CN116434603B - A lateral and longitudinal synchronization safety control method for autonomous driving fleet based on SSM - Google Patents

Abstract

The present invention discloses a method for lateral and longitudinal synchronization safety control of an autonomous driving fleet based on SSM, which includes the following steps: first, for a curve scene in an intelligent network environment, after the vehicles in the autonomous driving fleet receive the operation information of the front and rear vehicles and the road status information; then, the autonomous driving vehicle safety control target is constructed based on the SSM and the fleet spacing strategy; then, the vehicle path selection is dynamically planned and controlled in real time by combining model predictive control with vehicle dynamics. The present invention can ensure the efficiency of the autonomous driving fleet in a curve scene and improve its safety.

Description

A lateral and longitudinal synchronization safety control method for autonomous driving fleet based on SSM

Technical Field

The present invention relates to a method for controlling the synchronization safety of intelligent traffic, and in particular to a method for controlling the synchronization safety of a lateral and longitudinal directions of an autonomous driving fleet based on SSM.

Background technique

Existing studies have shown that more than 90% of vehicle collisions are caused by human error. In addition, according to the U.S. Department of Transportation's Research and Innovation Technology Administration, based on autonomous vehicle technology, about 80% of vehicle collisions can be reduced each year. Since the autonomous driving fleet can accurately perceive the surrounding environment, the reaction time is negligible, and it is not affected by distracted and fatigued driving, it can coordinate multiple vehicles to drive safely and compactly, improving traffic efficiency and safety. However, when the autonomous driving fleet is disturbed externally, it is difficult to maintain the pre-set vehicle spacing, increasing the risk of collision. At present, most autonomous driving fleet control algorithms assume that the vehicle is driving on a straight road, such as adaptive cruise control and cooperative adaptive cruise control. However, on a curved road, not only the longitudinal direction of the autonomous driving vehicle should be considered, but also the lateral dynamic changes. In addition, the collision risk of the autonomous driving fleet on a curved road is higher than that on a straight road. There is little research on solving the safety control of the autonomous driving fleet with external interference on a curved road.

On curved roads, autonomous vehicles must not only have the ability to track a predetermined path, but also need to avoid collisions and reduce the risk of collisions caused by external disturbances. For the longitudinal and lateral synchronization control of autonomous driving fleets, most current algorithms use a hierarchical approach to deal with the problem: first, only the position and velocity of the autonomous vehicle are considered to plan the trajectory, and then a simple feedback controller is used to track the planned trajectory. In order to achieve the reliability and safe maneuverability of autonomous vehicles, researchers have proposed a large number of motion planning strategies to optimize the driving path or trajectory of autonomous vehicles on various roads by tracking the lower-level feedback controller. However, the above studies either incorporate risk indicators (e.g., minimum time interval, minimum deceleration) into the control objectives to reduce the risk of collision, or consider safety constraints (e.g., minimum safety interval, minimum safety interval) to ensure that there is enough distance between vehicles. For example, some scholars have proposed a rolling horizon control method for autonomous driving systems, which provides a mechanism to minimize the safety risk of autonomous vehicles under traffic disturbances. With this mechanism, alternative safety measures (SSMs) can be easily incorporated into the safety control objectives of autonomous vehicles. Among various SSMs, time to collision (TTC) and its combined indicators, such as time to collision exposure (TET), time-integrated collision time (TIT), rear-end collision risk index (RCRI), difference in spatial and stopping distance (DSS), deceleration rate to avoid collision (DRAC), and post-entrapment time (PET), have been used for autonomous vehicle safety assessment. Although SSMs have been used to assess the safety impact of autonomous vehicles, no research has directly used SSM as a control objective for trajectory optimization of autonomous driving fleets.

In summary, in the trajectory optimization of autonomous driving fleets in curved road scenarios, the key technologies of optimal lateral and longitudinal synchronous safety control that directly consider SSM need to be studied urgently.

