A Triggered Model Prediction-Based Lane Change Trajectory Tracking Control Method Considering Driving Habits

Abstract

This invention provides a trigger-based model predictive trajectory tracking control method that considers driving habits, belonging to the field of intelligent driving control technology. The method mainly includes: establishing a two-degree-of-freedom vehicle dynamics model and a nonlinear kinematics model; transforming the vehicle dynamics model and vehicle kinematics model into discrete predictive models, and designing the objective function and constraints of the model predictive controller; designing the triggering conditions and control strategy of the tracking controller based on the driver's following error; and calculating the control input of the model predictive controller from the desired trajectory. This invention designs the triggering conditions and control logic of the model predictive controller based on the driver's road following characteristics. By controlling the front wheel steering angle of the vehicle to track the desired trajectory, it can ensure tracking accuracy while reducing the computational load of the controller and meeting the driver's driving needs.

Description

Technical Field

[0001] This invention belongs to the field of intelligent vehicle control technology, specifically relating to a trigger-based model prediction lane-changing trajectory tracking control method that takes into account driving habits. Background Technology

[0002] To improve driving safety and reduce driver workload, advanced driver assistance systems (ADAS) have become a hot research topic. Lane changing is a common operation in daily driving and a frequent scenario for traffic accidents. When a vehicle changes lanes according to instructions and follows the desired trajectory, the ADAS transmits vehicle status information to the controller via onboard sensors. The controller then calculates the appropriate front wheel steering angle to provide steering assistance to the driver, enabling the vehicle to complete the lane change smoothly.

[0003] Before fully automated driving is achieved, intelligent vehicles will remain in a state of partial autonomous driving or human-machine co-driving for a long time. However, when intelligent driving systems are operating, drivers inevitably experience tension due to external inputs. If this tension is not effectively managed, it may lead to potential dangers, reduce driver comfort, and thus affect drivers' acceptance of intelligent vehicles. Summary of the Invention

[0004] To address the shortcomings of the existing technology, this invention discloses a trigger-based model prediction lane-change trajectory tracking control method that considers driving habits. This method improves system computational efficiency while ensuring tracking performance and provides steering assistance that aligns with the driver's following habits. The implementation process of this invention is as follows:

[0005] A trigger-based model prediction-based lane-change trajectory tracking control method that considers driving habits includes the following steps:

[0006] 1) Perform force analysis on the vehicle and build a two-degree-of-freedom vehicle dynamics model:

[0007]

[0008] In the formula, m is the total vehicle mass; v x and v y These represent the vehicle's longitudinal and lateral speeds, respectively. Indicates the vehicle's heading angle; F xf F represents the longitudinal force of the front tire. yf and F yr These represent the lateral forces of the front and rear tires, respectively; δ f The steering angle of the vehicle's front wheels; I z Let l be the moment of inertia of the vehicle about the z-axis; f and l r These are the distances from the vehicle's center of gravity to the front and rear axles, respectively.

[0009] Considering the small steering angle of the vehicle's front wheels, sinδ f ≈δ f cosδ f ≈1, the front and rear tire slip angles can be approximated as:

[0010]

[0011] In the formula, α1 and α2 are the front and rear tire slip angles, respectively; β is the vehicle center of gravity slip angle; and γ is the vehicle yaw rate.

[0012] The lateral forces of the front and rear tires are:

[0013]

[0014] In the formula, k1 and k2 are the lateral stiffness of the front and rear tires of the vehicle, respectively;

[0015] Substituting into the above equation, the vehicle dynamics equation can be simplified to:

[0016]

[0017] 2) Vehicle motion in the global coordinate system can be represented as:

[0018]

[0019] In the formula, x o and y o These represent the longitudinal and lateral coordinates of the vehicle's center of mass, respectively. Assuming a constant longitudinal velocity and considering the vehicle's relatively small heading angle, therefore, we assume... Therefore, the vehicle kinematic model can be simplified to:

[0020]

[0021] 3) Based on the vehicle dynamics and kinematics model, construct the state equations:

[0022]

[0023] In the formula, the state quantity Control quantity u = δ f State matrix Control Matrix Output matrix C c =[1 1 00]; y is the output quantity;

[0024] 4) The nonlinear continuous prediction model (obtained after processing the state equations) is transformed into a discretized prediction model using the Euler method:

[0025]

[0026] In the formula, A d =1+T*A c B d =T*B;C d =C c T represents the sampling time; k represents the system's deviation length at time t.

[0027] 5) To limit the control increment, design new state variables. This yields a new state-space expression:

[0028]

[0029] In the formula, n is the dimension of the state variables, and s is the dimension of the control variables;

[0030] 6) The system's prediction model is:

[0031] Y=Ψξ+ΦΔU (10)

[0032] In the formula,

[0033] N c It controls the time domain, N p For prediction in the time domain;

[0034] 7) Establish the objective function and constraints of the model predictive controller.

