KR101164336B1 - Base plane height compensation method of mobile robot and generating method for elevation map using the same - Google Patents

Abstract

The present invention relates to a reference plane height correction method of a mobile robot having a distance sensor scanning a front bottom surface in a tilted state at a predetermined gaze angle and an altitude map preparation method using the same. The reference plane height correction method according to the present invention comprises the steps of: (a) obtaining a plurality of measured height values for the front bottom surface of the mobile robot according to the scan operation of the distance sensor; (b) extracting a plurality of effective height values from among the plurality of measured height values based on a reference height value currently registered with respect to the reference plane; (c) calculating a correction height value by applying the currently registered reference height value and the effective height value to a preset estimation algorithm; (d) updating the currently registered reference height value with the correction height value. Accordingly, by using the information from one distance sensor scanning the front bottom surface of the mobile robot in a tilted state at a constant gaze angle, the reference height value with respect to the reference plane, which is a reference for height measurement, is continuously maintained as the mobile robot travels. By being corrected and updated, it is possible to recognize the surrounding environment more accurately.

Description

BASE PLANE HEIGHT COMPENSATION METHOD OF MOBILE ROBOT AND GENERATING METHOD FOR ELEVATION MAP USING THE SAME}

The present invention relates to a reference plane height correction method of a mobile robot and a method for preparing an altitude map using the same, and more specifically, an obstacle using scan data from a distance sensor that scans a front bottom surface in a state inclined at a predetermined gaze angle. The present invention relates to a method of correcting a height of a reference plane of a mobile robot capable of obtaining more accurate driving information by correcting a height of a base plane, which is a criterion of height, and an altitude map using the same.

Robot technology has been developed along with the development of industrial robots to replace manual labor by manpower in the manufacturing industry, but recently, robots for providing various services such as guidance, serving, escort, cleaning, indoor patrol, etc. Research is ongoing.

As such, the robot for providing services such as guidance, serving, and escort is provided in the form of an autonomous mobile robot (hereinafter, referred to as a 'mobile robot'), which is more intelligent than an existing industrial robot. However, even in an unknown environment, it is required to be able to cope with the environment on its own without prior knowledge.

However, currently applied mobile robots have a problem of lack of cognitive ability and reasoning ability. Thus, many fields have been studied to increase the intelligence of mobile robots. As such, among the functions required for the mobile robot, the basic function is that the mobile robot should be able to move to a desired target point without a collision or fall situation.

In other words, in order to move freely in a living space, the mobile robot must be able to determine where it is located, and accurate recognition of the surrounding environment causing an unexpected situation such as an obstacle or a fall is required. Such functions can be performed by localization and mapping technology of mobile robots.

In general, in order for a robot to map a surrounding environment, a robot must basically include a sensor for recognizing the surrounding environment. Sensors mainly used in mobile robots include ultrasonic sensors, vision sensors, laser sensors, infrared sensors, and the like, and distance sensors such as laser sensors that generate distance information with surrounding environments due to relatively low price are widely used.

Mapping techniques for the surrounding environment can be largely divided into shape-based and grid-based. Shape-based mapping is a method of expressing the surrounding environment in a specific shape such as a line, point, or arc, and grid-based mapping method divides the robot's surroundings into small grids. It is also called stochastic lattice map as a method of probabilistically expressing the probability that an object exists in each lattice. Here, unlike the shape-based map, the grid-based map is easy to update and modify the map and has the advantage of efficiently expressing the presence or absence of the object regardless of the shape of the object. Therefore, various studies have been conducted to improve the grid-based mapping method.

As described above, the mobile robot accurately recognizes the surrounding environment, and in preparing the environment map, the mobile robot must accurately recognize the information from the sensor installed in the mobile robot. For example, if the mobile robot incorrectly recognizes the information from the sensor, it may misinterpret the height of the obstacle and make an error of crashing or other unnecessary route, and the wrong information is included in the environment map. Cause problems.

In order to solve such a problem, a technique for minimizing an error of misrecognizing information from a sensor by attaching a plurality of sensors of various types to a mobile robot has been proposed, but manufacturing cost due to the attachment of a plurality of sensors Due to the increase of the real robot is difficult to apply to the problem.

The present invention has been made to solve the above problems, it is possible to more accurately recognize the surrounding environment by using information from one distance sensor that scans the front bottom surface of the mobile robot in an inclined state at a constant gaze angle. It is an object of the present invention to provide a method for compensating the reference plane height of a mobile robot.

