CN115933662A - Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control - Google Patents
Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control Download PDFInfo
- Publication number
- CN115933662A CN115933662A CN202211594290.8A CN202211594290A CN115933662A CN 115933662 A CN115933662 A CN 115933662A CN 202211594290 A CN202211594290 A CN 202211594290A CN 115933662 A CN115933662 A CN 115933662A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- adaptive
- wheel
- model
- follows
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention discloses an intelligent automobile track tracking and stability control system and method based on adaptive model predictive control, and provides an adaptive preview error model, an adaptive forgetting factor recursive least square estimator and an integrated controller, aiming at the problem that a conventional preview model only changes the preview distance according to the automobile speed, the influence of a road surface adhesion coefficient on an intelligent automobile is considered, two preview distance switching modes based on the automobile speed and the road surface adhesion coefficient are designed, and switching soft constraints are made at the switching points of the automobile speed and the road surface adhesion coefficient, so that the repeated switching mode is avoided, and the system stability is improved; the method has the advantages that the tire cornering stiffness and the road adhesion coefficient are estimated on line aiming at the variable road condition scene which the intelligent automobile may face actually, the balance between the coordination convergence speed and the identification error of the adaptive forgetting factor is introduced, the precision of a prediction model is effectively improved, meanwhile, the control quantity constraint is adaptively adjusted based on the road adhesion coefficient, and the stability of the automobile is improved.
Description
Technical Field
The invention relates to the technical field of intelligent automobile control, in particular to an intelligent automobile trajectory tracking and stability control system and method based on self-adaptive model prediction control.
Background
With the rapid development of sensor technology, on-board computers and artificial intelligence, in this context, the automatic driving technology has been a worldwide research hotspot in the last decade. Automatic driving integrates perception, planning and motion control, wherein trajectory tracking control is a key technology. The main purpose of trajectory tracking is to accurately track a reference trajectory by eliminating tracking deviations. The trajectory tracking control method mainly comprises feed-forward feedback control, proportional-integral-derivative control, sliding mode control and the like, wherein the model prediction control has a prediction characteristic, can solve the problems of multi-objective optimization and constraint, and has strong robustness, so that the trajectory tracking control method is widely applied.
At present, fixed model parameters are generally adopted for model prediction control, and the actual working conditions of the intelligent automobile are complex and changeable and are influenced by various factors such as road curvature, road surface adhesion coefficient, tire cornering stiffness, automobile speed and the like, so that the accuracy of the model can be reduced, and the fixed constraint is adopted all the time to be not beneficial to the control of the intelligent automobile, so that the capability of tracking the automobile track can be reduced. In addition, the conventional preview error model for changing the preview distance according to the vehicle speed does not consider the change of the road adhesion coefficient, and different road conditions have different requirements on the preview distance, so that the self-adaptive control capability of the intelligent vehicle needs to be improved.
Disclosure of Invention
Aiming at the problems, the invention provides an intelligent automobile track tracking and stability control system based on adaptive model predictive control. The method mainly comprises four parts, namely an adaptive preview error model, an adaptive forgetting factor recursive least square estimator, a transverse controller and a torque distribution controller.
Wherein, the adaptive preview error model: different requirements of road curvature, road surface adhesion coefficient and vehicle speed on the preview distance are comprehensively considered, a switching mode and soft constraints are designed, and a self-adaptive preview error model is constructed. And a vehicle physical model is constructed through some simplifications and assumptions, the two models are combined to be used as a prediction model of the model prediction controller, and the control quantity is obtained by solving the model.
An adaptive forgetting factor recursive least squares estimator: under the actual application scene of the intelligent automobile, the road surface attachment coefficient and the tire cornering stiffness can change at any time along with different working conditions, so that the tire cornering stiffness and the road surface attachment coefficient are estimated by adopting the self-adaptive forgetting factor recursive least square estimator, and the self-adaptive forgetting factor is introduced to coordinate the balance between the convergence speed and the identification error. Accurate tire cornering stiffness estimation is beneficial to improving the accuracy of a model prediction controller model, meanwhile, the constraint of the control quantity can be changed in real time by estimating the road adhesion coefficient, and the tracking performance and the safety of the model prediction controller are improved.
A transverse controller: and a model prediction controller is adopted for carrying out transverse control, the function of the model prediction controller is to collect the state quantity of the vehicle, based on the established prediction model, the optimal solution is carried out under the condition of meeting the designed objective function and constraint, the front wheel turning angle is solved and acts on the actual vehicle, so that the minimum error between the actual running track and the reference track of the vehicle is realized, and meanwhile, the additional yaw moment is solved and used as an expected value to be output to torque distribution control for carrying out torque distribution and realizing stability control.
A torque distribution controller: the additional yaw moment and the longitudinal force output by the controller are subjected to torque optimized distribution under the condition of meeting the designed objective function and constraint, the influence of the tire slip rate is considered, further optimization is carried out, and the additional yaw moment and the longitudinal force act on an actual vehicle, so that stability control is realized, the trajectory tracking of the vehicle is assisted, and the error of the trajectory tracking is reduced.
