KR20190123494A - Method for generating an optimal path for an unmanned ground vehicle and apparatus therefor - Google Patents

Abstract

According to one embodiment of the present invention, in the generation of an optimal path for an unmanned vehicle, the safety of the unmanned vehicle during autonomous driving may be improved by generating a plurality of candidate paths for wide path points arranged at predetermined intervals on a wide path, calculating the least expensive path among the candidate paths based on continuous space mobility information and state information such as an initial/terminal velocity and a direction angle of the unmanned vehicle, and generating the optimal path for wide path.

Description

METHOD FOR GENERATING AN OPTIMAL PATH FOR AN UNMANNED GROUND VEHICLE AND APPARATUS THEREFOR}

The present invention relates to autonomous driving of unmanned vehicles, and more particularly to a method and apparatus for generating an optimal route for an unmanned vehicle.

In general, in operating an unmanned vehicle, the operator establishes a preliminary operation plan such as setting N main target points on the satellite map in order to move the unmanned vehicle to any desired target point.

In addition, when using the path search method based on a mobility map including grid-like driving possibilities generated based on digital terrain information (FDB), a wide path from a starting point to a target point may be generated on a wide grid map. .

As described above, the wide paths generated on the wide grid map generate discontinuous wide paths as shown in FIG. 1 according to the resolution of the unit grid. Thus, the wide paths are used as near-field targets for the regional path planning. This is accompanied by a problem of poor stability in autonomous driving.

To solve this problem, spline technique can be used to generate smooth wide path points based on discontinuous wide path point information. There was a problem that could not be overcome and additional post-processing was required.

Republic of Korea Patent No. 10-1581286 (Registration date December 23, 2015)

Accordingly, in an embodiment of the present invention, after generating a plurality of candidate paths for the wide path points for the wide path points arranged at predetermined intervals on the wide path, the continuous spatial mobility information and the initial / end speed of the unmanned vehicle are generated. Optimal route for unmanned vehicles that can increase stability during autonomous driving of unmanned vehicles by calculating the least expensive route among a plurality of candidate routes based on state information such as direction and direction angle. It is intended to provide a method and apparatus for generating the same.

)을 생성하는 단계와, 기설정된 기준점(

)과 상기 생성된 제 1 궤적(

)에 대해 종말 위치가 free 인 복수개의 후보 경로를 생성하는 단계와, 상기 연속 공간 가동성 정보에 기초하는 상기 복수개의 후보 경로 각각에 대한 비용을 산출하는 단계와, 상기 복수개의 후보 경로 중에서 가장 적은 비용이 산출된 후보 경로를 최적 경로로서 선택하는 단계를 포함한다.A method of generating an optimal route for an unmanned vehicle according to an embodiment of the present invention, the method comprising: generating continuous spatial mobility information based on navigation information driven by an unmanned vehicle; The first trajectory between the first wide path point and the second wide path point (

) And a predetermined reference point (

) And the generated first trajectory (

Generating a plurality of candidate paths having an end position of free for each of the plurality of candidate paths, calculating a cost for each of the plurality of candidate paths based on the continuous spatial mobility information, and having the lowest cost among the plurality of candidate paths. Selecting the calculated candidate path as an optimal path.

The first trajectory may be generated based on an initial speed and an initial direction angle of the unmanned vehicle at the first wide path point and an end speed and an end direction angle of the unmanned vehicle at the second wide path point. It features.

)는

를 만족하고, 이 경우

는 상기 제 1 궤적(

)과 상기 기설정된 기준점(

)에 대한 경로 제어 매개 변수(path control parameter)인 것을 특징으로 한다.In addition, the plurality of candidate paths (

)

Satisfies this case,

Is the first trajectory (

) And the preset reference point (

It is characterized in that the path control parameter (path control parameter) for.

)의 초기값은 랜덤값이며, 소정의 횟수 동안 상기 제어 매개 변수를 갱신하면서 상기 복수개의 후보 경로가 반복적으로 생성되는 것을 특징으로 한다.In addition, the path control parameter (

) Is a random value, and the plurality of candidate paths are repeatedly generated while updating the control parameter for a predetermined number of times.

In addition, the candidate path having the lowest cost among the plurality of repeatedly generated candidate paths is selected as the optimal path.

The selecting of the optimal path may include using gradient information of the continuous spatial mobility information.