In addition, an object of the present invention is to provide an altitude mapping method capable of creating a more accurate 2.5D altitude map for a surrounding environment using information from a distance sensor.

According to the present invention, in the reference plane height correction method of a mobile robot having a distance sensor for scanning the front bottom surface in a state inclined at a predetermined gaze angle, (a) the movement in accordance with the scan operation of the distance sensor Obtaining a plurality of measured height values for the front bottom surface of the robot; (b) extracting a plurality of effective height values from among the plurality of measured height values based on a reference height value currently registered with respect to the reference plane; (c) calculating a correction height value by applying the currently registered reference height value and the effective height value to a preset estimation algorithm; and (d) updating the currently registered reference height value with the correction height value.

Here, a Kalman filter algorithm may be applied as the estimation algorithm.

The currently registered reference height value may be registered by performing steps (b) to (d) with respect to a plurality of previous measurement height values obtained according to a previous scan operation of the distance sensor.

Further, in the step (b), the measured height value included in the ± effective value range of the reference height value among the plurality of measured height values is extracted as the effective height value; The effective value may be determined based on a previous cumulative dispersion value calculated based on a previous effective height value extracted by performing step (b) with respect to the previous measured height value.

Here, the step (c) may include: (c1) calculating a deviation value between the currently registered reference height value and the respective effective height values; (c2) calculating a current dispersion value of the calculated deviation values; (c3) calculating a current cumulative dispersion value based on the current dispersion value and the previous cumulative dispersion value with respect to the previous measured height value; (c4) calculating the correction height value based on the current cumulative dispersion value, the currently registered reference height value, the previous cumulative dispersion value, the current average value of the effective height values, and the current dispersion value. have.

And, the current cumulative variance value is

Where φ _t is the current cumulative variance value, φ _{0: t-1} is the previous cumulative variance value, and φ is the current variance value; The correction height value is

Where h _t is the correction height value, h _{0: t-1} is the currently registered reference height value, and z _t is an average value of the effective height value.

On the other hand, the above object is, according to another embodiment of the present invention, in the altitude map generation method of the mobile robot having a distance sensor for scanning the front bottom surface in a state inclined at a predetermined gaze angle, (A) of the distance sensor Performing a reference plane height correction method according to any one of claims 1 to 6 in units of scan operations to set an updated reference height value according to the reference plane height correction method as a height value of the reference plane; (B) correcting the measured height value obtained in the scan operation unit of the distance sensor based on the height value of the reference plane set in the scan operation unit of the distance sensor; (C) generating an altitude map around the mobile robot based on the corrected measured height value.

Here, the method may further include coordinate conversion of map information constituting the altitude map based on a change in pose of the mobile robot.

According to the present invention, according to the present invention, the reference height value with respect to the reference plane as a reference for the height measurement is moved by using information from one distance sensor scanning the front bottom surface of the mobile robot in an inclined state at a constant gaze angle The reference plane height correction method of a mobile robot that can recognize the surrounding environment more accurately by continuously correcting and updating according to the driving of the robot is provided.

According to the present invention, there is provided an altitude mapping method capable of creating a more accurate 2.5D altitude map for the surrounding environment based on information from a reference distance sensor.

1 is a view conceptually illustrating a driving state of a mobile robot according to the present invention;
2 is a view showing the configuration of a mobile robot according to the present invention;
3 is a view for explaining a reference plane height correction method of a mobile robot according to the present invention,
4 is a view showing a change in the reference height value according to the reference plane height correction method of the mobile robot according to the present invention,
5 is a view showing an example of the altitude recognition result according to whether or not the reference plane height correction method according to the present invention,
6 is a view for explaining a method for generating an altitude map of a mobile robot according to the present invention.

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

1 is a view conceptually showing a driving state of the mobile robot 100 according to the present invention, and FIG. 2 is a diagram illustrating a configuration of the mobile robot 100 according to the present invention. 1 and 2, the mobile robot 100 according to the present invention includes a distance sensor 10, a scan data generation unit 11, a measurement height value calculation unit 12, and a reference plane height correction unit 13. ), And the main control unit 15. In addition, the mobile robot 100 according to the present invention may include a map generator 16, a robot driver 17, and a data storage 14.

As shown in FIG. 1, the distance sensor 10 is installed in the mobile robot 100 to be positioned at a predetermined installation height. Here, the distance sensor 10 according to the present invention scans the front bottom surface of the mobile robot 100 in an inclined state at a predetermined gaze angle. In the present invention, the laser distance sensor 10 is applied to the distance sensor 10 as an example. The laser distance sensor 10 having a resolution of 1 ° is applied to scan data in units of 1 ° from 0 ° to 180 °. Take acquisition as an example.

