Quadruped robot motion control method and device, electronic equipment, and storage medium

By obtaining the movement data of the four-legged robot, generating the target state space equation and calculating the target position data, the problem of unstable movement of the four-legged robot in complex environments is solved, and stable control is achieved under the target motion mode is improved, and users are encouraged to use the robot Debugging and optimizing capabilities of sports.

CN115454109BActive Publication Date: 2024-11-12CHINA TELECOM CORP LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210982395.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-08-16
Publication Date
2024-11-12
Estimated Expiration
2042-08-16

AI Technical Summary

Technical Problem

The existing four-legged robots are unstable in complex environments and the merchants do not provide complete motion control programs, making it difficult for users to debug or optimize the stability of the robot.

Method used

By obtaining the foot position and posture data of the robot in the target motion mode, generating the target state space equation, calculating the target position data, and controlling the robot's movement based on these data to achieve stable motion.

Benefits of technology

Without knowing the original motion equation of the robot, the motion control of the four-legged robot is optimized to ensure that the robot runs stably under the target motion mode, and improves the flexibility and adaptability of motion control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115454109B_ABST
    Figure CN115454109B_ABST
Patent Text Reader

Abstract

The present application provides a quadruped robot motion control method and device, electronic device, and storage medium, wherein the method includes: obtaining first position data of each foot of the robot and first posture data of the robot in a target motion mode; generating a target state space equation of the robot based on the first position data and the first posture data of each foot; determining the target position data of each foot of the robot based on the target posture data and the target state space equation in the target motion mode; controlling the robot motion based on the target position data of each foot, so that in the target motion mode, each foot of the robot moves according to the corresponding target position data, and a motion effect consistent with the target posture data can be achieved; wherein the target posture data refers to the posture data of the robot when it is running stably in the target motion mode.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of automatic control technology, and in particular to a quadruped robot motion control method and device, electronic equipment, and storage medium. Background Art

[0002] Quadruped mammals in nature are found on all continents of the earth. They have significant advantages in terms of adaptability to terrain and environment, flexibility and efficiency of movement. They can move on almost any land, and can not only move quickly in rugged and complex terrain, but some also have good load-bearing capacity. They adapt to environmental changes in different ways and survive and reproduce in nature.

[0003] Drawing on the unique movement characteristics of quadrupeds, quadruped robots with strong terrain and environment adaptability, high dynamic movement capabilities, and large load capacity are gaining more and more attention in the field of robotics. These characteristics make quadruped robots have good application prospects in the fields of material transportation in unstructured terrain environments, post-disaster rescue, field exploration, and high-risk environment operations.

[0004] Existing bionic quadruped robots can perform basic movements such as standing and walking, and have a certain degree of stability under certain conditions. However, in more complex situations, such as being subjected to large external impacts, ups and downs, and changes in terrain, the original stability of the quadruped robot will be affected, making the movement of the quadruped robot unable to show good stability. Since merchants do not provide complete motion control programs when quadruped robot products are delivered, owners who purchase quadruped robots cannot know the internal motion equations of the quadruped robot, making it difficult to debug or improve the operation of the robot, which in turn results in the inability to optimize the stability of the robot under different conditions after delivery.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] In view of the above problems, the present application is proposed to provide a quadruped robot motion control method and device, electronic device, and storage medium that overcome the above problems or at least partially solve the above problems, including:

[0007] A quadruped robot motion control method, the method comprising:

[0008] Acquire first position data of each foot of the robot in a target motion mode and first posture data of the robot;

[0009] generating a target state space equation of the robot according to the first position data of each foot and the first posture data;

[0010] Determining target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation;

[0011] The robot movement is controlled based on the target position data of each foot.

[0012] Optionally, the obtaining of first position data of each foot of the robot and first posture data of the robot in the target motion mode includes:

[0013] Controlling the robot to move according to a target movement mode;

[0014] During the movement of the robot, the position of each foot of the robot and the position of the target part of the robot are measured to obtain the first position data of each foot of the robot in the target movement mode and the first posture data of the robot.

[0015] Optionally, generating the target state space equation of the robot according to the first position data of each foot and the first posture data includes:

[0016] Using the first position data of each foot as input and the first posture data as output, multiple fitting methods are used to perform fitting respectively to obtain multiple initial state space equations;

[0017] A target state-space equation for the robot is determined from the plurality of initial state-space equations.

[0018] Optionally, determining the target state space equation of the robot from the multiple initial state space equations comprises:

[0019] Acquire second position data of each foot of the robot in the target motion mode and second posture data of the robot;

[0020] Acquire initial posture data outputted by each of the initial state spaces when the second position data is used as input;

[0021] Calculating the difference between the initial posture data outputted by each of the initial state spaces and the second posture data, and determining the minimum difference;

[0022] The initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

[0023] Optionally, determining the initial state space equation corresponding to the minimum difference as the target state space equation of the robot further includes:

[0024] When the minimum difference is less than a preset difference threshold, the initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

[0025] Optionally, determining target position data of each foot of the robot according to the target posture data in the target motion mode and the target state space equation includes:

[0026] Converting the target state space equation into a corresponding transfer function;

[0027] Optimizing the transfer function to obtain a target transfer function;

[0028] The target position data of each foot of the robot is determined according to the target posture data and the target transfer function.

