Unmanned motorcycle and its steering-cmg collaborative control method and system

Abstract

This invention discloses an unmanned motorcycle and its steering-CMG cooperative control method, comprising: acquiring a steering-CMG extended-dimensional dynamic model with actuator angular velocity commands as the unified input; establishing an optimal feedback controller based on the model and solving for the basic control law including steering and CMG actuator feedback gains; constructing a steering-CMG dual-modal cooperative control strategy, dividing attitude adjustment into attitude establishment and attitude maintenance stages; controlling the CMG and steering actuator to cooperate in opposite phases during the attitude establishment stage to accelerate the equilibrium point migration; and controlling them to cooperate in phases during the attitude maintenance stage to enhance recovery capability and damping; constructing continuous cooperative weights based on Lyapunov energy, fusing the in-phase and out-of-phase cooperative outputs to obtain the final CMG cooperative control law to control the vehicle. This invention achieves a unified control model and engineering interface, balances attitude establishment speed and maintenance stability, and ensures a smooth transition of cooperative control.

Description

Technical Field

[0001] This invention belongs to the field of vehicle intelligent control technology, specifically relating to an unmanned motorcycle and its steering-CMG cooperative control method and system. Background Technology

[0002] Two-wheeled unmanned motorcycles are typical underactuated dynamic unstable systems. Their attitude cannot be stabilized by the structure itself and requires active control to maintain balance. Compared with four-wheeled mobile platforms, two-wheeled structures have advantages such as high mobility and strong passability, but their attitude degrees of freedom and steering degrees of freedom are strongly coupled, resulting in complex system dynamics.

[0003] In existing technologies, lateral acceleration is often generated by controlling the steering angle or angular velocity of the front wheels to adjust the vehicle's attitude. However, this method relies on lateral dynamic response, suffers from dynamic lag, and lacks sufficient lateral force under low-speed or continuous disturbance conditions, limiting attitude recovery capability. To enhance attitude adjustment capability, some solutions introduce a control moment gyroscope (CMG) as an auxiliary actuator, but single CMG control suffers from limited output torque and precession singularities. To further improve performance, existing technologies attempt to coordinate the steering actuator with the CMG, but the following main technical problems remain: The model and engineering control interface are inconsistent. Existing dynamic models mostly use actuator output torque as the input variable, while steering motors and CMG universal joint motors in actual engineering systems mostly use angular velocity or speed commands as the control interface. Existing methods do not uniformly consider the dynamic characteristics of actuators and the actual interface form at the model level, resulting in a disconnect between model design and engineering implementation, making it difficult to directly apply the control algorithm in engineering.

[0004] There is a lack of cooperative structural design based on system dynamics. Existing cooperative control systems mostly focus on simple superposition or fixed ratio allocation of control signals, without systematically modeling and analyzing the coupling relationship between steering and CMG from the perspective of system dynamics. This makes it difficult to achieve coordinated optimization of system dynamic characteristics at the closed-loop dynamics level.

[0005] There is a lack of differentiated coordination mechanisms for different operational phases. Existing cooperative control systems mostly adopt a unified control structure, failing to adjust the cooperative relationships of multiple actuators based on the dynamic characteristics during attitude adjustment. In the attitude establishment phase and the disturbance rejection maintenance phase during equilibrium point migration, the system's requirements for response speed and stability are drastically different, making it difficult for unified control to simultaneously address both maneuverability and disturbance rejection stability. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing unmanned motorcycle steering-CMG cooperative control method in achieving low efficiency of cooperative control of multiple actuators in different attitude adjustment stages, thereby providing an unmanned motorcycle and its steering-CMG cooperative control method and system.

[0007] A steering-CMG cooperative control method, comprising: Obtain a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; Based on the aforementioned steering-CMG extended dynamics model, an optimal feedback controller is established, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving the problem. Based on the aforementioned fundamental control law, a steering-CMG dual-modal cooperative control strategy is constructed, comprising: dividing the vehicle's attitude adjustment process into an attitude establishment phase and an attitude maintenance phase; in the attitude establishment phase, controlling the CMG actuator and the steering actuator to cooperate in opposite phases, so that the direction of the gyroscopic coupling torque generated by the CMG is opposite to the direction of the roll adjustment action generated by the steering actuator; in the attitude maintenance phase, controlling the CMG actuator and the steering actuator to cooperate in the same phase, so that the direction of the gyroscopic coupling torque generated by the CMG is the same as the direction of the roll adjustment action generated by the steering actuator. Based on Lyapunov energy construction, continuous cooperative weights are used to fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, and the final CMG cooperative control law is output. The vehicle is controlled based on the cooperative control law.

