Control method and system of wheel-legged balance robot

Abstract

A control method of a wheel-legged balance robot comprises the following steps that pose information of the robot during movement is obtained, pose changes between adjacent moments are extracted, and the pose changes serve as a first feature; the first feature is input into a pre-trained machine learning model, the posture change trend of the robot in the future is predicted, and the posture change trend comprises natural posture change and large-amplitude posture change; and judging whether the posture change trend is a large-amplitude posture change or not, and if yes, adjusting the advancing speed of the robot at the next moment according to the predicted posture change trend and the real-time advancing speed of the robot. The performance of the robot in a dynamic environment is optimized on the whole, the instability risk is reduced, the reliability of task execution and the user experience are improved, and a solid foundation is laid for wide application of the wheel-legged balance robot.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a control method and system for a wheel-legged balancing robot. Background Art

[0002] A wheel-legged balancing robot is a robot that combines wheeled and legged drive structures, offering excellent maneuverability and stability. Its design typically includes four or more wheels and multiple adjustable leg supports. The wheels provide efficient linear motion and steering capabilities, while the legs help the robot maintain balance on uneven surfaces. In terms of control, the key to a wheel-legged balancing robot is coordinating the movements of the wheels and legs to maintain stability, especially during dynamic motion. The control system typically obtains real-time information about the robot's posture and velocity through sensors (such as gyroscopes, accelerometers, and force sensors), and uses feedback control algorithms (such as PID control and LQR control) to adjust the coordination between the wheels and legs, ensuring that the robot can adjust its posture in a timely manner according to changes in the ground during motion to avoid imbalance. Furthermore, advanced control strategies may also incorporate machine learning algorithms to cope with uncertainty and dynamic changes in complex environments.

[0003] Wheel-legged balancing robots typically move according to speed commands from the user, with the control system adjusting its motion by receiving instructions from a handle, remote control, or other human-machine interface device. However, existing technologies often fail to intelligently adjust the robot's speed when it encounters large changes in posture. If the robot continues to move at the speed set by the user, it may experience drastic fluctuations in its posture. If the speed is not adjusted in time, the robot may not be able to quickly regain balance, leading to loss of control and a fall. This can not only damage the robot itself, but also the surrounding environment and even personnel. This is particularly serious in high-risk scenarios such as military reconnaissance or post-disaster rescue, where the potential harm caused by imbalance is even more severe. Summary of the Invention

[0004] In response to the above-mentioned defects, the purpose of the present invention is to propose a control method and system for a wheel-legged balancing robot. The existing technology is unable to perceive and respond to posture changes in a timely manner, and is unable to adjust the robot's travel speed in a timely manner.

[0005] To achieve this object, the present invention adopts the following technical solution: a control method for a wheel-legged balancing robot, comprising the following steps:

[0006] Step S1: Obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature;

[0007] Step S2: Inputting the first feature into a pre-trained machine learning model to predict the future posture change trend of the robot, wherein the posture change trend includes natural posture change and large-scale posture change;

[0008] Step S3: Determine whether the posture change trend is a large-scale posture change. If so, adjust the robot's next-moment travel speed based on the predicted posture change trend and the robot's real-time travel speed.

[0009] Preferably, the posture change in step S1 includes displacement data and angle data;

[0010] The displacement data is the displacement change rate of the robot's center of gravity in the horizontal and vertical directions;

[0011] The angle data is the rate of change of the robot's tilt angle over time.

[0012] Preferably, before executing step S2, the first feature needs to be processed, wherein the processing process includes: strengthening the displacement data in the form of power exponents and strengthening the angle data in the form of exponential operations;

[0013] The formula for enhancing the displacement data in power exponential form is as follows:

[0014]

[0015] where x i,i+1 ,y i,i+1 are the displacement change rates of the center of gravity in the horizontal and vertical directions, respectively, β is an exponential factor greater than 1, and i represents the moment;

[0016] The enhanced formula for performing exponential operations on angle data is as follows:

[0017] S θi =exp(α·|θ i | δ );

[0018] Among them, α, δ represent the adjustment parameters, θ i Tilt angle change rate.