Summary of the invention

Purpose of the invention: The purpose of the present invention is to provide a SSM-based lateral and longitudinal synchronous safety control method for an autonomous driving fleet that improves the safety of vehicles in the autonomous driving fleet in a curved scenario without sacrificing vehicle operating efficiency.

Technical solution: The method for controlling the horizontal and vertical synchronization safety of an autonomous driving fleet of the present invention comprises the following steps:

S1, obtaining road information of a curved scene;

S2, obtaining the initial operating status of all vehicles in the fleet;

S3, setting the vehicle safety distance control strategy in the convoy, and controlling the convoy with a fixed headway strategy;

S4, taking the time period from the current moment to the set value as the prediction range of the model predictive control, and designing the sampling time and the control time;

S5, designing the vehicle control objective function based on the selected SSM index and the target headway of the designed vehicle;

S6, using quadratic programming to find the optimal solution of the vehicle control objective function;

S7, according to the optimal solution obtained in step S6, the first solution of all vehicles in the fleet is used as a control input to control all vehicles;

S8, updating the running status of all vehicles;

S9: If not all vehicles have passed the curve scene, repeat steps S3 to S8 until all vehicles have passed the curve scene.

Further, in step S3, the vehicle safety distance control strategy in the convoy is set, and the steps of controlling the convoy with a fixed headway strategy are as follows:

S31, the lateral and longitudinal states of vehicle i satisfy the following equations:

Where X is the longitudinal displacement of the center point of the autonomous driving vehicle i; Y is the lateral displacement of the center point of the autonomous driving vehicle i. is its derivative; v _x represents the longitudinal velocity of the center point of the autonomous driving vehicle i, v _y represents the lateral velocity of the center point of the autonomous driving vehicle i; ψ represents the direction angle of the center point of the autonomous driving vehicle i, is its derivative; r represents the yaw rate of the center point i of the autonomous driving vehicle, is its derivative; β is the sideslip angle of the autonomous vehicle i, is its derivative; _Fxr represents the longitudinal force of the rear wheel of the autonomous driving vehicle i, _Fyf represents the lateral force of the front wheel of the autonomous driving vehicle i; _Fyr represents the lateral force of the rear wheel of the autonomous driving vehicle i; M is the vehicle mass; _Iz is the yaw inertia of the center point; _lr represents the distance from the center point of the autonomous driving vehicle to the rear wheel; _lf represents the distance from the center point of the autonomous driving vehicle to the front wheel; δ represents the steering angle of the autonomous driving vehicle i, is its derivative, δ′ represents the ideal steering angle of the autonomous vehicle i; F _x represents the longitudinal force of the autonomous vehicle i, is its derivative, and F′ _x represents the ideal longitudinal force of the autonomous driving vehicle i. Where δ′ and F′ _x are the control parameters input by the vehicle.

S32, solve the vehicle state equation, the expression of the vehicle state equation is as follows:

dx＝fdt+G·udt

in,

x＝[X Yψβr v _x v _y δF _x ] ^T

G＝[0 0 0 0 0 0 0 10 10] ^T

S33, select the fixed headway strategy, then the expression is:

τ ^* v _i,x (t)+l _f +l _r = (β _i-1,y (t)-β _i,y (t))×R

Where τ ^* represents the headway distance between vehicles; R represents the radius of curvature of the curve.

Further, in step S5, based on the selected SSM index and the designed vehicle target headway, the step of designing the objective function includes:

S51, determine the state constraint conditions of the vehicle:

-23deg≤δ≤23deg

|F′ _x |≤8600N

S52, the vehicle control target is expressed as:

min q(x,u)＝α ₁ (τ ^* v _i,x (t)+l _f +l _r -(β _i-1,y (t)-β _i,y (t))×Rs ₀ ) ² + α ₂ d ²

+α ₃ (v _i,x (t)-v _i-1,x (t)) ² +α ₄ (SSM ^* )

Wherein, α ₁ , α ₂ , α ₃ , α ₄ are weight coefficients; s ₀ is the stationary spacing; d is the curve geometry influence coefficient; SSM ^* is the control target designed according to the selected SSM.