[0008] Furthermore, it also includes constructing a Lyapunov energy function based on the optimal feedback controller, and identifying whether the vehicle is currently in the attitude establishment phase or the attitude maintenance phase based on the Lyapunov energy function.

[0009] Furthermore, in the step of obtaining the linear dynamic model of torque input in the steering-CMG extended-dimensional dynamic model with actuator angular velocity command as the unified input, the following steps are included: Based on the Kane method, a multibody dynamics model including the frame, fork, front and rear wheels and dual CMG devices is established under the wheel-ground nonholonomic constraint. The dual CMG device is subjected to differential mode equivalent processing, and the multibody dynamics model is linearized at the uniform straight-line operating point to obtain the linear dynamics model of the torque input.

[0010] Furthermore, in the step of obtaining the extended-dimensional dynamic model of the steering-CMG with the actuator angular velocity command as the unified input, the step of obtaining the extended-dimensional dynamic model based on the linear dynamic model of the torque input includes: A first-order dynamic model of the actuator is introduced to convert the torque input in the linear dynamic model of the torque input into the actuator angular velocity command input; By dynamically incorporating the actuators into the system through state expansion, the steering-CMG extended-dimensional dynamic model with the actuator angular velocity command as the unified input is obtained.

[0011] Furthermore, the differential mode equivalent processing of the dual CMG device includes: By setting the precession angles of the front and rear CMGs to be equal in magnitude and opposite in sign, a differential mode precession variable and a corresponding differential mode precession torque are introduced, and the dual CMG device is equivalent to a single precession degree of freedom.

[0012] Furthermore, the optimal feedback controller is established based on the steering-CMG extended-dimensional dynamics model, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving it, including: Discrete speed sweeps are performed within a given vehicle speed range, and the extended-dimensional dynamic model is linearized at each vehicle speed operating point to obtain a vehicle speed parameterized state space model. For each vehicle speed operating point, a quadratic performance index function of system state and control input is constructed, the optimal feedback control problem is solved, and the control gain matrix corresponding to different vehicle speeds is obtained. A vehicle speed-control gain lookup table is constructed. Based on the real-time vehicle speed, interpolation calculations are performed in the lookup table to obtain the feedback control gain at the current vehicle speed, which is then used to construct the basic control law.

[0013] Furthermore, the step of identifying whether the vehicle is currently in the attitude establishment phase or the attitude maintenance phase based on the Lyapunov energy function includes: Calculate the rate of change of the Lyapunov energy function; When the Lyapunov energy function and its rate of change are both less than their respective thresholds, and the turning reference rate of change meets the preset conditions, the recognition system enters the attitude maintenance phase; otherwise, the recognition system is in the attitude establishment phase.

[0014] Furthermore, the construction of continuous cooperative weights based on Lyapunov energy includes: Define a normalized energy and construct a continuous cooperative weight function based on the Sigmoid function; where, when the normalized energy is greater than a preset threshold, the continuous cooperative weight corresponds to the attitude establishment stage; when the normalized energy is less than the preset threshold, the continuous cooperative weight corresponds to the attitude maintenance stage. The continuous collaborative weights are subjected to first-order filtering, and the control outputs of the anti-phase collaboration and the in-phase collaboration are fused based on the filtered continuous collaborative weights.

[0015] A steering-CMG cooperative control system for implementing the above-mentioned steering-CMG cooperative control method includes: The dynamics model building module is used to build a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; The steering-CMG dual-modal cooperative control module is used to construct a steering-CMG dual-modal cooperative control strategy based on the basic control law. The final cooperative control module is used to construct continuous cooperative weights based on Lyapunov energy, fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, output the final CMG cooperative control law, and control the vehicle based on the cooperative control law.

[0016] An unmanned motorcycle includes a controller, a CMG actuator, a rear wheel, a frame, a steering actuator, a front fork, and a front wheel. The controller controls the CMG actuator and the steering actuator through the aforementioned steering-CMG coordinated control method.

[0017] Beneficial Effects: This invention discloses a steering-CMG cooperative control method. By acquiring a steering-CMG extended-dimensional dynamic model with actuator angular velocity commands as the unified input, the traditional model with torque as the input is reconstructed into an angular velocity command input form consistent with the actual servo control interface. This solves the inconsistency between model design and engineering implementation, improving the engineering feasibility and accuracy of the control algorithm. By constructing a steering-CMG dual-modal cooperative control strategy, in the attitude establishment phase, the CMG actuator and steering actuator are controlled to cooperate in opposite phases, actively disrupting the original balance to accelerate the roll equilibrium point migration, thus improving attitude response speed. In the attitude maintenance phase, they are controlled to cooperate in phases, enhancing the system's recovery capability and damping, improving disturbance rejection stability, thereby achieving effective coordination between maneuver response and disturbance rejection stability. Based on Lyapunov energy construction of continuous cooperative weights, the control outputs of the opposite-phase and in-phase cooperative modes are fused, avoiding abrupt changes in control input caused by rigid switching between different control modes, achieving a smooth transition between the two modes, and further improving the smoothness of the system's dynamic response and overall control performance. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of the main method steps of the present invention; Figure 2 This is a schematic diagram of the dual CMG differential mode precession mechanism of the present invention; Figure 3 This is a schematic diagram of the position vector relationship of key points in the system of the present invention; Figure 4 This is a schematic diagram of the steering-CMG dual-modal cooperative control framework of the present invention; Figure 5 This is a schematic diagram of the unmanned motorcycle structure of the present invention.