[0019] Preferably, the steps of predicting the future robot posture change trend in step S2 are as follows:

[0020] Inputting the first feature into a pre-trained machine learning model, and generating a posture change index through the machine learning model;

[0021] Determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

[0022] Preferably, the specific steps of step S3 are as follows:

[0023] Obtaining an evaluation coefficient based on the robot's current travel speed and posture change index;

[0024] The current speed of the robot is adjusted by the evaluation coefficient and the posture change index to obtain the next speed of the robot;

[0025] The evaluation coefficient is obtained as follows:

[0026] Δθ speed-impact =g(V current ,Posture Change );

[0027] Where g() is the functional relationship, V current is the current speed of the robot, Posture Change is the attitude change index;

[0028] The formula for obtaining the robot's next moment's speed is as follows:

[0029] Where k′ is the proportional adjustment coefficient, Δθ ref is the reference threshold of the posture change index.

[0030] A control system for a wheel-legged balancing robot, using the control method for a wheel-legged balancing robot, comprising a feature acquisition module, a judgment module, and an adjustment module;

[0031] The feature acquisition module is used to obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature;

[0032] The judgment module is used to input the first feature into a pre-trained machine learning model to predict the future trend of the robot's posture change;

[0033] The adjustment module is used to determine whether the posture change trend is a large-scale posture change. If so, the robot's next-moment travel speed is adjusted according to the predicted posture change trend and the robot's real-time travel speed.

[0034] Preferably, the feature acquisition module includes a first extraction submodule and a second extraction submodule;

[0035] The first extraction submodule is used to obtain the displacement change rate of the robot's center of gravity in the horizontal and vertical directions;

[0036] The second extraction submodule is used to obtain the rate of change of the robot's tilt angle over time.

[0037] Preferably, an optimization module is further included, and the optimization module is used to enhance the displacement data in a power exponential form and enhance the angle data in an exponential operation form.

[0038] Preferably, the judgment module includes an input submodule and a prediction submodule;

[0039] The input submodule is used to input the first feature into a pre-trained machine learning model, and generate a posture change index through the machine learning model;

[0040] The prediction submodule is used to determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

[0041] Preferably, the regulation module includes an evaluation submodule and an adjustment submodule;

[0042] The estimation module is used to obtain an evaluation coefficient based on the current travel speed and posture change index of the robot;

[0043] The adjustment submodule is used to adjust the current travel speed of the robot through the evaluation coefficient and the posture change index to obtain the robot's travel speed at the next moment.

[0044] One of the above technical solutions has the following advantages or beneficial effects: By acquiring the robot's posture data in real time and extracting key features, the present invention dynamically predicts future posture change trends using a pre-trained machine learning model, enabling early identification of potential balance risks and the implementation of targeted strategies. When posture changes fall within a natural range, the user's control intent is accurately preserved, ensuring the robot's smooth and stable movement. When posture changes significantly, the robot's speed is intelligently adjusted, allowing the robot sufficient time and control margin to correct its posture and avoid loss of control and falls. This not only effectively addresses the shortcomings of existing technologies, which are unable to timely perceive and respond to posture changes, but also significantly enhances the robot's adaptability in high-risk scenarios, such as military reconnaissance, post-disaster rescue, or operations on uneven terrain. Overall, the robot's performance in dynamic environments is optimized, the risk of instability is reduced, and the reliability of task execution and user experience are improved, laying a solid foundation for the widespread application of wheel-legged balancing robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of an embodiment of the method of the present invention.

[0046] Figure 2 It is a structural diagram of an embodiment of the system of the present invention. DETAILED DESCRIPTION

[0047] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention.

[0048] In the description of the embodiments of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the described features. In the description of the embodiments of the present invention, "plurality" means two or more, unless otherwise specifically specified.

[0049] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise specified, "plurality" means two or more. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0050] like Figures 1-2 As shown, a control method for a wheel-legged balancing robot includes the following steps:

[0051] Step S1: Obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature;

[0052] Step S2: Inputting the first feature into a pre-trained machine learning model to predict the future posture change trend of the robot, wherein the posture change trend includes natural posture change and large-scale posture change;

[0053] Step S3: Determine whether the posture change trend is a large-scale posture change. If so, adjust the robot's next-moment travel speed based on the predicted posture change trend and the robot's real-time travel speed.

