Vehicle lane changing decision-making and planning control method based on dynamic risk field

Abstract

The invention discloses a vehicle lane changing decision and planning control method based on a dynamic risk field, and the method comprises the steps: obtaining vehicle state parameters of an automatic driving vehicle and an obstacle vehicle under a Frenet coordinate system through coordinate transformation, and building an automatic driving vehicle lane changing decision function according to a preset driving style, the optimal target lane and the expected lane changing speed are solved, a decision instruction is sent to the planning control module, and a trajectory planning and tracking control integrated motion control strategy of the automatic driving vehicle is established by combining the decision instruction and the dynamic risk field model of lane changing of the automatic driving vehicle; the optimal front wheel steering angle increment and the optimal speed increment in the lane changing process of the automatic driving vehicle are obtained through optimization solution and serve as control signals to be sent to a braking or driving controller to conduct motion control over the vehicle, and finally lane changing operation of the automatic driving vehicle is completed. According to the invention, the adaptability of the automatic driving vehicle to a complex road environment is enhanced, and the lane changing efficiency, comfort and safety of the vehicle are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile automatic driving, and in particular to a vehicle lane changing decision and planning control method based on a dynamic risk field. Background Art

[0002] With the rapid development of urbanization and technology, the automotive industry has gradually become an important pillar industry of the national economy. The rapid growth of social economy has promoted people's pursuit of high-quality lifestyles, and the automotive industry has continued to evolve in the direction of electrification, networking, intelligence and sharing. At the same time, technological innovation in traffic safety has increasingly become the focus of attention of governments and all sectors of society. As a convenient means of travel, the number of cars has been increasing year by year. In 2023, the number of private cars has reached 293,568,900, an increase of 5.6% compared with 2022. This rapid growth has led to a sharp increase in the pressure on urban traffic, and traffic flow has gradually become saturated. However, due to differences in driving experience and driving habits, road traffic accidents occur frequently, posing a serious threat to property and life safety, and also having an adverse impact on travel efficiency and car demand. According to the data from the China Statistical Yearbook, a total of 256,409 traffic accidents occurred nationwide in 2022, causing 60,676 deaths. Although passive safety technologies (such as airbags, seat belts, head and neck protection devices, etc.) have played a certain role in reducing the harm of traffic accidents, it is difficult to effectively prevent improper driving behaviors such as distracted driving and fatigue driving by passive safety technologies alone. In this context, autonomous driving technology has gradually shown its importance. By collecting road environment information and sharing driving data, the autonomous driving system can help vehicles complete operations such as following vehicles and changing lanes, which can not only reduce the occurrence of traffic accidents, but also reduce traffic flow pressure and improve travel convenience, safety and efficiency. Therefore, promoting the development of autonomous driving technology is not only crucial to improving vehicle driving safety, but also has far-reaching significance for alleviating the growing traffic pressure and improving people's travel experience.

[0003] Autonomous vehicles have unreasonable lane-changing behaviors in their lane-changing behavior decisions, which leads to risks such as slippage, collision, and rollover during the lane-changing process. In addition, the current behavior decision-making system mainly prioritizes safety, which reduces the risk of accidents but also reduces driving efficiency to a certain extent. In terms of planning and control, autonomous vehicles lack efficient and smooth path planning and precise tracking control, resulting in poor comfort during driving. In addition, during the lane-changing process, some emergencies can make it difficult for the established planned path to adapt to the rapid changes in the environment, which may cause traffic accidents. Summary of the invention

[0004] The purpose of the present invention is to provide a vehicle lane changing decision and planning control method based on a dynamic risk field, which solves the adaptability of autonomous driving vehicles to emergencies and complex traffic, while improving the smoothness and comfort of the lane changing process.

[0005] The technical solution to achieve the purpose of the present invention is: a vehicle lane change decision and planning control method based on a dynamic risk field, the steps are as follows:

[0006] Step S1: Use sensors to obtain the position information and motion parameters of the autonomous driving vehicle and the obstacle vehicle in the Cartesian coordinate system, calculate the vehicle state information of all vehicles in the Frenet coordinate system through coordinate transformation, and proceed to step S2.

