KR101620911B1 - Auto Pilot Vehicle based on Drive Information Map and Local Route Management Method thereof - Google Patents

Abstract

The unmanned vehicle of the present invention is provided with a geographic information providing unit 30 for converting information available for recognition and path planning of an unmanned vehicle from connection information and attributes of a space included in a FDB (Feature Database) or DRM (Digital Road Map) A travel route estimating unit 40 for calculating an estimated travel route from the linear connection information on the space searched by the geographic information providing unit 30, a time varying characteristic of the sensor for providing the location information by comparing the recognized travel route types And an error estimating unit 50 for calculating a position error from the error. In the unmanned vehicle using such a configuration, the connection information and attributes of the space are detected by the geographical information of the traveling space in which the location and the traveling route are recognized, The position error between the estimated traveling route and the obtained traveling route is estimated while estimating the traveling route from the linear connection information for the curved region of the space, As a result yirueojim path that calculates and outputs the traveling speed limit value has a characteristic that the accuracy is greatly improved for a drive that contains a path control while the curve of the unmanned vehicle.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0002] The present invention relates to a map management system,

The present invention relates to an unmanned vehicle, and more particularly, to an unmanned vehicle based on a travel information map capable of generating and providing information necessary for performing a regional route plan by using a travel information map in a road including a curve, Management method.

Generally, autonomous driving of an unmanned vehicle is required to establish a safe route plan based on sufficient recognition information about the moving space to be moved. For this, it is necessary to recognize the driving environment by the sensor, A search is performed.

For example, when establishing a route plan in a limited running space, such as a road or a non-running road, sufficient information is obtained about a forward running route to which the onboard sensor should travel, and a direction in which the unmanned vehicle must move And the path planning using the extracted spatial information is established, thereby moving the unmanned vehicle.

Korean Patent Publication No. 10-2009-0108509 (October 15, 2009)

However, in the path planning of the unmanned vehicle using the mounted sensor, it is inevitable that the forward traveling road condition to be moved must be acquired with sufficient recognition information in the mounted sensor.

For example, when an unmanned vehicle is performing a route planning for a road with a curved section, there is a lack of recognition information on the road due to limitations of the detection area due to the performance of the onboard sensors, Will face.

Therefore, if the unmanned vehicle is confronted with a space including a curved road, it is impossible to plan a route to a space including a curved road due to a lack of information only by recognizing the movable area from the mounted sensor The situation is bound to happen.

In view of the above, the present invention is based on a curved road traveling route estimation based on travel information, and a location correction is performed through comparison with curve travel route recognition information obtained from an onboard sensor, so that traveling on a road including a curve is smooth And to provide a route planning management method using the same.

In order to accomplish the above object, there is provided a method of managing a local route plan of an unmanned vehicle using a running information map of the present invention, the method comprising the steps of: (A) detecting a location and a traveling space from an unmanned vehicle, Detecting information and attributes; (B) estimating a traveling route from the linear connection information on a curve section of the space in which the unmanned vehicle is searched; (C) calculating a position error generated between the estimated travel route and the travel route acquired by the onboard sensor; (D) applying the position error to the travel map creation together with the estimated travel route information and the on-board sensor travel route information; (E) calculating route points from the running map; (F) calculating a traveling speed limit value using the curvature of the estimated traveling route; (G) applying the route point and the travel speed limit value to execute the route control of the unmanned vehicle; As shown in FIG.

The error calculation in the step (C) is performed to correct the error due to the travel route position information displayed on the travel information map and the time-varying characteristics of the position sensor acquired during the running of the unmanned vehicle.

