A hierarchical control method for vehicle lateral following based on virtual connections

Abstract

This invention discloses a hierarchical control method for vehicle lateral following based on virtual connections, relating to the field of intelligent control technology for mobile platforms. It solves the technical problem of inaccurate vehicle lateral following control methods under complex and changing traffic conditions. The key technical solution is a two-layer control system. The upper layer control is based on a vehicle lateral following model with virtual connections, introducing vehicle-to-everything (V2X) technology to improve the vehicle's environmental perception. Based on the perceived information of the surrounding environment and its own state, the vehicle adjusts the stiffness and damping parameters of the virtual connections to obtain the desired lateral acceleration that meets the vehicle's lateral following stability requirements. The lower layer control establishes a two-degree-of-freedom lateral model of the vehicle with front wheel sideslip angle and yaw rate as state variables, and designs a decoupled control strategy based on front wheel active steering. This achieves decoupling of lateral motion from longitudinal and yaw motions, enabling tracking of the desired lateral acceleration under different vehicle speeds.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology for mobile platforms, and in particular to a hierarchical control method for vehicle lateral following based on virtual connections. Background Technology

[0002] Breakthroughs and advancements in communication, computer, and artificial intelligence technologies have made the application of technologies such as autonomous driving and intelligent connectivity possible in the automotive industry. Vehicle following control is a crucial function of autonomous driving, and by adjusting the distance between the current vehicle and the vehicle in front in real time, it positively impacts driving safety, fuel economy, and performance. Lateral control ensures that the vehicle travels along a specific trajectory during autonomous driving. Clearly, the positive effects of distance control require the coordination and optimization of lateral control to be truly realized. Current lateral control methods for vehicle following mainly focus on lane keeping, which struggles to handle scenarios such as lane changes and obstacle avoidance. Furthermore, because this method relies on lane detection technology, it is difficult to apply to operating conditions such as inclement weather or unclear lane markings. In addition, current lateral control algorithms are typically based on linear two-degree-of-freedom models, which are based on constant vehicle speed conditions. The effectiveness of lateral control largely depends on the operating speed range and model uncertainties, making it difficult to adapt to the complex and ever-changing traffic conditions in real-world scenarios, as well as application scenarios involving longitudinal and lateral dynamic coupling, such as emergency lane changes and obstacle avoidance.

[0003] Therefore, the above problems urgently need to be solved. Summary of the Invention

[0004] This application provides a hierarchical vehicle lateral following control method based on virtual connection. Its technical objective is to provide a vehicle lateral following control method that is applicable to highway conditions and meets different vehicle speed conditions.

[0005] The above-mentioned technical objective of this application is achieved through the following technical solution:

[0006] A hierarchical vehicle lateral following control method based on virtual connectivity is disclosed. This hierarchical control method is implemented through a hierarchical control architecture, which includes a state perception module, a vehicle-to-vehicle communication module, and a hierarchical control module. The hierarchical control module includes an upper-level controller and a lower-level controller. The method includes:

[0007] S1: The upper-level controller constructs a vehicle lateral following model based on virtual connections;

[0008] S2: Obtain the desired lateral acceleration of the current vehicle based on the vehicle lateral following model;

[0009] S3: The lower-level controller constructs a two-degree-of-freedom model of the vehicle and designs a decoupling control strategy based on the current vehicle's desired lateral acceleration and the two-degree-of-freedom model of the vehicle.

[0010] S4: The decoupling control strategy is used to track the desired lateral acceleration under different vehicle speed conditions, thereby achieving hierarchical control of the vehicle's lateral following.

[0011] The beneficial effects of this application are as follows: (1) A two-layer control architecture is adopted. The upper layer control obtains the desired control parameters based on the perception information, and the lower layer control controls the vehicle driving according to the desired control parameters, thereby enhancing the robustness of the system to different operating conditions; (2) The upper layer control is based on the vehicle lateral following model with virtual connection, and introduces the Internet of Vehicles to improve the vehicle's perception of the environment. The vehicle obtains the desired lateral acceleration that meets the requirements of the vehicle's lateral following stability by adjusting the stiffness and damping parameters of the virtual connection based on the perception information of the surrounding environment and its own state; (3) The lower layer control establishes a two-degree-of-freedom lateral model of the vehicle with the front wheel slip angle and yaw rate as state variables. Based on this vehicle model, a decoupling control strategy is designed to realize the decoupling of lateral motion with longitudinal motion and yaw motion, and to achieve the tracking of the desired lateral acceleration under different vehicle speed conditions. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating the workflow of each module in this application;

[0013] Figure 2 This is a schematic diagram of a vehicle lateral following model based on virtual connections;

[0014] Figure 3 The upper-level control flowchart for vehicle lateral following layered control;

[0015] Figure 4 This is a flowchart of the lower-level control layer for vehicle lateral following layer control. Detailed Implementation

[0016] The technical solution of this application will be described in detail below with reference to the accompanying drawings.

