A vehicle longitudinal and lateral optimization control method suitable for a curved road cruising scene

Abstract

The application is suitable for the technical field of vehicle control, and provides a vehicle longitudinal and lateral optimization control method suitable for a curved road cruise scene, which is established based on a model under a Frenet coordinate system, and the control effect is better than that of a traditional control under a geodetic coordinate system. The application not only solves the curved road adaptive cruise problem by using an LQR algorithm, but also takes into account the comfort index and the most important real-time index in engineering application. The longitudinal control strictly tracks the changing safe distance in the case of keeping a small relative speed in the safe distance processing, and properly handles the two contradictory indexes of keeping the relative speed between vehicles and the relative distance between vehicles, and the hierarchical control used in the longitudinal control is better than the traditional straight-through control.

Description

Technical Field

[0001] This invention belongs to the field of vehicle control technology, and in particular relates to a vehicle lateral and longitudinal optimization control method suitable for curve cruise scenarios. Background Technology

[0002] In recent years, with the improvement of people's living standards, the number of cars in my country has been increasing year by year. As a result, traffic accidents caused by human factors have also been increasing. The proactive development of vehicle driver assistance systems is imminent, especially for long-distance driving conditions of passenger cars. Adaptive Cruise Control (ACC), as an important component of vehicle driver assistance, has received widespread attention. However, at present, most research on ACC systems is limited to straight roads. Due to the difficulty in designing lateral controllers when cornering, research on cornering cruise control is relatively scarce.

[0003] Current research on cornering control mostly uses a combination of path tracking and straight-line adaptive cruise control. Path tracking uses PID or Active Disturbance Rejection Control (ADRC). PID and ADRC are robust and fast in adjustment, but the front wheel steering angle calculated by these two algorithms is abrupt. These algorithms only "brutally" adjust the error and do not consider the smoothness of the front wheel steering angle change, thus failing to meet the comfort indicators required for vehicle control. Some researchers use the MPC algorithm. The MPC algorithm needs to consider information in the prediction time domain, and a new calculation is performed at each sampling time, which is extremely computationally expensive and has poor real-time performance. Summary of the Invention

[0004] The purpose of this invention is to provide a vehicle lateral and longitudinal optimization control method suitable for curve cruise scenarios, aiming to solve the problems mentioned in the background art.

[0005] The present invention is implemented as follows: a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios includes:

[0006] Acquire the status information of this vehicle and the status information of the vehicle in front, and record them as the first motion information; acquire the road position information, road curvature and heading angle information, and record them as the second motion information.

[0007] Based on the first motion information and the second motion information, calculate and output throttle opening information, brake pedal information and steering wheel angle information;

[0008] Based on throttle opening information, brake pedal information, and steering wheel angle information, the vehicle's lateral and longitudinal control is performed:

[0009] In the longitudinal control of the vehicle, a variable vehicle spacing is selected as the safe vehicle spacing strategy, and the relative speed, relative displacement, vehicle speed and vehicle acceleration of the inter-vehicle are selected as state variables. A model predictive control algorithm is used to establish the inter-vehicle model and obtain the desired vehicle spacing and desired acceleration.

[0010] Based on the workshop model, a reverse longitudinal dynamics model of the vehicle is established, a suitable acceleration is optimized, and the acceleration is tracked. The desired acceleration is converted into the corresponding throttle opening and brake master cylinder pressure, which are applied to the vehicle itself to complete the longitudinal control task.

[0011] When performing lateral control of the vehicle, a global path is obtained based on the second motion information. The global path is then smoothed to obtain the vehicle driving reference line information, so that the vehicle travels along the vehicle driving reference line information.

[0012] A further technical solution employs a fixed inter-vehicle time-distance algorithm for vehicle spacing design, aiming to update the safe vehicle distance in real time based on the vehicle's speed.

[0013] d des =τv h +d0

[0014] In the formula d des τ is the expected safe following distance, d0 is the inter-vehicle time distance, v is the minimum maintaining distance, and τ is the minimum maintaining distance. h That is the speed of this vehicle.

