A risk assessment method and personalized decision-making method for intelligent connected vehicles

Abstract

This invention discloses a risk assessment method and personalized decision-making method for intelligent connected vehicles, comprising: calculating the driving risks posed by each surrounding vehicle to the main vehicle based on the main vehicle data and surrounding vehicle data using a driving safety field model; calculating the following risk and lane-changing risk borne by the main vehicle based on the driving risks posed by each surrounding vehicle; selecting a driving style and comparing the driving risks borne by the main vehicle with the driving risk thresholds corresponding to the driving style to determine the driving state of the main vehicle; this invention provides accurate driving risk assessment, comprehensive risk consideration dimensions, and can meet the comfort needs of different types of drivers or passengers.

Description

Technical Field

[0001] This invention belongs to the field of automotive driving decision-making technology, specifically relating to a risk assessment method and a personalized decision-making method for intelligent connected vehicles. Background Technology

[0002] Intelligent connected vehicles based on next-generation communication technologies can effectively address the technical bottlenecks faced by single-vehicle autonomous driving and are currently a hot research topic in the fields of autonomous driving and intelligent transportation. Intelligent decision-making, as a crucial component of intelligent connected vehicles, has a vital impact on vehicle safety and road traffic safety. How to effectively assess the driving risks posed by surrounding traffic participants in complex driving scenarios and then make reasonable and effective decisions is one of the key challenges currently facing intelligent connected vehicles.

[0003] Existing research on vehicle intelligent decision-making based on driving risk assessment mostly considers only the speed and distance factors between the vehicle and surrounding vehicles when assessing driving risks, with evaluation indicators such as collision time and headway. However, this decision-making method has certain problems: First, it considers too few influencing factors when conducting driving risk assessment, resulting in inaccurate risk assessment results; second, it only considers single-dimensional driving risks in the longitudinal or lateral direction, resulting in a limited risk assessment dimension; third, different types of drivers or passengers have different psychological thresholds for driving risks, and making vehicle intelligent decisions based on the same driving risk standard cannot meet the needs of drivers or passengers for driving comfort.

[0004] This shows that existing vehicle intelligent decision-making methods and related technologies based on risk assessment are insufficient to meet the development needs of intelligent connected vehicles. Summary of the Invention

[0005] In view of this, the present invention provides a method for risk assessment and personalized decision-making for intelligent connected vehicles, which makes driving risk assessment accurate, considers risk dimensions comprehensively, and can meet the comfort needs of different types of drivers or passengers.

[0006] This invention is achieved through the following technical solution:

[0007] A method for risk assessment of intelligent connected vehicles includes: using data acquired by sensors and vehicle-road cooperative perception facilities equipped in intelligent connected vehicles under intelligent connected environment for driving risk assessment, the data including main vehicle data and surrounding vehicle data, and using a driving safety field model to calculate the driving risk posed by each surrounding vehicle to the main vehicle.

[0008] The main vehicle data includes the main vehicle's physical mass, lateral coordinates, longitudinal coordinates, speed, and acceleration; the surrounding vehicle data includes the physical mass, lateral coordinates, longitudinal coordinates, length, width, speed, and acceleration of each surrounding vehicle.

[0009] Furthermore, based on the driving risks posed by surrounding vehicles to the main vehicle, the following risk and lane-changing risk borne by the main vehicle are calculated separately using weighted averages.

[0010] Furthermore, the method of calculating the driving risk posed by each surrounding vehicle to the main vehicle using the driving safety field model is as follows:

[0011] Step S11: Calculate the risk mass of the surrounding vehicle A based on its physical mass and speed.

[0012] Step S12: On the one hand, the potential energy risk field strength of the surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, and width of the surrounding vehicle A; on the other hand, the kinetic energy risk field strength of the surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, width, and speed of the surrounding vehicle A.

[0013] Step S13: Combine the potential energy risk field strength and kinetic energy risk field strength of the surrounding vehicle A obtained in step S12 to obtain the risk field strength of the surrounding vehicle A.

[0014] Step S14: Based on the risk field strength of the surrounding vehicle A obtained in step S13, calculate the driving risk posed by the surrounding vehicle A to the main vehicle by combining the physical mass, speed and acceleration of the main vehicle and the speed of the surrounding vehicle A.

[0015] Step S15: Calculate the driving risks posed by each surrounding vehicle to the main vehicle based on steps S11-S14.

