KR20180025611A - Autonomous driving system and autonomous driving method thereof - Google Patents

Abstract

The present invention relates to an autonomous driving system of a vehicle, and an autonomous driving method using the same, comprising: a sensing module (10) for sensing a physical state of an object vehicle; a power module (30) for controlling an engine torque of the vehicle, and controlling braking pressure of each wheel of the vehicle; and a control module (20) for controlling the sensing module (10) and the power module (30). The control module (20) can define a state of the vehicle based on a global coordinate system for a guide to destination and a fixed coordinate system of the vehicle so as to calculate a non-linear vehicle module with six degrees of freedom. Moreover, the control module (20) can attach a virtual safety limiting ellipse, calculated by a relation between the vehicle and the object vehicle measured by the sensing module (10), to the object vehicle.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to an autonomous driving system using an ellipse,

More particularly, the present invention relates to an autonomous driving system employing a nonlinear model predictive control method for path planning and motion planning, and an autonomous driving method using the autonomous driving system.

Recently, as the performance of various electronic devices and sensors of a vehicle is advanced and control technology is developed, an autonomous driving technology in which a driver perceives a driving environment by himself and operates to a target point is rapidly developed. Such an autonomous driving technique plans an optimal path without collision with other obstacles from a starting point to a target point, and carries out a motion planning process to follow the generated path.

This exercise plan is divided into a path planning for planning and generating a route to be driven by the vehicle and a path following for following the generated route. In general, the path planning is to measure the surrounding environment in real time through the Global Path Planning (Global Path Planning), which generates an optimal path based on the terrain and obstacle information of all the regions before driving, and the mounted environment sensor, And local path planning that partially regenerates the path through the path.

Such an autonomous driving technique is important to allow a vehicle to accurately and safely follow a path generated through various methods. Conventional path following methods include longitudinal control for controlling the speed of the vehicle and control of the steering angle And can be divided into lateral control. In general, the path follow-up uses a vehicle model or a kinematic model between the road and the vehicle to follow the generated path according to dynamic characteristics and constraints (lanes, obstacles, etc.) of the surrounding environment.

However, the conventional model predictive control technique guarantees the accuracy and stability of the control when the model used here perfectly matches the actual vehicle. By non-linear dynamics, the parameters are changed in real time There is an inevitable difference in performance that occurs when applied.

In addition, in the related art, uncertainty of the model is taken into consideration in performing the control through the model and it is not reflected in the autonomous driving method, and there is a problem that it is difficult to actively apply to the moving obstacle as it is designed considering only the static obstacle.

Korean Patent No. 10-1177374

SUMMARY OF THE INVENTION It is an object of the present invention to provide a nonlinear model-based control scheme for path planning and path estimation in a motion planning process of an autonomous vehicle, And control is performed to simulate close to the actual running.

Another object of the present invention is to calculate a virtual safety restriction ellipse for a target vehicle in an obstacle in real time and apply it to enhance stability in the autonomous driving process.

Another object of the present invention is to reduce the error between the actual vehicle operation and the model by applying the IMM technique or the like.

According to an aspect of the present invention, there is provided a vehicular drive system including a sensing module for sensing a physical state of a target vehicle, a power module for controlling an engine torque of the vehicle, And a control module for controlling the sensing module and the power module, wherein the control module defines a vehicle state based on a global coordinate system for guiding the vehicle to a destination and a fixed coordinate system of the vehicle, And the control module assigns a virtual safety restriction ellipse calculated by the relationship between the subject vehicle and the target vehicle measured by the sensing module to the target vehicle.

The state of the child vehicle

Is defined as the k denotes an axle brake rate, X, Y, V _x, V _y denotes a longitudinal / transverse distance and the speed of the global coordinate system, respectively, Ψ, γ denotes a respective yaw direction angle and angular velocity, m, I _mm , C _{f, r} , F _{xf, and x r} are the vehicle parameters and represent the mass, mass moment of inertia, front / rear tire rotational stiffness coefficient, and front / rear wheel longitudinal tire force, respectively.