The error calculation is performed by (c-1) dividing the calculated travelable area by a certain lattice size based on the traveling direction of the unmanned vehicle, and then connecting the center points thereof to extract the travel route center point for the travelable area, (C-3) calculating a slope and a slope of a straight line of the points constituting the set of N points, (c-4) calculating a slope and a slice of the straight line of the points constituting the set of N points, (C-5) if the estimated distance is larger than the reference distance, a new set of N points is constructed, and then the estimated distance is compared with the reference distance, (C-6) a point existing on a similar straight line is included in the set of N points when the estimated distance is smaller than the reference distance, while (c-6) Until we find a group of sets Repeatedly performed for a group of points existing on a straight line as the center point to a group and travel. (c-7) collective information of points classified into points on the same straight line at the center point of the travel route and segmental linear connection information on the travel information map are defined as a cost function, and from the cost function, And the position error is calculated by calculating a value that minimizes the rotational and translational distance between line segments estimated from the center of the road.

According to another aspect of the present invention, there is provided an unmanned vehicle based on a running information map including a position estimating unit estimating a position of an unmanned vehicle; An estimated running road recognizing unit that recognizes a running road in a space where the unmanned vehicle runs; A geographic information providing unit for converting information available for recognition and path planning of the unmanned vehicle from connection information and attributes of a space included in an FDB (Feature Database) or a DRM (Digital Road Map); A driving route estimating unit for calculating an estimated driving route from the linear connection information for the space searched in the geographic information providing unit; An error estimator for calculating a position error from an error by a time-varying characteristic of a sensor that compares a travel route type recognized by the estimated travel route recognizing unit and provides position information to the position estimating unit; A running map generating unit for generating a running map by the position error and the estimated running route; A route path planner for establishing a route point for a traveling route using the traveling map; A curvature estimator for calculating a curvature with respect to a curve section of the traveling passage; a traveling speed limiting unit for calculating a traveling speed limit value based on the curvature calculation value; A path controller for performing path control using the path point establishment and the traveling speed limit value; Is included.

The position estimator is linked to a Global Positioning System (GPS) / Inertial Navigation System (INS). The traveling road recognizing unit is connected to a camera, a lidar, and a radar. The geographical information providing unit includes a driving information management unit for searching space information based on a location in cooperation with a map search unit and providing driving information of a certain space in which the unmanned vehicle is located.

The present invention is advantageous in that information can be generated in a region including a traveling route of a curve lacking recognition information on a traveling space from a mounted sensor based on the traveling information map when establishing the local route planning of the unmanned vehicle .

Further, according to the present invention, the available information generated in the area including the curved road is provided to the local route planning of the unmanned vehicle, thereby greatly improving the safety of the unmanned vehicle traveling according to the local route plan.

2 is a flowchart of a method for managing a regional route plan of an unmanned vehicle based on a travel information map according to the present invention, and FIG. 3 FIG. 4 is a flowchart illustrating a method of classifying a traveling route center according to the present invention into a group of points existing on the same straight line according to the present invention, and FIG. 5 FIG. 6 is a detailed procedure of FIG. 4, and FIG. 6 is a flowchart illustrating a detailed procedure of the rotation and translational movement, which minimizes the error by comparing the segmented linear travel route estimated from the travel route central point of the travel route recognition section, This is an example of the process of calculating the value.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings, which illustrate exemplary embodiments of the present invention. The present invention is not limited to these embodiments.

1 shows a configuration of a traveling information map based unmanned vehicle equipped with a local route planning management device according to the present embodiment.

As shown in the figure, the unmanned vehicle includes a position estimating unit 10, a traveling road recognizing unit 20, a geographic information providing unit 30, a traveling road estimating unit 40, an error estimating unit 50, A local route planning unit 70, a curvature estimating unit 80, a traveling speed restricting unit 90, and a path control unit 100 are included.

The position estimating unit 10 estimates the position of the unmanned vehicle. To this end, the position estimating unit 10 is connected to a GPS (Global Positioning System) / INS (Inertial Navigation System) 10-1, And operates in conjunction with information.