Further, the assumption that the vehicle physical model provided by the invention is a vehicle dynamic model is as follows: the vertical, lateral and pitching motion of the vehicle is not considered, the vehicle is simplified into a two-degree-of-freedom dynamic model, and the vehicle only moves in an x-o-y plane. The vehicle is turned by the front wheel, a coordinate system of the vehicle body is positioned in a plane of bilateral symmetry of the vehicle, the origin of the center of mass of the vehicle is o, the x axis is the longitudinal axis of the vehicle, the positive direction of the x axis is the direction of the vehicle head, the y axis points to the lateral direction of the vehicle body, the positive direction meets the right-hand rule, and the positive direction of the z axis is vertical to the upper direction. Therefore, according to newton's second law, the equations of rotational balance and force balance are established at the centroid of the vehicle, resulting in the following expression:
wherein m is the vehicle mass, V x Is the longitudinal speed of the mass center under the coordinate system of the vehicle body, beta is the side slip angle of the mass center of the vehicle, gamma is the yaw angular velocity of the vehicle, I z Is the moment of inertia of the vehicle about the z-axis,/ f And l r Distances from the center of mass of the vehicle to the front axle and the rear axle, respectively, F yfl F yfr F yrl F yrr Respectively lateral forces acting on four wheels of the vehicle, M z An additional yaw moment.
From the above assumptions, it can be found that the tire lateral force is proportional to the tire slip angle, resulting in the following expression:
wherein, F yf And F yr Resultant of lateral forces acting on the wheels of the front axle and rear axle of the vehicle, respectively, C f And C r Cornering stiffness, alpha, of the front and rear wheels of the vehicle, respectively f And alpha r Slip angles, delta, of the front and rear wheels of the vehicle, respectively f Is the corner of the front wheel.
The expression for the two-degree-of-freedom model is obtained as follows:
in general, β and
the value is relatively small, close to 0, and the vehicle yaw angle can be considered to be approximately equal to the heading angle.
The preview error model expression is as follows:
wherein, the first and the second end of the pipe are connected with each other,
is the desired heading angle of the pre-aim point>
Is the current heading angle of the vehicle, and>
is the deviation of the heading angle of the pre-aiming point, is the difference between the desired heading angle and the current heading angle of the vehicle, e d For lateral deviation, e d0 Is the distance between the extension line of the pre-aiming point and the mass center of the current vehicle in the y-axis direction of the vehicle coordinate system, L p Is the pre-aiming distance.
Derived from the above equation:
where κ is the road curvature of the preview point.
The decision-making module collects information of road curvature, road adhesion and vehicle speed, comprehensively considers calculation burden, stability and safety of the controller, and respectively designs two pre-aiming distance switching modes based on a road adhesion coefficient and a vehicle speed.
With reference to the actual vehicle experimental data, the road adhesion coefficient was divided into three types of road surfaces, and a dry road surface was set when the road adhesion coefficient was 0.7 to 1, a wet road surface was set when the road adhesion coefficient was 0.5 to 0.7, and an icy and snowy road surface was set when the road adhesion coefficient was 0.3 to 0.5.
On a dry road surface, the intelligent automobile has good tracking performance and control capability, in order to prevent the calculation burden of a controller from being too heavy, a pre-aiming distance mode based on the speed of the automobile is adopted, and the expression is as follows:
wherein L is pmin The minimum pre-aiming distance is set to be 6m pmax Setting the maximum pre-aiming distance to be 20m min And V max Respectively minimum vehicle speed and maximum vehicle speed, respectively designed to be 6m/s and 20m/s, T p The preview time is set to 1s.
When the road surface is slippery and icy or snowy, stability and safety need to be considered, the control capability of the intelligent automobile is reduced, the curvature of the road at the long pre-aiming distance is also considered even under the condition of low speed, and the response is made in advance, so that a pre-aiming distance mode based on the road surface adhesion coefficient is adopted, and the expression is as follows:
wherein L is p1 And L p2 The two preview distances of the switching mode are respectively set to be 20m and 23m.
In order to prevent the influence of frequent switching of the preview distance mode on control, boundary fuzzy control is established at the boundary of the road adhesion coefficient, and soft constraint of +/-0.1 is formed on the road adhesion coefficient; and establishing boundary fuzzy control at the vehicle speed boundary to form +/-2 m/s soft constraint on the vehicle speed. After the soft constraints are formed, the soft constraints must be crossed to switch the preview distance mode.
The prediction equation of the model prediction controller can be obtained by integrating the two-degree-of-freedom dynamic model and the preview error model, and the expression is as follows:
and respectively estimating the tire cornering stiffness and the road adhesion coefficient by adopting an adaptive forgetting factor recursive least square estimator. The formula of the recursive least squares with forgetting factor is as follows:
z(k)＝h T (k)θ(k)+e(k)
wherein z (K) is the system output at time K, h (K) is the system input, θ (K) is the parameter to be identified, e (K) is the measurement noise, K (K) is the algorithm gain, P (K) is the covariance matrix, λ m Is a forgetting factor.