[0015] A further technical solution involves performing longitudinal vehicle control, where the equation of motion for the vehicle during longitudinal acceleration is shown below:

[0016] ma des =F t -F f -F w

[0017] Where F t It is the driving force on the tires generated by the engine's driving action; F f It is the rolling resistance from the ground; F w It is air resistance, and the effect of rotating mass is ignored.

[0018] A further technical solution is to determine the relationship between braking force and braking pressure P when the braking force is not greater than the maximum braking force on the road surface. des The following relationship must be satisfied:

[0019]

[0020] In the formula, F xb T represents the braking force on the wheel. bf It is the sum of the braking torques of the front wheels, Tbr It is the sum of the braking torques of the rear wheels, K b This is the conversion coefficient between braking force and braking pressure, where r is the tire rolling radius; the desired braking pressure is calculated based on the vehicle's longitudinal dynamics equations as follows:

[0021]

[0022] Meanwhile, considering that acceleration and braking will not occur simultaneously during vehicle operation, the drive-brake switching logic is designed using the vehicle acceleration-speed-brake drag curve. To prevent continuous logic switching during vehicle control, a buffer zone with a width of 0.2g is designed.

[0023] A further technical solution involves incorporating the desired vehicle spacing into the relative error of the workshop when establishing the workshop model. That is, when the relative error of the workshop is 0, the actual vehicle spacing is the desired vehicle spacing, as shown below:

[0024] Δd=x lead -x h -d des

[0025] Δv=v lead -v h

[0026] Where x lead It is the displacement of the vehicle in front, x h This is the displacement of the vehicle, d des It is the expected workshop distance, v lead It's the speed of the car in front, v h Δd is the vehicle speed, Δv is the relative error between the two vehicles, and Δv is the relative speed between the two vehicles.

[0027] By combining the velocity and acceleration equations, the longitudinal control state space equations of the vehicle are constructed as follows:

[0028]

[0029] State variable x = [Δd Δv a] h v h ] T Control quantity u = a des Interference The coefficient matrix is ​​as follows:

[0030]

[0031] Discretize it using the forward Euler method. The discretized system is shown below:

[0032]

[0033] In the formula T is the sampling time.

[0034] A further technical solution involves using the desired acceleration as the control variable for the discretized system, and constructing a system with the acceleration change as the control variable:

[0035]

[0036] A new state-space expression is derived:

[0037]

[0038] In the formula Δu(k) is the control increment at time k; and the output equation is constructed as follows:

[0039] η(k+1)=Dξ(k+1)

[0040] In the formula Output the inter-vehicle error and the relative speed between the two vehicles;

[0041] Define N p For prediction in the time domain, N c To control the time domain, and N p ≥N c Based on the iterative derivation of η(k+1)=Dξ(k+1), we get:

[0042]

[0043] In the formula:

[0044]

[0045]

[0046]

[0047] The interference is the acceleration of the vehicle in front. Since the acceleration of the vehicle in front cannot be obtained within the prediction time domain, the interference of the vehicle in front within the entire prediction time domain is not available. constant.

[0048] Further technical solutions, considering the comfort of the ride during control and the energy efficiency of vehicle control, therefore, considering the performance indicators of comfort, tracking performance, and fuel economy, define the following performance indicators:

[0049]

[0050] Where η ref The expected output value is expressed as follows: Rewrite the above equation in matrix form:

[0051] J = [YY] ref ] TQ[YY ref ]+Δu T RΔu

[0052] In the formula, Q and R are the output weight matrix and the control weight matrix, respectively;

[0053] Will Substituting into the formula J = [YY] ref ] T Q[YY ref ]+Δu T RΔu, and let Summarized as follows:

[0054] J = Δu T (θ T Qθ+R)Δu+2E T QΔθu+E T QE

[0055] Let H = 2(θ) T Qθ+R), f T =2E T Qθ then has:

[0056]

