A trajectory planning method for intelligent vehicles considering chassis dynamics boundaries

By constructing a driving safety field model that considers the chassis dynamics and using the vehicle dynamics model, the intelligent vehicle trajectory planning is optimized in real time, and the problem of failure to effectively consider the boundary changes of the chassis dynamics in the existing technology is solved, and the intelligent vehicle is improved Driving reliability and safety.

CN117246344BInactive Publication Date: 2024-11-29NANJING UNIV OF AERONAUTICS & ASTRONAUTICS +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202311396732.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-10-26
Publication Date
2024-11-29
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing intelligent vehicle trajectory planning methods fail to effectively consider the dynamic boundary of the vehicle chassis, which makes it difficult to ensure the reliability and safety of trajectory planning when driving on changing road surfaces.

Method used

By collecting bicycle and environmental information in real time, a driving safety field model that considers the chassis dynamics is built, the s-shaped curve is used to optimize the generation trajectory, and the chassis dynamics constraints are added based on the vehicle dynamics model, and the trajectory planning is updated in real time to adapt to the trajectory. Pavement changes.

Benefits of technology

It improves the driving reliability and safety of smart cars, reduces control risks, and can better adapt to changing driving environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117246344B_ABST
    Figure CN117246344B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for trajectory planning of an intelligent vehicle considering chassis dynamics boundaries, and the steps are as follows: real-time collection of vehicle information and environmental vehicle information required for trajectory planning; estimation of the current vehicle state; construction of a driving safety field considering chassis dynamics, using S-curve optimization, defining chassis dynamics boundaries, and realizing generation of a planning trajectory under chassis dynamics constraints; driving according to the planning trajectory obtained by the above planning, and performing trajectory planning again in combination with current vehicle information and environmental information. The method of the present invention calculates the chassis dynamics boundary, adds the chassis dynamics boundary to the driving safety field, and performs trajectory planning by adding constraints during the trajectory planning process, thereby further improving the reliability and safety of the intelligent vehicle.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent vehicle path planning, and in particular relates to an intelligent vehicle trajectory planning method taking chassis dynamic boundaries into consideration. Background Art

[0002] In the field of intelligent vehicles, trajectory planning is a key technology used to determine the optimal trajectory of a vehicle under different environments and driving tasks. Traditional trajectory planning methods are usually based on static environment models and do not take into account the limitations of the vehicle chassis dynamics boundaries. However, in real driving, the chassis dynamics boundaries of the vehicle have an important impact on trajectory planning. The chassis dynamics boundaries include limitations such as the vehicle's acceleration, braking force, and maximum steering angle, which directly affect the stability and safety of the vehicle while driving.

[0003] The main obstacle avoidance trajectory planning methods now include artificial potential field method, etc. The artificial potential field method regards the movement of the vehicle in the surrounding environment as the movement of the vehicle in an artificially established virtual field. The target point generates gravity, the obstacle generates repulsion, and the vehicle moves in the combined force of the two. The vehicle movement is controlled according to the combined force of gravity and repulsion to produce an optimal trajectory without collision. The traditional artificial potential field does not take into account the changes in the chassis dynamic boundary, and the vehicle is driving on a changing road surface, so it cannot achieve good results when the chassis dynamic boundary changes.

[0004] At present, there has been some research on the trajectory planning of intelligent vehicles. For example, the improved artificial potential field method in the Chinese invention patent application number CN201610079992.0, entitled “A robot trajectory planning method and research based on an improved artificial potential field method”, solves the local minimum point problem in the trajectory planning of the traditional potential field method, improves the real-time performance of trajectory planning, and has high environmental adaptation efficiency. However, the patent does not consider the changes in vehicle dynamics during driving; the Chinese invention patent application number CN202310720471.9, entitled “Method and device for adjusting a vehicle based on road tire adhesion coefficient”, provides a method for identifying a road adhesion coefficient, and adjusts the wheel torque by changing the road adhesion coefficient. However, the patent only uses the obtained tire adhesion coefficient to adjust the vehicle wheel tracking method, and does not consider the impact of the complete chassis dynamics changes.

[0005] In summary, most of the existing trajectory planning methods do not consider the changes of chassis dynamic boundaries during vehicle driving, and cannot achieve good results when driving on changing road surfaces. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a smart car trajectory planning method that takes into account the chassis dynamics boundary, so as to solve the problem that the trajectory planning method in the prior art cannot adapt to the changes in the road surface during driving, resulting in the difficulty in ensuring the reliability of the path obtained by the smart car trajectory planning. The method of the present invention calculates the chassis dynamics boundary, adds the chassis dynamics boundary to the driving safety field, and performs trajectory planning by adding constraints during the trajectory planning process, thereby further improving the reliability and safety of the smart car.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] The present invention provides a smart car trajectory planning method considering chassis dynamics boundaries, comprising the following steps:

[0009] (1) Real-time collection of vehicle information and surrounding vehicle information required for trajectory planning;

[0010] (2) Estimate the current vehicle state using the vehicle information obtained in step (1);

[0011] (3) constructing a driving safety field model, by inputting the current vehicle state obtained in step (2) into the driving safety field model, constructing a driving safety field that takes chassis dynamics into consideration, using S-curve optimization to define the chassis dynamics boundary, and generating a planning trajectory under the chassis dynamics constraint;

[0012] (4) Drive along the planned trajectory obtained in step (3) above, and plan the trajectory again every Δt seconds based on the current vehicle information and environment information.

