KR101553898B1 - System and method for estimating position of autonomous vehicle using position information of geographic feature - Google Patents

Abstract

The present invention relates to a system and method for estimating an autonomous moving vehicle capable of correcting a navigation error using positional information of a feature point, the apparatus comprising: a travel information device configured by a navigation sensor to estimate travel information of an autonomous moving vehicle; A feature recognition device for recognizing features and absolute positions of surrounding environmental features; A feature storage device for storing features and absolute positions of the recognized feature; An integrated position error determination unit for determining an integrated position error using a position error based on the navigation sensor and a position error based on the feature information measured through the travel information of the travel information device; And an integrated navigation unit for correcting the INS error using the integrated position error output from the integrated position error determination unit.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for estimating an autonomous moving vehicle using position information of a feature point,

The present invention relates to a system and method for estimating an autonomous moving vehicle, which can correct a navigation error using position information of a feature point.

GPS (Global Positioning System) / INS (Inertial Navigation System) is used to estimate the position of an autonomous moving vehicle in a terrestrial environment. However, in the ground environment, the GPS signal is easily disconnected due to the surrounding environment (trees, buildings, etc.), and various sensors (image sensor, RIDA sensor, etc.) are being researched to correct the error of the INS. However, the image sensor and the Lidar sensor are widely used for position estimation in a limited environment mainly due to a large change in the position estimation performance depending on the surrounding environment. In the all-weather environment unaffected by the GPS, There is a limiting factor.

On the other hand, the speedometer is widely used to compensate the INS error in the environment where GPS is not available. However, when the INS error is corrected using the speedometer in an environment where GPS is not available, the cumulative error increases with time have.

It is an object of the present invention to provide an autonomous moving vehicle position estimation system and method capable of more accurately estimating a position estimation of an autonomous moving vehicle even in an environment where GPS is not available for a long time by correcting a navigation error using positional information of the feature point .

According to an aspect of the present invention, there is provided a system for estimating an autonomous moving vehicle, including: a traveling information device configured to include traveling sensors to estimate traveling information of an autonomous moving vehicle; A feature recognition device for recognizing features and absolute positions of surrounding environmental features; A feature storage device for storing features and absolute positions of the recognized feature; An integrated position error determining unit for determining an integrated position error using a position error based on the navigation sensor and a position error based on the feature information measured through the travel information of the travel information device; And an integrated navigation unit for correcting the INS error using the integrated position error output from the integrated position error determination unit.

The integrated position error determination unit may determine the integrated position error based on the navigation sensor in an environment where the GPS is available and the integrated position error based on the feature information based on the feature information in an environment in which GPS is not available.

The integrated position error determination unit calculates the difference between the absolute position of the detected feature and the integrated navigation position measured through the navigation sensor and determines the absolute position of the detected feature as the integrated position error if the calculated difference is larger than the tolerance , The integrated navigation position is determined as the integrated position error.

According to another aspect of the present invention, there is provided a method for estimating a position of an autonomous mobile vehicle, the method comprising: storing feature information and an absolute position of the feature; Detecting a feature of the corresponding feature by recognizing the surrounding feature feature during driving of the autonomous moving vehicle in an environment where GPS can not be used; Determining an integrated position error by comparing the absolute position corresponding to the feature of the detected feature and the integrated navigation position using the navigation sensor; And correcting a navigation error of the INS using the determined integrated position error.

Wherein the determining the integrated position error comprises: calculating a difference between an absolute position of the feature detected and an integrated navigation position measured using the navigation sensor; Comparing the calculated position difference with a tolerance; And determining the absolute position of the detected feature as the combined position error if the calculated position error is greater than the tolerance, and determining the integrated navigation position as the combined position error if the calculated position error is smaller.

The present invention relates to an autonomous mobile vehicle that performs navigation based on a navigation sensor such as a GPS, an INS (Inertial Measurement Unit) / INS and a speedometer, and corrects a navigation error using position information of periodic or non- It is possible to more accurately estimate the autonomous mobile vehicle position estimation even in an environment where the vehicle is not available for a long time.

In other words, the present invention not only improves the reliability of the autonomous driving due to a great reduction in the dependence on the GPS in the autonomous traveling of the autonomous moving vehicle, but also estimates the position estimation of the autonomous moving vehicle more accurately in the all- .

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a configuration diagram of a position estimation system for an autonomous moving vehicle according to the present invention; FIG.
FIG. 2 is a conceptual diagram of feature recognition using sensor information of an autonomous moving vehicle. FIG.
3 is a flowchart showing a method of storing features and location information of a feature.
4 is a flow chart for retrieving location information of a feature stored in a storage device;
5 illustrates integrated navigation position correction using feature location information;
6 is a conceptual diagram of an integrated position estimating apparatus utilizing multiple information.
7 is a conceptual diagram for performing position error correction using the position estimation system of an autonomous mobile vehicle according to the present invention.