[0020] Explanation of reference numerals in the attached diagram: 1. Rear wheel contact point; 2. Rear wheel center of gravity; 3. Rear CMG center of gravity; 4. Frame center of gravity; 5. CMG base center of gravity; 6. Front CMG center of gravity; 7. Steering hinge point; 8. Front fork center of gravity; 9. Front wheel center of gravity; 10. Front wheel contact point; 11. Unmanned motorcycle controller; 12. Dual CMG actuators; 13. Rear wheel; 14. Frame; 15. Steering actuator; 16. Front fork; 17. Front wheel. Detailed Implementation

[0021] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0022] In the description of this application, it should be understood that 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 indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0023] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0024] Example 1: Reference Figure 1As shown, this embodiment provides a steering-CMG cooperative control method, including: Obtain a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; Based on the aforementioned steering-CMG extended dynamics model, an optimal feedback controller is established, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving the problem. Based on the aforementioned fundamental control law, a steering-CMG dual-modal cooperative control strategy is constructed, comprising: dividing the vehicle's attitude adjustment process into an attitude establishment phase and an attitude maintenance phase; in the attitude establishment phase, controlling the CMG actuator and the steering actuator to cooperate in opposite phases, so that the direction of the gyroscopic coupling torque generated by the CMG is opposite to the direction of the roll adjustment action generated by the steering actuator; in the attitude maintenance phase, controlling the CMG actuator and the steering actuator to cooperate in the same phase, so that the direction of the gyroscopic coupling torque generated by the CMG is the same as the direction of the roll adjustment action generated by the steering actuator. Based on Lyapunov energy construction, continuous cooperative weights are used to fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, and the final CMG cooperative control law is output. The vehicle is controlled based on the cooperative control law.

[0025] Specifically, in step S1, the steps for obtaining the linear dynamic model of torque input in the steering-CMG extended-dimensional dynamic model with actuator angular velocity command as the unified input include: Based on the Kane method, a multibody dynamics model including the frame, front fork, front wheel, rear wheel and dual CMG device is established under the wheel-ground nonholonomic constraint. The dual CMG device is subjected to differential mode equivalence processing to reduce the system degrees of freedom, so that the dual CMG device can be equivalent to a single precession degree of freedom, thereby simplifying the expression of system dynamics.

[0026] Furthermore, the multibody dynamics model is linearized near the uniform straight-line working point to obtain a linear dynamics model of the torque input, which is expressed in second-order matrix form as follows: ; The state-space form is represented as: ; The above linear dynamics model provides a unified dynamic basis for the design of the steering-CMG cooperative controller.

[0027] In this embodiment, the modeling method includes the following steps: Establish generalized coordinates, generalized velocity, and input / perturbation variables: The system's generalized coordinates are defined to include vehicle translation, vehicle attitude, and wheel system rotation, and corresponding Kane generalized velocity variables are introduced.

[0028] The generalized coordinates and velocity of the translation are: ; in, This indicates the position coordinates of the frame reference point in the inertial coordinate system. This represents the translational velocity of the frame reference point in the inertial frame.

[0029] The attitude and gear train generalized coordinates are represented as follows: ; in, These represent the vehicle's yaw angle, roll angle, rear wheel spin angle, chassis pitch angle, front wheel steering angle, and front wheel spin angle, respectively.

[0030] The corresponding Kane generalized velocity is: ; in, These represent the Kane generalized velocity corresponding to the yaw angle, the Kane generalized velocity corresponding to the roll angle, the rear wheel rotation angular velocity, the frame pitch angular velocity, the front wheel steering angular velocity, and the front wheel rotation angular velocity, respectively.

[0031] The precession degrees of freedom of the dual CMG differential mode are: ; in, Indicates the CMG differential mode precession angle. This indicates the CMG differential mode precession angular velocity.

[0032] System control inputs and disturbances are defined as follows: ; in This indicates the driving torque of the CMG gimbal. Indicates the driving torque of the steering actuator. This indicates the driving torque of the drive wheel motor. This indicates the external rolling disturbance torque.

