CN115743174A - Autonomous driving vehicle trajectory planning and tracking control method considering active safety - Google Patents
Autonomous driving vehicle trajectory planning and tracking control method considering active safety Download PDFInfo
- Publication number
- CN115743174A CN115743174A CN202211439635.2A CN202211439635A CN115743174A CN 115743174 A CN115743174 A CN 115743174A CN 202211439635 A CN202211439635 A CN 202211439635A CN 115743174 A CN115743174 A CN 115743174A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- potential field
- max
- ltr
- pedestrian
- 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
Landscapes
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
An autonomous driving vehicle track planning and tracking control method considering active safety relates to the technical field of automobile safe driving, and solves the problems of vehicle safety, pedestrian safety and the like when an autonomous driving vehicle which runs at a high speed in complex scenes such as a wet and slippery road surface, an emergency obstacle avoidance and an emergency pedestrian avoidance avoids obstacles; and setting constraint conditions, adopting a prediction model to predict the vehicle state at the k + i moment according to the vehicle state at the k moment and the control quantity at the k moment, and then taking the minimum objective function value as an optimization target and the set constraint conditions to obtain the control quantity meeting the constraint conditions. The invention improves the side-longitudinal safety of the vehicle, ensures the safety of pedestrians, has good real-time performance and realizes the safe obstacle avoidance of the vehicle in various complex scenes.
Description
Technical Field
The invention relates to the technical field of safe driving of automobiles, in particular to an autonomous driving vehicle trajectory planning and tracking control method considering active safety.
Background
With the increasing of automobile reserves year by year, the road traffic safety problem is gradually highlighted, and because passive safety technologies such as anti-collision beams, bumpers, pedestrian anti-collision foams and the like only reduce loss as much as possible when an accident happens, people pay more and more attention to the active safety technology of the vehicle, namely, the vehicle actively takes measures to reduce safety risks, so that the traffic accident is avoided. The safety risks of a running vehicle mainly include: firstly, in complex traffic environments such as pedestrians, multi-obstacle vehicles and the like, the traffic environment with dynamic change is not sufficiently analyzed, so that collision with other vehicles is caused; secondly, under wet and slippery road surfaces and emergency conditions in rainy and snowy days, the vehicles running at high speed are easy to have the dangers of rear-end collision, sideslip, side turning, lane departure and the like; thirdly, the pedestrian crosses the lane without rules, and the vehicle can not avoid in time, resulting in collision accidents. Although various active safety technologies are widely applied to vehicles, such as adaptive cruise (ACC) and Automatic Emergency Braking (AEB) for improving the longitudinal driving safety of the vehicle, and control units such as an Electronic Stability Program (ESP) and an Emergency Steering Assist (ESA) for improving the lateral stability of the vehicle, the control units are mainly concentrated on an execution level, under the large background of the development of automatic driving, the active safety control needs to coordinate the track planning of an upper layer, and the interactive coupling and mutual restriction between the control units respectively used for longitudinal and lateral safety cause that the overall performance of the vehicle cannot be optimized.
Huang proposes a method that uses a Model Predictive Controller (MPC) in combination with an artificial potential field, based on a vehicle kinematics model. The method adds the traffic environment potential field into the objective function, solves through multi-objective multi-constraint optimal control, simultaneously performs trajectory planning and tracking control, and realizes the side-longitudinal coupling control of the vehicle. The Xu Yang and Li H adopt a vehicle dynamic model on the basis of Huang, and improve the high-speed stability of the vehicle while realizing the control of the lateral-longitudinal coupling motion of the vehicle. The Snapper designs a Gaussian-like function to describe the potential field of the traffic environment on the basis of Huang, the potential field function is approximately processed by using a second-order Taylor formula before optimization solution, the nonlinear programming problem is converted into a standard quadratic programming problem to be solved, optimization solution time is shortened, and obstacle avoidance real-time performance is improved.
The method simultaneously carries out obstacle avoidance trajectory planning and tracking control, realizes the side-longitudinal coupling control of the vehicle, improves the dynamic obstacle avoidance capability of the vehicle, does not consider the safety problem of the vehicle and pedestrians when the vehicle is autonomously driven to avoid obstacles under a complex traffic scene, and influences the optimization solving efficiency of the MPC due to a complex potential field function, thereby being not beneficial to timely obstacle avoidance of the vehicle. Although the calculation efficiency can be improved to a certain extent by adopting a second-order Taylor formula approximation method, the control precision can be directly influenced by the generated error, and the safety of people and vehicles in a complex traffic environment is not facilitated.
Disclosure of Invention
The invention provides an autonomous driving vehicle track planning and tracking control method considering active safety, aiming at solving the problems of vehicle safety, pedestrian safety and the like when an autonomous driving vehicle which runs at a high speed in complex scenes such as a wet and slippery road surface, an emergency obstacle avoidance and an emergency pedestrian avoidance avoids obstacles.
An autonomous driving vehicle trajectory planning and tracking control method considering active safety is realized by the following steps:
acquiring traffic environment information by a vehicle-mounted sensor, establishing a traffic environment potential field according to the traffic environment information, and adding the traffic environment potential field into a target function of an MPC controller;
secondly, setting constraint conditions in the MPC controller, wherein the constraint conditions comprise the constraint on lateral stability indexes and control quantities of the vehicle;
thirdly, the MPC controller takes a dynamic model of the vehicle as a prediction model, the prediction model predicts the vehicle state at the k + i moment according to the vehicle state at the k moment and the control quantity at the k moment, then takes the minimum objective function value as an optimization target, obtains the control quantity meeting the constraint conditions, namely the front wheel rotation angle and the expected acceleration of the vehicle, through the optimization solution of the MPC according to the constraint conditions set in the second step, and finally controls the vehicle to complete safe obstacle avoidance;
and taking the minimum value of the objective function as an objective function of an optimization objective, and expressing the minimum value of the objective function as follows:
in the formula, gamma 1 、Γ 2 、Γ 3 、Γ 4 R, S, ρ is the weight of each subentry; k is the current time, N p And N c Respectively control time domain and prediction time domain, P road As a function of the road potential field, P car,j As a function of the potential field of the obstacle vehicle, P ped,p As a function of the pedestrian potential field, P goal As a function of the potential field of the target point, v x Is the main velocity, v ref For a desired barrier-free driving speed, Δ u is a control increment, ε q Relaxation factors for parameter terms in constraints
The invention has the beneficial effects that: the control method provided by the invention is based on a vehicle dynamics model, considers the safety targets of vehicles and pedestrians to design a traffic environment potential field and an MPC, combines the potential field and the MPC, adopts SQP to carry out optimization solution in the MPC to directly obtain a control quantity, enables the trajectory planning and tracking control of the autonomous driving vehicle to be carried out simultaneously, and simultaneously coordinates a plurality of side-longitudinal safety targets under various complex scenes, thereby effectively solving the safety problems of the vehicles and the pedestrians when the autonomous driving vehicle avoids obstacles under the complex scenes, and has the following advantages:
1. aiming at the problem of lateral stability of a vehicle running at a high speed under complex scenes such as a wet and slippery road surface, an emergency obstacle avoidance and an emergency pedestrian avoidance, lateral stability constraints such as a transverse load transfer rate of the vehicle are added into the MPC, the risks of side turning and side slipping of the vehicle are reduced, and the lateral safety of the vehicle is improved.
2. The method aims at the longitudinal safety problems of rear-end collision and the like of the vehicle in a complex scene, and considers the longitudinal safety distance of the vehicle to design the potential field of the obstacle avoidance vehicle, so that the autonomous driving vehicle always keeps the safety distance with the front vehicle in the obstacle avoidance process, and the longitudinal safety of the vehicle is guaranteed.
3. Aiming at the problem of pedestrian collision avoidance when pedestrians pass through the lane illegally, the pedestrian potential field is designed by considering the safe distance between the pedestrians and the vehicles, and the pedestrians are avoided by adopting a steering and braking synergistic action mode, so that the vehicles always keep the safe distance with the pedestrians in front in the obstacle avoidance process, and the safety of the pedestrians is guaranteed.
4. In order to enable the vehicle to avoid the obstacle in time under the emergency situation, the invention adopts the Gaussian function with continuous gradient and the quadratic polynomial to establish the potential field of the traffic environment, which is beneficial to the optimized solution of the MPC, and the MPC adopts the SQP to carry out the optimized solution, thereby ensuring the control precision, reducing the calculation time of the optimized solution and improving the real-time performance of the vehicle obstacle avoidance.
5. The invention improves the side-longitudinal safety of the vehicle, ensures the safety of pedestrians, has good real-time performance and further realizes the safe obstacle avoidance of the vehicle in various complex scenes.
Drawings
FIG. 1 is a schematic block diagram of an autonomous driving vehicle trajectory planning and tracking control method considering active safety according to the present invention.
Detailed Description
The embodiment is described with reference to fig. 1, and an active and safe autonomous driving vehicle trajectory planning and tracking control method is considered, in which a control quantity meeting constraint conditions is obtained through an MPC optimization solution based on a vehicle dynamics model and an improved traffic environment potential field, trajectory decision and tracking control are performed simultaneously, and a plurality of side-longitudinal safety targets in the potential field model and the constraint conditions are coordinated, so that safety requirements such as inter-vehicle safety distance, inter-human-vehicle safety distance, lateral stability and the like are met.
The specific process of the embodiment is as follows:
acquiring traffic environment information by a vehicle-mounted sensor, establishing a traffic environment potential field according to the traffic environment information, and adding the traffic environment potential field into a target function of an MPC controller;
the traffic environment information comprises road boundaries, lane lines, obstacle vehicle information and pedestrian information in the surrounding traffic environment, the traffic environment potential field is established by combining the information and considering the safe distance between vehicles and between people and vehicles, and the traffic environment potential field is added into an objective function of the MPC controller.
In the embodiment, the traffic environment potential field of the safety target is mainly considered to describe the interaction between the host vehicle (i.e. the autonomous driving vehicle) and the surrounding environment thereof, when an obstacle exists, the host vehicle is guided to smoothly avoid the obstacle and move to the target point, and when no obstacle exists, the host vehicle is guided to run on the center line of the lane. In the past, many potential field function designs are complex and have discontinuous gradients, so that discontinuous changes of vehicle dynamics are caused, and the efficiency of subsequent optimization solution is influenced.
A Gaussian function and a quadratic function with continuous and simple gradients are adopted as potential field functions, so that a complex traffic environment can be described, and a smooth obstacle avoidance track can be planned, and the efficiency of subsequent optimization solution can be improved. In addition, in order to realize the safety target of avoiding rear-end collision between the main vehicle and the front vehicle and avoiding collision between illegal pedestrians, the potential field of the obstacle vehicle and the potential field of the pedestrians are designed respectively by considering the safety distance between the vehicle and the pedestrian during modeling, namely the action domain of the potential field is set according to the safety distance.
The building of the traffic environment potential field comprises an obstacle vehicle potential field considering safe vehicle distance and a pedestrian potential field considering safe distance between people and vehicles. The specific process of the obstacle potential field considering the safe distance of the vehicle is as follows:
the obstacle potential field considering the safe distance of the vehicles guides the main vehicle to change the lane in the obstacle avoidance process of the main vehicle so as to exceed the slower obstacle in front and keep the safe distance between the vehicles. The closer the host vehicle is to the obstacle vehicle, the larger the value of the potential field function should be. The longitudinal and lateral reach of the obstacle vehicle should be different, and the host vehicle may be closer to the obstacle vehicle in the lateral direction, and the reach is set in the longitudinal direction in consideration of the safe inter-vehicle distance of the vehicle.
And describing the potential field of the obstacle vehicle by using a two-dimensional Gaussian function with the longitudinal coordinate X and the lateral coordinate Y of the obstacle vehicle as variables, wherein the potential field function is continuous and has continuous gradient. The level set of the potential field function is an ellipse, and the main vehicle can be better guided to smoothly change the lane, because the trajectory planning is carried out along the outer contour of the ellipse level set when the main vehicle avoids the obstacle. When the distance between the main vehicle and the obstacle vehicle is reduced, the obstacle vehicle potential field value is exponentially increased, and then the track optimization will try to make the vehicle far away from the obstacle vehicle. The longitudinal and lateral ranges of the obstacle vehicles may be passed through a convergence factor δ in the potential field function X And delta Y To adjust. The obstacle vehicle potential field function is as follows:
wherein eta is car Representing the potential field coefficient of the obstacle vehicle, determining the maximum value, delta, of the potential field Y Is the lateral convergence factor of the potential field of the obstacle vehicle, affecting the lateral scope of the potential field, X and X car,j Indicating the position of the host vehicle and the obstacle vehicle in the X direction, Y and Y, respectively car,j Respectively indicate the positions of the host vehicle and the obstacle vehicle in the Y direction. Delta. For the preparation of a coating X Is a longitudinal convergence factor of the potential field of the obstacle vehicle, influences the longitudinal scope of action of the potential field, and has a value of a distance scaling factor k car And the product of the reference safe vehicle distance D:
δ X ＝k car D (2)
the reference safe vehicle distance D may be expressed as follows:
in the formula, v x As the speed of the host vehicle, a max For main braking deceleration, v obs For the speed of the obstacle vehicle, a max,obs Braking deceleration for obstacle vehicles, t 1 For sensing and response time, d 0 Is the minimum safe vehicle distance.
In this embodiment, the pedestrian potential field considering the safety distance between the people and the vehicle is specifically:
the traditional pedestrian potential field adopts a potential field function with a circular level set, when a longitudinal action area is too small, a vehicle cannot decelerate and turn to avoid obstacles earlier, so that the vehicle can only react when being close to a pedestrian, the turning is over-urgent, and instability is easy to occur; when the lateral scope of action is too large, the reasonable steering obstacle avoidance trajectory cannot be obtained through optimization. The pedestrian potential field function with the level set being elliptic and the longitudinal and lateral action areas being adjustable is adopted, so that the safety distance between the main vehicle and pedestrians can be ensured, and the main vehicle is guided to avoid the pedestrians in advance. The pedestrian potential field function is as follows:
wherein eta is ped Representing the potential field coefficient of the pedestrian, determining the maximum value, delta, of the potential field py Is the lateral convergence factor of the pedestrian potential field, affecting the lateral scope of action of this potential field, δ px Is the longitudinal convergence factor of the pedestrian potential field, affecting the longitudinal scope of action, X and X, of the potential field Ped,p Respectively showing the positions of the host vehicle and the pedestrian in the X direction, Y and Y Ped,p The positions of the host vehicle and the pedestrian in the Y direction are indicated, respectively.
δ py Receive lateral scaling factor k py Safe distance D from the side of the man car py The influence, the relationship between them can be expressed as follows:
δ py ＝k py D py (5)
wherein the lateral safety distance D py Radius of the person receiving the traffic R p (with a circular area representing pedestrian occupancy of the road) and a vehicle width W c Vehicle to pedestrian lateral safety margin D m And (4) influence.
δ px By longitudinal distance scaling factor k px Longitudinal arrangement with man and vehicleFull distance D px The influence, the relationship between them can be expressed as follows:
δ px ＝k px D px (7)
wherein the longitudinal safety distance D px Speed v of the subject vehicle x And the lateral acceleration a of the main vehicle y Critical lateral safety distance D py Influence.
Secondly, setting constraint conditions in the MPC controller, wherein the constraint conditions comprise the constraint on lateral stability indexes and control quantities of the vehicle;
in this embodiment, the MPC controller considering the safety objective specifically includes:
vehicles in a complex scene often run at high speed, and can meet the complex conditions of rain, snow, pedestrians, multi-obstacle vehicles and the like, the dynamic nonlinear constraint conditions such as control input to a vehicle execution mechanism, slippage caused by friction between tires and the ground, side inclination caused by lateral acceleration and the like are more severe than those in a common scene, and the requirements on side-longitudinal coupling motion control of the vehicles are higher; and the autonomous driving vehicle has higher real-time requirement on the controller when running at high speed, and the vehicle needs to complete trajectory planning and tracking control in a sampling period in the face of a surrounding dynamically-changed traffic environment. The conventional obstacle avoidance method has insufficient consideration on actual conditions and safety, a single scene is relatively obtained, and the real-time obstacle avoidance consideration is insufficient.
On the basis of the established traffic environment potential field model, lateral stability indexes such as a cross-load transfer rate and the like and control quantity are restrained in the MPC, and a safety target for avoiding the vehicle from generating side turning and sideslip under various complex scenes is achieved. In the solving process, the nonlinear dynamics model and the nonlinear constraint are linearized, and finally the SQP is selected for optimization solving, so that the real-time performance of the MPC controller is improved. Because the optimized solution directly obtains the corner and the acceleration of the front wheel as the control quantity, the side-longitudinal coupling motion control of the vehicle is realized.
In this embodiment, the lateral stability constraint considering the vehicle is specifically:
the control target of the MPC controller is to ensure that the vehicle can avoid obstacles in time and simultaneously give consideration to the active safety of the vehicle, and the MPC controller converts the active safety of the vehicle into the constraint on the main state of the vehicle and the output of the controller. The calculation of the optimal control quantity is completed through the optimization solution of the performance index constraint condition, and a relaxation factor epsilon is introduced into the inequality constraint to ensure that a feasible solution can be solved under the constraint condition.
Vehicle lateral stability indicators include: lateral load transfer rate LTR, lateral acceleration a y The centroid slip angle beta and the yaw angular velocity r, and the inequality constraint forms are as follows:
LTR min +ε LTR V LTRmin ≤LTR≤LTR max +ε LTR V LTR max (9)
β min +εγV βmin ≤β≤β max +εβV βmax (11)
r min +ε r V r min ≤r≤r max +ε r V r max (12)
in the formula, LTR min Is the minimum value of the transverse load transfer rate LTR, epsilon LTR Relaxation factor, ε, for the transverse load transfer rate LTR V LTRmin The minimum relaxation term for the cross-load transfer rate; LTR max The maximum value of the cross-loading transfer rate LTR; epsilon LTR V LTR max The maximum relaxation term for the cross-load transfer rate; to ensure that the controller can obtain a feasible solution.
a ymin Is a lateral acceleration a y The minimum value of (a) is determined,
relaxation factor, a, for lateral acceleration y max Is a lateral acceleration a y Maximum value of (2)；
Is the minimum relaxation term for the lateral acceleration,
a maximum relaxation term for lateral acceleration;
β min is the minimum value of the centroid slip angle beta, epsilon β Relaxation factor of centroid slip angle beta, beta max Is the maximum value of the centroid slip angle beta, epsilon β V βmin Is the minimum relaxation term of centroid slip angle, epsilon β V βmax The maximum relaxation term for the centroid slip angle;
r min is the minimum value of the yaw rate r, epsilon r Relaxation factor r of yaw rate r max Is the maximum value of the yaw rate r, ε r V r min Is the minimum relaxation term of the yaw rate, ε r V r max The maximum relaxation term for the yaw rate.
In the present embodiment, the constraints of the controlled variable are the following inequality constraints to be satisfied, in which the output η, the controlled variable u, and the control increment Δ u of the prediction model are taken into consideration:
η min ≤η≤η max (13)
u min +ε u V u min ≤u≤u max +ε u V u max (14)
Δu min +ε Δu V Δu min ≤Δu≤Δu max +ε Δu V Δu max (15)
in the formula eta min And η max Minimum and maximum values of the output η, u min And u max Respectively, the minimum value and the maximum value, epsilon, of the control quantity u u To control the relaxation factor of the quantity u,. Epsilon u V u min And ε u V u max A minimum relaxation term and a maximum relaxation term of the control quantity u are respectively;
Δu min and Δ u max Minimum and maximum values, ε, of the control increment Δ u, respectively Δu To control the incremental delta u relaxation factor, epsilon Δu V Δu min And ε Δu V Δu max The control increments Δ u minimum slack term and maximum slack term, respectively.
And step three, the MPC controller takes a dynamic model of the vehicle as a prediction model, the prediction model predicts the vehicle state at the future moment according to the current vehicle state and the future control output, takes the established minimum objective function value as an optimization target and meets constraint conditions, and obtains control quantities, namely a front wheel corner and an expected acceleration through optimization solution, so that the vehicle is controlled to complete safe obstacle avoidance. The constraint conditions not only need to consider the dynamic constraint of the vehicle, but also need to consider the lateral stability constraint of the vehicle, namely, the lateral stability indexes such as the lateral load transfer rate, the lateral acceleration, the mass center lateral deviation angle, the yaw angular velocity and the like are constrained, so that the risks of side turning and side slipping of the vehicle are reduced, and the safety and the stability of the vehicle are improved. The linear transmission ratio is formed between the front wheel rotating angle and the steering wheel, and the acceleration controller adopts a fuzzy PID algorithm to obtain the accelerator opening and the brake pressure according to the existing results to realize acceleration control.
Therefore, the obstacle avoidance trajectory planning and tracking control method is essentially characterized in that vehicle safety obstacle avoidance and pedestrian collision avoidance are taken as main control targets, stability constraints such as cross-load transfer rate and the like are considered, and the optimal control solving problem with multiple targets and multiple constraints is achieved by taking the front wheel rotation angle and the longitudinal acceleration as control quantities.
In the present embodiment, the objective function having the minimum objective function value as the optimization objective is expressed by the following equation:
the objective function is mainly divided into four parts: the first part comprises the first four terms which are constituent elements of the main traffic environment potential field, and a prediction time domain N is calculated by combining a prediction model with a potential field function p The potential field value acting on the main vehicle. By introducing a traffic environment potential field to the targetIn the calibration function, a track with the minimum potential field value is searched to obtain an obstacle avoidance track, and obstacle avoidance track planning mainly depends on the potential field, so that the corresponding weight should be larger. The second section ensures that the host vehicle maintains the desired vehicle speed while traveling without obstruction. The third part is used for limiting the control increment, and aims to prevent the control increment from being greatly changed and reduce the motion amplitude of the actuator. The fourth part is a relaxation factor term in order to ensure that the controller can get a feasible solution. In the formula, gamma 1 、Γ 2 、Γ 3 、Γ 4 R, S, rho are the weights of each subentry; k is the current time, N p And N c Respectively control time domain and prediction time domain, P road As a function of the road potential field, P car,j As a function of the potential field of the obstacle vehicle, P ped,p As a function of the pedestrian potential field, P goal As a function of the potential field of the target point, v x Is the main velocity, v ref For a desired barrier-free driving speed, Δ u is a control increment, ε q Relaxation factors for parameter terms in constraints
(k + i | k) represents the prediction of time k to time k + i, and k following the symbol "|" represents that the current time is k.
Optimal control increment delta u at k moment is calculated by carrying out optimization solution by combining constraint conditions * (k) And further obtaining the optimal control quantity as follows:
u * (k)＝u * (k-1)+Δu * (k) (17)
in the formula u * (k) Control quantity for k time, u * And (k-1) is a control quantity at the time k-1.
In the embodiment, firstly, a Gaussian function and a quadratic function with continuous and simple gradients are adopted as potential field functions, so that smooth obstacle avoidance tracks can be planned, and meanwhile, in order to realize safe targets for avoiding rear-end collisions between a main vehicle and front vehicles and avoiding collisions between illegal pedestrians, safe distances between vehicles and between people and vehicles are respectively considered to design obstacle vehicle potential fields and pedestrian potential fields.
Secondly, lateral stability indexes such as cross-load transfer rate and the like and control quantity are restrained in the MPC, and the safety target of avoiding the vehicle from side turning and sideslip under a complex scene is achieved.
Finally, due to the adoption of the potential field function with continuous gradient, the SQP is adopted for optimization solution after the nonlinear dynamics model and the nonlinear constraint condition are linearized, and therefore the real-time performance of the MPC controller is improved. Simulation experiments verify that the method has good universality in different scenes, and the vehicles running at high speed can meet the cooperative requirements of multiple safety targets in multiple complex scenes when avoiding obstacles, so that the safe obstacle avoidance of the vehicles is realized.
Claims (6)
1. An autonomous driving vehicle track planning and tracking control method considering active safety is characterized in that: the method is realized by the following steps:
acquiring traffic environment information by a vehicle-mounted sensor, establishing a traffic environment potential field according to the traffic environment information, and adding the traffic environment potential field into a target function of an MPC controller;
secondly, setting constraint conditions in the MPC controller, wherein the constraint conditions comprise the constraint on lateral stability indexes and control quantity of the vehicle;
thirdly, the MPC controller takes a dynamic model of the vehicle as a prediction model, the prediction model predicts the vehicle state at the k + i moment according to the vehicle state at the k moment and the control quantity at the k moment, then takes the minimum objective function value as an optimization target, obtains the control quantity meeting the constraint conditions, namely the front wheel rotation angle and the expected acceleration of the vehicle, through the optimization solution of the MPC according to the constraint conditions set in the second step, and finally controls the vehicle to complete safe obstacle avoidance;
and taking the minimum value of the objective function as an objective function of an optimization objective, and expressing the minimum value of the objective function as follows:
in the formula, gamma 1 、Γ 2 、Γ 3 、Γ 4 、R、S and rho are weights of all the sub-items; k is the current time, N p And N c Respectively control time domain and prediction time domain, P road As a function of the road potential field, P car,j As a function of the potential field of the obstacle vehicle, P ped,p As a function of the pedestrian potential field, P goal As a function of the potential field of the target point, v x Is the main velocity, v ref For a desired barrier-free driving speed, Δ u is a control increment, ε q Is the relaxation factor of each parameter item in the constraint condition.
2. The active safety-considered autonomous driving vehicle trajectory planning and tracking control method of claim 1, characterized in that: the traffic environment information acquired in the first step comprises road boundaries, lane lines, obstacle vehicle information and pedestrian information.
3. The active safety-considered autonomous driving vehicle trajectory planning and tracking control method of claim 1, characterized in that: in the first step, the traffic environment potential field comprises an obstacle vehicle potential field considering a safe vehicle distance and a pedestrian potential field considering a safe distance between people and a vehicle;
the obstacle vehicle potential field considering the safe vehicle distance adopts a two-dimensional Gaussian function taking a longitudinal coordinate X and a lateral coordinate Y of the obstacle vehicle as variables as an obstacle vehicle potential field function, and the obstacle vehicle potential field function is expressed by the following formula:
in the formula eta car Is the potential field coefficient of the obstacle car, delta Y Lateral convergence factors, X and X, for the potential field of the obstacle vehicle car,j The positions of the main and obstacle vehicles in the X direction, Y and Y, respectively car,j The positions of the main vehicle and the obstacle vehicle in the Y direction respectively; delta X Is the longitudinal convergence factor of the potential field of the obstacle vehicle, and has a value of a distance scaling factor k car And reference safe vehicle distance D:
the pedestrian potential field considering the safety distance between the pedestrians and the vehicles is represented by a pedestrian potential field function with the level set being an ellipse and the longitudinal and lateral action areas being adjustable, and is represented by the following formula:
in the formula eta Ped Is the pedestrian potential field coefficient, δ py Is the lateral convergence factor, delta, of the pedestrian potential field px Longitudinal convergence factor, X, of pedestrian potential field Ped,p Position of pedestrian in X direction, Y Ped,p The position of the pedestrian in the Y direction.
4. The active safety-considered autonomous driving vehicle trajectory planning and tracking control method of claim 3, characterized in that:
the lateral convergence factor delta of the pedestrian potential field py And a lateral scaling factor k py And a safety distance D between the man and the vehicle py Is expressed by the following formula:
δ py ＝k py D py
a longitudinal convergence factor δ of the pedestrian potential field px Expansion factor k from longitudinal distance px And a safety distance D in the longitudinal direction of the man and the vehicle px Is expressed by the following formula:
δ px ＝k px D px 。
5. the active safety-considered autonomous driving vehicle trajectory planning and tracking control method of claim 1, characterized in that: in the second step, the vehicle lateral stability constraint comprises vehicle transverse load transfer rate LTR and lateral acceleration a y A centroid slip angle beta and a yaw angular velocity r; the specific constraint form is as follows:
LTR min +ε LTR V LTRmin ≤LTR≤LTR max +ε LTR V LTR max
β min +ε β V βmin ≤β≤β max +ε β V βmax
r min +ε r V r min ≤r≤r max +ε r V r max
in the formula, LTR min Is the minimum value of the transverse load transfer rate LTR, epsilon LTR Relaxation factor, ε, for the transverse load transfer rate LTR V LTRmin The minimum relaxation term for the cross-load transfer rate; LTR max The maximum value of the cross-loading transfer rate LTR; epsilon LTR V LTR max The maximum relaxation term of the cross-bearing transfer rate LTR;
a ymin is a lateral acceleration a y The minimum value of (a) is determined,
relaxation factor, a, for lateral acceleration y max Is a lateral acceleration a y Maximum value of (d);
is a lateral acceleration a y The minimum slack term of (a) is,
is a lateral acceleration a y The maximum relaxation term of (c);
β min is the minimum value of the centroid slip angle beta, epsilon β Relaxation factor β being the centroid slip angle β max Is the maximum value of the centroid slip angle beta, epsilon β V βmin Is the minimum relaxation term, ε, of the centroid slip angle β β V βmax The maximum relaxation term for the centroid slip angle β;
r min is the minimum value of the yaw rate r, ε r Relaxation factor r of yaw rate r max Is the maximum value of the yaw rate r, epsilon r V r min Is the minimum relaxation term, ε, of the yaw rate r r V r max The maximum relaxation term of the yaw rate r.
6. The active safety-considered autonomous driving vehicle trajectory planning and tracking control method of claim 1, wherein: in the second step, the constraints of the control quantity include an output quantity η, a control quantity u and a control increment Δ u of the prediction model, and the specific constraint form is as follows:
η min ≤η≤η max
u min +ε u V u min ≤u≤u max +ε u V u max
Δu min +ε Δu V Δu min ≤Δu≤Δu max +ε Δu V Δu max
in the formula eta min And η max Minimum and maximum values of the output η, u min And u max Respectively, the minimum value and the maximum value, epsilon, of the control quantity u u To control the relaxation factor of the quantity u,. Epsilon u V u min And epsilon u V u max A minimum relaxation term and a maximum relaxation term of the control quantity u are respectively;
Δu min and Δ u max Minimum and maximum values, ε, of the control increment Δ u, respectively Δu For controlling the delta u relaxation factor, ∈ Δ u V Δu min And ε Δu V Δu max The control increments Δ u minimum slack term and maximum slack term, respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211439635.2A CN115743174A (en) | 2022-11-17 | 2022-11-17 | Autonomous driving vehicle trajectory planning and tracking control method considering active safety |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211439635.2A CN115743174A (en) | 2022-11-17 | 2022-11-17 | Autonomous driving vehicle trajectory planning and tracking control method considering active safety |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115743174A true CN115743174A (en) | 2023-03-07 |
Family
ID=85372574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211439635.2A Pending CN115743174A (en) | 2022-11-17 | 2022-11-17 | Autonomous driving vehicle trajectory planning and tracking control method considering active safety |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115743174A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116203971A (en) * | 2023-05-04 | 2023-06-02 | 安徽中科星驰自动驾驶技术有限公司 | Unmanned obstacle avoidance method for generating countering network collaborative prediction |
-
2022
- 2022-11-17 CN CN202211439635.2A patent/CN115743174A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116203971A (en) * | 2023-05-04 | 2023-06-02 | 安徽中科星驰自动驾驶技术有限公司 | Unmanned obstacle avoidance method for generating countering network collaborative prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107943071B (en) | Formation maintaining control method and system for unmanned vehicle | |
| CN113386795B (en) | Intelligent decision-making and local track planning method for automatic driving vehicle and decision-making system thereof | |
| CN110187639B (en) | Trajectory planning control method based on parameter decision framework | |
| CN109032131B (en) | Dynamic overtaking obstacle avoidance method applied to unmanned automobile | |
| CN110471408B (en) | Unmanned vehicle path planning method based on decision process | |
| Yi et al. | Real time integrated vehicle dynamics control and trajectory planning with MPC for critical maneuvers | |
| Wurts et al. | Collision imminent steering at high speed using nonlinear model predictive control | |
| CN109669461A (en) | Automatic driving vehicle decision system and its method for planning track under a kind of complex working condition | |
| CN110703754A (en) | Path and speed highly-coupled trajectory planning method for automatic driving vehicle | |
| CN112947469A (en) | Automobile track-changing track planning and dynamic track tracking control method | |
| CN112904849A (en) | Integrated automatic driving automobile lane change track planning and tracking control method and system | |
| CN115743174A (en) | Autonomous driving vehicle trajectory planning and tracking control method considering active safety | |
| Song et al. | Longitudinal and lateral control methods from single vehicle to autonomous platoon | |
| Li et al. | Automatic lane change maneuver in dynamic environment using model predictive control method | |
| CN114510032A (en) | Multi-vehicle cooperative control method and device for guaranteeing rapid formation stability of vehicles | |
| Li et al. | Spatial-based predictive control for vehicle collision avoidance by steering maneuvers | |
| Zhang et al. | A novel simultaneous planning and control scheme of automated lane change on slippery roads | |
| Laheeb et al. | Online algorithm for controlling a cruise system under uncertainty in design parameters and environmental conditions using Monte-Carlo simulation | |
| Falsone et al. | Coordinated lane change in autonomous driving: a computationally aware solution | |
| Fan et al. | Intelligent vehicle lane-change trajectory planning on slippery road surface using nonlinear programming | |
| Ghariblu et al. | Trajectory planning of autonomous vehicle in freeway driving | |
| Ezeoma et al. | Model Predictive Control of Autonomous Vehicles to Enhance Driving Performance and Safety. | |
| Lelkó et al. | Integration of Robust Control with Reinforcement Learning for Safe Autonomous Vehicle Motion | |
| Bienemann et al. | Model Predictive Control for Autonomous Vehicle Following | |
| CN116300970B (en) | Autonomous vehicle formation method and device |
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 |