[0013] Furthermore, in step (1), the information required for trajectory planning is collected through millimeter wave radar, lidar, camera, IMU, GPS and vehicle sensors (such as vehicle speed sensor, yaw rate sensor, center of mass sideslip angle sensor and front wheel steering angle sensor).

[0014] Furthermore, the information required for trajectory planning collected in step (1) is the vehicle lateral, longitudinal acceleration and vehicle yaw rate information of the self-vehicle, as well as the state of the first type of stationary objects, the second type of stationary objects, the moving objects, the type of moving objects, the position of the moving objects and the state and position of the smart vehicle required for trajectory planning; the first type of stationary objects refers to objects that will cause great losses in the event of a collision; the second type of stationary objects refers to objects that will not collide with the vehicle but will restrict the driver's driving behavior; moving objects refer to vehicles or pedestrians moving on the road.

[0015] Furthermore, the specific steps of step (2) are as follows:

[0016] (21) Build vehicle dynamics model:

[0017] Taking the vehicle body coordinate system as the reference system, the vehicle dynamics model is established:

[0018] Longitudinal Movement:

[0019]

[0020] Lateral Movement:

[0021]

[0022] Yaw motion:

[0023]

[0024] Where m is the vehicle mass; J Z is the moment of inertia of the vehicle around the z-axis; V x is the longitudinal speed; V y is the lateral vehicle speed; is the yaw angular velocity; F xi and F yi are the longitudinal force and lateral force of the tire, respectively, where the subscripts i = fl, fr, rl, rr represent the front left wheel, front right wheel, rear left wheel, and rear right wheel, respectively; l a and l b Respectively represent the distance from the center of mass to the front axle and the rear axle, c is the wheelbase of the left and right wheels of the vehicle, and the front and rear wheelbases of the vehicle are consistent, δ is the front wheel turning angle;

[0025] (22) Estimate the vehicle state based on the vehicle dynamics model:

[0026] The sensor is used to measure the vehicle's lateral and longitudinal acceleration and the vehicle's yaw rate information to estimate the vehicle's current state.

[0027] Assume the state quantity of the observation system is x 1 =[β V x ] T , and its nonlinear state equation is as follows:

[0028]

[0029] Assume the state quantity of the observation system is y 1 =a y , and its nonlinear state equation is as follows:

[0030]

[0031] In the formula, f 1 () and h 1() are the vehicle lateral nonlinear state equation expressions and the vehicle longitudinal nonlinear state equation expressions, β is the road adhesion coefficient, a x and a y are the lateral acceleration and longitudinal acceleration of the vehicle, r is the wheel radius, w 1 and v 1 are the process noise vector and the measurement noise vector, respectively, both of which are zero-mean Gaussian white noise;

[0032] Convert equation (4) and equation (5) into nonlinear discrete form:

[0033]

[0034] In the formula, x 1,k ,y 1,k To predict the lateral and longitudinal vehicle states at time k, the vehicle dynamics model is used to estimate the position after obtaining it through GPS, x 1,k+1 is the predicted state quantity at the next moment.

[0035] In the process of state estimation, β and V x The value of is predicted at each discrete time step, and the output value is a y The measured value is compared with that obtained by the vehicle sensor to complete the correction of the state quantity; the road adhesion coefficient of the current road surface and the current state quantity of the vehicle are obtained through linear discretization.

[0036] Furthermore, the step (3) specifically includes:

[0037] (31) Construct a driving safety field model considering chassis dynamics;

[0038] (32) Calculate the attraction and repulsion of the driving safety field;

[0039] (33) Generate planning trajectory by optimizing S-curve and adding chassis dynamic constraints;

[0040] (34) Consider the constraints imposed by vehicle chassis dynamics on the planned path to improve vehicle stability during path tracking.

[0041] Furthermore, the step (31) specifically includes:

[0042] (311) The driving safety field considering chassis dynamics is composed of the potential energy field generated by stationary objects or roads, the kinetic energy field determined by moving vehicles, the behavior field determined by the driver's psychology / physiology, and the vehicle-road field caused by different chassis dynamics; the driving safety model expression is as follows:

[0043] E S =E R +E V +ED +E C (7)

[0044] In the formula, E S For driving safety field strength; E R is the potential energy field strength; E V is the kinetic energy field strength; E D is the behavioral field strength; E C is the field strength of the vehicle road;

[0045] (312) Potential energy field modeling;

[0046] The potential field is a physical field generated by the impact of objects or road conditions in the traffic environment on driving safety. Its size and direction are determined by the properties of the stationary object itself, its relative position, and road conditions. i ,y i ) is expressed by equation (8):

[0047]

[0048] In the formula, (x i ,y i ) is the coordinate of the center of mass of the stationary object i; the x-axis is the longitudinal direction of the road; the y-axis is the vertical direction of the road; E R_ij is the risk degree that the stationary object i will cause to the surrounding environment under certain road conditions. The larger the value, the greater the risk. The direction and r of the potential energy field strength ij Same direction; r ij =(x j -x i ,y j -y i ) is the surrounding environment (x j ,y j ) and the vector distance between the stationary object; k 1 , K is a constant to be determined; M i is the equivalent mass of stationary object i; R i is (x i ,y i ) Road condition influencing factors;

[0049] Equivalent mass M i Determined by the properties of the stationary object i itself, the risk level it poses depends on the type, mass and motion state of the stationary object; the function expression is as follows:

[0050]

[0051] In the formula, m i is the mass of the stationary object; H i is the type of stationary object; ak ,β k is a constant to be determined; v i is the speed of a stationary object;

[0052] (313) Kinetic energy field modeling;

[0053] The kinetic energy field is a physical field generated by the impact of moving objects (mainly surrounding vehicles) on driving safety in the traffic environment. The magnitude and direction of the field depend on the relative mass, position, speed and direction of the surrounding vehicles. i ,y i ) position of a moving object s in its surroundings (x j ,y j ) is formed as follows:

[0054]

[0055] In the formula, E V_sj is the moving object s around (x s ,y s ) forms a kinetic energy field, the direction of the field intensity vector is the same as r sj same;

[0056] r sj =(x j -x s ,y j -y s ) is the surrounding environment (x j ,y j ) and the vector distance between the moving object; R s is the road condition influencing factor; M s is the equivalent mass of the moving object s; v s is the speed of the moving object s; θ s is the velocity direction of the moving object s and r sj The angle between G and k 1 ,k 2 is a constant; exp[k 2 ·v s ·cos(θ s )] represents the law of change of the driving risk posed by the moving object s to the approaching vehicle as the moving vehicle approaches in the direction and speed;

[0057] (314) Behavioral field modeling;

[0058] The behavior field is a physical field generated by the driver's physiological / psychological factors affecting driving safety. Its field strength depends on the driver's behavior. n ,y n ) in the surrounding environment (x j ,yj ) formed in the behavioral field E D_nj The expression is as follows:

[0059] E D_nj =E V_nj ·D rn (11)

[0060] In the formula, E V_nj is the vehicle n in the surrounding environment (x j ,y j ) formed in the kinetic energy field, D rn is the risk factor for the driver;

[0061] (315) Vehicle road field modeling;

[0062] The vehicle-road field is a physical field that is affected by the road adhesion coefficient and vehicle speed during driving. The field strength depends on the road adhesion coefficient and vehicle speed. i ,y i ) The vehicle road field expression formed by the vehicle with a vehicle speed v at this position is:

[0063] E D_j =C j ·ln[β j ]·exp[v]+δ k (12)

[0064] Where v is the current speed of the vehicle; β ij is (x i ,y i ) at the road surface adhesion coefficient; C ij is the chassis dynamics factor; δ k is a constant to be determined;

[0065] (316) The driving safety field strength is formed by the superposition of the above-mentioned potential energy field, kinetic energy field, behavior field and vehicle road field.

[0066] Furthermore, the step (32) specifically includes:

[0067] (321) The generation of the trajectory requires that the target point has an attractive force on the ego vehicle, with the direction from the ego vehicle to the target point, and that the obstacle has a repulsive force on the ego vehicle, with the direction from the obstacle to the ego vehicle. The definition of the attractive force field is as follows:

[0068] Q Trgl =a(YY Trgl ) 2 (13)

[0069] Where a is the shape coefficient of the attraction potential field of the target lane, which is used to adjust the amplitude of the attraction potential field; Y is the lateral position of the vehicle; Y Trgl is the lateral position of the centerline of the target lane;

[0070] (322) The driving safety field is the repulsive force potential field of the current working condition, namely:

[0071] Q S =E S (14)

[0072] In the formula, Q S is the repulsive force potential field; E S It is a driving safety area;

[0073] (323) The direction of motion is generated by the negative gradient of the force field function, as shown in Equations (14) and (15):

[0074] F Trql (q)=-kY Trgl (15)

[0075]

[0076] Where p(q) is the distance between the vehicle and the obstacle; p 0 is the influence range, and the field strength beyond this influence range will not affect the generation of repulsive force; k and η are constants; F Trql (q) is the gravitational force generated by the specified target trajectory; F S (q) is the repulsive force exerted by the driving safety field on the vehicle;

[0077] (324) The combined field strength of the gravitational field and the driving safety field is the current total field. The gradient of the current total field is the direction in which the vehicle should move, as shown in equations (16) and (17):

[0078] Q(q)=Q Trgl (q)+Q Rep (q) (17)

[0079] F(q)=-▽Q(q) (18)

[0080] The vehicle position at the next moment is:

[0081] [X 1 ,Y 1 ]=[X 0 ,Y 0 ]+k path F(q)·Δt (19)

[0082] In the formula, [X 1 ,Y 1]The vehicle's position at the next moment; [X 0 ,Y 0 ] is the current position of the vehicle; Δt is the step length of trajectory prediction; k path is a constant.

[0083] Furthermore, the step (33) specifically includes:

[0084] (331) The function selected to generate the s-shaped curve of the planning trajectory is as follows:

[0085]

[0086] Where P a , P b and P c P is the relevant parameter for shaping the S-shaped curve. a is the maximum slope, P b is the amplitude coefficient, P c is the central symmetric point and the point of maximum slope;

[0087] (332) After selecting the S-curve parameterization equation, the tangent direction of the S-curve is calculated at each path point according to the direction of the resultant force vector; based on the tangent direction and the path point, the parameter value of the S-curve is calculated; the parameter will define the position of the point on the S-curve; by gradually adjusting the parameter value, transitioning from one path point to the next path point, a smooth S-curve is formed.