In the estimation of the cornering stiffness of the tire, z (k) is F yf And F yr H (k) are each alpha f And alpha r And each of θ (k) is C f And C r . The tire lateral force and the cornering angle are input as known quantities to an estimator, and then an estimated value of cornering stiffness is obtained from the estimated quantities.
When the tire is in the low slip ratio range, the road surface adhesion ratio and the slip ratio are approximately proportional, so the following expression can be obtained:
wherein, F x Is the tire longitudinal force, F z Is the tire vertical force, k r Is the slope of the adhesion versus slip ratio curve in the low slip ratio range, and s is the tire slip ratio.
The slip ratio is defined as follows:
wherein, R is the radius of the wheel, omega is the rolling angular velocity of the wheel, and v is the velocity of the center of the wheel. In the estimation of the road surface adhesion coefficient,
h (k) is s, and theta (k) is k r The tire longitudinal force, the vertical force and the tire slip ratio are calculated as known amounts, and after the slope of the adhesion ratio-slip ratio curve is calculated, the road surface adhesion coefficient is calculated by the following formula:
μ＝k r s m p
wherein μ is a road surface adhesion coefficient, s m P is a proportionality coefficient of the maximum road surface adhesion coefficient to the peak road surface adhesion coefficient in the linear region.
The forgetting factor is used for distributing the weight of new and old data, and is usually 0.98, but when the error of the identification parameter is very small, the error of the online identification parameter can be increased by introducing the forgetting factor, and when the error of the online identification is very large, the forgetting factor needs to be optimized, so that the online identification has faster convergence speed, and the identification error is reduced, therefore, the forgetting factor should be designed to be capable of adaptively changing along with the error, and the following expression for calculating the adaptive forgetting factor is provided for meeting the above requirements:
λ(k)＝λ min +(1-λ min )h ε(k)
wherein λ (k) is a forgetting factor at time k, λ min Is the minimum value of the forgetting factor, and h is a sensitivity coefficient which represents the sensitivity of the forgetting factor to errors. e (k) is the error at time k, e base Is an allowable error reference.
By replacing λ by λ (k) m Finally obtainThe expression to the adaptive forgetting factor recursive least squares is as follows:
the transverse controller adopts a model prediction controller, and writes a prediction equation into a state equation form:
wherein, the first and the second end of the pipe are connected with each other,
is the rate of change of the system state variable, x is the system state variable, y is the system output variable, u is the control input, w is the disturbance variable, A c B c D c C c For the coefficient matrix, the specific calculation is as follows:
u＝[δ f ,M z ] T
w＝κ
discretizing the state equation can obtain the following expression:
wherein, A = I + A c T s ，B＝B c T s ，D＝D c T s ，T s The controller sample time.
In order to control and constrain the control increment conveniently, the state quantity and the control quantity are combined to be used as a new system state variable to obtain a new state equation, and the expression is as follows:
wherein the content of the first and second substances,
for the new coefficient matrix, the specific calculation is as follows:
the prediction equation for the vehicle future state and system output is as follows:
in the above-mentioned formula, the compound of formula,
ΔU(k)＝[Δu(k),Δu(k+1),...Δu(k+N c -1)] T
W(k)＝[w(k),w(k+1),…,w(k+N c -1)] T
wherein, N p Predicting time domain, N, for a model predictive controller c The time domain is controlled for the model predictive controller.
To ensure that the lateral controller can accurately track the trajectory and ensure stability, the following objective function is established:
wherein Y is ref (k) Indicating the expected value of the system output at time k, Y ref (k) Y (k) represents the system output error, Q and R are the weights of the error and control increment, respectively, p is the weight coefficient, and ε is the relaxation factor.
The controller constraint design is:
δ fmin ≤δ f ≤δ fmax Δδ fmin ≤Δδ≤Δδ max
M zmin ≤M z ≤M zmax ΔM zmin ≤ΔM z ≤ΔM zmax
wherein, delta fmin And delta fmax Minimum and maximum values of the angle of rotation of the front wheel, M zmin And M zmax Minimum and maximum values of the additional yaw moment, delta, respectively fmin And Δ δ max Minimum and maximum values, respectively, of the front wheel steering angle increment, Δ M zmin And Δ M zmax Respectively being an additional cross barMinimum and maximum pendulum moment increments.
The final objective function and constraint expression are as follows:
wherein, delta U min And Δ U max Minimum and maximum control increment, U min And U max The minimum value and the maximum value of the control quantity are respectively.
