CN117584953A - Automatic driving-oriented horizontal and vertical integrated prediction control method - Google Patents

Abstract

The invention discloses an automatic driving horizontal and vertical integrated prediction control method, which comprises the following steps: acquiring current vehicle state parameters through sensors: predicting the future state of the vehicle according to the vehicle state parameters to obtain longitudinal predicted vehicle speed, lateral predicted vehicle speed, predicted yaw rate and predicted centroid side deviation angle at each moment in the vehicle prediction time; calculating a vehicle prediction stability margin; predicting the states of obstacles around the vehicle, and determining a collision-free risk space; planning a vehicle track in a collision-free risk space to obtain target lateral displacement, target lateral speed, target lateral acceleration, target longitudinal displacement, target longitudinal speed and longitudinal acceleration of the vehicle; if the vehicle is in an understeer state in the future, the lateral acceleration is lifted in the track planning process; the vehicle is in an oversteering state in the future, and lateral acceleration is reduced in the track planning process; and controlling steering wheel rotation angle and distributing torque of the four-wheel motor according to the vehicle track planning result.

Description

Automatic driving-oriented horizontal and vertical integrated prediction control method

Technical Field

The invention belongs to the technical field of automatic driving, and particularly relates to a transverse and longitudinal integrated predictive control method for automatic driving.

Background

With the rapid development of economy and science and technology for decades, the intelligent level of an automobile is greatly improved, the intelligent driving system of the L2 level is widely applied to various automobile types at present, such as a AEB (Autonomous Emergency Braking) system, a millimeter wave radar is utilized to detect the relative distance and the relative speed between the intelligent driving system and an obstacle in front of the automobile, the dangerous degree of the automobile collision is judged, and partial braking or full braking is adopted to the automobile, so that the collision is avoided or the severity of the collision is reduced; the NOA (Navigate on Autopilot) system is further upgraded on the basis of the traditional ACC (Adaptive Cruise Control) system, and functions of autonomous lane changing overtaking, lane insertion and the like can be realized. However, when a dangerous situation or an extreme working condition is faced, the current intelligent driving system needs to be taken over by a driver, and the level of intelligence still needs to be further improved.

Disclosure of Invention

The invention aims to provide an automatic driving horizontal and vertical integrated prediction control method, which constructs a vehicle driving safety space together by comprehensively predicting the stable state of a vehicle and the collision risk with an obstacle, and performs track planning and vehicle motion control in the safety space by considering the constraint of the stable state of the vehicle, so that the running safety and riding comfort of an intelligent vehicle can be improved.

The technical scheme provided by the invention is as follows:

an automatic driving horizontal and vertical integrated prediction control method comprises the following steps:

acquiring current vehicle state parameters through sensors: longitudinal displacement X, lateral displacement Y, longitudinal vehicle speed v _x Lateral speed v of vehicle _y Longitudinal acceleration a _x Lateral acceleration a _y Heading angle θ, steering wheel angle δ _sw The method comprises the steps of carrying out a first treatment on the surface of the Predicting the future state of the vehicle according to the current vehicle state parameters to obtain longitudinal predicted vehicle speed, lateral predicted vehicle speed, predicted yaw rate and predicted centroid side deviation angle at each moment in the vehicle prediction time;

calculating a vehicle predictive stability margin κ _stable (k)；

κ _stable (k)＝max(κ _ueb (k),κ _unt (k),κ _untchange (k))；

Wherein, kappa _ueb (k) Indicating understeer factor, κ _unt (k) Indicating oversteer factor, κ _untchange (k) Indicating a corrected oversteer factor;

wherein s is _fl 、s _fr 、s _rl Sum s _rr Slip ratios of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel respectively;delta as steering wheel angle influence coefficient _sw Delta for steering wheel angle _swcri Is the critical rotation angle;

predicting the states of obstacles around the vehicle, and determining a collision-free risk space;

planning a vehicle track in the collision-free risk space to obtain target lateral displacement, target lateral speed, target lateral acceleration, target longitudinal displacement, target longitudinal speed and longitudinal acceleration of the vehicle;

wherein whenAnd->When the target lateral acceleration is corrected to

When (when)Or mean(κ _untchange (k) And->When the target lateral acceleration is corrected to +.>

In the method, in the process of the invention,for the lateral acceleration increment coefficient omega _des Representing a target yaw rate; ω is expressed as actual yaw rate, a _ytar Representing a target lateral acceleration; kappa (kappa) _uebmax Is an understeer factor maximum;

and controlling steering wheel rotation angle and distributing torque of the four-wheel motor according to the vehicle track planning result.