[0035] Define the objective function:

[0036]

[0037] In the formula, Q and R are the weights of the output and control quantities, respectively, ρ is the weight of the relaxation factor, and ε is the relaxation factor.

[0038] Design constraints:

[0039] Control constraints (i.e., physical actuator constraints):

[0040] u min ≤u t+k ≤u max ,0≤k≤N c -1 (12)

[0041] Control Incremental Constraints:

[0042] Δu min ≤u t+k -u t+k-1 ≤Δu max ,0≤k≤N c -1 (13)

[0043] Output constraints:

[0044] Y min ≤Ψξ+ΦΔU≤Y max (14)

[0045] 8) When driving on the road, drivers may not always stay in the middle of the road due to different personal habits and needs, and may therefore deviate to the left or right.

[0046] Traditional time-triggered model predictive controllers perform optimization at every moment to calculate the best control sequence that fits the current situation, but only use the first control variable of the control sequence to input to the control system.

[0047] To enable the controller to provide additional steering angles that match the driver's tracking habits, mimic the driver's driving, and reduce the computational load of the model predictive controller, an event-triggered model predictive controller based on the driver's driving habits was designed. The control strategy is as follows:

[0048] The controller only performs calculations to solve for a new control sequence when the error between the actual trajectory and the reference trajectory exceeds a certain threshold or when the controller's optimal control sequence is exhausted; at other times, the control quantities are input to the control system one by one using the control sequence calculated last time.

[0049] 9) Design the triggering conditions for the model predictive controller:

[0050]

[0051] In the formula, σ is the trigger threshold, determined by the driver's sensitivity to vehicle deviation; k max This represents the maximum capacity of the control sequence, where k represents the maximum capacity of the control sequence. max =N c ;

[0052] 10) Through the above steps, the controller can provide the driver with the desired cornering assistance, keeping the error between the vehicle's actual trajectory and the reference trajectory within an acceptable range for the driver. This ensures the tracking effect while reducing the computational load of the control system to a certain extent, thus meeting the driver's needs.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. This invention can provide steering assistance to drivers in accordance with their road following habits. It analyzes the driver's driving data to obtain the driver's following error sensitivity. Based on this, it designs a model to predict the trigger threshold of the control. When the system's lateral error reaches the threshold or the control sequence is exhausted, the tracking controller starts to calculate a new control sequence to provide the driver with more suitable steering operations and improve driving comfort.

[0055] 2. Compared with traditional time-triggered model predictive control strategies, tracking control by designing event-triggered conditions can reduce the computational burden on the controller while ensuring tracking performance. Attached Figure Description

[0056] Figure 1 Schematic diagram of the process of the present invention;

[0057] Figure 2 This invention addresses the lateral error of tracking a reference trajectory at 60 km / h. Detailed Implementation

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] Example:

[0060] The flowchart of this embodiment is as follows: Figure 1 As shown, this embodiment provides a trigger-based model prediction lane-change trajectory tracking control method that considers driving habits. A two-degree-of-freedom vehicle dynamics model and a nonlinear kinematics model are established. The prediction model of the model predictive controller, its objective function, and constraints are designed. Based on the driver's road following characteristics, the triggering conditions and control strategy of the tracking controller are designed, including the following steps:

[0061] 1. Perform force analysis on the vehicle and build a two-degree-of-freedom vehicle dynamics model:

[0062]

[0063] In the formula, m is the total vehicle mass, 1903 kg; v x and v y These represent the vehicle's longitudinal and lateral speeds, respectively. Indicates the vehicle's heading angle; F xf F represents the longitudinal force of the front tire. yf and F yr These represent the lateral forces of the front and rear tires, respectively; δf The steering angle of the vehicle's front wheels; I z The vehicle's moment of inertia about the z-axis is 1536.7 kg / m. 2 ,;l f and l r These are the distances from the vehicle's center of gravity to the front and rear axles, respectively, which are 1.232m and 1.468m.

[0064] Considering the small steering angle of the vehicle's front wheels, sinδ f ≈δ f cosδ f ≈1, the front and rear tire slip angles can be approximated as:

[0065]

[0066] In the formula, α1 and α2 are the front and rear tire slip angles, respectively; β is the vehicle center of gravity slip angle; and γ is the vehicle yaw rate.

[0067] The lateral forces of the front and rear tires are:

[0068]

[0069] In the formula, k1 and k2 are the lateral stiffness of the front and rear tires of the vehicle, respectively.