And the transverse controller obtains the optimal front wheel corner and the optimal additional yaw moment according to the objective function and the constraint solution. The control quantity constraint and the control increment constraint in the invention can be changed along with the road adhesion coefficient estimated by the self-adaptive forgetting factor recursive least square method. The road surface is dry and the constraint design is U min ，U max ，ΔU min ，ΔU max The road surface is wet and slippery, and the restraint is updated to 0.8U min ，0.8U max ，0.7ΔU min ，0.7ΔU max The road surface is ice and snow, and the restriction is updated to 0.5U min ，0.5U max ，0.4ΔU min ，0.4ΔU max 。
In order to take into account the stability of the vehicle while tracking the trajectory and taking into account the influence of the slip ratio of the tires, the invention proposes a torque distribution controller whose objective function is designed taking into account three points, first requiring that the sum of the torques of each wheel is equal to the total torque required to track the longitudinal speed, the total torque being the torque required to maintain the current vehicle longitudinal speed, determined by the error between the desired longitudinal speed and the actual longitudinal speed. Secondly, the driving force of each wheel should meet the additional yaw moment provided by the model predictive controller, and finally, the low energy consumption requirement of the intelligent automobile in the driving process is considered. The objective function for the optimal torque distribution is therefore designed as follows:
wherein HT-V representsTotal torque and additional yaw moment, sigma and Q, satisfied 1 For its weighting factor, T is in the form of a matrix of four wheel torques, T min And T max Minimum and maximum values of torque to meet elliptical limits of tire friction, H, T, V, R 1 The matrix expression of (c) is calculated as follows:
wherein d is the tread.
And converting the objective function into a quadratic programming problem, and solving the torque of each wheel. While the slip ratio is used to regulate the final torque output. The wheel longitudinal force has a stable region which increases from small to large and an unstable region which gradually decreases from a peak value along with the change of the slip ratio, and the optimal wheel slip ratio is estimated according to the characteristic, wherein the expression is as follows:
wherein, F x (k) And s (k) are the wheel longitudinal force and slip ratio, Δ F, respectively, at the moment k after discretization x And Δ s is the difference between the current time and the previous time when Δ F x S is equal to 0 and Δ s is greater than 0 d For optimal wheel slip rate.
The expression for the desired wheel roll angular velocity is derived from the optimal wheel slip ratio as follows:
wherein, ω is d The desired wheel roll angular velocity. The torque is adjusted by using sliding mode control, and the design expression and the approach law of the sliding mode surface are as follows:
wherein s is ω Is a slip-form surface and is characterized in that,
derivative of slip form surface, sat(s) ω ) For the saturation function, it is designed as follows:
where Δ is a small design parameter.
In the design of the approach law, the reduction of the coefficient epsilon can reduce the system buffeting phenomenon, but can lead the system to tend to be stable and reduce the speed. The torque at which the wheel slip ratio can ultimately be considered is expressed as follows:
T d ＝J w [ω d -ε ω sat(s ω )]+T
wherein, T d For the final action on the wheel torque of the intelligent vehicle, J w Is the moment of inertia of the wheel. Epsilon ω A coefficient greater than 0.
The invention has the beneficial effects that:
1. the self-adaptive preview error model provided by the invention considers the influence of a road adhesion coefficient on an intelligent automobile aiming at the problem that a conventional preview model only changes the preview distance according to the automobile speed, designs two preview distance switching modes based on the automobile speed and the road adhesion coefficient, participates in solving as a part of a model prediction controller prediction equation, and performs switching soft constraint at the switching point of the automobile speed and the road adhesion coefficient, thereby avoiding repeated switching modes and improving the system stability. The model is suitable for trajectory tracking under different vehicle speeds, takes into account the additional requirements of the intelligent vehicle on the pre-aiming distance under a low-adhesion road surface besides the vehicle speed, and improves the safety of the intelligent vehicle.
2. The adaptive forgetting factor recursive least square estimator provided by the invention estimates the tire cornering stiffness and the road adhesion coefficient on line aiming at the changeable road condition scene which the intelligent automobile may face actually, introduces the balance between the coordinated convergence speed and the identification error of the adaptive forgetting factor, effectively improves the accuracy of a prediction model, updates the control quantity and the constraint of the control increment along with the change of the road adhesion coefficient, and obviously improves the adaptive track tracking capability and the safety of the intelligent automobile under various working conditions.
3. The torque distribution controller provided by the invention provides a new torque distribution strategy aiming at the characteristics of four-wheel intelligent automobiles and tires, adjusts the output of the torque according to the slip rate, and improves the stability of the automobile while ensuring the tracking precision of the intelligent automobile compared with the conventional average torque distribution strategy.
Drawings
FIG. 1 is a two degree of freedom kinetic model of a vehicle;
FIG. 2 is an adaptive predictive error model;
FIG. 3 is a flow chart of the intelligent vehicle trajectory tracking and stability control method based on adaptive model predictive control according to the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings.
