Dynamic instability risk assessment method for driving track of intelligent driving vehicle

Abstract

The invention relates to the field of risk assessment of intelligent driving vehicles, in particular to a dynamic instability risk assessment method for a driving track of an intelligent driving vehicle, a dynamic instability risk assessment system for the driving track of the intelligent driving vehicle, computer equipment and a storage medium. And the future vehicle trajectory information given by the trajectory planning and predicting module is used as input, and the front wheel turning angle and the vehicle speed required by the vehicle tracking trajectory are obtained. And then the model is used as the input of a nonlinear two-degree-of-freedom vehicle dynamic model based on a UniTile-Ctrl tire model to describe the nonlinear dynamic response of the vehicle under the complex working condition, and the prediction of the dynamic state of the vehicle is realized. On the basis, a vehicle stability evaluation method based on a phase plane is introduced, and evaluation and pre-judgment of the dynamic instability risk of the vehicle driving track are achieved. The method is of great significance in solving the dynamic instability risk evaluation of the driving track of the intelligent driving vehicle under the complex working condition.

Description

Technical Field

[0001] The present invention relates to the field of risk assessment of intelligent driving vehicles, and in particular to a method, system, computer equipment and storage medium for dynamic instability risk assessment of a driving trajectory of an intelligent driving vehicle. Background Art

[0002] The dynamic instability risk assessment of the driving trajectory of intelligent driving vehicles is an important research field in the autonomous driving system, which is directly related to the driving safety and stability of the vehicle. For intelligent driving vehicles, the assessment of the dynamic instability risk of the driving trajectory requires not only accurate prediction of the trajectory, but also further analysis of the potential dynamic instability risk of the vehicle in combination with the dynamic response model of the vehicle, which places higher requirements on the accuracy of the vehicle dynamics model and the real-time performance of the algorithm.

[0003] At present, the dynamic instability risk assessment of the driving trajectory of intelligent driving vehicles has made a lot of progress, but it still faces problems such as insufficient prediction accuracy in complex working conditions and poor real-time performance. Specifically, most of the dynamic instability assessment methods for traditional vehicle driving trajectories use a two-degree-of-freedom single-track or double-track vehicle dynamics model based on a linear tire model. Although these methods can describe the stability of vehicles under conventional working conditions such as low speed, high adhesion, and small lateral acceleration, they ignore the nonlinear and dynamic coupling characteristics of vehicles in complex working conditions such as high-speed cornering, emergency obstacle avoidance, and rainy and snowy roads, and cannot accurately assess the dynamic instability risk of vehicles under complex working conditions. High-fidelity nonlinear tire models based on magic formula models, UniTire models, LuGre models, etc. are limited in practical applications due to their complex models and numerous parameters, which make it difficult to meet the real-time requirements of the controller. In addition, considering that vehicles are prone to unexpected dynamic instability such as skidding, tail swinging, and rollover under complex driving conditions such as high-speed cornering, emergency obstacle avoidance, and rainy and snowy roads, at this time, accompanied by significant changes in the vehicle body posture, the actual motion trajectory of the vehicle will deviate from the reference trajectory given by the motion planning layer. The traditional dynamic instability risk assessment method of the vehicle's driving trajectory based on road curvature or planned trajectory curvature can no longer accurately reflect the dynamic instability risk of the vehicle's actual driving trajectory. Therefore, for autonomous driving vehicles, how to effectively balance the contradiction between model accuracy and algorithm real-time performance, establish a control-oriented vehicle dynamics model and tire model, and accurately assess the dynamic instability risk of the vehicle's driving trajectory in complex environments has become one of the key challenges to ensure the safety of autonomous driving. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method, system, computer device and storage medium for evaluating the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle. The method for evaluating the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle described in the present invention comprises the following steps:

[0005] (1) The vehicle trajectory prediction information for the next 3-5 seconds output by the intelligent driving vehicle trajectory prediction module is used as the input of the vehicle dynamics state prediction model, denoted as: ,in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t It is the relative time of each predicted trajectory point in the next 3-5 seconds.

[0006] (2) The longitudinal speed of the vehicle as it travels along the predicted trajectory on the road V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second derivative Equal, that is: The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, and the relationship is expressed as: δ= atan( L / R ) = atan( Lk );in, d is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k。

[0007] (3) Establish a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model.

[0008] First, the dynamic equation based on the two-degree-of-freedom vehicle dynamics model is as follows:

[0009] ,in, β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l rare the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi =f ( α i , S xi , F zi , m i ), i = f , r Calculated.

[0010] m i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: .

[0011] Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 .

[0012] Secondly, a control-oriented composite working condition UniTire-Ctrl tire model is established. The longitudinal slip rate in the composite working condition UniTire-Ctrl tire model is S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. oh is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle.

[0013] Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: Sy = tan α ·( S x -1).

[0014] Relative longitudinal slip rate f x , relative lateral slip rate f y and the comprehensive slip ratio f They are defined as: ,in, K x and K y are the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, m x and m y are the longitudinal and lateral friction coefficients, respectively.