Preferably, the longitudinal predicted vehicle speed at each time in the vehicle prediction time domainLateral prediction of vehicle speed->Predicted yaw rate +.>And predicting centroid slip angle +.>The calculation method of (1) is as follows:

wherein,represents a longitudinal predicted vehicle speed predicted from the history data, and>represents a laterally predicted vehicle speed predicted from historical data, and->Representing the predicted yaw rate predicted from the history data,/->Representing a predicted centroid slip angle predicted from the historical data; />Represents a longitudinal predicted vehicle speed predicted from a linear vehicle dynamics model,/->Represents a lateral predicted vehicle speed predicted from a linear vehicle dynamics model,/->Representing predicted yaw rate predicted from linear vehicle dynamics model,/->Representing a predicted centroid slip angle predicted from a linear vehicle dynamics model; />Represents a longitudinally predicted vehicle speed predicted from a nonlinear vehicle dynamics model,/->Represents the predicted lateral speed of the vehicle predicted by the nonlinear vehicle dynamics model,/->Representing predicted yaw rate predicted from a nonlinear vehicle dynamics model,/and method for generating a target yaw rate>Representing a predicted centroid slip angle predicted from a nonlinear vehicle dynamics model;and->Each coefficient; a, a _yc Expressed as corrected lateral acceleration; g is gravitational acceleration.

Preferably, the understeer factor calculation formula is:

in the method, in the process of the invention,is the ratio influence coefficient of the understeer factor; />A vehicle speed influence coefficient that is an understeer factor;the road surface adhesion influence coefficient; a, a _ω For an understeer factorYaw rate error coefficient; a, a _β A centroid slip angle error coefficient that is an understeer factor; beta _des Is the target centroid slip angle; />Predicting a centroid slip angle for the moment k; />Predicting yaw rate for time k; />Predicting a longitudinal vehicle speed for the moment k; mu (mu) ₀ The road adhesion coefficient after correction.

Preferably, the oversteer factor is calculated as:

in the method, in the process of the invention,a vehicle speed influence coefficient that is an oversteer factor; />Road surface adhesion influence coefficient which is an oversteer factor; b _ω A yaw rate error coefficient that is an oversteer factor; b _β Is the centroid slip angle error coefficient of the oversteer factor.

Preferably, the method for determining the collision-free risk space is as follows:

traversing all individual obstacle Risk values Risk _i ；

If Risk _i ＜Risk _max No collision risk space constraint is required;

if Risk _i ≥Risk _max Predicting a collision risk space of the single obstacle i;

wherein, risk _max Threshold for collision risk；

Traversing collision risk spaces of all obstacles with collision risk, and differentiating all the collision risk spaces of the preliminary collision risk spaces to obtain residual collision-free risk spaces;

the preliminary collision risk space is a space in an effective lane boundary line or an effective driving area in a sensor sensing range.

Preferably, a single obstacle Risk value Risk _i Calculated by the following formula;

wherein,is a longitudinal risk influence coefficient; />Is a lateral risk influence coefficient; d, d _i Distance from the vehicle to the ith obstacle; l (L) _wai The longitudinal early warning distance from the vehicle to the obstacle i is set; l (L) _coi The longitudinal collision danger distance from the vehicle to the obstacle i is set; s is(s) _wai The lateral early warning distance from the vehicle to the obstacle i is set; s is(s) _coi A lateral collision danger distance from the vehicle to the obstacle i; deltav _xobsi For the longitudinal relative speed of the vehicle and the obstacle i, deltav _yobsi The lateral relative speed of the vehicle and the obstacle i.

Preferably, the vehicle emergency braking is controlled if there is no solution to the vehicle trajectory planning in the collision-free risk space.

Preferably, ifOr->Torque of the four-wheel motor is evenly distributed;

wherein T is _bfl 、T _bfr 、T _brl 、T _brr The motor torques of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are respectively; a is the distance from the centroid to the front axis; b is the distance from the centroid to the rear axis; m is the mass of the whole vehicle.