[0054] The present invention acquires the robot's posture data in real time and extracts key features. Combined with a pre-trained machine learning model, it dynamically predicts future posture change trends, enabling early identification of potential balance risks and the implementation of targeted strategies. When posture changes fall within a natural range, the user's control intent is accurately retained, ensuring the robot's smooth and stable movement. When posture changes significantly, the robot's speed is intelligently adjusted, allowing the robot sufficient time and control margin to correct its posture and avoid loss of control and falls. This not only effectively addresses the inability of existing technologies to perceive and respond to posture changes in a timely manner, but also significantly enhances the robot's adaptability in high-risk scenarios, such as military reconnaissance, post-disaster rescue, or tasks on uneven terrain. Overall, it optimizes the robot's performance in dynamic environments, reduces the risk of instability, and improves the reliability of task execution and user experience, laying a solid foundation for the widespread application of wheel-legged balancing robots.

[0055] Preferably, the posture change in step S1 includes displacement data and angle data;

[0056] The displacement data is the displacement change rate of the robot's center of gravity in the horizontal and vertical directions;

[0057] The angle data is the rate of change of the robot's tilt angle over time.

[0058] Displacement data uses the center of gravity offset quantification value to quantify the rate of change in the horizontal and vertical displacement of the robot's center of gravity due to posture changes or external disturbances during movement, reflecting the robot's posture stability. Angle data uses the tilt rate quantification value to quantify the rate at which the robot's tilt angle changes over time, reflecting the speed of the robot's tilt change during movement and further reflecting the robot's posture changes.

[0059] Preferably, before executing step S2, the first feature needs to be processed, wherein the processing process includes: strengthening the displacement data in the form of power exponents and strengthening the angle data in the form of exponential operations;

[0060] The formula for enhancing the displacement data in power exponential form is as follows:

[0061]

[0062] where x i,i+1 ,y i,i+1 are the displacement change rates of the center of gravity in the horizontal and vertical directions, respectively, β is an exponential factor greater than 1, and i represents the moment;

[0063] After obtaining the displacement change rates in the horizontal and vertical directions, the horizontal and vertical displacement change rates are combined to generate a comprehensive center of gravity offset quantization value, which reflects the displacement change rates of the center of gravity in the horizontal and vertical directions caused by posture changes or external disturbances during the robot's movement. In order to further enhance the sensitivity to large displacement changes, the power exponential form is used to strengthen the response to speed changes.

[0064] The enhanced formula for performing exponential operations on angle data is as follows:

[0065] S θi =exp(α·|θ i | δ );

[0066] Among them, α, δ represent the adjustment parameters, θ i Tilt angle change rate.

[0067] After obtaining the tilt angle change rate between each moment, the tilt angle change rate is converted into a tilt velocity quantization value, and the system's sensitivity to drastic posture changes is enhanced. The present invention specifically uses exponential operation to enhance the change rate.

[0068] Preferably, the steps of predicting the future robot posture change trend in step S2 are as follows:

[0069] Inputting the first feature into a pre-trained machine learning model, and generating a posture change index through the machine learning model;

[0070] Determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

[0071] These pre-trained machine learning models are typically based on supervised learning methods such as regression analysis, support vector machines (SVMs), random forests, or deep learning networks (e.g., LSTM networks). These algorithms extract valuable information from input features, such as quantized values ​​of center of gravity offset (displacement data) and tilt rate (angle data), establishing a mapping relationship between input and output. After training, the model can not only predict the robot's immediate posture changes but also infer future posture changes based on current state trends and generate a posture change index (a measure of the probability of posture change). This predictive model is built based on previously known training data. By drawing on a large amount of data from different scenarios, the model learns the robot's behavior patterns and responses in different states, effectively assessing potential risks and helping the robot adjust its posture control strategy in real time. In practical applications, when input features change, the model can immediately provide feedback on the potential risk level, enabling proactive measures to ensure the robot maintains a stable posture and avoid instability or falls caused by undetected risks.