[0057] Because E T QE is a constant term in the calculation and does not affect the solution of the optimal problem, so it is discarded, resulting in a standard quadratic programming problem:

[0058]

[0059] The vehicle's acceleration and its rate of change must be limited to meet comfort and fuel economy performance indicators. Research indicates that the optimal acceleration for human passengers is between -3 and 2. However, excessively drastic acceleration changes can also negatively impact comfort. The following constraints are designed, where u... min =-3, u max =2, representing the minimum and maximum acceleration values ​​respectively, with acceleration increment constraints ranging from -1.5 to 1.5, where Δu min =-1.5, Δu max =1.5, representing the minimum and maximum values ​​of the acceleration increment, respectively;

[0060]

[0061] Rewrite the control constraints as quadratic programming constraints:

[0062]

[0063] in,

[0064]

[0065] Control incremental constraints are

[0066] ΔU min ≤ΔU≤ΔU max

[0067] in,

[0068]

[0069] This invention provides a vehicle lateral and longitudinal optimization control method suitable for curve cruise scenarios. Using a model based on the Frenet coordinate system, the control performance is superior to that of traditional geodetic coordinate system control. This invention not only solves the curve adaptive cruise problem using the LQR algorithm but also considers comfort indicators as well as the most important real-time indicators in engineering applications. In terms of longitudinal control, it strictly tracks changing safe following distances while maintaining relatively low relative speeds, properly handling the conflicting indicators of maintaining relative speed and relative distance between vehicles. Furthermore, the use of hierarchical control in longitudinal control is superior to traditional direct-flow control. Attached Figure Description

[0070] Figure 1 A flowchart of the design of adaptive cruise control for curves, which is a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios, is provided for an embodiment of the present invention.

[0071] Figure 2 A flowchart illustrating the ACC longitudinal controller algorithm design in a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario, provided in an embodiment of the present invention.

[0072] Figure 3 A flowchart illustrating the design of a lateral controller algorithm in a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario, provided in an embodiment of the present invention.

[0073] Figure 4 This is a design diagram of a lower-level controller in a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario, provided by an embodiment of the present invention.

[0074] Figure 5 This invention provides a fitted drag curve in a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario, as provided in an embodiment of the present invention.

[0075] Figure 6 The projection diagram of the vehicle in the Frenet coordinate system in a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario provided by an embodiment of the present invention;

[0076] Figure 7This invention provides an algorithm solution time diagram for a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario, as provided in an embodiment of the present invention.

[0077] Figure 8 A comparison of the safe distance of a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by an embodiment of the present invention in straight-line adaptive cruise.

[0078] Figure 9 A speed comparison of a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by an embodiment of the present invention in straight-line adaptive cruise.

[0079] Figure 10 A comparison of longitudinal displacement in straight-line adaptive cruise using a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by an embodiment of the present invention;

[0080] Figure 11 A comparison of lateral displacement in straight-line adaptive cruise using a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by an embodiment of the present invention;

[0081] Figure 12 The vehicle trajectory diagram in straight-line adaptive cruise is provided by an embodiment of the present invention for a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios.

[0082] Figure 13 A comparison of safe distances in the roundabout condition of adaptive cruise driving in curves, based on a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by embodiments of the present invention.

[0083] Figure 14 A speed comparison of a vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios provided by an embodiment of the present invention in a curve adaptive cruise roundabout condition;

[0084] Figure 15 A comparison of longitudinal displacement in the roundabout condition of a vehicle lateral and longitudinal optimization control method for curve cruise scenarios provided in this embodiment of the invention.

[0085] Figure 16 A comparison of lateral displacement in a roundabout condition of adaptive cruise, provided by an embodiment of the present invention, of a vehicle lateral and longitudinal optimization control method applicable to a curve cruise scenario;

[0086] Figure 17 The vehicle trajectory diagram in the roundabout condition of the adaptive cruise, provided by an embodiment of the present invention, is an example of a vehicle lateral and longitudinal optimization control method applicable to the curve cruise scenario. Detailed Implementation

[0087] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0088] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0089] This invention provides a method for optimizing vehicle lateral and longitudinal directions in a curve cruise scenario, as one embodiment of the present invention.