FIG. 1 is a two-degree-of-freedom dynamic model of a vehicle, wherein the vehicle only moves in an x-o-y plane without considering the vertical, roll and pitch motions of the vehicle. The vehicle is turned by the front wheel, a coordinate system of the vehicle body is positioned in a plane of bilateral symmetry of the vehicle, the origin of the center of mass of the vehicle is o, the x axis is the longitudinal axis of the vehicle, the positive direction of the x axis is the direction of the vehicle head, the y axis points to the lateral direction of the vehicle body, the positive direction meets the right-hand rule, and the positive direction of the z axis is vertical to the upper direction. Therefore, according to newton's second law, equations for rotational balance and force balance are established at the center of mass of the vehicle, resulting in the following expression:
wherein m is the vehicle mass, V x Is the longitudinal speed of the mass center under the coordinate system of the vehicle body, beta is the side deviation angle of the mass center of the vehicle, gamma is the yaw velocity of the vehicle, I z Is the moment of inertia of the vehicle about the z-axis, l f And l r Distances from the center of mass of the vehicle to the front axle and the rear axle, respectively, F yfl F yfr F yrl F yrr Fl denotes a left front wheel, fr denotes a right front wheel, rl denotes a left rear wheel, rr denotes a right rear wheel, M z An additional yaw moment.
From the above assumptions, it can be considered that the tire lateral force is proportional to the tire slip angle, resulting in the following expression:
wherein, F yf And F yr Resultant forces of wheel lateral forces acting on the front axle and rear axle of the vehicle, respectively, C f And C r Cornering stiffness, alpha, of the front and rear wheels of the vehicle, respectively f And alpha r Slip angles, delta, of the front and rear wheels of the vehicle, respectively f Is the corner of the front wheel.
The expression of the two-degree-of-freedom model is obtained as follows:
FIG. 2 is a prediction error model, since in general, the centroid slip angle β and its derivatives
The value is relatively small, close to0, the vehicle yaw angle can be considered to be approximately equal to the heading angle, so the preview error model expression is as follows:
wherein the content of the first and second substances,
is the desired heading angle of the pre-aim point>
For the current course angle of the vehicle, in combination with the vehicle speed>
Is the deviation of the heading angle of the pre-aiming point, is the difference between the desired heading angle and the current heading angle of the vehicle, e d For lateral deviation, e d0 Is the distance between the extension line of the pre-aiming point and the mass center of the current vehicle in the y-axis direction of the vehicle coordinate system, L p For the pre-aiming distance
Derived from the above equation:
where κ is the road curvature of the preview point.
The prediction equation of the model prediction controller can be obtained by integrating the two-degree-of-freedom dynamic model and the preview error model, and the expression is as follows:
FIG. 3 is a flow chart of an intelligent vehicle trajectory tracking and stability control method based on adaptive model predictive control. When the intelligent automobile tracks, firstly, an adaptive forgetting factor recursive least square estimator calculates the lateral deviation rigidity and the road adhesion coefficient of wheels on line according to relevant information collected by a sensor arranged on the intelligent automobile, a prediction model, a control quantity and control increment constraints are updated in real time, then, an adaptive pre-aiming error model receives a reference track given by an upper track planning module and the road adhesion coefficient information of the estimator, a pre-aiming distance is determined through a decision module, a transverse error, a pre-aiming point course error and a pre-aiming point curvature are calculated and transmitted to a model prediction controller, then, the model prediction controller obtains a front wheel corner and an additional yaw moment through optimal solution, the longitudinal controller compares the difference between an expected longitudinal speed and an actual speed to calculate the longitudinal force of the automobile, finally, a torque distribution controller carries out torque distribution, the influence of the wheel slip rate is considered, the torque of each wheel is adjusted to obtain the final wheel torque, and the wheel torque and the front wheel rotation angle control the intelligent automobile to track tracking and stability control.
The above-listed series of detailed descriptions are merely specific illustrations of possible embodiments of the present invention, and they are not intended to limit the scope of the present invention, and all equivalent means or modifications that do not depart from the technical spirit of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. An intelligent automobile trajectory tracking and stability control system based on adaptive model prediction control is characterized by comprising: the system comprises an adaptive preview error model, an adaptive forgetting factor recursive least square estimator, a transverse controller and a torque distribution controller;
the adaptive preview error model: different requirements of road curvature, road surface adhesion coefficient and vehicle speed on the preview distance are comprehensively considered, a switching mode and soft constraints are designed, and a self-adaptive preview error model is constructed;
an adaptive forgetting factor recursive least squares estimator: the method is used for accurately estimating the tire cornering stiffness and the road adhesion coefficient, and introducing a self-adaptive forgetting factor to coordinate the balance between the convergence speed and the identification error;
a transverse controller: the method comprises the steps that a model prediction controller is adopted to conduct transverse control, the model prediction controller takes an adaptive predictive error model and a vehicle two-degree-of-freedom model as prediction models, controlled variables are obtained by solving the models, the model is used for collecting state quantities of a vehicle, optimal solution is conducted on the basis of the established prediction models under the condition that a designed objective function and constraint are met, front wheel turning angles are solved to act on an actual vehicle, so that the minimum error between an actual running track and a reference track of the vehicle is achieved, and meanwhile an additional yaw moment is solved to serve as an expected value to be output to a torque distribution controller to conduct torque distribution, and stability control is achieved;
a torque distribution controller: and performing optimized distribution of the torque under the condition that the additional yaw moment and the longitudinal force output by the transverse controller meet the designed objective function and constraint, and considering the influence of the tire slip rate to act on the actual vehicle after optimization so as to realize stability control and assist the trajectory tracking of the vehicle.