[0015] For longitudinal sliding stiffness K x , ignoring the influence of the longitudinal elasticity of the tire, the longitudinal sliding stiffness of the tire is approximately linearly related to the load, that is: K x =F z · l 1.

[0016] For cornering stiffness K y , the tire cornering stiffness shows nonlinear characteristics with load, and the relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 are the tire model related parameters obtained by identifying the experimental data, l The value range of 1 is 20-100, s 1. s 2. s The value range of 3 is 0.001-0.1.

[0017] The simplified longitudinal and lateral friction coefficient is uniformly expressed as: ,in, m 0 is the initial friction coefficient, a μ A coefficient that controls the degree of reduction in the friction coefficient.

[0018] Introducing Direction Factor l To describe the direction of the total tangential force, the expression is: ,in, f c and n are the model-related identification parameters.

[0019] Introducing Direction Factor l After that, the new relative longitudinal slip rate is obtained. f xn , the new relative lateral slip rate f yn and the new combined slip ratio f n : .

[0020] Introducing Direction Factor l After that, the total tangential force of the tire , longitudinal force F x and lateral force F y Respectively expressed as: ,in, E is the curvature factor.

[0021] Combining the above formula, the expression of lateral force can be written as: .

[0022] Will S y = tan α ·( S x -1) Substitute into the above formula and write the lateral force as: F y =K ye α ; In the formula K ye is the effective cornering stiffness, and its specific expression is: At this point, the construction of the control-oriented composite working condition UniTire-Ctrl tire model has been completed.

[0023] Finally, the nonlinear two-degree-of-freedom vehicle dynamics model is obtained by substituting it into the two-degree-of-freedom vehicle dynamics model. In order to ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the parameters of the UniTire-Ctrl tire model under composite working conditions using tire experimental data. At the same time, the actual vehicle model verification data is used to identify the vehicle model parameters, and the parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result.

[0024] (4) Based on the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds, the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model is used to calculate the front wheel turning angle required to track the predicted trajectory. d and longitudinal speed V x As model input, calculate the lateral forces on the front and rear axle tires F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r , predicting the vehicle dynamics state that characterizes the lateral and yaw stability of the vehicle.

[0025] As a preferred method, the front and rear wheel slip angles are used. α f —α r Phase plane vehicle stability evaluation index, quantitative evaluation of vehicle stability margin: According to the front and rear wheel side slip angles mentioned above α f 、α r The calculation formula is based on the analysis of the two variables in α f —α r The position in the phase plane is used to judge and measure the vehicle body stability of the driving trajectory; α f —α r Phase plane diagram, with the origin as the center and radius as R The circle with a value of 1 represents the stable region, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat are the saturated sideslip angles of the front and rear wheels respectively; with the origin as the center, The circle with a radius of is the critical stability boundary; with the origin as the center and the radius of The circle is the limit instability boundary, where ( α f,s , α r,s ) represents the saddle point coordinates of the front and rear wheel sideslip angles that are closer to the center of the circle.

[0026] when R 3> R 2: Vehicle stability index x writing: .

[0027] when R 3≤ R 2: Vehicle stability index x writing: .

[0028] x Describe the stability of the vehicle. x ≥0, the vehicle is in the stable zone, and when -0.5≤ x When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; x When <-0.5, the vehicle is in an unstable zone.

[0029] The present invention also provides a dynamic instability risk assessment system for the driving trajectory of an intelligent driving vehicle, comprising: a trajectory prediction module, a vehicle dynamic state prediction module, a vehicle instability risk prejudgment module, and a trajectory dynamic instability evaluation module.

[0030] The trajectory prediction module outputs the vehicle's trajectory prediction information in the next 3-5 seconds, recorded as: ,in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t It is the relative time of each predicted trajectory point in the next 3-5 seconds.

[0031] The vehicle dynamics state prediction module includes a vehicle kinematics model and a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model.

[0032] The vehicle kinematics model performs the following operations: the longitudinal speed of the vehicle as it travels along the predicted trajectory on the road V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second derivative Equal, that is: .

[0033] The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, and the relationship is expressed as: δ= atan( L / R ) = atan( Lk );in, d is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k .

[0034] The method for constructing a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model is as follows:

[0035] The dynamic equation based on the two-degree-of-freedom vehicle dynamics model is: ,in, β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l r are the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi = f ( α i , S xi , F zi , m i ), i = f , r Calculated.

[0036] m i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: .

[0037] Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 .

[0038] Longitudinal slip ratio in UniTire-Ctrl tire model for control-oriented composite conditions S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. oh is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle.

[0039] Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: S y = tan α ·( S x -1).

[0040] Relative longitudinal slip rate f x , relative lateral slip rate f y and the comprehensive slip ratio f They are defined as: ,in, K x and K y are the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, m x and m y are the longitudinal and lateral friction coefficients, respectively.

[0041] For longitudinal sliding stiffness K x , the tire longitudinal slip stiffness is approximately linear with the load, and the relationship is expressed as: K x=F z · l 1.