[0016] Furthermore, the calculation formula for the potential energy risk field strength of the surrounding vehicle A, obtained in step S12 by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, and width of the surrounding vehicle A, is as follows:

[0017]

[0018] Among them, E sv_sta M represents the potential energy risk field strength of surrounding vehicle A, where A is a field strength coefficient greater than 0. sv For the risk quality of surrounding vehicle A obtained in step S11, x ev The longitudinal coordinate of the main vehicle, x sv Let w be the longitudinal coordinate of the surrounding vehicle A. x L is the weighting factor for the length of surrounding vehicle A. sv Let y be the length of the surrounding vehicle A. ev The longitudinal coordinate of the main vehicle, y svLet w be the lateral coordinate of the surrounding vehicle A. y W is the weighting factor for the width of surrounding vehicle A. sv Let d be the width of the surrounding vehicle A, and d be the distance vector between the main vehicle and the surrounding vehicle A, where d = (x ev -x sv ,y ev -y sv ).

[0019] Furthermore, the calculation formula for the kinetic energy risk field strength of the surrounding vehicle A, obtained in step S12 by combining the lateral and longitudinal coordinates of the main vehicle and the lateral and longitudinal coordinates, length, width, and speed of the surrounding vehicle A, is as follows:

[0020]

[0021] Among them, E sv_va Let M be the kinetic energy risk field strength of the surrounding vehicle A, where A is a field strength coefficient greater than 0. sv For the risk quality of surrounding vehicle A obtained in step S11, x ev The longitudinal coordinate of the main vehicle, x sv Let v be the longitudinal coordinate of the surrounding vehicle A, α be the weighting coefficient related to the speed of the surrounding vehicle A, and v be the longitudinal coordinate of the surrounding vehicle A. sv Let w be the speed of the surrounding vehicle A. x L is the weighting factor for the length of surrounding vehicle A. sv Let y be the length of the surrounding vehicle A. ev The longitudinal coordinate of the main vehicle, y sv Let w be the lateral coordinate of the surrounding vehicle A. y W is the weighting factor for the width of surrounding vehicle A. sv Let d be the width of the surrounding vehicle A; θ be the angle between the velocity direction of the surrounding vehicle A and d, with clockwise direction being positive; γ be a weighting coefficient related to the angle θ; β be a weighting coefficient related to the acceleration of the surrounding vehicle A; and a be the width of the surrounding vehicle A. sv Let be the acceleration of a surrounding vehicle A, and d be the distance vector between the main vehicle and the surrounding vehicle A, d = (x ev -x sv ,y ev -y sv ).

[0022] Furthermore, the calculation formula for step S14 is as follows:

[0023]

[0024] Where, r ev E represents the driving risk posed by surrounding vehicle A to the main vehicle. sv For the risk field strength of the surrounding vehicle A obtained in step S13, m evThe physical mass of the main vehicle; w v The relative speed of the main vehicle and surrounding vehicle A |v ev -v sv | weighting coefficient; v ev Main vehicle speed; v sv w represents the speed of surrounding vehicles. a The weighting coefficient for the main vehicle acceleration; a ev The acceleration of the main vehicle.

[0025] Furthermore, the risk of being followed by the main vehicle is calculated by weighting the driving risks posed to the main vehicle by the surrounding vehicles in front, to the left front, and to the right front.

[0026] The lane-changing risk to the main vehicle is calculated by weighting the driving risks posed by the vehicles surrounding the main vehicle from the left front, left rear, right front, and right rear.

[0027] A personalized decision-making method for intelligent connected vehicles, based on the aforementioned intelligent connected vehicle risk assessment method, includes: selecting a driving style, comparing the driving risk experienced by the main vehicle with the driving risk threshold corresponding to the driving style, thereby determining the driving status of the main vehicle.

[0028] Furthermore, the driving risks include car-following risk and lane-changing risk, and the driving risk thresholds include car-following risk threshold and lane-changing risk threshold;

[0029] The method for determining the driving status of the main vehicle by comparing the driving risks experienced by the main vehicle with the driving risk thresholds corresponding to the driving style is as follows:

[0030] When the risk of following the car is less than or equal to the risk threshold, the car is driving normally.

[0031] When the risk of following the car exceeds the following risk threshold, the risk of left and right lane changes faced by the driver is compared with the lane change risk threshold corresponding to the driving style, and divided into the following four situations:

[0032] If the risk of changing lanes on the left and the risk of changing lanes on the right are both less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the left or the right.

[0033] If the risk of changing lanes on the left is less than or equal to the risk threshold for changing lanes and the risk of changing lanes on the right is greater than the risk threshold for changing lanes, then the main vehicle will change lanes on the left.

[0034] If the risk of changing lanes on the left is greater than the lane change risk threshold and the risk of changing lanes on the right is less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the right.

[0035] If the risk of changing lanes on the left and the risk of changing lanes on the right are both greater than or equal to the lane change risk threshold, the main vehicle will apply the brakes.