The present invention proposes a method of greatly improving the accuracy of position estimation by using the position of an autonomous moving vehicle (robot) using characteristic information of a surrounding feature type with a speedometer. That is, according to the present invention, an autonomous moving vehicle performs navigation based on a navigation sensor such as a GPS, an INS (Inertial Measurement Unit) / INS, a speedometer, etc., and corrects the navigation error using the position information of the periodical or non- It is possible to more accurately estimate the autonomous moving vehicle position estimation even in an environment in which the GPS is not available for a long time.

For example, in the present invention, the absolute position of a stop line / signal lamp is stored in advance in an autonomous traveling component of an autonomous moving vehicle in a road environment where the autonomous vehicle travels, / Traffic lights, and uses the stored absolute position as an assistant navigational method in integrated navigation, so that the position of the autonomous moving vehicle is estimated by fusing with the vehicle speed system even in an environment where GPS is not available for a long time.

1 is a configuration diagram of a position estimation system for an autonomous mobile vehicle according to the present invention.

As shown in FIG. 1, the system for estimating an autonomous moving vehicle according to the present invention includes a traveling information apparatus 100 (position estimating apparatus) that is composed of multiple sensors and estimates traveling information of an autonomous moving vehicle, A feature storage device 120 (feature database) for storing information (feature and absolute position) of the recognized feature, a feature recognition device 110 for recognizing the feature and the absolute position, An integrated position error determination unit 130 for estimating a position error of the image and the lidar based on the outputted driving information and comparing the positional errors based on the feature information to determine an integrated position error, And an integrated navigation unit 140 for correcting the INS error using the integrated position error output from the error determination unit 130.

The traveling information device 100 includes a navigation sensor such as a GPS, an Inertial Measurement Unit (IMU), a Compass, a speedometer (or a speedometer), an image sensor and a RIDAR sensor.

The integrated position error determination unit 130 determines whether or not the integrated position error determination unit 130 determines that the integrated position error determination unit 130 determines the integrated position error based on the current position of the autonomous vehicle based on the navigation sensor, To determine an integration error.

The integrated navigation unit 140 corrects the integrated navigation position using the traveling information (the speed system / IMU) of the traveling information device 100 by using the integrated position error output from the integrated position error determination unit 130. [

In another embodiment, the integrated position error determination unit 130 may be integrated with the integrated navigation unit 140.

The operation of the autonomous mobile vehicle position estimation system according to the embodiment of the present invention will now be described in detail.

FIG. 2 is a conceptual diagram for recognizing a feature item in a feature recognition apparatus using a sensor of an autonomous moving vehicle, and FIG. 3 is a flowchart illustrating an operation for recognizing and storing feature information and position information during a test run.

As shown in FIGS. 2 and 3, the operation environment of the autonomous moving vehicle may include a feature such as a traffic light / stop line. Therefore, according to the present invention, preliminary driving (test driving) is performed in a road environment in which an autonomous vehicle travels, and the feature information and the absolute position of features existing in the road environment are measured and stored as an autonomous running component (DB) May be stored in the device 120 in advance.
That is, if the feature is identified during the test run (S100), the feature information of the feature (s) is extracted using the recognition sensor of the feature recognition apparatus 110 (S110) The absolute position is extracted (S120). Accordingly, the autonomous mobile vehicle stores the feature information and the absolute position of the feature extracted by the feature recognizing device 120 in the autonomous running component DB, that is, the feature storage device 120 (S130).

4 is a flowchart for searching for location information of the feature stored in the storage device.

In order to increase the matching speed between the feature information of the feature information input from the sensor and the feature information of the feature information stored in the storage device 120 during the actual running of the autonomous mobile vehicle, And determines the absolute position of the feature in the limited search area and transmits the result to the integrated navigation part 140 periodically or non-periodically.

To do this, if the feature is detected during the movement of the autonomous moving vehicle, the feature recognition apparatus 110 extracts the feature of the feature through the recognition sensor (S200). In order to increase the matching speed between the feature information of the extracted feature and the feature information of the features stored in the storage device 120, the integrated position error determination unit 130 determines whether or not the self- A surrounding search area is determined based on the current position of the moving vehicle (S210). Then, an absolute position corresponding to the feature of the extracted feature in the determined search area is searched from the feature storage device 120, and an error between the absolute position of the searched feature and the integrated navigation position measured through the navigation sensor is calculated as a tolerance (S220 to S240). If the calculated difference is larger than the tolerance, the absolute position of the detected feature is determined as the integrated position error. If the calculated difference is smaller than the tolerance, the integrated navigation position is determined as the integrated position error and transmitted to the integrated navigation unit 140 (S250, S260).
Accordingly, the integrated navigation unit 260 corrects the navigation position using the traveling information (the speed system / IMU) of the traveling information device 100 using the integrated position error output from the integrated position error determination unit 130. [

delete

5 is a conceptual diagram of an integrated navigation position correction using position information of a feature point.

Referring to FIG. 5, the integrated navigation position error of the position data using the navigation sensor (IMU / speedometer) of the traveling information device 100 in an environment where GPS is not available for a long time increases as the time increases . Thus, the present invention improves the position accuracy of the autonomous moving vehicle by reducing the integrated navigation position error using the feature data. In an environment where GPS is available, the integrated navigation position error in an environment where the integrated navigation position error is within the GPS error range but GPS is not available deviates greatly from the GPS error even if IMU or speedometer etc. is used in the driving information device 100 Of course, the degree of error depends on the level of IMU used).