Preferably, ifAnd->

When the vehicle turns to the left in the future;

wherein k is _ueωl Correction coefficient, k, for left turn understeer omega _ueβl A correction coefficient for left-turn understeer beta; when the vehicle is turned to the right in the future,

wherein k is _ueωr Correction coefficient k for right turn understeer omega _ueβl And correcting the coefficient for right-turn understeer beta. It is preferred that the composition of the present invention,or mean (kappa) _untchange (k) And->When the vehicle turns to the left in the future:

wherein k is _unωl Correction coefficient, k, for left turn oversteer omega _unβl Correcting the coefficient for the over steering beta;

when the vehicle is turned to the right in the future,

wherein k is _unωr Correction coefficient, k, for right turn oversteer omega _unβl The beta correction factor is for right turn oversteer.

The beneficial effects of the invention are as follows:

according to the automatic driving-oriented horizontal and vertical integrated prediction control method, the vehicle driving safety space is constructed jointly by comprehensively predicting the vehicle stable state and the collision risk with the obstacle, track planning and vehicle motion control are performed in the safety space by considering the vehicle stable state constraint, the vehicle motion state and the environment collision risk can be predicted in advance, further pre-control is performed, the intelligent vehicle driving safety and riding comfort are improved, and the intelligent level of the intelligent vehicle is effectively improved.

Drawings

Fig. 1 is a frame diagram of an automatic driving horizontal and vertical integrated predictive planning control system according to the invention.

Fig. 2 is a flow chart of the automatic driving horizontal and vertical integrated prediction planning control method.

Fig. 3 is a crash risk space constraint diagram according to the present invention.

Detailed Description

The present invention is described in further detail below with reference to the drawings to enable those skilled in the art to practice the invention by referring to the description.

In order to improve autonomy and actual performance of an intelligent driving system under critical dangerous scenes and limiting working conditions, the invention provides an automatic driving horizontal and vertical integrated prediction control method for an intelligent distributed (four-wheel independent driving) electric automobile, which is implemented based on the following automatic driving horizontal and vertical integrated prediction control system.

As shown in FIG. 1, the automatic driving-oriented horizontal and vertical integrated prediction planning control system consists of a data acquisition module, a central processing module and an execution module. The data acquisition module consists of a laser radar sensor, a camera sensor, a GNNS (Global Navigation Satellite System) sensor and a steering wheel angle sensor. The laser radar is used for sensing barrier information of the surrounding environment of the vehicle; the camera sensor is used for sensing lane line information and barrier information of surrounding environment of the vehicle; the GNNS sensor is a global navigation satellite system and is used for positioning vehicle position information and detecting the longitudinal speed v of the vehicle _x Lateral speed v of vehicle _y Longitudinal acceleration a _x Lateral acceleration a _y Yaw rate ω, heading angle θ; the steering wheel angle sensor is used for detecting the steering wheel angle delta _f The method comprises the steps of carrying out a first treatment on the surface of the The central processing module consists of vehicle state prediction, driving risk prediction and driving behavior decision; the vehicle state prediction is used for predicting state information of the vehicle in a future period of time; the collision risk prediction is used for predicting the collision risk with surrounding obstacles in a future period of time; the driving behavior decision is carried out according to the result after the driving risk prediction, and a control instruction is sent to the execution module; the execution module comprises a steering wheel, a left front wheel motor, a right front wheel motor, a left rear wheel motor and a right rear wheel motor; the steering wheel is used for controlling the transverse movement direction of the vehicle; the left front wheel motor, the right front wheel motor, the left rear wheel motor and the right rear wheel motor are used for controlling four-wheel torque.

As shown in fig. 2, the specific method for the automatic driving-oriented horizontal-vertical integrated predictive planning control is as follows.

1. Vehicle state prediction

The GNSS sensor and the steering wheel angle sensor of the data acquisition module can respectively acquire the longitudinal displacement X, the lateral displacement Y and the longitudinal vehicle speed v of the vehicle _x Lateral speed v of vehicle _y Longitudinal acceleration a _x Lateral acceleration a _y Heading angle θ, steering wheel angle δ _sw ；

To promote T in the future _sp The accuracy of the vehicle state prediction in the (vehicle prediction horizon) time period is composed of three parts, namely, historical data prediction, linear vehicle dynamics model prediction and nonlinear vehicle dynamics model prediction.