[0007] Step S2: Establish a lane-changing decision cost function for the autonomous driving vehicle based on the vehicle status information of each vehicle in the Frenet coordinate system. The cost function includes three performance indicators: driving safety cost, riding comfort cost, and travel efficiency cost. The weight coefficients of the three performance indicators are adjusted according to the different driving styles of the autonomous driving vehicle to obtain the optimal target lane and expected lane-changing speed for the autonomous driving vehicle, and then proceed to step S3.

[0008] Step S3: Analyze the road structure and the vehicle status information of the obstacle vehicle, establish a dynamic risk field model for lane changing of the autonomous driving vehicle, and proceed to step S4.

[0009] Step S4: Based on the dynamic risk field model of the autonomous driving vehicle's lane change, combined with the autonomous driving vehicle's optimal target lane and expected lane change speed, establish an integrated motion control strategy for autonomous driving vehicle trajectory planning and tracking control, optimize and solve to obtain the optimal front wheel steering angle and acceleration during the vehicle's lane change process, and proceed to step S5.

[0010] Step S5: Send the optimal front wheel steering angle increment and the optimal speed increment as control signals to the brake or drive controller to control the vehicle's motion, thereby completing the lane changing operation of the autonomous driving vehicle.

[0011] Compared with the prior art, the present invention has the following significant advantages:

[0012] (1) Enhanced adaptability and more flexible lane changing: The present invention fully considers the different driving styles of autonomous vehicles in lane changing decisions, and can make behavioral decisions that are more in line with human driving habits based on specific driving scenarios, thereby improving the system's response speed and decision-making accuracy.

[0013] (2) Higher comfort and safety: The dynamic risk field model continuously optimizes the lane-changing trajectory, effectively avoiding sudden turns or sudden braking during the lane-changing process, and significantly improving the stability and smoothness of the vehicle's lane-changing. In addition, the use of a motion control strategy that integrates trajectory planning and tracking control further improves the accuracy of vehicle control and reduces the risk of trajectory deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a flow chart of the vehicle lane changing and planning control method based on dynamic risk field of the present invention.

[0015] Figure 2 It is a flow chart of lane-changing decision-making for autonomous vehicles.

[0016] Figure 3 It is a flow chart of the motion control strategy for an autonomous vehicle.

[0017] Figure 4 This is a schematic diagram of the obstacle vehicle risk field. DETAILED DESCRIPTION

[0018] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0019] Combination Figure 1 to Figure 3 The vehicle lane changing and planning control method based on dynamic risk field described in the present invention comprises the following steps:

[0020] Step S1: Use sensors to obtain the position information and motion parameters of the autonomous driving vehicle and the obstacle vehicle in the Cartesian coordinate system, and calculate the vehicle state information corresponding to all vehicles in the Frenet coordinate system through coordinate transformation, as follows:

[0021]

[0022] Among them, the coordinates of the autonomous driving vehicle P are (x, y), P i is the coordinate corresponding to the i-th point in the pre-acquired lane centerline point sequence, expressed as (x i ,y i ), M is the total number of point sequence coordinates, and the pseudo-projection point P of the autonomous driving vehicle in the Frenet coordinate system m The coordinates of (x m ,y m ), and P and P m Meet|PP m |=min{PP 1 ,PP 2 ,…PP M}, the actual projection point P of the autonomous driving vehicle r The coordinates of (x r ,y r), s is the distance traveled by the vehicle along the center line of the reference lane, d is the normal offset of the vehicle from the center line of the reference lane, is the vehicle pseudo-projection point P m The heading angle and θ at m Vehicle pseudo-projection point P m The tangent vector at .

[0023] Step S2: A lane-changing decision cost function for an autonomous vehicle is established based on the vehicle state information of each vehicle in the Frenet coordinate system. The cost function includes three performance indicators: driving safety cost, riding comfort cost, and travel efficiency cost. The weight coefficients of the three performance indicators are adjusted according to the different driving styles of the autonomous vehicle, so as to obtain the optimal target lane and lane-changing expected speed for the autonomous vehicle, as follows:

[0024] The expressions of the three performance indicators of the lane change decision cost function of the autonomous driving vehicle are:

[0025] Driving safety cost of autonomous vehicles C ds , including vertical safety costs C ds_log and horizontal security cost C ds_lat :

[0026] C ds =(β 2 -1) 2 C ds_log +β 2 C ds_lat

[0027]

[0028] Among them, α is the lane number of the lane where the autonomous driving vehicle is located, α={1,2,3} corresponds to {left lane, middle lane, right lane}, β represents the lane change decision behavior of the autonomous driving vehicle, β={-1,0,1} corresponds to {left lane change, lane keeping, right lane change}, ε is the number of the obstacle vehicle in the target lane, ε={1,2} corresponds to {obstacle vehicle in front of the target lane, obstacle vehicle behind the target lane}, and They correspond to the driving speed of the autonomous driving vehicle, the driving speed of the obstacle vehicle in the target lane, and the driving speed of the obstacle vehicle in the adjacent lane of the target lane. and They correspond to the driving distance of the autonomous driving vehicle, the driving distance of the obstacle vehicle in the target lane, and the driving distance of the obstacle vehicle in the adjacent lane of the target lane, k v_log is the longitudinal velocity weight coefficient, k v_lat is the lateral velocity weight coefficient, k s_log is the longitudinal distance weight coefficient, k s_lat is the horizontal distance weight coefficient, lα+β is the vehicle length, Δv 1 , Δs 1 , σ 1 , Δv 2 , Δs 2 , σ 2 All are intermediate variables.

[0029] The cost of ride comfort of autonomous vehicles C rc :

[0030]

[0031] in, is the longitudinal acceleration of the autonomous vehicle, is the lateral acceleration of the autonomous vehicle, is the longitudinal acceleration weight coefficient, is the weight coefficient of lateral acceleration, and δ is the front wheel turning angle of the autonomous vehicle.

[0032] The travel efficiency cost of autonomous vehicles C te :

[0033]

[0034] in, is the speed of the obstacle vehicle in front of the target lane, v max is the maximum speed of the target lane under the conditions permitted by traffic regulations, is an intermediate variable.

[0035] The cost function C of lane-changing decision of autonomous driving vehicle sum The expression is:

[0036] C sum =ω ds C ds +ω rc C rc +ω te C te

[0037] Among them, ω ds is the driving safety cost weight coefficient, ω rc is the riding comfort cost weight coefficient, ω te is the travel efficiency cost weight coefficient.

[0038] The weight coefficients of the three performance indicators of autonomous driving vehicles with different driving styles are shown in Table 1:

[0039] Table 1 Weight coefficients of different driving styles

[0040]

[0041] The optimal target lane and expected lane change speed are obtained through the lane change decision cost function of the autonomous driving vehicle. The decision process is expressed as:

[0042]

[0043] Among them, β * is the optimal target lane for the autonomous vehicle, v des is the optimal lane-changing speed of the autonomous vehicle, s des The best lane-changing position for autonomous vehicles. Select the interval for the speed, Select an interval for the lane change position.

[0044] Step S3: Analyze the road structure and the vehicle status information of the obstacle vehicle, and establish a dynamic risk field model for lane changing of the autonomous driving vehicle, as follows:

[0045] The dynamic risk field model includes the road risk field and the dynamic risk field of the obstacle vehicle;

[0046] The lane lines on the road are divided into two types: crossable and non-crossable. At the same time, it is also necessary to avoid continuous lane crossing when vehicles change lanes. The road risk field expression is:

[0047]

[0048] Among them, E L is the superimposed risk field of all lane lines acting on the autonomous driving vehicle, N is the total number of lane lines, μ j is the gain coefficient of the jth lane line acting on the vehicle. The farther the autonomous driving vehicle is from the lane line, the larger its value. In addition, the gain coefficient of the lane line that cannot be crossed is greater than the gain coefficient of the lane line that can be crossed. j is the risk field coefficient of the jth lane line, d line,j is the distance from the vehicle to the jth lane line, σ r is the lane line risk field convergence coefficient.

[0049] The dynamic risk field expression of the obstacle vehicle is:

[0050]

[0051] Among them, E V is the superimposed risk field of all obstacle vehicles on the road, C is the total number of obstacle vehicles, λ w is the risk field weight coefficient of the w-th obstacle vehicle, v w is the longitudinal velocity of the wth obstacle vehicle, a w is the longitudinal acceleration of the wth obstacle vehicle, (s w ,dw ) is the position coordinate of the w-th obstacle vehicle in the Frenet coordinate system, τ 1 is the longitudinal weight coefficient of the obstacle vehicle, τ 2 is the lateral weight coefficient of the obstacle vehicle, a 1 is the unknown coefficient related to acceleration, a 2 is the unknown coefficient related to speed.

[0052] Step S: Based on the dynamic risk field model of the autonomous driving vehicle lane change, combined with the optimal target lane and the expected lane change speed of the autonomous driving vehicle, a motion control strategy integrating vehicle trajectory planning and tracking control is established, and the optimal front wheel steering angle and acceleration during the vehicle lane change process are obtained by optimization, as follows:

[0053] The prediction model after vehicle discretization is established, and its formula is as follows:

[0054] x(k+1)=f(x(k),Δu(k))

[0055] Among them, x(k) is the control quantity of the kth step, expressed as is the yaw angle of the autonomous vehicle, v x is the longitudinal velocity of the autonomous vehicle, v y is the lateral velocity of the autonomous vehicle, γ is the yaw rate of the autonomous vehicle, and Δu(k) is the control increment of the kth step, expressed as Δu(k)=[Δv,Δδ] T , Δv is the speed increment of the autonomous vehicle, Δδ is the front wheel turning angle increment of the autonomous vehicle, and T represents transposition.

[0056] The objective function of vehicle lane changing is established, and its expression is as follows:

[0057]

[0058]

[0059] Among them, J 1 , J 2 , J 3 and J 4 They are lane line risk field objective function, obstacle vehicle dynamic risk field objective function, expected speed difference objective function and control increment objective function, J is the sum of the four objective functions, E L (k+p|k) is the risk field increment of the lane line superposition at the k+pth step, E V (k+p|k) is the incremental risk field of the obstacle vehicle superimposed on the k+pth step, N P is the prediction step length, N C To control the step size, p = 1, 2, 3, ..., N p, q=1,2,3,…,N C , ΔV(k+p|k) is the speed increment of the automatic driving vehicle in the k+pth step, Δu(k+q|k) is the increment of the control quantity in the k+qth step, ξ 1 , 2 , 3 , 4 J 1 , J 2 , J 3 , J 4 The weight coefficient, Q 1 , Q 2 , Q 3 , R are J 1 , J 2 , J 3 , J 4 The weight matrix of .

[0060] While satisfying vehicle dynamics, the stability and comfort of vehicle driving must also be considered. Therefore, constraints are imposed on the vehicle's control amount and control increment:

[0061] u(k+p|k)=u(k+p-1|k)+Δu(k+p|k)

[0062] u min ≤u(k+p|k)≤u max

[0063] Δu min ≤Δu(k+p|k)≤Δu max

[0064] Among them, u(k+p|k) is the control amount of the k+pth step predicted by the kth step, Δu(k+p|k) is the control increment of the k+pth step predicted by the kth step, [u min ,u max ] is the range of the control quantity, [Δu min , Δu max ] is the range of the control increment.

[0065] Through optimization, the optimal control increment Δu during the vehicle lane change process can be obtained. * , expressed as:

[0066] Δu * =[Δv * ,Δδ * ] T

[0067] Where Δv * is the optimal speed increment, Δδ * Optimal front wheel steering angle increment.

[0068] Step S5: Increase the optimal front wheel steering angle by Δδ * and the optimal speed increment Δv * The control signal is sent to the brake or drive controller to control the vehicle's motion, thereby completing the lane changing operation of the autonomous driving vehicle.

[0069] Example 1

[0070] The vehicle lane-changing decision and planning control method based on a dynamic risk field described in the present invention comprises the following steps:

[0071] Step S1: Use sensors to obtain the position information and motion parameters of the autonomous driving vehicle and the obstacle vehicle in the Cartesian coordinate system, and calculate the vehicle state information of all vehicles in the Frenet coordinate system through coordinate transformation. Figure 4 The vehicle status information of each vehicle in the Frenet coordinate system, including position, speed and acceleration information, is shown in the figure. The autonomous driving vehicle is located in the middle lane with a speed of v ev is 15m / s, acceleration a ev 0m / s 2 , the left front car is in the left lane, with a speed of v lf is 15m / s, acceleration a lf 2m / s 2 , the left rear vehicle is in the left lane at a speed v lr is 12m / s, acceleration a lr 0m / s 2 , the front car is in the middle lane, speed v f is 15m / s, acceleration a f -2m / s 2 , the right front car is in the right lane, speed v rf is 15m / s, acceleration a rf 0m / s 2 , go to step S2.

[0072] Step S2: According to Figure 4 The motion information of all vehicles in the , and the lane change decision cost function C of the autonomous driving vehicle are established sum , the cost function includes the driving safety cost C ds , Ride comfort cost C rc and travel efficiency cost C te The weight coefficients of the three performance indicators are adjusted according to the different driving styles of the autonomous driving vehicle to obtain the optimal target lane β of the autonomous driving vehicle. * and the desired lane change speed v des , go to step S3.

[0073] Step S3: Figure 4The left front vehicle, the left rear vehicle, the front vehicle and the right front vehicle are all obstacle vehicles. The road structure and vehicle status information of the obstacle vehicles are analyzed, and a dynamic risk field model for lane changing of the autonomous driving vehicle is established, and the process proceeds to step S4.

[0074] Step S4: Based on the dynamic risk field model of the autonomous driving vehicle lane change, combined with the optimal target lane and the expected lane change speed of the autonomous driving vehicle, a motion control strategy integrating trajectory planning and tracking control of the autonomous driving vehicle is established, and the optimal front wheel steering angle increment Δδ is obtained in the process of the vehicle lane change by optimization. * and the optimal speed increment Δv * , go to step S5.

[0075] Step S5: Increase the optimal front wheel steering angle by Δδ * and the optimal speed increment Δv * The control signal is sent to the brake or drive controller to control the vehicle's motion, thereby completing the lane changing operation of the autonomous driving vehicle.

Claims

1. A vehicle lane-changing decision and planning control method based on a dynamic risk field, characterized in that: Here are the steps: Step S1: Use sensors to obtain the position information and motion parameters of the autonomous driving vehicle and the obstacle vehicle in the Cartesian coordinate system, calculate the vehicle state information of all vehicles in the Frenet coordinate system through coordinate transformation, and then proceed to step S2; Step S2: Establish a lane-changing decision cost function for the autonomous driving vehicle based on the vehicle state information of each vehicle in the Frenet coordinate system. The cost function includes three performance indicators: driving safety cost, riding comfort cost, and travel efficiency cost. The weight coefficients of the three performance indicators are adjusted according to the different driving styles of the autonomous driving vehicle, so as to obtain the optimal target lane and lane-changing expected speed for the autonomous driving vehicle, and then proceed to step S3. Step S3: Analyze the road structure and the vehicle status information of the obstacle vehicle, establish a dynamic risk field model for lane changing of the autonomous driving vehicle, and proceed to step S4; Step S4: Based on the dynamic risk field model of the autonomous driving vehicle lane change, combined with the optimal target lane and the expected lane change speed of the autonomous driving vehicle, a motion control strategy integrating trajectory planning and tracking control of the autonomous driving vehicle is established, and the optimal front wheel steering angle increment and the optimal speed increment during the lane change process of the vehicle are obtained by optimization and solution, and then the process goes to step S5; Step S5: Send the optimal front wheel steering angle increment and the optimal speed increment as control signals to the brake or drive controller to control the vehicle's motion, thereby completing the lane changing operation of the autonomous driving vehicle.

2. The vehicle lane-changing decision and planning control method based on dynamic risk field according to claim 1 is characterized in that: In step S1, the position information and motion parameters of the autonomous driving vehicle and the obstacle vehicle in the Cartesian coordinate system are obtained by using sensors, and the vehicle state information corresponding to all vehicles in the Frenet coordinate system is calculated through coordinate transformation, as follows: The coordinates of the autonomous driving vehicle P are (x, y), P i is the coordinate corresponding to the i-th point in the pre-acquired lane centerline point sequence, expressed as (x i ,y i ), M is the total number of point sequence coordinates, and the pseudo-projection point P of the autonomous driving vehicle in the Frenet coordinate system m The coordinates of (x m ,y m ), and P and P m Meet|PP m |=min{PP1,PP2,…PP M }, the actual projection point P of the autonomous driving vehicle r The coordinates of (x r ,y r ), s is the distance traveled by the autonomous driving vehicle along the center line of the reference lane, d is the normal offset of the autonomous driving vehicle from the center line of the reference lane, is the pseudo projection point P of the autonomous driving vehicle m The heading angle at m is the pseudo projection point P of the autonomous driving vehicle m The tangent vector at .

3. The vehicle lane-changing decision and planning control method based on dynamic risk field according to claim 2 is characterized in that: In step S2, a lane-changing decision cost function for the autonomous driving vehicle is established based on the vehicle state information of each vehicle in the Frenet coordinate system. The cost function includes three performance indicators: driving safety cost, riding comfort cost, and travel efficiency cost. The specifics are as follows: Driving safety cost of autonomous vehicles C ds , including vertical safety costs C ds_log and horizontal security cost C ds_lat : C ds =(β 2 -1) 2 C ds_log +b 2 C ds_lat Among them, α is the lane number of the lane where the autonomous driving vehicle is located, α={1,2,3} corresponds to {left lane, middle lane, right lane}, β represents the lane change decision behavior of the autonomous driving vehicle, β={-1,0,1} corresponds to {left lane change, lane keeping, right lane change}, ε is the number of the obstacle vehicle in the target lane, ε={1,2} corresponds to {obstacle vehicle in front of the target lane, obstacle vehicle behind the target lane}, and They correspond to the driving speed of the autonomous driving vehicle, the driving speed of the obstacle vehicle in the target lane, and the driving speed of the obstacle vehicle in the adjacent lane of the target lane. and They correspond to the driving distance of the autonomous driving vehicle, the driving distance of the obstacle vehicle in the target lane, and the driving distance of the obstacle vehicle in the adjacent lane of the target lane, κ v_log is the longitudinal velocity weight coefficient, κ v_lat is the lateral velocity weight coefficient, κ s_log is the longitudinal distance weight coefficient, κ s_lat is the horizontal distance weight coefficient, l α+β is the vehicle length, Δv1, Δs1, σ1, Δv2, Δs2, σ2 are all intermediate variables; The cost of ride comfort of autonomous vehicles C rc : in, is the longitudinal acceleration of the autonomous vehicle, is the lateral acceleration of the autonomous vehicle, is the longitudinal acceleration weight coefficient, is the weight coefficient of lateral acceleration, δ is the front wheel turning angle of the autonomous vehicle; The travel efficiency cost of autonomous vehicles C te : in, is the speed of the obstacle vehicle in front of the target lane, v max is the maximum speed of the target lane under the conditions permitted by traffic regulations, is an intermediate variable; The cost function C of lane-changing decision of autonomous driving vehicle sum The expression is: C sum =ω ds C ds +oh rc C rc +oh te C te Among them, ω ds is the driving safety cost weight coefficient, ω rc is the riding comfort cost weight coefficient, ω te is the travel efficiency cost weight coefficient.

4. The vehicle lane-changing decision and planning control method based on dynamic risk field according to claim 3 is characterized in that: In step S2, the weight coefficients of the three performance indicators are adjusted according to the different driving styles of the autonomous driving vehicle, so as to obtain the optimal target lane and lane change expected speed of the autonomous driving vehicle, as follows: The weight coefficients of the three performance indicators of autonomous driving vehicles with different driving styles are shown in Table 1: Table 1 Weight coefficients of different driving styles The optimal target lane and expected lane change speed are obtained through the lane change decision cost function of the autonomous driving vehicle. The decision process is expressed as: Among them, β * is the optimal target lane for the autonomous vehicle, v des is the optimal lane-changing speed of the autonomous vehicle, s des The best lane-changing position for autonomous vehicles. Select the interval for the speed, Select the zone for the lane change position.