Therefore, in an environment where GPS is not available, the integrated navigation method corrects the position error using the feature-based position data as shown in the following equation (1).

[Equation 1]

IF P _integration- P _topography > ε,

P _integration = P _topography

Else

P _Integration = P _Integration

Where P is _integrated difference sokgye / IMU integrated navigation such as a location utilizing, P is the _terrain features based on location, ε is the tolerance.

That is, if the difference between the integrated navigation position and the feature-based error is greater than the tolerance, then the feature-based position is used. If the difference is less than the tolerance, the integrated navigation position is used.

FIG. 6 is a conceptual diagram of an integrated position estimating apparatus using multiple information, and FIG. 7 shows a concept of performing position error correction.

Aligns the initial position of the autonomous moving vehicle after the navigation alignment as shown in Fig. The initial position of the autonomous moving vehicle is determined by checking whether the GPS is normal, and the initial position of the autonomous moving vehicle is set by using the GPS or the feature based location. When the initial position is set, the navigation device corrects the INS error using the GPS, the feature storage device, the image, and the lidar.

As shown in FIG. 7, when performing the navigation for correcting the INS error, the integrated navigation is performed depending on the navigation sensor such as the GPS in the normal GPS environment and the image navigation (matching by the matching between the camera input images) , The INS error is corrected using the feature-based navigation as described above in addition to the terrain navigation (location estimation through matching between the Lada input data and the aerial map). At this time, the integrated position error determination unit 130 can apply each navigation according to the stored ratio.

As described above, the present invention provides an autonomous moving vehicle position estimation system using feature location information, thereby greatly reducing the dependence on the GPS in the autonomous traveling of the autonomous moving vehicle, thereby improving the reliability of autonomous traveling, It is possible to more accurately estimate the autonomous moving vehicle's position estimation even in an all-weather environment by utilizing multiple information including the device.

It will be appreciated that the configurations and methods of the embodiments described above are not to be limited and that the embodiments may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Therefore, it should be understood that the above-described embodiments are to be considered in all respects as illustrative and not restrictive.

100: travel information device 110: feature recognition device
120: Feature storage device 130: Integrated position error determination unit
140: Integrated Navigation Department

Claims (8)

A traveling information device for measuring an integrated navigation position of the autonomous moving vehicle through a navigation sensor;
A feature storage device for storing features and absolute positions of features recognized in the road environment during test driving;
A feature recognition device that recognizes features of a feature in a road environment during actual driving;
A first position error through matching between camera input images in the travel information device in a GPS inaccessible environment, a second position error through matching between the LADA input data and the air map, and a second position error through the measured integrated navigation position and the recognized feature An integrated position error determination unit for determining an integrated position error using a third position error between pre-stored absolute positions corresponding to the integrated position error; And
And an integrated navigation unit for correcting the integrated navigation position of the autonomous moving vehicle using the determined integrated position error,
Wherein the integrated position error determination unit determines the integrated position error by applying the first to third position errors according to the stored ratio. The navigation system as claimed in claim 1, wherein the navigation sensor
An IMU, a compass, a speedometer, a camera, and a lidar. delete 2. The apparatus of claim 1, wherein the integrated position error determination unit
Calculating a difference between the absolute position of the detected feature and the integrated navigation position measured through the navigation sensor and determining the absolute position of the detected feature as the integrated position error if the calculated position error is larger than the allowable error, And the navigation position is determined as an integrated position error. A method for estimating a position of an autonomous mobile vehicle that corrects a navigation error of a GPS Inertial Navigation System (INS) while performing a navigation using a plurality of navigation sensors,
Storing feature information and an absolute position of the feature measured in the test run;
Estimating a first position error through matching between camera input images if the GPS can not be used in actual driving, and estimating a second position error through matching between the Lada input data and the aerial map;
Calculating a third position error between the integrated navigation position of the autonomous moving vehicle measured by the navigation sensor and the stored absolute position corresponding to the surrounding feature;
Determining an integrated position error by applying the first to third position errors according to a pre-stored ratio;
And correcting an integrated navigation position of the autonomous moving vehicle by using the determined integrated position error. 6. The method of claim 5, wherein the plurality of navigation sensors
An IMU, a camera, a speedometer, a camera, and a lidar. 6. The method of claim 5, wherein the determined integrated position error is
Wherein the information is transmitted periodically or non-periodically to an integrated navigation unit. 6. The method of claim 5, wherein determining the integrated position error comprises:
Calculating a difference between the absolute position of the detected feature and the integrated navigation position measured using the navigation sensor;
Comparing the calculated position difference with a tolerance; And
Determining an absolute position of the detected feature as an integrated position error if the calculated position error is greater than a tolerance and determining the integrated navigation position as an integrated position error if the calculated position error is smaller than the tolerance; Position estimation method.

Images