[0033] Establish the recurrence relation of the reference frame: Establish an inertial reference frame And construct the yaw, roll, frame, head tube, fork and wheel reference system in a recursive manner.

[0034] The rotation operator is expressed as: ; Representing a frame of reference Around the axis Rotation angle A new frame of reference is obtained.

[0035] The recurrence relation of the reference frame is: ; ; ; ; ; ; Reference Figure 2 As shown, differential mode equivalent processing is performed on the dual CMG device, including: By setting the precession angles of the front and rear CMGs to be equal in magnitude and opposite in sign, a differential mode precession variable and a corresponding differential mode precession torque are introduced, and the dual CMG device is equivalent to a single precession degree of freedom.

[0036] To reduce the model's degrees of freedom while preserving the synergistic effect of the two CMGs, a differential mode equivalence method is adopted: the precession angles of the front and rear CMGs are equal in magnitude but opposite in sign, and only one differential mode precession variable is introduced. : ; The differential precession torque is similarly set with the opposite sign: ; Based on this, a reference system for CMG before and after is established: ; And set the rotor spin angular velocity to a constant value. , used to generate gyroscope coupling terms.

[0037] Reference Figure 3 As shown, the keypoint location vector chain is defined: Define the position vector relationships of key points in the unmanned motorcycle system, and establish spatial relationships between key points through vector recursion.

[0038] With the origin of inertia Define the frame reference point as a benchmark : ; based on Recursively establish the centers of gravity of the rear wheel, steering hinge point, frame center of gravity, fork center of gravity, front wheel center of gravity, front wheel contact point, and all CMG centers of gravity: The rear wheel center is indicated as: ; The steering hinge point is represented as: ; The frame center of gravity is represented as: ; The center of gravity of the fork and the center of gravity of the front wheel are represented as follows: ; ; The front wheel contact point is constructed such that it is located at the wheel radius along the wheel-to-ground normal direction. The location is represented as: ; in It is a unit vector determined by the direction of the wheel plane and the ground normal.

[0039] The CMG centroid is represented as: ; ; Define the recursive relationship between angular velocity and linear velocity The angular velocities of each reference frame are directly defined using Kane's generalized velocity, for example: ; The two-point velocity theorem for rigid body kinematics is: ; in, For point In an inertial frame Decrease speed.

[0040] Let the translational velocity of the frame reference point be... ; By recursively applying the two-point velocity theorem, the velocity expressions for key points such as the frame, fork, front wheel, and CMG can be obtained, thus establishing the KD equation: Establish the kinematic differential equations (KD equations): ; ; Establish wheel-ground nonholonomic constraints and ground-contact constraints Define the unit vector of the rolling direction (cross product of the wheel plane direction and the ground normal): ; Establish wheel-to-ground constraints (velocity layer constraints): Pure rolling (velocity along the rolling direction equals radius multiplied by rotational angular velocity) is represented as: ; The absence of sideslip constraints (lateral velocity is zero) is represented as: ; Ground-hugging constraint (vertical velocity is zero) is expressed as: ; And establish geometric ground-hugging constraints (contact point height is zero, position layer constraints) as follows: ; Construct a multi-rigid-body object with inertia parameters, and apply loads and control torques: The frame, load, fork, front and rear wheels, CMG mount / universal frame / rotor are all considered rigid bodies, and their masses are defined separately. The position of the center of mass and the inertia tensor .

[0041] The main loads of the system include: Gravity is represented as: ; The disturbance overturning moment is expressed as: ; Steering torque (applied to the front fork steering shaft) is expressed as: ; The differential CMG precession torque (front + rear −) is expressed as: ; The constrained dynamic equations are derived using the Kane method: In this embodiment, the Kane method is used to construct the dynamic equations. For each independent generalized velocity... ,Establish: ; in, It is a generalized active force (external force / external torque projected onto part of the velocity). It is the generalized inertial force (inertial force / inertial torque projected onto a portion of the velocity).

[0042] The inertial force and moment of inertia of a rigid body are expressed as follows: ; ; The generalized force projection form is expressed as: ; ; By combining velocity constraints and geometric constraints, the dependent velocities and dependent coordinates are eliminated to obtain the minimum system of equations.

[0043] Linearization at the uniform straight-line working point yields... With state-space model: The equilibrium working point for uniform straight-line movement is selected and represented as: ; ; ; ; ; Linearizing the dynamic equations at this operating point yields a second-order matrix form, as shown below: ; Further written in first-order state-space form, it can be represented as: ; ; In step S1, Indicates the forward speed of the vehicle. Indicates the first A rigid body mass, Indicates the first rigid body inertia tensor Indicates the rear wheel radius. Indicates the radius of the front wheel. Indicates the frame reference point. Indicates the steering hinge point. Indicates the frame reference frame. Indicates the front fork reference frame. Represents the rear wheel reference frame. Indicates the front wheel reference frame. Indicates the former CMG reference frame. Indicates the post-CMG reference frame. Point In an inertial frame The speed in the middle, Point relative point The position vector, Representing a rigid body relative inertial frame angular velocity, Represents gravitational acceleration. This represents the state-space system matrix.