[0029] Optionally, optimizing the transfer function to obtain a target transfer function includes:

[0030] Obtaining a feedback gain matrix of the transfer function by a pole placement method;

[0031] The transfer function is optimized according to the feedback gain matrix to obtain a target transfer function.

[0032] A quadruped robot motion control device, the device comprising:

[0033] A data acquisition module, used to acquire first position data of each foot of the robot and first posture data of the robot in a target motion mode;

[0034] An equation generating module, used for generating a target state space equation of the robot according to the first position data of each foot and the first posture data;

[0035] A data determination module, used to determine the target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation;

[0036] A motion control module is used to control the motion of the robot based on the target position data of each foot.

[0037] Optionally, the data acquisition module includes:

[0038] A first control submodule, used to control the robot to move according to a target movement mode;

[0039] The first acquisition submodule is used to measure the position of each foot of the robot and the position of the target part of the robot during the movement of the robot, so as to obtain the first position data of each foot of the robot in the target movement mode and the first posture data of the robot.

[0040] Optionally, the equation generating module comprises:

[0041] A data fitting submodule, used for taking the first position data of each foot as input and the first posture data as output, and performing fitting respectively by using a plurality of fitting methods to obtain a plurality of initial state space equations;

[0042] A target selection module is used to determine the target state space equation of the robot from the multiple initial state space equations.

[0043] Optionally, the target selection module includes:

[0044] A verification data acquisition submodule, used to acquire second position data of each foot of the robot and second posture data of the robot in the target motion mode;

[0045] A simulation output determination submodule, used for obtaining initial posture data outputted by each of the initial state spaces when the second position data is used as input;

[0046] A difference calculation submodule, used for calculating the difference between the initial posture data outputted by each of the initial state spaces and the second posture data, and determining the minimum difference;

[0047] The target equation determination submodule is used to determine the initial state space equation corresponding to the minimum difference as the target state space equation of the robot.

[0048] Optionally, the target equation determination submodule is specifically used to determine the initial state space equation corresponding to the minimum difference as the target state space equation of the robot when the minimum difference is less than a preset difference threshold.

[0049] Optionally, the data determination module includes:

[0050] A conversion submodule, used to convert the target state space equation into a corresponding transfer function;

[0051] An optimization submodule, used for optimizing the transfer function to obtain a target transfer function;

[0052] The position determination submodule is used to determine the target position data of each foot of the robot according to the target posture data and the target transfer function.

[0053] Optionally, the optimization submodule includes:

[0054] A feedback determination unit, used for obtaining a feedback gain matrix of the transfer function by a pole placement method;

[0055] Based on the feedback optimization unit, it is used to optimize the transfer function according to the feedback gain matrix to obtain the target transfer function.

[0056] An electronic device comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein the computer program implements the steps of the quadruped robot motion control method as described above when executed by the processor.

[0057] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the quadruped robot motion control method as described above are implemented.

[0058] This application has the following advantages:

[0059] In an embodiment of the present application, by acquiring first position data of each foot of the robot and first posture data of the robot in a target motion mode; generating a target state space equation of the robot according to the first position data and the first posture data of each foot; determining the target position data of each foot of the robot according to the target posture data and the target state space equation in the target motion mode; controlling the movement of the robot based on the target position data of each foot, so that in the target motion mode, each foot of the robot moves according to the corresponding target position data, thereby achieving a motion effect consistent with the target posture data; wherein the target posture data refers to the posture data of the robot when it is running stably in the target motion mode. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the description of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0061] Figure 1 A flowchart of a quadruped robot motion control method according to an embodiment of the present application;

[0062] Figure 2 A logic diagram for constructing a target state space equation in an example of the present application;

[0063] Figure 3 A schematic diagram of a process for constructing a target state space equation in an example of the present application;

[0064] Figure 4 This is a schematic diagram of an image of first position data collected in an example of the present application;

[0065] Figure 5 This is a schematic diagram of an image of first posture data collected in an example of the present application;

[0066] Figure 6 A schematic diagram of a unified identification window in an example of the present application;

[0067] Figure 7 A schematic diagram of a simulation environment for verifying the rationality of a target state space equation in an example of the present application;

[0068] Figure 8 for Figure 7 The simulation verification difference diagram shown;

[0069] Fig. 9 A logic diagram of motion control optimization in an example of the present application;

[0070] Fig.10 A schematic diagram of a flow chart of motion control optimization in an example of the present application;

[0071] Fig.11 A schematic diagram of a simulation environment for motion control optimization in an example of the present application;

[0072] Fig.12 for Fig.11 The simulation verification difference diagram shown;

[0073] Fig.13 This is a structural block diagram of a quadruped robot motion control device according to an embodiment of the present application. DETAILED DESCRIPTION

[0074] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0075] For the communications industry, communications operators are facing the transformation from traditional operators to innovative technology companies, and quadrupedal bionic robots are a hot and cutting-edge topic in robotics research at home and abroad. The inventors found that the application scenarios of quadrupedal robots themselves can also be well combined with operators, such as park inspections, special environment detection, data collection, etc. Therefore, the study of quadrupedal robots is also an important part of scientific and technological innovation.