[0090] 1. Safe following distance control strategy:

[0091] A fixed vehicle spacing algorithm is used for vehicle spacing design, which avoids the problem that the fixed vehicle spacing algorithm is simple but has poor adaptability. The expected safe vehicle spacing is updated in real time based on the vehicle speed, as shown in the following formula:

[0092] d des =τv h +d0 (0.1)

[0093] In the formula d des τ is the expected safe following distance, d0 is the inter-vehicle time distance, v is the minimum maintaining distance, and τ is the minimum maintaining distance. h That is the speed of this vehicle.

[0094] 2. ACC Vertical Controller Design:

[0095] 2.1 Lower-level controller:

[0096] The lower-level controller in the ACC longitudinal controller mainly establishes the vehicle's inverse longitudinal dynamics model and receives the desired acceleration 'a' from the output signal from the upper-level controller. des This is converted into corresponding throttle opening and brake master cylinder pressure, which act on the vehicle itself to complete the control task. To enable the lower-level controller to better follow the desired acceleration, this invention employs a feedforward and feedback PID controller design. The control scheme is as follows: Figure 4 As shown.

[0097] The vehicle drive calculation module is based on Newton's second law, which calculates the vehicle's drive torque by inversely using vehicle acceleration. It obtains the throttle opening by looking up the engine MAP curve. The basic principle is as follows:

[0098] The equation of motion for the car during longitudinal acceleration is shown below:

[0099] ma des =F t -F f -F w (1.2)

[0100] Where F t It is the driving force on the tires generated by the engine's driving action; F f It is the rolling resistance from the ground; F w This refers to air resistance; the effect of rotating mass is ignored. For gasoline vehicles equipped with an automatic transmission, the driving force F... t The following relationship exists between the driving torque and the driving torque:

[0101]

[0102] Where the proportionality coefficient K d This represents the amplification factor of the torque from the engine output to the wheels, and is related to the transmission's gear ratio, efficiency, and the final drive ratio. After obtaining the engine drive torque, the throttle opening value corresponding to the current engine speed and desired engine torque can be obtained by looking up the MAP table of throttle opening, engine speed, and torque, and used as a feedforward value.

[0103] The vehicle braking calculation module sets the throttle opening to 0 when the drive and braking module issues a braking command. The lower-level controller then executes the braking operation. When the braking force is not greater than the maximum braking force on the road surface, the braking force and braking pressure P are compared. des The following relationship must be satisfied:

[0104]

[0105] In the formula, F xb T represents the braking force on the wheel. bf It is the sum of the braking torques of the front wheels, T br It is the sum of the braking torques of the rear wheels, K b is the conversion coefficient between braking force and braking pressure, and r is the tire rolling radius. The desired braking pressure is calculated based on the vehicle's longitudinal dynamics equations as follows:

[0106]

[0107] The drive-brake switching logic module primarily considers that acceleration and braking cannot occur simultaneously during vehicle operation. Generally, deceleration only requires releasing the accelerator, and braking is achieved through frictional resistance, air resistance, and engine drag resistance. If this braking effect is insufficient, active braking can be initiated using the brakes. Based on the above analysis, the drive-brake switching logic is designed using the vehicle's acceleration-velocity braking drag curve. To prevent continuous logic switching during vehicle control, a buffer zone with a width of 0.2g is designed. The fitted braking drag curve is shown below. Figure 5 As shown.

[0108] 2.2 Upper-level controller design:

[0109] The upper-level controller is designed using a model predictive control algorithm, fully considering the relative speed and displacement between the vehicles, the vehicle's speed, and the vehicle's acceleration. These four quantities are selected as state variables. When building the workshop model, the desired vehicle spacing is added to the workshop relative error. That is, when the workshop relative error is 0, the actual vehicle spacing is the desired vehicle spacing, as shown below:

[0110]

[0111] Where x lead It is the displacement of the vehicle in front, x h This is the displacement of the vehicle, d des It is the expected workshop distance, v lead It's the speed of the car in front, v h Δd is the vehicle speed, Δd is the relative error between the two vehicles, and Δv is the relative speed between the two vehicles.

[0112] By combining the velocity and acceleration equations, the longitudinal control state space equations of the vehicle are constructed as follows:

[0113]

[0114] State variable x = [Δd Δv a] h v h ] T Control quantity u = a des Interference The coefficient matrix is ​​as follows:

[0115]

[0116] Discretize it using the forward Euler method. The discretized system is shown below:

[0117]

[0118] In the formula T is the sampling time. The control variable of the system shown in the above formula is the desired acceleration. If the above system is controlled, only the acceleration can be constrained, not the acceleration increment. In actual control, sudden acceleration changes may occur, which is detrimental to the ride and does not meet the comfort performance index. In actual control, it is desirable for the change in acceleration to be as gradual as possible to meet the requirements of the comfort index. Therefore, a control variable is constructed using the change in acceleration as the control variable:

[0119]

[0120] A new state-space expression is derived:

[0121]

[0122] In the formula Δu(k) is the control increment at time k. The output equation is constructed as follows:

[0123] η(k+1)=Dξ(k+1) (1.12)

[0124] In the formula Output the inter-vehicle error and the relative speed between the two vehicles.

[0125] Define N p For prediction in the time domain, N c To control the time domain, and N p ≥N c According to the iterative derivation of equation (1.12), we get

[0126]

[0127] In the formula

[0128]

[0129]

[0130]

[0131] The interference is the acceleration of the vehicle in front. Since the acceleration of the vehicle in front cannot be obtained within the prediction time domain, the interference of the vehicle in front within the entire prediction time domain is not available. constant.

[0132] The design of an ACC system must consider not only the vehicle's following performance, but also its comfort and fuel efficiency. Therefore, the performance indicators must satisfy comfort, tracking performance, and fuel economy. The following performance indicators are defined:

[0133]

[0134] Where η ref The expected output value is expressed as follows: Rewrite the above equation in matrix form:

[0135] J = [YY] ref ] T Q[YY ref ]+Δu T RΔu (1.15)

[0136] In the formula, Q and R are the output weight matrix and the control weight matrix, respectively.

[0137] Substituting equation (1.13) into equation (1.15), and letting Summarized as follows:

[0138] J = Δu T (θ T Qθ+R)Δu+2ET QθΔu+E T QE (1.16)

[0139] Let H = 2(θ) T Qθ+R), f T =2E T Qθ then has:

[0140]

[0141] Because E T QE is a constant term in the calculation and does not affect the solution of the optimal problem, so it is discarded, resulting in a standard quadratic programming problem:

[0142]

[0143] The vehicle's acceleration and its rate of change must be limited to meet comfort and fuel economy performance indicators. Research indicates that the optimal acceleration for human passengers is between -3 and 2. However, excessively drastic acceleration changes can also negatively impact comfort. The following constraints are designed, where u... min =-3, u max =2, representing the minimum and maximum acceleration values ​​respectively, with acceleration increment constraints ranging from -1.5 to 1.5, where Δu min =-1.5, Δu max =1.5, representing the minimum and maximum values ​​of the acceleration increment, respectively.

[0144]

[0145] Rewrite the control constraints as quadratic programming constraints:

[0146]

[0147] in,

[0148]

[0149] Control incremental constraints are

[0150] ΔU min ≤ΔU≤ΔU max (1.21)

[0151] in,

[0152]

[0153] 2.3 LQR-based horizontal controller design:

[0154] The LQR algorithm combines the advantages of PID and ADRC, such as rapid adjustment and strong robustness, and can also handle multi-objective optimization problems. Therefore, LQR is used to design the lateral controller of a vehicle.

[0155] When modeling vehicle dynamics, the following assumptions need to be made:

[0156] (1) Ignore vertical motion;

[0157] (2) Assume the suspension system is rigid and ignore its effects;

[0158] (3) Only consider tire lateral slip characteristics;

[0159] (4) Ignore the effect of wheel track on turning radius.

[0160] (5) Use a two-degree-of-freedom bicycle model to describe the vehicle model.

[0161] When performing vehicle control, the road curvature and heading angle calculation module can obtain a global path from a high-precision map, smooth the global path to obtain vehicle driving reference line information. The obtained reference line is a series of discrete points, x = [x1 x2 … x n ], y = [y1 y2 … y n Let x and y be the x and y coordinates of each discrete point, and then perform a difference operation on them:

[0162]

[0163] In the formula, dx is the difference with respect to x, and dy is the difference with respect to y. Using the midpoint Euler method on dx and dy, we obtain dx. final and dy final The heading angle is obtained by calculating the arctangent in the four quadrants:

[0164]

[0165] In the formula, heading is the heading angle, and arctan2 is the arctangent in the four quadrants. The road curvature is:

[0166]

[0167] In the formula, k is the road curvature, and dheading is the difference in heading angle.

[0168] The two-degree-of-freedom vehicle model used in this invention is

[0169]

[0170] In the formula, m is the total mass of the vehicle, and I is the moment of inertia. It is the rate of change of yaw rate.

[0171] like Figure 6 The figure shows the projection of the vehicle state in the Frenet coordinate system.

[0172] Point A in the diagram is the vehicle's location, and point B is the projection of the vehicle's location onto the curve in the Frenet coordinate system. It is the position vector of the vehicle in the geodetic coordinate system. It is the position vector of the point on the curve projected onto the vehicle's position, n s It is the vehicle velocity normal vector, τ s It is the tangent vector in the velocity direction, n r τ is the normal vector at the vehicle's projection point. r It is the tangent vector at the projection point. θ is the speed at the projection point, and θ is the vehicle's heading angle. r d is the heading angle at the projection point, and d is the lateral error between points A and B.

[0173] According to equation (1.25) and Figure 6 The state-space equations of the vehicle lateral controller are derived jointly from the positional relationships shown.

[0174]

[0175] in,

[0176]

[0177]

[0178] Among them, e d Represents lateral error. It is the rate of change of lateral error. It is the heading angle error. It is the rate of change of heading angle error, v x It is the vehicle's longitudinal speed, C αf It is the front wheel lateral stiffness, C αr It refers to the rear wheel lateral stiffness. δ is the disturbance quantity, and δ is the front wheel steering angle, which is the control quantity.

[0179] Using the LQR algorithm for lateral controller design, temporarily ignoring the disturbance in equation (1.26), and discretizing it using the midpoint Euler method, we can obtain the equation...

[0180]

[0181] in

[0182] Construct a performance function to solve the LQR problem.

[0183]

[0184] Q1 and R1 are the state variable weight matrix and the control variable weight matrix, respectively.

[0185] This problem is an optimization problem with equality constraints. The discrete LQR problem is solved using the Lagrange multiplier method, yielding the discrete Riccati equations.

[0186] P k-1 =Q1+A T P k (I+BR1 -1 B T P k ) -1 A (1.29)

[0187] The convergence matrix P is obtained by iterating over Riccati, therefore the control quantity u1 is

[0188] u1=-Kerr(k) (1.30)

[0189] Where -K = -(R + B) T PB) -1 B T PA.

[0190] Introducing feedforward δ in the LQR algorithm f This is because the standard LQR solution cannot make the system error zero, so feedforward is needed to compensate for the steady-state error and eliminate the influence of disturbances.

[0191] After introducing feedforward

[0192]

[0193] The feedforward expression is obtained by calculating equation (1.31).

[0194]

[0195] In the formula, K3 is the third feedback coefficient in the K matrix.