[0036] Furthermore, the following risk threshold and lane change risk threshold are determined as follows: based on a publicly available natural driving dataset, cluster analysis is performed on the following risk and lane change risk of drivers in real-world driving situations, dividing them into three categories. The cluster centers of the three categories are respectively used as the following risk threshold and lane change risk threshold for aggressive, normal, and conservative driving styles.

[0037] Beneficial effects:

[0038] (1) In this invention, the main vehicle data includes the main vehicle's physical mass, lateral coordinate, longitudinal coordinate, speed, and acceleration; the surrounding vehicle data includes the physical mass, lateral coordinate, longitudinal coordinate, length, width, speed, and acceleration of each surrounding vehicle. Based on a driving safety field model, this invention comprehensively considers multiple factors such as the main vehicle's physical mass, lateral coordinate, longitudinal coordinate, speed, and acceleration, as well as the surrounding vehicles' physical mass, lateral coordinate, longitudinal coordinate, length, width, speed, acceleration, and heading angle, thereby accurately calculating the driving risks posed by surrounding vehicles to the main vehicle.

[0039] (2) This invention calculates the following risk and lane-changing risk of the main vehicle by weighting the driving risks posed by surrounding vehicles. By weighting the driving risks posed by individual surrounding vehicles, this invention obtains the longitudinal following risk and lateral lane-changing risk of the main vehicle, comprehensively considering driving risks in both longitudinal and lateral dimensions, thus further ensuring driving safety.

[0040] (3) This invention calculates the potential energy risk field strength of surrounding vehicle A by combining the risk mass of surrounding vehicle A with the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, and width of surrounding vehicle A. In reality, the larger the vehicle size, the greater the risk it generates, and the different effects it has on different directions. This invention introduces the vehicle's geometric properties, solving the problem that the original potential energy field model treats the vehicle as a point mass with only mass and no size, and the potential energy field is circular, without considering the vehicle's geometric dimensions. Furthermore, it uses the peak of the fourth-order central moment flattening function based on the two-dimensional Gaussian function, thus making the potential energy field of stationary vehicles more accurate.

[0041] (4) This invention calculates the kinetic energy risk field strength of surrounding vehicle A by combining the risk mass of surrounding vehicle A with the lateral and longitudinal coordinates of the main vehicle and the lateral and longitudinal coordinates, length, width, and speed of surrounding vehicle A. In addition to incorporating vehicle geometric properties for calculation, it also considers that the speed of surrounding vehicles has different effects on the lateral and longitudinal directions. The speed of surrounding vehicles is introduced into the kinetic field model to calculate the influence of the speed of surrounding vehicles in different lateral and longitudinal directions, further improving the accuracy of the calculation of the kinetic energy risk field strength of surrounding vehicles.

[0042] (5) Based on the risk field strength of the surrounding vehicle A obtained in step S13, this invention calculates the driving risk of the surrounding vehicle A to the main vehicle by combining the physical mass, speed and acceleration of the main vehicle and the speed of the surrounding vehicle A. Considering that the risk faced by the main vehicle is related to the relative speed of the two vehicles, rather than only to the speed of the main vehicle, the relative speed of the vehicles is introduced into the final formula for calculating the driving risk of the surrounding vehicle to the main vehicle. At the same time, the acceleration of the main vehicle is also introduced to further improve the accuracy of risk assessment.

[0043] (6) This invention calculates the following risk to the main vehicle based on the weighted average of the driving risks posed by the surrounding vehicles in front, to the left front, and to the right front; and calculates the lane-changing risk to the main vehicle based on the weighted average of the driving risks posed by the surrounding vehicles to the main vehicle to the left front, to the left rear, to the right front, and to the right rear. Because the magnitude of the following risk to the main vehicle is mainly affected by the surrounding vehicles in front, to the left front, and to the right front, there is no need to consider the driving risks posed by other surrounding vehicles to the main vehicle; and because the magnitude of the lane-changing risk to the main vehicle is mainly affected by the vehicles in the adjacent lanes, there is no need to consider the driving risks posed by the surrounding vehicles in front.

[0044] (7) This invention determines the driving status of the vehicle by selecting a driving style and comparing the driving risks borne by the main vehicle with the driving risk thresholds corresponding to each driving style. After the driver or passenger selects a driving style, this invention compares the corresponding threshold with the driving risks faced by the main vehicle to determine the driving status of the main vehicle, thus realizing personalized intelligent decision-making for the vehicle and ensuring that the decision meets the driver's or passenger's requirements for a comfortable driving style.