[0042] For cornering stiffness K y , the tire cornering stiffness shows nonlinear characteristics with load, and the relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 are the tire model related parameters obtained by identifying the experimental data.

[0043] The simplified longitudinal and lateral friction coefficient is uniformly expressed as: .

[0044] Introducing Direction Factor l To describe the direction of the total tangential force, the expression is: ,in, f c and n are the model-related identification parameters.

[0045] Introducing Direction Factor l After that, the new relative longitudinal slip rate is obtained. f xn , the new relative lateral slip rate f yn and the new combined slip ratio f n : .

[0046] Introducing Direction Factor l After that, the total tangential force of the tire , longitudinal force F x and lateral force F y Respectively expressed as: ,in, E is the curvature factor.

[0047] The expression for the lateral force is written as: ,Will S y = tan α ·( S x -1) and write the lateral force as: F y =K yeα， In the formula K ye is the effective cornering stiffness, expressed as: .

[0048] After the control-oriented composite working condition UniTire-Ctrl tire model is built, it is substituted into the two-degree-of-freedom vehicle dynamics model to obtain the nonlinear two-degree-of-freedom vehicle dynamics model. In order to ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the parameters of the composite working condition UniTire-Ctrl tire model using tire experimental data. At the same time, the actual vehicle model verification data is used to identify the vehicle model parameters, and the parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result.

[0049] The vehicle instability risk prediction module uses the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds output by the trajectory prediction module, and uses the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model in the vehicle dynamics state prediction module to track the front wheel turning angle required for the predicted trajectory. d and longitudinal speed V x As model input, calculate the lateral forces on the front and rear axle tires F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r , predicting the vehicle dynamics state that characterizes the lateral and yaw stability of the vehicle.

[0050] The trajectory dynamics instability evaluation module adopts the front and rear wheel side slip angle α f —α r Phase plane vehicle stability evaluation index, quantitative evaluation of vehicle stability margin: According to the front and rear wheel side slip angles mentioned above α f 、α r The calculation formula is based on the analysis of the two variables in α f —α r The vehicle body stability of the driving trajectory is judged and measured by its position in the phase plane.

[0051] according to αf —α r Phase plane diagram, with the origin as the center and radius as R The circle with a value of 1 represents the stable region, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat are the saturated sideslip angles of the front and rear wheels respectively; with the origin as the center, The circle with a radius of is the critical stability boundary; with the origin as the center and the radius of The circle is the limit instability boundary, where ( α f,s , α r,s ) represents the saddle point coordinates of the front and rear wheel sideslip angles that are closer to the center of the circle.

[0052] when R 3> R 2: Vehicle stability index x writing: .

[0053] when R 3≤ R 2: Vehicle stability index x writing: .

[0054] x Describe the stability of the vehicle. x ≥0, the vehicle is in the stable zone, and when -0.5≤ x When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; x When <-0.5, the vehicle is in an unstable zone.

[0055] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle as described above is implemented.

[0056] The present invention also provides a storage medium containing computer executable instructions, and the computer executable instructions are executed by a computer processor to implement the above-mentioned method for assessing the risk of dynamic instability of the driving trajectory of an intelligent driving vehicle.

[0057] Beneficial effects of the present invention:

[0058] The present invention is based on the vehicle Ackerman steering kinematic model, and converts the trajectory information of the next 3-5 seconds output by the trajectory planning module or the trajectory prediction module of the intelligent driving vehicle into the kinematic information such as the vehicle's turning angle and speed; on this basis, a nonlinear two-degree-of-freedom vehicle dynamics model based on the control-oriented composite working condition UniTire-Ctrl tire model is established, and the vehicle dynamics state is predicted according to the vehicle trajectory prediction information, so as to realize the effective deduction from the vehicle trajectory information to the vehicle dynamics state; in addition, the front and rear wheel sideslip angle phase planes are introduced to evaluate and predict the dynamic instability risk of the vehicle's driving trajectory. The method for evaluating the dynamic instability risk of the driving trajectory of the intelligent driving vehicle proposed in the present invention introduces a nonlinear two-degree-of-freedom vehicle dynamics model based on the tire model of the composite working condition UniTire-Ctrl, which can accurately describe the nonlinear and dynamic coupling characteristics of the vehicle under complex working conditions such as high-speed cornering, emergency obstacle avoidance, rainy and snowy roads, and improve the accuracy of the vehicle dynamic instability risk evaluation; at the same time, the control-oriented composite working condition UniTire-Ctrl has the advantages of few model parameters, high precision, concise expression, and easy controller use, which can effectively coordinate the contradiction between model accuracy and algorithm real-time performance, and better meet the requirements of the vehicle controller for real-time performance; in addition, the proposed method for evaluating the dynamic instability risk of the driving trajectory of the intelligent driving vehicle under complex working conditions can quantitatively evaluate and predict whether the candidate trajectory meets the vehicle dynamic stability constraint based on the future trajectory prediction information as input, and judge whether the vehicle will be dynamically unstable when the intelligent driving vehicle tracks the predicted trajectory, avoid risks in advance, and ensure the stable and safe operation of the intelligent driving vehicle. The present invention is of great significance for solving the dynamic instability risk evaluation of the driving trajectory of the intelligent driving vehicle under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 Schematic diagram of the overall flow of the evaluation method of the present invention.