A history data prediction section:

record the current time T _nf If the current time is less than T _nf (initial stage time threshold) no prediction is made;

the recorded data are:

wherein Δt is _nc Is the sampling period; v is set as _x (t-T _nf I T) as an example, the previous T < th > representing the current time _nf The speed of the vehicle at the moment;

the historical centroid slip angle can be calculated from the historical data,

the predictive model formula:

……

wherein s (k+ 1|t) represents a predicted value of the next cycle time at the current time, and s (k+T) _sp /Δt _nc I T) represents the future k+t at the current time _sp /Δt _nc Predicted values for the period time; taking k=t; s (k-j|t) represents a value of a period j before the current time, which is history data; in the subsequent prediction, the previous cycle predicted value is used as historical data;

will v _x 、v _y Predicting states of omega and beta by using the prediction model to obtain a prediction array based on historical data(longitudinally predicted vehicle speed),>(lateral predicted vehicle speed),%>(predicted yaw rate)>(predicting centroid slip angle);

linear vehicle dynamics model prediction section:

the discrete linear vehicle dynamics prediction model is as follows:

……

in the method, in the process of the invention,δ _f i is the front wheel rotation angle _sw T is the sampling period of the prediction model, k, for the transmission ratio from the steering wheel to the front wheel _f For yaw stiffness of front axle, k _r For the cornering stiffness of the rear axle, m is the mass of the whole vehicle, a is the distance from the mass center to the front axle, and b is the distance from the mass center to the rear axle;

array prediction from linear vehicle dynamics model(longitudinally predicted vehicle speed),>(predicted vehicle speed sideways),(predicted yaw rate)>(predicting centroid slip angle);

nonlinear vehicle dynamics model prediction section:

the discrete nonlinear vehicle dynamics model is as follows:

……

wherein F is _cf For the side force of the front wheel F _cr The side force of the rear wheel is calculated by a tire magic formula;

predicting an array from a nonlinear vehicle dynamics model(longitudinally predicted vehicle speed),>(lateral predicted vehicle speed),%>(predicted yaw rate)>(predicting centroid slip angle);

the three-part prediction result is fitted according to the following formula,

wherein a is _yc For corrected lateral acceleration, a _yc The following conditions are satisfied:

g is gravity acceleration;

other fitting coefficients satisfy the condition that,

2. vehicle predictive stability margin calculation

In order to better realize the prediction of the stable state of the vehicle, help is provided for the subsequent behavior decision, and the current stable state of the vehicle needs to be calculated;

firstly, calculating the yaw rate and the centroid slip angle of the target:

for a pair ofAnd->Integrating to obtain beta _des And omega _des ；

The understeer factor calculation formula is as follows:

in the method, in the process of the invention,is the proportion influence coefficient of the understeer factor, and has a value range of 1-5; />The value range is 0-1 for the vehicle speed influence coefficient; />The value range is 0-1 for the road surface adhesion influence coefficient; a, a _ω The value range is 0-1 for yaw rate error coefficient; a, a _β The value range is 0-1 for the error coefficient of the centroid side deflection angle;

wherein mu is ₀ The following limitations are met,

wherein μ is road adhesion coefficient, μ ₀ The road surface adhesion coefficient after correction;

the oversteer factor is calculated as follows:

the oversteer factor is as follows:

takes the value of the vehicle speed influence coefficientThe range is 0 to 1; />The value range is 0-1 for the road surface adhesion influence coefficient; b _ω The value range is 0-1 for yaw rate error coefficient; b _β The value range is 0-1 for the error coefficient of the centroid side deflection angle;

if the vehicle speed v is simultaneously satisfied _x Greater than the low adhesion critical vehicle speed v _xcri0 Steering wheel angle delta _sw Is greater than critical angle delta _swcri ，μ ₀ A critical value mu smaller than the low adhesion coefficient _cri0 Slip ratio s of four wheels _fl (left front wheel), s _fr (Right front wheel), s _rl (left rear wheel), s _rr (rear right wheel) in which any one is greater than the critical slip ratio s _cri Indicating that the vehicle is in a low adhesion risk steering condition, further correction of the oversteer factor is required, the corrected oversteer factor is as follows,

in the method, in the process of the invention,the value range is 0-1 for the steering wheel angle influence coefficient;

vehicle steady state is κ _stable (k)，κ _stable (k)＝max(κ _ueb (k),κ _unt (k),κ _untchange (k))。

