Method and device for determining grounding state, robot and storage medium

Abstract

The invention provides a method and a device for determining a ground contact state, a robot and a storage medium. The method comprises the following steps: determining a counter-acting force of a foot of the robot to the ground; according to a probability model obtained through pre-training, estimating the probability of stable contact between the foot and the ground under the condition that the reacting force is known; and according to the probability, obtaining the ground contact state of the foot and the ground. The method can reduce the dependence of the detection of the grounding state on the contact sensor.

Description

Technical Field

[0001] The present application relates to the field of robots, and more specifically, to a method, device, robot and storage medium for determining a ground contact state in the field of robots. Background Art

[0002] At present, in order to detect the contact status between the robot's feet and the ground, it is usually necessary to install additional contact sensors on the bottom of the robot's feet. However, due to objective factors such as the robot's movement mode and environment, the contact sensors are usually easily damaged. Therefore, how to reduce the dependence of the detection of the ground contact status on the contact sensors has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] The present application provides a method, device, robot and storage medium for determining a ground contact state, which can reduce the dependence of ground contact state detection on a contact sensor.

[0004] In a first aspect, a method for determining a ground contact state is provided, the method comprising: determining a reaction force of a robot's foot on the ground; estimating, based on a pre-trained probability model, the probability that the foot is in stable contact with the ground when the reaction force is known; and obtaining, based on the probability, the ground contact state of the foot and the ground.

[0005] In the above technical scheme, by determining the reaction force of the robot's foot on the ground, based on the pre-trained probability model, the probability of the robot's foot being in stably contact with the ground when the above reaction force is known is estimated, so that the contact state of the robot's foot and the ground is directly obtained based on the probability, without relying on the detection of the contact sensor, so that the contact state of the robot's foot and the ground can be detected as much as possible without relying on the contact sensor.

[0006] In combination with the first aspect, in some possible implementations, the above-mentioned probability model is trained in the following manner: collecting multiple reaction forces of the robot's foot on the ground under different movement modes and the contact labels corresponding to the above-mentioned multiple reaction forces; wherein the above-mentioned contact label includes a first label for characterizing the stable contact between the above-mentioned foot and the above-mentioned ground and a second label for characterizing the unstable contact between the above-mentioned foot and the above-mentioned ground; the above-mentioned probability model is trained according to the above-mentioned multiple reaction forces and the contact labels corresponding to the above-mentioned multiple reaction forces.

[0007] In the above technical solution, by collecting multiple reaction forces of the robot's feet on the ground in different motion modes and the contact labels corresponding to the above multiple reaction forces, it is beneficial to enable the probability model trained based on these data to accurately detect the robot's contact status in different motion modes, that is, the probability model can be applicable to the detection of the robot's contact status in different motion modes, which is beneficial to improve the scope of application of the probability model.

[0008] In combination with the first aspect, in some possible implementations, the above-mentioned probability model is trained based on the above-mentioned multiple reaction forces and the contact labels corresponding to the above-mentioned multiple reaction forces, including: obtaining the above-mentioned probability model by training on the collected data set through maximum likelihood estimation; wherein the above-mentioned data set includes the above-mentioned multiple reaction forces and the contact labels corresponding to the above-mentioned multiple reaction forces.

[0009] In conjunction with the first aspect, in some possible implementations, the above probability model is expressed as follows:

[0010]

[0011] Among them, p t (L=1|F t ) is the value of the given F t The probability that the foot is in stable contact with the ground is, F t is the reaction force of the foot on the ground, Represents F t The normal component of , β and β0 are weight parameters.

[0012] In combination with the first aspect, in some possible implementations, the above-mentioned determination of the reaction force of the robot's foot on the ground includes: obtaining the joint position and joint velocity of the above-mentioned robot; and determining the reaction force of the robot's foot on the ground according to the above-mentioned joint position and joint velocity and the forward kinematic equation of the above-mentioned robot.

[0013] In combination with the first aspect, in some possible implementations, before determining the reaction force of the robot's feet on the ground, the method further includes: determining whether the robot's feet are provided with contact sensors; if the robot's feet are provided with contact sensors, determining the contact state of the robot's feet with the ground through the contact sensors; if the robot's feet are not provided with contact sensors, determining the reaction force of the robot's feet on the ground.