[0060] Figure 2 Schematic diagram of the Ackerman steering principle.

[0061] Figure 3 The present invention is based on α f —α r Schematic diagram of the stability boundary demarcation of the phase plane.

[0062] Figure 4 It is a schematic diagram of vehicle status and stability indicators of the present invention. DETAILED DESCRIPTION

[0063] like Figure 1 As shown, the present invention provides a method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle, comprising the following steps:

[0064] (1) The vehicle trajectory prediction information for the next 3-5 seconds output by the intelligent driving vehicle trajectory prediction module is used as the input of the vehicle dynamics state prediction model, denoted as: ;in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t It is the relative time of each predicted trajectory point in the next 3-5 seconds.

[0065] (2) When a vehicle is traveling along a predicted trajectory on the road, the longitudinal speed of the vehicle is usually considered to be V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second derivative Equal, that is: .

[0066] The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, such as Figure 2 As shown, the relationship is expressed as: δ= atan( L / R ) = atan( Lk );in, d is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k .

[0067] (3) Establish a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model and track the front wheel turning angle required for the predicted trajectory d and longitudinal speed V x As model input, the lateral acceleration, center of mass sideslip angle and yaw rate that characterize the lateral and yaw stability of the vehicle are calculated to achieve vehicle state prediction based on the predicted trajectory information.

[0068] Specifically, the dynamic equation based on the two-degree-of-freedom vehicle dynamics model is as follows:

[0069] ,in,β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l r are the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi =f ( α i , S xi , F zi , m i ), i = f , r Calculated.

[0070] m i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: .

[0071] Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 .

[0072] At present, in most theoretical research and engineering applications, in order to reduce the complexity of the controller and ensure the real-time performance of the algorithm, a pure side slip tire model (linear tire model) is usually used to simplify the controller design; that is, it is usually assumed that within a certain range, the tire lateral force and the side slip angle are linearly related. This type of method has a good performance in lateral acceleration a. y <0.4 g Good control effects can be achieved in normal operating conditions. However, when the vehicle passes through some large curvature curves at a higher speed, the lateral acceleration will increase significantly. Once the lateral acceleration a y >0.4 g, when the vehicle enters the nonlinear region, the nonlinear and longitudinal and lateral coupling characteristics of the tire will become particularly significant. At this time, the traditional linear tire model cannot accurately describe the tire force, and there is a large model error, which affects the control effect.

[0073] In addition, considering emergency conditions such as accelerating to change lanes for overtaking, decelerating to change lanes for obstacle avoidance, etc., the vehicle will enter a compound slip condition of combined steering and drive / braking control. At this time, the coupling characteristics of the longitudinal and lateral tire forces are obvious, which further accelerates the tire forces to approach the nonlinear region. At this time, the pure working condition tire model cannot accurately describe the nonlinear coupling characteristics of the tire forces, which will directly affect the final control effect. Therefore, facing the controller design requirements, establishing a concise and accurate tire model to describe the tire characteristics of pure working conditions and compound working conditions is of great significance for improving the control effect of the model-based control algorithm under complex working conditions and improving the control performance of the whole vehicle. To this end, the present invention establishes a control-oriented compound working condition UniTire-Ctrl tire model, and on this basis, realizes the prediction of the dynamic state of intelligent driving vehicles under complex working conditions.

[0074] The longitudinal slip rate in the UniTire-Ctrl tire model under the composite working condition S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. oh is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle.

[0075] Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: S y = tan α ·( S x -1).

[0076] Relative longitudinal slip rate f x , relative lateral slip rate f y and the comprehensive slip ratio f They are defined as: ,in, K x and K yare the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, m x and m y are the longitudinal and lateral friction coefficients, respectively.

[0077] Changes in vertical load will directly affect the longitudinal slip stiffness of the tire. K x and cornering stiffness K y , for the longitudinal sliding stiffness K x , the influence of the longitudinal elasticity of the tire can be ignored. Therefore, for the sake of simplicity, it is assumed that the longitudinal sliding stiffness of the tire is approximately linearly related to the load, that is: K x =F z · l 1.

[0078] For the cornering stiffness K y For tires, the influence of carcass elasticity cannot be ignored, and the tire cornering stiffness shows nonlinear characteristics with load, so the simplified relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 are tire model related parameters obtained by identifying the experimental data, l The value range of 1 is 20-100, s 1. s 2. s The value range of 3 is 0.001-0.1.

[0079] Compared with the traditional UniTire tire model, the present invention selects a more concise expression to describe the dynamic friction characteristics of the tire, ignoring the influence of the vertical load on the friction coefficient, and assuming that the longitudinal and lateral friction coefficients are approximately equal. Therefore, the simplified longitudinal and lateral friction coefficients are uniformly expressed as: ;in, m 0 is the initial friction coefficient, or the peak friction coefficient, a μ A coefficient that controls the degree of reduction in the friction coefficient.