The traveling road recognizing unit 20 recognizes a traveling route of a space in which the unmanned vehicle travels. To this end, the traveling road recognizing unit 20 includes a camera 20-1, a Lidar 20 -2) and a radar (Radar) 20-3, and operates in conjunction with the information provided therefrom. Generally, a camera 20-1, a Lidar 20-2, and a radar 20-3 are referred to as mounted sensors of an unmanned vehicle.

The geographical information providing unit 30 is divided into a driving information map unit 30-1 and a map searching unit 30-2.

The driving information leadership unit 30-1 extracts from the connection information and attributes of the space included in the FDB (Feature Database) and DRM (Digital Road Map), which are kinds of geographical information, information conversion . In particular, the driving information map unit 30-1 searches the space information based on the position in cooperation with the map searching unit 30-2, thereby providing driving information of a certain space in which the unmanned vehicle is currently located.

The travel route estimating unit 40 estimates the traveling route from the linear connection information for the space searched from the traveling information leader 30-1.

The error estimating unit 50 estimates the shape of the traveling route recognized by the camera 20-1, Lidar 20-2, and Radar 20-3, which are mounted sensors, And calculates the position error caused by the error of the GPS / INS 10-1.

The travel map generating unit 60 generates the travel map information based on the position error calculated by the error estimating unit 50 and the travel route information extracted from the travel information map by the travel route estimating unit 40, , The Lidar (20-2), and the Radar (20-3).

The regional route planning unit 70 receives the driving map generated by the driving map generating unit 60, and plans the route point through the received driving map. In addition, the traveling speed limiting unit 90 receives the limiting value of the traveling speed.

The curvature estimator 80 estimates the curvature of the estimated traveling path.

The traveling speed restricting unit 90 sets a traveling speed at which the unmanned vehicle can travel safely from the traveling path curvature estimated by the curvature estimating unit 80 and sets the traveling speed as a limit value to the local path planning unit 70 .

The route control unit 100 receives the planned route point provided by the local route planning unit 70 and the traveling speed limit value provided by the traveling speed limiting unit 90 and performs route control of the unmanned vehicle using these information .

Meanwhile, FIG. 2 shows a flowchart of a method of managing a regional route plan of an unmanned vehicle based on a travel information map according to the present invention.

As in S10, the unmanned vehicle recognizes its own position and travel route. This includes a position estimator 10 connected to a Global Positioning System (GPS) / INS (Integrated Navigation System) 10-1, a camera 20-1, a Lidar 20-2, , And a runway recognition unit 20 associated with the radar 20-3.

Then, as shown in S20, the unmanned vehicle detects connection information and attributes of the space from the geographical information, and then searches for the space where the unmanned vehicle is currently located, such as S30. This is accomplished through a geographic information providing unit 30 divided into a driving information lead unit 30-1 and a map searching unit 30-2. Particularly, a FDB (Feature Database), a DRM Map and so on are used.

Then, as in S40, the unmanned vehicle estimates the traveling route to the searched space. This is done by estimating the traveling route from the linear connection information for the space searched by the traveling road estimation unit 40 from the traveling information leadership unit 30-1. This example is illustrated in FIG. As shown in the figure, the traveling path 200 is a curve section, and the curved section of the traveling path 200 is divided into space connection information 300 that is connected to a line, so that linear information on the spatial connection of the running information map, Is estimated.

Thereafter, as in S50, the unmanned vehicle calculates an error between the estimated traveling route and the traveling route acquired from the onboard sensor. The process of calculating this error is illustrated in FIGS. 4 to 6, the travel route location information displayed on the travel information map and the location information by the location estimation unit 10 of the unmanned vehicle traveling in the same position as the GPS / INS 10-1 It is based on the existence of error due to the time-varying characteristics of the sensor.

Therefore, the error estimating unit 50 determines whether or not the traveling road recognizing unit 20 is a camera 20-1, a Lidar 20-2, a Radar 20-3, And the travel route estimated by the travel route estimating unit 40 are compared with each other.