Further, the cost for each of the candidate paths is characterized in that it is also based on at least one of kinematic constraints and boundary conditions.

In addition, an end position of the optimal path may be used as an initial position for generating an optimal path for the second wide path point and the third wide path point.

)을 생성하고, 기설정된 기준점(

)과 상기 생성된 제 1 궤적(

)에 대해 종말 위치가 free 인 복수개의 후보 경로를 생성하고, 상기 연속 공간 가동성 정보에 기초하는 상기 복수개의 후보 경로 각각에 대한 비용을 산출하고, 상기 복수개의 후보 경로 중에서 가장 적은 비용이 산출된 후보 경로를 최적 경로로서 선택하는 것을 특징으로 한다.In addition, an apparatus for generating an optimal route for an unmanned vehicle according to an embodiment of the present invention, a continuous spatial information generation unit for generating continuous spatial mobility information based on the navigation information traveled by the unmanned vehicle, and the optimum route generation unit The optimal path generation unit may include a first trajectory between a first wide path point and a second wide path point among a plurality of wide path points on the wide path;

), And the preset reference point (

) And the generated first trajectory (

Generate a plurality of candidate paths having an end position free of each other, calculate a cost for each of the plurality of candidate paths based on the continuous spatial mobility information, and calculate the least cost among the plurality of candidate paths. The route is selected as an optimal route.

The first trajectory may be generated based on an initial speed and an initial direction angle of the unmanned vehicle at the first wide path point and an end speed and an end direction angle of the unmanned vehicle at the second wide path point. It features.

)는

를 만족하고, 이 경우

는 상기 제 1 궤적(

)과 상기 기설정된 기준점(

)에 대한 경로 제어 매개 변수(path control parameter)인 것을 특징으로 한다.In addition, the plurality of candidate paths (

)

Satisfies this case,

Is the first trajectory (

) And the preset reference point (

It is characterized in that the path control parameter (path control parameter) for.

)의 초기값은 랜덤값이며, 소정의 횟수 동안 상기 제어 매개 변수를 갱신하면서 상기 복수개의 후보 경로가 반복적으로 생성되는 것을 특징으로 한다.In addition, the path control parameter (

) Is a random value, and the plurality of candidate paths are repeatedly generated while updating the control parameter for a predetermined number of times.

In addition, the candidate path having the lowest cost among the plurality of repeatedly generated candidate paths is selected as the optimal path.

In addition, the cost for each of the candidate paths may be calculated based on at least one of kinematic constraints and boundary conditions.

In addition, an end position of the optimal path may be used as an initial position for generating an optimal path for the second wide path point and the third wide path point.

)을 생성하는 단계와, 기설정된 기준점(

)과 상기 생성된 제 1 궤적(

)에 대해 종말 위치가 free 인 복수개의 후보 경로를 생성하는 단계와, 상기 연속 공간 가동성 정보에 기초하여, 상기 복수개의 후보 경로 각각에 대한 비용을 산출하는 단계와, 상기 복수개의 후보 경로 중에서 가장 적은 비용이 산출된 후보 경로를 최적 경로로서 선택하는 단계를 수행하게 하는 명령어를 포함하는 프로그램이 기록되어 있는 것을 특징으로 한다.In addition, as a computer-readable recording medium according to an embodiment of the present invention, the processor for generating a continuous spatial mobility information based on the navigation information traveled by the unmanned vehicle, and a plurality of wide path points on the wide path The first trajectory between the first wide path point and the second wide path point (

) And a predetermined reference point (

) And the generated first trajectory (

Generating a plurality of candidate paths having an end position free for the step C); calculating a cost for each of the plurality of candidate paths based on the continuous spatial mobility information; A program is recorded that includes instructions for performing a step of selecting a candidate path for which cost is calculated as an optimal path.

According to an embodiment of the present invention, in generating an optimal path for an unmanned vehicle, after generating a plurality of candidate paths for the wide area path points for the wide area path points arranged at predetermined intervals on the wide path, Stability during autonomous driving of unmanned vehicles by calculating the least expensive route among a plurality of candidate routes based on spatial mobility information and state information such as initial / end speed and direction angle of unmanned vehicles. To increase.