[0080] In order to further improve the accuracy of the model under composite conditions, the directional factor is introduced. lTo describe the direction of the total tangential force. In the extreme combined working conditions of steering and braking combined control, accompanied by severe load transfer, the directional factor will change significantly, directly affecting the force on the tire. The specific expression is: ,in, f c and n are the model-related identification parameters.

[0081] Introducing Direction Factor l After that, the new relative longitudinal, lateral and comprehensive slip rates are obtained: .

[0082] Therefore, the introduction of the directional factor l After that, the total tangential force of the tire , longitudinal force F x and lateral force F y Respectively expressed as: ,in, E is the curvature factor, which is used to adjust the curvature of the tire total tangential force curve in the transition zone. In the simplified model, the differences in the longitudinal and lateral mechanical properties of the tire are ignored, and the longitudinal and lateral curvature factors are equal and equal to the curvature factor of the total tangential force.

[0083] In order to further simplify the design of the controller, the expression of the lateral force is written as: .

[0084] Will S y = tan α ·( S x -1) Substitute into the above formula and assume that when the sideslip angle is small enough, tan α ≈ α , the lateral force can be rewritten as: F y =K ye α， In the formula K ye is the effective cornering stiffness, which is different from the normal cornering stiffness. K ye It can express the tire's cornering stiffness in the linear and nonlinear regions, as well as the influence of the longitudinal slip rate on the tire's cornering characteristics. The specific expression is: .

[0085] At this point, the construction of the control-oriented composite working condition UniTire-Ctrl tire model has been completed. Substituting it into the two-degree-of-freedom vehicle dynamics model, a nonlinear two-degree-of-freedom vehicle dynamics model is obtained.

[0086] In order to ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the parameters of the UniTire-Ctrl tire model under composite working conditions using tire experimental data. At the same time, the vehicle model parameters are identified using the actual vehicle verification data. The parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result.

[0087] (4) Based on the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds, the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model is used to realize the lateral force of the front and rear axle tires. F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r Prediction of vehicle dynamics states characterizing the lateral and yaw stability of the vehicle.

[0088] This embodiment adopts the front and rear wheel side slip angle α f —α r The vehicle stability evaluation index of the phase plane quantitatively evaluates the vehicle stability margin. According to the front and rear wheel side slip angles given above α f 、α r The calculation formula is based on the analysis of the two variables in α f — α r The vehicle body stability of the driving trajectory is judged and measured by its position in the phase plane. α f —α r Phase plane Figure 3 As shown, the green circular area represents the stable area, which is centered at the origin and has a radius of R 1 circle, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat They are the saturated side slip angles of the front and rear wheels respectively; the blue area is centered at the origin. The circle is a circle with a radius of . In this area, the rear wheels of the vehicle enter the descending area. Therefore, when the vehicle is in this area, there is a risk of slipping or even instability. This boundary is called the critical stability boundary. The red area is centered at the origin and has a radius of The circle, the point ( α f,s , α r,s ) represents a saddle point closer to the origin. When the state is in the red area, the vehicle has become unstable. This boundary is called the limit instability boundary.

[0089] Based on the above analysis, define the vehicle stability index ξ:

[0090] when R 3> R 2: Vehicle stability index x writing: .

[0091] when R 3≤ R 2: Vehicle stability index x writing: .

[0092] like Figure 4 As shown, x A more detailed description of vehicle stability is x ≥0, the vehicle is in the stable zone, and when -0.5≤ x When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; x When <-0.5, the vehicle is in an unstable zone.

[0093] This embodiment also provides a dynamic instability risk assessment system for the driving trajectory of an intelligent driving vehicle, including: a trajectory prediction module, a vehicle dynamic state prediction module, a vehicle instability risk prediction module, and a trajectory dynamic instability evaluation module.

[0094] The trajectory prediction module outputs the vehicle's trajectory prediction information in the next 3-5 seconds, recorded as: ,in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t It is the relative time of each predicted trajectory point in the next 3-5 seconds.

[0095] The vehicle dynamics state prediction module includes a vehicle kinematics model and a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model.

[0096] The vehicle kinematics model performs the following operations: the longitudinal speed of the vehicle as it travels along the predicted trajectory on the road V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second derivative Equal, that is: .

[0097] The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, and the relationship is expressed as: δ= atan( L / R ) = atan( Lk );in, d is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k .

[0098] The method for constructing a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model is as follows:

[0099] The dynamic equation based on the two-degree-of-freedom vehicle dynamics model is: ,in, β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l r are the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi = f ( α i , S xi , Fzi , m i ), i = f , r Calculated.

[0100] m i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: .

[0101] Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 .

[0102] Longitudinal slip ratio in UniTire-Ctrl tire model for control-oriented composite conditions S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. oh is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle.

[0103] Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: S y = tan α ·( S x -1).

[0104] Relative longitudinal slip rate f x , relative lateral slip rate f y and the comprehensive slip ratio f They are defined as: ,in, K x andK y are the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, m x and m y are the longitudinal and lateral friction coefficients, respectively.