[0017] The vehicle lateral following hierarchical control method based on virtual connections described in this application is implemented through a hierarchical control architecture, such as... Figure 1 As shown, this hierarchical control architecture includes a workshop communication module, a status awareness module, and a hierarchical control module. The workshop communication module includes a V2V communication device, which acquires the lateral speed v of the vehicle ahead. y,leadThe state perception module includes an onboard radar, an onboard camera, a vehicle speed sensor, and an inertial measurement unit. The hierarchical control module includes an upper-level controller and a lower-level controller. The upper-level controller calculates the desired trajectory parameters of the vehicle by sensing the surrounding environment and the vehicle's state, while the lower-level controller controls the vehicle to travel along the desired trajectory.

[0018] The vehicle lateral following hierarchical control method based on virtual connection described in this application includes:

[0019] S1: The upper-level controller builds a vehicle lateral following model based on virtual connections.

[0020] Specifically, step S1 includes:

[0021] S11: Based on the virtual connection structure, the virtual lateral force applied to the current vehicle is obtained, and this virtual lateral force is expressed as:

[0022] F y，vir =K p ΔL y (t-τ y )+K i ∫ΔL y (t-τ y )dt+C p (v y，lead (t-τ y )-v y (t-τ y ));

[0023] Among them, K p and K i All are virtual stiffness coefficients; K p The proportional gain represents the virtual spring elongation; K i The integral gain representing the virtual spring elongation; C p Indicates the virtual damping coefficient; ΔL y =YY n Y = y lead -y represents the lateral distance between the current vehicle and the vehicle in front. lead Indicates the lateral position of the vehicle in front, Y n Y represents the rated length of the virtual spring. n =T gap v y v y T represents the current lateral speed of the vehicle. gap Indicates the lateral following time interval; v y,lead τ represents the lateral velocity of the vehicle in front; for simplicity, τ y Take the maximum value of the time delay for the fusion processing of workshop communication and radar / camera information.

[0024] Using the lateral position of the preceding vehicle as a reference, the lateral distance Y between the current vehicle and the preceding vehicle is obtained, and the rated length Y of the virtual connection is defined. n Let v be the current lateral speed of the vehicle. y The function, i.e., Y n =T gap v y The elongation of the virtual connection can then be defined as YY. n denoted as ΔL y This is obtained through the positioning information of the preceding vehicle and the current vehicle. The lateral velocity v of the preceding vehicle is introduced through the vehicle-to-vehicle communication module. y,lead .

[0025] S12: Considering that vehicle lateral control is achieved by relying on the lateral force generated by the vehicle steering system, and taking into account the limitations of road friction coefficient and vehicle performance, the virtual lateral force F is... y,vir The design is based on the saturation boundary value of the virtual spring-damped system, expressed as:

[0026]

[0027] Where μ represents the road surface friction coefficient; ± indicates the sign and F y,vir The values ​​are the same; m represents the current vehicle mass.

[0028] S13: Based on the virtual lateral force F within the saturation boundary value range y,vir Calculate the acceleration of the current vehicle after the virtual connection is applied. Get to The output vehicle lateral following model, such as Figure 2 As shown, the lateral following model of the vehicle is represented as follows:

[0029]

[0030] S2: Obtain the desired lateral acceleration of the current vehicle based on the vehicle lateral following model.

[0031] Specifically, step S2 includes:

[0032] S21: Based on the vehicle lateral following model, the transfer function with the lateral velocity of the preceding vehicle as input and the lateral velocity of the following vehicle as output is obtained, expressed as:

[0033]

[0034] S22: The virtual stiffness coefficient K is obtained through the transfer function. p K i With virtual damping coefficient C pAdjustments are made to ensure stable operation of the vehicle's lateral following system. Once the lateral following system is stable, the desired lateral acceleration of the current vehicle is obtained.

[0035]

[0036] C(s)=C p ;

[0037] H(s) = 1 + T gap S;

[0038] Where K(s) represents the rigidity of the virtual connection; C(s) represents the damping of the virtual connection; H(s) is used to obtain the ideal lateral position of the preceding vehicle under the current vehicle following state, and then the elongation of the virtual connection is obtained by comparing it with the lateral position of the preceding vehicle.