1 is a diagram illustrating a wide path and a wide path point on a conventional wide grid map;
Figure 2 is a detailed block diagram of an optimum path generation device for an unmanned vehicle according to an embodiment of the present invention.
3 is a detailed block diagram of a continuous spatial information generating unit according to an embodiment of the present invention.
4 to 7 are conceptual diagrams illustrating an optimal path generation operation between wide path points according to an embodiment of the present invention.
8 is an exemplary view of an optimal path on the entire wide area path according to an embodiment of the present invention.
9 is a flowchart illustrating an operation of an optimum path generation apparatus for an unmanned vehicle according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings will be described in detail the operating principle of the present invention. In the following description of the present invention, when it is determined that a detailed description of a known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. Terms to be described later are terms defined in consideration of functions in the present invention, and may be changed according to intentions or customs of users or operators. Therefore, the definition should be made based on the contents throughout the specification.

FIG. 2 illustrates a detailed block diagram of an optimum path generation device for an unmanned vehicle according to an embodiment of the present invention.

Hereinafter, an operation of each component of the optimum path generation device 200 for an unmanned vehicle according to an embodiment of the present invention will be described in detail with reference to FIG. 2.

First, the continuous spatial information generation unit 210 generates continuous spatial possibility information based on at least one of wide path navigation information and environment recognition information on which an unmanned vehicle travels.

As shown in FIG. 3, the continuous spatial information generator 210 may include details such as the navigation information generator 300, the environment recognition information generator 302, the mobility data generator 304, and the continuous spatial mobility information generator 306. It may include components.

The navigation information generator 300 obtains the absolute coordinates of the path traveled by the unmanned vehicle and generates the navigation information of the unmanned vehicle when the unmanned vehicle autonomously travels according to the wide path planning.

In this case, the navigation information may refer to global positioning system (GPS) based location information on a route on which the unmanned vehicle travels. That is, the navigation information generator 300 is equipped with a GPS system to obtain GPS information about the current position of the unmanned vehicle and the route traveled by the unmanned vehicle, and to calculate absolute coordinates of the route on which the unmanned vehicle travels based on the GPS information. Create

The environmental recognition information generator 302 is connected with environmental recognition apparatuses such as a camera, a lidar, a radar, and the like to recognize environmental information on a surrounding area of a predetermined range on a route on which an unmanned vehicle travels. Create

At this time, the lidar device of the environment recognition apparatus as described above detects the terrain characteristics and obstacles by scanning the terrain of the predetermined area in front of the driverless vehicle. The size of the region may be set to about 40m × 40m, but is not limited thereto.

Accordingly, the environment recognition information generator 302 generates environment awareness information about a path around the unmanned vehicle based on the terrain characteristics and obstacles in front of the driverless vehicle provided from the lidar apparatus. Such environment recognition information may mean, for example, information about an area in which the vehicle can travel among the surrounding areas of the path on which the unmanned vehicle travels.

The mobility data generator 304 may generate absolute coordinates corresponding to a path and a surrounding area in which an unmanned vehicle travels based on the navigation information and the environmental recognition information generated by the navigation information generator 300 and the environmental awareness information generator 302, respectively. Recognize and generate label information indicating whether the unmanned vehicle can travel by absolute coordinates. In addition, the mobility data generator 304 stores label information for each absolute coordinate in the mobility DB 306.

The continuous space mobility information generator 308 performs machine learning based on the label information for each absolute coordinate stored in the mobility DB 306 to define the probability of driving in the continuous space between the absolute coordinates.

)와 주행 불가능 데이터(

)가 정의될 수 있다.In addition, the continuous space mobility information generator 308 generates mobility information on the continuous space by generating label information indicating whether the vehicle can travel by absolute coordinates corresponding to the continuous space based on the driving probability defined above. Stored in the DB 306. For example,

) And non-driving data (

) Can be defined.

과 학습이 필요한 가중치

로 구성된 로지스틱 회귀 분류기(Logistic regression classifier)를 통해 기계 학습을 수행하여 무인 차량이 주행한 경로와 경로 주변의 주행 가능한 영역에 대응되는 절대 좌표들 사이의 연속 공간에 대한 주행 가능 확률을 정의한다. 이때, 연속 공간 가동성 정보 생성기(308)는 연속 공간상에 가동성 정보를 정의하기 위해 Hilbert maps 기법을 활용할 수 있으나, 이에 한정되는 것은 아니다.That is, the continuous space mobility information generator 308 is a sparse kernel as shown in [Equation 1] below based on the driving enable / disable data in the two-dimensional space obtained from the movable DB 306.