Furthermore, when the collision time TTC with the preceding vehicle is selected as the control target, since TTC is expressed as:

but

SSM ^* _TTC = (v _i,x (t)-v _i-1,x (t)) ² -((β _i-1,y (t)-β _i,y (t))×R-(l _f +l _r )) ² .

Furthermore, when the collision avoidance deceleration DRAC is selected as the control target, DRAC is expressed as:

Where p _i (t) represents the position of vehicle i; l represents the length of the vehicle.

but

SSM ^* _DRAC = (v _i,x (t)-v _i-1,x (t)) ² - (β _i-1,y (t)-β _i,y (t))×R.

Furthermore, when the emergency deceleration collision potential index PICUD is selected as the control target, since PICUD is expressed as:

In the formula, a represents acceleration; Δt represents the sampling time interval;

Then we have:

Compared with the prior art, the present invention has the following significant effects:

The present invention uses SSM to design the objective function of vehicles in the fleet, and realizes effective and safe control of the fleet in the horizontal and vertical directions of the autonomous driving fleet traveling on a curved road in a vehicle-road cooperative environment.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the lateral and longitudinal variables of a vehicle in the present invention;

Fig. 2 is a flow chart of the present invention;

FIG3( a ) is a schematic diagram of the simulation effect of the spatial displacement change of the center point of the autonomous driving vehicle of the present invention.

FIG3( b ) is a schematic diagram of the simulation effect of the longitudinal displacement change of the center point of the autonomous driving vehicle of the present invention.

FIG3(c) is a schematic diagram of the simulation effect of the lateral displacement change of the center point of the autonomous driving vehicle of the present invention.

FIG3(d) is a schematic diagram of the simulation effect of the change of the direction angle of the center point of the autonomous driving vehicle of the present invention.

FIG3(e) is a schematic diagram of the simulation effect of the side slip angle change of the automatic driving vehicle of the present invention.

FIG3( f ) is a schematic diagram showing the simulation effect of the longitudinal velocity change of the center point of the autonomous driving vehicle of the present invention.

FIG3(g) is a schematic diagram of the simulation effect of the lateral velocity change of the center point of the autonomous driving vehicle of the present invention.

FIG3(h) is a schematic diagram of the simulation effect of the yaw rate change of the center point of the autonomous driving vehicle of the present invention.

FIG3(i) is a schematic diagram of the simulation effect of the steering angle change of the automatic driving vehicle of the present invention,

FIG3(j) is a schematic diagram of the simulation effect of the longitudinal force change of the autonomous driving vehicle of the present invention.

Detailed ways

The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

It should be noted that the terms such as "upper", "lower", "left", "right", "front", "back", etc. cited in the invention are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments in their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

In the intelligent network environment, when the autonomous driving fleet is driving on a curve, the present invention first uses the curve information and the initial state of the vehicle according to the V2V and V2I communication technologies, and then uses the selected SSM (Surrogate Safety Measures) indicator and the vehicle fixed headway strategy to formulate the objective function, and dynamically plans the path selection of the emergency vehicle through the idea of model predictive control of the rolling time domain, so as to improve the safety of the vehicles in the autonomous driving fleet in the curve scenario without losing the vehicle operation efficiency.

As shown in Figure 1, it is a schematic diagram of the lateral and longitudinal variables of the vehicle in an embodiment of the present invention; as shown in Figure 2, it is a framework diagram of the lateral and longitudinal synchronization safety control method of the autonomous driving fleet in the present invention. This embodiment is applicable to the case where the safe optimization trajectory of the autonomous driving fleet in the curve scene is dynamically planned by a server and other equipment.