[0105] For longitudinal sliding stiffness K x , the tire longitudinal slip stiffness is approximately linear with the load, and the relationship is expressed as: K x =F z · l 1.

[0106] For cornering stiffness K y , the tire cornering stiffness shows nonlinear characteristics with load, and the relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 are the tire model related parameters obtained by identifying the experimental data.

[0107] The simplified longitudinal and lateral friction coefficient is uniformly expressed as: .

[0108] Introducing Direction Factor l To describe the direction of the total tangential force, the expression is: ,in, f c and n are the model-related identification parameters.

[0109] Introducing Direction Factor l After that, the new relative longitudinal slip rate is obtained. f xn , the new relative lateral slip rate f yn and the new combined slip ratio f n : .

[0110] Introducing Direction Factor l After that, the total tangential force of the tire , longitudinal force F x and lateral force F yRespectively expressed as: ,in, E is the curvature factor.

[0111] The expression for the lateral force is written as: ,Will S y = tan α ·( S x -1) and write the lateral force as: F y =K ye α， In the formula K ye is the effective cornering stiffness, expressed as: .

[0112] After the control-oriented composite working condition UniTire-Ctrl tire model is built, it is substituted into the two-degree-of-freedom vehicle dynamics model to obtain the nonlinear two-degree-of-freedom vehicle dynamics model. In order to ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the parameters of the composite working condition UniTire-Ctrl tire model using tire experimental data. At the same time, the actual vehicle model verification data is used to identify the vehicle model parameters, and the parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result.

[0113] The vehicle instability risk prediction module uses the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds output by the trajectory prediction module, and uses the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model in the vehicle dynamics state prediction module to track the front wheel turning angle required for the predicted trajectory. d and longitudinal speed V x As model input, calculate the lateral forces on the front and rear axle tires F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r , predicting the vehicle dynamics state that characterizes the lateral and yaw stability of the vehicle.

[0114] The trajectory dynamics instability evaluation module adopts the front and rear wheel side slip angle α f—α r Phase plane vehicle stability evaluation index, quantitative evaluation of vehicle stability margin: According to the front and rear wheel side slip angles mentioned above α f 、α r The calculation formula is based on the analysis of the two variables in α f —α r The vehicle body stability of the driving trajectory is judged and measured by its position in the phase plane.

[0115] according to α f —α r Phase plane diagram, with the origin as the center and radius as R The circle with a value of 1 represents the stable region, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat are the saturated sideslip angles of the front and rear wheels respectively; with the origin as the center, The circle with a radius of is the critical stability boundary; with the origin as the center and the radius of The circle is the limit instability boundary, where ( α f,s , α r,s ) represents the saddle point coordinates of the front and rear wheel sideslip angles that are closer to the center of the circle.

[0116] when R 3> R 2: Vehicle stability index x writing: .

[0117] when R 3≤ R 2: Vehicle stability index x writing: .

[0118] x Describes the stability of the vehicle. x ≥0, the vehicle is in the stable zone, and when -0.5≤ x When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; x When <-0.5, the vehicle is in an unstable zone.

[0119] This embodiment also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle as described above is implemented.

[0120] This embodiment also provides a storage medium containing computer executable instructions, and the computer executable instructions are executed by a computer processor to implement the above-mentioned method for assessing the risk of dynamic instability of the driving trajectory of an intelligent driving vehicle.

Claims

1. A method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle, characterized in that: The following steps are involved: (1) The vehicle trajectory prediction information for the next 3-5 seconds output by the intelligent driving vehicle trajectory prediction module is used as the input of the vehicle dynamics state prediction model, denoted as: ,in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t is the relative time of each predicted trajectory point in the next 3-5 seconds; (2) The longitudinal speed of the vehicle as it travels along the predicted trajectory on the road V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second derivative Equal, that is: ; The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, and the relationship is expressed as: δ = atan( L / R ) = atan( LqCy ); in, δ is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k ; (3) Establish a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model: First, the dynamic equation based on the two-degree-of-freedom vehicle dynamics model is as follows: ,in, β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l r are the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi =f ( α i , S xi , F zi , μ i ), i = f , r Calculated; μ i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: ； Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 ; Secondly, a control-oriented composite working condition UniTire-Ctrl tire model is established. The longitudinal slip rate in the composite working condition UniTire-Ctrl tire model is S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. ω is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle; Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: S y = tan α ·( S x -1); Relative longitudinal slip rate φ x , relative lateral slip rate φ y and the comprehensive slip ratio φ They are defined as: ,in, K x and K y are the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, μ x and μ y are longitudinal and lateral friction coefficients, respectively; For longitudinal sliding stiffness K x , ignoring the influence of the longitudinal elasticity of the tire, the longitudinal sliding stiffness of the tire is approximately linearly related to the load, that is: K x =F z · l 1; For cornering stiffness K y , the tire cornering stiffness shows nonlinear characteristics with load, and the relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 is the tire model related parameters obtained by identifying the experimental data; The simplified longitudinal and lateral friction coefficient is uniformly expressed as: ,in, μ 0 is the initial friction coefficient, a μ A coefficient that controls the degree of reduction in the friction coefficient; Introducing Direction Factor λ To describe the direction of the total tangential force, the expression is: ,in, φ c and n are the identification parameters related to the model; Introducing Direction Factor λ After that, the new relative longitudinal slip rate is obtained. φ xn , the new relative lateral slip rate φ yn and the new combined slip ratio φ n : ； Introducing Direction Factor λ After that, the total tangential force of the tire , longitudinal force F x and lateral force F y Respectively expressed as: ,in, E is the curvature factor; Combining the above formula, the expression of lateral force can be written as: ； Will S y = tan α ·( S x -1) Substitute into the above formula and write the lateral force as: F y =K ye α ; In the above formula K ye is the effective cornering stiffness, and its specific expression is: ; At this point, the construction of the UniTire-Ctrl tire model for control-oriented composite working conditions has been completed; Finally, the nonlinear two-degree-of-freedom vehicle dynamics model is obtained by substituting it into the two-degree-of-freedom vehicle dynamics model. To ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the parameters of the UniTire-Ctrl tire model under the composite working condition using tire test data. At the same time, the vehicle model parameters are identified using the actual vehicle verification data, and the parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result. (4) Based on the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds, the nonlinear two-degree-of-freedom vehicle dynamics model based on the composite working condition UniTire-Ctrl tire model is used to calculate the front wheel turning angle required to track the predicted trajectory. δ and longitudinal speed V x As model input, calculate the lateral forces on the front and rear axle tires F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r , predicting the vehicle dynamics state that characterizes the lateral and yaw stability of the vehicle.