[0014] In the above technical solution, by adding a step of determining whether the robot's feet are provided with contact sensors, different ground contact state detection methods can be provided based on different determination results, which is beneficial to improving the flexibility of ground contact state detection.

[0015] In combination with the first aspect, in some possible implementations, the robot is a multi-legged robot, the probability that the foot is in stably contact with the ground includes the probability that multiple feet of the multi-legged robot are respectively in stably contact with the ground, and the method further includes: determining the speed of each leg of the multi-legged robot; determining the speed of the multi-legged robot according to the speed of each leg of the multi-legged robot and the probability that multiple feet of the multi-legged robot are respectively in stably contact with the ground.

[0016] In a second aspect, a device for predicting a ground contact state is provided, the device comprising: a determination module for determining the reaction force of a robot's foot on the ground; an estimation module for estimating the probability of the foot being in stable contact with the ground when the reaction force is known based on a pre-trained probability model; and an acquisition module for acquiring the ground contact state of the foot with the ground based on the probability.

[0017] In combination with the second aspect, in some possible implementations, the device also includes: a training module, used to collect multiple reaction forces of the robot's foot on the ground under different movement modes and contact labels corresponding to the above-mentioned multiple reaction forces; wherein the above-mentioned contact labels include a first label for characterizing the stable contact between the above-mentioned foot and the above-mentioned ground and a second label for characterizing the unstable contact between the above-mentioned foot and the above-mentioned ground; the above-mentioned probability model is trained based on the above-mentioned multiple reaction forces and the contact labels corresponding to the above-mentioned multiple reaction forces.

[0018] In combination with the second aspect, in some possible implementations, the training module is specifically used to obtain the above-mentioned probability model by training on the collected data set through maximum likelihood estimation; wherein the above-mentioned data set includes the above-mentioned multiple reaction forces and contact labels corresponding to the above-mentioned multiple reaction forces respectively.

[0019] In conjunction with the second aspect, in some possible implementations, the above probability model is expressed as follows:

[0020]

[0021] Among them, p t (L=1|F t ) is the value of the given F t The probability that the foot is in stable contact with the ground is, F t is the reaction force of the foot on the ground, Represents F t The normal component of , β and β0 are weight parameters.

[0022] In combination with the second aspect, in some possible implementations, the determination module is specifically used to: obtain the joint positions and joint velocities of the above-mentioned robot; and determine the reaction force of the robot's feet on the ground based on the above-mentioned joint positions and joint velocities and the forward kinematics equations of the above-mentioned robot.

[0023] In combination with the second aspect, in some possible implementations, the device also includes: a judgment module, used to determine whether the foot of the above-mentioned robot is provided with a contact sensor; the determination module is specifically used to determine the contact state of the foot of the above-mentioned robot with the above-mentioned ground through the above-mentioned contact sensor when the foot of the above-mentioned robot is provided with a contact sensor; when the foot of the above-mentioned robot is not provided with a contact sensor, determine the reaction force of the robot's foot on the ground.

[0024] In combination with the second aspect, in some possible implementations, the robot is a multi-legged robot, and the probability that the feet are in stably contact with the ground includes the probability that the multiple feet of the multi-legged robot are respectively in stably contact with the ground, and the device also includes: a body speed determination module, used to determine the speed of each leg of the multi-legged robot; determine the speed of the multi-legged robot according to the speed of each leg of the multi-legged robot and the probability that the multiple feet of the multi-legged robot are respectively in stably contact with the ground.

[0025] In a third aspect, a robot is provided, comprising a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the robot executes the method in the first aspect or any possible implementation of the first aspect.

[0026] In a fourth aspect, a computer program product is provided, comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method in the first aspect or any possible implementation of the first aspect.