3. Obstacle state prediction in effective road space

According to the obstacle information, lane line information and road boundary line information acquired by the laser radar and the camera, extracting the obstacle information in the perception range and the effective road space according to the current common algorithm;

the extracted information includes longitudinal relative vehicle speed Deltav _xobs i. Lateral relative vehicle speed Deltav _yobsi Longitudinal relative acceleration Δa _xobsi Lateral relative accelerationDegree Deltaa _yobsi Longitudinal relative displacement DeltaX _obsi Relative lateral displacement Δy _obsi I ranges from 0 to N _obsmax ，N _obsmax Is the maximum number of perceived obstacles;

in the prediction period T _sp In the method, the obstacle information is predicted by using a vehicle kinematic model,

……

4. collision risk detection

In the prediction time domain T _sp Predicting the risk of collision between surrounding obstacles and the vehicle;

the risk status of the individual obstacle is calculated according to the following formula,

in the method, in the process of the invention,the value range is 0-1 for the longitudinal risk influence coefficient; />The value range is 0-1 for the lateral risk influence coefficient; d, d _i Distance from the vehicle to the ith obstacle; l (L) _wai The longitudinal early warning distance from the vehicle to the ith obstacle vehicle is provided; l (L) _coi The longitudinal collision dangerous distance from the vehicle to the ith obstacle vehicle is set; s is(s) _wai From the host vehicle to the ithThe lateral early warning distance of the obstacle vehicle; s is(s) _coi The side collision dangerous distance from the vehicle to the ith obstacle vehicle is set; l (L) _wai ，l _coi ，s _wai ，s _coi The specific expression is as follows:

wherein τ _i Is the critical collision time of the vehicle and a single obstacle,l _wa0 the critical stopping distance is 3-5 m; c _sw The value range is 0-1 for the lateral early warning adjustment coefficient; />The value range is 0-1 for the lateral collision adjusting coefficient; w (w) _d Is the lane width, is determined by the perceived lane information;

5. driving behavior decision

A collision risk space constraint

Single obstacle Risk condition Risk calculated from four (collision Risk detection) _i Checking Risk value of traversal, if Risk _i ＜Risk _max The Risk value of the obstacle does not exceed the Risk threshold value, no collision Risk space constraint is needed, and Risk is shown _max Value range 5 for collision risk threshold10; otherwise, the risk value of the obstacle exceeds a risk threshold value, the risk of collision exists, the collision risk space constraint is needed, the collision risk space constraint process is as follows,

assuming that the collision risk space is taken as a preliminary collision risk space in the sensing range of the sensor and in an effective lane boundary line or an effective driving area, predicting the collision risk space of the obstacle with collision risk, traversing the spaces of all the obstacles with collision risk, and making differences between all the collision risk spaces and the preliminary collision risk space to obtain the rest collision risk-free space;

taking the scenario shown in fig. 3 as an example, the obstacle 1 is the obstacle vehicle No. 1, the obstacle 2 is the obstacle vehicle No. 2, and the longitudinal and transverse lengths of the collision risk spaces of the single obstacle are respectively:

l _ki ＝l _wai +l _cari +d _carw ；

s _ki ＝s _wai +w _cari +w _carw ；

wherein, I _cari Is the length of the obstacle vehicle, w _cari Is the width of the obstacle, is obtained by a sensing part, d _carw For reserving the distance longitudinally, the value range is 0.2 to 0.5m, w _carw The reserved distance is reserved for the transverse direction, and the value range is 0.1-0.2 m;

taking the obstacle 1 as an example, the longitudinal length and the transverse length of the collision risk control are respectively,

l _k1 ＝l _wa1 +l _car1 +d _carw ；

s _k1 ＝s _wa1 +w _car1 +w _carw ；

the space indicates that in the current vehicle state, if the vehicle enters the space, the risk of collision with the obstacle 1 exists, and if the vehicle does not exist in the space, the risk of collision with the obstacle 1 does not exist;

after checking the obstacle 1, checking collision risk spaces of other obstacles in sequence, and merging and integrating all the obstacle risk spaces to obtain a total collision risk space;

if there is an obstacleThe object vehicle and the self-vehicle have a certain included angle, the included angle degree is more than 2 degrees, and the relative vehicle speed increment Deltav _yi If the distance between the collision risk space boundary near the side of the own vehicle and the vehicle boundary is more than 2 degrees, as shown in the obstacle vehicle 1 in fig. 3, the obstacle vehicle 1 and the own vehicle have a certain included angle of more than 0.02m/sThe other side is->Wherein s is _wa1 Representing the lateral early warning distance from the vehicle to the 1 st obstacle vehicle; otherwise, the two sides are equally distributed.