2. The intelligent vehicle trajectory tracking and stability control system based on adaptive model predictive control as claimed in claim 1, wherein the vehicle two-degree-of-freedom model is:
the method is characterized in that the vehicle is turned by front wheels, a vehicle body coordinate system is located in a plane of bilateral symmetry of the vehicle, the original point of the center of mass of the vehicle is o, an x axis is a longitudinal axis of the vehicle, the positive direction of the x axis is the direction of a vehicle head, a y axis points to the lateral direction of the vehicle body, the positive direction of the y axis meets a right hand rule, the positive direction of a z axis is vertical and upward, and a rotation balance equation and a force balance equation are established at the center of mass of the vehicle according to a Newton second law to obtain the following expression:
wherein m is the vehicle mass, V x Is the longitudinal speed of the mass center under the coordinate system of the vehicle body, beta is the side deviation angle of the mass center of the vehicle, gamma is the yaw velocity of the vehicle, I z Is the moment of inertia of the vehicle about the z-axis, l f And l r Distances from the center of mass of the vehicle to the front axle and the rear axle, respectively, F yfl F yfr F yrl F yrr Respectively lateral forces acting on four wheels of the vehicle, M z An additional yaw moment;
the tire lateral force is proportional to the tire slip angle, and the following expression is obtained:
wherein, F yf And F yr Resultant of lateral forces acting on the wheels of the front axle and rear axle of the vehicle, respectively, C f And C r Cornering stiffness, alpha, of the front and rear wheels of the vehicle, respectively f And alpha r Slip angles, delta, of the front and rear wheels of the vehicle, respectively f Is the corner of the front wheel;
the expression for the two-degree-of-freedom model is obtained as follows:
wherein beta and
with values close to 0, it can be obtained that the vehicle yaw angle is approximately equal to the heading angle.
3. The intelligent automobile trajectory tracking and stability control system based on adaptive model prediction control as claimed in claim 1, wherein the preview error model expression is as follows:
wherein the content of the first and second substances,
is the expectation of the preview pointAngle of course->
Is the current heading angle of the vehicle, and>
is the deviation of the heading angle of the pre-aiming point, is the difference between the desired heading angle and the current heading angle of the vehicle, e d For lateral deviation, e d0 Is the distance between the extension line of the pre-aiming point and the current center of mass of the vehicle in the y-axis direction of the vehicle coordinate system, L p The pre-aiming distance is used;
from the above formula, one can obtain:
where κ is the road curvature of the preview point.
4. The system of claim 1, wherein the adaptive model predictive control-based intelligent vehicle trajectory tracking and stability control system comprises two pre-aiming distance switching modes based on road adhesion coefficient and vehicle speed:
dividing the road surface adhesion coefficient into three road surfaces, namely a dry road surface, a wet and slippery road surface and an ice and snow road surface;
on a dry road surface, a pre-aiming distance mode based on the vehicle speed is adopted, and the expression is as follows:
wherein L is pmin Is the minimum pre-aiming distance, L pmax Is the maximum preview distance, V min And V max Respectively minimum and maximum vehicle speed, T p Is the preview time;
when the road surface is slippery and icy, a pre-aiming distance mode based on the road surface adhesion coefficient is adopted, and the expression is as follows:
wherein L is p1 And L p2 Set to 20m and 23m, respectively.
5. The system of claim 4, further comprising: establishing boundary fuzzy control at the road surface adhesion coefficient boundary to form soft constraint of +/-0.1 on the road surface adhesion coefficient; and establishing boundary fuzzy control at the vehicle speed boundary, forming soft constraint of +/-2 m/s on the vehicle speed, and switching the preview distance mode only by crossing the soft constraint after the soft constraint is formed.