[0044] This state-space model, i.e., the linear dynamic model of the torque input, is used for subsequent controller design.

[0045] Specifically, in step S2, based on the linear dynamic model obtained in step S1 in this embodiment, the first-order dynamics of the actuator are introduced and the control input is dynamically reconstructed, so that the model input is uniformly converted from force rectangular form to velocity form, thereby obtaining a unified interface model that can be directly used for engineering controller implementation.

[0046] In obtaining the extended-dimensional dynamic model of the steering-CMG model with the actuator angular velocity command as the unified input, the steps of obtaining the extended-dimensional dynamic model based on the linear dynamic model with the torque input include: Step S2.1: Introduce the first-order dynamic model of the actuator to convert the torque input in the linear dynamic model of the torque input into the actuator angular velocity command input; The linear model of torque input is expressed as: After linearizing the uniform straight-line operating point, the torque input state-space model for control design is obtained: ; in, The system state vector (including roll, yaw, steering, CMG precession and its derivatives, etc.); The equivalent input torque vector of the actuator can be expressed as: ; First-order dynamics of the actuators are constructed. In actual systems, steering actuators and CMG gimbal actuators are typically implemented using "speed loop servo" or "speed closed loop," with their upper-level control interface being angular velocity commands. To unify the input format, a first-order dynamic model is introduced for each actuator in this embodiment, mapping "speed commands" to "torque outputs."

[0047] Steering actuator dynamics: Let the angular velocity of the steering shaft be: ; Its speed servo dynamics can be expressed as: ; in The equivalent time constant of the steering actuator, This is the steering angular velocity command.

[0048] The actuator output torque and speed tracking error can be expressed by the equivalent gain as: ; in The equivalent torque-speed error gain of the steering actuator (which can be obtained from motor / drive parameters or identification).

[0049] The above expression is equivalent to "torque output of the closed-loop velocity loop", encapsulating the lower-level torque loop as an equivalent first-order velocity loop, thus allowing the upper-level controller to only need to output... .

[0050] CMG precession actuator dynamics: Let the CMG differential mode precession angular velocity be: ; Similarly, the dynamic representation of the CMG gimbal speed servo is as follows: ; in The equivalent time constant of the gimbal actuator. This is the precession angular velocity command.

[0051] The differential precession torque is also expressed by the equivalent torque generated by the speed error: ; in The equivalent torque-velocity error gain for the CMG gimbal actuator.

[0052] Dynamic reconstruction: Substitute the above equation into the torque obtained in step S1 and input it into the model: ; The equivalent input form for the speed command can be obtained.

[0053] Recorded as: ; And the equivalent actuator gain matrix: ; Then we have: ; Substituting into the original model, we get: ; Step S2.2: To make the model input directly... It is necessary to Incorporating the extended state. In this embodiment, the actuator is dynamically incorporated into the system through state extension, resulting in the steering-CMG extended dynamic model with the actuator angular velocity command as the unified input.

[0054] Extended-dimensional state-space model: Construct an extended-dimensional state, represented as: ; The first-order dynamics of the actuator are written as follows: ; in: ; The extended-dimensional model for obtaining the unified speed command input is as follows: ; in: ; here For "from the original state" The selection matrix for "extracting the corresponding angular velocity components" (e.g., from...) Pull out , ). In implementation It can be directly determined by the state definition.

[0055] By expanding the dimensions, the actuators are dynamically incorporated into the system model, unifying the controller input to... and This ensures consistency with the engineering servo control interface; at the same time, the model retains the actuator response hysteresis characteristics, improving the accuracy of closed-loop prediction / design.

[0056] In step S3, based on the extended-dimensional dynamic model of the unified speed command input obtained in step S2, this embodiment proposes an optimal feedback cooperative control method based on sweep speed scheduling. By linearizing the sweep speed, a vehicle speed-related dynamic model is established, and the optimal feedback control gain is solved at each vehicle speed operating point, thereby constructing a cooperative control law that varies with vehicle speed, and realizing stable attitude control of the unmanned motorcycle under different vehicle speed conditions.