And weights that require learning

Machine learning is performed through a logistic regression classifier, which defines a probable probability of continuous space between absolute coordinates corresponding to a path driven by an unmanned vehicle and a driveable area around the path. In this case, the continuous space mobility information generator 308 may use the Hilbert maps technique to define mobility information on the continuous space, but is not limited thereto.

Therefore, when the mobility information on the continuous space is used as described above, the wide path can be displayed to have continuity, so even if the resolution of the wide path map is increased, there is no inconsistency between the navigation information mounted on the unmanned vehicle and the actual road environment. .

Meanwhile, the optimal path generation unit 220 uses the continuous spatial mobility information generated by the continuous spatial information generation unit 210 to generate a grid-based global path 100 having discontinuities generated on a wide-area grid-based map. By generating a wide path having a continuity, the problem is that the wide path 100 does not match the actual road.

In addition, the optimum path generation unit 220 generates an optimal path connecting the respective wide path points with respect to the wide path points 102 and 104 disposed at predetermined intervals on the wide path 100.

At this time, in generating the optimal path as described above, the optimal path generation unit 220 optimizes the path between the wide area path points (102, 104) arranged at a predetermined interval on the wide path 100 in accordance with the actual road environment To generate a path, a first trajectory is generated between a first wide path point 102 and a second wide path point 104 among a plurality of wide path points on the wide path, and is generated in a predetermined reference point and the generated first trajectory. After generating a plurality of S candidate paths having a free end position for the S, a path having the least cost is generated as an optimal path. Here, in generating candidate paths connecting the wide path points 102 and 104, the optimum path generator 220 generates a plurality of stop points between the two wide path points 102 and 104 and passes through the plurality of stop paths. Candidate paths can be generated by concatenating the points.

That is, in one embodiment of the present invention, since the wide path 100 as described above may be defined as a path having continuity based on continuous spatial mobility information, the optimal path generation unit 220 may minimize the cost of the path control parameter. It is possible to calculate the PCP (path control parameter) more accurately, and the path control parameter (PCP) can be more accurately matched with the actual road environment. In this case, the above path control parameters and costs will be described later in detail in the operation of the optimum path generator 220 referring to FIGS. 4 to 7.

In this case, the optimum path generation unit 220 may use a subspace trajectory optimization in generating an optimal path between each of the wide area path points, but is not limited thereto.

4 to 7 illustrate an optimal path by calculating the least expensive candidate path among a plurality of candidate paths for the first wide path point 102 and the second wide path point 104 according to an embodiment of the present invention. An operation concept of setting is shown in FIG. 6, which is a drawing when a predetermined cost calculation repetition number is performed once, 50 times, 100 times, and 150 times, respectively.

4 to 7, it can be seen that S candidate paths 400, 500, 600, and 700 are generated between two wide path points 102 and 104 according to the path control parameter λ.

In addition, it can be seen that these S candidate paths 400, 500, 600, and 700 gradually converge closer to the wide-area path 100 in accordance with the number of update of the path control parameter according to the cost calculation.

)을 생성한다. 또한, 기설정된 기준점(

)(410)과 Prey 궤적(

)에 대해 종말 위치가 free 인 S개의 후보 경로(400, 500, 600, 700)를 생성하고, 연속 공간 가동성 정보에 기초하는 S개의 후보 경로(400, 500, 600, 700) 각각에 대한 비용을 산출한다. 이때, 종말 위치가 free라는 의미는 생성되는 후보 경로의 종말 위치가 기준점(

)과 제2 광역 경로점(104)을 잇는 직선위의 임의의 지점이 될 수 있음을 의미한다.At this time, the optimal path generation unit 220 of the plurality of wide path points (102, 104) on the wide path 100 between the first wide path point 102 and the second wide path point (104)

) In addition, the preset reference point (

) 410 and Prey trajectory (

We generate S candidate paths (400, 500, 600, 700) whose end positions are free, and calculate the cost for each of the S candidate paths (400, 500, 600, 700) based on continuous spatial mobility information. Calculate In this case, the end position is free means that the end position of the generated candidate path is a reference point (

) May be any point on a straight line connecting the second wide path point 104.