The method for controlling the horizontal and vertical synchronization safety of an autonomous driving vehicle fleet of the present invention comprises the following steps:

Step 1: Obtain the road information of the curved scene, including lane style, curvature radius, number of vehicles and other environmental parameters.

Step 2: Obtain the initial operating status of all vehicles in the fleet, including vehicle position, speed and other status parameters.

Step 3: Set the vehicle safety distance control strategy in the convoy and select the fixed headway strategy for convoy control. The specific implementation steps are as follows:

Step 31, the motion state of vehicle i changes as shown in FIG1 , and its lateral and longitudinal states satisfy equations (1)-(9):

Where X is the longitudinal displacement of the center point of the autonomous driving vehicle i; Y is the lateral displacement of the center point of the autonomous driving vehicle i, is its derivative; v _x represents the longitudinal velocity of the center point of the autonomous driving vehicle i, v _y represents the lateral velocity of the center point of the autonomous driving vehicle i; ψ represents the direction angle of the center point of the autonomous driving vehicle i, is its derivative; r represents the yaw rate of the center point i of the autonomous driving vehicle, is its derivative; β is the sideslip angle of the autonomous vehicle i, is its derivative; _Fxr represents the longitudinal force of the rear wheel of the autonomous driving vehicle i, _Fyf represents the lateral force of the front wheel of the autonomous driving vehicle i; _Fyr represents the lateral force of the rear wheel of the autonomous driving vehicle i; M is the vehicle mass; _Iz is the yaw inertia of the center point; _lr represents the distance from the center point of the autonomous driving vehicle to the rear wheel; _lf represents the distance from the center point of the autonomous driving vehicle to the front wheel; δ represents the steering angle of the autonomous driving vehicle i, is its derivative, δ′ represents the ideal steering angle of the autonomous vehicle i; F _x represents the longitudinal force of the autonomous vehicle i, is its derivative, and F′ _x represents the ideal longitudinal force of the autonomous driving vehicle i. Where δ′ and F′ _x are the control parameters input by the vehicle.

Step 32, according to equations (1) to (9), the vehicle state equation can be listed as:

dx＝fdt+G·udt (10)

in,

x＝[XY ψ β rv _x v _y δ F _x ] ^T

G＝[0 0 0 0 0 0 0 10 10] ^T

Where T represents the matrix transpose.

Step 33: Under the conditions of step 32, the fixed headway strategy between the i-th vehicle and the i-1-th vehicle is:

τ ^* v _i,x (t)+l _f + l _r = (β _i-1,y (t)-β _i,y (t))×R (11)

Wherein, τ ^* represents the headway between vehicles; vi _,x (t) represents the longitudinal speed of the i-th autonomous driving vehicle; l _r represents the distance from the center point of the autonomous driving vehicle to the rear wheel; l _f represents the distance from the center point of the autonomous driving vehicle to the front wheel; β _i,y (t) represents the lateral sideslip angle of the autonomous driving vehicle i; and R represents the radius of curvature of the curve.

Step 4: Take the time period from the current moment to 10 seconds later as the prediction range of the model predictive control, and design the sampling time and control time to be 1 second.

Step 5, designing the vehicle control objective function based on the selected SSM index and the target headway of the designed vehicle;

Step 51, determine the state constraints of the vehicle:

-23deg≤δ≤23deg (12)

|F′ _x |≤8600N (13)

Step 52, the vehicle control target can be expressed as:

min q(x,u)＝α ₁ (τ ^* v _i,x (t)+l _f +l _r -(β _i-1,y (t)-β _i,y (t))×Rs ₀ ) ² +α ₂ d ² +α ₃ (v _i,x (t)-v _i-1,x (t)) ² +α ₄ (SSM ^* ) (14)

Wherein, α ₁ , α ₂ , α ₃ , α ₄ are weight coefficients; s ₀ is the stationary spacing; d is the curve geometry influence coefficient; SSM ^* is the control target designed according to the selected SSM.