[0072] The machine learning model is not specifically limited here. Any machine learning model that can perform comprehensive analysis of displacement data and angle to generate a posture change index is acceptable. In order to implement the technical solution of the present invention, the present invention provides a specific implementation method; the expression for generating the posture change index is:

[0073]

[0074] where v x , v y are respectively preset proportional coefficients. It can be seen from the posture change index that the larger the performance value of the center of gravity offset quantization value generated after analyzing the displacement change rate of the robot's center of gravity in the horizontal and vertical directions between adjacent moments, the larger the performance value of the tilt rate quantization value generated after analyzing the rate of change of the robot's tilt angle over time between adjacent moments. That is, the larger the performance value of the posture change index generated when predicting the future posture change trend of the robot through a pre-trained machine learning model, the greater the posture change amplitude of the robot between adjacent moments, and vice versa.

[0075] Preferably, the specific steps of step S3 are as follows:

[0076] Obtaining an evaluation coefficient based on the robot's current travel speed and posture change index;

[0077] The current speed of the robot is adjusted by the evaluation coefficient and the posture change index to obtain the next speed of the robot;

[0078] The evaluation coefficient is obtained as follows:

[0079] Δθ speed-impact =g(V current ,Posture Change );

[0080] Where g() is the functional relationship, V current is the current speed of the robot, Posture Change is the attitude change index;

[0081] The function g(V current ,Posture Change ) is used to evaluate the influence of current speed on posture change, function g(V current ,Posture Change ) can be obtained through experimental data, physical models, and empirical analysis. First, a large amount of experimental data is collected to analyze the robot's motion performance under different travel speeds and posture change indices, and the impact of travel speed on posture change is analyzed. This data can help derive the form of the function g(), which is usually a functional relationship that can be linear, nonlinear, or piecewise. Through deep learning or regression analysis techniques, an expression for this function can be further fitted to accurately describe the impact of travel speed on posture change under different conditions.

[0082] The formula for obtaining the robot's next moment's speed is as follows:

[0083] Where k′ is the proportional adjustment coefficient, Δθ ref is the reference threshold of the posture change index.

[0084] When the robot's posture change is identified as a large change, the robot's current travel speed is adjusted through the evaluation coefficient and the posture change index. The main function of this step is to ensure that the robot can maintain balance in the case of drastic posture changes or unstable external environment, to avoid losing control or falling. Specifically, large posture changes usually indicate that the robot is facing large dynamic disturbances or complex terrain environments. At this time, too fast speed may increase the risk of imbalance, while too slow speed may lead to reduced mission efficiency. Through intelligent speed adjustment, the system can dynamically slow down or speed up the robot's travel speed, giving it enough time to correct its posture while maintaining stability in complex scenarios. This process not only improves the safety of the robot, but also enhances its adaptability in dynamic environments, ensuring that the robot can smoothly complete its mission objectives.

[0085] A control system for a wheel-legged balancing robot, using the control method for a wheel-legged balancing robot, comprising a feature acquisition module, a judgment module, and an adjustment module;

[0086] The feature acquisition module is used to obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature;

[0087] The judgment module is used to input the first feature into a pre-trained machine learning model to predict the future trend of the robot's posture change;

[0088] The adjustment module is used to determine whether the posture change trend is a large-scale posture change. If so, the robot's next-moment travel speed is adjusted according to the predicted posture change trend and the robot's real-time travel speed.

[0089] Preferably, the feature acquisition module includes a first extraction submodule and a second extraction submodule;

[0090] The first extraction submodule is used to obtain the displacement change rate of the robot's center of gravity in the horizontal and vertical directions;

[0091] The second extraction submodule is used to obtain the rate of change of the robot's tilt angle over time.

[0092] Preferably, an optimization module is further included, and the optimization module is used to enhance the displacement data in a power exponential form and enhance the angle data in an exponential operation form.

[0093] Preferably, the judgment module includes an input submodule and a prediction submodule;

[0094] The input submodule is used to input the first feature into a pre-trained machine learning model, and generate a posture change index through the machine learning model;

[0095] The prediction submodule is used to determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

[0096] Preferably, the regulation module includes an evaluation submodule and an adjustment submodule;

[0097] The estimation module is used to obtain an evaluation coefficient based on the current travel speed and posture change index of the robot;

[0098] The adjustment submodule is used to adjust the current travel speed of the robot through the evaluation coefficient and the posture change index to obtain the robot's travel speed at the next moment.