The safety restriction ellipse is an elliptical shape in which the distance between the subject vehicle and the opponent vehicle is formed along the major axis, and the avoidable area by the safety restriction ellipse is

, And a and b are

Respectively,

here

,

,

to be.

The control module changes the local vehicle into the local route driving mode when the child vehicle enters the safety restriction tower.

When the relative speed between the vehicle and the opponent vehicle is equal to or greater than the reference value in the regional route running mode, the control module accelerates the vehicle in the longitudinal direction or in the lateral direction according to the driving state of the target vehicle.

The relative distance and the relative speed between the subject vehicle and the target vehicle are estimated through the IMM technique.

(I) the constant speed travel, (ii) the longitudinal direction acceleration / deceleration travel, or (iii) the lateral direction acceleration / deceleration travel according to the relative motion with the target vehicle, Transition pro- bability is defined through the Markov chain model.

In the lateral travel control of the child vehicle, the limit distance of the lateral movement is set as the lateral lane distance that occurs when the lane width of the road and one lane are changed, and the limit magnitude of the yaw rate Is set based on the maximum lateral acceleration and the speed of the subject vehicle.

According to another aspect of the present invention, there is provided a method for automatically running an automobile,

A step of setting a global path for guiding the subject vehicle to a destination, setting a global coordinate system for guiding the subject vehicle to a destination and a fixed coordinate system of the subject vehicle, Calculating a 6-degree-of-freedom non-linear vehicle module from the vehicle state, calculating a non-linear vehicle module from the calculated 6-degree of freedom non-linear vehicle module, calculating a traveling distance of the vehicle along the global route, And a step of starting a local route driving mode in which the subject vehicle is driven according to the driving of the target vehicle when the subject vehicle enters the safe limit ellipse do.

In the case of the local route running mode, when the relative speed between the subject vehicle and the target vehicle is equal to or higher than the reference value and there is no lane change of the target vehicle, the lateral acceleration running of the subject vehicle is performed, When there is a lane change of the target vehicle which is equal to or higher than the reference value, the vehicle is accelerated longitudinally in the longitudinal direction.

The autonomous vehicle driving system using the safety restriction ellipse according to the present invention as described above and the autonomous driving method using the same have the following effects.

According to the present invention, a nonlinear model-based control method for path planning and path estimation is used in the motion planning process of an autonomous vehicle, and calculation and control are performed at every sampling time to approximate the actual running, It is possible to bring about an autonomous driving effect close to the actual driving.

In addition, in the present invention, the safety of the vehicle can be improved by setting a safe limit tower between the subject vehicle and the target vehicle (dynamic obstacle). In particular, since the safety restriction tower is set by monitoring in real time the relative distance and relative speed between the target vehicle and the subject vehicle, more active and safe autonomous traveling can be achieved.

Also, in the present invention, the IMM technique and the Markov chain model are introduced to simulate the autonomous driving simulation as close as possible to the actual vehicle driving, thereby improving the autonomous driving accuracy.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic view showing components constituting an embodiment of an autonomous vehicle running system according to the present invention; Fig.
FIG. 2 (a) is a conceptual diagram showing a nonlinear model predictive control exercise plan used in a vehicle autonomous driving system using a safety restriction ellipse according to the present invention.
FIG. 2 (b) is a conceptual view showing a state in which a safety restriction ellipse used in an autonomous vehicle driving system using a safety restriction ellipse according to the present invention is applied.
3 is a flowchart sequentially showing an autonomous running method of a vehicle according to the present invention.
FIG. 4 is a flowchart sequentially showing a local route driving mode followed by an autonomous travel method of a vehicle according to the present invention shown in FIG. 3;
5 is an exemplary view showing an embodiment of local path autonomous traveling of a vehicle using the autonomous traveling method of the vehicle according to the present invention.
Fig. 6 is an exemplary diagram showing a simulation of the autonomous driving shown in Fig. 5 sequentially; Fig.
FIG. 7 is an exemplary view showing another embodiment of a local path autonomous traveling of a vehicle using the autonomous traveling method of the vehicle according to the present invention; FIG.
Fig. 8 is an exemplary view showing a simulation of the autonomous running shown in Fig. 7 sequentially; Fig.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference symbols as possible even if they are shown in different drawings. In the following description of the embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the understanding why the present invention is not intended to be interpreted.