Next, the optimal path generation unit 220 selects the candidate paths with the lowest cost among the S candidate paths 400, 500, 600, and 700 as the optimal paths 402, 502, 602, and 702.

Here, the first trajectory is generated based on the initial speed and the initial direction angle of the unmanned vehicle at the first wide path point 102 and the end speed and the end direction angle of the unmanned vehicle at the second wide path point 104. Can be.

)는 아래의 [수학식 2]와 같이 나타낼 수 있다.In addition, S candidate paths (400, 500, 600, 700) (

) Can be expressed as Equation 2 below.

는 Prey 궤적(

)과 기설정된 기준점(

)에 대한 경로 제어 매개 변수(path control parameter)를 의미한다. 이때, 경로 제어 매개 변수(

)의 초기값은 랜덤값이며, 이러한 경로 제어 매개 변수는 소정의 횟수 동안 갱신되면서 복수개의 후보 경로가 반복적으로 생성될 수 있다.In Equation 2 above,

Is the Prey trajectory (

) And the preset reference point (

Means a path control parameter for. At this time, the path control parameters (

) Is a random value, and this path control parameter is updated for a predetermined number of times, and a plurality of candidate paths may be repeatedly generated.

The optimal path generator 220 selects the candidate path having the lowest cost among the plurality of candidate paths repeatedly generated as described above as the optimal path.

In addition, the optimal path generation unit 220 may calculate the cost for each candidate path based on at least one of kinematic constraints and boundary conditions of the unmanned vehicle.

In addition, when the optimal path generation unit 220 calculates an optimal path for the first wide path point 102 and the second wide path point 104 as described above, the optimal path generation unit 220 determines the end position of the optimum path through the second wide path point 104. ) And the third wide path point 106 after the second wide path point 104.

In addition, when the optimal path generation unit 220 generates all the optimal paths 402, 502, 602, and 702 between each of the wide path points 102 and 104, the optimal path generation unit 220 connects all the optimal paths as shown in FIG. 8. This will create an overall optimal path 800 between the wide path points of the unmanned vehicle that minimizes phase costs.

Hereinafter, an operation of generating an optimal path in which the cost between the wide path points 102 and 104 is minimized in the optimum path generation unit 220 will be described in more detail.

), 기준점

(410)에 대해 경로 제어 매개 변수

를 조절하여 S개의 후보 경로(

)(400, 500, 600, 700)를 위 [수학식 2]를 통해 생성시킬 수 있다.That is, the optimal path generation unit 220 is based on a subspace trajectory optimization technique and has a given Prey trajectory (

), Benchmark

Path control parameters for 410

To adjust the S candidate paths (

) (400, 500, 600, 700) can be generated through the above [Equation 2].

In addition, the first and second derivatives for Equation 1 above may be obtained as shown in Equation 3 below.

를 최소화하는 경로 제어 매개 변수

를 찾는 문제로 정의할 수 있다.In addition, the subspace path optimization problem satisfies the inequality constraints of [Equation 5] and the equation constraints of [Equation 6] below, and the cost of [Equation 4].

Path control parameters to minimize

Can be defined as the problem of finding.

On the other hand, in one embodiment of the present invention can be considered as the kinematic model constraint condition of the unmanned vehicle in Equation 7 below as the equation constraint condition of Equation 6 above.

과 아래의 [수학식 8]의 관계로부터 무인 차량의 속도, 방향각, 제어입력을 얻을 수 있다.here

The speed, direction angle, and control input of the unmanned vehicle can be obtained from the relationship between and [Equation 8] below.

사이의 Legendre-Gauss-Lobatto(LGL) 점들을 이용하여 [수학식 4]~[수학식 6]은 이산화(Discrete)된 부분 공간 상에서의 경로 제어 매개 변수

를 최적화하는 문제로 아래의 [수학식 9]~[수학식 11]에서와 같이 재정의 할 수 있다.At this time,

Using Legendre-Gauss-Lobatto (LGL) points between Equations 4 and 6 are path control parameters on discrete subspaces.

The problem of optimizing can be overridden as shown in [Equation 9] to [Equation 11] below.

와 1차 및 2차 미분은 아래의 [수학식 12]에서와 같이 정의할 수 있다.Then, the candidate path trajectories on the discretized subspace

And the first and second derivatives can be defined as in Equation 12 below.