6. The intelligent vehicle tracking and stability control system based on adaptive model predictive control of claim 1, wherein the predictive equation of the model predictive controller is as follows:
7. the intelligent vehicle trajectory tracking and stability control system based on adaptive model predictive control of claim 1, wherein the adaptive forgetting factor recursive least squares estimator:
the formula of the recursive least squares with forgetting factor is as follows:
z(k)＝h T (k)θ(k)+e(k)
wherein z (K) is the system output at time K, h (K) is the system input, θ (K) is the parameter to be identified, e (K) is the measurement noise, K (K) is the algorithm gain, P (K) is the covariance matrix, λ (K) is the covariance matrix m Is a forgetting factor;
in the estimation of the cornering stiffness of the tire, z (k) is F yf And F yr H (k) are each alpha f And alpha r And each of θ (k) is C f And C r The tire lateral force and the cornering angle are used as known quantities and input to an estimator, and then an estimated value of the cornering stiffness is obtained through the estimator;
when the tire is in the low slip ratio range, the road surface adhesion ratio and the slip ratio are approximately proportional, so the following expression can be obtained:
wherein, F x Is a tire longitudinal force, F z Is the vertical force of the tire, k r The slope of the adhesion rate-slip rate curve in the low slip rate range is shown, and s is the tire slip rate;
the slip ratio is defined as follows:
where R is the wheel radius, ω is the wheel roll angular velocity, v is the velocity of the wheel center, in the estimation of the road adhesion coefficient,
h (k) is s, and theta (k) is k r After the gradient of the adhesion rate-slip rate curve is calculated as the tire longitudinal force, the vertical force and the tire slip rate are known, the road surface adhesion coefficient is calculated by the following formula:
μ＝k r s m p
wherein μ is a road surface adhesion coefficient, s m Is the maximum tire slip ratio in the linear region, and p is the proportionality coefficient of the maximum road surface adhesion coefficient and the peak road surface adhesion coefficient in the linear region;
the forgetting factor is used for distributing the weight of new and old data, and is designed to be adaptively changed along with the error, and the expression is specifically calculated:
λ(k)＝λ min +(1-λ min )h ε(k)
wherein, λ (k) is forgetting factor at k time, λ min Is the minimum value of the forgetting factor, and h is a sensitivity coefficient representing the sensitivity of the forgetting factor to errors. e (k) is the error at time k, e base Is an allowable error reference;
by replacing λ by λ (k) m Finally, the expression of the adaptive forgetting factor recursive least squares is obtained as follows:
8. the intelligent automobile trajectory tracking and stability control system based on adaptive model predictive control as claimed in claim 1 or 6, wherein the lateral controller adopts a model predictive controller, and writes a predictive equation into a state equation form:
wherein the content of the first and second substances,
is the rate of change of the system state variable,x is the system state variable, y is the system output variable, u is the control input, w is the disturbance variable, A c B c D c C c For the matrix, the specific calculation is as follows:
u＝[δ f ,M z ] T
w＝ρ
discretizing the state equation can obtain the following expression:
wherein, A = I + A c T s ，B＝B c T s ，D＝D c T s ；
Combining the state quantity and the control quantity as a new system state variable to obtain a new state equation, wherein the expression is as follows:
wherein the content of the first and second substances,
for the new matrix, the specific calculation is as follows:
the prediction equation for the vehicle future state and system output is as follows:
in the above-mentioned formula, the compound of formula,
ΔU(k)＝[Δu(k),Δu(k+1),…Δu(k+N c -1)] T
W(k)＝[w(k),w(k+1),…,w(k+N c -1)] T
the following objective function is established:
wherein, Y ref (k) Indicating the expected value of the system output at time k, Y ref (k) -Y (k) represents the system output error, Q and R are the weights of the error and control increment, respectively, p is the weight coefficient, and e is the relaxation factor;
the lateral controller constraint design is as follows:
δ fmin ≤δ f ≤δ fmax Δδ fmin ≤Δδ≤Δδ max
M zmin ≤M z ≤M zmax ΔM zmin ≤ΔM z ≤ΔM zmax
wherein, delta fmin And delta fmax Minimum and maximum values of the angle of rotation of the front wheel, M zmin And M zmax Minimum and maximum values of the additional yaw moment, delta, respectively fmin And delta max Minimum and maximum values, respectively, of the front wheel steering angle increment, Δ M zmin And Δ M zmax Minimum and maximum values of the additional yaw moment increment, respectively;
the final objective function and constraint expression are as follows:
s.t.ΔU min ≤ΔU≤ΔU max
U min ≤U≤U max
wherein, delta U min And Δ U max Minimum and maximum control increment, U min And U max Respectively a control quantity minimum value and a control quantity maximum value;
and the transverse controller obtains the optimal front wheel corner and the optimal additional yaw moment according to the objective function and the constraint solution.
9. An intelligent vehicle trajectory tracking and stability control system based on adaptive model predictive control as claimed in claim 1, wherein the objective function of the torque distribution controller is calculated as follows:
s.t.T min ≤T≤T max
where HT-V represents the total torque and additional yaw moment, σ and Q, required to be satisfied 1 For its weighting factor, T is in the form of a matrix of four wheel torques, T min And T max Minimum and maximum values of torque to meet elliptical limits of tire friction, H, T, V, R 1 The matrix expression of (a) is calculated as follows:
T＝[T fl T fr T rl T rr ] T V＝[T M z ] T
wherein d is the tread.
Converting the objective function into a quadratic programming problem, solving the torque of each wheel, adjusting the final torque output by using the slip ratio, along with the change of the slip ratio, the longitudinal force of the wheel has a stable region which is increased from small to large and an unstable region which is gradually reduced from a peak value, and estimating the optimal wheel slip ratio according to the characteristic, wherein the expression is as follows:
wherein, F x (k) And s (k) are the wheel longitudinal force and slip ratio, Δ F, respectively, at the moment k after discretization x And Δ s is the difference between the current time and the previous time, when Δ F x S is equal to 0 and Δ s is greater than 0 d The optimal wheel slip rate is obtained;
the expression for the desired wheel roll angular velocity is derived from the optimal wheel slip ratio as follows:
wherein, ω is d For expecting the rolling angular speed of the wheel, the sliding mode control is used for adjusting the torque, and the design expression and the approaching law of the sliding mode surface are as follows:
in the design of the approach law, the reduction of the coefficient epsilon can reduce the system buffeting phenomenon, but the system tends to be reduced in stable speed, and finally the torque of the wheel slip rate can be considered, wherein the expression is as follows:
T d ＝J w [ω d -ε ω sat(s ω )]+T
wherein, T d Wheel torque for ultimate effect on smart car, J w Is the moment of inertia of the wheel.