[0088] Furthermore, the step (34) specifically includes:

[0089] (341) The stability boundary of the sideslip angle is determined by the saturated tire slip angles of the front and rear tires. The boundary expression of chassis dynamics is:

[0090]

[0091]

[0092] Where, γ represents the sideslip angle boundary; δ f Represents the front wheel angle; α i represents the slip angle corresponding to the peak tire force of the tire, i=fl,fr,rl,rr represent the left front tire, right front tire, left rear tire, and right rear tire respectively; l f , l r Respectively represent the front wheelbase and rear wheelbase;

[0093] (342) According to equation (21), the stability region of the vehicle is determined, which limits the growth of the yaw rate and the sideslip angle. The stability region can be transformed into the vehicle dynamics boundary constraint in equation (23):

[0094]

[0095] Where K is the definition constraint; β ij is (x i ,y i ) at the road surface adhesion coefficient; β cos is the road adhesion coefficient constraint of the planned path considering smoothness and vehicle stability in path planning; v is the vehicle speed; a is the vehicle acceleration; ω is the yaw rate;

[0096] (343) By constraining the stability region, a safe planning trajectory that the vehicle can pass through is generated.

[0097] Furthermore, in step (4), when the planned trajectory is updated, if the current trajectory can satisfy all constraints, the smart car is controlled to continue to travel along the current trajectory; if the current trajectory cannot satisfy all constraints, the planned trajectory is replanned and the planned trajectory of the smart car is updated. After the update, the smart car will travel along the updated trajectory.

[0098] Beneficial effects of the present invention:

[0099] The present invention fully considers the situation that the vehicle road may change during the driving process of the vehicle, improves the feasibility of the practical application of the smart car, and reduces the control risk of the vehicle.

[0100] The present invention uses a driving safety field that takes chassis dynamics into consideration, and improves vehicle driving safety by increasing the influence of the road adhesion coefficient and the current speed of the vehicle on the field strength. Compared with traditional driving safety fields, it can better reflect the impact of the environment during driving.

[0101] The present invention calculates the chassis dynamics state through a vehicle dynamics model, obtains the current road risk potential field by considering the driving safety field of the chassis dynamics, uses an S-shaped curve to optimize the planned curve, and designs a vehicle chassis dynamics boundary to constrain the planned curve of the vehicle. The present invention fully considers road surface factors, can better ensure the driving safety of the vehicle, and enables the vehicle to adapt to the changing driving environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 Schematic diagram of the method of the present invention.

[0103] Figure 2a It is a vertical schematic diagram of the 3D driving safety field.

[0104] Figure 2b This is a schematic diagram of a 3D driving safety field.

[0105] Figure 3 Generate a schematic diagram for the optimized trajectory. DETAILED DESCRIPTION

[0106] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0107] Reference Figure 1-Figure 3 As shown, a smart car trajectory planning method considering chassis dynamics boundaries of the present invention comprises the following steps:

[0108] (1) Real-time collection of vehicle information and surrounding vehicle information required for trajectory planning;

[0109] Among them, in the step (1), the information required for trajectory planning is collected through millimeter wave radar, laser radar, camera, IMU, GPS and vehicle sensors (such as vehicle speed sensor, yaw rate sensor, center of mass sideslip angle sensor and front wheel steering angle sensor).

[0110] The information required for trajectory planning is the lateral and longitudinal acceleration and yaw rate information of the ego vehicle, as well as the status of the first-category stationary objects, second-category stationary objects, moving objects, the type of moving objects, the position of moving objects and the status and position of the smart vehicle required for trajectory planning; the first-category stationary objects refer to objects that will cause great losses in the event of a collision; the second-category stationary objects refer to objects that will not collide with the vehicle but will restrict the driver's driving behavior; moving objects refer to vehicles or pedestrians moving on the road.

[0111] (2) Estimate the current vehicle state using the vehicle information obtained in step (1); the specific steps are as follows:

[0112] (21) Build vehicle dynamics model:

[0113] Taking the vehicle body coordinate system as the reference system, the vehicle dynamics model is established:

[0114] Longitudinal Movement:

[0115]

[0116] Lateral Movement:

[0117]

[0118] Yaw motion:

[0119]

[0120] Where m is the vehicle mass; J Z is the moment of inertia of the vehicle around the z-axis; V x is the longitudinal speed; V y is the lateral vehicle speed; is the yaw angular velocity; Fxi and F yi are the longitudinal force and lateral force of the tire, respectively, where the subscripts i = fl, fr, rl, rr represent the front left wheel, front right wheel, rear left wheel, and rear right wheel, respectively; l a and l b Respectively represent the distance from the center of mass to the front axle and the rear axle, c is the wheelbase of the left and right wheels of the vehicle, and the front and rear wheelbases of the vehicle are consistent, δ is the front wheel turning angle;

[0121] (22) Estimate the vehicle state based on the vehicle dynamics model:

[0122] The sensor is used to measure the vehicle's lateral and longitudinal acceleration and the vehicle's yaw rate information to estimate the vehicle's current state.

[0123] Assume the state quantity of the observation system is x 1 =[β V x ] T , and its nonlinear state equation is as follows:

[0124]

[0125] Assume the state quantity of the observation system is y 1 =a y , and its nonlinear state equation is as follows:

[0126]

[0127] In the formula, f 1 (·) and h 1 (·) are the vehicle lateral nonlinear state equation expressions and the vehicle longitudinal nonlinear state equation expressions, β is the road adhesion coefficient, a x and a y are the lateral acceleration and longitudinal acceleration of the vehicle, r is the wheel radius, w 1 and v 1 are the process noise vector and the measurement noise vector, respectively, both of which are zero-mean Gaussian white noise;

[0128] Convert equation (4) and equation (5) into nonlinear discrete form:

[0129]

[0130] In the formula, x 1,k ,y 1,k To predict the lateral and longitudinal vehicle states at time k, the vehicle dynamics model is used to estimate the position after obtaining it through GPS, x 1,k+1 is the predicted state quantity at the next moment.