[0076] Existing bionic quadruped robots can achieve the most basic movements such as standing and walking, and can have a certain degree of stability under certain conditions. However, in more complex situations, such as being subjected to large external impacts, ups and downs, terrain changes, etc., the movement of quadruped robots cannot show good stability. Since merchants do not provide complete motion control programs when quadruped robot products are delivered, owners who purchase quadruped robots cannot know the internal motion equations of the quadruped robots, making it difficult to debug or improve the operation of the robots, which in turn results in the inability to optimize the stability of the robots under different conditions after delivery.

[0077] In view of this, the embodiment of the present application proposes a quadruped robot motion control method and device, electronic device, and storage medium. One of the main technical concepts of the embodiment of the present application is that when the robot cannot move stably in the target motion mode, the first position data of each foot and the first posture data of the robot are obtained during the actual motion of the robot in the target motion mode, and then the robot is motion modeled according to the first position data and first posture data of each foot, and the target state space equation of the robot is generated; then, according to the target posture data when the robot is expected to move stably in the target motion mode, and the generated target state space equation, the target position data of each foot of the robot when it moves stably in the target motion mode is calculated; finally, the robot motion is controlled according to the calculated target position data of each foot, so that each foot of the robot moves according to the corresponding target position data, so as to achieve stable motion in the target motion mode; the robot motion is optimized and controlled under the restriction that the original motion equation of the robot is unknown.

[0078] The quadruped robot motion control method provided in the embodiment of the present application can be applied to electronic devices, and can be used to perform motion modeling and optimization on a quadruped robot that has been put into use, so as to realize motion control of the quadruped robot; it can also be used to provide new research ideas for the study of quadruped robot motion control. Among them, the electronic device can be hardware or software. When the electronic device is hardware, it can be implemented as a cluster composed of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the electronic device is software, it can be installed in the hardware devices listed above.

[0079] Exemplarily, the quadruped robot motion control method provided in the embodiment of the present application can be installed in the control module of the robot in the form of software, or can also be installed in the form of software in an electronic device that can communicate with the robot, such as the robot's remote control device.

[0080] Reference Figure 1, shows a flow chart of the steps of a quadruped robot motion control method provided by an embodiment of the present application. In the embodiment of the present application, the method may include the following steps:

[0081] Step 101, obtaining the first position data of each foot of the robot in the target motion mode and the first posture data of the call number robot.

[0082] In the embodiment of the present application, by controlling the robot to move in a target motion mode, the first position data of each foot changing with time and the first posture data of the robot as a whole changing with time during the actual motion process of the robot can be obtained.

[0083] Step 102: Generate a target state space equation of the robot according to the first position data of each foot and the first posture data.

[0084] By collecting multiple sets of first position data and first posture data, the target state space equation of the robot can be calculated by fitting or building a robot dynamics model.

[0085] Step 103, determining the target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation.

[0086] The target posture data refers to the overall posture data of the robot changing over time when it moves stably in the target motion mode. After obtaining the state space equation of the robot, combined with the target posture data, the target position data of each foot of the robot corresponding to the target posture data in the target motion mode can be calculated; that is, when each foot of the robot moves according to the target position data, it can achieve stable motion in the target motion mode.

[0087] Step 104: Control the movement of the robot based on the target position data of each foot.

[0088] After determining the target position data of each foot of the robot corresponding to the target posture data, by controlling each foot of the robot to move according to the target position data, the robot can be made to move stably in the target motion mode, thereby optimizing the motion control of the robot.

[0089] The embodiment of the present application obtains the first position data of each foot of the robot and the first posture data of the robot in the target motion mode; generates the target state space equation of the robot according to the first position data and the first posture data of each foot; determines the target position data of each foot of the robot according to the target posture data and the target state space equation in the target motion mode; controls the movement of the robot based on the target position data of each foot, so that in the target motion mode, each foot of the robot moves according to the corresponding target position data, thereby achieving a motion effect consistent with the target posture data; wherein the target posture data refers to the posture data of the robot when it is running stably in the target motion mode.

[0090] By using the embodiments of the present application, when the robot moves unstably in the target motion mode and under the restriction condition that the robot's motion equation is unknown, the target state space equation of the robot can be generated by obtaining the actual motion state of the robot in the target motion mode, and then based on the target state space equation and the target posture data during stable motion in the target motion mode, the target position data of each leg of the robot when it is to achieve stable motion in the target motion mode can be obtained, and then by controlling each leg of the robot to move according to the target position data, the robot can achieve stable motion in the target motion mode.

[0091] Next, the motion control method of the quadruped robot in this exemplary embodiment will be further described.

[0092] In step 101, first position data of each foot of the robot in a target motion mode and first posture data of the robot are obtained.