In describing the components of the embodiment of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are intended to distinguish the constituent elements from other constituent elements, and the terms do not limit the nature, order or order of the constituent elements. When a component is described as being "connected", "coupled", or "connected" to another component, the component may be directly connected or connected to the other component, Quot; may be "connected," "coupled," or "connected. &Quot;

The autonomous traveling system of a vehicle according to the present invention is for effectively performing longitudinal and lateral control of a vehicle based on a nonlinear model predictive control technique for planning a motion of an autonomous vehicle. To this end, the present invention divides the route into two parts: a path planning for planning and creating a route to be traveled by the vehicle, and a path following for following the generated route. Global path planning, which generates an optimal route based on the terrain and obstacle information of all the areas to the destination before driving, and a sensing module 10 mounted on the vehicle, And local path planning that partially regenerates the path through measured and measured information.

Here, the global route plan is a general route plan to the destination, and the regional route plan is to carry out automatic running of the vehicle in relation to a dynamic obstacle, for example, another vehicle (hereinafter referred to as a 'destination vehicle'). Hereinafter, this concept will be described separately. A vehicle that is an object of automatic driving is referred to as a "vehicle", and a vehicle that is a dynamic obstacle is referred to as a "target vehicle".

The subject vehicle includes a sensing module 10, a power module 30, and a control module 20 for autonomous travel of the subject vehicle. The sensing module 10 is for sensing the state of the vehicle as well as the physical state of the target vehicle, and may include a RADAR, an LiDAR, an infrared sensor, a camera, and a GPS. In this case, the LiDAR method calculates the distance to the next vehicle by measuring the returning time of the laser beam, and it is effective at low speed and short distance. Therefore, the LiDAR sensor can be used to measure the relative distance and the relative speed with the target vehicle.

The power module 30 is for controlling the vehicle, and includes engine torque for speed control and braking pressure of each wheel. These power modules 30 are described again in the definition section for predicting the nonlinear model below .

In the present invention, the vehicle state is defined on the basis of the global coordinate system for guiding to the destination and the fixed coordinate system of the vehicle, and a 6-degree-of-freedom nonlinear vehicle module is calculated from the vehicle state. 6 DOF means linear and rotational motion in the X, Y, and Z axes, respectively. By constructing such a 6 DOF nonlinear vehicle model, it is possible to simulate more realistic vehicle control. For reference, FIG. 2 (a) is a conceptual diagram of a non-linear model predictive control motion plan used in an autonomous running system of a vehicle according to the present invention.

The state of the vehicle according to the present invention is as follows.

), 구동력(

), 제동력(

)이 있다. 이를 차량 모델에 제어 입력(

)으로 사용하기 위하여 구동력은 타이어의 유효반지름(

)과 변속기 및 최종감속기의 기어비(

)를 통해 엔진토크(

)로 변환하고, 동토크와 압력의 변환 상수(

)를 통해 각 바퀴의 제동압력(

)으로 바꾸어 사용하였다.The above equation represents the vehicle state equation (differential equation) and represents the vehicle model. It is used to apply the model-based control method. X, Y, V _x , and V _y denote longitudinal and transverse distances and velocities of the global coordinate system, respectively, where Ψ and γ denote yaw direction angles and angular velocities, respectively, and m, I _mm , C _{f, r} , F _{xf, and x r} are vehicle parameters, respectively, mass, mass moment of inertia, front / rear tire rotational stiffness coefficient, front / rear wheel longitudinal tire force. And u denotes a control input. The control input values include a front wheel steering angle (

), Driving force (

), Braking force

). This is input to the vehicle model (

), The driving force is the effective radius of the tire

Gear ratio of the transmission and the final reduction gear

) Through the engine torque

), And the conversion constant of the dynamic torque and the pressure (

) To the braking pressure of each wheel

).