2. The method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle according to claim 1, characterized in that: The front and rear wheel slip angles α f —α r Phase plane vehicle stability evaluation index, quantitative evaluation of vehicle stability margin: According to the front and rear wheel side slip angles mentioned above α f 、α r The calculation formula is based on the analysis of the two variables in α f —α r The position in the phase plane is used to judge and measure the vehicle body stability of the driving trajectory; according to α f —α r Phase plane diagram, with the origin as the center and radius as R The circle with a value of 1 represents the stable region, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat are the saturated slip angles of the front and rear wheels respectively; With the origin as the center, The circle with a radius of is the critical stability boundary; with the origin as the center and the radius of The circle is the limit instability boundary, where ( α f,s , α r,s ) represents the saddle point coordinates of the front and rear wheel sideslip angles that are closer to the center of the circle; when R 3> R 2: Vehicle stability index ξ writing: ； when R 3≤ R 2: Vehicle stability index ξ writing: ； ξ Describes the stability of the vehicle. ξ ≥0, the vehicle is in the stable zone, and when -0.5≤ ξ When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; ξ When <-0.5, the vehicle is in an unstable zone.

3. A dynamic instability risk assessment system for the driving trajectory of an intelligent driving vehicle, characterized in that: include: Trajectory prediction module, vehicle dynamics state prediction module, vehicle instability risk prediction module, trajectory dynamics instability evaluation module; The trajectory prediction module outputs the vehicle's trajectory prediction information in the next 3-5 seconds, recorded as: ,in, u is the trajectory prediction information set, are the arc length of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the arc length with time, are the lateral displacement of the predicted trajectory in the Frenet coordinate system, the first-order derivative and the second-order derivative of the lateral displacement with time, k To predict the curvature of the trajectory, t is the relative time of each predicted trajectory point in the next 3-5 seconds; The vehicle dynamics state prediction module includes a vehicle kinematics model and a nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model: The vehicle kinematics model execution operation includes: when the vehicle is traveling along the predicted trajectory on the road, the longitudinal speed of the vehicle V x and longitudinal acceleration a x The first-order derivative of arc length with time and the second-order derivative Equal, that is: ; The front wheel turning angle required for the vehicle to track the predicted trajectory is obtained based on the trajectory curvature information through the Ackerman steering geometry relationship, and the relationship is expressed as: δ = atan( L / R ) = atan( LqCy ); in, δ is the front wheel turning angle, L is the vehicle wheelbase, R is the radius of the predicted trajectory, according to the curvature of the predicted trajectory k The calculation is satisfied: R= 1 / k ; The method for constructing the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model is as follows: The dynamic equation based on the two-degree-of-freedom vehicle dynamics model is: ,in, β is the vehicle's sideslip angle, r is the vehicle yaw angular velocity, m is the vehicle mass, I z is the yaw moment of inertia, l f 、l r are the distances from the front and rear axles to the vehicle’s center of mass, F yf 、F yr The lateral forces of the front and rear axle tires are respectively calculated by the UniTire-Ctrl tire model F yi =f ( α i , S xi , F zi , μ i ), i = f , r Calculated; μ i is the road friction coefficient, the front and rear wheel slip angles α f 、α r Respectively expressed as: ； Without considering the influence of longitudinal and lateral acceleration, the static vertical load of the front and rear wheels is F zf and F zr They are: ,in, g is the acceleration due to gravity, g=9.8m / s 2 ; Establish a control-oriented composite working condition UniTire-Ctrl tire model, and the longitudinal slip rate in the composite working condition UniTire-Ctrl tire model S x and lateral slip rate S y Respectively expressed as: ,in, V sx and V sy are the longitudinal slip velocity and the lateral slip velocity, respectively. ω is the tire rotation angular velocity, R e is the effective radius of the tire, V is the moving speed of the tire, α is the tire slip angle; Under composite working conditions, the relationship between the longitudinal and lateral slip rates of tires can be expressed as: S y = tan α ·( S x -1); Relative longitudinal slip rate φ x , relative lateral slip rate φ y and the comprehensive slip ratio φ They are defined as: ,in, K x and K y are the tire longitudinal slip stiffness and lateral stiffness, F z is the vertical