The scan data generation unit 11 generates scan data based on the data acquired according to the scan operation of the laser distance sensor 10. Then, the measurement height value calculator 12 calculates the measurement height value for each scan data by using the scan data generated by the scan data generator 11.

Here, the scan data has information about the distance between the laser distance sensor 10, the bottom surface or obstacles and the laser distance sensor 10 is reflected, the installation height of the laser distance sensor 10 and the laser distance According to the information on the gaze angle of the sensor 10, it is possible to calculate the information on the measured height value for the floor or obstacle.

The reference plane height correction unit 13 corrects the reference height value with respect to the height of the reference plane using the measurement height value calculated by the measurement height value calculation unit 12. Here, the reference surface becomes a virtual surface which becomes a reference when the mobile robot 100 measures the height of the surrounding environment of the mobile robot 100 such as an obstacle or a bottom surface.

The reference plane height corrector 13 according to the present invention may include an effective height value extractor 13a and a correction height value calculator 13b, as shown in FIG. 2. Here, the effective height value extraction unit 13a extracts a plurality of effective height values from the measurement height values calculated by the measurement height value calculation unit 12. And the correction height value calculation part 13b correct | amends a reference height value using the some effective height value extracted by the effective height value extraction part 13a.

Hereinafter, a method of correcting the reference height value by the reference plane height corrector 13 according to the present invention will be described in detail with reference to FIG. 3.

First, upon initial driving of the mobile robot 100, an initial value is set in the mobile robot 100 (S30). In the reference plane height correction method according to the present invention, a value set as an initial value includes a reference height value and a cumulative dispersion value. Here, the reference height value is initially set to '0', and the cumulative dispersion value is initialized to an arbitrary value.

In the state where the initial value is set as described above, the front bottom surface of the mobile robot 100 is scanned according to the scan operation of the laser distance sensor 10, and as described above, the scan data generation unit 11 and the measured height value are calculated. A plurality of measurement height values are obtained through the unit 12 (S31).

When the measured height value is obtained as described above, the effective height value extracting unit 13a extracts a plurality of effective height values among the plurality of measured height values based on the reference height values currently registered with respect to the reference plane (S32). Here, the effective height value is a value that can be used to correct the reference height value. For example, the measured height value corresponding to the obstacle is not extracted as the effective height value.

In the present invention, an example in which the measured height value included in the ± effective value range of the reference height value among the plurality of measured height values is extracted as the effective height value. This may be expressed as in [Equation 1].

[Equation 1]

In Equation 1, h _t-1 is a reference height value corrected in the t-1 th scan operation of the laser distance sensor 10, and z _{t, i} is a value of the i th scan data extracted in the t th scan operation. Measured height value. 3σ is an effective value, which will be described later. Here, as described above, the initial state of the mobile robot 100 is that h _t-1 is initialized to '0' as described above.

When the extraction of the effective height value satisfying [Equation 1] is completed, the correction height value calculation unit 13b applies the extracted effective height value and the currently registered reference height value to a preset estimation algorithm to correct the height value. To calculate (S35). In the present invention, the Kalman filter algorithm is applied as an estimation algorithm.

More specifically, first, a deviation value between the currently registered reference height value and each effective height value is calculated (S33). This may be expressed as in [Equation 2].

[Equation 2]

는 편차값이고, z_t는 [수학식 1]을 통해 추출된 유효 높이값이다. 그리고, h_t-1은 레이저 거리 센서(10)의 t-1 번째 스캔 동작에서 보정된 기준 높이값, 즉, 레이저 거리 센서(10)의 이전 스캔 동작에 따라 보정되어 현재 등록되어 있는 기준 높이값이다.In [Equation 2]

Is the deviation value, and z _t is the effective height value extracted through [Equation 1]. And, h _t-1 is the reference height value corrected in the t-1 th scan operation of the laser distance sensor 10, that is, the reference height value currently corrected and registered according to the previous scan operation of the laser distance sensor 10. to be.

Then, the average value and the variance value of the deviation values calculated through Equation 2 are calculated (S33). The cumulative dispersion value of the current scan operation, that is, the t-th scan operation is calculated based on the calculated dispersion value and the cumulative dispersion value calculated in the previous scan operation, that is, the t-1 th scan operation (S34). Here, the cumulative dispersion value may be calculated through [Equation 3].