[0057] The optimal feedback controller is established based on the steering-CMG extended dynamics model, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving it, including: Discrete speed sweeps are performed within a given vehicle speed range, and the extended-dimensional dynamic model is linearized at each vehicle speed operating point to obtain a vehicle speed parameterized state space model. Because the dynamic characteristics of unmanned motorcycles change significantly with vehicle speed, the system matrix... , With vehicle speed There is a coupling relationship. To obtain a controller applicable to the entire vehicle speed range, this technology focuses on a given vehicle speed range. The speed is discretized by a certain step size, and the extended-dimensional dynamic model obtained in step S2 is linearized at each vehicle speed operating point to obtain a set of vehicle speed parameterized state-space models: ; in, For extended-dimensional state vectors, This is the actuator angular velocity control command.

[0058] For each vehicle speed operating point, a quadratic performance index function of system state and control input is constructed, the optimal feedback control problem is solved, and the control gain matrix corresponding to different vehicle speeds is obtained. For each operating speed point, construct a quadratic performance index function for the system state and control input: ; in, The state weight matrix is... To control the input weight matrix.

[0059] The feedback control law is obtained by solving the optimal control problem: ; in, These correspond to the feedback gain of the steering actuator and the CMG actuator, respectively.

[0060] Therefore, we can obtain: The steering angular velocity command is expressed as: ; The CMG precession angular velocity command is expressed as: .

[0061] A vehicle speed-control gain lookup table is constructed. Based on the real-time vehicle speed, interpolation calculations are performed in the lookup table to obtain the feedback control gain at the current vehicle speed, which is then used to construct the basic control law.

[0062] The control gain matrix corresponding to different vehicle speeds is obtained through offline speed sweep calculation. Construct a vehicle speed-control gain lookup table.

[0063] In actual control, based on the vehicle's real-time speed Interpolation calculations are performed in the lookup table to obtain the feedback control gain at the current vehicle speed: ; In step S3, This represents the feedback gain matrix that varies with vehicle speed. Represents the state error vector; The above-mentioned speed-scanning scheduling method can ensure that the unmanned motorcycle has good attitude stability and control performance across the entire speed range.

[0064] Specifically, refer to Figure 4As shown, in step S4, based on the optimal feedback control law based on vehicle speed scheduling constructed in step S3, a steering-CMG dual-modal cooperative control strategy is proposed. According to the dynamic characteristics of the vehicle attitude adjustment process, the control process is divided into an attitude establishment stage and an attitude maintenance stage. Dynamic cooperative control between the steering actuator and the CMG actuator is realized through an energy-driven continuous cooperative mechanism, thereby taking into account both attitude establishment speed and attitude stability performance.

[0065] Based on the aforementioned fundamental control law, a steering-CMG dual-modal cooperative control strategy is constructed, including: When the steering reference changes (e.g., steering angle reference) or path curvature When a step or slope change occurs, the vehicle's roll equilibrium point will shift with the steering input, and the relationship can be expressed as: ; in, For the roll balance angle, For steering angle, For vehicle speed.

[0066] When the input changes, the system state needs to shift from the original equilibrium point. Migrate to a new equilibrium point ; During the shift to the equilibrium point, the rapid movement of the steering channel introduces an equivalent disturbance term into the roll dynamics, which leads to the system requiring a large counter-steering compensation to suppress overshoot, while the attitude establishment process has a slow response speed.

[0067] To accelerate the system's migration from the original equilibrium point to the new equilibrium point, this technology introduces a CMG anti-phase cooperative control mechanism during the attitude establishment phase. This mechanism ensures that the direction of the gyroscopic coupling torque generated by the CMG is opposite to the direction of the roll adjustment action generated by the steering actuator, thereby actively disrupting the original equilibrium state and accelerating the system's transition to the new equilibrium state.

[0068] Based on the optimal feedback control law for vehicle speed scheduling obtained in step S3, the system control input is expressed as: ; in, These are the steering actuator angular velocity command and the CMG precession angular velocity command, respectively.

[0069] The corresponding basic control law is: Steering control is represented as: ; CMG basic control is represented as follows: ; To achieve dual-modal cooperative control, a sign factor and participation coefficient are introduced into the CMG channel: ; in, Representing different modes, This represents the collaborative participation coefficient.

[0070] The vehicle's attitude adjustment process is divided into an attitude establishment phase (Build) and an attitude maintenance phase (Hold). During the pose establishment phase, define Attitude maintenance mode, defined ; During the attitude establishment phase, the CMG actuator and the steering actuator are controlled to work in opposite phases, so that the direction of the gyroscopic coupling torque generated by the CMG is opposite to the direction of the roll adjustment action generated by the steering actuator, as shown below: ; During the attitude maintenance phase, the CMG actuator and the steering actuator are controlled to work in phase, so that the direction of the gyroscopic coupling torque generated by the CMG is the same as the direction of the roll adjustment action generated by the steering actuator, as expressed as: ; in , It can be preferably designed as a function that varies with vehicle speed.