[0027] In a fifth aspect, a computer-readable storage medium is provided, which stores a computer program code. When the computer program code runs on a computer, the computer executes the method in the above-mentioned first aspect or any possible implementation manner of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a schematic flow chart of a method for determining a ground contact state provided in an embodiment of the present application;

[0029] Figure 2 is a schematic flow chart of another method for determining a ground contact state provided in an embodiment of the present application;

[0030] Figure 3 is a structural schematic diagram of a device for determining a ground contact state provided in an embodiment of the present application;

[0031] Figure 4 It is a schematic diagram of the structure of a robot provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] The technical solution in the present application will be described clearly and in detail below in conjunction with the accompanying drawings. In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B: "and / or" in the text is only a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0033] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0034] Robots are usually expected to be able to perform real-time operations in places that are difficult for humans to reach, such as unmanned areas, mountainous areas, underground environments, etc. This requires accurate real-time estimation of their position, speed, direction and other information. However, robots often encounter slipping soles and uneven ground. Therefore, the traditional assumption that the soles of robots, i.e., their feet, are always in stable contact with the ground without slipping, is often untenable.

[0035] Almost all robot state estimation, motion planning and other functions that rely on the real-time state of the robot (such as position, speed, direction, etc.) will use the robot's leg kinematics and leg odometer (LO) technology. LO technology can extract the robot's body speed from the robot's leg speed (relative to the robot's body coordinate system). The input of LO technology usually includes joint encoders and leg kinematic models. However, LO technology relies on at least one of the robot's feet being in stable contact with the ground. But as mentioned above, the assumption that the robot's sole is always in stable contact with the ground without slipping is often untenable.

[0036] At present, in order to detect the contact state between the robot's feet and the ground, that is, the ground contact state of the robot's feet, it is usually necessary to install additional contact sensors on the soles of the robot's feet. However, due to objective factors such as the robot's movement mode and environment, the contact sensors are usually easy to be damaged. Therefore, how to reduce the dependence of the ground contact state detection on the contact sensors has become a technical problem that needs to be solved urgently.

[0037] In order to solve the above technical problems, the embodiment of the present application provides a method for determining the ground contact state, which is conducive to reducing the dependence of the detection of the ground contact state on the contact sensor, so as to reduce the hardware cost to a certain extent. The above method for determining the ground contact state can be applied to a robot, which can be a multi-legged robot, and the robot can be integrated with a control module, and the control module can include a detection algorithm for the ground contact state. The detection algorithm can be as follows: Figure 1 The process shown is implemented. Through the method in this embodiment, the ground contact state of the robot's feet and the ground can be detected without relying on contact sensors as much as possible.

[0038] Figure 1 It is a schematic flowchart of a method for determining a ground contact state provided in an embodiment of the present application.

[0039] For example, Figure 1 As shown, the method includes:

[0040] Step 101, determining the reaction force of the robot's foot on the ground.

[0041] Step 102 , based on a pre-trained probability model, estimate the probability of the foot being in stable contact with the ground when the reaction force is known.

[0042] Step 103: acquiring the contact state between the foot and the ground according to the probability.

[0043] exist Figure 1 In the illustrated embodiment, the reaction force of the robot's foot on the ground is determined and, based on a pre-trained probability model, the probability of the robot's foot being in stable contact with the ground when the above-mentioned reaction force is known is estimated, thereby directly obtaining the ground contact state of the robot's foot based on the probability without relying on the detection of the contact sensor, thereby enabling the detection of the ground contact state of the robot's foot without relying on the contact sensor as much as possible.

[0044] Below Figure 1 The specific implementation of each step in the embodiment shown is described as follows:

[0045] In step 101 , the reaction force of the robot's foot on the ground is determined.

[0046] The reaction force of the foot on the ground is the reaction force (Ground Reaction Force, GRF) of the robot's foot on the ground when the robot's foot contacts the ground.

[0047] It is understandable that a stable foothold depends on the magnitude of the friction force applied by the robot's foot to the ground, and the friction force is proportional to the normal component of the ground reaction force GRF and the foot-ground static friction coefficient. Generally, the foot-ground static friction coefficient is unknown, and when the foot is not equipped with a contact sensor, it is difficult to directly estimate the GRF. Therefore, in this embodiment, when the robot's foot is not equipped with a contact sensor, the following method for determining the GRF is provided:

[0048] First, the joint positions and joint velocities of the robot are obtained, and then the reaction force of the robot's feet on the ground is determined based on the joint positions and joint velocities and the robot's forward kinematics equation. In the embodiment, the joint positions and joint velocities of the robot can be detected based on the position sensors and velocity sensors set at the joints of the robot.

