KR101581286B1 - System and method for path planning for autonomous navigation of driverless ground vehicle - Google Patents

Abstract

The present invention relates to a route tracking system and method for autonomous driving of an unmanned driving vehicle, and more particularly, to a route tracking method for autonomous driving of an unmanned driving vehicle, the method comprising the steps of: (a) (B) generating a reference path parallel to the true path using the current position of the vehicle, (c) using the current position of the vehicle as a starting point, Generating at least one candidate path by generating a number of virtual path points equal in number to the reference points of the path, (d) examining whether or not the generated candidate paths are obstructive interference, (E) selecting an optimal path based on an index and an elapsed time of the target point among the modified candidate paths.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for autonomous driving of an unmanned driving vehicle,

The present invention relates to a route tracking system and method for autonomous driving of an unmanned driving vehicle, and more particularly, to a system and method for autonomous navigation of an unmanned driving vehicle, in which a reference route parallel to a true route is generated using a current position measured by a position measuring apparatus, At least one candidate path is created by the same number of virtual path points as the number of the candidate paths, the candidate paths are modified based on the environment information provided from the surrounding environment sensing device, and an optimal path is selected from among the modified candidate paths And more particularly, to a system and method for tracking a route for autonomous driving of an unmanned driving vehicle.

2. Description of the Related Art [0002] As is well known, an automatic traveling apparatus built in various traveling means and performing an automatic traveling through a traveling position search is mainly applied to a ship, an airplane, , Road congestion, and the like to the user through a monitor, or the automatic traveling device drives the vehicle by itself or controls the running state.

However, the automatic traveling device has not yet been commercialized. For example, in the case of a moving object moving at high speed such as a vehicle, the automatic traveling device recognizes the traveling environment (for example, front vehicle recognition, obstacle detection, Or each of the emergency functions corresponding thereto by judging the driving environment in real time, and a high-performance processor capable of processing a large amount of data in real time is required for this purpose.

In addition, in the moving state, since it is necessary to recognize the driving environment and to detect the surrounding vehicle, it is difficult to accurately divide the object and the background of the vehicle, so that errors occur frequently in tracking the object. Often occurs.

In recent years, studies have been actively conducted on unmanned automatic traveling devices. Such unmanned automatic traveling devices detect a travelable section using, for example, a radar, a lidar, and an image sensor , And the present position of the unmanned automatic traveling device is determined by using a global positioning system (GPS), and then the vehicle is automatically traveled to the destination by generating a traveling route of the vehicle in real time, Can be performed.

In order to carry out such unmanned automatic traveling, a function of grasping a travelable section by using various sensors (for example, radar, lidar, image sensor, etc.) is indispensably required. In addition, A function of generating a detailed travel route in any one of the sections is required and a function of detecting and avoiding obstacles in the travel section is required.

There are A *, D *, and non-Zygote curve generation algorithms for local path generation for detailed travel path generation. The A * algorithm is used for automatic parking of a vehicle because it generates a path that can be matched to a given global route with the shortest calculation speed. However, since the A * and D * algorithms tend to follow a given global path, there is a problem that S-shaped paths with a large fluctuation in the lateral direction are frequently generated when the position error of the GPS reception data is large. Therefore, it is difficult to apply the A * algorithm to the current situation.

In order to accurately follow a given route point, the vehicle's self-position recognition must be accurate. The data provided by a low-cost GPS receiver currently available on the market always includes a position error of about 5 m, There is a drawback that can not be.

Korean Patent No. 10-1214474, entitled " Navigation device and method of providing travel route information using the same, automatic traveling system and method thereof "

SUMMARY OF THE INVENTION It is an object of the present invention to create a route capable of performing autonomous travel on a road where various types of obstacles are present with self position data of a vehicle accompanied by a large error, And to provide a route tracking system and method for autonomous driving of an unmanned driving vehicle.

Another object of the present invention is to provide a method and apparatus for generating a plurality of paths parallel to a path connecting given path points within an area including a given road width and a nominal position error of a GPS receiver, The present invention also provides a system and method for autonomous navigation of an unmanned driving vehicle that minimizes the S-shaped path generation and allows smooth and rapid travel even when the GPS position error is large.

It is still another object of the present invention to provide an unmanned driving vehicle capable of generating a plurality of candidate routes within a given road width at regular time intervals to select a route that does not interfere with the obstacle even when an unexpected obstacle occurs, And to provide a path tracking system and method for autonomous driving.

It is another object of the present invention to provide a system and method for tracking an autonomous vehicle traveling along an unmanned driving vehicle that changes the path so that the vehicle can follow the obstacle.

According to an aspect of the present invention, there is provided a route tracking method for autonomous driving of an unmanned driving vehicle, the method comprising the steps of: (a) supplementing supplementary route points to a route point of a previously provided global route, (B) generating a reference path parallel to the true path using the current position of the vehicle, (c) using the current position of the vehicle as a starting point, and Generating at least one candidate path by generating virtual path points in a predetermined number of paths, (d) checking whether the generated candidate paths are obstructive interference, and correcting candidate paths when there is no obstacle interference, e) selecting an optimal path based on the index of the target point and the elapsed time among the modified candidate paths. The path tracking method is provided.

Wherein the step (a) comprises the steps of: calculating an increment of supplementary path points to be supplemented between path points existing on the global path, generating a supplementary path point by using the calculated increment value, And creating a true path by connecting the point and the previously provided path point.

Wherein the step (b) includes the steps of finding a first path point closest to a current position of the vehicle, obtaining a distance d between a line connecting the first path point and a next path point and a current position, A second path point that is separated by the detection distance (R) of the environmental sensing device is found, and a line connecting itself and a path point subsequent to itself is found for each of the smoothed path points from the first path point to the second path point Generating new reference points using the created line and the distance d, and connecting the new reference points to generate a reference path.

Wherein the step (c) includes the steps of: calculating an increment for determining virtual path points to be located between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path; Determining a virtual path point at a position spaced apart by the increment value, and generating at least one candidate path by connecting the determined virtual path points.

Wherein the step (d) comprises the steps of: checking whether the minimum square surrounding the vehicle and the obstacle on the hazard map are interfered; if there is no interference with the obstacle, Searching a virtual path point having the largest index among virtual path points within a certain distance as a target point, calculating a traveling direction to a target point when a target point exists, calculating a traveling direction of the vehicle using the calculated traveling direction Updating the vector to obtain a new passing point, and calculating and storing the speed of the vehicle at the new passing point.

 Wherein calculating the driving direction to the target point includes calculating an angle between a vector from a first passing point to a target point and a vector of a current heading, calculating a steering angle using the obtained angle, And calculating a running direction to the target point.

According to another aspect of the present invention, there is provided an apparatus for tracking a vehicle, comprising: (a) supplementing auxiliary path points to a path point of a previously provided global path to generate a true path; (b) Generating a reference path parallel to the true path, (c) generating a virtual path point in a number equal to the reference points of the reference path, starting from the current position of the vehicle, and generating at least one candidate path (D) checking whether the generated candidate paths are obstructive interference, and correcting candidate paths when there is no obstacle interference, (e) determining an index of the target point and an elapsed time from among the corrected candidate paths, And a step of selecting an optimum path on the basis of the path information of the unmanned driving vehicle A recording medium is provided.

According to another aspect of the present invention, a true path correcting unit for generating a true path by supplementing auxiliary path points at a path point of a previously provided global path, and a reference path parallel to the true path using the current position of the vehicle A candidate path generating unit for generating at least one candidate path by generating virtual path points having the same number as the reference points of the reference path with the current position of the vehicle as a starting point; A candidate path correcting unit for displaying an obstacle in a danger map when an obstacle exists in the provided surrounding environment information, inspecting the candidate paths for obstacle interference, and correcting candidate paths in the absence of obstacle interference, And an optimal path selection unit for selecting an optimal path based on the index of the target point and the elapsed time among the candidate paths It is provided by the path tracking device.

Wherein the true path correcting unit calculates an increment value of supplementary path points to be supplemented between path points existing on the global path and generates a supplementary path point using the calculated increment value, You can modify the true path by concatenating the provided path points.

The reference path generation unit finds a first path point closest to the current position of the vehicle, calculates a distance d between a line connecting the first path point and a next path point and a current position, A second path point that is separated by the detection distance R of the apparatus is searched, and a line connecting itself and a path point subsequent to itself is created for each of the smoothed path points from the first path point to the second path point , It is possible to generate new reference points by using the created line and the distance d, and create reference paths by connecting new reference points.

Wherein the candidate path generation unit calculates an increment value for determining virtual path points to be positioned between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path, A virtual path point may be determined at an incremental distance, and at least one or more candidate paths may be created by connecting the determined virtual path points.

The candidate path modifying unit checks whether or not the minimum square surrounding the vehicle and the obstacle on the hazard map interfere with each other. If there is no interference with the obstacle, the candidate route modifying unit determines that the distance between the minimum square and the turn circle is within a certain distance Searching a virtual path point having the largest index among the virtual path points as a target point, calculating a traveling direction to the target point, updating a state vector of the vehicle using the calculated traveling direction to obtain a new passing point, The speed of the vehicle at the passing point can be calculated and stored.

The candidate path correcting unit calculates an angle between a vector from a first passing point to a target point and a vector of a current heading, calculates a steering angle using the obtained angle, and calculates a running direction to a target point using the calculated steering angle .

According to another aspect of the present invention, there is provided an apparatus for measuring a position of a vehicle, the apparatus comprising: a position measuring device for measuring a current position of the vehicle; sensing information about the surrounding environment when the vehicle travels through the provided global path; A reference path that is parallel to the true path using the current position measured by the position measuring device, and generates at least one or more reference paths by the same number of virtual path points as the reference points of the reference path, A path-following device for generating a candidate path, modifying the generated candidate paths based on the environment information provided from the environment-sensing device, and selecting an optimal path from among the modified candidate paths, And a steering device for driving the vehicle along the optimal path. A follow-up system is provided.

The ambient environment sensing device may include at least one of a laser scanner, a camera, and an image recognition sensor that senses an obstacle and an environmental characteristic of the surroundings of the vehicle when the vehicle starts running, and provides the sensor to the path following device.

According to the present invention, it is possible to generate a path capable of performing autonomous traveling on a road in which various types of obstacles exist, using the self-position data of the vehicle accompanied by a large error, Can be minimized.

Also, by generating a plurality of routes parallel to a route connecting given route points within an area obtained by adding a given road width and a nominal position error of the GPS receiver, and selecting a route that can reach the destination in the shortest time, It is possible to minimize the generation of the S-shaped path and smooth and fast traveling.

In addition, it is possible to generate a plurality of candidate routes within a given road width at a predetermined period of time to select a route that does not interfere with the obstacle even if an unexpected obstacle occurs, thereby driving the vehicle.

In addition, the path can be modified so that the vehicle follows the obstacle presence check.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a view illustrating a route tracking system for autonomous driving of an unmanned driving vehicle according to an embodiment of the present invention; FIG.
2 is an exemplary diagram for explaining an obstacle scenario and a route division according to an embodiment of the present invention;
3 is a block diagram schematically showing a configuration of a path estimating apparatus according to an embodiment of the present invention.
FIG. 4 is an exemplary diagram for explaining auxiliary path points and path point smoothing according to an embodiment of the present invention; FIG.
5 is an exemplary diagram for explaining a method of generating a reference path according to an embodiment of the present invention;
FIG. 6 is an exemplary diagram illustrating a method for generating a candidate path according to an embodiment of the present invention; FIG.
FIG. 7 is an exemplary diagram for explaining a method of selecting a target point according to an embodiment of the present invention; FIG.
8 is an exemplary diagram illustrating a method for determining a new pass-through point in accordance with an embodiment of the present invention.
9 is a flowchart illustrating a route tracking method for autonomous driving of an unmanned driving vehicle according to an embodiment of the present invention.
10 illustrates an example of an obstacle displayed on a hazard map according to an embodiment of the present invention.

The foregoing and other objects, features, and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG.

Each functional unit represented in the present specification is only an example of implementation of the present invention. Therefore, other functional units may be used in other embodiments of the present invention without departing from the spirit and scope of the present invention. In addition, each functional unit may be implemented solely in hardware or software configuration, but may be implemented in combination of various hardware and software configurations that perform the same function.

In the embodiment of the present invention, the vehicle may be collectively referred to as a robot or an unmanned vehicle.

FIG. 1 is a diagram illustrating a route tracking system for autonomous driving of an unmanned driving vehicle according to an embodiment of the present invention. FIG. 2 is an exemplary view for explaining an obstacle scenario and a route division according to an embodiment of the present invention.

Referring to FIG. 1, a route tracking system for autonomous driving of an unmanned driving vehicle includes a position measuring device 100, an environment sensing device 200, a path following device 300, and a steering device 400.

The position measuring apparatus 100 measures the current position of the vehicle and transmits the measured current position to the path following apparatus 300. [ The position measuring apparatus 100 may be, for example, a GPS receiver. The GPS receiver calculates the position value of the vehicle using satellite signals received from the plurality of satellites through the antenna, and transmits the position value to the path following device.

The position measuring apparatus 100 recognizes the heading and the self position of the vehicle and transmits the recognition result to the path following device 300. [

The surrounding environment sensing device 200 senses information about the surrounding environment when traveling through the provided global path and transmits the sensed surrounding environment information to the path following device 300. [ That is, the surrounding environment sensing device 200 may sense obstacles and environmental characteristics of the surroundings of the vehicle when the vehicle starts traveling, and then provide the sensed obstacle and environmental characteristics to the path following device 300. Examples of such an environment sensing device include a laser scanner, a stereo camera, and an image recognition sensor. Here, the obstacle has various shapes such as a lane changing vehicle, a stopping vehicle, a gate, a narrow road, a haystack, and a central obstacle from other lanes depending on the road environment.

In addition, the surrounding environment sensing device 200 acquires geographical information such as an inclination, a roughness, and a friction coefficient of the terrain from a terrain sensing device such as a binocular and a laser distance measuring device mounted on a vehicle and transmits the geographical information to the path following device 300 It is possible.

The path following device 300 generates a reference path parallel to the true path using the current position measured by the position measuring device 100 and generates at least one or more reference paths by the same number of virtual path points as the reference points of the reference path. Generates candidate paths, modifies the generated candidate paths based on the environment information provided from the surrounding environment sensing device 200, and selects an optimal path from among the modified candidate paths.

The position measured by the position measuring apparatus 100 is an error, the vehicle can not be driven to a position coinciding with the true path due to the position error thereof, the obstacle must be avoided if there is an obstacle in front of the vehicle, If you do not get out of the road and there is no clash with obstacles in the corners, you can get out of the route. In order to cope with this situation, the path following device 300 must generate a path that the vehicle can follow. Particularly, in order to avoid obstacles, the route is divided into a change section and a travel section as shown in FIG. The change section refers to a section that adjusts the true path and interval to avoid an obstacle when there is an obstacle ahead, or a section that adjusts the true path and interval to change the travel path to shorten the travel time in the track. The traveling section refers to a section that maintains a constant distance parallel to the true path.

Autonomous driving can be viewed as waypoint navigation because the width of the road to be driven and the position of the center of the road are given in advance as waypoints. Therefore, in the path planning, it is assumed that the global path is given. The presence or absence of an obstacle ahead of the vehicle being driven is discriminated by the surrounding environment sensing device 200, and the obstacle detection result is displayed on the map on which the route is to be set. In this case, it is the creation of a local path to follow the path point in the obstacle detection area and to create a path to avoid the obstacle. The route follow-up device 300 increases the local route update rate to generate a path through which the vehicle can travel without departing from the road even if a position error of the GPS occurs largely or an obstacle appears.

The path following device 300 generates one or more candidate paths for obstacle avoidance based on the environment information about the vehicle provided from the surrounding environment sensing device 200 and selects an optimal path among the generated candidate paths do. That is, the path following device 300 detects the position of the obstacle based on the information provided from the surrounding environment sensing device 200, and generates an optimal path for avoiding the obstacle.

The path following device 300 is a device for dynamically correcting the provided global path using obstacle information around the vehicle and includes an environment sensing device 200 capable of measuring a distance to a surrounding obstacle with laser, The obstacle sensing information is sensed through the map information DB and the map information is received through the map information DB to compare the obstacle sensing information with the obstacle sensing information. Lt; / RTI > Here, the obstacle information is information for distinguishing an area where the vehicle can go from the area on which the vehicle can not go, and is calculated based on the obstacle information obtained in real time using the original map information and the surrounding environment sensing device 200 . Of course, it is possible to extract obstacle information by using any one of the map information or the surrounding environment sensing device 200, but simultaneous use of the map information and the surrounding environment sensing device 200 can yield more accurate information.

The path correction through the path following device 300 calculates both free spaces in a direction perpendicular to the path at a certain distance from the current position of the vehicle in the previously provided global path and then calculates the center of the calculated free space By correcting the path to pass through the point.

A detailed description of such a path following device 300 will be given with reference to FIG.

The steering apparatus 400 drives the vehicle along the optimum path selected in the path following apparatus 300. The steering apparatus 400 may be an electric motor for autonomous steering of the vehicle.

FIG. 3 is a block diagram schematically showing a configuration of a path estimating apparatus according to an embodiment of the present invention. FIG. 4 is an exemplary view for explaining a sub path point and path point smoothing according to an embodiment of the present invention. FIG. 6 is an exemplary diagram for explaining a method of generating a candidate path according to an embodiment of the present invention. FIG. 7 is a view for explaining a method of generating a reference path according to an embodiment of the present invention FIG. 8 is an exemplary diagram for explaining a method for determining a new passing point according to an embodiment of the present invention.

3, the path following apparatus 300 includes a map information storage unit 310, a true path correcting unit 320, a reference path generating unit 330, a candidate path generating unit 340, 350, an optimal path selection unit 360, and a control unit 370.

The map information storage unit 310 stores map data. The map information storage unit 310 is a storage medium including, for example, a hard disk drive (HDD), and stores various map information for providing route search and navigation service, Information and the like are stored and managed, and they are extracted and provided to the controller 370 as needed.

True path corrector 320 corrects the true path by supplementing the auxiliary path points to the path point of the previously provided global path. Here, the waypoint is a radar coordinate with respect to the position at which the vehicle should travel, and the route point information is provided not only with the position as shown in Table 1, but also with the width of the road at that position. The path connecting these path points may be a true path.

sign Information p Order of path points

p ^th location of the path point

road width at the p ^th path point (m) - Others (speed limit, road shape, etc.)

The true path remover 320 calculates the increment of the supplementary path points to be supplemented between the path points existing on the global path, generates the supplementary path point using the calculated increment value, Modify the true path by concatenating the path points.

The route points play an important role in the steering or speed control of the vehicle because the route form can predict the shape of the road the vehicle will travel in the future. However, the route points on the global route are so sparse that it is often necessary to turn the vehicle suddenly when following the route in which the route points are linearly connected. Therefore, in order to realize smooth steering, auxiliary path points are supplemented between the path points provided as shown in Fig. 4 (a). Then, a gentle curved true path is created as shown in FIG. 4 (b).

,

사이에 보충할 보조 경로점들의 간격을 0.5m 이하가 되도록 하여 참경로를 수정하는 방법에 대해 설명하기로 한다.Hereinafter, two path points given on the global path

,

A description will be given of a method of correcting the true path by making the interval of auxiliary path points to be supplemented to be less than 0.5 m.

를 수학식 1을 이용하여 구한다. The true path corrector 320 determines the increment of each of the auxiliary path points so that the interval of the supplementary path points to be supplemented between the two given path points is less than 0.5 m

Is obtained by using Equation (1).

은 두 점 사이의 유클리디안 거리이다. Here, 0.5 m is an experimentally determined value,

Is the Euclidian distance between two points.

를 수학식 2를 이용하여 구한다. When the increment value is calculated according to Equation (1), the true path correcting unit 320 uses the increment value to calculate the qth auxiliary path point

Is obtained by using Equation (2).

이다. From here,

to be.

의 평활화를 수행한다. 이때, p번째 경로점

의 평활화는 이 경로점 전후로 r개의 경로점들을 택하여 수학식 3과 같이 수행한다.When the auxiliary path point is generated using Equation (2), the true path correcting unit 320 adds the pth path point

. At this time, the pth path point

The smoothing is performed as shown in Equation (3) by selecting r path points before and after this path point.

는 p번째 평활된 경로점이고, r은 주어진 경로점들의 거리에 따라 달라질 수 있다.here

Is the pth smoothed path point, and r can vary depending on the distance of the given path points.

The path connecting the smoothed path points according to Equation (3) may be a modified true path.

The reference path generation unit 330 generates a reference path that is parallel to the true path modified by the true path modification unit 320 using the current position.

The reference path generating unit 330 finds the first path point closest to the current position of the vehicle and obtains the distance d between the line connecting the first path point and the next path point and the current position. Then, the reference path generating unit 330 finds a second path point that is separated by the detection distance R of the laser scanner mounted on the vehicle, and calculates a reference point for each of the smoothed path points from the first path point to the second path point Make a line connecting yourself and the path point following yourself. Then, the reference path generation unit 330 generates new reference points using the created line and the distance d, and connects the new reference points to generate a reference path.

와 가장 가까운 경로점

을 찾고,

와 다음 경로점인

을 연결하는 선과 차량의 현재 위치 사이의 거리(d)를 구한다. 그런 후, 기준 경로 생성부(330)는 차량에 장착된 레이저 스캐너의 검지거리 R을 이용하여

로부터 R만큼 떨어진 경로점

을 찾는다. 일례로 R이 50m라면 평활된 경로점 사이의 거리가 0.5m이므로 m은 100이 된다. 그런 후 기준 경로 생성부는

를 포함하여 (m+1)개의 평활화된 경로점들 각각에 대해 자신과 자신 다음에 오는 경로점을 연결하는 선을 만들고 이 선과 상기 구해진 d 를 이용하여 새로운 기준점들

을 생성한다. 이와 같이 만든 기준점들을 연결한 경로가 기준경로(reference path)이다.The reference path generation unit 330 sets a reference path parallel to the true path using the current position of the vehicle as shown in FIG. First, the reference path generation unit 330 generates a reference path

The closest path point to

Looking for,

And the next route point

And the current position of the vehicle. Then, the reference path generating unit 330 uses the detection distance R of the laser scanner mounted on the vehicle

Lt; RTI ID = 0.0 > R &

. For example, if R is 50 m, then m is 100 because the distance between smooth path points is 0.5 m . Then, the reference path generating unit

For each of the (m + 1) smoothed path points, including a line connecting itself and a path point following itself, and using this line and the obtained d ,

. The reference path is the path connecting these reference points.

The candidate path generating unit 340 generates at least one candidate path by generating virtual path points having the same number as the reference points of the reference path, with the current position of the vehicle as a starting point.

The candidate path generation unit 340 calculates an increment value for determining virtual path points to be located between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path, A virtual path point is determined at the incremental value, and the determined virtual path points are connected to generate a candidate path.

When the vehicle meets an obstacle while driving, it must follow the lane partitioned by the paint mark, it may be difficult to follow the given route point if the GPS position error is large. At this time, the vehicle follow-up device 300 must provide a path through which the vehicle can travel. Since there is a GPS position error, it is indispensable to generate a path through which the vehicle can travel. For this, the candidate path generation unit 340 generates a plurality of candidate paths within a range of the road width given to the left and right based on the true path and the positional error provided by the specification of the GPS receiver in use. At this time, the road width is not a large variable in the proposed path planning algorithm, and if the road width is narrow, the number of the candidate paths is small and the number is large.

The interval of the candidate paths to be generated by the candidate path generating unit 340 may be arbitrarily set (for example, 0.5 m), and each candidate path is divided into a change period and a running period as described in FIG.

로 정하고, 후보경로 l의 변경구간을 형성하기 위해 설정된 변경구간의 끝에 위치한 기준경로의 점을

라 한다. 후보 경로 생성부(340)는

과 이에 대응하는 경로 l의 가상 경로점

사이에 위치할 가상 경로점들을 정하기 위한 증분치를 수학식 4를 이용하여 구한다.A method by which the candidate path generating unit 340 generates a candidate path will be described with reference to FIG. The candidate path generation unit 340 generates virtual path points of the same number as the reference points of the reference path to generate candidate paths. At this time, the starting point of all the candidate paths is the current position of the vehicle

And the point of the reference path located at the end of the change period set to form the change period of the candidate path l

. The candidate path generating unit 340

And virtual path points of path l corresponding thereto

The incremental value for determining the virtual path points to be located between the virtual path points is obtained by using Equation (4).

은 기준경로와 후보경로 l사이의 거리를 나타낸다. 도 6에 보인 바와 같이 기준점들의 법선 방향으로

떨어진 곳에 가상 경로점을 정한다. 후보경로 l의 주행구간의 가상 경로점들은 도 6에 보인 바와 같이 기준점들의 법선방향으로 0.5m의 배수만큼 떨어진 곳에 정한다. Where superscript l is the index of the candidate path, n +1 is the number of reference points in the change interval,

Represents the distance between the reference path and the candidate path l . As shown in FIG. 6, in the normal direction of the reference points

Set virtual path points away. The virtual path points of the running path of the candidate path l are set at a distance of a multiple of 0.5 m in the normal direction of the reference points as shown in Fig.

The candidate path remover 350 displays an obstacle in the danger map when there is an obstacle in the surrounding environment information provided according to the running of the vehicle, examines whether or not the obstacle is interfered with the candidate paths, Modify the paths.

That is, the candidate path correcting unit 350 checks whether or not there is interference between the minimum square surrounding the vehicle and the obstacle on the hazard map, and if there is no interference with the obstacle, Finds the virtual path point having the largest index among the virtual path points within a certain distance as the target point. If there is a target point, the candidate path correcting unit 350 calculates the running direction to the target point, updates the state vector of the vehicle using the calculated running direction, calculates the speed of the vehicle at the new passing point, do.

는 모든 후보경로들의 시작점(x _t , y _t ), 즉 첫 통과점이 된다. Since the candidate paths generated by the candidate path generating unit 340 are generated without consideration of the steering angle, heading, presence of an obstacle, etc., the vehicle may not be able to follow. Therefore, the candidate path correcting unit 350 performs an obstacle presence check on each of the candidate paths, and corrects the candidate path so that the vehicle can follow it based on the inspection result. In the revision process, the vehicle passing model and the goal point method used in pure pursuit are used to calculate the passing points that the vehicle can pass through. At this time, the possibility of collision with the obstacle is determined on the risk map in which the position of the obstacle is indicated. Also, the reference path is regarded as one of the candidate paths, and the current position

Is the starting point ( x _t , y _t ) of all candidate paths, i.e., the first pass point.

을 찾는다. 그 이유는 차량이 주행할 수 있는 자세를 취할 수 있도록 하고, 각이 진 코너에서 경로가 꼬이는 것을 방지하기 위해서다. 상기 찾아진

가 목표점일 수 있고, 변수

에 인덱스 k를 저장한다. 만약, 목표점이 존재하지 않으면, 후보 경로 수정부(350)는 현 경로의 수정을 종료한다. The candidate path corrector 350 checks whether there is interference between the bounding box surrounding the vehicle and obstacles on the hazard map. If interference is found as a result of the inspection, the candidate path correcting unit 350 ends the correction of the path. If there is no interference, the candidate path generating unit 350 generates a candidate path having the largest index among the virtual path points within a distance of 0.5 m from the rotation circle outside the minimum turning radius given by the vehicle specification as shown in FIG. 7

. The reason for this is to allow the vehicle to be able to travel and to prevent the path from twisting at the angled corners. The found

May be the target point, and the variable

≪ / RTI > If the target point does not exist, the candidate path correcting unit 350 ends the modification of the current path.

)을 수학식 5를 이용하여 구한다. When the target point is found, the candidate path correcting unit 350 sets the heading (heading) to the target point

) Is obtained by using Equation (5).

일 수 있다. here,

Lt; / RTI >

( x _t , y _t ) to the target point and the vector of the current heading are set as f and g , respectively, and the angle between f and g is obtained using Equation (6).

)을 구한다.The candidate path correcting unit 350 uses the equation 7 to calculate the steering angle < RTI ID = 0.0 >

).

However, since the steering angle of the vehicle has the maximum steering limit, it is determined as shown in Equation (8).

)을 이용한 수학식 5를 이용하여 목표점까지의 주행방향(헤딩)을 계산한다.The candidate path correcting unit 350 calculates the steering angle < RTI ID = 0.0 >

(Heading) to the target point using Equation (5) using the equation (5).

의 영향을 받고, 통과점 수량을 늘리려면 작은 값을 취하고, 수량을 줄이려면 큰 값을 취한다.The candidate path corrector 350 then updates the state vector of the vehicle to obtain a new pass point. Referring to Fig. 8, (x _{t +1} , y _{t +1} ) of the updated state vector becomes the next pass point. Therefore, the number of passing points on the path is

, A small value is taken to increase the number of passing points, and a large value to reduce the number of passing points.

The candidate path correcting unit 350 obtains the speed of the vehicle at a new passing point using Equation (9).

이며,

는 속력 안정계수로 곡률이 k인 도로를 안정적으로 주행할 수 있도록 실험적으로 선택한다. 그리고 새로운 통과점까지의 경과 시간(lap time)은

이다.From here,

Lt;

Is experimentally selected so that the road with the curvature k can be stably traveled with the speed stability coefficient. And the lap time to the new pass point is

to be.

The candidate path corrector 350 stores the pass point (x _{t + 1} , y _{t + 1} ).

에 저장된 목표점의 인덱스와 경과시간

에 의해 하나의 후보경로를 최종 경로로 선정한다. 최종 경로 선정과정에

이

보다 우선한다. 그 이유는

의 값이 크다는 것은 차량이 장애물과의 충돌없이 목적지까지 경로를 잘 추종한다는 의미이기 때문이다.

값이 같은 후보 경로가 여러개 있을 경우, 그 가운데

이 가장 작은 경로가 선정되고, 이 경로상의 통과점들이 차량 제어기에 입력된다.The optimal path selection unit 360 selects an optimal path based on the index of the target point and the elapsed time among the candidate paths corrected by the candidate path correction unit 350. [ That is, the optimal path selection unit 360 selects a candidate path among the candidate paths modified in the candidate path modification unit 350

The index and elapsed time of the target point stored in

A candidate path is selected as a final path. In the final route selection process

this

. The reason is that

The reason for this is that the vehicle follows the route to the destination without collision with the obstacle.

If there are several candidate paths with the same value,

The smallest path is selected, and passing points on this path are input to the vehicle controller.

The control unit 370 includes a map information storage unit 310, a true path correction unit 320, a reference path generation unit 330, a candidate path generation unit 340, a candidate path correction unit 350, 360, and the like.

FIG. 9 is a flowchart illustrating a route tracking method for autonomous driving of an unmanned driving vehicle according to an embodiment of the present invention. FIG. 10 illustrates an obstacle displayed on a danger map according to an embodiment of the present invention.

Referring to FIG. 9, the path-following apparatus replaces the path points of the provided global path to generate a true path (S902). That is, the path following apparatus calculates the increment of the auxiliary path points to be supplemented between the path points existing on the global path, generates the auxiliary path point by using the calculated increment value, To create a true path.

Then, the path following device generates a reference path parallel to the true path using the current position of the vehicle (S904). That is, the path following device finds the first path point closest to the current position of the vehicle, and obtains the distance d between the line connecting the first path point and the next path point and the current position. Then, the path-following apparatus finds a second path point that is separated by the detection distance R of the laser scanner mounted on the vehicle, and then finds each of the smoothed path points from the first path point to the second path point, Make a line connecting the path points to. Then, the path following apparatus generates new reference points using the created line and the distance d, and connects the new reference points to generate a reference path.

The path-following apparatus generates at least one candidate path by generating virtual path points in the number equal to the reference points of the reference path, starting from the current position of the vehicle (S906). That is, the path-following apparatus calculates an increment value for determining virtual path points to be located between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path, A virtual path point is determined at an incremental distance, and the determined virtual path points are connected to generate a candidate path.

The path-following apparatus checks whether or not obstacles are interfered with the candidate paths, and modifies the candidate paths when there is no obstacle interference (S908). At this time, the path following apparatus displays the obstacles detected by the surrounding environment sensing apparatus on the danger map as shown in FIG.

The path-following apparatus selects one optimal path by the index of the target point and the elapsed time among the modified candidate paths (S910).

According to another aspect of the present invention, there is provided a route tracking system comprising: (a) generating a true route by supplementing auxiliary route points to a route point of a previously provided global route; (b) (C) generating virtual path points in a number equal to the reference points of the reference path with the current position of the vehicle as a starting point, thereby generating at least one candidate path (D) checking whether the generated candidate paths are obstructive interference, and correcting candidate paths when there is no obstacle interference, (e) determining an index and an elapsed time of a target point among the corrected candidate paths, And selecting an optimal path based on the time when the vehicle is parked in the unmanned driving vehicle. The recording medium is provided.

The route tracking method for autonomous driving of the unmanned driving vehicle can be programmed, and the codes and code segments constituting the program can be easily deduced by programmers in the field. In addition, a program for a route tracking method for autonomous driving of an unmanned driving vehicle may be stored in an information storage medium (Readable Media) readable by the electronic device, and read and executed by the electronic device.

Thus, those skilled in the art will appreciate that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the embodiments described above are to be considered in all respects only as illustrative and not restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

100: Position measuring device 200: Ambient environment sensing device
300: Path follower 310: Map information storage unit
320: True path correcting unit 330: Reference path generating unit
340: candidate path generating section 350: candidate path determining section
360: Optimum path selection unit 370:
400: Steering device

Claims (15)

A route tracking method for autonomous driving of an unmanned driving vehicle,
(a) calculating an increment of supplementary path points to be supplemented between path points of the provided global path, generating a supplementary path point using the calculated increment value, and comparing the generated supplementary path point and the provided path point Connecting to create a true path;
(b) generating a reference path parallel to the true path using the current position of the vehicle;
(c) generating at least one candidate path by generating virtual path points in a number equal to the reference points of the reference path, with the current position of the vehicle as a starting point;
(d) checking the generated candidate paths for obstacle interference, and correcting candidate paths in the absence of obstacle interference; And
(e) selecting an optimal path based on an index and an elapsed time of the target point among the modified candidate paths;
Wherein the method comprises the steps of:
delete The method according to claim 1,
The step (b)
Finding a first path point closest to a current position of the vehicle and obtaining a distance d between a line connecting the first path point and a next path point and a current position;
A second path point that is separated by a detection distance (R) of the surrounding environment sensing device mounted on the vehicle, and for each of the smoothed path points from the first path point to the second path point, To form a line connecting the electrodes; And
Generating new reference points by using the created line and the distance d, and connecting the new reference points to generate a reference path.
The method according to claim 1,
The step (c)
Calculating an increment value for determining virtual path points to be located between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path;
Determining a virtual path point at a distance of the increment value in the normal direction of the reference path path points;
And generating at least one candidate path by connecting the determined virtual path points to each other.
The method according to claim 1,
The step (d)
Checking whether the minimum square surrounding the vehicle and the obstacle on the hazard map are interfering with each other;
Searching for a virtual path point having the largest index among the virtual path points within a certain distance from the rotation circle outside the minimum turning radius given by the vehicle specification as a target point when there is no interference with the obstacle;
Calculating a running direction to a target point when the target point exists;
Updating a state vector of the vehicle using the calculated travel direction to obtain a new passing point;
Calculating a speed of the vehicle at the new passing point and storing the speed of the vehicle at the new passing point.
6. The method of claim 5,
Wherein the step of calculating the running direction to the target point comprises:
Calculating an angle between a vector from the first pass point to the target point and a vector of the current heading;
Calculating a steering angle using the obtained angle; And
And calculating a traveling direction to a target point by using the calculated steering angle.
When executed by the path following device,
(a) calculating an increment of supplementary path points to be supplemented between path points of the provided global path, generating a supplementary path point using the calculated increment value, and comparing the generated supplementary path point and the provided path point Connecting to create a true path;
(b) generating a reference path parallel to the true path using the current position of the vehicle;
(c) generating at least one candidate path by generating virtual path points in a number equal to the reference points of the reference path, with the current position of the vehicle as a starting point;
(d) checking the generated candidate paths for obstacle interference, and correcting candidate paths in the absence of obstacle interference; And
(e) selecting an optimal path based on an index and an elapsed time of the target point among the modified candidate paths; and (e) reading a computer program containing a program for executing a path following method for autonomous driving of an unmanned vehicle, Possible recording medium.
Calculating an increment value of auxiliary path points to be supplemented between path points existing on the provided global path, generating a secondary path point using the calculated increment value, and comparing the generated secondary path point and the previously provided path point A true path metric that connects and creates a true path;
A reference path generation unit for generating a reference path parallel to the true path using the current position of the vehicle;
A candidate path generation unit for generating at least one candidate path by generating a number of virtual path points equal to the reference points of the reference path, starting from the current position of the vehicle;
A method of displaying an obstacle in a risk map when an obstacle exists in surrounding environment information provided according to a running of the vehicle, checking whether the obstacle is interfered with the candidate paths, government; And
An optimal path selection unit for selecting an optimal path based on the index of the target point and the elapsed time among the modified candidate paths;
.
delete 9. The method of claim 8,
The reference path generation unit finds a first path point closest to the current position of the vehicle, calculates a distance d between a line connecting the first path point and a next path point and a current position, A second path point that is separated by the detection distance R of the apparatus is searched, and a line connecting itself and a path point subsequent to itself is created for each of the smoothed path points from the first path point to the second path point , Generates new reference points using the created line and the distance (d), and connects the new reference points to generate a reference path.
9. The method of claim 8,
Wherein the candidate path generation unit calculates an increment value for determining virtual path points to be positioned between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path, A virtual path point is set at an incremental distance, and at least one candidate path is created by connecting the determined virtual path points.
9. The method of claim 8,
The candidate path modifying unit checks whether or not the minimum square surrounding the vehicle and the obstacle on the hazard map interfere with each other. If there is no interference with the obstacle, the candidate route modifying unit determines that the distance between the minimum square and the turn circle is within a certain distance Searching a virtual path point having the largest index among the virtual path points as a target point, calculating a traveling direction to the target point, updating a state vector of the vehicle using the calculated traveling direction to obtain a new passing point, And calculates and stores the speed of the vehicle at the passing point.
13. The method of claim 12,
The candidate path correcting unit calculates an angle between a vector from a first passing point to a target point and a vector of a current heading, calculates a steering angle using the obtained angle, calculates a running direction to a target point using the calculated steering angle, Wherein the path-following apparatus comprises:
A position measuring device for measuring a current position of the vehicle;
An ambient environment sensing device for sensing information on the surrounding environment when the vehicle travels through the provided global path and transmitting the sensed environmental information to the path following device;
Generating a reference path parallel to the true path by using the current position measured by the position measuring device and generating at least one candidate path by the same number of virtual path points as the reference points of the reference path, A path tracking device for modifying the generated candidate paths based on the environment information provided from the environment sensing device and selecting an optimal path among the modified candidate paths; And
And a steering device for driving the vehicle along the optimal path selected in the path following device,
The path following apparatus calculates an increment for determining virtual path points to be located between a path point of the reference path located at the end of the modification period and a virtual path point of the corresponding candidate path, Wherein a virtual path point is determined at a position separated by the increment, and at least one candidate path is created by connecting the determined virtual path points.
15. The method of claim 14,
Wherein the ambient environment sensing device includes at least one of a laser scanner, a camera, and an image recognition sensor that senses an obstacle and an environmental characteristic of the surroundings of the vehicle when the vehicle starts running, and provides the sensor to the path following device. A Path Tracking System for Autonomous Driving of Vehicles.

Images