[0093] In the embodiment of the present application, the target motion mode can be set by the user according to actual needs. When the embodiment of the present application is used in a scenario where the robot's internal motion equation is unknown and the robot is found to be unable to move stably, the target motion mode generally selects the motion mode when the robot is unable to move stably as expected.

[0094] Exemplarily, the robot's movement modes may include linear movement, curved movement, uphill movement, etc.; when the robot is performing linear movement, an offset occurs and it cannot perform linear movement stably. At this time, in order to optimize the control of the robot's linear movement, the linear movement can be determined as the target movement mode.

[0095] Both the first position data and the first posture data are related to time. That is to say, in the embodiment of the present application, the first position data of each foot and the first posture data of the robot when the robot actually moves in the target motion mode can be obtained at specified time intervals.

[0096] The first posture data is used to represent the motion state of the robot as a whole, and the motion state of the robot as a whole is associated with the motion state of each foot of the robot, that is, the first posture data can be calculated by the first position data of each foot and the preset corresponding relationship. In one example, the motion state of a part (target part) on the robot body can be used to represent the motion state of the robot as a whole, for example, the target part can be the center of gravity of the robot.

[0097] In an optional embodiment of the present application, the obtaining of the first position data of each foot of the robot and the first posture data of the robot in the target motion mode may include:

[0098] Controlling the robot to move according to a target movement mode;

[0099] During the movement of the robot, the position of each foot of the robot and the position of the target part of the robot are measured to obtain the first position data of each foot of the robot in the target movement mode and the first posture data of the robot.

[0100] The robot can be controlled by a remote control device, that is, the robot can be connected to the remote control device in communication to receive instructions from the remote control device, and then perform corresponding movements according to the received instructions. For example, the remote control device can pre-store instructions corresponding to different movement modes, and after determining the target movement mode, the remote control device can send a target instruction corresponding to the target movement mode to the robot, so that the robot moves according to the target movement mode.

[0101] During the movement of the robot, the position of each foot of the robot and the position of the target part are continuously measured at specified time intervals to obtain the first position data of each foot of the robot and the first posture data of the robot in the target movement mode.

[0102] Exemplarily, a device executing the quadruped robot motion control method of the present application can communicate with sensors inside the robot, and the sensors inside the robot collect the position of each foot and the position of the target part during the robot's motion process to obtain the first position data of each foot of the robot in the target motion mode and the first posture data of the robot.

[0103] In step 102, a target state space equation of the robot is generated according to the first position data of each foot and the first posture data.

[0104] After collecting the first position data of each foot of the robot in the target motion mode and the first posture data of the robot, the corresponding state space equation can be obtained by data fitting.

[0105] Exemplarily, the above-mentioned generating the target state space equation of the robot according to the first position data of each foot and the first posture data may include:

[0106] Using the first position data of each foot as input and the first posture data as output, multiple fitting methods are used to perform fitting respectively to obtain multiple initial state space equations;

[0107] A target state-space equation for the robot is determined from the plurality of initial state-space equations.

[0108] In this embodiment, a variety of fitting methods can be used to fit the collected first position data and first posture data to obtain multiple initial state space equations, and then one of the multiple initial state space equations is selected as the target state space equation of the robot, thereby improving the rationality of the obtained target state space equation.

[0109] Exemplarily, the above-mentioned determining the target state space equation of the robot from the multiple initial state space equations may include:

[0110] Acquire second position data of each foot of the robot in the target motion mode and second posture data of the robot;

[0111] Acquire initial posture data outputted by each of the initial state spaces when the second position data is used as input;

[0112] Calculating the difference between the initial posture data outputted by each of the initial state spaces and the second posture data, and determining the minimum difference;

[0113] The initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

[0114] In this embodiment, the first position data and the first posture data can be understood as a test data set, or training data set, for constructing an initial state space equation, and the second position data and the second posture data can be understood as a verification data set for verifying the rationality of the constructed initial state space equation. The method for obtaining the second position data and the second posture data is the same as the method for obtaining the first position data and the first posture data, which will not be repeated here.

[0115] By obtaining the second position data of each foot of the robot under the target motion mode and the second posture data of the robot, the second position data is used as the input of the constructed initial state-space equation, and the output corresponding to the initial state-space equation (i.e., the initial posture data) is calculated. Then, the output of the initial state-space equation is compared with the corresponding second posture data, and the rationality of the initial space equation is verified according to the comparison result.

[0116] Since there may be multiple initial state-space equations generated by this embodiment, the difference between the initial posture data output by each initial state-space equation and the corresponding second posture data can be calculated. The smaller the difference, the more accurate the fitting result. Then, the minimum difference is determined from the differences corresponding to each initial state-space equation. Finally, the initial state-space equation corresponding to the minimum difference is determined as the target state-space equation of the robot.

[0117] It should be noted that the deviations between each initial state space equation and the first position data and the first posture data can also be compared in a visual form to determine the target state space equation with the smallest deviation from the multiple initial state space equations. Alternatively, the first position data can be directly used as the input of each initial state space equation, the predicted posture data output by each initial state space equation is calculated, the predicted posture data is subtracted from the corresponding first posture data, and the initial state space equation with the smallest overall difference is determined as the target state space equation.