[0045] (8) The present invention determines the driving status of the main vehicle by comparing the driving risk faced by the main vehicle with the driving risk threshold corresponding to the driving style: when the following risk faced by the main vehicle is less than or equal to the following risk threshold, the main vehicle drives normally; when the following risk faced by the main vehicle is greater than the following risk threshold, the left and right lane change risks faced by the main vehicle are compared with the lane change risk threshold corresponding to the driving style, and are divided into the following four situations: if both the left and right lane change risks are less than or equal to the lane change risk threshold, the main vehicle performs a left lane change or a right lane change; if the left lane change risk is less than or equal to the lane change risk threshold and the right lane change risk is greater than the lane change risk threshold, the main vehicle performs a left lane change; if the left lane change risk is greater than the lane change risk threshold and the right lane change risk is less than or equal to the lane change risk threshold, the main vehicle performs a right lane change; if both the left and right lane change risks are greater than or equal to the lane change risk threshold, the main vehicle applies the brakes. The present invention first compares the following risk and then compares the lane change risk, ensuring personalized decision-making while ensuring the driving safety of the main vehicle.

[0046] (9) Based on publicly available natural driving datasets, this invention uses cluster analysis to assess the following risk and lane-changing risk of drivers in real-world road driving situations, classifying them into three categories. The cluster centers of these three categories serve as the following risk threshold and lane-changing risk threshold for aggressive, normal, and conservative driving styles, respectively. Furthermore, cluster analysis can be used to further classify the types beyond the three categories to meet the needs of different driving styles. Attached Figure Description

[0047] Figure 1 A diagram illustrating a personalized decision-making framework for intelligent connected vehicles based on multidimensional risk assessment, provided for this invention.

[0048] Figure 2 This is a schematic diagram showing the distribution of vehicles around the present invention;

[0049] Figure 3 This is a flowchart of the personalized intelligent decision-making process for vehicles according to the present invention. Detailed Implementation

[0050] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0051] Example 1:

[0052] This embodiment provides a risk assessment method for intelligent connected vehicles. (See appendix) Figure 1 This includes: data acquired by sensors and vehicle-road cooperative perception facilities of intelligent connected vehicles in an intelligent connected environment for driving risk assessment. The data includes data of the main vehicle and data of surrounding vehicles. The driving risk posed by each surrounding vehicle to the main vehicle is calculated using a driving safety field model. The main vehicle data includes the physical mass, lateral coordinate, longitudinal coordinate, speed and acceleration of the main vehicle. The surrounding vehicle data includes the physical mass, lateral coordinate, longitudinal coordinate, length, width, speed and acceleration of each surrounding vehicle.

[0053] This embodiment is based on a driving safety field model, and comprehensively considers multiple factors such as the physical mass, lateral coordinates, longitudinal coordinates, speed, and acceleration of the main vehicle, as well as the physical mass, lateral coordinates, longitudinal coordinates, length, width, speed, acceleration, and heading angle of surrounding vehicles, so as to accurately calculate the driving risks posed by surrounding vehicles to the main vehicle.

[0054] For further details, please see the appendix. Figure 2 The surrounding vehicles include: vehicles in front, vehicles to the left front, vehicles to the left rear, vehicles to the right front, and vehicles to the right rear. The maximum number of surrounding vehicles is 5, and the minimum number is 0.

[0055] Vehicles in front and around f v The vehicle is defined as the vehicle with the smallest longitudinal distance located within 120m in front of the lane where the main vehicle is located;

[0056] Vehicles around the left front v The vehicle is defined as the one with the smallest longitudinal distance located within 120m ahead of the lane adjacent to the left of the main vehicle.

[0057] Vehicles around the left rear v The vehicle is defined as the vehicle with the smallest longitudinal distance located within 120m behind the adjacent lane on the left side of the main vehicle.

[0058] Vehicles around the right front v The vehicle is defined as the one with the smallest longitudinal distance located within 120m ahead of the adjacent lane to the right of the main vehicle.

[0059] Vehicles around the right rear v A vehicle is defined as the one with the smallest longitudinal distance located within 120m behind the adjacent lane on the right side of the main vehicle.