&Quot; (3) "

In Equation 3, φ _t is the current cumulative dispersion value, the cumulative dispersion value at the t th scan operation, and φ _{0: t-1} is the previous cumulative dispersion value, that is, the cumulative dispersion value at the t-1 th scan operation. , φ is the variance of the deviation values calculated through [Equation 2]. Here, as described above, φ ₀ is initialized to an arbitrary value during initial driving of the mobile robot 100.

And phi _{0: t-1} is applied to determination of the above-mentioned effective value. That is, in the present invention, it is determined based on the previous cumulative dispersion value calculated based on the previous scan operation, that is, the effective height value extracted in the t-1 th scan operation. For example, using a standard deviation corresponding to the variance value.

&Quot; (4) "

That is, in the present invention, the reference height value registered in the t-1 th scan operation by using the cumulative dispersion value calculated for the effective measured value in the t th scan operation for the effective measured value in the t-1 th scan operation Will be reflected in the extraction of effective measurements.

Meanwhile, when the cumulative dispersion value is calculated through the above process, the correction height value is calculated based on the calculated cumulative dispersion value, the currently registered reference height value, the previous cumulative dispersion value, the average value of the effective height values, and the dispersion value. (S35). This may be expressed as in [Equation 5].

[Equation 5]

In Equation 5, h _t is a correction height value of the t th scan operation, and h _{0: t-1} is a reference height currently registered, that is, a correction height value of the t-1 th scan operation, as described above. And z _t is the average value of the effective height values. And, as described above, φ is the variance value for the deviation values, and φ _{0: t-1} is the previous cumulative variance value, that is, the cumulative variance value in the t-1 th scan operation.

Through the above process, when the correction height value is calculated in the t-th scan operation, the currently registered reference height value is updated with the calculated correction height value (S36). That is, in the t-th scan operation, the correction height value is registered as the reference height value. Through this process, the reference height value is continuously updated until the driving of the mobile robot 10 is terminated (S37).

4 is a diagram illustrating a change in a reference height value through five scan operations after setting '0' as an initial reference height value through the above process. As shown in Figure 4, it can be seen that the reference height value is converged to 55mm. Accordingly, the mobile robot 100 more accurately recognizes the surrounding environment such as an obstacle detected by the laser distance sensor 10.

For example, assuming that an actual obstacle of 77mm exists, in a state where the reference plane height correction method according to the present invention is not applied, the height may be recognized as 15mm and may be recognized as an obstacle that can be passed. On the other hand, when the reference plane height correction method according to the present invention is applied, since the height of the reference plane is corrected to 55mm, the height of the obstacle is more accurately recognized as 77mm.

In addition, when the mobile robot 100 rises along the inclined plane, when the front surface of the floor on which the mobile robot 100 is driven increases, when the correction for the reference plane is not performed, the mobile robot 100 recognizes it as an obstacle or a low obstacle on the inclined plane. Despite this existence, it is possible to eliminate an error that is recognized as an obstacle that cannot be overcome.

FIG. 5A is a diagram illustrating a 2.5D elevation map generated in a state in which the reference plane height correction method according to the present invention is not applied. FIG. 5A illustrates an example in which the area A is recognized as a low altitude area. 3 (b) illustrates an example in which an accurate reference plane is displayed by compensating a region having low altitude as the reference plane height correction method according to the present invention is applied.

Hereinafter, the altitude map generation process of the mobile robot 100 according to the present invention will be described in detail with reference to FIGS. 2 and 6.

First, as described above, when the reference height value is updated, the reference height value is set to the height value of the current reference plane (S60). At this time, the main control unit 15 controls the robot driving unit 17 so that the traveling of the mobile robot 100 is controlled based on the updated reference height value corrected by the reference plane height correcting unit 13 and updated. Here, the reference height value corrected by the reference plane height correction unit 13 may be stored in the data storage unit 14, and the main controller 15 may refer to the reference height value stored in the data storage unit 14 to refer to the robot. The driving unit 17 can be controlled.

The map generator 16 may measure the measured height value calculated by the measured height value calculator 12, the scan data generated by the scan data generator 11, and the reference under the control of the main controller 15. An altitude map is generated using the height value (S62). Here, the map generator 16 may include a measurement height value corrector 16a and a map converter 16b.

The measurement height value corrector 16a scans the laser distance sensor 10 based on the height value of the reference plane set as the scan operation unit of the laser distance sensor 10, that is, the reference height value updated through the above-described process. The measured height value obtained in the operation unit is corrected (S61).