Track planning taking into account the steady state of the vehicle

The average steady state of the vehicle in the predicted horizon is calculated,

if it isAnd->Indicating that the vehicle is in understeer state in future, κ _uebmax The value range is 0.25-0.5, the lateral acceleration can be properly promoted in the track planning process, and the target lateral displacement, the target lateral velocity and the target lateral acceleration are [ Y ] _tar v _ytar a _ytar +Δa _ytar ]Target longitudinal displacement, target longitudinal velocity, target longitudinal acceleration [ X ] _tar v _xtar a _xtar ]；

Δa _ytar For the target lateral acceleration increment,

the value range is 0-2 for the lateral acceleration increment coefficient;

and carrying out transverse and longitudinal planning by using a fifth-order polynomial respectively in the transverse and longitudinal directions:

the lateral initial constraint being the current vehicle lateral state [ Y v ] _y a _y ]The terminal state is the target quantity [ Y _tar v _ytar a _ytar +Δa _ytar ]Solving the coefficient a according to the initial constraint and the terminal target ₀ ，a ₁ ，a ₂ ，a ₃ ，a ₄ ，a ₅ ，

Wherein t is the time of path planning, and the value range is [0, T _pend ],T _pend Planning terminal time;

longitudinal alike, longitudinal initiation constraint is the current vehicle longitudinal state [ X v ] _x a _x ]The terminal state is the target quantity [ X _tar v _xtar a _xtar ]Solving the coefficient b according to the initial constraint and the terminal target ₀ ，b ₁ ，b ₂ ，b ₃ ，b ₄ ，b ₅ ；

If it isOr mean (kappa) _untchange (k) And->Indicating that the vehicle is in an oversteer condition in the future, κ _uebmax The value range is 0.25-0.5, the lateral acceleration can be properly reduced in the track planning process, and the target lateral direction is setThe lateral velocity and lateral acceleration are [ Y _tar v _ytar a _ytar -Δa _ytar ]The target longitudinal displacement, longitudinal velocity and longitudinal acceleration are [ X ] _tar v _xtar a _xtar ]；

Δa _ytar For the target lateral acceleration increment,

the value range is 0-2 for the lateral acceleration increment coefficient;

and carrying out transverse and longitudinal planning by using a fifth-order polynomial respectively in the transverse and longitudinal directions:

on the contrary, ifOr->Indicating that the vehicle is in a stable state in the future, the track planning target state is not required to be adjusted, and the target lateral displacement, the lateral speed and the lateral acceleration are [ Y ] _tar v _ytar a _ytar ]The target longitudinal displacement, longitudinal velocity and longitudinal acceleration are [ X ] _tar v _xtar a _xtar ]And carrying out horizontal and longitudinal planning by using a fifth-order polynomial respectively in the horizontal and longitudinal directions:

if there is no solution in the collision risk constraint space, it indicates that in the feasible space, the collision cannot be avoided by active steering or partial braking, and the vehicle has a large collision risk, and at this time, the degree of collision needs to be avoided by emergency braking or reduced to the greatest extent.

6. Execution control

Calculating a target lateral movement track and a longitudinal movement track in five (driving behavior decision), tracking the target lateral track by using a transverse PID controller, calculating a target steering wheel corner, and controlling and following by steering wheel control; tracking the longitudinal speed of the target by using a longitudinal PID controller, and calculating the total torque M of the target _tarall The four-wheel motor is dynamically adjusted and distributed, and the dynamic adjustment and distribution steps are as follows:

if the future motion state of the vehicle is in a steady stateOr->) The four-wheel motor torque is distributed according to the average distribution,

T _bfl 、T _bfr 、T _brl 、T _brr the motor torques of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are respectively;

if the future motion state of the vehicle is under-turnedAnd->) And the vehicle turns left in the future, the right side wheel increases the torque, the left side wheel decreases the torque,

k _ueωl correction coefficient, k, for left turn understeer omega _ueβl A correction coefficient for left-turn understeer beta;

if the future motion state of the vehicle is under-turnedAnd->) And the vehicle turns right in the future, the left side wheel increases the torque, the right side wheel decreases the torque,

k _ueωr correction coefficient k for right turn understeer omega _ueβl A correction coefficient for the right turn understeer beta;

if the future motion state of the vehicle is oversteeredOr mean (kappa) _untchange (k) Is) and) And the vehicle turns left in the future, the right side wheel reduces the torque, the left side wheel increases the torque,

k _unωl correction coefficient, k, for left turn oversteer omega _unβl Correcting the coefficient for the left turn oversteer beta;

if the future motion state of the vehicle is oversteeredOr mean (kappa) _untchange (k) Is) and) And the vehicle turns right in the future, the left side wheel reduces the torque, the right side wheel increases the torque,

k _unωr correction coefficient, k, for right turn oversteer omega _unβl The beta correction factor is for right turn oversteer.