10. An intelligent automobile track tracking and stability control method based on adaptive model prediction control is characterized in that when an intelligent automobile tracks, firstly, an adaptive forgetting factor recursive least square estimator calculates wheel cornering stiffness and road adhesion coefficient on line according to relevant information collected by a sensor carried by the intelligent automobile, a prediction model, a control quantity and control increment constraint are updated in real time, then, an adaptive pre-aiming error model receives a reference track given by an upper track planning module and road adhesion coefficient information of the estimator, a pre-aiming distance is determined through a decision module, a transverse error, a pre-aiming point course error and a pre-aiming point curvature are calculated and transmitted to a model prediction controller, the model prediction controller obtains a front wheel corner and an additional yaw moment through optimal solution, the longitudinal controller compares a difference value between an expected longitudinal speed and an actual speed, calculates a vehicle longitudinal force and transmits the vehicle longitudinal force to a torque distribution controller, and finally, the torque distribution controller performs torque distribution and considers the influence of a wheel slip rate, adjusts the torque of each wheel to obtain final wheel torque, and controls the intelligent automobile to track tracking and stability control together with the front wheel corner.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211594290.8A CN115933662A (en) | 2022-12-13 | 2022-12-13 | Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211594290.8A CN115933662A (en) | 2022-12-13 | 2022-12-13 | Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115933662A true CN115933662A (en) | 2023-04-07 |
Family
ID=86652111
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211594290.8A Pending CN115933662A (en) | 2022-12-13 | 2022-12-13 | Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115933662A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117806175A (en) * | 2024-03-01 | 2024-04-02 | 北京理工大学 | Error self-learning track tracking control method and system for distributed driving vehicle model |
-
2022
- 2022-12-13 CN CN202211594290.8A patent/CN115933662A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117806175A (en) * | 2024-03-01 | 2024-04-02 | 北京理工大学 | Error self-learning track tracking control method and system for distributed driving vehicle model |
| CN117806175B (en) * | 2024-03-01 | 2024-04-30 | 北京理工大学 | Error self-learning track tracking control method and system for distributed driving vehicle model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109318905B (en) | Intelligent automobile path tracking hybrid control method | |
| CN111806427B (en) | Comprehensive control method for four-hub motor driven vehicle | |
| CN111890951B (en) | Intelligent electric automobile trajectory tracking and motion control method | |
| CN109795502B (en) | Intelligent electric vehicle path tracking model prediction control method | |
| CN110356404B (en) | Intelligent driving system with autonomous lane changing function and improved lateral safety | |
| CN107161207B (en) | Intelligent automobile track tracking control system and control method based on active safety | |
| Lee et al. | Adaptive vehicle traction force control for intelligent vehicle highway systems (IVHSs) | |
| CN109291932B (en) | Feedback-based electric vehicle yaw stability real-time control device and method | |
| CN111923908A (en) | Stability-fused intelligent automobile path tracking control method | |
| CN104773170A (en) | Stability integrated control method of vehicle | |
| CN111959500B (en) | Automobile path tracking performance improving method based on tire force distribution | |
| CN114967475B (en) | Unmanned vehicle trajectory tracking and stability robust control method and system | |
| CN113320542A (en) | Tracking control method for automatic driving vehicle | |
| CN113221257B (en) | Vehicle transverse and longitudinal stability control method under extreme working condition considering control area | |
| CN114684199A (en) | Vehicle dynamics series hybrid model driven by mechanism analysis and data, intelligent automobile trajectory tracking control method and controller | |
| CN116552550A (en) | Vehicle track tracking control system based on parameter uncertainty and yaw stability | |
| CN115933662A (en) | Intelligent automobile trajectory tracking and stability control system and method based on adaptive model prediction control | |
| Li et al. | Adaptive sliding mode control of lateral stability of four wheel hub electric vehicles | |
| CN114987537A (en) | Neural network dynamics-based road adaptive drift control system and method for automatic driving vehicle | |
| CN114454871A (en) | Unmanned platform stable tracking control method for four-wheel independent drive | |
| CN113183953B (en) | Active safety control method and system for vehicle after collision based on distributed driving chassis | |
| CN112829766B (en) | Adaptive path tracking method based on distributed driving electric vehicle | |
| CN116834754A (en) | Transverse and longitudinal cooperative control method for self-adaptive speed regulation of automatic driving vehicle | |
| CN116560371A (en) | Self-adaptive model predictive control-based automatic driving vehicle path tracking method | |
| Shen et al. | Stability and Maneuverability Guaranteed Torque Distribution Strategy of ddev in handling limit: a novel lstm-lmi approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |