Wheel-legged robot reconstruction stability control method considering foot wheel rigidity influence

Abstract

The invention discloses a wheel-legged robot reconstruction stability control method considering foot wheel rigidity influence. The method comprises the following steps: 1, establishing a wheel-legged robot coordinate system; 2, the deviation between the ideal ZMP and the actual ZMP and the deviation change rate are input into a fuzzy controller, and the ideal displacement of the movement of the mass center adjusting sliding block in the wheel-legged robot reconstruction process is calculated and output; 3, differentiating the ideal displacement of the sliding block to obtain an ideal speed, inputting the ideal displacement and the ideal speed in the reconstruction process into an LQG controller to obtain a control voltage, and inputting the control voltage into a dynamic model of a centroid adjusting mechanism to obtain the actual displacement of the centroid adjusting sliding block in the reconstruction process; and 4, calculating an actual ZMP value of the wheel-legged robot, making a difference between the actual ZMP value and the ideal ZMP value, and feeding the difference back to the fuzzy controller to form stable closed-loop control. According to the method, the actual ZMP position in the robot reconstruction process can approach to the ideal position under the influence of the foot wheel rigidity, so that the stability of the system during reconstruction is improved.

Description

Technical Field

[0001] This invention belongs to the technical field of robot stability control, specifically relating to a stability control method for the reconfiguration process of a wheeled robot. Background Technology

[0002] This wheel-legged robot is a novel type of wheel-legged robot based on an automotive structure design, capable of transforming between automotive and humanoid configurations. It can travel at high speeds on flat urban roads in automotive mode, and then reconstruct into a humanoid mode to enter buildings, using gait and stair climbing to reach rooms quickly, achieving point-to-point access. Currently, research directions for wheel-legged robots mainly include: mechanical structure and drive design, motion planning and control, environmental perception and intelligent decision-making, and embodied intelligence. Traditional wheel-legged robots undergo partial reconstruction between wheeled and legged configurations, meaning there are no significant differences in shape, structure, or center of mass before and after reconstruction. Many scholars only analyze and control the wheeled or legged movement of wheel-legged robots, with limited research on the motion stability issues during the switching between different motion configurations. Regarding the adjustment of robot stability, existing research mostly achieves balance by adjusting the rotation law of the joint motors to deform the leg mechanism and adjust the robot's center of mass. However, the robot's leg joints are also responsible for the unfolding motion during the reconstruction process, which leads to the robot occupying a larger envelope space and consuming more energy. In terms of stability control criteria for wheeled and legged robots, the current mainstream approach considers the robot's rigidity based on the zero-moment point method, neglecting the changes in the robot's center of mass caused by foot elasticity, which renders traditional criteria ineffective. Regarding stability control strategies for wheeled and legged robots, most rely on sensor-collected information such as the robot's tilt angle, employing traditional feedback control or model-based predictive control, which suffers from poor robustness or excessive computational demands. Summary of the Invention

[0003] The present invention addresses the shortcomings of the prior art by providing a stability control method for the reconfiguration motion of a wheeled robot. This method aims to determine and adjust the movement of the slider in the center of mass mechanism based on the motion parameters of the wheeled robot, thereby changing the position of the center of mass of the wheeled robot and improving the stability of its reconfiguration motion.

[0004] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: This invention discloses a stability control method for reconfiguration of a wheeled-legged robot considering the influence of wheel stiffness. The wheeled-legged robot includes: a vehicle body, a horizontal lifting mechanism, a center-of-gravity adjustment mechanism, and a foldable leg mechanism. The vehicle body consists of a front body and a rear body. The horizontal lifting mechanism, the center-of-gravity adjustment mechanism, and the foldable leg mechanism are arranged on the top plate of the rear body. The horizontal lifting mechanism adopts a symmetrical arrangement on the left and right sides. The lower rod of the horizontal lifting mechanism is connected to the front body, and the upper rod forms a sliding pair with an electric push rod. The end of the upper rod is directly fixed to the output shaft of the lifting motor of the rear body. The center-of-gravity adjustment mechanism consists of horizontal and vertical ball screw slides, a drive motor, and sliders mounted on the slides. The foldable leg mechanism adopts a symmetrical design, with both legs having the same structure, consisting of a thigh, a calf, and a supporting foot with four wheels, and correspondingly equipped with hip joints, knee joints, and ankle joints. The stability control method is characterized by the following steps: Step 1: Establish the coordinate system for the reconfiguration process of the wheeled robot; Step 2: Establish the displacement of the slider in the centroid adjustment mechanism according to equation (1). With the control voltage of the drive motor Electromechanical model between; (1) In equation (1), This indicates the equivalent current constant of the centroid mechanism being adjusted. This indicates the equivalent torque constant of the adjustment center of mass mechanism; The transmission ratio of the horizontal and vertical ball screws is given, and , For the lead of the horizontal and vertical ball screws; and These are the equivalent rotational inertia and equivalent viscous damping coefficient of the center-of-mass mechanism, respectively. To adjust the acceleration of the center-of-mass slider, To adjust the speed of the center of mass slider; Step 3: Establish a reconstructed nonlinear vibration model of the wheeled robot, and calculate the vertical elastic force of the wheel wheels to obtain an improved wheeled robot considering wheel stiffness. ZMP Criterion; Step 4: Based on the improved ZMP The criterion is obtained during the reconfiguration process of the wheeled robot. The actual zero torque point at time t. ; Step 5: Take the center of the foot support area of ​​the wheeled robot as the ideal zero torque point position. ,Will Position of the ideal zero torque point deviation and the rate of change of deviation The input is sent to the fuzzy controller, and the output is used to adjust the slider in the centroid mechanism. Ideal displacement at time ,Will After differentiation, the slider is obtained at... Ideal speed at any moment ; Step Six: Ideal displacement of the slider at any time and ideal speed The input is processed by the Kalman state observer in the LQG controller. Observed values ​​of slider displacement at time t and velocity observations The deviation between the two is used to calculate and adjust the center of gravity mechanism drive motor. Control voltage at any time Thus, we can obtain information about wheeled robots in... The actual zero torque point position at time t. ; Step 7: Assign to Then, return to step four and execute sequentially until the wheeled robot is reconfigured. Approaching the ideal position, thereby achieving stability control during the reconfiguration process of the wheeled robot.

[0005] The stability control method for reconfiguration of a wheeled robot considering the influence of wheel stiffness, as described in this invention, is characterized in that step one is performed as follows: Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot is located. ,Will The origin It is positioned in the middle of the two foot plates. The positive direction of the axis is the forward direction of the wheeled robot, and the basic coordinate system is... The positive direction of the axis is perpendicular to the horizontal plane and upwards, in the basic coordinate system. The positive direction of the axis is perpendicular to shaft and The plane containing the axle points outward from the vehicle body; Step 1.2: Sequentially establish the attachment coordinate systems for the left ankle joint, left knee joint, and left hip joint; the attachment coordinate systems for the right ankle joint, right knee joint, and right hip joint; and the attachment coordinate system for the horizontal lifting mechanism. Denote any one of these attachment coordinate systems as the [number missing]. a A physical coordinate system Let any one of the following joints—left ankle, left knee, left hip, right ankle, right knee, right hip, and the lifting joint corresponding to the horizontal lifting mechanism—be denoted as the first joint. a One joint, .

[0006] Furthermore, step three is performed as follows: Step 3.1: Establish the reconstructed nonlinear vibration model of the wheel-legged robot according to equation (2); (2) In equation (2), It is the total mass of the wheeled robot. , These are the locations of the wheeled robot's center of mass during the reconstruction process. Changes in acceleration and displacement along the axial direction, , These are the actual vertical displacement and vertical acceleration of the wheeled robot. , These are the pitch angle and pitch acceleration of the wheeled robot. The center of mass of the wheeled robot is in The acceleration along the axial direction, where g is the acceleration due to gravity. , It refers to the linear stiffness of the front and rear wheels. , It is the non-linear coefficient of the front and rear wheels. It is the moment of inertia of the wheeled robot relative to its center of mass. The distance between the center of gravity and the front wheel axle. This is the distance between the center of gravity and the rear wheel axle; Step 3.2, and After assigning values, the deformation of the front wheel is obtained by solving equation (2). Deformation of the rear wheel .

[0007] Step 3.3: Calculate the vertical spring force of the front wheel according to equation (3). Vertical spring force of the rear wheel : (3) In equation (3), and The vertical elastic force of the front and rear wheels, and For the linear stiffness of the front and rear wheels, and The nonlinear coefficients for the front and rear wheels are... and This represents the vertical deformation of the front and rear wheels.

[0008] Step 3.4: Establish the improved wheeled robot according to equation (4). ZMP Criterion: (4) In equation (4), For any one of the following components: the horizontal lifting mechanism, the slider of the center of gravity adjustment mechanism, the robot's two thighs, two calf legs, and the supporting feet of the two legs... The mass of each component It is the number of components, and =8, , These are the first wheeled robots Each component Axial and Axial acceleration of the center of mass, , These are the first wheeled robots The centroid of each component lies in the basic coordinate system. Axial and Axial coordinates; and These are the vertical elastic forces of the front wheel and the rear wheel, respectively.

[0009] Furthermore, step six is ​​performed as follows: Step 6.1, The actual displacement of the slider at any given time and actual speed as well as Control voltage at any time The input is processed in the Kalman state observer to obtain... Observed displacement of the slider at time and observation speed ; Step 6.2, place the slider in Ideal displacement at time With the slider Observed displacement at time deviation as well as Ideal speed at any moment With observation speed deviation The input is entered into the LQG controller, thereby using equation (5) to obtain the drive motor's position. Control voltage at any time : (5) In equation (5), The optimal feedback gain matrix; for The error state vector at time t, and ; Step 6.3, Substituting into equation (2), we obtain the position of the slider in the center-of-gravity adjustment mechanism. Actual displacement at time t and actual speed ; Step 6.4, Input Improved ZMP The criterion is used to output the wheeled robot's... The actual zero torque point position at time t. ; The present invention provides an electronic device, including a memory and a processor, characterized in that the memory is used to store a program supporting the processor in executing the stability control method, and the processor is configured to execute the program stored in the memory.

[0010] The present invention provides a computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, performs the steps of the stability hierarchical control method.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention utilizes Newton-Euler theory combined with a nonlinear vibration model of the foot wheel to establish a dynamic model of a wheeled robot, thereby deriving a robot reconfiguration stability criterion that considers the stiffness of the foot wheel. This criterion is used to control stability, thus mitigating the influence of the nonlinear vibration of the elastic foot wheel at the foot end on the robot's motion behavior and stability during reconfiguration.

[0012] 2. This invention uses the center-of-gravity adjustment mechanism installed on the rear body of a wheeled robot as the control object and designs a stability control method for the reconfiguration process. The fuzzy controller takes the deviation between the actual ZMP and the ideal ZMP and the rate of change of the deviation as input, and outputs the ideal displacement of the slider. The slider position is adjusted in real time to change the ZMP position of the wheeled robot. The LQG controller is used to track the output of the fuzzy controller, ultimately achieving closed-loop control of the entire robot system and improving reconfiguration stability.

[0013] 3. This invention innovatively designs a center-of-gravity adjustment mechanism for stability control during robot reconstruction, while the robot's leg joints are only responsible for the unfolding action during the reconstruction process. The two functions are completed by two different actuators, which greatly reduces energy consumption. Compared with traditional robots that achieve both leg movement and stability control through joint motors, the control response is faster and the real-time performance is better. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the structure and configuration of the wheeled-legged robot involved in this invention; Figure 2 This is a schematic diagram of the center-of-gravity adjustment mechanism of the wheeled robot involved in this invention; Figure 3 This is a schematic diagram illustrating the stability control principle of the reconstructed motion of the wheel-legged robot based on the adjustment center of mass mechanism of the present invention. Figure 4 This is a schematic diagram of the coordinate system of the wheeled robot involved in this invention; Figure 5a The figures show the motion results of the wheeled robot involved in this invention before and after reconfiguration stability control when the foot wheel stiffness is 150kN / m. Figure 5b The figures show the motion results of the wheeled robot involved in this invention before and after reconfiguration stability control when the wheel stiffness is 300 kN / m. Detailed Implementation

[0015] The present invention will be further described below with reference to the accompanying drawings and specific embodiments: The wheeled-legged robot involved in this invention is a complex nonlinear motion-coupled electromechanical system, such as... Figure 1 The diagram shows the structure and configuration of the wheeled-legged robot of this invention. The robot includes a vehicle body, a horizontal lifting mechanism, a center-of-gravity adjustment mechanism, and a foldable leg mechanism. The vehicle body consists of a front body and a rear body, each equipped with a pair of hub motors and a pair of MacPherson strut suspensions. The rear body houses the foldable legs, the lifting mechanism, and the center-of-gravity adjustment mechanism. The foldable legs are a series mechanism, allowing for 6 degrees of freedom of spatial movement at the ankle, knee, and hip joints. In vehicle mode, they can be folded into the vehicle body; in humanoid mode, they can be unfolded to support the entire robot. The lifting mechanism uses a lifting motor and an electric push rod to keep the front body horizontal during robot reconfiguration. The center-of-gravity adjustment mechanism consists of a ball screw slide rail, a drive motor, and a slider mounted on the slide rail.

[0016] Adjusting the specific structure and layout of the center of mass mechanism, such as Figure 2 As shown, the function of the center-of-gravity adjustment mechanism is mainly to adjust the position of the center of gravity in the X and Y directions through a cross-shaped layout. During the reconstruction process, the center-of-gravity adjustment mechanism can dynamically adjust the slider position based on real-time zero-torque point (ZMP) position feedback, thereby effectively improving the stability of the robot reconstruction process.

[0017] In this embodiment, a stability control method for reconstructing motion in a wheeled-legged robot is described, specifically, as follows: Figure 3As shown. This method mainly consists of a fuzzy controller and an LQG controller. The fuzzy controller, based on fuzzy theory, calculates the ideal displacement of the center-of-gravity adjustment mechanism based on the real-time zero-torque point position of the wheeled robot. The LQG controller, on the other hand, tracks the ideal displacement of the slider by calculating the control voltage of the center-of-gravity adjustment mechanism, thereby making the actual zero-torque point position of the wheeled robot approach the ideal zero-torque point position, thus improving the stability of robot reconfiguration. Specifically, this stability control method proceeds according to the following steps: Step 1: Establish the coordinate system for the reconfiguration process of the wheeled robot; Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot is located. ,Will The origin It is positioned in the middle of the two foot plates. The positive direction of the axis is the forward direction of the wheeled robot, and the basic coordinate system is... The positive direction of the axis is perpendicular to the horizontal plane and upwards, in the basic coordinate system. The positive direction of the axis is perpendicular to shaft and The plane containing the axle points outward from the vehicle body; Step 1.2: Sequentially establish the attachment coordinate systems for the left ankle joint, left knee joint, and left hip joint; the attachment coordinate systems for the right ankle joint, right knee joint, and right hip joint; and the attachment coordinate system for the horizontal lifting mechanism. Denote any one of these attachment coordinate systems as the [number missing]. a A physical coordinate system Let any one of the following joints—left ankle, left knee, left hip, right ankle, right knee, right hip, and the lifting joint corresponding to the horizontal lifting mechanism—be denoted as the first joint. a One joint, Specifically, such as Figure 4 As shown.

[0018] Using the ankle joint coordinate system Relative to the base coordinate system Taking the homogeneous coordinate transformation matrix as an example: Therefore, the homogeneous coordinate transformation matrix of any two coordinate systems can be expressed as: .

[0019] Step 2: Establish the displacement of the adjusting centroid slider according to equation (1). With the control voltage of the drive motor Electromechanical model between; (1) In equation (1), This indicates the equivalent current constant of the centroid mechanism being adjusted. This indicates the equivalent torque constant of the adjustment center of mass mechanism; Let be the transmission ratio of the ball screw, and , The lead of the ball screw; and These are the equivalent rotational inertia and equivalent viscous damping coefficient of the center-of-mass mechanism, respectively. To adjust the acceleration of the center-of-mass slider, To adjust the speed of the center of mass slider.

[0020] Step 3: Establish a reconstructed nonlinear vibration model of the wheeled robot, and calculate the vertical elastic force of the wheel wheels to obtain an improved wheeled robot considering wheel stiffness. ZMP Criterion; Step 3.1: Establish the reconstructed nonlinear vibration model of the wheel-legged robot according to equation (2); (2) In equation (2), It is the total mass of the wheeled robot. , These are the locations of the wheeled robot's center of mass during the reconstruction process. Changes in acceleration and displacement along the axial direction, , These are the actual vertical displacement and vertical acceleration of the wheeled robot. , These are the pitch angle and pitch acceleration of the wheeled robot. The center of mass of the wheeled robot is in The acceleration along the axial direction, where g is the acceleration due to gravity. , It refers to the linear stiffness of the front and rear wheels. , It is the non-linear coefficient of the front and rear wheels. It is the moment of inertia of the wheeled robot relative to its center of mass. The distance between the center of gravity and the front wheel axle. This is the distance between the center of gravity and the rear wheel axle.

[0021] Step 3.2, and After assigning values, the deformation of the front wheel is obtained by solving equation (5). Deformation of the rear wheel This allows for the calculation of the vertical spring force of the front wheel. Vertical spring force of the rear wheel .

[0022] Step 3.3: Calculate the vertical spring force of the front wheel according to equation (3). Vertical spring force of the rear wheel : (3) In equation (3), and The vertical elastic force of the front and rear wheels, and For the linear stiffness of the front and rear wheels, and The nonlinear coefficients for the front and rear wheels are... and This represents the vertical deformation of the front and rear wheels.

[0023] Step 3.4: Establish the improved wheeled robot according to equation (4). ZMP Criterion: (4) In equation (4), For any one of the following components: the horizontal lifting mechanism, the slider of the center of gravity adjustment mechanism, the robot's two thighs, two calf legs, and the supporting feet of the two legs... The mass of each component It is the number of components, and =8, , These are the first wheeled robots Each component Axial and Axial acceleration of the center of mass, , These are the first wheeled robots The centroid of each component lies in the basic coordinate system. Axial and Axial coordinates; and These are the vertical elastic forces of the front wheel and the rear wheel, respectively.

[0024] Step 4: Based on the improved ZMP The criterion is obtained during the reconfiguration process of the wheeled robot. The actual zero torque point at time t. ; Step 5: Take the center of the foot support area of ​​the wheeled robot as the ideal zero torque point position. ,Will Position of the ideal zero torque point deviation and the rate of change of deviation The input is sent to the fuzzy controller, and the output is used to adjust the slider in the centroid mechanism. Ideal displacement at time ,Will After differentiation, the slider is obtained at... Ideal speed at any moment ; The actual zero torque point With the ideal zero torque point deviation and rate of change of deviation As input to the fuzzy controller, the output is Through blurring processing and Then, fuzzy inference and declarative analysis were performed using a fuzzy controller to obtain the results. .

[0025] Step 5.1.1: Determine the universe of discourse for the fuzzy controller: Input variables The domain of discourse is [-0.08, 0.08]. The universe of discourse is [-0.2, 0.2]. The output variable of the fuzzy controller is the ideal displacement of the slider. The domain of discourse is [-0.55, 0.55]. Step 5.1.2: Determine the fuzzy set: The fuzzy sets for both the fuzzy control input and output are set to {NB, NM, NS, O, PS, PM, PB}, which represent {negative large, negative medium, negative small, zero, positive small, positive medium, positive large} respectively; Step 5.1.3: Determine the membership function: Triangle membership function; Step 5.1.4: Establish fuzzy rules; and The fuzzy rules are shown in Table 1; Table 1 and Fuzzy rule table

[0026] Step Six: Ideal displacement of the slider at any time and ideal speed The input is processed by the Kalman state observer in the LQG controller. Observations of slider displacement and velocity at time points , The control voltage of the center-of-gravity mechanism drive motor is adjusted by calculating the deviation between the two. .

[0027] Step 6.1, The actual displacement of the slider at any given time and actual speed as well as Control voltage at any time The input is processed in the Kalman state observer to obtain... Observed displacement of the slider at time and observation speed .

[0028] Step 6.1.1: Establish the state-space expression of the electromechanical dynamics model; definition For system state variables, To control the input variables, A is the system matrix of the center-of-gravity adjustment mechanism system, B is the system input matrix, and C is the output matrix. Based on formula (5), the electromechanical model of the center-of-gravity adjustment mechanism can be described as a state-space equation: (5) In equation (5): .

[0029] In actual control processes, various noise interferences can affect the accuracy of the model. Therefore, after modifying equation (5), we obtain: (6) In equation (6), and It consists of process noise and measurement noise, both of which are Gaussian white noise. Noise covariance data: Step 6.1.2: Establish the Kalman state observer according to equation (7); (7) In equation (7), Here is the Kalman gain matrix. For observed state variables.

[0030] Define observation error Subtracting equation (7) from equation (6) yields equation (8): (8) Step 6.1.3: Define the error covariance , The term represents the measurement residuals. According to the optimal estimation condition, the residuals must be orthogonal to past observations, i.e. =0, differentiating it yields equation (9): (9) Differentiating equation (9) with respect to L and setting the derivative to 0, we obtain equation (10): (10) Substituting the calculated L into equation (7) yields the observed values ​​of the slider displacement and velocity. and .

[0031] Step 6.2, place the slider in Ideal displacement at time With the slider Observed displacement at time deviation and Ideal speed at any moment With observation speed deviation Input into the LQG controller, and thus obtain the drive motor's position using equation (11). Control voltage at any time : (11) In equation (11), The optimal feedback gain matrix; for The error state vector at time t, and .

[0032] Step 6.2.1: Define the quadratic performance index function according to equation (12). ; (12) In equation (12), Let be the system state vector. For control vectors, The weighted matrix of the state vectors. The weighting matrices control the inputs. The Q matrix and R matrix can be represented as: (13) In equation (13), It is the weighting coefficient of the controller for the slider displacement. It is the weighting coefficient of the controller for the slider speed. It is the weighting coefficient of the control quantity.

[0033] To minimize the quadratic performance index function, we need to solve the algebraic Riccati equation (ARE) to obtain the K matrix: (14) In equation (14), P is obtained from the ARE equation: (15) Step 6.2.2: Use a genetic algorithm to minimize the quadratic performance index function. Thus, the optimized gain matrix is ​​obtained. ; Step 6.2.3: In order to track the ideal slider displacement ,Will and As a state variable, the control voltage is: (16) In equation (16), .

[0034] Step 6.3, Substituting into equation (2), we obtain the position of the slider in the center-of-gravity adjustment mechanism. Actual displacement at time t and actual speed ; Step 6.4, Input Improved ZMP The criterion is used to output the wheeled robot's... The actual zero torque point position at time t. ; Step 7: Assign to Then, return to step four and execute sequentially until the wheeled robot is reconfigured. Approaching the ideal position, thereby achieving stability control during the reconfiguration process of the wheeled robot.

[0035] 5. An electronic device comprising a memory and a processor, characterized in that the memory is used to store a program supporting the processor to execute any one of the stability control methods of claims 1-4, and the processor is configured to execute the program stored in the memory.

[0036] 6. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, performs the steps of the stability hierarchical control method according to any one of claims 1-4.

[0037] The system motion results of the wheeled robot before and after reconfiguration stability control using the center-of-gravity actuator are as follows: Figure 5a and Figure 5b As shown. Figure 5aThe figure shows the ZMP variation curves during the robot reconfiguration process when the wheel stiffness is 150 kN / m. As shown, the uncontrolled ZMP curve initially exhibits significant jitter, oscillating for a period before reaching the center of the support domain in about 5 seconds and moving towards the other boundary of the support domain, eventually stopping at a dangerous position near the boundary. The ZMP curve under fuzzy PID control rapidly approaches the center of the support domain, but then moves away from it, remaining near the center for 2.5 seconds until the end of reconfiguration, exhibiting a small error. Under optimized LQG control, the initial jitter of the ZMP curve continues to decrease, reaching the center of the support domain in about 1.5 seconds, and the overshoot is only 0.0001 m compared to the 0.003 m of the unoptimized LQG control, indicating a more stable reconfiguration of the wheel-legged robot.

[0038] Figure 5b The figure shows the ZMP variation curves during the robot reconfiguration process when the wheel stiffness is 300 kN / m. As shown, without control, the system experiences more severe shaking in the initial stage of reconfiguration, greatly increasing the possibility of the wheeled robot tipping over and becoming unstable. The ZMP under fuzzy PID control exhibits greater error and overshoot. Unoptimized LQG control mitigates the shaking phenomenon, while the ZMP curve shows better reduction in shaking under parameter-optimized LQG control. This indicates that the performance index function value of the control system is smaller after parameter optimization, and the system error state variable approaches 0 faster. Comparison with the reconfiguration simulation results with a stiffness of 150 kN / m shows that the LQG control optimized by the genetic algorithm has better robustness to stiffness changes.

[0039] The tracking error and maximum error of the ZMP curve under the three control methods are shown in Table 2.

[0040] Table 2 Tracking Errors of Three Control Methods

[0041] As shown in Table 2, the average error of the parameter-optimized LQG control does not exceed 0.0029m, and the maximum error does not exceed 0.0067m. In summary, the parameter-optimized LQG control performs well in tracking the ideal displacement during the reconstruction process and has a good control effect on the reconstruction stability of the wheel-legged robot.

Claims

1. A stability control method for reconfiguration of a wheeled-legged robot considering the influence of wheel stiffness, the wheeled-legged robot comprising: The vehicle body, horizontal lifting mechanism, center of gravity adjustment mechanism, and foldable leg mechanism; The vehicle body consists of a front body and a rear body. A horizontal lifting mechanism, a center-of-gravity adjustment mechanism, and a foldable leg mechanism are arranged on the roof of the rear body. The horizontal lifting mechanism is arranged symmetrically on the left and right sides. The lower rod of the horizontal lifting mechanism is connected to the front body, and the upper rod forms a sliding pair with an electric push rod. The end of the upper rod is directly fixed to the output shaft of the lifting motor of the rear body. The center-of-gravity adjustment mechanism consists of horizontal and vertical ball screw slides, a drive motor, and sliders mounted on the slides. The foldable leg mechanism adopts a symmetrical design, with identical leg structures, each consisting of a thigh, a calf, and a supporting foot with four wheels, and correspondingly equipped with hip, knee, and ankle joints. The stability control method is characterized by the following steps: Step 1: Establish the coordinate system for the reconfiguration process of the wheeled robot; Step 2: Establish the displacement of the slider in the centroid adjustment mechanism according to equation (1). With the control voltage of the drive motor Electromechanical model between; (1) In equation (1), This indicates the equivalent current constant of the centroid mechanism being adjusted. This indicates the equivalent torque constant of the adjustment center of mass mechanism; The transmission ratio of the horizontal and vertical ball screws is given, and , For the lead of the horizontal and vertical ball screws; and These are the equivalent rotational inertia and equivalent viscous damping coefficient of the center-of-mass mechanism, respectively. To adjust the acceleration of the center-of-mass slider, To adjust the speed of the center of mass slider; Step 3: Establish a reconstructed nonlinear vibration model of the wheeled robot, and calculate the vertical elastic force of the wheel wheels to obtain an improved wheeled robot considering wheel stiffness. ZMP Criterion; Step 4: Based on the improved ZMP The criterion is obtained during the reconfiguration process of the wheeled robot. The actual zero torque point at time t. ; Step 5: Take the center of the foot support area of ​​the wheeled robot as the ideal zero torque point position. ,Will Position of the ideal zero torque point deviation and the rate of change of deviation The input is sent to the fuzzy controller, and the output is used to adjust the slider in the centroid mechanism. Ideal displacement at time ,Will After differentiation, the slider is obtained at... Ideal speed at any moment ; Step Six: Ideal displacement of the slider at any time and ideal speed The input is processed by the Kalman state observer in the LQG controller. Observed values ​​of slider displacement at time t and velocity observations The deviation between the two is used to calculate and adjust the center of gravity mechanism drive motor. Control voltage at any time Thus, we can obtain information about wheeled robots in... The actual zero torque point position at time t. ; Step 7: Assign to Then, return to step four and execute sequentially until the wheeled robot is reconfigured. Approaching the ideal position, thereby achieving stability control during the reconfiguration process of the wheeled robot.

2. The stability control method for reconfiguration of a wheeled robot considering the influence of wheel stiffness as described in claim 1, characterized in that, The first step is performed as follows: Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot is located. ,Will The origin It is positioned in the middle of the two foot plates. The positive direction of the axis is the forward direction of the wheeled robot, and the basic coordinate system is... The positive direction of the axis is perpendicular to the horizontal plane and upwards, in the basic coordinate system. The positive direction of the axis is perpendicular to shaft and The plane containing the axle points outwards from the vehicle body; Step 1.2: Sequentially establish the attachment coordinate systems for the left ankle joint, left knee joint, and left hip joint; the attachment coordinate systems for the right ankle joint, right knee joint, and right hip joint; and the attachment coordinate system for the horizontal lifting mechanism. Denote any one of these attachment coordinate systems as the [number missing]. a A physical coordinate system Let any one of the following joints—left ankle, left knee, left hip, right ankle, right knee, right hip, and the lifting joint corresponding to the horizontal lifting mechanism—be denoted as the first joint. a One joint, .

3. The stability control method for reconfiguration of a wheeled robot considering the influence of wheel stiffness according to claim 2, characterized in that, Step three is performed as follows: Step 3.1: Establish the reconstructed nonlinear vibration model of the wheel-legged robot according to equation (2); (2) In equation (2), It is the total mass of the wheeled robot. , These are the locations of the wheeled robot's center of mass during the reconstruction process. Changes in acceleration and displacement along the axial direction, , These are the actual vertical displacement and vertical acceleration of the wheeled robot. , These are the pitch angle and pitch acceleration of the wheeled robot. The center of mass of the wheeled robot is in The acceleration along the axial direction, where g is the acceleration due to gravity. , It refers to the linear stiffness of the front and rear wheels. , It is the non-linear coefficient of the front and rear wheels. It is the moment of inertia of the wheeled robot relative to its center of mass. The distance between the center of gravity and the front wheel axle. This is the distance between the center of gravity and the rear wheel axle; Step 3.2, and After assigning values, the deformation of the front wheel is obtained by solving equation (2). Deformation of the rear wheel ; Step 3.3: Calculate the vertical spring force of the front wheel according to equation (3). Vertical spring force of the rear wheel : (3) In equation (3), and The vertical elastic force of the front and rear wheels, and For the linear stiffness of the front and rear wheels, and The nonlinear coefficients for the front and rear wheels are... and This represents the vertical deformation of the front and rear wheels; Step 3.4: Establish the improved wheeled robot according to equation (4). ZMP Criterion: (4) In equation (4), For any one of the following components: the horizontal lifting mechanism, the slider of the center of gravity adjustment mechanism, the robot's two thighs, two calf legs, and the supporting feet of the two legs... The mass of each component It is the number of components, and =8, , These are the first wheeled robots Each component Axial and Axial acceleration of the center of mass, , These are the first wheeled robots The centroid of each component lies in the basic coordinate system. Axial and Axial coordinates; and These are the vertical elastic forces of the front wheel and the rear wheel, respectively.

4. The stability control method for reconfiguration of a wheeled robot considering the influence of wheel stiffness according to claim 3, characterized in that, Step six is ​​performed as follows: Step 6.1, The actual displacement of the slider at any given time and actual speed as well as Control voltage at any time The input is processed in the Kalman state observer to obtain... Observed displacement of the slider at time and observation speed ; Step 6.2, place the slider in Ideal displacement at time With the slider Observed displacement at time deviation as well as Ideal speed at any moment With observation speed deviation The input is entered into the LQG controller, thereby using equation (5) to obtain the drive motor's position. Control voltage at any time : (5) In equation (5), The optimal feedback gain matrix; for The error state vector at time t, and ; Step 6.3, Substituting into equation (2), we obtain the position of the slider in the center-of-gravity adjustment mechanism. Actual displacement at time t and actual speed ; Step 6.4, Input Improved ZMP The criterion is used to output the wheeled robot's... The actual zero torque point position at time t. .

5. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports the processor in executing any of the stability control methods described in claims 1-4, and the processor is configured to execute the program stored in the memory.

6. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program, when executed by the processor, performs the steps of any of the stability hierarchical control methods described in claims 1-4.

Images