KR20240096070A - SLAM NAVIGATON SYSTEM for Autonomous Forklift Truck based on ROS - Google Patents

Abstract

The SLAM navigation system for a ROS-based unmanned forklift is introduced. To this end, the present invention includes a LIDAR measurement sensor installed on a forklift, an IMU sensor 20 that senses the position of the forklift, and a device that receives information sensed by the LIDAR measurement sensor 10 and the IMU sensor 20. It includes a control unit 100, wherein the control unit 100 includes a driving record unit 200 that can sense the relative position of the forklift when it moves, and a SLAM unit that senses the movement of the forklift in real time and displays a map. It further includes 300, wherein the slam unit 300 uses a Gmapping program, and implements a map using the lidar measurement sensor 10, the IMU sensor 20, and the odometer. It is characterized by obtaining distance and position values for the map, and RViz is used when creating the map, and the created map is saved as a yaml file.

Description

SLAM NAVIGATON SYSTEM for Autonomous Forklift Truck based on ROS}

The present invention relates to a slam navigation system for a ROS-based unmanned forklift.

SLAM (Simultaneous Localization and Mapping) refers to a technique used especially in self-driving vehicles to create a map of the surrounding environment and at the same time recognize the current location of the vehicle within the created environment map.

At this time, self-driving mobile objects include vehicles, robots, electronic devices, etc. that can move on their own through autonomous driving.

The most important technology in SLAM-based mobile robot autonomous navigation is robot localization technology that determines the robot's current location.

Meanwhile, in terms of logistics efficiency, market demand for unmanned forklift systems is gradually increasing, and many related studies are being conducted. Most of the studies implement PC-based autonomous driving algorithms using sensors, etc.

However, since the self-driving algorithm is implemented based on Intel PC, it is difficult to enter the market due to its high price, so a low-cost control solution is urgently needed.

Meanwhile, when designing a controller for an unmanned forklift system, an Autonomous Forklift Controller (AFC) must be additionally installed separately from the existing Electronic Control Unit (ECU).

Because of this, there is a problem in that the design of a system for unmanned forklifts becomes complicated.

Accordingly, the present invention realizes cost reduction of unmanned forklifts through the control logic implemented in the present invention of SLAM Navigation and various HW, which are already commercialized and used by overseas advanced companies, and provides optimal autonomous driving technology even without forklift experience. It has a purpose.

Korean Patent Publication No. 10-2010-0132269 (2012.07.02.)

The purpose of the present invention is to provide a slam navigation system for a ROS-based unmanned forklift that combines LiDAR sensor technology capable of 360-degree 2D scanning that is ultra-small/lightweight/low-cost and capable of autonomous driving through AMR-based slam navigation technology.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned can be clearly understood by those skilled in the art from the description below.

The SLAM navigation system for a ROS-based unmanned forklift is introduced.

For this purpose, the present invention includes a LIDAR measurement sensor 10 installed on a forklift, an IMU sensor 20 that senses the position of the forklift, and information sensed by the LIDAR measurement sensor 10 and the IMU sensor 20. It includes a control unit 100 that receives a, wherein the control unit 100 includes a driving record unit 200 that can sense the relative position of the forklift when it moves, and a slam (200) that senses the movement of the forklift in real time and displays a map. It further includes a SLAM unit 300, wherein the SLAM unit 300 uses a Gmapping program, and the LiDAR measurement sensor 10, the IMU sensor 20, and the driving record unit 200 It is characterized by obtaining distance values and location values to implement a map using , and when creating a map, RViz is used, and the created map is saved as a yaml file.

According to the present invention consisting of the above configuration, various effects as follows are realized.

First, there is the advantage of using LIDAR to provide an autonomous driving algorithm that takes into account the special functions of ROS-based unmanned forklifts.

Second, it has the advantage of being directly commanded using Command-Line Tools without a GUI.

Third, there is the advantage of 3D visualization using Rviz.

Fourth, by using RQT, various effects are realized, such as measuring the position of the forklift by increasing the degree of freedom in routing and learning, increasing precision, and dramatically reducing the time to find the optimal path.

The effects according to the technical idea of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below.