[0060] Furthermore, the method for calculating the driving risk posed by each surrounding vehicle to the main vehicle using the driving safety field model is as follows:

[0061] Step S11: Calculate the risk mass of surrounding vehicle A based on its physical mass and speed, using the following formula:

[0062]

[0063] Among them, M sv For the risk quality of surrounding vehicle A, m sv Let v be the physical mass of the surrounding vehicle A. sv The speed of surrounding vehicle A;

[0064] Step S12: On the one hand, the potential energy risk field strength of the surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, and width of the surrounding vehicle A. The calculation formula is as follows:

[0065]

[0066] Among them, E sv_sta M represents the potential energy risk field strength of surrounding vehicle A, where A is a field strength coefficient greater than 0. sv For the risk quality of surrounding vehicle A obtained in step S11, x ev The longitudinal coordinate of the main vehicle, x sv Let w be the longitudinal coordinate of the surrounding vehicle A. x L is the weighting factor for the length of surrounding vehicle A. sv Let y be the length of the surrounding vehicle A. ev The longitudinal coordinate of the main vehicle, y sv Let w be the lateral coordinate of the surrounding vehicle A.y W is the weighting factor for the width of surrounding vehicle A. sv Let d be the width of the surrounding vehicle A, and d be the distance vector between the main vehicle and the surrounding vehicle A, where d = (x ev -x sv ,y ev -y sv );

[0067] On the other hand, the risk mass of surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle with the lateral and longitudinal coordinates, length, width, and speed of surrounding vehicle A. The calculation formula is as follows:

[0068]

[0069] Among them, E sv_va Let v be the kinetic energy risk field strength of surrounding vehicle A, α be the weighting coefficient related to the speed of surrounding vehicle A, and v be the velocity of surrounding vehicle A. sv Let be the velocity of the surrounding vehicle A; θ be the angle between the velocity direction of the surrounding vehicle A and d, with clockwise direction being positive; γ be a weighting coefficient related to the angle θ; β be a weighting coefficient related to the acceleration of the surrounding vehicle A; and a ∈ 0. sv Let A be the acceleration of a vehicle A in the vicinity.

[0070] Step S13: Combine the potential energy risk field strength and kinetic energy risk field strength of the surrounding vehicle A obtained in step S12 to obtain the risk field strength of the surrounding vehicle A. The calculation formula is as follows:

[0071] E sv =E sv_sta +E sv_va Formula (4)

[0072] Among them, E sv Let E be the risk field strength of a certain vehicle A in the vicinity. sv_sta E represents the potential energy risk field strength of the surrounding vehicle A obtained in step S12. sv_va The kinetic energy risk field strength of the surrounding vehicle A obtained in step S12;

[0073] Step S14: Based on the risk field strength of surrounding vehicle A obtained in step S13, and combining the physical mass, speed, and acceleration of the main vehicle with the speed of surrounding vehicle A, calculate the driving risk posed by surrounding vehicle A to the main vehicle. The calculation formula is as follows:

[0074]

[0075] Where, r ev E represents the driving risk posed by surrounding vehicle A to the main vehicle. sv For the risk field strength of the surrounding vehicle A obtained in step S13, m evThe physical mass of the main vehicle; w v The relative speed of the main vehicle and surrounding vehicle A |v ev -v sv | weighting coefficient; v ev Main vehicle speed; v sv w represents the speed of surrounding vehicles. a The weighting coefficient for the main vehicle acceleration; a ev The acceleration of the main vehicle.

[0076] Step S15: Calculate the driving risks posed by each surrounding vehicle to the main vehicle based on steps S11-S14 (i.e., the driving safety field model).

[0077] Furthermore, the surrounding vehicle data also includes the heading angles of the surrounding vehicles. When the heading angle of surrounding vehicle A is not 0°, the coordinates of surrounding vehicle A need to be transformed. The coordinate transformation formula is:

[0078]

[0079] Where, x' sv Let y' be the transformed longitudinal coordinate of the surrounding vehicle A. sv Let A be the lateral coordinate of the surrounding vehicle A after the transformation. Let A be the heading angle of the surrounding vehicle A, with counterclockwise as positive.

[0080] Transform the coordinates x' of the surrounding vehicle A. sv y' sv Replace the coordinates x of the original surrounding vehicle A respectively sv y sv Perform the calculation.

[0081] In real-world driving environments, there are situations where a vehicle's heading angle is not 0°, such as when changing lanes and cutting in, where the vehicle body is not parallel to the road direction. In such cases, the risk field of a moving vehicle needs to be deflected at a certain angle to more accurately describe the driving risks posed by surrounding vehicles to the main vehicle.

[0082] Example 2:

[0083] Another embodiment of this application discloses a risk assessment method for intelligent connected vehicles, see appendix. Figure 1 It also includes: calculating the following risk and lane-changing risk of the main vehicle by weighting the driving risks posed by surrounding vehicles to the main vehicle.

[0084] This embodiment calculates the driving risks posed by individual surrounding vehicles by weighting the risks to the main vehicle, including longitudinal following risks and lateral lane-changing risks. By comprehensively considering the driving risks in both longitudinal and lateral dimensions, driving safety is further ensured.