[0118] In order to further improve the rationality of the target state space equation, in the process of determining the initial state space equation corresponding to the minimum difference as the target state space equation of the robot, the following steps may also be included:

[0119] When the minimum difference is less than a preset difference threshold, the initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

[0120] In this embodiment, the initial state space equation with the minimum difference, i.e., the candidate state space equation, can be first determined from multiple initial state space equations, and then it can be determined whether the difference corresponding to the candidate state space equation is less than the preset difference threshold. If so, it means that the candidate state space equation can reasonably represent the motion state of the robot currently executing the target motion mode, so the candidate state space equation is determined as the target state space equation of the robot. If not, it is necessary to re-execute the above process of collecting data, fitting data, and screening state space equations. .

[0121] In step 103, target position data of each foot of the robot is determined according to the target posture data in the target motion mode and the target state space equation.

[0122] After determining the target state space equation of the robot in the target motion mode, the target posture data is used as the output of the target state space equation, and the input corresponding to the target state space equation, that is, the target position data of each foot of the robot, can be calculated. It can be understood that the target posture data is a sign of the robot's stable motion in the target motion mode, and the calculated target position data is the premise for the robot to achieve stable motion in the target motion mode. That is to say, when each foot of the robot moves according to the corresponding target position data, stable motion in the target motion mode can be achieved.

[0123] In an optional embodiment of the present application, the target position data of each foot of the robot may be determined according to the target posture data and the target state space equation under the target motion mode, and may include:

[0124] Converting the target state space equation into a corresponding transfer function;

[0125] Optimizing the transfer function to obtain a target transfer function;

[0126] The target position data of each foot of the robot is determined according to the target posture data and the target transfer function.

[0127] In order to facilitate the calculation and optimize the control of the robot's motion, after obtaining the robot's target state space equation, the target state space equation can be converted into the form of a corresponding transfer function; then the transfer function is optimized to obtain the target transfer function; finally, the target position data of each foot of the robot is determined based on the target posture data and the target transfer function.

[0128] In an optional embodiment, the process of optimizing the transfer function to obtain the target transfer function may include:

[0129] Obtaining a feedback gain matrix of the transfer function by a pole placement method;

[0130] The transfer function is optimized according to the feedback gain matrix to obtain a target transfer function.

[0131] In the process of optimizing the transfer function, the present embodiment uses the pole placement method to obtain the feedback gain matrix to optimize the transfer function. Compared with other feedback control methods, this is more convenient and simpler and can achieve better results.

[0132] In step 104, the movement of the robot is controlled based on the target position data of each foot.

[0133] When each foot of the robot moves according to the corresponding target position data, it can ensure that the robot moves stably in the target motion mode. Taking the target motion mode as a linear motion mode as an example, when each foot of the robot moves according to the corresponding target position data, it can ensure that the robot keeps walking in a straight line in the linear motion mode. Therefore, after determining the target position data of each foot, the robot movement is controlled according to the target position data, so that the robot can move stably in the target motion mode, that is, the movement of the robot is optimized.

[0134] In order to make each foot of the robot move according to the corresponding target position data, in one example, after determining the target position data of each foot, the target motor parameters corresponding to the target position data can be determined based on the relationship between the motor parameters of the motor driving the robot to move and the position data of each foot, and then the motor is controlled to drive the robot to move according to the target motor parameters, so that each foot of the robot moves according to the corresponding target position data.

[0135] By using the embodiments of the present application, when the robot moves unstably in the target motion mode and under the restriction condition that the robot's motion equation is unknown, the target state space equation of the robot can be generated by obtaining the actual motion state of the robot in the target motion mode, and then based on the target state space equation and the target posture data during stable motion in the target motion mode, the target position data of each leg of the robot when it is to achieve stable motion in the target motion mode can be obtained, and then by controlling each leg of the robot to move according to the target position data, the robot can achieve stable motion in the target motion mode.

[0136] In order to facilitate the understanding of the present invention by those skilled in the art, Figure 2-Figure 12 The motion control method of a quadruped robot provided in one embodiment of the present application is exemplarily described and illustrated.

[0137] This embodiment performs motion optimization control on a quadruped robot whose internal motion equation cannot be known due to software or hardware restrictions. When the quadruped robot cannot stably move in the target motion mode, the motion of the quadruped robot can be optimized by collecting data, constructing the target state space equation of the quadruped robot, and using a motion control method based on the target state space equation, thereby achieving optimal control of the motion of the quadruped robot when the algorithm and control method of the quadruped robot are restricted by the manufacturer.

[0138] Therefore, the motion control method of the quadruped robot involved in this embodiment mainly includes two modules: constructing the target state space equation and motion control optimization.

[0139] like Figure 2-Figure 3 As shown, the logic of constructing the target state space equation is as follows Figure 2As shown in the figure, the collected data is divided into a test data set and a verification data set, a state space equation is generated based on the test data set, the input data of the verification data set is input into the state space equation, a simulation is performed, and the corresponding simulation output is obtained. The rationality of the generated state space equation is evaluated by comparing the simulation output with the data in the verification data set. Specifically, Figure 3 Steps 301-303 shown:

[0140] Step 301, collect data.