Although embodiments of the present invention have been disclosed above, it is not limited to the details and embodiments shown and described, it is well suited to various fields of use for which the invention would be readily apparent to those skilled in the art, and accordingly, the invention is not limited to the specific details and illustrations shown and described herein, without departing from the general concepts defined in the claims and their equivalents.

Claims (10)

1. An automatic driving horizontal and vertical integrated prediction control method is characterized by comprising the following steps:

acquiring current vehicle state parameters through sensors: longitudinal displacement X, lateral displacement Y, longitudinal vehicle speed v _x Lateral speed v of vehicle _y Longitudinally addSpeed a _x Lateral acceleration a _y Heading angle θ, steering wheel angle δ _sw The method comprises the steps of carrying out a first treatment on the surface of the Predicting the future state of the vehicle according to the current vehicle state parameters to obtain longitudinal predicted vehicle speed, lateral predicted vehicle speed, predicted yaw rate and predicted centroid side deviation angle at each moment in the vehicle prediction time;

calculating a vehicle predictive stability margin κ _stable (k)；

κ _stable (k)＝max(κ _ueb (k),κ _unt (k),κ _untchange (k))；

Wherein, kappa _ueb (k) Indicating understeer factor, κ _unt (k) Indicating oversteer factor, κ _untchange (k) Indicating a corrected oversteer factor;

wherein s is _fl 、s _fr 、s _rl Sum s _rr Slip ratios of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel respectively;delta as steering wheel angle influence coefficient _sw Delta for steering wheel angle _swcri Is the critical rotation angle;

predicting the states of obstacles around the vehicle, and determining a collision-free risk space;

planning a vehicle track in the collision-free risk space to obtain target lateral displacement, target lateral speed, target lateral acceleration, target longitudinal displacement, target longitudinal speed and longitudinal acceleration of the vehicle;

wherein whenAnd->When it willThe target lateral acceleration is corrected to a _ytar1 ′；

When (when)Or mean (kappa) _untchange (k) And->When the target lateral acceleration is corrected to a _ytar2 ′；/>

In the method, in the process of the invention,for the lateral acceleration increment coefficient omega _des Representing a target yaw rate; ω is expressed as actual yaw rate, a _ytar Representing a target lateral acceleration; kappa (kappa) _uebmax Is an understeer factor maximum;

and controlling steering wheel rotation angle and distributing torque of the four-wheel motor according to the vehicle track planning result.

2. The automated lateral and longitudinal integrated predictive control method for vehicle according to claim 1, wherein the longitudinal predicted vehicle speed at each time in the vehicle prediction time domainLateral prediction of vehicle speed->Predicted yaw rate +.>And pre-heatingMeasuring centroid side deviation angle->The calculation method of (1) is as follows:

wherein,represents a longitudinal predicted vehicle speed predicted from the history data, and>represents a laterally predicted vehicle speed predicted from historical data, and->Representing the predicted yaw rate predicted from the history data,/->Representing a predicted centroid slip angle predicted from the historical data; />Representation ofLongitudinal predicted vehicle speed predicted from linear vehicle dynamics model, and method for generating a linear vehicle dynamics model>Represents a lateral predicted vehicle speed predicted from a linear vehicle dynamics model,/->Representing predicted yaw rate predicted from linear vehicle dynamics model,/->Representing a predicted centroid slip angle predicted from a linear vehicle dynamics model;represents a longitudinally predicted vehicle speed predicted from a nonlinear vehicle dynamics model,/->Represents the predicted lateral speed of the vehicle predicted by the nonlinear vehicle dynamics model,/->Representing predicted yaw rate predicted from a nonlinear vehicle dynamics model,/and method for generating a target yaw rate>Representing a predicted centroid slip angle predicted from a nonlinear vehicle dynamics model; ρ _vxdata 、ρ _vxdyn 、ρ _vydata 、ρ _vydyn 、ρ _ωdata 、ρ _ωdyn 、ρ _βdata And ρ _βdyn Each coefficient; a, a _yc Expressed as corrected lateral acceleration; g is gravitational acceleration.