Fig. 4 (A) shows a case where the travelable area calculated by the travel route recognizing unit 20 is divided into a constant lattice size based on the traveling direction of the unmanned vehicle, and the center point of the travel route for the travelable area is extracted .

Fig. 4 (B) is a view for explaining a method for calculating an error from the center of the traveling route extracted from Fig. 4 (A) and the traveling route information on the running information map, .

FIG. 5 is a concrete example of FIG. 4 (B), and a detailed procedure of a process of classifying the center of a traveling route from the area into a group of points existing on the same straight line can be seen.

As shown, if there are points of D _s , ..., D _j on the same straight line, the set of points is defined as G _i (S51). Next, the slope and slice of the straight line of the points constituting G _i are obtained (S52). Thereafter, it is assumed that the distance between the next point D _{j + 1} and L _i is r _{j + 1} (S53). Next, the distance _{rj + 1} and the reference value r are compared with each other (S54). Next, if the distance _{rj + 1} is larger than the reference value r, it is classified as a point that is not on the line L _i to construct a new set of points G _j , and whether L _{j + 1} is on the same line A set of points to be obtained is obtained (S55-1). In contrast, when the distance r _{j + 1} is smaller than the reference value r D _{j + 1} is added to D _{j + 1} to G _i to be in the L _i and performs a check for the next D _{j + 2} (S55-2 ).

Thus, by performing S51 to S55-1 or S55-2 with respect to the center point of the running route, a group of points existing on the similar straight line can be found.

Then, in the error estimating unit 50, it is determined whether or not the set of straight lines formed by the group of points on the similar straight line found from the running route center point for the travelable area of the travel route recognizing unit matches the piecewise linear spatial information in the running information map . FIG. 6 shows an example of information required for this search process.

As shown in the figure, a set of points classified into the same straight line at the center of the traveling route is defined as G _i , and a piecewise linear connection information on the driving information map is defined as M _k .

Then, let d _{jj be} the distance between the straight lines L _i and M _k formed by the points included in G _j , l _{i be} the length of the line segment of the straight line L _i , σ _i ^{2 be} the variance between the points included in G _i and L _j The following cost function is defined (Equation 1).

------- (수식1)

------- (Equation 1)

By using the cost function (Equation 1), a value that minimizes the rotation Φ and the translational movement distance P (p _y , p _y ) between the line segment M _k running on the running information map and the line segment L _j estimated from the running route center Is obtained.

Then, the obtained? And P represent the position error between the running route in the running information map and the running route in the running route recognizing unit, and the error estimating unit 50 transfers this value to the regional route planning unit 70 The local path planning unit 70 is used for path planning.

Referring again to FIG. 2, S60 is a step of generating a traveling map based on the calculated position error, estimated travel route information, and on-board sensor travel route information calculated in S50. This is done through the travel map generator 60 linked to the error estimator 50 and the travel route estimator 40.

Step S70 is a step of calculating route points from the running map. This is done through the local route planning unit 70 which receives the running map generated by the running map generating unit 60.

S80 is a step of determining whether or not there is a traveling speed limitation. This is done through the traveling speed restricting unit 90 which receives the curvature estimated by the curvature estimating unit 80 with respect to the estimated traveling path. The traveling speed limiting unit 90 transmits the calculated traveling speed to the local path planning unit 70 as a limit value.

In step S100, the unmanned vehicle carries out the path control, which is performed by the route control unit 100, which receives the planned route point provided by the local route planning unit 70 and the traveling speed limit value provided by the traveling speed limiting unit 90 .