[0141] In this step, the collected data includes a test data set for constructing the state space equation of the quadruped robot and a verification data set for verifying the rationality of the constructed state space equation. Wherein, the test data set and the verification data set are both collected during the movement process of the quadruped robot performing the target movement mode, including the position data of the four legs relative to the body of the quadruped robot (i.e., the first position data and the second position data), and the posture data of the robot (i.e., the first posture data and the second posture data). Wherein, the first position data and the first posture data constitute the test data set, and the second position data and the second posture data constitute the verification data set.

[0142] Specifically, the remote controller can be used to control the quadruped robot to move in the target motion mode. For ease of understanding, the following description is made using the target motion mode being a linear motion mode as an example, and the collected data is filled in Table 1.

[0143] Table 1

[0144] Column 1 Column 2 Column 3 Column 4 Column 5 Column 6 Column 7 Column 8 Column 9 Column 10 Column 11 Column 12 Column 13 Column 14 Column 15

[0145] In Table 1, each row is used to record a set of data collected at a specified time interval. A set of data includes 15 data, and the collected position data of the four feet of the robot, each foot position data includes three directions, that is, the collected position data of the four feet of the quadruped robot in three directions are filled in the columns corresponding to columns 1 to 12 of Table 1, and the collected robot posture data, the posture data includes three directions, that is, the collected posture data of the quadruped robot in three directions are recorded in the columns corresponding to columns 13 to 15 of Table 1.

[0146] In one example, the collected data can be output as a matrix to the MATLAB Workspace, where the input u is an n*12 matrix, the output y is an n*3 matrix, and n is the number of data sets.

[0147] For example, when the quadruped robot is controlled to move in a straight line, the image of the first position data collected is as follows: Figure 4 As shown; the image of the first posture data collected is as follows Figure 5 shown. Figure 4 In the figure, the horizontal axis represents the number of data, and the vertical axis represents the position of each foot relative to the body, that is, the first position data of each foot, wherein line 1 represents: the x direction of the left front foot; line 2 represents: the x direction of the right front foot; line 3 represents: the y direction of the right back foot; line 4 represents: the y direction of the right front foot; line 5 represents: the y direction of the left back foot; line 6 represents: the y direction of the left front foot; line 7 represents: the x direction of the right back foot; line 8 represents: the x direction of the left back foot; line 9 represents: the z direction of the left front foot; line 10 represents: the z direction of the right back foot; line 11 represents: the z direction of the right front foot; line 12 represents: the direction of the left back foot. Figure 5 In FIG. 1 , the horizontal axis represents the number of data, and the vertical axis represents the posture of the quadruped robot, that is, the first posture data, wherein line 1 represents the x direction, line 2 represents the y direction, and line 3 represents the z direction.

[0148] Step 302: fitting state space equation parameters.

[0149] The general representation of the state-space equation is as follows:

[0150]

[0151] y=Cx(t)+Du(t);

[0152] Where u(t) is the input, x(t) is the state quantity, y is the output, and A, B, C, and D are the parameters that need to be determined.

[0153] For quadruped robots, since input and output need to be directly or indirectly related, the input, state and output have the following meanings:

[0154] Input u(t): The position data of the four feet of the quadruped robot in the target running state, that is, the position of the four feet relative to the body (that is, the distance relative to the main coordinate system, including the three directions of x, y, and z), which is a function that changes with time t. There are 12 in total, representing the position of each foot in three directions.

[0155] State quantity x(t): The posture data of the quadruped robot, which is equal to y when the noise is not considered when establishing the state space equation.

[0156] Output y: The posture data of the quadruped robot (including the three directions of x, y, and z), which is a function that changes with time t. There are 3 in total, representing the overall position of the quadruped robot in space.

[0157] In one example, the collected first position data and first posture data can be input into MATLAB software, and the system identification window can be opened using the "system identification" command provided by MATLAB software. Figure 6 The collected first position data and first posture data are imported into this function to calculate the parameters A, B, C, and D of the state space equation, thereby obtaining the state space equation of the robot in the target running state.

[0158] Since the parameters of the state space equation calculated by different fitting methods are generally different, multiple sets of parameters A, B, C, D can be obtained, that is, multiple initial state space equations can be obtained. According to the data fitting procedures of each initial state space equation, the target state space equation with the best data fitting degree is determined.

[0159] Step 303: verify the rationality of the target state space equation through simulation.

[0160] The parameters of the target state space equation and the input of the verification data set (i.e., the second position data) are placed in a visual simulation tool (such as Simulink) to build a simulation environment. The output (i.e., the initial posture data) is calculated through the target state space equation, and finally compared with the output of the verification data set (i.e., the second posture data) to evaluate the rationality of the equation.