For example, when TTC (Time-to-Collision) is selected as the control target, since TTC is expressed as:

but

SSM*TTC=(v _i,x (t)-v _i-1,x (t)) ² -((β _i-1,y (t)-β _i,y (t))×R-(l _f + l _r )) ² (16)

When DRAC (Deceleration Rate to Avoid a Crash) is selected as the control target, DRAC is expressed as:

Where p _i (t) represents the position of vehicle i; _vi (t) represents the speed of vehicle i; vi _-1 (t) represents the speed of vehicle i-1; l represents the length of the vehicle.

but

SSM ^* _DRAC = (v _i,x (t)-v _i-1,x (t)) ² - (β _i-1,y (t)-β _i,y (t))×R (18)

When PICUD (Potential Index for Collision with Urgent Deceleration) is selected as the control target, PICUD is expressed as:

Where a represents acceleration and Δt represents the sampling time interval.

but

Step 6: Use quadratic programming to find the optimal solution of the vehicle objective function.

Step 7: According to the optimal solution in step 6, the solutions of the first control step state of all vehicles in the fleet (ie, δ′ and F′ _x ) are used as control inputs to perform iterative control on all vehicles.

Step 8: Repeat steps 1-7 to update the operating status of all vehicles.

Step 9: If not all vehicles have passed the curve scene, that is, all vehicles have reached the end point, repeat steps 3 to 8 until all vehicles have passed the curve scene.

Simulation:

The parameters of the fleet consisting of 6 autonomous vehicles are shown in Table 1.

Table 1 Simulation data design of a fleet of 6 autonomous vehicles

Table 2 shows the results of the fleet safety indicators (minimum TTC value, maximum DRAC value, minimum PICUD value) and the fleet stability state (full-course speed change variance, full-course acceleration change variance) under three fixed headway time strategies (τ*=0.5s, τ*=1.0s, τ*=1.5s) of the intelligent connected and autonomous vehicle (CAV) and different SSM control.

Table 2 Safety performance of CAV fleet on circular road under different headway strategies

Figure 3(a) to (j) are schematic diagrams of simulation results under three different SSMs (TTC, DRAC, and PICUD).

It can be seen that the use of the present invention can improve the safety of the autonomous driving fleet in the curve scene to a certain extent. Compared with the fleet without considering SSM, its safety indicators are improved, and the speed and acceleration variances are relatively small, that is, the fleet can ensure a smoother driving process.

The above are only preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims (1)

1. An automatic driving fleet horizontal and vertical synchronous safety control method based on SSM is characterized by comprising the following steps:

s1, obtaining road information of a curve scene;

s2, acquiring initial running states of all vehicles in a motorcade;

S3, setting a vehicle safety distance control strategy in a motorcade, and controlling the motorcade by using a fixed headway strategy;

S4, taking a time period from the current moment to a set value as a prediction range of model prediction control, and designing sampling time and control time;

s5, designing a vehicle control objective function based on the selected SSM index and the designed vehicle target headstock distance;

S6, solving an optimal solution of a vehicle control objective function by using quadratic programming;

S7, controlling all vehicles by taking the first solution of all vehicles in the motorcade as control input according to the optimal solution obtained in the step S6;

S8, updating all vehicle running states;

S9, if the vehicles do not all pass through the curve scene, repeating the steps S3 to S8 until all the vehicles pass through the curve scene;

in step S3, a vehicle safety distance control strategy in the fleet is set, and the fleet control is performed by using the fixed headway strategy as follows:

s31, the lateral-longitudinal state of the vehicle i satisfies the following equation:

Wherein X is the longitudinal displacement of the central point of the automatic driving vehicle i; y is the lateral displacement of the central point of the automatic driving vehicle i; Is a derivative thereof; v _x denotes the longitudinal speed of the centre point of the autonomous vehicle i, v _y denotes the lateral speed of the centre point of the autonomous vehicle i; psi denotes the direction angle of the center point of the autonomous vehicle i, Is a derivative thereof; r denotes the yaw rate of the centre point of the autonomous vehicle i,Is a derivative thereof; beta is the slip angle of the autonomous vehicle i,Is a derivative thereof; f _xr represents the longitudinal force of the rear wheels of the autonomous vehicle i, F _yf represents the lateral force of the front wheels of the autonomous vehicle i; f _yr represents the lateral force of the rear wheels of the autonomous vehicle i; m is the mass of the vehicle; i _z is yaw inertia of the center point; l _r denotes the distance from the centre point of the autonomous vehicle to the rear wheels; l _f denotes the distance from the centre point of the autonomous vehicle to the front wheels; delta represents the steering angle of the autonomous vehicle i,For its derivative, δ' represents the ideal steering angle of the autonomous vehicle i; f _x denotes the longitudinal force of the autonomous vehicle i,For its derivative, F' _x represents the ideal longitudinal force of the autonomous vehicle i; wherein δ 'and F' _x are control parameters input by the vehicle;

s32, solving a vehicle state equation, wherein the expression of the vehicle state equation is as follows:

dx＝fdt+G·udt

Wherein,

x＝[X Y ψ β r v_x v_y δ F_x]^T

G＝[0 0 0 0 0 0 0 10 10]^T

S33, selecting a fixed headway strategy, wherein the expression is as follows:

τ^*v_i,x(t)+l_f+l_r＝(β_i-1,y(t)-β_i,y(t))×R

wherein τ ^* represents the inter-vehicle head space; v _i,x (t) represents the longitudinal speed of the i-th autonomous vehicle; β _i,y (t) represents the lateral slip angle of the autonomous vehicle i; beta _i-1,y (t) represents the lateral sideslip angle of the autonomous vehicle i-1; r represents the radius of curvature of the curve;

In step S5, the SSM indicators include a time to collision with the preceding vehicle TTC, a deceleration to collision DRAC, and an emergency deceleration collision potential indicator PICUD; based on the selected SSM index and the designed vehicle target head space, the step of designing the objective function includes:

s51, determining a state constraint condition of the vehicle:

-23deg≤δ≤23deg

|F′_x|≤8600N

s52, the vehicle control target is expressed as:

min q(x,u)＝α₁(τ^*v_i,x(t)+l_f+l_r-(β_i-1,y(t)-β_i,y(t))×R-s₀)²+α₂d²+α₃(v_i,x(t)-v_i-1,x(t))²+α₄(SSM^*)

Wherein, alpha ₁、α₂、α₃、α₄ is a weight coefficient; s ₀ is the rest distance; d is the geometric influence coefficient of the curve; v _i-1,x (t) represents the longitudinal speed of the i-1 st autonomous vehicle; SSM ^* is a control target designed according to the selected SSM;

When the time to collision TTC with the preceding vehicle is selected as the control target, since TTC is expressed as:

Then

SSM^* _TTC＝(v_i,x(t)-v_i-1,x(t))²-((β_i-1,y(t)-β_i,y(t))×R-(l_f+l_r))²;

When the collision avoidance deceleration DRAC is selected as the control target, since the DRAC is expressed as:

Where p _i (t) represents the position of the vehicle i; p _i-1 (t) represents the position of the vehicle i-1; v _i (t) denotes the speed of the vehicle i; v _i-1 (t) represents the speed of the vehicle i-1; l represents the length of the vehicle;

Then

SSM^* _DRAC＝(v_i,x(t)-v_i-1,x(t))²-(β_i-1,y(t)-β_i,y(t))×R；

When the emergency deceleration collision potential index PICUD is selected as the control target, since PICUD is expressed as:

Wherein a represents acceleration; Δt represents a sampling time interval;

Then