[0071] Based on the optimal feedback controller, a Lyapunov energy function is constructed, and the Lyapunov energy function is used to identify whether the vehicle is currently in the attitude establishment phase or the attitude maintenance phase.

[0072] Calculate the rate of change of the Lyapunov energy function; When the Lyapunov energy function and its rate of change are both less than their respective thresholds, and the turning reference rate of change meets the preset conditions, the recognition system enters the attitude maintenance phase; otherwise, the recognition system is in the attitude establishment phase.

[0073] Based on the optimal controller obtained in step S3, the Lyapunov energy function is constructed and expressed as: ; in The solution to the Riccati equation corresponding to the LQR controller is expressed as: ; This energy function can be used to characterize the degree to which a system deviates from its equilibrium point.

[0074] Simultaneously calculate the rate of energy change, expressed as: ; When the system satisfies and And the turning reference rate of change satisfies When the system is in the attitude maintenance phase, it enters the attitude establishment phase; otherwise, it remains in the attitude maintenance phase.

[0075] To avoid frequent mode switching, a hysteresis mechanism is preferred, expressed as: This improves the stability of modality recognition.

[0076] Based on Lyapunov energy construction, continuous cooperative weights are used to fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, and the final CMG cooperative control law is output. The vehicle is controlled based on the cooperative control law.

[0077] To avoid abrupt changes in control input caused by mode switching, this technique constructs a continuous cooperative weight based on an energy function, expressed as: ; The continuous collaborative weights are subjected to first-order filtering, and the control outputs of the anti-phase collaboration and the in-phase collaboration are fused based on the filtered continuous collaborative weights.

[0078] Define a normalized energy and construct a continuous cooperative weight function based on the Sigmoid function; where, when the normalized energy is greater than a preset threshold, the continuous cooperative weight corresponds to the attitude establishment stage; when the normalized energy is less than the preset threshold, the continuous cooperative weight corresponds to the attitude maintenance stage. Define normalized energy as follows: ; And construct the weight function, expressed as: ; in, This is the Sigmoid function.

[0079] when hour, The system is in the attitude establishment phase; when hour, The system is in the attitude maintenance phase.

[0080] To ensure smooth control, a first-order filter is applied to the weights: ; The final CMG control law is uniformly expressed as: ; in, This indicates CMG anti-coordination (attitude establishment phase); This indicates CMG in-phase coordination (attitude maintenance phase).

[0081] In step S4, Represents the solution matrix of the Riccati equation. Represents the LQR weight matrix. Denotes the Lyapunov energy function. This represents the energy normalization reference value. Indicates the CMG collaborative control weights. This represents the Sigmoid function. This represents the time constant of the weighted filtering.

[0082] Through the aforementioned energy-driven continuous coordinated control strategy, dynamic coordination between the steering actuator and the CMG actuator can be achieved at different attitude adjustment stages, thereby significantly improving the attitude establishment speed, stability, and anti-disturbance capability of the unmanned motorcycle across the entire vehicle speed range.

[0083] Example 2: This embodiment provides a steering-CMG cooperative control system for implementing the steering-CMG cooperative control method described in Embodiment 1, including: The dynamics model building module is used to build a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; The steering-CMG dual-modal cooperative control module is used to construct a steering-CMG dual-modal cooperative control strategy based on the basic control law. The final cooperative control module is used to construct continuous cooperative weights based on Lyapunov energy, fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, output the final CMG cooperative control law, and control the vehicle based on the cooperative control law.

[0084] Example 3 Reference Figure 5 As shown, this embodiment provides an unmanned motorcycle, including a controller, a CMG actuator, a rear wheel, a frame, a steering actuator, a front fork, and a front wheel. The controller controls the CMG actuator and the steering actuator through the steering-CMG cooperative control method described in Embodiment 1.

[0085] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A steering-CMG cooperative control method, characterized in that, include: Obtain a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; Based on the aforementioned steering-CMG extended dynamics model, an optimal feedback controller is established, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving the problem. Based on the aforementioned fundamental control law, a steering-CMG dual-modal cooperative control strategy is constructed, comprising: dividing the vehicle's attitude adjustment process into an attitude establishment phase and an attitude maintenance phase; in the attitude establishment phase, controlling the CMG actuator and the steering actuator to cooperate in opposite phases, so that the direction of the gyroscopic coupling torque generated by the CMG is opposite to the direction of the roll adjustment action generated by the steering actuator; in the attitude maintenance phase, controlling the CMG actuator and the steering actuator to cooperate in the same phase, so that the direction of the gyroscopic coupling torque generated by the CMG is the same as the direction of the roll adjustment action generated by the steering actuator. Based on Lyapunov energy construction, continuous cooperative weights are used to fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, and the final CMG cooperative control law is output. The vehicle is controlled based on the cooperative control law.