[0161] Taking Simulink as an example, the simulation environment is as follows Figure 7 As shown. Wherein, u_ is the input of the validation data set, and y_ is the output of the validation data set. The input of the validation data set is used as the input of the target state space equation, and the oscilloscope 1 is used to display the simulation output of the target state space equation based on the input of the validation data set. The oscilloscope 2 is used to display the output of the validation data set. The rationality of the equation is evaluated by the difference method, and the oscilloscope 3 is used to display the difference between the output of the validation data set and the simulation output. In one example, the difference image obtained is as follows: Figure 8 shown. Figure 8 In the figure, the horizontal axis represents time, the vertical axis represents difference; line 1 represents the x direction, line 2 represents the y direction, and line 3 represents the z direction. Figure 8 It can be seen from the comparison that the difference between the simulation output and the output of the verification data set is basically close to 0, indicating that the established target state space equation basically conforms to the actual motion trajectory.

[0162] After constructing the target state space equation of the quadruped robot, the quadruped robot motion is further optimized based on the target state space equation. In order to make the output of the equation more accurate, basic feedback control is adopted. The feedback control logic is as follows: Fig. 9As shown in the figure, the basic idea of ​​feedback control is: the feedback gain matrix K is obtained by the pole configuration method, the difference between the output of the target state space equation and the expected value is multiplied by the feedback gain matrix K and C and fed back to the input of the target state space equation, thereby correcting the input of the target state space equation to correct the output of the target state space equation so that the difference between it and the actual output is minimized.

[0163] The motion control optimization process, i.e. the feedback control process, specifically includes the following: Fig.10 Steps 1001-1003 shown

[0164] Step 1001, deriving a transfer function based on a target state space equation.

[0165] After determining the target state space equation, the target state space equation can be converted into the form of a transfer function. In one example, this can be achieved by the command "[num, den] = ss2tf (A, B, C, D, J)" in MATLAB. Among them, num is the numerator parameter of the transfer function, den is the denominator parameter of the transfer function, A, B, C, D are the parameters of the space state equation, and J is the number of inputs: 12. In one example, the transfer function calculated by MATLAB is as follows:

[0166]

[0167] Step 1002: Obtain an output feedback gain matrix by pole placement method.

[0168] In order to further improve the accuracy of the target state-space equation, the transfer function needs to be optimized. Specifically, the feedback control method can be used to optimize the transfer function. The pole configuration method is more convenient and simple to obtain the feedback gain matrix, and can achieve better results. Therefore, in this example, the output feedback gain matrix is ​​obtained by the pole configuration method; of course, in other examples, the transfer function can also be optimized by other feedback control methods.

[0169] In one example, it can be implemented by the command "K=place(A, B, p)" in MATLAB, where K is the feedback gain matrix, A and B are the parameters of the spatial state equation, and p is the pole vector. K calculated by MATLAB is as follows:

[0170]

[0171] Step 1003: simulation verification.

[0172] The feedback gain matrix, the target state space equation and the input of the verification data set (i.e., the second position data) are placed in a visual simulation tool (such as Simulink) to build a simulation environment. The input of the target state space equation is jointly determined by the input of the verification data set and the feedback input controlled by the feedback gain matrix. The simulation output is calculated by the target state space equation. Finally, the target state space equation is optimized based on the comparison between the simulation output and the output of the verification data set (i.e., the second posture data), that is, the transfer function is optimized to obtain the target transfer function.

[0173] Taking Simulink as an example, the simulation environment is as follows Fig.11 As shown. Where u_ is the input of the verification data, y_ is the output of the verification data, the simulation output calculated by the target state space equation is displayed by oscilloscope 4; the output of the verification data is displayed by oscilloscope 5, the difference between the simulation output and the output of the verification data is displayed by oscilloscope 6; the feedback input is displayed by oscilloscope 7. In one example, the difference image obtained is as follows Fig.12 In the figure, the horizontal axis represents time, the vertical axis represents difference; line 1 represents the x direction, line 2 represents the y direction, and line 3 represents the z direction.

[0174] from Fig.12 It can be seen that through feedback control, the output of the target state space equation is closer to the actual output. Among them, the difference between line 1 and line 3 is almost 0, indicating that under feedback control, the output of the target state space equation in the x and z directions is basically consistent with the actual output, and the posture of the quadruped robot can be corrected according to the actual output feedback.

[0175] It should be noted that, for the method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited by the described order of actions, because according to the embodiments of the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present application.

[0176] Reference Fig.13 , shows a structural block diagram of an embodiment of a quadruped robot motion control device of the present application, corresponding to the above-mentioned quadruped robot motion control method embodiment, in the embodiment of the present application, the device may include the following modules:

[0177] A data acquisition module 1301 is used to acquire first position data of each foot of the robot and first posture data of the robot in a target motion mode;

[0178] An equation generating module 1302, configured to generate a target state space equation of the robot according to the first position data of each foot and the first posture data;

[0179] A data determination module 1303 is used to determine the target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation;

[0180] The motion control module 1304 is used to control the motion of the robot based on the target position data of each foot.

[0181] Optionally, the data acquisition module 1301 may include:

[0182] A first control submodule, used to control the robot to move according to a target movement mode;

[0183] The first acquisition submodule is used to measure the position of each foot of the robot and the position of the target part of the robot during the movement of the robot, so as to obtain the first position data of each foot of the robot in the target movement mode and the first posture data of the robot.

[0184] Optionally, the equation generating module 1302 may include:

[0185] A data fitting submodule, used for taking the first position data of each foot as input and the first posture data as output, and performing fitting respectively by using a plurality of fitting methods to obtain a plurality of initial state space equations;

[0186] A target selection module is used to determine the target state space equation of the robot from the multiple initial state space equations.

[0187] Optionally, the target selection module may include:

[0188] A verification data acquisition submodule, used to acquire second position data of each foot of the robot and second posture data of the robot in the target motion mode;

[0189] A simulation output determination submodule, used for obtaining initial posture data outputted by each of the initial state spaces when the second position data is used as input;

[0190] A difference calculation submodule, used for calculating the difference between the initial posture data outputted by each of the initial state spaces and the second posture data, and determining the minimum difference;

[0191] The target equation determination submodule is used to determine the initial state space equation corresponding to the minimum difference as the target state space equation of the robot.

[0192] Optionally, the target equation determination submodule is specifically used to determine the initial state space equation corresponding to the minimum difference as the target state space equation of the robot when the minimum difference is less than a preset difference threshold.

[0193] Optionally, the data determination module 1303 may include:

[0194] A conversion submodule, used to convert the target state space equation into a corresponding transfer function;

[0195] An optimization submodule, used for optimizing the transfer function to obtain a target transfer function;

[0196] The position determination submodule is used to determine the target position data of each foot of the robot according to the target posture data and the target transfer function.

[0197] Optionally, the optimization submodule may include:

[0198] A feedback determination unit, used for obtaining a feedback gain matrix of the transfer function by a pole placement method;

[0199] Based on the feedback optimization unit, it is used to optimize the transfer function according to the feedback gain matrix to obtain the target transfer function.

[0200] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0201] An embodiment of the present application also discloses an electronic device, including a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program implements the steps of the quadruped robot motion control method as described above when executed by the processor.

[0202] The embodiment of the present application also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the quadruped robot motion control method as described above are implemented.

[0203] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0204] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, devices, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0205] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0206] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0207] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0208] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0209] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0210] The above is a detailed introduction to a quadruped robot motion control method and device, electronic device and storage medium provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A quadruped robot motion control method, characterized in that: The method comprises: Acquire first position data of each foot of the robot in a target motion mode and first posture data of the robot; generating a target state space equation of the robot according to the first position data of each foot and the first posture data; Determining target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation; Controlling the movement of the robot based on the target position data of each foot; The step of generating the target state space equation of the robot according to the first position data of each foot and the first posture data comprises: Using the first position data of each foot as input and the first posture data as output, multiple fitting methods are used to perform fitting respectively to obtain multiple initial state space equations; A target state-space equation for the robot is determined from the plurality of initial state-space equations.

2. The method according to claim 1, characterized in that The step of obtaining first position data of each foot of the robot and first posture data of the robot in the target motion mode includes: Controlling the robot to move according to a target movement mode; During the movement of the robot, the position of each foot of the robot and the position of the target part of the robot are measured to obtain the first position data of each foot of the robot in the target movement mode and the first posture data of the robot.

3. The method according to claim 1, characterized in that Determining a target state space equation of the robot from the plurality of initial state space equations comprises: Acquire second position data of each foot of the robot in the target motion mode and second posture data of the robot; Acquire initial posture data outputted by each of the initial state spaces when the second position data is used as input; Calculating the difference between the initial posture data outputted by each of the initial state spaces and the second posture data, and determining the minimum difference; The initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

4. The method according to claim 3, characterized in that The initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot, further comprising: When the minimum difference is less than a preset difference threshold, the initial state space equation corresponding to the minimum difference is determined as the target state space equation of the robot.

5. The method according to any one of claims 1 to 4, characterized in that Determining target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation includes: Converting the target state space equation into a corresponding transfer function; Optimizing the transfer function to obtain a target transfer function; The target position data of each foot of the robot is determined according to the target posture data and the target transfer function.

6. The method according to claim 5, characterized in that The step of optimizing the transfer function to obtain a target transfer function includes: Obtaining a feedback gain matrix of the transfer function by a pole placement method; The transfer function is optimized according to the feedback gain matrix to obtain a target transfer function.

7. A quadruped robot motion control device, characterized in that: The device comprises: A data acquisition module, used to acquire first position data of each foot of the robot and first posture data of the robot in a target motion mode; An equation generating module, used for generating a target state space equation of the robot according to the first position data of each foot and the first posture data; A data determination module, used to determine the target position data of each foot of the robot according to the target posture data under the target motion mode and the target state space equation; A motion control module, used for controlling the motion of the robot based on the target position data of each foot; The equation generating module comprises: A data fitting submodule, used for taking the first position data of each foot as input and the first posture data as output, and performing fitting respectively by using a plurality of fitting methods to obtain a plurality of initial state space equations; A target selection module is used to determine the target state space equation of the robot from the multiple initial state space equations.

8. An electronic device, characterized in that: The method comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, the steps of the quadruped robot motion control method as claimed in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the quadruped robot motion control method according to any one of claims 1 to 6 are implemented.