As described above, in the unmanned vehicle according to the present embodiment, information conversion that can be used for recognition and path planning of the unmanned vehicle is performed from connection information and attributes of a space included in the FDB (Feature Database) or DRM (Digital Road Map) The travel route estimating unit 40 calculates an estimated travel route from the linear connection information for the space searched by the geographic information providing unit 30 and the geographic information providing unit 30, And an error estimator 50 for calculating a position error from the error due to the time-varying characteristic of the provided sensor. In the unmanned vehicle using this configuration, the traveling space in which the position and the travel route are recognized is used as geographical information, And then estimates the traveling route from the linear connection information for the curved region of the searched space and calculates a position between the estimated traveling route and the obtained traveling route After calculating the difference, as this yirueojim running map generation and path calculation point and calculation accuracy of the travel speed limit value is included in the route control when the curve of the unmanned vehicle is traveling is greatly improved.

10: Position estimation unit 10-1: GPS / INS
20: Driving road recognition unit 20-1: Camera
20-2: Lidar 20-3: Radar
30: Providing geographic information 30-1: Driving information leadership
30-2: map search unit 40:
50: Error estimating unit 60: Travel map generating unit
70: local path planning unit 80: curvature estimation unit
90: traveling speed limiting unit 100: path control unit
200: Driveway 300: Space connection information

Claims (6)

(A) detecting connection information and attributes of a space after a location and a traveling space recognized by an unmanned vehicle are searched for a space from geographical information;
(B) estimating a traveling route from the linear connection information on a curve section of the space in which the unmanned vehicle is searched;
(C) calculating a value that minimizes the rotation and translational distance between the estimated travel route segment and the line segment estimated from the travel route center point of the unmanned vehicle, and calculating the calculated value from the estimated travel route and the travel Calculating a position error generated between the first and second electrodes;
(D) applying the position error to the travel map creation together with the estimated travel route information and the on-board sensor travel route information;
(E) calculating route points from the running map;
(F) calculating a traveling speed limit value using the curvature of the estimated traveling route;
(G) applying the route point and the travel speed limit value to execute the route control of the unmanned vehicle;
Wherein the route guidance is performed by using the map of the running route of the unmanned vehicle.

The method as claimed in claim 1, wherein the error calculation in the step (C) is performed to correct an error caused by the travel route position information displayed on the travel information map and the time-varying characteristic of the position sensor acquired during running of the unmanned vehicle A Method of Regional Route Planning Management of Unmanned Vehicle Using Driving Information Map.
A position estimating unit estimating a position of the unmanned vehicle on which the local path planning management method of the unmanned vehicle using the running information map of claim 1 or 2 is performed;
An estimated running road recognizing unit that recognizes a running road in a space where the unmanned vehicle runs;
A geographic information providing unit for converting information available for recognition and path planning of the unmanned vehicle from connection information and attributes of a space included in an FDB (Feature Database) or a DRM (Digital Road Map);
A driving route estimating unit for calculating an estimated driving route from the linear connection information for the space searched in the geographic information providing unit;
An error estimator for calculating a position error from an error by a time-varying characteristic of a sensor that compares a travel route type recognized by the estimated travel route recognizing unit and provides position information to the position estimating unit;
A running map generating unit for generating a running map by the position error and the estimated running route;
A route path planner for establishing a route point for a traveling route using the traveling map;
A curvature estimator for calculating a curvature with respect to a curve section of the traveling passage; a traveling speed limiting unit for calculating a traveling speed limit value based on the curvature calculation value;
A path controller for performing path control using the path point establishment and the traveling speed limit value;
Wherein the vehicle information is a map of the vehicle.
4. The unmanned vehicle according to claim 3, wherein the position estimating unit is associated with a Global Positioning System (GPS) / Inertial Navigation System (INS).
4. The unmanned vehicle according to claim 3, wherein the traveling road recognizing unit is associated with a camera, a lidar, and a radar.
The geographical information providing system according to claim 3, wherein the geographical information providing unit includes a driving information guidance unit in which space information is searched based on a location in cooperation with a map search unit, and travel information of a certain space in which the unmanned vehicle is located is provided. Information map based unmanned vehicle.

Images