[0196] To verify the speed advantage of LQR over MPC, a comparison of actual computation times for solving transverse control problems using MPC and LQR was conducted. The experimental results are as follows: Figure 7 As shown, the calculation time is in the millisecond range, but the solution time of MPC is approximately 8 to 16 ms, while the solution time of LQR is approximately 1 to 2 ms. Under the same working conditions, the solution efficiency of online LQR is much greater than that of MPC.

[0197] Combined with appendix Figure 8-17This system displays various driving data under straight-line adaptive cruise control and roundabout conditions (road adhesion coefficient of 0.5) using this method. This LQR lateral controller not only enables lane keeping but also smooths front wheel steering angle changes, meeting comfort performance indicators—something PID and ADRC cannot achieve. Furthermore, because this invention uses a vehicle model built in the Frenet coordinate system, the lateral and longitudinal control of the vehicle are decoupled and do not interfere with each other, further improving control accuracy and reducing control difficulty. Another major advantage of LQR is offline computation. Compared to other algorithms that can only perform online computation, LQR greatly improves the utilization of computing resources and solves problems quickly. In practical engineering, the biggest problem in vehicle control is real-time performance, and LQR effectively solves this problem.

[0198] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A vehicle longitudinal and lateral optimization control method suitable for a curved cruising scenario, characterized in that, include: Acquire the status information of this vehicle and the status information of the vehicle in front, and record them as the first motion information; acquire the road position information, road curvature and heading angle information, and record them as the second motion information. Based on the first motion information and the second motion information, calculate and output throttle opening information, brake pedal information and steering wheel angle information; Based on throttle opening information, brake pedal information, and steering wheel angle information, the vehicle's lateral and longitudinal control is performed: In the longitudinal control of the vehicle, a variable vehicle spacing is selected as the safe vehicle spacing strategy, and the relative speed, relative displacement, vehicle speed and vehicle acceleration of the inter-vehicle are selected as state variables. A model predictive control algorithm is used to establish the inter-vehicle model and obtain the desired vehicle spacing and desired acceleration. Based on the workshop model, a reverse longitudinal dynamics model of the vehicle is established, a suitable acceleration is optimized, and the acceleration is tracked. The desired acceleration is converted into the corresponding throttle opening and brake master cylinder pressure, which are applied to the vehicle itself to complete the longitudinal control task. When performing lateral control of the vehicle, a global path is obtained based on the second motion information. The global path is then smoothed to obtain the vehicle driving reference line information, so that the vehicle travels along the vehicle driving reference line information.

2. The vehicle longitudinal and lateral optimization control method suitable for a curved road cruising scenario according to claim 1, characterized in that, The fixed-distance algorithm is used to design the vehicle spacing, and the expected safe distance is updated in real time based on the vehicle's speed. In the formula is the desired safety vehicle distance, is the inter-vehicle time distance, is the minimum maintained inter-vehicle distance, is the host vehicle speed.

3. The vehicle longitudinal and lateral optimization control method suitable for a curved road cruising scenario according to claim 2, characterized in that, When performing longitudinal vehicle control, the equation of motion for the car during longitudinal acceleration is as follows: wherein is the driving force of the tire due to the driving action of the engine; is the rolling resistance from the ground; is the air resistance, ignoring the effect of the rotational mass.

4. The vehicle lateral and longitudinal optimization control method applicable to curve cruise scenarios according to claim 3, characterized in that, When the braking force is not greater than the maximum road braking force, the braking force and the braking pressure satisfy the following relationship: In the formula, Indicates the braking force on the wheels. It is the sum of the braking torques of the front wheels. It is the sum of the braking torques of the rear wheels. This is the conversion coefficient between braking force and braking pressure, where r is the tire rolling radius; the desired braking pressure is calculated based on the vehicle's longitudinal dynamics equations as follows: Meanwhile, considering that acceleration and braking will not occur simultaneously during vehicle operation, the drive-brake switching logic is designed using the vehicle acceleration-speed-brake drag curve. To prevent continuous logic switching during vehicle control, a buffer zone with a width of 0.2g is designed.

Images