3. The automated driving horizontal and vertical integrated predictive control method according to claim 2, wherein the understeer factor calculation formula is:

in the method, in the process of the invention,is the ratio influence coefficient of the understeer factor; />A vehicle speed influence coefficient that is an understeer factor; />The road surface adhesion influence coefficient; a, a _ω Yaw rate error coefficient which is an understeer factor; a, a _β A centroid slip angle error coefficient that is an understeer factor; beta _des Is the target centroid slip angle; />Predicting a centroid slip angle for the moment k; />Predicting yaw rate for time k; />Predicting a longitudinal vehicle speed for the moment k; mu (mu) ₀ The road adhesion coefficient after correction.

4. The automated lateral and longitudinal integrated predictive control method according to claim 3, wherein the oversteer factor is calculated by the formula:

wherein b is _vx A vehicle speed influence coefficient that is an oversteer factor; b _μ0 Road surface adhesion influence coefficient which is an oversteer factor; b _ω A yaw rate error coefficient that is an oversteer factor; b _β Is the centroid slip angle error coefficient of the oversteer factor.

5. The automated lateral and longitudinal integrated predictive control method for vehicle driving according to claim 3 or 4, wherein the method for determining collision-free risk space is:

traversing all individual obstacle Risk values Risk _i ；

If Risk _i ＜Risk _max No collision risk space constraint is required;

if Risk _i ≥Risk _max Predicting a collision risk space of the single obstacle i;

wherein, risk _max Is a collision risk threshold;

traversing collision risk spaces of all obstacles with collision risk, and differentiating all the collision risk spaces of the preliminary collision risk spaces to obtain residual collision-free risk spaces;

the preliminary collision risk space is a space in an effective lane boundary line or an effective driving area in a sensor sensing range.

6. The automated lateral and longitudinal integrated predictive control method for a vehicle of claim 5, wherein a single obstacle Risk value Risk _i Calculated by the following formula;

wherein,is a longitudinal risk influence coefficient; />Is a lateral risk influence coefficient; d, d _i Distance from the vehicle to the ith obstacle; l (L) _wai The longitudinal early warning distance from the vehicle to the obstacle i is set; l (L) _coi The longitudinal collision danger distance from the vehicle to the obstacle i is set; s is(s) _wai The lateral early warning distance from the vehicle to the obstacle i is set; s is(s) _coi A lateral collision danger distance from the vehicle to the obstacle i; deltav _xobsi For the longitudinal relative speed of the vehicle and the obstacle i, deltav _yobsi The lateral relative speed of the vehicle and the obstacle i.

7. The automated guided transverse and longitudinal integrated predictive control method of claim 6, wherein the vehicle emergency braking is controlled if the vehicle trajectory plan in the collision risk free space is free of solutions.

8. The automated lateral and longitudinal integrated predictive control method of claim 7, wherein ifOr->Torque of the four-wheel motor is evenly distributed;

wherein T is _bfl 、T _bfr 、T _brl 、T _brr The motor torques of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are respectively; a is the distance from the centroid to the front axis; b is the distance from the centroid to the rear axis; m is the mass of the whole vehicle.

9. The automated driving oriented of claim 8A method for controlling the integrated prediction of the transverse and longitudinal directions of a vehicle is characterized byAnd->

When the vehicle turns to the left in the future;

wherein k is _ueωl Correction coefficient, k, for left turn understeer omega _ueβl A correction coefficient for left-turn understeer beta;

when the vehicle is turned to the right in the future,

wherein k is _ueωr Correction coefficient k for right turn understeer omega _ueβl And correcting the coefficient for right-turn understeer beta.

10. The automated driving oriented horizontal and vertical integrated predictive control method of claim 9, wherein,or mean (kappa) _untchange (k) And->

When the vehicle turns to the left in the future:

wherein k is _unωl Correction coefficient, k, for left turn oversteer omega _unβl Correcting the coefficient for the over steering beta;

when the vehicle is turned to the right in the future,

wherein k is _unωr Correction coefficient, k, for right turn oversteer omega _unβl The beta correction factor is for right turn oversteer.