[0085] Furthermore, since the magnitude of the following risk experienced by the main vehicle is primarily influenced by the surrounding vehicles in front, to the left, and to the right, the following risk experienced by the main vehicle is calculated by weighting the driving risks posed by these surrounding vehicles. The specific calculation method is as follows:

[0086] r cf =w lf r sv_lf +w mf r sv_mf +w rf r sv_rf Formula (7)

[0087] Where, r cf The risk of being followed by the main vehicle, w lf r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the left front. sv_lf To mitigate the driving risks posed by surrounding vehicles to the left front of the vehicle, w mf r represents the driving risk weighting coefficient corresponding to the surrounding vehicles ahead. sv_mf To mitigate the driving risks posed by surrounding vehicles to the main vehicle, w rf r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the vehicle to the right front. sv_rf The driving risks posed to the main vehicle by surrounding vehicles on the right front;

[0088] Furthermore, since the magnitude of the lane-changing risk borne by the main vehicle is primarily influenced by vehicles in the adjacent lanes, the lane-changing risk borne by the main vehicle is calculated by weighting the driving risks posed by vehicles surrounding the main vehicle on its left front, left rear, right front, and right rear. The specific calculation method is as follows:

[0089]

[0090] Where, r lc_l The lane-changing risk faced by the main vehicle in the left lane, w lf r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the left front. sv_lf To mitigate the driving risks posed by surrounding vehicles to the left front of the vehicle, w lr r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the left rear. sv_lr The driving risk posed to the main vehicle by vehicles surrounding it on the left rear; lc_r To mitigate the lane-changing risks associated with the right lane, w rf r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the vehicle to the right front. sv_rf To mitigate the driving risks posed by surrounding vehicles to the right front of the vehicle, w rr r represents the driving risk weighting coefficient corresponding to the vehicles surrounding the right rear. sv_rrThis is to mitigate the driving risks posed to the main vehicle by vehicles surrounding the right rear.

[0091] Example 3:

[0092] Another embodiment of this application provides a personalized decision-making method for intelligent connected vehicles, based on the intelligent connected vehicle risk assessment method described in Embodiments 1 and 2. (See appendix.) Figure 1 This includes: selecting a driving style, comparing the driving risks experienced by the main vehicle with the driving risk thresholds corresponding to the driving style, thereby determining the driving status of the main vehicle.

[0093] In this embodiment, after the driver or passenger selects a driving style, the corresponding threshold is compared with the driving risks faced by the main vehicle to achieve personalized intelligent vehicle decision-making that takes into account the comfort of the driver or passenger.

[0094] Furthermore, the driving risks include car-following risk and lane-changing risk, and the driving risk thresholds include car-following risk threshold and lane-changing risk threshold;

[0095] See appendix Figure 3 The method for determining the driving status of the main vehicle by comparing the driving risks experienced by the main vehicle with the driving risk thresholds corresponding to the driving style is as follows:

[0096] When the risk of following the car is less than or equal to the risk threshold, the car is driving normally.

[0097] When the risk of following the car exceeds the following risk threshold, the risk of left and right lane changes faced by the driver is compared with the lane change risk threshold corresponding to the driving style, and divided into the following four situations:

[0098] If the risk of changing lanes on the left and the risk of changing lanes on the right are both less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the left or the right.

[0099] If the risk of changing lanes on the left is less than or equal to the risk threshold for changing lanes and the risk of changing lanes on the right is greater than the risk threshold for changing lanes, then the main vehicle will change lanes on the left.

[0100] If the risk of changing lanes on the left is greater than the lane change risk threshold and the risk of changing lanes on the right is less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the right.

[0101] If the risk of changing lanes on the left and the risk of changing lanes on the right are both greater than or equal to the lane change risk threshold, the main vehicle will apply the brakes.

[0102] Furthermore, the following risk threshold and lane change risk threshold are determined as follows: Based on publicly available natural driving datasets, cluster analysis is performed on the following risk and lane change risk under real-world driving conditions, dividing them into three categories. The cluster centers of the three categories are respectively used as the following risk threshold and lane change risk threshold for aggressive, normal, and conservative driving styles. The specific steps are as follows:

[0103] Step S31: Based on the existing publicly available natural driving dataset, calculate the following risk of the driver in real road driving conditions using formulas (1)-(7), that is, the following risk at the initial moment of the driver performing a lane change; calculate the lane change risk of the driver in real road driving conditions using formulas (1)-(6) and (8), that is, the lane change risk at the initial moment of the driver performing a lane change.

[0104] Step S32: Input the following risk and lane-changing risk obtained in Step S31 under real-world driving conditions into the clustering analysis algorithm. Set the number of clusters to 3. Define the centers of the three clusters as the following risk threshold and lane-changing risk threshold corresponding to the three driving styles of aggressive, normal, and conservative drivers or passengers, respectively. See Table I for details.

[0105] Table 1 Risk thresholds for each driving style

[0106] conservative normal type radical <![CDATA[Following risk threshold R cf > <![CDATA[R cf_c ]]> <![CDATA[R cf_m ]]> <![CDATA[R cf_r ]]> <![CDATA[Lane - changing risk threshold R lc > <![CDATA[R lc_c ]]> <![CDATA[R lc_m ]]> <![CDATA[R lc_r ]]>

[0107] Example 4:

[0108] This application provides a specific embodiment of a risk assessment method and personalized decision-making method for intelligent connected vehicles in a one-way three-lane scenario, based on a selected conservative driving style. The specific steps are as follows:

[0109] Step 1: In an intelligent connected environment, based on the sensors and vehicle-road cooperative perception facilities equipped by the intelligent connected vehicle itself, the physical mass, lateral coordinates, longitudinal coordinates, speed and acceleration of the main vehicle are obtained; the physical mass, lateral coordinates, longitudinal coordinates, length, width, speed, acceleration and heading angle of the surrounding vehicles are obtained.

[0110] Step 2: Calculate the driving risks posed by surrounding vehicles to the main vehicle according to formulas (1)-(6);

[0111] Step 3: Based on the driving risks posed by the surrounding vehicles to the main vehicle in Step 2, the following risk r borne by the main vehicle is calculated by weighted average using formula (7). cf On the other hand, the lane-changing risk r faced by the main vehicle in the left and right lanes is calculated by weighted average calculation using formula (8). lc_l and r lc_r ;

[0112] Step 4: The conservative following risk threshold is R. cf_c The lane-changing risk threshold is R. lc_c ;

[0113] The following risk r obtained in step three cf The conservative following risk threshold is R cf_c Comparison:

[0114] When the risk of following the car is less than or equal to the risk threshold, i.e. r cf ≤R cf_c At that time, the main vehicle was being driven normally;

[0115] When the risk of following the car is greater than the risk threshold, i.e. r cf >R cf_c At that time, the risk of lane changing faced by the left and right lanes will be considered. lc_l and r lc_r Each is compared with the lane change risk threshold R. lc_c The comparison and judgment are divided into the following four situations:

[0116] If r lc_l ≤R lc_c or r lc_r ≤R lc_c If so, the main vehicle will change lanes to the left or right;

[0117] If r lc_l ≤R lc_c And r lc_l >R lc_c If so, the main vehicle will change lanes to the left;

[0118] If r lc_l >R lc_c And r lc_r ≤R lc_c Then the main vehicle will change lanes to one side;

[0119] If r lc_l >R lc_c And r lc_l >R lc_c Then the main vehicle will apply the brakes.

