KR101807397B1 - Method and device for avoiding obstacles based on risk of obstacles - Google Patents

Abstract

A method and an apparatus for avoiding obstacles that determine a passing point between obstacles in consideration of the risk of collision between a moving object and an obstacle are disclosed. The disclosed obstacle avoidance method includes: generating a collision risk value for each of a first obstacle and a second obstacle; Comparing collision risk values of the first and second obstacles to determine a passing point between the first and second obstacles such that the collision risk value is close to a relatively small obstacle; And generating a movement path of the moving object including the passing point.

Description

TECHNICAL FIELD The present invention relates to a method and an apparatus for avoiding obstacles based on the risk of an obstacle,

The present invention relates to a method and apparatus for avoiding an obstacle based on the risk of an obstacle, and more particularly, to a method and apparatus for avoiding obstacles that determine a passing point between obstacles in consideration of the risk of collision between the obstacle and a moving object.

Recently, autonomous vehicles, unmanned robots, drones, and unmanned vehicles have been actively studied. In order to commercialize such unmanned mobile objects, consideration of safety is essential. The main threat to safe driving is a collision with an obstacle while driving, so research on how to detect and avoid obstacles is also being conducted in parallel.

There is a method of using a SVM (Support Vector Machine) as a method of avoiding existing obstacles and generating a path of a moving object. SVM is an algorithm widely used in the field of machine learning. As shown in FIG. 1, the SVM determines a support vector at the outermost data among a plurality of data, Classify the data.

The obstacle avoiding method using the SVM avoids obstacles by considering the above-described data as obstacles and determining a line passing through the support vectors as a moving path of the moving object.

However, since the above-described method does not consider the movement of the obstacle, it does not reflect the traveling environment of the actual moving object, and there is a high possibility of collision between the obstacle and the moving object.

A related prior art document is Korean Patent Publication No. 2013-0118116.

SUMMARY OF THE INVENTION The present invention provides an obstacle avoidance method and apparatus for determining a passing point between obstacles in consideration of the risk of collision between a moving object and an obstacle and generating an obstacle avoidance path.

According to another aspect of the present invention, there is provided a method for avoiding an obstacle in a mobile object based on a risk of an obstacle, the method comprising: generating a collision risk value for each of the first obstacle and the second obstacle; Comparing collision risk values of the first and second obstacles to determine a passing point between the first and second obstacles such that the collision risk value is close to a relatively small obstacle; And generating a movement path of the moving object including the passing point.

According to another aspect of the present invention, there is provided a method for avoiding an obstacle in a mobile object based on a risk of an obstacle, the method comprising: generating a collision risk value for each of the first obstacle and the second obstacle; Determining a passing point between the first obstacle and the second obstacle according to the collision risk value; And generating the curved path using a predetermined point between the current position of the moving object and the passing point, the current position of the moving object, and the passing point.

According to another aspect of the present invention, there is provided an apparatus for avoiding obstacles based on the risk of an obstacle, the apparatus including a sensor for generating a collision risk value for each of the first obstacle and the second obstacle, A risk determination section; A passing point determining unit that determines a passing point between the first obstacle and the second obstacle according to the collision risk value; And a movement path generation unit for generating a movement path including the passing point.

According to the present invention, it is possible to reduce the possibility of a vehicle colliding with an obstacle by generating a path such that the risk moves relatively far away from the obstacle and relatively far from the obstacle.

Also, according to the present invention, the obstacle avoidance path can be created in consideration of the dynamic state of the obstacle, thereby reducing the risk of collision between the obstacle and the vehicle even in an unexpected situation.

1 is a diagram for explaining an SVM algorithm.
2 is a diagram for explaining the concept of the obstacle avoidance method based on the risk of an obstacle according to the present invention.
3 is a diagram for explaining an obstacle avoidance apparatus based on the risk of an obstacle according to an embodiment of the present invention.
4 and 5 are views for explaining a risk value determination method according to an embodiment of the present invention.
6 and 7 are views for explaining a method of determining a passing point between obstacles according to an embodiment of the present invention.
FIG. 8 is a diagram for explaining a curved path generation method according to an embodiment of the present invention. Referring to FIG.
FIG. 9 is a view for explaining an obstacle avoidance method based on the risk of an obstacle according to an embodiment of the present invention.
10 is a view for explaining an obstacle avoidance method based on the risk of an obstacle according to a specific embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

2 is a diagram for explaining the concept of the obstacle avoidance method based on the risk of an obstacle according to the present invention.

Hereinafter, the case where the moving object is a vehicle is described as an embodiment, but the present invention can be applied to various moving objects such as an autonomous vehicle, a mobile robot, and a drone.

The present invention creates a path of travel of a vehicle passing between obstacles while avoiding obstacles. At this time, the present invention determines the passage point of the vehicle between the obstacles in consideration of the collision risk against the vehicle and the obstacle, and determines the passing point between the obstacles so that the collision risk is close to the relatively small obstacle. Since the obstacle may be a dynamic state as well as a static state, the present invention determines the collision risk in consideration of the moving state of the obstacle.

2, when the risk of collision between the vehicle 240 and the first obstacle 210 and the risk of collision between the vehicle 240 and the second obstacle 220 are the same, the present invention can be applied to the first obstacle 210 210 and 220 may be determined as the passing point of the vehicle 240. The distances between the first and second obstacles 210 and 220 at the center point 230 are the same.

However, when the risk of collision between the vehicle 240 and the first obstacle 210 is greater than the risk of collision with the second obstacle 220, the present invention can be applied to the case where the passing point of the vehicle 240 is the second obstacle 220, (230). Conversely, when the risk of collision between the vehicle 240 and the first obstacle 210 is less than the risk of collision with the second obstacle 220, the present invention can prevent the vehicle 240 from passing through the first obstacle 210 and the center point 230). ≪ / RTI >

As a result, according to the present invention, it is possible to reduce the possibility of the vehicle colliding with an obstacle by generating a path such that the risk moves relatively far from the obstacle and relatively far from the obstacle.

3 is a diagram for explaining an obstacle avoidance apparatus based on the risk of an obstacle according to an embodiment of the present invention.

Referring to FIG. 3, the obstacle avoidance apparatus according to the present invention includes a risk determination unit 310, a passing point determination unit 320, and a movement path generation unit 330.

The risk determination unit 310 uses the sensor to generate a collision risk value for each of the first obstacle and the second obstacle. The risk determining unit 310 detects the velocity, acceleration, and the like of the first and second obstacles using the sensing value, and generates a collision risk value with the vehicle using the detected information.

The sensor is a sensor for detecting an obstacle, and may be an LIDAR or an ultrasonic sensor as an example. The sensor may be included in an obstacle avoidance device or vehicle.

The passing point determining unit 320 determines a passing point between the first obstacle and the second obstacle according to the collision risk value. At this time, the passing point determining unit 320 determines a passing point between the first and second obstacles so that the collision risk value is close to the relatively small obstacle.

The movement path generation unit 330 generates a movement path of the vehicle including the passing point. At this time, the movement path generation unit 330 generates a curved movement path for the natural running of the vehicle.

Meanwhile, the obstacle avoidance apparatus according to the present invention estimates the position of the obstacle after the collision prediction time by using the acceleration of the obstacle and the expected collision time between the vehicle and the obstacle, and if the distance between the predicted position and the vehicle is less than the threshold value , A collision risk value is generated, and a series of operations for obstacle avoidance can be performed.

According to the present invention, by creating an obstacle avoidance route in consideration of the dynamic state of the obstacle, the risk of collision between the obstacle and the vehicle can be reduced even in an unexpected situation.

4 and 5 are views for explaining a risk value determination method according to an embodiment of the present invention.

The risk determining unit 310 according to the present invention can determine the collision risk value using the acceleration of the first and second obstacles and the expected collision time with the vehicle. At this time, the expected collision time can be determined according to the distance between the vehicle and the obstacle and the relative speed.

As described above, the acceleration and speed of the obstacle can be determined through the sensing value of the sensor, and the distance and relative speed between the vehicle and the obstacle can be determined based on the current speed and position of the vehicle. The expected collision time is proportional to the distance between the vehicle and the obstacle and inversely proportional to the relative speed.

As an example, the collision risk value can be defined by 7 values as shown in FIG. 4, and the larger the value, the higher the collision risk. The collision risk value is proportional to the magnitude of the acceleration (ACC) and may be inversely proportional to the estimated collision time (TTC). The collision risk value can be calculated based on the center point between the obstacles. In this case, a minus (-) in FIG. 3 indicates that the obstacle moves in a direction away from the center point. If the acceleration or the collision expected time is minus (-), the collision risk value becomes smaller because the distance between the vehicle and the obstacle becomes longer.

For example, if the first and second obstacles 510 and 520 move to the state shown in FIG. 5 and the vehicle 540 moves between the first and second obstacles 510 and 520, Since the vehicle 540 and the second obstacle 520 are moving in the direction of the center point 530 while the acceleration and the velocity of the obstacle 520 are smaller than the acceleration and the velocity of the first obstacle 510, Generates a collision risk value for the obstacle so that the collision risk value of the second obstacle 520 is greater than the collision risk value of the first obstacle 510. [

In Fig. 5, solid arrows indicate speed, and dashed arrows indicate acceleration. The length of the arrow indicates the magnitude of the velocity or acceleration. The acceleration of the first obstacle 510 and the estimated collision time are negative values.

6 and 7 are views for explaining a method of determining a passing point between obstacles according to an embodiment of the present invention.

The passing point determination unit 320 according to the present invention can determine the passing point of the vehicle using the difference between the collision risk value and the collision risk value between the vehicle and the obstacle. In one embodiment, the passing point may be closer to the relatively small obstacle value of the collision risk value, in proportion to the difference between the collision risk value and the collision risk value.

For example, when the collision risk value of the vehicle and the first obstacle 610 is smaller than the collision risk value of the second obstacle 620, the passing point is greater than the collision risk value of the second obstacle 620 1 < / RTI > At this time, as the collision risk value of the first and second obstacles 610 and 620 is large and the difference of the collision risk values is large, the passing point can be determined to be closer to the first obstacle 610.

6, point B is a passing point when the collision risk value of the first and second obstacles 610 and 620 is greater than that of point A, and point C is the passing point of the first and second obstacles 610 and 620, 620 may have a large difference in collision risk value.

At this time, the passing point determining unit 320 can generate a virtual straight line 640 between the first and second obstacles to determine a passing point located on a virtual straight line. Here, the imaginary straight line is a straight line connecting the first and second obstacles, and the red dot in FIG. 7 represents the start point and the end point of the shortest straight line connecting between the obstacles. The red dot in FIG. 7 may correspond to the support vector of the SVM algorithm described above.

According to the embodiment, when the collision risk value of the first and second obstacles 610 and 620 is equal to or less than the threshold value, the passing point determining unit 320 determines the first and second obstacles 610 and 620 for fast calculation, The central point 630 between the first and second obstacles 610 and 620 can be determined as a passing point without comparing the collision risk values of the first and second obstacles 610 and 620.

On the other hand, the passing point determining unit 320 can use the cost function to determine whether the vehicle 750 passes through the passage point X between obstacles when two or more obstacles exist, as shown in FIG. 7 . That is, the passing point determining unit 320 may use a cost function to select one path among a plurality of red paths. 7, the obstacle includes guardrail obstacles 730 and 740 of the road in addition to the vehicle obstacles 710 and 720, and the cost function C _i may be equal to Equation 1 as an example.

는 각 경로에서 장애물 사이의 통과 지점에 대한 충돌 위험도 값을 나타내며,

는 도로의 식별 표지와 차량의 이동 경로 간 거리 차이 값을 나타낸다. 도로의 식별 표지는 예를 들어, 차선, 중앙선 등일 수 있다.Here,? Is a constant determined from 0 to 1,

Represents the collision risk value for the passage point between obstacles in each path,

Represents the distance difference value between the identification mark of the road and the movement path of the vehicle. The identification mark of the road may be, for example, a lane, a center line, and the like.

That is, the passing point determining unit 320 can determine that the vehicle passes through the obstacles indicating the minimum cost by using the cost function, and passes through the identification mark of the road, The point can be determined.

If it is decided to pass between obstacles, the vehicle can move via the passing point of the determined path.

FIG. 8 is a diagram for explaining a curved path generation method according to an embodiment of the present invention. Referring to FIG.

The movement path generation unit 330 according to the present invention can generate a curved movement path using a predetermined point between the current position of the vehicle and the passing point, the current position of the vehicle, and the passing point.

The movement path generation unit 330 may generate a curve by using a Line and arc algorithm, a Spline curve algorithm, a Clothoid curve algorithm, a Bezier curve algorithm, or the like. In the case of using the Bezier curve algorithm, ] To generate the movement route.

Here, P ₀ represents the current position 810 of the vehicle, and P ₁ represents a predetermined point 820 between the current position of the vehicle and the passing point at the passing point determining section 320. And P ₂ represents the passing point 830 determined by the passing point determining unit 320. P ₁ can be variously determined so that the movement path has a curved shape.

A curved line in the form of a parabola can be generated according to a quadratic function (Equation 2), where t is a variable when t is 0 and the end point of a curve when t is 1.

8 (b), when the pass point 830 is determined between the first and second dynamic obstacles 840 and 850, the predetermined point P ₁ is between the first dynamic obstacle 840 and the static obstacle 860 As shown in FIG. Alternatively, the predetermined point P ₁ may be a passing point between the first dynamic obstacle 840 and the static obstacle 860 determined according to the collision risk value in the passing point determining unit 320.

FIG. 9 is a view for explaining an obstacle avoidance method based on the risk of an obstacle according to an embodiment of the present invention.

The obstacle avoidance method according to the present invention can be performed in the above-described obstacle avoiding device or moving object, and in Fig. 9, a vehicle obstacle avoiding method will be described as an embodiment.

 The vehicle according to the present invention generates a collision risk value for each of the first obstacle and the second obstacle (S910). At this time, the vehicle can determine the collision risk value using the acceleration of the first and second obstacles and the expected collision time with the vehicle.

Then, the vehicle determines the passing point between the first obstacle and the second obstacle (S920) according to the collision risk value, and compares the collision risk values of the first and second obstacles so that the collision risk value is relatively small So as to determine a passing point between the first and second obstacles. The passing point may be a point at which the collision risk value approaches a relatively small obstacle in proportion to the difference between the collision risk value and the collision risk value.

The vehicle can create a virtual straight line between the first and second obstacles to determine a passing point located on an imaginary straight line where the imaginary straight line is the shortest distance connecting the first and second obstacles It is a straight line.

Also, the vehicle may determine a center point between the first and second obstacles as a passing point when the collision risk value of the first and second obstacles is equal to or less than the threshold value.

When the passing point is determined, the vehicle generates a traveling path of the vehicle including the passing point (S930). At this time, the vehicle can generate a curved travel path using a predetermined point between the current position of the moving object and the passing point, the current position of the moving object, and the passing point.

10 is a view for explaining an obstacle avoidance method based on the risk of an obstacle according to a specific embodiment of the present invention.

The vehicle according to the present invention estimates the positions of the first and second obstacles (S1020) after the estimated collision time using the acceleration and the estimated collision time of the obstacle by collecting data on the obstacle using the LIDAR sensor (S1010) do. If the distance between the predicted position and the vehicle is less than or equal to the threshold value, the process proceeds to step S1030.

When the LIDAR sensor is used, the obstacle is detected in a plurality of dot shapes along the portion reflecting the laser. The vehicle according to the present invention determines the support vector among the plurality of points using the SVM algorithm, It is determined whether or not it can pass (S1030). As an example, if the shortest distance between obstacles is greater than or equal to a threshold value, the vehicle can be determined to be able to pass between obstacles.

Then, the vehicle calculates the acceleration and the collision prediction time of the obstacle based on the center point between the obstacles and generates the collision risk value for the obstacle (S1040). The vehicle is weighted according to the risk value to determine the passing point of the vehicle (S1050). Depending on the weight, the passing point of the vehicle can be close to the first obstacle or the second obstacle.

If there are two or more obstacles less than the critical distance, the vehicle uses the cost function to determine which obstacle to move, that is, which route to select (S1060), smoothens the selected route in a curved shape, And finally generates a moving route of the vehicle (S1070).

The above-described technical features may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, the present invention has been described with reference to particular embodiments, such as specific elements, and specific embodiments and drawings. However, it should be understood that the present invention is not limited to the above- And various modifications and changes may be made thereto by those skilled in the art to which the present invention pertains. Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, belong to the scope of the present invention .

Claims (12)

A method for avoiding an obstacle of a mobile object based on a risk of an obstacle,
Generating a collision risk value for each of the first obstacle and the second obstacle;
Comparing collision risk values of the first and second obstacles to determine a passing point between the first and second obstacles such that the collision risk value is close to a relatively small obstacle; And
And creating a movement path of the moving object including the passing point,
The passage point
In proportion to the difference between the collision risk value and the collision risk value, the collision risk value approaches a relatively small obstacle,
The step of determining the pass point
Determining a center point between the first and second obstacles as the passing point when the collision risk value of the first and second obstacles is less than or equal to a threshold value
Obstacle avoidance method.
The method according to claim 1,
The step of generating the collision risk value
Determining the collision risk value using the acceleration of the first and second obstacles and the expected collision time with the moving object,
The estimated collision time is
The distance between the moving object and the obstacle, and the relative speed
Obstacle avoidance method.
3. The method of claim 2,
Further comprising: predicting a position of the first and second obstacles after the collision expected time using the acceleration and the collision expected time,
The step of generating the collision risk value
And when the distance between the predicted position and the moving object is equal to or less than a threshold value,
Obstacle avoidance method.
delete delete The method according to claim 1,
The step of determining the pass point
Creating a virtual straight line between the first and second obstacles to determine the passing point located on the imaginary straight line,
The imaginary straight line
A straight line of the shortest distance connecting the first and second obstacles
Obstacle avoidance method.
The method according to claim 1,
The step of generating the movement path
Using the predetermined position between the current position of the moving body and the passing point, the current position of the moving body and the passing point,
Obstacle avoidance method.
delete delete An obstacle avoidance apparatus based on the risk of an obstacle,
A risk determination unit for generating a collision risk value for each of the first obstacle and the second obstacle using the sensor;
A passing point determining unit for comparing a collision risk value of the first and second obstacles to determine a passing point between the first obstacle and the second obstacle so that the collision risk value approaches a relatively small obstacle; And
And a movement path generation unit for generating a movement path including the passage point,
The passage point
In proportion to the difference between the collision risk value and the collision risk value, the collision risk value approaches a relatively small obstacle,
Wherein the passing point determining unit determines the center point between the first and second obstacles as the passing point when the collision risk value of the first and second obstacles is equal to or less than the threshold value
Obstacle avoidance device.
delete 11. The method of claim 10,
The movement path generation unit
The moving path in a curved shape is generated by using the predetermined position between the current position of the moving object using the obstacle avoiding device and the passing point, the current position of the moving object, and the passing point
Obstacle avoidance device.

Images