Figure 1 is a diagram illustrating the overall configuration of a slam navigation system for a ROS-based unmanned forklift according to some embodiments of the present invention.
Figure 2 is a flowchart of the operation logic of the navigation unit, which is a component of the present invention.
Figure 3 is a diagram of an example Rviz implementation.
Figure 4 is a diagram of an example RTQ implementation.
Figure 5 is a conceptual diagram of the navigation unit, which is a component of the present invention.
Figure 6 is an example diagram for Gmapping-based map creation.
Figure 7 is an example diagram for Amcl-based location estimation.
Figures 8 and 9 are diagrams of the obstacle collision area and collision condition configuration.
Figure 10 is a diagram of the development technology of an unmanned forklift for smart logistics to which the present invention can be applied.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. The advantages and features of the present invention and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms. The present embodiments are merely intended to ensure that the disclosure of the present invention is complete, and that the present invention is not limited to the embodiments disclosed below and is provided by those skilled in the art It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used with meanings that can be commonly understood by those skilled in the art to which the present invention pertains. Additionally, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless clearly specifically defined.

Additionally, in this specification, the singular form may also include the plural form unless specifically stated in the phrase. As used herein, “comprises” and/or “comprising” refers to the presence of one or more other components, steps, operations and/or elements. or does not rule out addition.

First, prior to a detailed description of the present invention, the purpose of the present invention is as follows.

The purpose is to provide ROS-based SLAM Navigation technology for autonomous driving of forklifts. To this end, provision of ROS-based driving algorithm using 2D LiDAR, obstacle recognition and collision prevention algorithm using safety scanner, LiDAR sensor and Odometry The purpose is to provide a signal processing algorithm, a map matching-based location recognition algorithm, and finally a slam navigation system for an AMR-capable ROS-based unmanned forklift.

In order to achieve the above object, a detailed description of the present invention is as follows.

Figure 1 is an overall configuration diagram of the present invention, Figure 2 is a flow chart of the operation logic of the navigation unit 400, which is a component of the present invention, Figure 3 is a diagram of an example of Rviz implementation, and Figure 4 is a diagram of an example of RTQ implementation. Figure 5 is a conceptual diagram of the navigation unit, which is a component of the present invention, Figure 6 is an example diagram of Gmapping-based map creation, Figure 7 is an example diagram of Amcl-based location estimation, Figures 8 to 8 9 is a diagram of the obstacle collision area and collision condition configuration, and FIG. 10 is a diagram of the technology for developing an unmanned forklift for smart logistics to which the present invention can be applied.

As shown in the drawing, the present invention is a LIDAR measurement sensor 10 installed on a forklift, an IMU sensor 20 that senses the position of the forklift, and a LIDAR measurement sensor 10 and an IMU sensor 20. It includes a control unit 100 that receives the sensed information, and the control unit 100 includes a driving record unit 200 that can sense the relative position of the forklift when it moves, and a map that senses the movement of the forklift in real time and displays a map. It further includes a SLAM unit 300, and the SLAM unit 300 is characterized by using a Gmapping program, and uses a lidar measurement sensor 10, an IMU sensor 20, and a driving record unit 200. It is characterized by obtaining distance values and location values to implement the map, and RViz is used when creating the map, and the created map is saved as a yaml file.

At this time, the control unit 100 further includes a navigation unit 400 that implements the shortest path for the forklift to the target point of the forklift on the map implemented by the slam unit 300, and the navigation unit 400 uses an odometer reading. A step of measuring the distance to the target point while information is updated through (S100), a step of estimating the current location of the forklift on the map implemented through the slam unit 300 (S200), and the distance from the current location of the forklift to the target point. It consists of a step of generating a movement trajectory (S300), and a step of moving to the target point while avoiding obstacles while moving along the movement trajectory to the target point (S400).

ROS used in the present invention is specifically implemented with the control logic of 1) Command-Line Tools, 2) Rviz, and 3) RQT.

In other words, the Command-Line Tools operated in the control unit 100 of the present invention can control the operation of the forklift and use ROS functions only with commands provided by ROS without the conventional GUI.

At this time, as shown in Figure 3, it is characterized by 3D visualization through Rviz, and the appearance and planned operation of the forklift are expressed after visualizing sensor data such as LiDAR and cameras.

In addition, RQT, one of the control logics implemented in the control unit 100 of the present invention, provides a Qt-based framework for GUI development as shown in Figure 4, and using this tool, users can freely specify routes and learn. Its characteristic is that it can be used by specifying a high level.

Meanwhile, the map creation in the slam unit 300, which is a component of the present invention, matches the distance data output from the lidar measurement sensor 10 mounted at a point at the top of the forklift, and the driving record unit 200 It is characterized by setting the driving record information of the forklift recorded in and the data sensed by a separate distance sensor installed in the forklift as input values of the Gmapping program.

Additionally, SLAM Navigation is a function that measures and estimates the location of a forklift. It uses the Runge-Kutta formula to check the approximate x, y and rotation angle θ of the moved location, and determines the optimal route to the destination based on a map containing road and obstacle information. Implements the function to calculate and drive.