[0131] In the process of state estimation, β and Vx The value of is predicted at each discrete time step, and the output value is a y The measured value is compared with that obtained by the vehicle sensor to complete the correction of the state quantity; the road adhesion coefficient of the current road surface and the current state quantity of the vehicle are obtained through linear discretization.

[0132] (3) Constructing a driving safety field model, by inputting the current vehicle state obtained in step (2) into the driving safety field model, constructing a driving safety field that takes chassis dynamics into consideration, using S-curve optimization to define the chassis dynamics boundary, and generating a planning trajectory under the chassis dynamics constraint; specifically, including:

[0133] (31) Construct a driving safety field model considering chassis dynamics;

[0134] (32) Calculate the attraction and repulsion of the driving safety field;

[0135] (33) Generate planning trajectory by optimizing S-curve and adding chassis dynamic constraints;

[0136] (34) Consider the constraints imposed by vehicle chassis dynamics on the planned path to improve vehicle stability during path tracking.

[0137] Specifically, the step (31) comprises:

[0138] (311) The driving safety field considering chassis dynamics is composed of the potential energy field generated by stationary objects or roads, the kinetic energy field determined by moving vehicles, the behavior field determined by the driver's psychology / physiology, and the vehicle-road field caused by different chassis dynamics; the driving safety model expression is as follows:

[0139] E S =E R +E V +E D +E C (7)

[0140] In the formula, E S For driving safety field strength; E R is the potential energy field strength; E V is the kinetic energy field strength; E D is the field strength of the behavior field; E C is the field strength of the vehicle road;

[0141] (312) Potential energy field modeling;

[0142] The potential field is a physical field generated by the impact of objects or road conditions in the traffic environment on driving safety. Its size and direction are determined by the properties of the stationary object itself, its relative position, and road conditions. i ,y i) is expressed by equation (8):

[0143]

[0144] In the formula, (x i ,y i ) is the coordinate of the center of mass of the stationary object i; the x-axis is the longitudinal direction of the road; the y-axis is the vertical direction of the road; E R_ij is the risk degree that the stationary object i will cause to the surrounding environment under certain road conditions. The larger the value, the greater the risk. The direction and r of the potential energy field strength ij Same direction; r ij =(x j -x i ,y j -y i ) is the surrounding environment (x j ,y j ) and the vector distance between the stationary object; k 1 , K is a constant to be determined; M i is the equivalent mass of stationary object i; R i is (x i ,y i ) Road condition influencing factors;

[0145] Equivalent mass M i Determined by the properties of the stationary object i itself, the risk level it poses depends on the type, mass and motion state of the stationary object; the function expression is as follows:

[0146]

[0147] In the formula, m i is the mass of the stationary object; H i is the type of stationary object; a k ,β k is a constant to be determined; v i is the speed of a stationary object;

[0148] (313) Kinetic energy field modeling;

[0149] The kinetic energy field is a physical field generated by the impact of moving objects (mainly surrounding vehicles) on driving safety in the traffic environment. The magnitude and direction of the field depend on the relative mass, position, speed and direction of the surrounding vehicles. i ,y i ) position of a moving object s in its surroundings (x j ,y j ) is formed as follows:

[0150]

[0151] In the formula, E V_sj is the moving object s around (x s ,y s ) forms a kinetic energy field, the direction of the field intensity vector is the same as r sj Same; r sj =(x j -x s ,y j -y s ) is the surrounding environment (x j ,y j ) and the vector distance between the moving object; R s is the road condition influencing factor; M s is the equivalent mass of the moving object s; v s is the speed of the moving object s; θ s is the velocity direction of the moving object s and r sj The angle between G and k 1 ,k 2 is a constant; exp[k 2 ·v s ·cos(θ s )] represents the law of change of the driving risk posed by the moving object s to the approaching vehicle as the moving vehicle approaches in the direction and speed;

[0152] (314) Behavioral field modeling;

[0153] The behavior field is a physical field generated by the driver's physiological / psychological factors affecting driving safety. Its field strength depends on the driver's behavior. n ,y n ) in the surrounding environment (x j ,y j ) formed in the behavioral field E D_nj The expression is as follows:

[0154] E D_nj =E V_nj ·D rn (11)

[0155] In the formula, E V_nj is the vehicle n in the surrounding environment (x j ,y j ) formed in the kinetic energy field, D rn is the risk factor for the driver;

[0156] (315) Vehicle road field modeling;

[0157] The vehicle-road field is a physical field that is affected by the road adhesion coefficient and vehicle speed during driving. The field strength depends on the road adhesion coefficient and vehicle speed.i ,y i ) The vehicle road field expression formed by the vehicle with a vehicle speed v at this position is:

[0158] E D_j =C j ·ln[β j ]·exp[v]+δ k (12)

[0159] Where v is the current speed of the vehicle; β ij is (x i ,y i ) at the road surface adhesion coefficient; C ij is the chassis dynamics factor; δ k is a constant to be determined;

[0160] (316) The driving safety field strength is formed by the superposition of the above-mentioned potential energy field, kinetic energy field, behavior field and vehicle road field.