2. The steering-CMG coordinated control method according to claim 1, characterized in that, It also includes, Based on the optimal feedback controller, a Lyapunov energy function is constructed, and the Lyapunov energy function is used to identify whether the vehicle is currently in the attitude establishment phase or the attitude maintenance phase.

3. The steering-CMG cooperative control method according to claim 1, characterized in that, In obtaining the extended-dimensional dynamic model of the steering-CMG with the actuator angular velocity command as the unified input, the steps for obtaining the linear dynamic model with torque input include: Based on the Kane method, a multibody dynamics model including the frame, fork, front and rear wheels and dual CMG devices is established under the wheel-ground nonholonomic constraint. The dual CMG device is subjected to differential mode equivalent processing, and the multibody dynamics model is linearized at the uniform straight-line operating point to obtain the linear dynamics model of the torque input.

4. The steering-CMG coordinated control method according to claim 3, characterized in that, In obtaining the extended-dimensional dynamic model of the steering-CMG with the actuator angular velocity command as the unified input, the step of obtaining the extended-dimensional dynamic model based on the linear dynamic model with the torque input includes: A first-order dynamic model of the actuator is introduced to convert the torque input in the linear dynamic model of the torque input into the actuator angular velocity command input; By dynamically incorporating the actuators into the system through state expansion, the steering-CMG extended-dimensional dynamic model with the actuator angular velocity command as the unified input is obtained.

5. The steering-CMG coordinated control method according to claim 3, characterized in that, The differential mode equivalent processing of the dual CMG device includes: By setting the precession angles of the front and rear CMGs to be equal in magnitude and opposite in sign, a differential mode precession variable and a corresponding differential mode precession torque are introduced, and the dual CMG device is equivalent to a single precession degree of freedom.

6. The steering-CMG cooperative control method according to claim 1, characterized in that, The optimal feedback controller is established based on the steering-CMG extended dynamics model, and the basic control law containing the feedback gains of the steering actuator and the CMG actuator is obtained by solving it, including: Discrete speed sweeps are performed within a given vehicle speed range, and the extended-dimensional dynamic model is linearized at each vehicle speed operating point to obtain a vehicle speed parameterized state space model. For each vehicle speed operating point, a quadratic performance index function of system state and control input is constructed, the optimal feedback control problem is solved, and the control gain matrix corresponding to different vehicle speeds is obtained. A vehicle speed-control gain lookup table is constructed. Based on the real-time vehicle speed, interpolation calculations are performed in the lookup table to obtain the feedback control gain at the current vehicle speed, which is then used to construct the basic control law.

7. The steering-CMG cooperative control method according to claim 2, characterized in that, The process of identifying whether a vehicle is currently in the attitude establishment or attitude maintenance phase based on the Lyapunov energy function includes: Calculate the rate of change of the Lyapunov energy function; When the Lyapunov energy function and its rate of change are both less than their respective thresholds, and the turning reference rate of change meets the preset conditions, the recognition system enters the attitude maintenance phase; otherwise, the recognition system is in the attitude establishment phase.

8. The steering-CMG coordinated control method according to claim 1, characterized in that, The construction of continuous cooperative weights based on Lyapunov energy includes: Define a normalized energy and construct a continuous cooperative weight function based on the Sigmoid function; where, when the normalized energy is greater than a preset threshold, the continuous cooperative weight corresponds to the attitude establishment stage; when the normalized energy is less than the preset threshold, the continuous cooperative weight corresponds to the attitude maintenance stage. The continuous collaborative weights are subjected to first-order filtering, and the control outputs of the anti-phase collaboration and the in-phase collaboration are fused based on the filtered continuous collaborative weights.

9. A steering-CMG cooperative control system, used to implement the steering-CMG cooperative control method according to any one of claims 1-8, characterized in that, include: The dynamics model building module is used to build a steering-CMG extended-dimensional dynamics model with actuator angular velocity commands as the unified input; The steering-CMG dual-modal cooperative control module is used to construct a steering-CMG dual-modal cooperative control strategy based on the basic control law. The final cooperative control module is used to construct continuous cooperative weights based on Lyapunov energy, fuse the control outputs of the anti-phase cooperative and the in-phase cooperative based on the continuous cooperative weights, output the final CMG cooperative control law, and control the vehicle based on the cooperative control law.

10. An unmanned motorcycle, characterized in that, The vehicle includes a controller, a CMG actuator, a rear wheel, a frame, a steering actuator, a front fork, and a front wheel. The controller controls the CMG actuator and the steering actuator using a steering-CMG coordinated control method according to any one of claims 1-8.

Images