는 아래의 [수학식 13]에서와 같이 매핑(mapping)을 통해 정의된다.here,

Is defined by mapping as shown in [Equation 13] below.

(410)과 Prey의 궤적

(420)에 대해

번째 경로 제어 매개 변수

를 조절하여 후보 경로 궤적

(400)를 위 [수학식 12]에 의해 생성할 수 있다.That is, as shown in Figure 4 predefined reference point

410 and Prey's trajectory

About 420

Path control parameters

The candidate path trajectory by adjusting the

400 may be generated by Equation 12 above.

In addition, in an embodiment of the present invention, in order to generate a Prey trajectory 420 that satisfies the initial direction angle / speed and the end direction angle / speed of the unmanned vehicle, a third order polynomial function is generated as shown in Equation 14 below. Can be used.

,

)과, [수학식 2]에서

인 경우,

인 관계를 통해 위 [수학식 14]에서의 계수(

)는 아래의 [수학식 15]에서와 같이 구할 수 있다. Where the initial / end time of the driverless vehicle (

,

), And in [Equation 2]

If is

The coefficients in Equation 14

) Can be obtained as shown in Equation 15 below.

), 속도(

), 방향각(

)에 따라

를 아래의 [수학식 16]에서와 같이 얻을 수 있다.In Equation 15 above, the position at the initial time and the end time,

), speed(

), Direction angle (

)Depending on the

Can be obtained as shown in Equation 16 below.

Referring again to FIG. 4, the Prey trajectory 420 considering the initial speed / direction angle of the current unmanned vehicle and the terminal speed / direction angle of the sampled wide path points 102 and 104 may be identified.

(410)과

에 대해 종말 위치 프리(Free)인 후보 경로(400)를 생성하기 위해 아래의 [수학식 17]에서와 같이

번째 노드의 경로 제어 매개 변수

를

개의 하모닉 함수의 조합 및 3차 다항함수와 균등 임의의 수를 통해 조절하여

개의 후보 경로(400)들을 랜덤(random)하게 생성한다.Where given

410 and

In order to generate the candidate path 400 that is the end position Free for the Equation 17, as shown in Equation 17 below.

Control parameters on the first node

To

Combination of harmonic functions and three random polynomials and equal random numbers

Randomly generate two candidate paths 400.

개의 후보 경로

(400)들을 확인 할 수 있다.That is, as shown in Figure 4

Candidate paths

You can check the 400.

개의 후보 경로(400) 별 비용을 계산하기 위해 종말 시간

고정인 비용은 무인 차량의 가동성, 제어입력, 경계 조건에 대한 지표와 각 지표의 영향도를 조절하는 인자는

로 아래의 [수학식 18]과 같이 정의한다.Where

End time to calculate costs per two candidate paths 400

Fixed costs are factors that control the parameters of the unmanned vehicle's mobility, control inputs, boundary conditions, and the influence of each indicator.

It is defined as [Equation 18] below.

에 대한 편미분은 연쇄법칙(Chain rule)에 의해 아래의 [수학식 20]에서와 같이 정의할 수 있다.In order to find the solution of the optimization problem defined so far, optimization is performed by applying Particle Swarm Optimization (PSO). In addition, in an embodiment of the present invention, since continuous spatial mobility information is used, mobility gradient information may be utilized in a PSO update process. For this purpose, the path control parameter of [Equation 19] expressing the mobility information.

The partial derivative for can be defined as in [Equation 20] by the chain rule.

에 대한 편미분

는 아래의 [수학식 21]에서와 같이 정의한다.Here, the sparse kernel defined by Equation 22 below

Partial differential for

Is defined as in Equation 21 below.

In addition, Equation 21 may be defined as Equation 23 below by the chain law, and each partial derivative may be obtained through Equation 24 and Equation 25 below.

에 대한 편미분

는 아래의 [수학식 26]에서와 같이 정의할 수 있다.And, in [Equation 2]

Partial differential for

May be defined as in Equation 26 below.

FIG. 9 illustrates an operation control flow in the optimum path generation device 200 for an unmanned vehicle according to an embodiment of the present invention. Hereinafter, an embodiment of the present invention will be described in detail with reference to FIGS. 1 to 9.

First, the optimum path generation apparatus 200 receives continuous spatial mobility information for forming the wide path 100 having continuity (S900).

In addition, the optimum path generation device 200 receives the current state information of the unmanned vehicle including the initial speed, the end speed, the direction angle, the kinematic constraint condition, the boundary condition, etc. of the unmanned vehicle so that the unmanned vehicle is on the wide path 100. Check the information, such as the current speed, direction at the location (S902).

In addition, the optimum path generating apparatus 200 may determine the optimum path 402 between each of the wide path points 102 and 104 with respect to the wide path points 102 and 104 arranged at predetermined intervals on the wide path 100. Randomly generates a value of the initial path control parameter for generating (S904).

The optimum path generator 200 generates S candidate paths 400 for the two wide path points 102 and 104 for which the optimum path 402 is to be obtained using the initial path control parameters ( S906).

)과 Prey 궤적(

)에 대해 종말 위치가 free 인 복수개의 후보 경로를 생성할 수 있다.That is, the optimum path generating apparatus 200 generates a predetermined reference point in generating the candidate path 400.

) And Prey trajectory (

For example, a plurality of candidate paths having free terminal positions may be generated.

At this time, in generating the candidate paths as described above, the optimum path generation device 200 generates a plurality of waypoints 404 between the two wide path points 102 and 104 as shown in FIG. The candidate path 400 may be generated by connecting the points 404.

In operation S908, the optimal path generation apparatus 200 calculates the costs for the S candidate paths 400. In this case, the cost is a value indicating whether it is appropriate for the unmanned vehicle to proceed in the corresponding path, and the above-described calculation of the cost reflects the continuous space mobility information and the state information of the unmanned vehicle. Based on the points 102 and 104, the goodness of fit of the candidate path 400, which is formed near the area in which the wide-area path 100 is formed or formed in a direction similar to the direction in which the driverless vehicle is directed, may be calculated.

In addition, the optimum path generating apparatus 200 checks whether the number of cost calculations for the candidate path 400 as described above exceeds a preset reference iteration number, and updates the path control parameter until the reference iteration number is reached. The cost calculation for the candidate paths 400 formed between the wide path points 102 and 104 is repeatedly performed (S910 and S912).

According to the repetition of the above cost calculation, the S candidate paths 400 formed between the wide path points 102 and 104 are gradually increased with respect to the wide path points 102 and 104 as shown in FIGS. 4 to 7. It is close to the actual wide path.

When the number of reference repetitions exceeds the optimal path generating apparatus 200, the optimal path generating apparatus 200 detects candidate paths having the least cost among the S candidate paths 400 (S914), and detects the detected candidate paths according to the corresponding wide path point 102. The optimal path 402 is set to 104 (S916).

As described above, when all the optimal paths between all the wide path points on the wide path 100 are generated, as shown in FIG. 8, the optimal path 800 of the unmanned vehicle that minimizes the cost on the entire wide path 100 may be generated. do.

As described above, according to one embodiment of the present invention, in generating an optimal path for an unmanned vehicle, a plurality of candidate paths for the wide area path points for the wide area path points arranged at predetermined intervals on the wide path; After generating, we calculate the least cost among the candidate paths based on continuous space mobility information and initial information such as initial / end speed and direction angle of unmanned vehicle to generate the optimal path for wide area path. To improve the stability of autonomous driving of the vehicle.

Combinations of the steps of each flowchart attached to the present invention may be performed by computer program instructions. These computer program instructions may be mounted on a processor of a general purpose computer, special purpose computer, or other programmable data processing equipment such that the instructions performed through the processor of the computer or other programmable data processing equipment are described in each step of the flowchart. It will create a means to perform them. These computer program instructions may be stored in a computer usable or computer readable memory that can be directed to a computer or other programmable data processing equipment to implement functionality in a particular manner, and thus the computer usable or computer readable memory. It is also possible for the instructions stored therein to produce an article of manufacture containing instruction means for performing the functions described in each step of the flowchart. Computer program instructions may also be mounted on a computer or other programmable data processing equipment, such that a series of operating steps may be performed on the computer or other programmable data processing equipment to create a computer-implemented process to create a computer or other programmable data. Instructions for performing the processing equipment may also provide steps for executing the functions described in each step of the flowchart.

In addition, each step may represent a module, segment or portion of code that includes one or more executable instructions for executing a specified logical function (s). It should also be noted that in some alternative embodiments, the functions noted in the steps may occur out of order. For example, the two steps shown in succession may in fact be performed substantially simultaneously or the steps may sometimes be performed in the reverse order, depending on the function in question.

Meanwhile, in the above description of the present invention, specific embodiments have been described, but various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the invention should be determined by the claims rather than by the described embodiments.

210: continuous spatial information generation unit 220: optimal path generation unit

Claims (16)

As a method of generating an optimal route for an unmanned vehicle,
Generating continuous spatial mobility information based on the navigation information driven by the unmanned vehicle,
A first trajectory between the first wide path point and the second wide path point among the plurality of wide path points on the wide path;

),
The preset reference point (

) And the generated first trajectory (

Generating a plurality of candidate paths of which the end position is free for.
Calculating a cost for each of the plurality of candidate paths based on the continuous spatial mobility information;
Selecting a candidate path having the lowest cost among the plurality of candidate paths as an optimal path;
How to create the best route. The method of claim 1,
The first trajectory is generated based on an initial speed and an initial direction angle of the unmanned vehicle at the first wide path point and an end speed and an end direction angle of the unmanned vehicle at the second wide path point.
How to create the best route. The method of claim 1,
The plurality of candidate paths (

)

Satisfies this case,

Is the first trajectory (

) And the preset reference point (

Path control parameter for
How to create the best route. The method of claim 3, wherein
The path control parameters (

) Is the random value,
The plurality of candidate paths are repeatedly generated while updating the control parameter for a predetermined number of times.
How to create the best route. The method of claim 4, wherein
The candidate path having the lowest cost among the plurality of repeatedly generated candidate paths is selected as the optimal path.
How to create the best route. The method of claim 3, wherein
The step of selecting as the optimal path uses the gradient information of the continuous spatial mobility information.
How to create the best route. The method of claim 1,
The cost for each of the candidate paths is also based on at least one of kinematic constraints and boundary conditions.
How to create the best route. The method of claim 5,
The end position of the optimal path is used as an initial position for generating an optimal path for the second wide path point and the third wide path point.
How to create the best route. A device for generating an optimal route for an unmanned vehicle,
A continuous spatial information generator for generating continuous spatial mobility information based on navigation information driven by an unmanned vehicle,
Including the optimal path generator,
The optimum path generation unit,
A first trajectory between the first wide path point and the second wide path point among the plurality of wide path points on the wide path;

),
The preset reference point (

) And the generated first trajectory (

Generate a plurality of candidate paths with the end position free for
Calculate a cost for each of the plurality of candidate paths based on the continuous spatial mobility information,
Selecting a candidate path having the lowest cost among the plurality of candidate paths as an optimal path
Optimal path generation device. The method of claim 9,
The first trajectory is generated based on an initial speed and an initial direction angle of the unmanned vehicle at the first wide path point and an end speed and an end direction angle of the unmanned vehicle at the second wide path point.
Optimal path generation device. The method of claim 9,
The plurality of candidate paths (

)

Satisfies this case,

Is the first trajectory (

) And the preset reference point (

Path control parameter for
Optimal path generation device. The method of claim 11,
The path control parameters (

) Is the random value,
The plurality of candidate paths are repeatedly generated while updating the control parameter for a predetermined number of times.
Optimal path generation device. The method of claim 12,
The candidate path having the lowest cost among the plurality of repeatedly generated candidate paths is selected as the optimal path.
Optimal path generation device. The method of claim 9,
The cost for each of the candidate paths is also calculated based on at least one of kinematic constraints and boundary conditions.
Optimal path generation device. The method of claim 13,
The end position of the optimal path is used as an initial position for generating an optimal path for the second wide path point and the third wide path point.
Optimal path generation device. A computer readable recording medium,
Let the processor
Generating continuous spatial mobility information based on the navigation information driven by the unmanned vehicle,
A first trajectory between the first wide path point and the second wide path point among the plurality of wide path points on the wide path;

),
The preset reference point (

) And the generated first trajectory (

Generating a plurality of candidate paths of which the end position is free for.
Calculating a cost for each of the plurality of candidate paths based on the continuous spatial mobility information;
Selecting a candidate path having the lowest cost among the plurality of candidate paths as an optimal path;
The program containing the instructions to be executed is recorded.
Computer-readable recording media.

Images