[0039] Specifically, based on the Routh-Hurwitz stability criterion, the virtual stiffness coefficient K in the virtual spring-damper is... p K i With virtual damping coefficient C p Adjustments are made to ensure that lateral velocity disturbances occurring during vehicle operation eventually converge, thus ensuring the stable operation of the vehicle lateral following system. Therefore, the desired lateral acceleration of the current vehicle can be obtained through the aforementioned vehicle lateral following model based on virtual connections.

[0040] S3: The lower-level controller constructs a two-degree-of-freedom model of the vehicle and designs a decoupling control strategy based on the current vehicle's desired lateral acceleration and the two-degree-of-freedom model of the vehicle.

[0041] Specifically, considering the highway driving conditions involved in this application, where the vehicle travels at a constant speed and the sideslip angle and steering wheel angle are both relatively small, the front wheel sideslip angle β is established. f For a two-degree-of-freedom vehicle model where the yaw rate r is the state variable, the two-degree-of-freedom vehicle model can be expressed as:

[0042]

[0043] Where l represents the vehicle's wheelbase; f Indicates the distance from the front axle to the center of gravity; l r α represents the distance from the rear axle to the center of mass. f α represents the slip angle of the front tire. f =δ f -β f ;α r α represents the rear tire slip angle. r =-β f +lr / v;f ff represents the lateral force on the front wheel. r represents the lateral force on the rear wheel; v represents the longitudinal velocity.

[0044] Based on the two-degree-of-freedom model of the vehicle, and using the front wheel steering angle as the control input, a decoupled control strategy for front wheel active steering was designed, expressed as:

[0045]

[0046] Among them, a y,f This indicates the lateral acceleration of the front wheels. k s Indicates the acceleration feedback gain; δ f Indicates the steering angle of the front wheels.

[0047] S4: The decoupling control strategy is used to track the desired lateral acceleration under different vehicle speed conditions, thereby achieving hierarchical control of the vehicle's lateral following.

[0048] Specifically, Substituting into the two-degree-of-freedom model of the vehicle, we get This achieves decoupling of lateral movement from longitudinal movement and yaw movement at the front axle of the vehicle. The lateral decoupling control strategy can track the desired lateral acceleration under different vehicle speed conditions.

[0049] like Figure 3 As shown, in this application, the upper-level controller is based on a vehicle lateral following model with virtual connections. Vehicle networking is introduced to improve the vehicle's environmental perception. The vehicle combines its own state feedback with the perception information obtained from the vehicle-to-vehicle communication module, and adjusts the stiffness and damping parameters of the virtual spring-damper to obtain the desired lateral acceleration that meets the vehicle's lateral following stability requirements. Here, K(s) represents the rigidity characteristic of the virtual connection, K(s) = K p +K i / s,K p K i Both represent virtual stiffness coefficients, which are the proportional gain and integral gain of the virtual elastic connection elongation, respectively; C(s) represents the damping characteristic of the virtual connection, C(s) = C p C p Represents the virtual damping coefficient; H(s) = 1 + T gap The ideal lateral position of the preceding vehicle under the current following state is obtained through H(s), and then compared with the lateral position of the preceding vehicle to obtain the elongation of the virtual connection. Based on the Routh-Hurwitz stability criterion, the virtual stiffness coefficient K in the virtual connection is... p K i With virtual damping coefficient C pAdjustments are made to ensure that lateral velocity disturbances occurring during vehicle operation eventually converge, thereby ensuring the stable operation of the vehicle's lateral following system.

[0050] like Figure 4 As shown, the lower-level controller designs a decoupled control law for front wheel steering based on the vehicle's lateral dynamics model. In actual control, this decoupled control strategy depends on the desired control parameter 'a' output by the upper-level controller. y,des The current feedback quantities of the vehicle are longitudinal velocity v, yaw rate r, and lateral acceleration a at the front axle. y,f In this case, the longitudinal velocity v acts on the lateral acceleration a. y,f The feedback gain is included. This feedback control strategy decouples the lateral, longitudinal, and yaw motions at the front axle of the vehicle, enabling tracking of the desired lateral acceleration under different vehicle speeds.

