensors on the feet, comparing the force sensor readings with a pre-set threshold to determine ground contact. Specifically, if the force sensor reading exceeds the threshold, it's considered to be in contact with the ground; otherwise, it's considered not in contact. However, this approach requires additional sensors for the robot, increasing operating costs. Furthermore, these sensors are prone to aging and damage, resulting in low accuracy in ground contact detection.

[0073] In view of this, embodiments of this application provide a ground contact detection method, apparatus, computer-readable storage medium, and robot to solve the problems of high cost and low accuracy of existing ground contact detection methods.

[0074] In this embodiment, the foot support reaction force of the robot can be determined based on the data information of the robot's leg joint encoder, and the foot support reaction force can be processed using a preset ground contact detection model to obtain the ground contact detection result of the robot. There is no need to configure additional sensors for the robot, which reduces the robot's operating cost. Moreover, it is not affected by sensor aging and damage and can maintain a high accuracy rate.

[0075] The executing entity in the embodiments of this application can be a robot, including but not limited to industrial robots, home service robots, commercial service robots, and other types of robots.

[0076] Please refer to Figure 1. One embodiment of a ground contact detection method in this application may include:

[0077] Step S101: Obtain data information from the robot's leg joint encoder.

[0078] The data information of the leg joint encoder may include, but is not limited to, leg joint rotation angle, leg joint speed, and leg joint motor current.

[0079] Step S102: Determine the foot support reaction force of the robot based on the data information from the leg joint encoder.

[0080] In this embodiment, the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term can be determined based on the leg joint rotation angle, leg joint rotation speed, and leg joint motor current, respectively. Furthermore, the robot's foot ground reaction force (GRF) can be determined based on these parameters, as shown in the following formula:

[0081] Where, q l For leg joint angles, J is the rotational speed of the leg joint. l (q l J is the Jacobian matrix for the leg, which can be calculated based on the leg joint rotation angle. l T (q l ) is the transpose of the Jacobian matrix of the leg. The terms representing the centrifugal force, Coriolis force, and gravity of the leg can be calculated based on the leg joint rotation angle and the leg joint rotation speed, τ. l Let f be the torque vector of the leg joint, which has an approximate linear relationship with the leg joint motor current and can be calculated from the leg joint motor current. Let f be the reaction force at the end of the robot's foot. l,x f l,y and fl,z These are the components of the foot support reaction force on the x-axis, y-axis, and z-axis, respectively.

[0082] Step S103: Use a preset ground contact detection model to process the foot support reaction force to obtain the robot's ground contact detection results.

[0083] In this embodiment of the application, the corresponding metric coefficient can be determined based on the foot support reaction force, as shown in the following formula:

[0084] Where, μ f This is a metric used to measure the reliability of the foot's contact with the ground. When the foot begins to slide, this metric is equal to the static friction coefficient μ. s .

[0085] Because different surfaces have different coefficients of friction, μ s The value of is unknown, but as long as it is in μ f <μ s In this case, the foot is firmly in contact with the ground, and μ f The smaller the value, the stronger the contact with the ground. For numerical stability and calculation simplification, we can only consider the component of the foot support reaction force on the z-axis and ignore other components. We compare the component of the foot support reaction force on the z-axis with a specific threshold. When the component of the foot support reaction force on the z-axis is greater than the specific threshold, we determine that the robot's foot is in contact with the ground; otherwise, we determine that the robot's foot is not in contact with the ground.

[0086] Based on this understanding, embodiments of this application can use a preset ground contact detection model to process the foot support reaction force, thereby obtaining the robot's ground contact detection result. The ground contact detection model is a Logistic Regression Classifier (LRC) model trained based on a preset ground contact detection sample set, as shown in the following equation:

[0087] Among them, T l Let {0,1} be the set of events where the foot touches the ground, and T l =1 indicates that the foot touches the ground, T l =0 indicates that the foot does not touch the ground, β is the model parameter of the ground contact detection model, which may include the first model parameter β1 and the second model parameter β2, P(T l =1|f l,z ;β) is the component f of the foot support reaction force on the z-axis. l,z The ground contact probability is determined by the model parameter β of the ground contact detection model, and will be simplified and denoted as P in the following text.

[0088] In this embodiment, the touchdown detection model can be pre-trained to determine its parameters. The training process for the touchdown detection model may include the steps shown in Figure 2:

[0089] Step S201: Obtain the ground contact detection sample set.

[0090] Each contact test sample in the contact test sample set can include the foot support reaction force and the corresponding contact test result label. Here, the contact test sample set can be denoted as... Where k is the index of the contact tracing sample, 1≤k≤n, and n is the number of contact tracing samples in the contact tracing sample set, (f l,z,k ,T k Let f be the k-th ground contact detection sample in the ground contact detection sample set. l,z,k T represents the z-axis component of the foot support reaction force in the k-th ground contact detection sample. k This is the ground contact detection result label in the k-th ground contact detection sample. A value of 1 indicates that the foot is in contact with the ground, and a value of 0 indicates that the foot is not in contact with the ground.

[0091] Step S202: Determine the log-likelihood function corresponding to the ground contact detection sample set.

[0092] In this embodiment of the application, a likelihood function (LF) can be constructed based on the ground contact detection sample set as shown in the following formula:

[0093] Taking the natural logarithm of the above likelihood function, we obtain the log-likelihood function (LLF) as shown in the following equation:

[0094] Among them, f k (T k ) for T k The probability of occurrence, i.e., the label of the touchdown detection result in the k-th touchdown detection sample is T. k The probability, P k Let f be the z-axis component of the foot support reaction force in the k-th ground contact detection sample. l,z,k The contact probability is determined by the model parameter β of the contact detection model, f(T) = f(T1,T2,…,T). n Let T be the probability of occurrence, i.e., T1, T2, ..., T. n The joint probability of occurrence, where each T k Since they are all independent and have the same logic density function, the joint density function can be written as the product of each density function.

[0095] Step S203: With the goal of maximizing the log-likelihood function, estimate the model parameters of the ground contact detection model to obtain the model parameters of the ground contact detection model.

[0096] In this embodiment of the application, the model parameters of the ground contact detection model can be estimated by maximum likelihood estimation (MLE), that is, by maximizing the probability of the observed T, the unknown model parameters can be obtained.

[0097] Specifically, the partial derivative of the log-likelihood function with respect to the model parameters of the contact detection model can be obtained. Based on the partial derivative and the preset learning rate, the adjustment amount of the model parameters of the contact detection model can be determined. The model parameters of the contact detection model can then be iteratively updated according to the adjustment amount to obtain the iteratively updated model parameters, as shown in the following formula:

[0098] Where h is the learning rate, and its specific value can be flexibly set according to the actual situation. This application does not impose a specific limitation on it. Through this gradient descent method, the model parameter β can be gradually adjusted to maximize the log-likelihood function.

[0099] After each parameter iteration update, it can be determined whether the preset parameter iteration update termination condition is met. The parameter iteration update termination condition can be flexibly set according to the actual situation. For example, it can be that the number of iterations is greater than the preset number of iterations threshold, or the adjustment amount of the model parameters is less than the preset adjustment amount threshold, etc. This application embodiment does not specifically limit it.

[0100] If the parameter iteration update termination condition is not met, the next parameter iteration update can continue. This involves returning to the step of calculating the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model, and subsequent steps, until the parameter iteration update termination condition is met. If the parameter iteration update termination condition is met, the iterative update of the model parameters can end, and the model parameters after the last iteration update are determined as the final model parameters of the ground contact detection model.

[0101] Figure 3 shows a schematic diagram of the training process of the ground contact detection model. The horizontal axis represents the number of iterations, and the vertical axis represents the model error of the ground contact detection model. As shown in the figure, the model error of the ground contact detection model decreases continuously with the increase of the number of iterations until the iteration terminates.

[0102] Figure 4 shows a schematic diagram of the final trained ground contact detection model. The horizontal axis represents the component of the foot reaction force on the z-axis, in Newtons (N), and the vertical axis represents the ground contact probability P(T). l =1|f l,zAs shown in the figure, the ground contact detection model can map the foot support reaction force to the interval (0,1) to characterize the robot's ground contact probability.

[0103] After the ground contact detection model is trained, it can be used to perform actual ground contact detection. This involves inputting the foot support reaction force into the ground contact detection model and obtaining the robot's ground contact probability output by the ground contact detection model.

[0104] In one specific implementation of this application, the ground contact probability can be directly determined as the robot's ground contact detection result.

[0105] In another specific implementation of this application, the binary ground contact information of the robot can be determined based on the ground contact probability and a preset ground contact probability threshold. The specific value of the ground contact probability threshold can be flexibly set according to actual conditions; for example, it can be set to 0.5 or other values, and this application does not specifically limit it. The binary ground contact information can include ground contact (denoted as 1) or no ground contact (denoted as 0). If the ground contact probability is greater than or equal to the ground contact probability threshold, the binary ground contact information can be determined as ground contact. Conversely, if the ground contact probability is less than the ground contact probability threshold, the binary ground contact information can be determined as no ground contact.

[0106] Figure 5 shows a comparative diagram of the ground contact detection results. In the figure, the horizontal axis represents the time axis (seconds), the vertical axis represents the component of the foot reaction force on the z-axis (Newtons, N), and the vertical axis of the upper figure represents the ground contact detection results. The figure shows four different ground contact detection results: result 1 is the true value, result 2 is the ground contact probability output by the ground contact detection model in this embodiment, result 3 is the binary ground contact information in this embodiment, and result 4 is the ground contact detection result obtained through existing technology. It can be seen that the ground contact detection results obtained in this embodiment have a high accuracy rate.

[0107] In summary, the embodiments of this application can determine the foot support reaction force of the robot based on the data information of the robot's leg joint encoder, and process the foot support reaction force using a preset ground contact detection model to obtain the robot's ground contact detection result. There is no need to configure additional sensors for the robot, which reduces the robot's operating cost. Moreover, it is not affected by sensor aging and damage and can maintain a high accuracy rate.

[0108] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0109] Corresponding to the ground contact detection method described in the above embodiments, Figure 6 shows a structural diagram of an embodiment of a ground contact detection device provided in this application.

[0110] In this embodiment, a ground contact detection device may include:

[0111] The data acquisition module 601 is used to acquire data information from the robot's leg joint encoders;

[0112] The foot support reaction force determination module 602 is used to determine the foot support reaction force of the robot based on the data information of the leg joint encoder;

[0113] The ground contact detection module 603 is used to process the foot support reaction force using a preset ground contact detection model to obtain the ground contact detection result of the robot.

[0114] The ground contact detection model is a logistic regression classifier model trained based on a preset ground contact detection sample set for ground contact detection. Each ground contact detection sample in the ground contact detection sample set includes the foot reaction force and the corresponding ground contact detection result label.

[0115] In one specific implementation of this application embodiment, the ground contact detection device may further include:

[0116] A ground contact detection sample set acquisition module is used to acquire the ground contact detection sample set;

[0117] The log-likelihood function determination module is used to determine the log-likelihood function corresponding to the ground contact detection sample set;

[0118] The model parameter estimation module is used to estimate the model parameters of the ground contact detection model with the goal of maximizing the log-likelihood function, so as to obtain the model parameters of the ground contact detection model.

[0119] In one specific implementation of this application, the model parameter estimation module can be specifically used to: calculate the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model; determine the adjustment amount of the model parameters of the ground contact detection model based on the partial derivative and a preset learning rate; iteratively update the model parameters of the ground contact detection model based on the adjustment amount to obtain the iteratively updated model parameters; if the preset parameter iteration update termination condition is not met, return to the step of calculating the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model and its subsequent steps; if the parameter iteration update termination condition is met, determine the iteratively updated model parameters as the model parameters of the ground contact detection model.

[0120] In one specific implementation of this application embodiment, the ground contact detection module may include:

[0121] The model interaction unit is used to input the foot support reaction force into the ground contact detection model and obtain the ground contact probability of the robot output by the ground contact detection model.

[0122] The first result determination unit is used to determine the ground contact probability as the ground contact detection result of the robot.

[0123] In one specific implementation of this application embodiment, the ground contact detection module may include:

[0124] The model interaction unit is used to input the foot support reaction force into the ground contact detection model and obtain the ground contact probability of the robot output by the ground contact detection model.

[0125] The binary ground contact information determination unit is used to determine the robot's binary ground contact information based on the ground contact probability and a preset ground contact probability threshold; wherein, the binary ground contact information includes whether the robot is in contact with the ground or not.

[0126] The second result determination unit is used to determine the binary ground contact information as the ground contact detection result of the robot.

[0127] In one specific implementation of this application, the binary ground contact information determination unit can be specifically used to: determine the binary ground contact information as ground contact when the ground contact probability is greater than or equal to the ground contact probability threshold; and determine the binary ground contact information as no ground contact when the ground contact probability is less than the ground contact probability threshold.

[0128] In one specific implementation of this application embodiment, the data information of the leg joint encoder includes the leg joint rotation angle, the leg joint rotation speed, and the leg joint motor current;

[0129] The foot support reaction force determination module can be specifically used to: determine the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term based on the leg joint rotation angle, leg joint rotation speed, and leg joint motor current, respectively; and determine the robot's foot support reaction force based on the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term.

[0130] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0131] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0132] Figure 7 shows a schematic block diagram of a robot provided in an embodiment of this application. For ease of explanation, only the parts related to the embodiment of this application are shown.

[0133] As shown in FIG7, the robot 7 of this embodiment includes: a processor 70, a memory 71, and a computer program 72 stored in the memory 71 and executable on the processor 70. When the processor 70 executes the computer program 72, it implements the steps in the various ground contact detection method embodiments described above, such as steps S101 to S103 shown in FIG1. ​​Alternatively, when the processor 70 executes the computer program 72, it implements the functions of each module / unit in the various device embodiments described above, such as the functions of modules 601 to 603 shown in FIG6.

[0134] For example, the computer program 72 may be divided into one or more modules / units, which are stored in the memory 71 and executed by the processor 70 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 72 in the robot 7.

[0135] Those skilled in the art will understand that Figure 7 is merely an example of robot 7 and does not constitute a limitation on robot 7. It may include more or fewer components than shown, or combine certain components, or different components. For example, robot 7 may also include input / output devices, network access devices, buses, etc.

[0136] The processor 70 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0137] The memory 71 can be an internal storage unit of the robot 7, such as a hard drive or memory. The memory 71 can also be an external storage device of the robot 7, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 71 can include both internal and external storage units of the robot 7. The memory 71 is used to store the computer program and other programs and data required by the robot 7. The memory 71 can also be used to temporarily store data that has been output or will be output.

[0138] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0139] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0140] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0141] In the embodiments provided in this application, it should be understood that the disclosed devices / robots and methods can be implemented in other ways. For example, the device / robot embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0142] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0143] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0144] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the content included in the computer-readable storage medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable storage medium does not include electrical carrier signals and telecommunication signals.

[0145] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for detecting ground contact, characterized in that, include: Acquire data from the robot's leg joint encoders; The foot support reaction force of the robot is determined based on the data information from the leg joint encoder; The foot support reaction force is processed using a preset ground contact detection model to obtain the ground contact detection result of the robot; The ground contact detection model is a logistic regression classifier model trained based on a preset ground contact detection sample set for ground contact detection. Each ground contact detection sample in the ground contact detection sample set includes the foot reaction force and the corresponding ground contact detection result label.

2. The ground contact detection method according to claim 1, characterized in that, Before processing the foot support reaction force using a preset ground contact detection model to obtain the robot's ground contact detection result, the following steps are also included: Obtain the ground contact detection sample set; Determine the log-likelihood function corresponding to the ground contact detection sample set; With the goal of maximizing the log-likelihood function, the model parameters of the ground contact detection model are estimated to obtain the model parameters of the ground contact detection model.

3. The ground contact detection method according to claim 2, characterized in that, The step of estimating the model parameters of the ground contact detection model with the objective of maximizing the log-likelihood function, to obtain the model parameters of the ground contact detection model, includes: The partial derivatives of the log-likelihood function with respect to the model parameters of the ground contact detection model are obtained; Based on the partial derivative results and the preset learning rate, the adjustment amount of the model parameters of the ground contact detection model is determined; The model parameters of the ground contact detection model are iteratively updated based on the adjustment amount to obtain the iteratively updated model parameters; If the preset parameter iteration update termination condition is not met, return to the step of calculating the partial derivative of the log-likelihood function with respect to the model parameters of the ground contact detection model and its subsequent steps. If the parameter iteration update termination condition is met, the iteratively updated model parameters are determined as the model parameters of the ground contact detection model.

4. The ground contact detection method according to claim 1, characterized in that, The process of using a preset ground contact detection model to process the foot support reaction force to obtain the robot's ground contact detection result includes: The foot support reaction force is input into the ground contact detection model, and the ground contact probability of the robot output by the ground contact detection model is obtained; The ground contact probability is determined as the ground contact detection result of the robot.

5. The ground contact detection method according to claim 1, characterized in that, The process of using a preset ground contact detection model to process the foot support reaction force to obtain the robot's ground contact detection result includes: The foot support reaction force is input into the ground contact detection model, and the ground contact probability of the robot output by the ground contact detection model is obtained; Based on the ground contact probability and a preset ground contact probability threshold, the robot's binary ground contact information is determined; wherein, the binary ground contact information includes whether the robot is in contact with the ground or not. The binary ground contact information is determined as the ground contact detection result of the robot.

6. The ground contact detection method according to claim 5, characterized in that, The step of determining the robot's binary ground contact information based on the ground contact probability and a preset ground contact probability threshold includes: If the ground contact probability is greater than or equal to the ground contact probability threshold, the binary ground contact information is determined to be ground contact. If the ground contact probability is less than the ground contact probability threshold, the binary ground contact information is determined to be no ground contact.

7. The ground contact detection method according to any one of claims 1 to 6, characterized in that, The data information of the leg joint encoder includes the leg joint rotation angle, leg joint speed, and leg joint motor current; Determining the foot support reaction force of the robot based on the data information from the leg joint encoder includes: Based on the leg joint rotation angle, the leg joint rotation speed, and the leg joint motor current, the robot's leg Jacobian matrix, leg joint torque vector, leg centrifugal force term, Coriolis force term, and gravity term are determined respectively. The foot reaction force of the robot is determined based on the Jacobian matrix of the robot's legs, the torque vector of the leg joints, the centrifugal force term of the legs, the Coriolis force term, and the gravity term.

8. A ground contact detection device, characterized in that, include: The data acquisition module is used to acquire data information from the robot's leg joint encoders; The foot support reaction force determination module is used to determine the foot support reaction force of the robot based on the data information of the leg joint encoder; The ground contact detection module is used to process the foot support reaction force using a preset ground contact detection model to obtain the ground contact detection result of the robot. The ground contact detection model is a logistic regression classifier model trained based on a preset ground contact detection sample set for ground contact detection. Each ground contact detection sample in the ground contact detection sample set includes the foot reaction force and the corresponding ground contact detection result label.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the ground contact detection method as described in any one of claims 1 to 7.

10. A robot comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the ground contact detection method as described in any one of claims 1 to 7.

Images