[0120] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A risk assessment method for intelligent connected vehicles, characterized in that, include: Based on the data obtained by the sensors and vehicle-road cooperative perception facilities of intelligent connected vehicles in the intelligent connected environment for driving risk assessment, including the data of the main vehicle and the data of surrounding vehicles, the driving risk of each surrounding vehicle to the main vehicle is calculated by using the driving safety field model. The main vehicle data includes the main vehicle's physical mass, lateral coordinates, longitudinal coordinates, speed, and acceleration; the surrounding vehicle data includes the physical mass, lateral coordinates, longitudinal coordinates, length, width, speed, and acceleration of each surrounding vehicle. The method of calculating the driving risk to the main vehicle from each surrounding vehicle using a driving safety field model is as follows: Step S11: Calculate the risk mass of the surrounding vehicle A based on its physical mass and speed. Step S12: On the one hand, the potential energy risk field strength of the surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, and width of the surrounding vehicle A; on the other hand, the kinetic energy risk field strength of the surrounding vehicle A is calculated by combining the lateral and longitudinal coordinates of the main vehicle and the lateral, longitudinal coordinates, length, width, and speed of the surrounding vehicle A. Step S13: Combine the potential energy risk field strength and kinetic energy risk field strength of the surrounding vehicle A obtained in step S12 to obtain the risk field strength of the surrounding vehicle A. Step S14: Based on the risk field strength of the surrounding vehicle A obtained in step S13, and combined with the physical mass, speed and acceleration of the main vehicle and the speed of the surrounding vehicle A, calculate the driving risk posed by the surrounding vehicle A to the main vehicle. Step S15: Calculate the driving risks posed by each surrounding vehicle to the main vehicle based on steps S11-S14. The formula for calculating the potential energy risk field strength of the surrounding vehicle A in step S12 is as follows: The formula for calculating the kinetic energy risk field strength of the surrounding vehicle A in step S12 is as follows: The calculation formula for step S14 is as follows: in, Let the potential energy risk field strength be the strength of the surrounding vehicle A. A The electric field strength coefficient is greater than 0. The risk quality of the surrounding vehicle A obtained in step S11, The longitudinal coordinate of the main vehicle Let be the longitudinal coordinate of the surrounding vehicle A. The weighting coefficients for the length of surrounding vehicle A. Let A be the length of the surrounding vehicle A. The lateral coordinate of the main vehicle. Let be the lateral coordinate of the surrounding vehicle A. The weighting factor for the width of surrounding vehicle A. The width of the surrounding vehicle A, d Let A be the distance vector between the main vehicle and surrounding vehicle A. , The kinetic energy risk field strength of the surrounding vehicle A. α The weighting coefficient is related to the speed of the surrounding vehicle A. Let A be the speed of the surrounding vehicle A. θ Let the velocity direction of a certain surrounding vehicle A be relative to... d The angle between them is positive in the clockwise direction; γ For the included angle θ Relevant weighting coefficients; β The weighting coefficients are related to the acceleration of the surrounding vehicle A. Let A be the acceleration of a vehicle A in the surrounding area. The driving risks posed to the main vehicle by surrounding vehicle A. The risk field strength of the surrounding vehicle A obtained in step S13, The physical mass of the main vehicle; The relative speed between the main vehicle and surrounding vehicle A Weighting coefficients; Main vehicle speed; The weighting coefficient for the acceleration of the main vehicle; The acceleration of the main vehicle; The surrounding vehicle data also includes the heading angles of the surrounding vehicles. When the heading angle of surrounding vehicle A is not 0°, the coordinates of surrounding vehicle A need to be transformed. The coordinate transformation formula is: in, Here are the transformed longitudinal coordinates of the surrounding vehicle A. Let A be the lateral coordinate of the surrounding vehicle A after the transformation. Let A be the heading angle of the surrounding vehicle A, with counterclockwise being positive; Transform the coordinates of the surrounding vehicle A , Replace the coordinates of the surrounding vehicle A respectively , Perform the calculation.

2. The intelligent connected vehicle risk assessment method as described in claim 1, characterized in that, Based on the driving risks posed by surrounding vehicles to the main vehicle, the following risk and lane-changing risk borne by the main vehicle are calculated separately using weighted averages.

3. The intelligent connected vehicle risk assessment method as described in claim 2, characterized in that, The risk of being followed by another vehicle is calculated by weighting the driving risks posed to the vehicle by the surrounding vehicles in front, to the left front, and to the right front. The lane-changing risk to the main vehicle is calculated by weighting the driving risks posed by the vehicles surrounding the main vehicle from the left front, left rear, right front, and right rear.

4. A personalized decision-making method for intelligent connected vehicles, based on the intelligent connected vehicle risk assessment method according to any one of claims 1-3, comprising: By selecting a driving style, the driving risks experienced by the main vehicle are compared with the driving risk thresholds corresponding to each driving style, thereby determining the driving status of the main vehicle.

5. The personalized decision-making method for intelligent connected vehicles as described in claim 4, characterized in that: The driving risks include following risks and lane-changing risks, and the driving risk thresholds include following risk thresholds and lane-changing risk thresholds. The method for determining the driving status of the main vehicle by comparing the driving risks experienced by the main vehicle with the driving risk thresholds corresponding to the driving style is as follows: When the risk of following the car is less than or equal to the risk threshold, the car is driving normally. When the risk of following the car exceeds the following risk threshold, the risk of left and right lane changes faced by the driver is compared with the lane change risk threshold corresponding to the driving style, and divided into the following four situations: If the risk of changing lanes on the left and the risk of changing lanes on the right are both less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the left or the right. If the risk of changing lanes on the left is less than or equal to the risk threshold for changing lanes and the risk of changing lanes on the right is greater than the risk threshold for changing lanes, then the main vehicle will change lanes on the left. If the risk of changing lanes on the left is greater than the lane change risk threshold and the risk of changing lanes on the right is less than or equal to the lane change risk threshold, then the main vehicle will change lanes on the right. If the risk of changing lanes on the left and the risk of changing lanes on the right are both greater than or equal to the lane change risk threshold, the main vehicle will apply the brakes.

6. The personalized decision-making method for intelligent connected vehicles as described in claim 5, characterized in that, The following risk threshold and lane change risk threshold are determined as follows: Based on the publicly available natural driving dataset, cluster analysis is performed on the following risk and lane change risk of drivers in real road driving situations, and they are divided into three categories. The cluster centers of the three categories are respectively used as the following risk threshold and lane change risk threshold for aggressive, normal and conservative driving styles.

Images