In addition, the location of the forklift is estimated using the Amcl node of ROS, and is characterized by setting information from the driving recorder 200 and data sensed from a distance sensor separately installed on the forklift as input values of Amcl.

More specifically, as shown in Figure 5, a map is created using the LiDAR measurement sensor of the forklift, an IMU sensor 20 is used for Odometry, and data from the LiDAR sensor and Odometry are transmitted to the PC through WiFi. It is delivered to the gmapping (OpenSLAM package) node, and the gmapping node creates a map through internal algorithm calculation, which can be checked in real time through Rviz, and the created map data is saved as a map.yalm file through the map_server node. Its characteristic is that it is implemented.

Meanwhile, as shown in Figures 6 and 7, the map creation is carried out using the Gmapping node of ROS, and the distance data output from the LiDAR mounted on the top of the forklift is matched and reconstructed into distance data centered on the robot coordinate system, and then the forklift itself Odometry information and distance sensor data are set to be used as input for Gmapping, and the forklift is moved through joystick operation to generate map data for the environment and save it as a file.

At this time, location estimation is performed using the Amcl node of ROS, and the odometry information, distance sensor data, and map data of the forklift itself are set to be used as input to Amcl, and the settings of the omnidirectional unmanned forklift are adjusted when the Amcl node is executed.

Meanwhile, as shown in FIGS. 8 and 9, the existing logic that can detect obstacles using LiDAR for obstacle recognition and collision prevention is disclosed by the present invention. The LiDAR sensor has a polar coordinate system centered on the coordinate system of the forklift. The reference distance is received and converted to a value in the Cartesian coordinate system.

In the case of obstacle avoidance, a point where a forklift and an obstacle will not collide is selected based on the location information of the obstacle recognized through a sensor, and the forklift is moved to that point. The point where a collision will not occur is determined by the radius of the forklift from the end point of the corner of the obstacle. Set it to a location further away.

At this time, when moving in a straight line, set the obstacle collision area and stop or proceed with obstacle avoidance when an obstacle exists in the collision area. When moving in the x-axis or y-axis around the coordinate system of the forklift, the obstacle collision area and collision conditions are set during straight movement. When moving on the x-axis and y-axis simultaneously, it is configured to use a mixture of the relevant conditions. During rotational movement, it is configured to set an obstacle collision area and stop or proceed with obstacle avoidance when an obstacle exists in the collision area.

Specifically, the logic to prevent collision when an obstacle occurs is as follows.

In the definition described later, Px is the sensing area coordinate value along the longitudinal direction (x-axis direction) of the forklift, Py is the sensing area coordinate value along the width direction of the forklift (y-axis direction orthogonal to the x-axis direction), l ₁ is the distance along the y-axis from the center point of the forklift to the side extremities of the forklift (including wheels), and l ₂ is the distance along the y-axis from the center point of the forklift to the front or rear end of the forklift (including wheels). The distance according to the direction, α represents the safety margin value of the corner area of the forklift, and β represents the safety margin value of the front, rear, and side areas of the forklift.

Meanwhile, α may be a fixed constant, but β is preferably a variable that varies in real time. Specifically, β can have a variable value according to the equation below, thereby further maximizing the stability of collision avoidance performance and driving space efficiency.

(Here, s represents the instantaneous speed of the forklift, FT_w represents the total weight of the forklift, FT_y represents the width of the forklift body (i.e., the width of the forklift excluding the wheels), and A represents the weight constant.)

When the forklift moves in a straight line when moving to the target point, the front area of the forklift is defined by the following equation,

Px 〉l ₂ ＆ -(l ₁ + α)〈 Py〈(l ₁ + α)

The collision conditions for forklift obstacles are defined as follows,

Px 〈 (l ₂ + β)

It is characterized by obstacle avoidance logic.

With similar logic, when the forklift moves in a straight line when moving to the target point, the rear area of the forklift is defined by the following equation,

Px 〈 - l ₂ ＆ -(l ₁ + α)〈 Py〈 (l ₁ + α)

The collision conditions for forklift obstacles are defined as follows,

Px 〉 -( l ₂ + β)

It is characterized by obstacle avoidance logic.

In addition, when the forklift moves in a straight line when moving to the target point, the left area of the forklift is defined by the following equation,

Py 〉 l ₁ ＆ -( l ₂ + α)〈 Px 〈 ( l ₂ + α)

The collision conditions for forklift obstacles are defined as follows,

Py 〈 (l ₁ + β)

It is characterized by obstacle avoidance logic.

When the forklift moves in a straight line when moving to the target point, the right area of the forklift is defined by the following equation,

Py 〉- l ₁ ＆ -( l ₂ + α)〈 Px 〈 ( l ₂ + α)

The collision conditions for forklift obstacles are defined as follows,

Py 〉- (l ₁ + β)

It is characterized by obstacle avoidance logic.

On the other hand, when the forklift rotates while moving to the target point, the rotation area value (r) of the forklift is defined by the following equation,

r= sqrt(Px × Px + Py × Py)

The collision conditions for forklift obstacles are defined as follows,

r 〈 (l + α)

It is characterized by obstacle avoidance logic.

Here, l is defined as sqrt(l ₁ × l ₁ + l ₂ × l ₂ ).

Those skilled in the art related to this embodiment will understand that the above-described substrate can be implemented in a modified form without departing from the essential characteristics. Therefore, the disclosed methods should be considered from an explanatory rather than a restrictive perspective. The scope of the present invention is indicated in the claims rather than the foregoing description, and all differences within the equivalent scope should be construed as being included in the present invention.

100: Control unit 200: Driving record book
300: Slam unit 400: Navigation unit

Claims (7)

A LIDAR measurement sensor installed on a forklift and an IMU sensor that senses the location of the forklift; A control unit 100 that receives information sensed by the lidar measurement sensor 10 and the IMU sensor 20,
The control unit 100,
A driving recorder 200 that can sense the relative position of the forklift when it moves;
It further includes a SLAM unit 300 that senses the movement of the forklift in real time and displays a map,
The slam unit 300 is characterized by using the Gmapping program, and includes a distance value for implementing a map using the lidar measurement sensor 10, the IMU sensor 20, and the driving record unit 200. A slam navigation system for a ROS-based unmanned forklift, characterized by obtaining position values, using RViz when creating a map, and saving the generated map as a yaml file.
According to claim 1,
The control unit 100 further includes a navigation unit 400 that implements the shortest path for the forklift to the target point of the forklift on the map implemented by the slam unit 300,
The navigation unit 400,
Measuring the distance to the target point while updating information through the odometer;
estimating the current location of the forklift on a map implemented through the slam unit 300;
Generating a movement trajectory from the current location of the forklift to the target point;
A slam navigation system for a ROS-based unmanned forklift, characterized in that it is implemented by moving to the target point while avoiding obstacles while moving along the movement trajectory to the target point.
According to clause 2,
Creating a map in the slam unit 300,
The distance data output from the lidar measurement sensor 10 mounted at a point at the top of the forklift is matched, and the driving record information of the forklift recorded in the driving record unit 200 is sensed by a separate distance sensor installed in the forklift. Slam navigation system for ROS-based unmanned forklift, characterized in that data is set as input value of the Gmapping program.
According to clause 2,
The location of the forklift is estimated using the Amcl node of ROS.
A slam navigation system for a ROS-based unmanned forklift, characterized in that the information in the driving record unit 200 and the data sensed from a distance sensor separately installed on the forklift are set as input values of Amcl.
According to clause 4,
If the forklift moves in a straight line when moving to the target point,
The front area of the forklift is defined by the following equation,
Px 〉l ₂ ＆ -(l ₁ + α)〈 Py〈(l ₁ + α)
The collision conditions for forklift obstacles are defined as follows,
Px 〈 (l ₂ + β)
Characterized by obstacle avoidance logic,
The rear area of the forklift is defined by the following equation,
Px 〈 - l ₂ ＆ -(l ₁ + α)〈 Py〈 (l ₁ + α)
The collision conditions for forklift obstacles are defined as follows,
Px 〉 -( l ₂ + β)
A slam navigation system for a ROS-based unmanned forklift, characterized by obstacle avoidance logic.
According to clause 4,
If the forklift moves in a straight line when moving to the target point,
The left area of the forklift is defined by the following equation,
Py 〉 l ₁ ＆ -( l ₂ + α)〈 Px 〈 ( l ₂ + α)
The collision conditions for forklift obstacles are defined as follows,
Py 〈 (l ₁ + β)
Characterized by obstacle avoidance logic,
The right area of the forklift is defined by the following equation,
Py 〉- l ₁ ＆ -( l ₂ + α)〈 Px 〈 ( l ₂ + α)
The collision conditions for forklift obstacles are defined as follows,
Py 〉- (l ₁ + β)
A slam navigation system for a ROS-based unmanned forklift, characterized by obstacle avoidance logic.
According to clause 4,
If the forklift makes a rotating movement while moving to the target point,
r= sqrt(Px × Px + Py × Py)
The collision conditions for forklift obstacles are defined as follows,
r 〈 (l + α)
A slam navigation system for a ROS-based unmanned forklift, characterized by obstacle avoidance logic.