Here, the engine torque and the braking pressure are as follows.

The equation defines the force the vehicle will travel in the longitudinal direction, where forward is converted to torque from the engine, and reverse or stop is converted to brake pressure.

However, since the nonlinear model predictive control technique has a large number of variables to be determined, it is disadvantageous to the computation time. Therefore, in the present invention, the nonlinear model predictive control method performs linearization using the nonlinear model state as the action point, and can use it as a constraint condition for the model predictive control.

In this embodiment, the lateral drive control of the subject vehicle can be set as follows. The limit distance of the lateral motion of the subject vehicle is set to a lateral movement distance that occurs when the lane width of the road and the one lane are changed and the restricted size of the yaw rate of the subject vehicle is the maximum lateral acceleration, As shown in FIG.

The above equation is an optimization formula used in the control method. It is an equation that minimizes the J (k) term and constructs an optimization problem that finds the optimal value satisfying the avoidable region below.

On the other hand, the control module of the present invention gives a virtual safety restriction ellipse calculated by the relationship between the subject vehicle and the target vehicle measured by the sensing module 10 to the target vehicle. The safety restraint tower is a virtual space that allows a self vehicle to safely and automatically run without risk of collision with the target vehicle, and is a long oval shape in the longitudinal direction. And the safety range for the safety restriction tower, that is, the avoidance area is as follows.

, And a and b are

Respectively,

here

,

,

to be.

That is, hx and hy denote the longitudinal time interval and the transverse time interval, respectively, and Dx and Dy are safety factors, which are constants, respectively. As shown in the above equation, the safety restriction ellipse is calculated by the relative distance and the relative speed with the target vehicle. FIG. 2 (b) shows a state where a safety restraining tower is applied to a target vehicle. Consequently, the above equation is an expression for expressing the outside of the ellipse, and is a formula showing the relationship between the distance between the vehicle and the obstacle vehicle and the ellipse.

In order to set the inequality constraint, we can linearize the above equation using the state of the nonlinear model at the current point as a point of action.

Meanwhile, the relative distance and the relative speed between the subject vehicle and the target vehicle are estimated through the IMM technique, thereby improving the accuracy. The IMM (Interacting Multiple Model Filter) technique is an interactive multi-model filter that estimates the position of a vehicle according to each model so as to more accurately estimate the position of a moving vehicle. The interaction multi-model filter includes a Kinematic Model Extended Kalman Filter for analyzing a model in which slip hardly occurs, and a dynamic model extension for analyzing a model in which slip occurs more frequently And a Kalman filter (Dynamic Model Extended Kalman Filter) so as to estimate the position of the vehicle in accordance with the driving environment of the vehicle. Since the IMM technique itself is general, a detailed description will be omitted.

In this embodiment, the local route operation control of the subject vehicle is performed in accordance with the relative movement with the target vehicle (i) at the constant speed, (ii) the longitudinal direction acceleration / deceleration, or (iii) And the transition probability between these operation patterns is defined through the Markov chain model.

Hereinafter, a method of automatically running an automobile according to the present invention will be described in order.

First, a step of setting a global route to guide a child vehicle to a destination is preceded (S100). Next, the state of the child vehicle is defined for nonlinear model prediction (S110). At this time, the vehicle state is defined on the basis of the global coordinate system for guiding the subject vehicle to the destination and the fixed coordinate system of the subject vehicle, and a six-degree-of-freedom nonlinear vehicle module is calculated from the formula described above .

Subsequently, the host vehicle travels along the global route. For example, the vehicle is autonomously driven along a predetermined route, and the information about the entire route and the fixed obstacle has already been input, so that autonomous travel is performed based on the information.

It is determined whether the dynamic obstacle is detected during the autonomous operation (S130). The detection of such an obstacle can be performed by the sensor module described above. If a dynamic obstacle is detected, a virtual safety restriction ellipse calculated by the relationship between the subject vehicle and the target vehicle is given to the target vehicle (S140). The safety restriction ellipse is derived from the foregoing equation, and the value is stored in the control module of the subject vehicle.

Then, it is determined in real time whether the subject vehicle falls within the range of the safety restriction ellipse (S150). If it is within the safety restriction ellipse, the mode is changed from the global route travel mode to the local route travel mode (S160). The regional route driving mode means that the automatic driving is performed actively according to the relationship between the target vehicle and the subject vehicle.

If it does not enter the safety restriction ellipse, the travel is made along the global path as before (S155) and the safety restriction ellipse is recalculated. Safety limit ellipses are calculated in real time, at regular time intervals. This is because the relative distance between the subject vehicle and the target vehicle and the relative speed change in real time.

When the mode is changed to the local route driving mode, the own vehicle is automatically driven according to the driving of the target vehicle (S170). The process includes a total of three modes of operation. That is, the local route travel control of the subject vehicle is selected as one of (i) constant speed travel, (ii) longitudinal direction acceleration / deceleration travel, or (iii) lateral direction acceleration / deceleration travel, . And the probability of mode change between them is defined by the Markov chain model.

More specifically, the relative speed between the host vehicle and the opponent vehicle is measured first. Then, it is determined whether the value is equal to or greater than a reference value (S180). Here, "above the reference value" means that the vehicle between the subject vehicle and the target vehicle approaches a certain speed or more. For example, if the speed of the vehicle is 80 km / h and the speed of the target vehicle is 60 km / h, the relative speed is -20 km. If the reference value is 20 km, the next step is reached.

If the relative speed is equal to or greater than the reference value, it is determined whether the lane of the target vehicle is changed (S190). If the relative speed between the subject vehicle and the target vehicle is equal to or higher than the reference value and there is a lane change of the target vehicle, the longitudinal acceleration running of the subject vehicle is performed (S200). On the contrary, when the relative speed between the subject vehicle and the target vehicle is equal to or higher than the reference value and there is no change in the lane of the target vehicle, the vehicle is accelerated in the lateral direction (S210).

That is, when the target vehicle changes the lane, the subject vehicle accelerates to the same lane, leading to the target vehicle, and the results of simulating the process are shown in FIG. 5 and FIG. When the target vehicle maintains the lane, the subject vehicle changes the lane to avoid the target vehicle, and the results of simulating the process are shown in FIGS. 7 and 8. FIG.

In the drawing, reference numerals 10A to 10C denote a vehicle, and 20A to 20C denote a target vehicle, respectively.

Of course, in the process of overtaking the target vehicle, the safety limit tower is calculated in real time, and the target vehicle is controlled not to enter the safety limit ellipse.

At this time, the time point of changing from the global route travel mode to the local route travel mode may be based on entry of the vehicle into the safe limit ellipse. That is, the local route driving mode in which the subject vehicle travels in accordance with the travel of the destination vehicle is started.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. That is, within the scope of the present invention, all of the components may be selectively coupled to one or more of them. Furthermore, the terms "comprises", "comprising", or "having" described above mean that a component can be implanted unless otherwise specifically stated, But should be construed as including other elements. All terms, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted to be consistent with the contextual meanings of the related art, and are not to be construed as ideal or overly formal, unless expressly defined to the contrary.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the scope of the present invention but to limit the scope of the technical idea of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

Claims (14)

A sensing module for sensing a physical state of the target vehicle,
A power module for controlling the engine torque of the vehicle and the braking pressure of each wheel,
And a control module for controlling the sensing module and the power module,
The control module defines a vehicle state based on a global coordinate system for guiding to a destination and a fixed coordinate system of the vehicle, calculates a six-degree-of-freedom nonlinear vehicle module from the vehicle state,
Wherein the control module uses a safety restriction ellipse that gives a virtual safety restriction ellipse calculated by the relationship between the subject vehicle and the target vehicle measured by the sensing module to the target vehicle.
2. The method of claim 1, wherein the state of the child vehicle

Lt; / RTI >
Wherein k denotes an axle brake _{rate, X, Y, V x,} V y denotes a longitudinal / transverse distance and the speed of the global coordinate system, respectively, Ψ, γ denotes a respective yaw direction angle and angular velocity, m, I _mm , C _{f, r} , F _{xf, and x r} are vehicle parameters, each of which is a safety limiting ellipse representing mass, mass moment of inertia, front / rear wheel torsional stiffness coefficient, and front / rear longitudinal tire force.
[3] The method according to claim 1 or 2, wherein the safety restriction ellipse is an elliptical shape in which a distance between the subject vehicle and the opponent vehicle is formed along a major axis,

, And a and b are

Respectively,
here

,

Autonomous vehicle system using safety limit ellipses.
4. The system of claim 3, wherein the control module uses a safety restriction ellipse to change the vehicle to a local route running mode when the vehicle enters the safety restriction tower.
5. The method according to claim 4, wherein, when the relative speed between the subject vehicle and the opponent vehicle is equal to or greater than a reference value in the regional route running mode, the control module controls the longitudinal acceleration Autonomous travel system of a vehicle using safety limit ellipse to travel.
6. The system of claim 5, wherein the relative distances and relative speeds of the vehicle and the target vehicle are estimated using an IMM technique.
The vehicle control system according to claim 6, wherein the control of the vehicle is performed by any one of (i) constant speed running, (ii) longitudinal direction acceleration / deceleration driving, or (iii) lateral direction acceleration / , And the transition probability between these operation patterns is defined by the Markov chain model.
The vehicle according to claim 7, wherein, in the lateral travel control of the child vehicle, the limit distance of the lateral movement is set to a lateral movement distance generated when changing the lane width of the road and one lane, The limiting size of the angular velocity at the time of driving is the autonomous driving system of the vehicle using the safety limit ellipse which is set based on the maximum lateral acceleration and the speed of the vehicle.
A method of automatically running an automobile,
A step of setting a global route for guiding the subject vehicle to a destination,
In order to predict the nonlinear model, the vehicle state is defined based on the global coordinate system for guiding the subject vehicle to its destination and the fixed coordinate system of the subject vehicle, and a 6-degree-of-freedom nonlinear vehicle module is calculated Step,
The vehicle traveling along the global route;
When a dynamic obstacle is detected, a virtual safety restriction ellipse calculated by a relationship between the host vehicle and the target vehicle is given to the target vehicle;
And starting a local route driving mode in which the subject vehicle travels in accordance with travel of the destination vehicle when the subject vehicle enters the safe limit ellipse.
10. The method of claim 9, wherein in the local route driving mode,
When the relative speed between the subject vehicle and the target vehicle is equal to or greater than the reference value and there is no change in the lane of the target vehicle, the lateral acceleration running of the subject vehicle is performed,
Wherein the longitudinal acceleration of the subject vehicle occurs when the relative speed between the subject vehicle and the target vehicle is equal to or greater than a reference value and there is a lane change of the subject vehicle.
11. The method of claim 10, wherein a relative distance and a relative speed between the subject vehicle and the target vehicle are estimated through an IMM technique.
The vehicle control system according to claim 11, wherein the control of the vehicle is performed by any one of (i) constant-speed running, (ii) longitudinal acceleration / deceleration, or (iii) lateral acceleration / deceleration , And the transition probability between these operating patterns is defined by the Markov chain model.
13. The control method according to claim 12, wherein, in the lateral travel control of the child vehicle, the limit distance of the lateral movement is set to a lateral travel distance which occurs when the lane width of the road and one lane are changed, The limiting size of the angular velocity at the time of driving is the automatic traveling method using the safety limit ellipse which is set based on the maximum lateral acceleration and the speed of the vehicle.
14. The vehicle occupant protection system according to any one of claims 9 to 13, wherein the safety restriction ellipse is an elliptical shape in which the distance between the subject vehicle and the opponent vehicle is formed along a major axis,

, And a and b are

Respectively,
here

A method of automatic driving of a vehicle using a safety limit ellipse.

Images