[0049] Exemplarily, GRF can be calculated by the following formula:

[0050]

[0051] Where p is the joint position, is the joint velocity, J(p) is the partial derivative of the robot's forward kinematics equation, g b represents the acceleration due to gravity, represents the vector determined by centrifugal force, Coriolis force, and gravity, and τ is the control torque of the robot, which can be understood as the control torque of the robot joint. The forward kinematic (FK) equation of the robot leg can be summarized as follows:

[0052] p f =FK(φ,ρ)

[0053] Among them, p f is the position of the foot in the robot body coordinate system, φ is the rotation angle of all joints on one leg of the robot, and ρ represents the generalized kinematic parameters, including the length of the connecting rod between the joints, the axis offset, etc.

[0054] In step 102, based on the probability model obtained through pre-training, the probability of the foot being in stable contact with the ground when the reaction force is known is estimated. Thus, in step 103, the ground contact state of the foot and the ground can be obtained based on the probability of the foot being in stable contact with the ground. The ground contact state of the foot and the ground can be a state of stable contact between the foot and the ground or a state of unstable contact between the foot and the ground. If the estimated probability of stable contact between the foot and the ground is greater than or equal to a preset probability threshold, it can be determined that the ground contact state is a state of stable contact between the foot and the ground. If the estimated probability of stable contact between the foot and the ground is less than the preset probability threshold, it can be determined that the ground contact state is a state of unstable contact between the foot and the ground.

[0055] Exemplarily, the above-mentioned preset probability threshold can be set according to actual needs. For example, the higher the detection accuracy requirement for the ground contact state is, the larger the probability threshold can be set, and the lower the detection accuracy requirement for the ground contact state is, the smaller the probability threshold can be set. Optionally, the probability threshold can be set to be greater than or equal to 0.5, for example, it can be specifically set to 0.6 or 0.7, but this embodiment does not specifically limit this.

[0056] Exemplarily, the above-mentioned preset probability threshold value can also be determined according to the current motion state of the robot within a period of time. If the motion state of the robot is a steady motion, such as walking on a flat ground, the probability threshold value can be relatively small. If the motion state of the robot is a non-steady motion, such as running and jumping on the ground, the probability threshold value can be relatively large. That is to say, a first probability threshold value and a second probability threshold value can be pre-stored inside the robot, and the first probability threshold value is greater than the second probability threshold value. If the current motion state of the robot is a steady motion, the ground contact state of the foot and the ground can be obtained according to the second probability threshold value and the estimated probability of the foot and the ground being in stable contact. If the estimated probability of the foot and the ground being in stable contact is greater than or equal to the second probability threshold value, the ground contact state of the foot and the ground can be determined to be a stable contact state, otherwise the ground contact state of the foot and the ground can be determined to be an unstable contact state. If the current motion state of the robot is a non-steady motion, the ground contact state of the foot and the ground can be obtained according to the first probability threshold value and the estimated probability of the foot and the ground being in stable contact. If the estimated probability of the foot being in stable contact with the ground is greater than or equal to the first probability threshold, the ground contact state of the foot and the ground can be determined to be a stable contact state; otherwise, the ground contact state of the foot and the ground can be determined to be an unstable contact state.

[0057] By setting different probability thresholds for different motion states, and setting a smaller probability threshold for smooth motion and a larger probability threshold for non-steady motion, it is helpful to accurately estimate the ground contact state.

[0058] In this embodiment, the reaction force GRF of the robot's foot on the ground can be used as an input of the probability model, so that the probability model can output the probability of the foot being in stable contact with the ground when the reaction force is known. The probability of the foot being in stable contact with the ground can reflect the actual contact state of the foot with the ground.

[0059] In an exemplary embodiment, the above probability model can be obtained by training from S1 to S2 as follows:

[0060] S1, collecting multiple reaction forces of the robot's feet on the ground under different movement modes and contact labels corresponding to the multiple reaction forces.

[0061] S2: training a probability model according to the multiple reaction forces and the contact labels corresponding to the multiple reaction forces.

[0062] In this embodiment, by collecting multiple reaction forces of the robot's feet on the ground in different motion modes and the contact labels corresponding to the above-mentioned multiple reaction forces, it is beneficial to enable the probability model trained based on these data to accurately detect the robot's contact status in different motion modes, that is, the probability model can be applicable to the detection of the robot's contact status in different motion modes.

[0063] In the above S1, the contact label includes a first label for characterizing stable contact between the foot and the ground and a second label for characterizing unstable contact between the foot and the ground. For example, the first label can be represented by L=1, and the second label can be represented by L=0. The above S1 can be understood as: collecting a number of data samples for training the probability model, and labeling the data samples with sample labels respectively. The several data samples can be understood as the above-mentioned multiple reaction forces of the foot on the ground under different movement modes, and the sample labels can be understood as the above-mentioned contact labels.

[0064] In a specific implementation, in order to further make the applicability of the trained probability model more extensive, during data collection, multiple reaction forces of the robot's feet on different types of ground under different movement modes can be collected. The different types of bottom surfaces may include grass, mountain roads, highways, etc.

[0065] Exemplarily, when determining the contact tags corresponding to the above-mentioned multiple reaction forces, a contact sensor can be used to output the contact tags corresponding to the above-mentioned multiple reaction forces to indicate the ground contact state of the robot's feet represented by different reaction forces of the robot's feet on the ground.

[0066] In the above S2, the multiple reaction forces collected in S1 and the contact labels corresponding to the multiple reaction forces can be used as a data set to train a probability model on the data set.

[0067] Exemplarily, the implementation of S2 may be: obtaining a probability model by training on a collected data set through maximum likelihood estimation, where the data set includes the collected multiple reaction forces and contact tags corresponding to the multiple reaction forces.

[0068] Exemplarily, the expression of the above probability model is as follows:

[0069]

[0070] Among them, p t (L=1|F t ) is the value of the given F t The probability that the foot is in stable contact with the ground, F t is the reaction force of the foot on the ground, Represents F t The normal component of , β and β0 are weight parameters. This probability model can be understood as: a probability model of stable contact between the robot's foot and the ground.

[0071] In a specific implementation, the multiple reaction forces of the robot's feet on the ground under different motion modes collected in S1 and the contact labels corresponding to the multiple reaction forces can be brought into the expression of the above probability model, so as to train the above β and β0. After the above β and β0 are obtained through training, the unknown quantity in the expression of the above probability model is F t The normal component of Therefore, after obtaining the current reaction force of the robot's foot on the ground at the current moment, the normal component of the current reaction force can be determined. Then the normal component of the current reaction force Substitute it into the expression of the above probability model to calculate the probability that the foot is in stable contact with the ground when the current reaction force is known, and then determine the ground contact state of the robot's foot at the current moment based on this probability.

[0072] In this embodiment, the theoretical basis for judging whether the robot's foot is in stable contact with the ground is that a stable foothold depends on the magnitude of the friction force applied by the robot's foot to the ground, and the friction force is proportional to the normal component of the ground reaction force GRF and the foot-ground static friction coefficient. Combining this theoretical basis, the expression of the above-mentioned probability model can be constructed.

[0073] In an exemplary embodiment, before determining the reaction force of the robot's feet on the ground as mentioned above, it may also include: judging whether the robot's feet are provided with contact sensors; if the robot's feet are provided with contact sensors, determining the contact state of the robot's feet with the ground through the contact sensors; if the robot's feet are not provided with contact sensors, determining the reaction force of the robot's feet on the ground.

[0074] Specifically, before executing the above step 101, it can be determined whether the robot's foot is provided with a contact sensor. If a contact sensor is provided, the contact tag output by the contact sensor can be directly obtained. When the contact tag output by the contact sensor is L=1, it can be determined that the contact state of the robot's foot and the ground is a stable contact state. When the contact tag output by the contact sensor is L=0, it can be determined that the contact state of the robot's foot and the ground is an unstable contact state. In other words, if it is determined that the robot's foot is provided with a contact sensor, it is not necessary to perform steps 101 to 103, but to directly detect the contact state based on the contact tag output by the contact sensor. If it is determined that the robot's foot is provided with a contact sensor, steps 101 to 103 can be performed again to estimate the probability of the robot's foot and the ground being in stable contact under the condition of a known reaction force through a probability model, thereby combining the probability to obtain the contact state of the foot and the ground.

[0075] This embodiment can add a step of determining whether the robot's feet are provided with contact sensors, and provide different ground contact state detection methods based on different determination results, which is beneficial to improving the flexibility of ground contact state detection.

[0076] In an exemplary embodiment, the robot is a multi-legged robot, such as a bipedal robot, a quadrupedal robot, etc. The probability that the foot is in stable contact with the ground includes the probability that the multiple feet of the multi-legged robot are in stable contact with the ground, and the method further includes: determining the speed of each leg of the multi-legged robot; determining the speed of the multi-legged robot according to the speed of each leg of the multi-legged robot and the probability that the multiple feet of the multi-legged robot are in stable contact with the ground.

[0077] Specifically, the above steps 101 to 102 can be used to determine the probability that the foot of each leg is in stable contact with the ground. In addition, the speed of each leg can be detected by a speed sensor provided on each leg of the multiple robots. Thus, the multi-legged robot can determine the speed of the multi-legged robot according to the speed of each leg of the multi-legged robot and the probability that the multiple feet of the multi-legged robot are in stable contact with the ground. The speed of the multi-legged robot can be understood as the body speed of the multi-legged robot.

[0078] For example, the speed of the multi-legged robot can be determined by performing a weighted average calculation based on the speed of each leg of the multi-legged robot and the probability that each of the legs of the multi-legged robot is in stable contact with the ground. For example, the body speed of the multi-legged robot can be calculated by the following formula:

[0079]

[0080] in, is the reaction force of the foot of the i-th leg of the known multi-legged robot on the ground at the current time t The probability that the foot of the i-th leg is in stable contact with the ground is, is the speed of the i-th leg of the multi-legged robot at the current time t, i represents the index number of the leg of the multi-legged robot, and C represents the index number set of the leg of the multi-legged robot.

[0081] It should be noted that in this embodiment, the estimated probability of the foot being in stable contact with the ground is used to estimate the robot's body speed. In a specific implementation, it can also be applied to other tasks, such as real-time online state estimation, motion planning and other tasks.

[0082] In an exemplary embodiment, a flowchart of a method for determining a ground contact state can be found at Figure 2 ,include:

[0083] Step 201: Start of ground contact state modeling. The start of ground contact state modeling can be understood as the start of ground contact state detection. For example, a modeling start instruction can be sent to the robot to trigger the robot to start ground contact state detection.

[0084] Step 202: Determine whether the robot's foot is provided with a contact sensor. If yes, execute step 203; otherwise, execute step 204.

[0085] Step 203: Determine the contact state of the robot's foot and the ground through the contact sensor. For example, the contact sensor can directly output a contact tag. If the output contact tag is 1, it means that the contact state is a stable contact state between the foot and the ground. If the output contact tag is 0, it means that the contact state is an unstable contact state between the foot and the ground. In other words, if the robot's foot is provided with a contact sensor, the contact tag is directly read from the contact sensor. The contact tag L=1 represents stable contact or a stable foothold, and the contact tag L=0 represents unstable contact or an unstable foothold.

[0086] Step 204: Execute the algorithm to determine the process of contacting the tag. The process of the algorithm to determine the process of contacting the tag may be steps 101 to 103 described above.

[0087] Step 205: Use the obtained contact tags for subsequent tasks, such as robot body velocity estimation, real-time online state estimation, motion planning, and the like.

[0088] The method for detecting the ground contact state provided in this embodiment is independent of whether the robot's feet are equipped with contact sensors and can detect two common ground contact states: stable contact and unstable contact. This not only reduces the hardware cost, but can also be used as a module to be easily integrated into other multi-legged robot state estimation, motion planning and other frameworks / algorithms without the need for additional hardware equipment, and has strong portability and a wide range of applications.

[0089] The above figures are only schematic illustrations of the steps included in the method according to the exemplary embodiment of the present specification, and are not intended to be limiting. It is easy to understand that the steps shown in the above figures do not indicate or limit the time sequence of these steps. In addition, it is also easy to understand that these steps can be performed synchronously or asynchronously, for example, in multiple modules.

[0090] Figure 3 It is a structural schematic diagram of a device for determining a ground contact state provided in an embodiment of the present application.

[0091] For example, Figure 3 As shown, the device 300 includes: a determination module 301, used to determine the reaction force of the robot's foot on the ground; an estimation module 302, used to estimate the probability of the above-mentioned foot being in stable contact with the above-mentioned ground when the above-mentioned reaction force is known based on a pre-trained probability model; an acquisition module 303, used to acquire the contact state of the above-mentioned foot with the above-mentioned ground based on the above-mentioned probability.

[0092] In one possible implementation, the device 300 also includes: a training module, used to collect multiple reaction forces of the robot's foot on the ground under different movement modes and contact labels corresponding to the above-mentioned multiple reaction forces; wherein the above-mentioned contact label includes a first label for characterizing the stable contact between the above-mentioned foot and the above-mentioned ground and a second label for characterizing the unstable contact between the above-mentioned foot and the above-mentioned ground; according to the above-mentioned multiple reaction forces and the contact labels corresponding to the above-mentioned multiple reaction forces, the above-mentioned probability model is trained.

[0093] In a possible implementation, the training module is specifically used to obtain the above-mentioned probability model by training on the collected data set through maximum likelihood estimation; wherein the above-mentioned data set includes the above-mentioned multiple reaction forces and contact labels corresponding to the above-mentioned multiple reaction forces respectively.

[0094] In a possible implementation, the above probability model is expressed as follows:

[0095]

[0096] Among them, p t (L=1|F t ) is the value of the given F t The probability that the foot is in stable contact with the ground is, F t is the reaction force of the foot on the ground, Represents F t The normal component of , β and β0 are weight parameters.

[0097] In a possible implementation, the determination module 301 is specifically used to: obtain the joint positions and joint velocities of the robot; and determine the reaction force of the robot's feet on the ground based on the joint positions and joint velocities and the forward kinematics equations of the robot.

[0098] In one possible implementation, the device 300 also includes: a judgment module, used to determine whether the foot of the above-mentioned robot is provided with a contact sensor; the determination module 301 is specifically used to determine the contact state of the foot of the above-mentioned robot with the above-mentioned ground through the above-mentioned contact sensor when the foot of the above-mentioned robot is provided with a contact sensor; when the foot of the above-mentioned robot is not provided with a contact sensor, determine the reaction force of the robot's foot on the ground.

[0099] In one possible implementation, the robot is a multi-legged robot, and the probability that the feet are in stably contact with the ground includes the probability that the multiple feet of the multi-legged robot are respectively in stably contact with the ground. The device 300 also includes: a body speed determination module, used to determine the speed of each leg of the multi-legged robot; and determine the speed of the multi-legged robot according to the speed of each leg of the multi-legged robot and the probability that the multiple feet of the multi-legged robot are respectively in stably contact with the ground.

[0100] Figure 4 It is a schematic diagram of the structure of a robot provided in an embodiment of the present application.

[0101] For example, Figure 4 As shown, the robot includes: a memory 401 and a processor 402, wherein the memory 401 stores executable program codes, and the processor 402 is used to call and execute the executable program codes to perform a method for determining a ground contact state.

[0102] In this embodiment, the robot can be divided into functional modules according to the above method example. For example, each functional module can be corresponded, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0103] In the case of dividing each functional module according to each function, the robot may include: a determination module, an estimation module, an acquisition module, etc. It should be noted that all relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, which will not be repeated here.

[0104] The robot provided in this embodiment is used to execute the above-mentioned method for determining the ground contact state, and thus can achieve the same effect as the above-mentioned implementation method.

[0105] In the case of an integrated unit, the robot may include a processing module and a storage module. The processing module may be used to control and manage the robot's actions. The storage module may be used to support the robot in executing mutual program codes and data.

[0106] The processing module may be a processor or a controller, which may implement or execute various exemplary logic blocks, modules and circuits represented in combination with the contents disclosed in this application. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module may be a memory.

[0107] This embodiment also provides a computer-readable storage medium, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement a method for determining a ground contact state in the above-mentioned embodiment.

[0108] This embodiment further provides a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the above-mentioned related steps to implement a method for determining a ground contact state in the above-mentioned embodiment.

[0109] In addition, the robot provided in the embodiments of the present application may specifically be a chip, a component or a module, and the robot may include a connected processor and a memory; wherein the memory is used to store instructions, and when the robot is running, the processor may call and execute the instructions so that the chip executes a method for determining a ground contact state in the above-mentioned embodiments.

[0110] Among them, the robot, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0111] Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0112] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0113] The above contents are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for determining a ground contact state, characterized in that: The method comprises: Determine the reaction force of the robot's feet on the ground; estimating, based on a pre-trained probability model, a probability that the foot is in stable contact with the ground when the reaction force is known; A ground contact state between the foot and the ground is obtained according to the probability.

2. The method according to claim 1, characterized in that The probability model is trained in the following way: Collect multiple reaction forces of the robot's foot on the ground under different motion modes and contact labels corresponding to the multiple reaction forces; wherein the contact label includes a first label for characterizing stable contact between the foot and the ground and a second label for characterizing unstable contact between the foot and the ground; The probability model is obtained by training according to the multiple reaction forces and the contact labels respectively corresponding to the multiple reaction forces.

3. The method according to claim 2, characterized in that The training to obtain the probability model according to the multiple reaction forces and the contact labels respectively corresponding to the multiple reaction forces includes: The probability model is obtained by training on the collected data set through maximum likelihood estimation; wherein the data set includes the multiple reaction forces and the contact tags corresponding to the multiple reaction forces respectively.

4. The method according to claim 3, characterized in that The expression of the probability model is as follows: Among them, p t (L=1|F t ) is the value of the given F t The probability that the foot is in stable contact with the ground, F t is the reaction force of the foot on the ground, Represents F t The normal component of , β and β0 are weight parameters.

5. The method according to claim 1, characterized in that Determining the reaction force of the robot's foot on the ground includes: Obtaining joint positions and joint velocities of the robot; The reaction force of the robot's foot on the ground is determined according to the joint positions and joint velocities and the forward kinematics equation of the robot.

6. The method according to claim 1, characterized in that Before determining the reaction force of the robot's foot on the ground, the method further includes: Determining whether a contact sensor is provided on the foot of the robot; In the case where a contact sensor is provided on the foot of the robot, determining the contact state of the foot of the robot with the ground through the contact sensor; In the case where the robot's foot is not provided with a contact sensor, the reaction force of the robot's foot on the ground is determined.

7. The method according to any one of claims 1 to 6, characterized in that: The robot is a multi-legged robot, the probability that the foot is in stable contact with the ground includes the probability that multiple feet of the multi-legged robot are respectively in stable contact with the ground, and the method further includes: determining a velocity of each leg of the multi-legged robot; The speed of the multi-legged robot is determined according to the speed of each leg of the multi-legged robot and the probability that the multiple feet of the multi-legged robot are respectively in stable contact with the ground.

8. A device for predicting ground contact state, characterized in that: The device comprises: A determination module, used for determining the reaction force of the robot's foot on the ground; An estimation module, configured to estimate the probability of the foot being in stable contact with the ground when the reaction force is known, based on a pre-trained probability model; An acquisition module is used to acquire a ground contact state between the foot and the ground according to the probability.

9. A robot, characterized in that: The robot comprises: A memory for storing executable program codes; A processor, configured to call and run the executable program code from the memory, so that the robot executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the method according to any one of claims 1 to 7 is implemented.

Images