[0051] A two-layer control architecture is adopted. The upper-layer control obtains the desired control parameters based on perceived information, while the lower-layer control controls the vehicle's movement according to the desired control parameters, enhancing the system's robustness to different operating conditions. The upper-layer control is based on a virtual-connected vehicle lateral following model, introducing vehicle-to-everything (V2X) technology to improve the vehicle's environmental perception. The vehicle combines its own state feedback with perceived information obtained from the vehicle-to-vehicle communication module, adjusting the stiffness and damping coefficients of the virtual connection to obtain the desired lateral acceleration that meets the vehicle's lateral following stability requirements. The lower-layer control establishes a two-degree-of-freedom lateral model of the vehicle with front wheel sideslip angle and yaw rate as state variables. Based on this vehicle model, a decoupling control strategy is designed to decouple lateral motion from longitudinal and yaw motion, enabling tracking of the desired lateral acceleration under different vehicle speeds.

[0052] The above are exemplary embodiments of this application, and the scope of protection of this application is defined by the claims and their equivalents.

Claims

1. A virtual connection-based vehicle lateral following hierarchical control method, the hierarchical control method being implemented through a hierarchical control architecture, the hierarchical control architecture comprising a state perception module, an inter-vehicle communication module, and a hierarchical control module, the hierarchical control module comprising an upper controller and a lower controller, characterized in that, The method comprises: S1: a higher-level controller constructs a vehicle lateral following model based on virtual connection; S2: a desired lateral acceleration of a current vehicle is obtained according to the vehicle lateral following model; S3: a lower-level controller constructs a two-degree-of-freedom vehicle model, and designs a decoupling control strategy according to the desired lateral acceleration of the current vehicle and the two-degree-of-freedom vehicle model; S4: the desired lateral acceleration is tracked under different vehicle speeds through the decoupling control strategy, so as to realize hierarchical control of vehicle lateral following; In step S1, the following steps are included: S11: a virtual lateral force applied to the current vehicle is obtained based on the virtual connection structure, and the virtual lateral force is expressed as: ; wherein, and are virtual stiffness coefficients; represents a proportional gain of the virtual spring elongation; represents an integral gain of the virtual spring elongation; represents a virtual damping coefficient; Δ , represents a lateral distance between the current vehicle and the preceding vehicle, represents a lateral position of the preceding vehicle, represents a rated length of the virtual spring, , represents a lateral speed of the current vehicle, represents a lateral following distance; represents a lateral speed of the preceding vehicle; represents a maximum value of a time lag of the inter-vehicle communication and the radar, camera information fusion processing; S12: According to the road friction coefficient, the virtual lateral force The saturation boundary value based on the virtual spring-damping is designed, which is represented as: ; wherein, represents a road surface friction coefficient; represents a sign which is the same as the value of represents a mass of the current vehicle; S13: virtual lateral force within the saturation boundary value range calculating the acceleration of the current vehicle after the virtual connection force , obtaining a vehicle lateral following model with as output, the vehicle lateral following model being represented as: 。 2. The method of claim 1, wherein, The inter-vehicle communication module comprises a V2V communication device, the lateral speed of the preceding vehicle is acquired by means of the V2V communication device.

3. The method of claim 2, wherein, The step S2 comprises: S21: a transfer function with the lateral speed of a preceding vehicle as input and the lateral speed of a following vehicle as output is obtained according to the vehicle lateral following model, and is expressed as: ; S22: adjusting the virtual stiffness coefficient , and the virtual damping coefficient by the transfer function to ensure that the vehicle lateral following system is stable, the desired lateral acceleration of the current vehicle is obtained .

4. The method of claim 3, wherein, In step S3, the two-degree-of-freedom vehicle model is expressed as: ; wherein, denotes the wheelbase of the vehicle; denotes the distance of the front axle to the center of mass; denotes the distance of the rear axle to the center of mass; denotes the front wheel tire side slip angle, denotes the front wheel side slip angle; denotes the rear wheel tire side slip angle, denotes the front wheel lateral force; denotes the rear wheel lateral force; denotes the yaw rate; denotes the longitudinal velocity;​​​​​​ Based on the two-degree-of-freedom vehicle model, a decoupling control strategy of front-wheel active steering is designed with a front-wheel steering angle as a control input, and the decoupling control strategy of the front-wheel active steering is expressed as: ； wherein, represents a front wheel lateral acceleration, = a x (a x + b x) + c x (a x + b x)2+ d x (a x + b x)3; ; represents an acceleration feedback gain; represents a desired lateral acceleration; represents a front wheel steering angle.

5. The method of claim 4, wherein, In step S4, the Substituting into the vehicle 2-DOF model, we get , which realizes the decoupling of lateral and longitudinal motions and yaw motion at the front axle of the vehicle, thus realizing the tracking of the desired lateral acceleration under different vehicle speeds.

Images