[0070] Substituting the front and rear tire slip angles and the front and rear tire lateral forces into the vehicle dynamics model, the vehicle dynamics equations can be simplified to:

[0071]

[0072] 2. Vehicle motion in the global coordinate system can be represented as:

[0073]

[0074] In the formula, x o and y o These represent the longitudinal and lateral coordinates of the vehicle's center of gravity, respectively.

[0075] 3. Assuming the vehicle's longitudinal speed is constant, and considering the vehicle's relatively small heading angle, therefore, we assume... Therefore, the vehicle kinematic model can be simplified to:

[0076]

[0077] 4. Based on the vehicle dynamics and kinematics model, construct the state equations:

[0078]

[0079] 5. Select the state variable as The control quantity is u = δ f .

[0080] In the formula, the matrix Control Matrix Output matrix C c = [1 1 0 0]; y is the output quantity.

[0081] 6. By discretizing the nonlinear continuous prediction model using the Euler method, we can obtain:

[0082]

[0083] In the formula, A d =1+T*A c B d =T*B;C d =C c T is the sampling time, which is set to 0.01 here; k represents the discrete step size of the system at time t, which is set to 0 here.

[0084] 7. To limit the control increment Δu(t) (i.e., the control increment at time t in ΔU), design new state variables. x(k|t) and u(k-1|t) represent the state variable and control quantity at time t, respectively, thus obtaining a new state-space expression:

[0085]

[0086] In the formula, n is the dimension of the state variables, and s is the dimension of the control variables.

[0087] 8. The system's prediction model is:

[0088] Y=Ψξ+ΦΔU

[0089] In the formula,

[0090] N c This controls the time domain; set it to 5. p Set to 20 for the time domain prediction;

[0091] 9. Establish the objective function and constraints of the model predictive controller.

[0092] Define the objective function:

[0093]

[0094] Here, ρ is the relaxation factor weight, which is set to 1000, and ε is the relaxation factor, which is set to 10.

[0095] Design constraints:

[0096] Control constraints (i.e., physical actuator constraints):

[0097] u min ≤u t+k ≤u max ,0≤k≤N c -1

[0098] Control Incremental Constraints:

[0099] Δu min ≤u t+k -u t+k-1 ≤Δu max ,0≤k≤N c -1

[0100] Output constraints:

[0101] Y min ≤Ψξ+ΦΔU≤Y max

[0102] 10. An event-triggered model predictive controller based on driver driving habits, with the following control strategy:

[0103] The controller only performs calculations to solve for a new control sequence when the error between the actual trajectory and the reference trajectory exceeds a certain threshold or when the controller's optimal control sequence is exhausted; at other times, the control quantities are input to the control system one by one using the control sequence calculated last time.

[0104] Design the triggering conditions for the model predictive controller:

[0105]

[0106] Here σ = 0.253m, k max =5.

[0107] Through the above steps, the controller can provide the driver with the desired cornering assistance, keeping the error between the vehicle's actual trajectory and the reference trajectory within an acceptable range for the driver. This ensures tracking performance while reducing the computational load on the control system to some extent, thus meeting the driver's needs.

[0108] like Figure 2The figure shows the lateral error of the present invention when tracking the reference trajectory at 60 km / h. Here, tMPC is a traditional time-triggered model predictive controller, while eMPC is a trigger-based model predictive controller improved in this paper, illustrating the tracker's tracking performance. Table 1 below shows the controller triggering conditions of the present invention when tracking the reference trajectory at 60 km / h and 80 km / h, respectively. Here, tMPC is a traditional time-triggered model predictive controller, while eMPC is a trigger-based model predictive controller improved in this paper, illustrating the tracker's control computation efficiency.

[0109] Table 1. Controller triggering status for tracking the reference trajectory at 60km / h and 80km / h.

[0110]

[0111] The embodiments of the present invention are for illustrative and descriptive purposes only. Many modifications and variations will be apparent to those skilled in the art. By selecting and describing these embodiments, we aim to better elucidate the principles and practical applications of the invention, enabling those skilled in the art to gain a thorough understanding of the invention and design various embodiments with various modifications according to specific needs.