5. The vehicle lane-changing decision and planning control method based on dynamic risk field according to claim 4 is characterized in that: In step S3, the road structure and the vehicle status information of the obstacle vehicle are analyzed to establish a dynamic risk field model for lane changing of the autonomous driving vehicle, as follows: The dynamic risk field model includes the road risk field and the dynamic risk field of the obstacle vehicle; Lane lines on the road are divided into two types: crossable and non-crossable. When an autonomous vehicle changes lanes, it must not only avoid leaving the road area, but also avoid continuous lane crossing. The road risk field expression is: Among them, E L is the superimposed risk field of all lane lines acting on the autonomous driving vehicle, N is the total number of lane lines, μ j is the gain coefficient of the jth lane line acting on the vehicle. The farther the autonomous driving vehicle is from the lane line, the larger its value. In addition, the gain coefficient of the lane line that cannot be crossed is greater than the gain coefficient of the lane line that can be crossed. j is the risk field coefficient of the jth lane line, d line,j is the distance from the autonomous driving vehicle to the jth lane line, σ r is the lane line risk field convergence coefficient; The dynamic risk field expression of the obstacle vehicle is: Among them, E V is the superimposed risk field of all obstacle vehicles on the road, C is the total number of obstacle vehicles, λ w is the risk field weight coefficient of the w-th obstacle vehicle, v w is the longitudinal velocity of the wth obstacle vehicle, a w is the longitudinal acceleration of the wth obstacle vehicle, (s w ,d w ) is the position coordinate of the w-th obstacle vehicle in the Frenet coordinate system, τ1 is the longitudinal weight coefficient of the obstacle vehicle, τ2 is the lateral weight coefficient of the obstacle vehicle, a1 is the undetermined coefficient related to acceleration, and a2 is the undetermined coefficient related to speed.

6. The vehicle lane-changing decision and planning control method based on dynamic risk field according to claim 5 is characterized in that: In step S4, based on the dynamic risk field model of the autonomous driving vehicle lane change, combined with the optimal target lane and the expected lane change speed of the autonomous driving vehicle, a motion control strategy integrating vehicle trajectory planning and tracking control is established, and the optimal front wheel steering angle increment and the optimal speed increment during the vehicle lane change process are obtained by optimization, as follows: The prediction model after vehicle discretization is established, and its formula is as follows: x(k+1)=f(x(k),Δu(k)) Among them, x(k) is the control quantity of the kth step, is the yaw angle of the autonomous vehicle, v x is the longitudinal velocity of the autonomous vehicle, v y is the lateral velocity of the autonomous vehicle, γ is the yaw rate of the autonomous vehicle, δ is the front wheel turning angle of the autonomous vehicle, Δu(k) is the control increment of the kth step, Δu(k)=[Δv,Δδ] T , Δv is the speed increment of the autonomous vehicle, Δδ is the front wheel turning angle increment of the autonomous vehicle; T represents transposition; The objective function of vehicle lane changing is established, and its expression is as follows: Among them, J1, J2, J3 and J4 are respectively the lane line risk field objective function, the obstacle vehicle dynamic risk field objective function, the expected speed difference objective function and the control increment objective function, J is the sum of the four objective functions, E L (k+p|k) is the superposition risk field value of the k+pth lane line, E V (k+p|k) is the superimposed risk field value of the obstacle vehicle in the k+pth step, N P is the prediction step length, N C To control the step size, p = 1, 2, 3, ..., N p , q=1,2,3,…,N C , ΔV(k+p|k) is the speed increment of the automatic driving vehicle in the k+pth step, Δu(k+q|k is the increment of the control amount in the k+qth step, ξ1, ξ2, ξ3, ξ4 are the weight coefficients of J1, J2, J3, J4 respectively, Q1, Q2, Q3, R are the weight matrices of J1, J2, J3, J4 respectively; While satisfying vehicle dynamics, the stability and comfort of vehicle driving must also be considered. Therefore, constraints are imposed on the vehicle's control amount and control increment: u(k+p|k)=u(k+p-1|k)+Δu(k+p|k) u min ≤u(k+p|k)≤u max Δu min ≤Δu(k+p|k)≤Δu max Among them, u(k+p|k) is the control amount of the k+pth step predicted by the kth step, Δu(k+p|k) is the control increment of the k+pth step predicted by the kth step, [u min ,u max ] is the range of the control quantity, [Δu min , Δu max ] is the range of the control increment; Through optimization, the optimal control increment Δu is obtained during the vehicle lane change process. * , expressed as: Thu * =[Δv * ,Dd * ] T Where Δv * is the optimal speed increment, Δδ * Optimal front wheel steering angle increment.

Images