[0161] Specifically, the step (32) comprises:

[0162] (321) The generation of the trajectory requires that the target point has an attractive force on the ego vehicle, with the direction from the ego vehicle to the target point, and that the obstacle has a repulsive force on the ego vehicle, with the direction from the obstacle to the ego vehicle. The definition of the attractive force field is as follows:

[0163] Q Trgl =a(YY Trgl ) 2 (13)

[0164] Where a is the shape coefficient of the attraction potential field of the target lane, which is used to adjust the amplitude of the attraction potential field; Y is the lateral position of the vehicle; Y Trgl is the lateral position of the centerline of the target lane;

[0165] (322) The driving safety field is the repulsive force potential field of the current working condition, namely:

[0166] Q S =E S (14)

[0167] In the formula, Q S is the repulsive force potential field; E S It is a driving safety area;

[0168] (323) The direction of motion is generated by the negative gradient of the force field function, as shown in Equations (14) and (15):

[0169] F Trql (q)=-kY Trgl (15)

[0170]

[0171] Where p(q) is the distance between the vehicle and the obstacle; p 0 is the influence range, and the field strength beyond this influence range will not affect the generation of repulsive force; k and η are constants; F Trql (q) is the gravitational force generated by the specified target trajectory; F S (q) is the repulsive force exerted by the driving safety field on the vehicle;

[0172] (324) The combined field strength of the gravitational field and the driving safety field is the current total field. The gradient of the current total field is the direction in which the vehicle should move, as shown in equations (16) and (17):

[0173] Q(q)=Q Trgl (q)+Q Rep (q) (17)

[0174] F(q)=-▽Q(q) (18)

[0175] The vehicle position at the next moment is:

[0176] [X 1 ,Y 1 ]=[X 0 ,Y 0 ]+k path F(q)·Δt (19)

[0177] In the formula, [X 1 ,Y 1 ]The vehicle's position at the next moment; [X 0 ,Y 0 ] is the current position of the vehicle; Δt is the step length of trajectory prediction; k path is a constant.

[0178] Specifically, the step (33) comprises:

[0179] (331) The function selected to generate the s-shaped curve of the planning trajectory is as follows:

[0180]

[0181] Where P a , P b and P c P is the relevant parameter for shaping the S-shaped curve. a is the maximum slope, P b is the amplitude coefficient, P c is the central symmetric point and the point of maximum slope;

[0182] (332) After selecting the S-curve parameterization equation, the tangent direction of the S-curve is calculated at each path point according to the direction of the resultant force vector; based on the tangent direction and the path point, the parameter value of the S-curve is calculated; the parameter will define the position of the point on the S-curve; by gradually adjusting the parameter value, transitioning from one path point to the next path point, a smooth S-curve is formed.

[0183] Specifically, the step (34) comprises:

[0184] (341) The stability boundary of the sideslip angle is determined by the saturated tire slip angles of the front and rear tires. The boundary expression of chassis dynamics is:

[0185]

[0186]

[0187] Where, γ represents the sideslip angle boundary; δ f Represents the front wheel angle; α i represents the slip angle corresponding to the peak tire force of the tire, i=fl,fr,rl,rr represent the left front tire, right front tire, left rear tire, and right rear tire respectively; l f , l r Respectively represent the front wheelbase and rear wheelbase;

[0188] (342) According to equation (21), the stability region of the vehicle is determined, which limits the growth of the yaw rate and the sideslip angle. The stability region can be transformed into the vehicle dynamics boundary constraint in equation (23):

[0189]

[0190] Where K is the definition constraint; β ij is (x i ,y i ) at the road surface adhesion coefficient; β cos is the road adhesion coefficient constraint of the planned path considering smoothness and vehicle stability in path planning; v is the vehicle speed; a is the vehicle acceleration; ω is the yaw rate;

[0191] (343) By constraining the stability region, a safe planning trajectory that the vehicle can pass through is generated.

[0192] (4) Drive along the planned trajectory obtained in step (3) above, and plan the trajectory again every Δt seconds based on the current vehicle information and environment information.

[0193] When updating the planned trajectory, if the current trajectory can meet all constraints, the smart car is controlled to continue to drive along the current trajectory. If the current trajectory cannot meet all constraints, the planned trajectory is re-planned and the planned trajectory of the smart car is updated. After the update, the smart car will drive according to the updated trajectory.

[0194] The present invention has many specific application paths. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements can be made without departing from the principle of the present invention. These improvements should also be regarded as the protection scope of the present invention.

Claims

1. A method for intelligent vehicle trajectory planning considering chassis dynamics boundaries, characterized in that the steps are as follows: (1) Real-time collection of vehicle information and surrounding vehicle information required for trajectory planning; (2) Estimate the current vehicle state using the vehicle information obtained in step (1); (3) constructing a driving safety field model, by inputting the current vehicle state obtained in step (2) into the driving safety field model, constructing a driving safety field that takes chassis dynamics into consideration, using S-curve optimization to define the chassis dynamics boundary, and generating a planning trajectory under the chassis dynamics constraint; (4) Driving along the planned trajectory obtained in step (3) above, and planning the trajectory again every Δt seconds based on the current vehicle information and environment information; The step (3) specifically includes: (31) Construct a driving safety field model considering chassis dynamics; (32) Calculate the attraction and repulsion of the driving safety field; (33) Generate planning trajectory by optimizing S-curve and adding chassis dynamic constraints; (34) Consider the constraints imposed by vehicle chassis dynamics on the planned path to improve vehicle stability during path tracking; The step (33) specifically comprises: (331) The function selected to generate the s-shaped curve of the planning trajectory is as follows: Where P a , P b and P c P is the relevant parameter for shaping the S-shaped curve. a is the maximum slope, P b is the amplitude coefficient, P c is the central symmetric point and the point of maximum slope; (332) After selecting the S-curve parameterization equation, the tangent direction of the S-curve is calculated at each path point according to the direction of the resultant force vector; based on the tangent direction and the path point, the parameter value of the S-curve is calculated; the parameter will define the position of the point on the S-curve; by gradually adjusting the parameter value, transitioning from one path point to the next path point, a smooth S-curve is formed.

2. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 1, It is characterized in that The information required for trajectory planning collected in step (1) is the vehicle lateral, longitudinal acceleration and vehicle yaw angular velocity information of the self-vehicle, as well as the state of the first-category stationary objects, the second-category stationary objects, the moving objects, the type of moving objects, the position of the moving objects and the state and position of the smart vehicle required for trajectory planning; the first-category stationary objects refer to objects that will cause great losses in the event of a collision; the second-category stationary objects refer to objects that will not collide with the vehicle but will restrict the driver's driving behavior; moving objects refer to vehicles or pedestrians moving on the road.

3. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 1, It is characterized in that The specific steps of step (2) are as follows: (21) Build vehicle dynamics model: Taking the vehicle body coordinate system as the reference system, the vehicle dynamics model is established: Longitudinal Movement: Lateral Movement: Yaw motion: Where m is the vehicle mass; J Z is the moment of inertia of the vehicle around the z axis; V x is the longitudinal speed; V y is the lateral vehicle speed; is the yaw angular velocity; F xi and F yi are the longitudinal force and lateral force of the tire, respectively, where the subscripts i = fl, fr, rl, rr represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; l a and l b Respectively represent the distance from the center of mass to the front axle and the rear axle, c is the wheelbase of the left and right wheels of the vehicle, and the front and rear wheelbases of the vehicle are consistent, δ is the front wheel turning angle; (22) Estimate the vehicle state based on the vehicle dynamics model: The sensor is used to measure the vehicle's lateral and longitudinal acceleration and the vehicle's yaw rate information to estimate the vehicle's current state. Assume the state quantity of the observation system is x 1 =[β V x ] T , and its nonlinear state equation is as follows: Assume the state quantity of the observation system is y 1 =a y , and its nonlinear state equation is as follows: In the formula, f 1 ( ) and h 1 ( ) are the vehicle lateral nonlinear state equation expressions and the vehicle longitudinal nonlinear state equation expressions, β is the road adhesion coefficient, a x and a y are the lateral acceleration and longitudinal acceleration of the vehicle, r is the wheel radius, w 1 and v 1 are the process noise vector and the measurement noise vector, respectively, both of which are zero-mean Gaussian white noise; Convert equation (4) and equation (5) into nonlinear discrete form: In the formula, x 1,k ,y 1,k To predict the lateral and longitudinal vehicle states at time k, the vehicle dynamics model is used to estimate the position after obtaining it through GPS, x 1,k+1 is the predicted state quantity at the next moment.

4. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 3, It is characterized in that In the process of state estimation, β and V x The value of is predicted at each discrete time step, and the output value is a y The measured value is compared with that obtained by the vehicle sensor to complete the correction of the state quantity; the road adhesion coefficient of the current road surface and the current state quantity of the vehicle are obtained through linear discretization.

5. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 1, It is characterized in that The step (31) specifically comprises: (311) The driving safety field considering chassis dynamics is composed of the potential energy field generated by stationary objects or roads, the kinetic energy field determined by moving vehicles, the behavior field determined by the driver's psychology / physiology, and the vehicle-road field caused by different chassis dynamics; the driving safety model expression is as follows: AND S =And R +E V +E D +E C (7) In the formula, E S For driving safety field strength; E R is the potential energy field strength; E V is the kinetic energy field strength; E D is the behavioral field strength; E C is the field strength of the vehicle road; (312) Potential energy field modeling; The potential field is a physical field generated by the impact of objects or road conditions in the traffic environment on driving safety. Its size and direction are determined by the properties of the stationary object itself, its relative position, and road conditions. i ,y i ) is expressed by equation (8): In the formula, (x i ,y i ) is the coordinate of the center of mass of the stationary object i; the x-axis is the longitudinal direction of the road; the y-axis is the vertical direction of the road; E R_ij is the risk degree that the stationary object i will cause to the surrounding environment under certain road conditions. The larger the value, the greater the risk. The direction and r of the potential energy field strength ij Same direction; r ij =(x j -x i ,y j -y i ) is the surrounding environment (x j ,y j ) and the vector distance between the stationary object; k 1 , K is a constant to be determined; M i is the equivalent mass of stationary object i; R i is (x i ,y i ) Road condition influencing factors; Equivalent mass M i Determined by the properties of the stationary object i itself, the risk level it poses depends on the type, mass and motion state of the stationary object; the function expression is as follows: In the formula, m i is the mass of the stationary object; H i is the type of stationary object; a k ,β k is a constant to be determined; v i is the speed of a stationary object; (313) Kinetic energy field modeling; The kinetic energy field is a physical field generated by the impact of moving objects on driving safety in the traffic environment. The magnitude and direction of the field depend on the relative mass, position, speed and direction of the surrounding vehicles. i ,y i ) position of a moving object s in its surroundings (x j ,y j ) is formed as follows: In the formula, E V_sj is the moving object s around (x s ,y s ) forms a kinetic energy field, the direction of the field intensity vector is the same as r sj same; r sj =(x j -x s ,y j -y s ) is the surrounding environment (x j ,y j ) and the vector distance between the moving object; R s is the road condition influencing factor; M s is the equivalent mass of the moving object s; v s is the speed of the moving object s; θ s is the velocity direction of the moving object s and r sj The angle between G and k 1 ,k 2 is a constant; exp[k 2 ·v s ·cos(θ s )] represents the law of change of the driving risk posed by the moving object s to the approaching vehicle as the moving vehicle approaches in the direction and speed; (314) Behavioral field modeling; The behavior field is a physical field generated by the driver's physiological / psychological factors affecting driving safety. Its field strength depends on the driver's behavior. n ,y n ) in the surrounding environment (x j ,y j ) formed in the behavioral field E D_nj The expression is as follows: AND D_nj =And V_nj ·D rn (11) In the formula, E V_nj is the vehicle n in the surrounding environment (x j ,y j ) formed in the kinetic energy field, D rn is the risk factor for the driver; (315) Vehicle road field modeling; The vehicle-road field is a physical field that is affected by the road adhesion coefficient and vehicle speed during driving. The field strength depends on the road adhesion coefficient and vehicle speed. i ,y i ) The vehicle road field expression formed by the vehicle with a vehicle speed v at this position is: E D_j =C j ·ln[β ij ]·exp[v]+δ k (12) Where v is the current speed of the vehicle; β ij is (x i ,y i ) at the road surface adhesion coefficient; C j is the chassis dynamics factor; δ k is a constant to be determined; (316) The driving safety field strength is formed by the superposition of the above-mentioned potential energy field, kinetic energy field, behavior field and vehicle road field.

6. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 5, It is characterized in that The step (32) specifically comprises: (321) The generation of the trajectory requires that the target point has an attractive force on the ego vehicle, with the direction from the ego vehicle to the target point, and that the obstacle has a repulsive force on the ego vehicle, with the direction from the obstacle to the ego vehicle. The definition of the attractive force field is as follows: Q Trgl =a(YY Trgl ) 2 (13) Where a is the shape coefficient of the attraction potential field of the target lane, which is used to adjust the amplitude of the attraction potential field; Y is the lateral position of the vehicle; Y Trgl is the lateral position of the centerline of the target lane; (322) The driving safety field is the repulsive force potential field of the current working condition, namely: Q S =E S (14) In the formula, Q S is the repulsive force potential field; E S It is a driving safety area; (323) The direction of motion is generated by the negative gradient of the force field function, as shown in Equations (14) and (15): F Trql (q)=-kY Trgl (15) Where p(q) is the distance between the vehicle and the obstacle; p 0 is the influence range, and the field strength beyond this influence range will not affect the generation of repulsive force; k and η are constants; F Trql (q) is the gravitational force generated by the specified target trajectory; F S (q) is the repulsive force exerted on the vehicle by the driving safety field; (324) The combined field strength of the gravitational field and the driving safety field is the current total field. The gradient of the current total field is the direction in which the vehicle should move, as shown in equations (16) and (17): Q(q)=Q Trgl (q)+Q s (q) (17) The vehicle position at the next moment is: [X 1 ,Y 1 ]=[X 0 ,Y 0 ]+k path F(q)·Δt (19) In the formula, [X 1 ,Y 1 ]The vehicle's position at the next moment; [X 0 ,Y 0 ] is the current position of the vehicle; Δt is the step length of trajectory prediction; k path is a constant.

7. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 1, It is characterized in that The step (34) specifically comprises: (341) The stability boundary of the sideslip angle is determined by the saturated tire slip angles of the front and rear tires. The boundary expression of chassis dynamics is: Where, γ represents the sideslip angle boundary; δ f Represents the front wheel angle; α i represents the slip angle corresponding to the peak tire force of the tire, i=fl,fr,rl,rr represent the left front tire, right front tire, left rear tire, and right rear tire respectively; l f , l r Respectively represent the front wheelbase and rear wheelbase; (342) According to equation (21), the stability region of the vehicle is determined, which limits the growth of the yaw rate and the sideslip angle. The stability region can be transformed into the vehicle dynamics boundary constraint in equation (23): Where K is the definition constraint; β ij is (x i ,y i ) at the road surface adhesion coefficient; β cos is the road adhesion coefficient constraint of the planned path considering smoothness and vehicle stability in path planning; v is the vehicle speed; a is the vehicle acceleration; ω is the yaw angular velocity; (343) By constraining the stability region, a safe planning trajectory that the vehicle can pass through is generated.

8. The intelligent vehicle trajectory planning method considering chassis dynamics boundaries according to claim 1, It is characterized in that In the step (4), when the planned trajectory is updated, if the current trajectory can satisfy all constraints, the smart car is controlled to continue to travel along the current trajectory; if the current trajectory cannot satisfy all constraints, the planned trajectory is replanned and the planned trajectory of the smart car is updated, and the smart car will travel along the updated trajectory.

Citation Information

Patent Citations

  • A robot path planning method and system based on improved artificial potential field method

    CN105629974B

  • Method and device for adjusting vehicle based on road surface tire adhesion coefficient

    CN116443022A