load, μ x and μ y are longitudinal and lateral friction coefficients, respectively; For longitudinal sliding stiffness K x , the tire longitudinal slip stiffness is approximately linear with the load, and the relationship is expressed as: K x =F z · l 1; For cornering stiffness K y , the tire cornering stiffness shows nonlinear characteristics with load, and the relationship is expressed as: ,in, , F z0 is the nominal load, l 1. s 1. s 2. s 3 is the tire model related parameters obtained by identifying the experimental data; The simplified longitudinal and lateral friction coefficient is uniformly expressed as: ,in, μ 0 is the initial friction coefficient, a μ A coefficient that controls the degree of reduction in the friction coefficient; Introducing Direction Factor λ To describe the direction of the total tangential force, the expression is: ,in, φ c and n are the identification parameters related to the model; Introducing Direction Factor λ After that, the new relative longitudinal slip rate is obtained. φ xn , the new relative lateral slip rate φ yn and the new combined slip ratio φ n : ； Introducing Direction Factor λ After that, the total tangential force of the tire , longitudinal force F x and lateral force F y Respectively expressed as: ,in, E is the curvature factor; Combining the above formula, the expression of lateral force can be written as: ； Will S y = tan α ·( S x -1) Substitute into the above formula and write the lateral force as: F y =K ye α ; In the above formula K ye is the effective cornering stiffness, and its specific expression is: ; After the control-oriented composite working condition UniTire-Ctrl tire model is built, it is substituted into the two-degree-of-freedom vehicle dynamics model to obtain a nonlinear two-degree-of-freedom vehicle dynamics model. To ensure that the response of the nonlinear two-degree-of-freedom vehicle dynamics model is consistent with the actual vehicle response, the least squares parameter identification method is used to identify the composite working condition UniTire-Ctrl tire model parameters using tire test data. At the same time, the actual vehicle model verification data is used to identify the vehicle model parameters, and the parameter combination that makes the model response most consistent with the actual vehicle response is found as the final result. The vehicle instability risk prediction module uses the predicted trajectory information of the intelligent driving vehicle in the next 3-5 seconds output by the trajectory prediction module, and uses the nonlinear two-degree-of-freedom vehicle dynamics model based on the UniTire-Ctrl tire model in the vehicle dynamics state prediction module to track the front wheel turning angle required for the predicted trajectory. δ and longitudinal speed V x As model input, calculate the lateral forces on the front and rear axle tires F yf 、F yr , vehicle center of mass sideslip angle β , lateral acceleration a y , yaw rate r , front and rear wheel slip angle α f 、α r , predicting the vehicle dynamics states that characterize the lateral and yaw stability of the vehicle; The trajectory dynamics instability evaluation module adopts the front and rear wheel side slip angle α f —α r Phase plane vehicle stability evaluation index, quantitative evaluation of vehicle stability margin: According to the front and rear wheel side slip angles mentioned above α f 、α r The calculation formula is based on the analysis of the two variables in α f —α r The position in the phase plane is used to judge and measure the vehicle body stability of the driving trajectory; according to α f —α r Phase plane diagram, with the origin as the center and radius as R The circle with a value of 1 represents the stable region, where R 1=max(| α f,sat |,| α r,sat |), α f,sat and α r,sat are the saturated sideslip angles of the front and rear wheels respectively; with the origin as the center, The circle with a radius of is the critical stability boundary; with the origin as the center and the radius of The circle is the limit instability boundary, where ( α f,s , α r,s ) represents the saddle point coordinates of the front and rear wheel sideslip angles that are closer to the center of the circle; when R 3> R 2: Vehicle stability index ξ writing: ； when R 3≤ R 2: Vehicle stability index ξ writing: ； ξ Describes the stability of the vehicle. ξ ≥0, the vehicle is in the stable zone, and when -0.5≤ ξ When <0, the vehicle is in the transition zone and there is a risk of vehicle instability; ξ When <-0.5, the vehicle is in an unstable zone.

4. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the method for assessing the risk of dynamic instability of the driving trajectory of an intelligent driving vehicle as described in any one of claims 1-2 is implemented.

5. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are executed by a computer processor to implement a method for assessing the risk of dynamic instability of a driving trajectory of an intelligent driving vehicle as described in any one of claims 1 to 2.

Images