Claims

1. A trigger-based model prediction lane-changing trajectory tracking control method considering driving habits, characterized in that, Includes the following steps: Step 1: Establish a two-degree-of-freedom dynamic model and a nonlinear kinematic model for the vehicle, perform force analysis on the vehicle, and construct the vehicle dynamic equations, expressed as follows: ， In the formula, and These refer to the lateral stiffness of the front and rear tires of the vehicle. The sideslip angle is the angle between the vehicle's center of gravity and its body. Let yaw rate be the vehicle's angular velocity. For the overall vehicle weight; Represents the longitudinal speed of the vehicle; Indicates the vehicle's heading angle; The steering angle of the vehicle's front wheels; For vehicles to bypass Moment of inertia of the shaft; and These are the distances from the vehicle's center of gravity to the front and rear axles, respectively. The vehicle kinematics model is constructed as follows: ， In the formula, and Let be the longitudinal and lateral coordinates of the vehicle's center of gravity, respectively, and represent them as follows: , , , Let yaw rate be the vehicle's angular velocity. Indicates the vehicle's heading angle. Represents the longitudinal speed of the vehicle. The sideslip angle is the angle between the vehicle's center of gravity and its body. Step 2: Based on the vehicle dynamics and kinematics model, establish the prediction model of the model predictive controller and transform it into a discrete prediction model. In the formula, This is the predicted output matrix of the model. The state matrix, For the state variable matrix, For the control matrix, To control the increment matrix, specifically, ； ； ； ; It controls the time domain. For prediction in the time domain; Step 3: Based on the vehicle's state variables and control increments, design a model to predict the controller's objective function, and set constraints based on the magnitude of the control variable, the control variable increment, and the output variable to obtain the optimal control sequence at the current moment. Step 4: Based on the driver's following error, design the triggering conditions and control strategy for the tracking controller. The triggering condition is set as follows: when the distance between the vehicle's actual position and the reference trajectory exceeds a trigger threshold or the control sequence reaches its maximum capacity, the controller is triggered to perform calculations. Specifically, the triggering conditions for the model prediction controller are designed as follows: ， In the formula, , The trigger threshold is determined by the driver's sensitivity to vehicle deviation; This represents the maximum capacity of the control sequence. The control strategy is set as follows: when the error between the actual trajectory and the reference trajectory exceeds a certain threshold or the controller's optimal control sequence is exhausted, the controller will perform calculations to solve for a new control sequence and input the first control quantity of the new control sequence to the control system. At other times, the control quantities are fed to the control system one by one, following the control sequence calculated in the previous operation.

2. The trigger-based model prediction lane-change trajectory tracking control method considering driving habits as described in claim 1, characterized in that, Step 2 includes: Based on the vehicle dynamics and kinematics model, the state equations are constructed as follows: ， In the formula, the state variables Control quantity State matrix Control Matrix Output matrix ; For output quantity; The nonlinear continuous prediction model is transformed into a discretized prediction model using the Euler method: ， In the formula, ; ; ; Sampling time; Indicates that the system is in The time spent apart from the long walk.

3. The trigger-based model prediction lane-change trajectory tracking control method considering driving habits as described in claim 1, characterized in that, In step 2, To control increment To impose constraints, design new state variables. , , They represent in The state variables and control variables at each moment are used to obtain a new state-space expression: ， In the formula, ; ; ; For the dimension of the state quantity, To control the dimensions of the quantity, express An identity matrix of dimension 1.

4. The trigger-based model prediction lane-change trajectory tracking control method considering driving habits according to claim 3, characterized in that, Step 3 includes: Define the objective function: ， In the formula, and These are the weights of the output and control variables, respectively. For relaxation factor weights, It is a relaxation factor; Design constraints: Control constraints: ， Control Incremental Constraints: ， Output constraints: 。 5. A trigger-based model prediction lane-change trajectory tracking control system that considers driving habits, characterized in that, The system implementing the trigger-based model prediction lane-changing trajectory tracking control method considering driving habits as described in claim 1 includes a vehicle kinematics and dynamics module and an event-triggered model prediction controller, wherein... The vehicle kinematics and dynamics module includes a two-degree-of-freedom vehicle model and a vehicle kinematics model, and outputs state variables. Where y is the output quantity. Let yaw rate be the vehicle's angular velocity. The sideslip angle is the angle at the vehicle's center of gravity. Indicates the vehicle's heading angle; The event-triggered model predictive controller includes a trigger module, a control sequence module, and a prediction module. The trigger module determines whether a triggering condition is met. If the condition is met, the prediction model obtains the optimal solution based on the objective function and constraints, and outputs an updated control sequence. To the control sequence module; The control sequence outputs a series of control quantities. The vehicle is given the ability to track and control the reference trajectory.

6. The trigger-based model prediction lane change trajectory tracking control system considering driving habits according to claim 5, characterized in that, Predictive Model , ； ； ； ; It controls the time domain. For prediction in the time domain.

7. The trigger-based model prediction lane change trajectory tracking control system considering driving habits according to claim 6, characterized in that, Objective function: ， In the formula, and These are the weights of the output and control variables, respectively. For relaxation factor weights, It is a relaxation factor; Constraints: Control constraints (i.e., physical actuator constraints): ， Control Incremental Constraints: ， Output constraints: 。

Images