[0099] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative uses of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0100] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. A control method for a wheel-legged balancing robot, characterized in that: The steps include: Step S1: Obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature; Step S2: Inputting the first feature into a pre-trained machine learning model to predict the future posture change trend of the robot, wherein the posture change trend includes natural posture change and large-scale posture change; Step S3: Determine whether the posture change trend is a large-scale posture change. If so, adjust the robot's next-moment travel speed based on the predicted posture change trend and the robot's real-time travel speed.

2. The control method of a wheel-legged balancing robot according to claim 1, characterized in that: The posture change in step S1 includes displacement data and angle data; The displacement data is the displacement change rate of the robot's center of gravity in the horizontal and vertical directions; The angle data is the rate of change of the robot's tilt angle over time.

3. The control method of a wheel-legged balancing robot according to claim 2, characterized in that: Before executing step S2, the first feature needs to be processed, wherein the processing process includes: strengthening the displacement data in the form of power exponents and strengthening the angle data by exponential operations; The formula for enhancing the displacement data in power exponential form is as follows: where x i,i+1 ,y i,i+1 are the displacement change rates of the center of gravity in the horizontal and vertical directions, respectively, β is an exponential factor greater than 1, and i represents the moment; The enhanced formula for performing exponential operations on angle data is as follows: S θi =exp(α·|θ i | δ ); Among them, α, δ represent the adjustment parameters, θ i Tilt angle change rate.

4. The control method of a wheel-legged balancing robot according to claim 2, characterized in that: The steps of predicting the future robot posture change trend in step S2 are as follows: Inputting the first feature into a pre-trained machine learning model, and generating a posture change index through the machine learning model; Determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

5. The control method of a wheel-legged balancing robot according to claim 4, characterized in that: The specific steps of step S3 are as follows: Obtaining an evaluation coefficient based on the robot's current travel speed and posture change index; The current speed of the robot is adjusted by the evaluation coefficient and the posture change index to obtain the next speed of the robot; The evaluation coefficient is obtained as follows: Δθ speed-impact =g(V current ,Posture Change ); Where g() is the functional relationship, V current is the current speed of the robot, Posture Change is the attitude change index; The formula for obtaining the robot's next moment's speed is as follows: Where k′ is the proportional adjustment coefficient, Δθ ref is the reference threshold of the posture change index.

6. A control system for a wheel-legged balancing robot, characterized in that: A control method for a wheel-legged balancing robot according to any one of claims 1 to 5, comprising a feature acquisition module, a judgment module, and an adjustment module; The feature acquisition module is used to obtain the posture information of the robot during movement, extract the posture changes between adjacent moments, and use the posture changes as the first feature; The judgment module is used to input the first feature into a pre-trained machine learning model to predict the future trend of the robot's posture change; The adjustment module is used to determine whether the posture change trend is a large-scale posture change. If so, the robot's next-moment travel speed is adjusted according to the predicted posture change trend and the robot's real-time travel speed.

7. The control system of a wheel-legged balancing robot according to claim 6, characterized in that: The feature acquisition module includes a first extraction submodule and a second extraction submodule; The first extraction submodule is used to obtain the displacement change rate of the robot's center of gravity in the horizontal and vertical directions; The second extraction submodule is used to obtain the rate of change of the robot's tilt angle over time.

8. The control system of a wheel-legged balancing robot according to claim 6, characterized in that: The method further comprises an optimization module, wherein the optimization module is used to perform power exponential enhancement on the displacement data and to perform exponential enhancement on the angle data.

9. The control system of a wheel-legged balancing robot according to claim 6, characterized in that: The judgment module includes an input submodule and a prediction submodule; The input submodule is used to input the first feature into a pre-trained machine learning model, and generate a posture change index through the machine learning model; The prediction submodule is used to determine whether the posture change index is greater than a preset posture change index reference threshold. If the posture change index is greater than the posture change index reference threshold, the current posture change is a large-scale posture change. If the posture change index is less than or equal to the posture change index reference threshold, the current posture change is a natural posture change.

10. The control system of a wheel-legged balancing robot according to claim 6, characterized in that: The regulation module includes an evaluation submodule and an adjustment submodule; The estimation module is used to obtain an evaluation coefficient based on the current travel speed and posture change index of the robot; The adjustment submodule is used to adjust the current travel speed of the robot through the evaluation coefficient and the posture change index to obtain the robot's travel speed at the next moment.

Images