Then, the altitude map around the mobile robot 100 is generated based on the corrected measured height value (S62). Here, in the case where the altitude map for the mobile robot 100 is registered in advance or previously moved, the main controller 15 may modify the altitude map with the corrected measured height value.

On the other hand, when the pose change of the mobile robot 100 occurs (S63), the map converter 16b may perform coordinate transformation of map information constituting the altitude map (S64). Here, the altitude map generated by the mobile robot 100 according to the present invention has an example of a grid map (Grid map).

Here, the map information constituting the altitude map may include h (i, j), x (i, j), y (i, j), z (i, j). Here, h (i, j) indicates whether the cell (i, j) is occupied, x (i, j) is the x coordinate value measured in the cell (i, j), and y (i, j) is (i , j) is the y-coordinate value measured in the cell, and z (i, j) is the z-coordinate value measured in the cell (i, j). And x (i, j), y (i, j), z (i, j) will have information only when the cell is occupied state, that is, h (i, j) = 1.

Equation 6 shows an example of a determinant for coordinate transformation of map information.

&Quot; (6) "

In Equation 6, Δθ _r is the angle change amount of the pose of the mobile robot 100, Δx _r is the x change amount of the pose of the mobile robot 100, and Δy _r is the y change amount of the pose of the mobile robot 100.

Through the above process, it is possible to create a more accurate altitude map through the correction of the reference height value, the coordinate transformation of the altitude map in accordance with the change of the pose of the mobile robot 100 is performed to the center of the mobile robot 100 Altitude mapping is possible.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, Of the right.

100: mobile robot 10: laser distance sensor
11: scan data generation unit 12: measurement height value calculation unit
13 reference plane height correction unit 14 data storage unit
15: main control unit 16: map generation unit
14: robot drive unit

Claims (8)

In the reference plane height correction method of a mobile robot having a distance sensor for scanning the front bottom surface while being inclined at a predetermined gaze angle,
(a) acquiring a plurality of measurement height values of the front bottom surface of the mobile robot according to a scan operation of the distance sensor;
(b) extracting a plurality of effective height values from among the plurality of measured height values based on a reference height value currently registered with respect to the reference plane;
(c) calculating a correction height value by applying the currently registered reference height value and the effective height value to a preset estimation algorithm;
(d) updating the currently registered reference height value with the correction height value,
The currently registered reference height value is registered by performing steps (b) to (d) with respect to a plurality of previous measurement height values obtained according to a previous scan operation of the distance sensor,
In the step (b), the measured height value included in the ± effective value range of the reference height value among the plurality of measured height values is extracted as the effective height value, and the effective value is determined by the ( b) the reference plane height correction method of the mobile robot, characterized in that it is determined based on the previous cumulative dispersion value calculated based on the previous effective height value extracted by performing the step. The method of claim 1,
A method of correcting the reference plane height of a mobile robot, characterized in that a Kalman filter algorithm is applied as the estimation algorithm. delete delete The method of claim 2,
In step (c),
(c1) calculating a deviation value between the currently registered reference height value and the respective effective height values;
(c2) calculating a current dispersion value of the calculated deviation values;
(c3) calculating a current cumulative dispersion value based on the current dispersion value and the previous cumulative dispersion value with respect to the previous measured height value;
(c4) calculating the correction height value based on the current cumulative dispersion value, the currently registered reference height value, the previous cumulative dispersion value, the current average value of the valid height values, and the current dispersion value. A reference plane height correction method of a mobile robot, characterized in that. The method of claim 5,
The current cumulative variance value is

Where φ _t is the current cumulative variance value, φ _{0: t-1} is the previous cumulative variance value, and φ is the current variance value;
The correction height value is

(Where h _t is the correction height value, h _{0: t-1} is the currently registered reference height value, and z _t is an average value of the effective height value) of the mobile robot. Reference plane height correction method. In the altitude map generation method of a mobile robot having a distance sensor for scanning the front floor surface inclined at a predetermined gaze angle,
(A) The reference height updated according to the reference plane height correction method by performing the reference plane height correction method according to any one of claims 1, 2, 5, and 6 as a unit of the scan operation of the distance sensor. Setting a value to a height value of the reference plane;
(B) correcting the measured height value obtained in the scan operation unit of the distance sensor based on the height value of the reference plane set in the scan operation unit of the distance sensor;
(C) generating an altitude map around the mobile robot based on the corrected measured height value. The method of claim 7, wherein
(D) a method of generating an altitude map of a mobile robot, the method further comprising coordinate transformation of map information constituting the altitude map based on a change in pose of the mobile robot.

Images