KR20240023008A - Moving robot, method and program for outputting estimated position data in space of moving robot using position tracking model - Google Patents

Abstract

The present disclosure provides an artificial intelligence-based method for locating an unmanned vehicle in space, comprising: detecting whether an abnormal reboot of the unmanned vehicle has occurred; If an abnormal reboot occurs based on the detected result, determining whether a charging signal from the docking station has been received; If the charging signal is not received, performing a kidney recovery mode to find a location in space; Collecting image data and LIDAR data to create an inference model by driving in a preset zigzag pattern from a preset first point to a second point so that the unmanned vehicle can find its position in space; Inputting the collected image data into an image extraction model to extract spatial data within the image; and inputting spatial data and lidar data in the image into a position tracking model to output expected position data in space of the unmanned moving object; may include.

Description

Mobile robot, method and program to output expected position data in space of a mobile robot using a position tracking model {MOVING ROBOT, METHOD AND PROGRAM FOR OUTPUTTING ESTIMATED POSITION DATA IN SPACE OF MOVING ROBOT USING POSITION TRACKING MODEL}

This disclosure relates to mobile robots. More specifically, the present disclosure relates to a method and program for outputting expected position data in space of a mobile robot using a position tracking model.

In the era of the 4th Industrial Revolution, the introduction of unmanned service vehicles is increasing in various spaces such as hospitals, airports, and malls.

Conventionally, unmanned vehicles may be rebooted due to lack of power or system errors while moving autonomously.

At this time, when the unmanned vehicle was turned off and the power was turned on again, the unmanned vehicle had difficulty determining where its current location was in space.

Therefore, in recent years, research on unmanned vehicles that can quickly and accurately find their current location in space has been continuously conducted even when an abnormal reboot occurs.

Republic of Korea Patent Publication No. 10-2014-0141265 (December 10, 2014)

The purpose of the embodiments disclosed in this disclosure is to provide a way to quickly and accurately find one's current location in space.

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

A method of finding a position in space of an artificial intelligence-based unmanned mobile object according to an aspect of the present disclosure for achieving the above-described technical problem includes: detecting whether an abnormal reboot has occurred; If the abnormal reboot occurs based on the detected result, determining whether a charging signal from the docking station has been received; If the charging signal is not received, performing a kidnap recovery mode to find a location in space; Collecting image data and LIDAR data to create an inference model by driving in a preset zigzag pattern from a preset first point to a second point so that the unmanned mobile device finds its position in the space; extracting spatial data within the image by inputting the collected image data into an image extraction model; and inputting the spatial data in the image and the LIDAR data into a position tracking model to output expected position data in space of the unmanned moving object. may include.

In addition, the collection step may be characterized by rotating 360 degrees at each of the plurality of points within the first point and the second point to further collect a unit number of image data and LIDAR data for each rotation angle.

Additionally, the expected position data may include the position coordinates of the unmanned mobile device and a rotation angle at which the unmanned mobile device faces at least one of front, side, and rear.

In addition, in the output step, clustering the expected location data between clusters, selecting a cluster in which the number of expected location data is greater than a preset number among the clustered clusters, and selecting the same number of clusters among the clustered clusters. If there is, it can be characterized by selecting a cluster with low variance.

In addition, the output step may be characterized in that, if the expected position data deviates from a preset error value, the LIDAR data is input to an Iterative Closet Point (ICP) algorithm to further output the corrected expected position data. there is.

Additionally, an unmanned mobile device according to another aspect of the present disclosure includes a communication unit that communicates with a docking station; and a processor that controls an operation to find a location in space, wherein the processor detects whether an abnormal reboot occurs, and when the abnormal reboot occurs based on the detected result, the docking station is connected to the docking station through the communication unit. Determine whether the charging signal has been received, and if the charging signal is not received, perform a kidney recovery mode to find the location in the space, and find the location in the space at a preset first point. By driving in a preset zigzag pattern to a second point, image data and LiDAR data are collected to create an inference model, the collected image data is input into an image extraction model to extract spatial data within the image, and the space within the image is extracted. Data and the lidar data may be input into a location tracking model to output expected location data within the space.

In addition, the processor may rotate 360 degrees at each of a plurality of points within the first point and the second point to further collect a unit number of image data and lidar data for each rotation angle.

In addition, the processor clusters the expected location data between clusters, selects a cluster in which the number of expected location data is greater than a preset number among the clustered clusters, and selects a cluster with the same number of clusters among the clustered clusters. If so, it can be characterized by selecting clusters with low variance.

In addition, the processor may input the LIDAR data into an Iterative Closet Point (ICP) algorithm if the expected position data deviates from a preset error value, and further output corrected expected position data. .

In addition to this, a computer program stored in a computer-readable recording medium for execution to implement the present disclosure may be further provided.

In addition, a computer-readable recording medium recording a computer program for executing a method for implementing the present disclosure may be further provided.

According to the means for solving the above-described problem of the present disclosure, one can quickly and accurately find one's current location in space.

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

1 is a diagram showing the configuration of an unmanned mobile vehicle according to the present disclosure.
FIG. 2 is a diagram illustrating map information on a space stored in the memory of FIG. 1 as an example.
3 to 5 are flowcharts showing a method for finding the location of an unmanned moving object in space according to the present disclosure.
FIG. 6 is a diagram illustrating an example of a process in which an unmanned mobile device finds a position in space through the processor of FIG. 1.
FIG. 7 is a diagram showing an example of a process for storing expected position data and corrected expected position data in the memory of FIG. 1 .
FIG. 8 is a diagram illustrating an example of the process of clustering expected location data between clusters and selecting the optimal cluster using the processor of FIG. 1.

Like reference numerals refer to like elements throughout this disclosure. The present disclosure does not describe all elements of the embodiments, and general content or overlapping content between the embodiments in the technical field to which the present disclosure pertains is omitted. The term 'part, module, member, block' used in the specification may be implemented as software or hardware, and depending on the embodiment, a plurality of 'part, module, member, block' may be implemented as a single component, or It is also possible for one 'part, module, member, or block' to include multiple components.

Throughout the specification, when a part is said to be “connected” to another part, this includes not only direct connection but also indirect connection, and indirect connection includes connection through a wireless communication network. do.

Additionally, when a part "includes" a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

Throughout the specification, when a member is said to be located “on” another member, this includes not only cases where a member is in contact with another member, but also cases where another member exists between the two members.

Terms such as first and second are used to distinguish one component from another component, and the components are not limited by the above-mentioned terms.

Singular expressions include plural expressions unless the context clearly makes an exception.

The identification code for each step is used for convenience of explanation. The identification code does not explain the order of each step, and each step may be performed differently from the specified order unless a specific order is clearly stated in the context. there is.

Hereinafter, the operating principle and embodiments of the present disclosure will be described with reference to the attached drawings.

In this specification, the control unit of the unmanned mobile device according to the present disclosure includes various devices that can perform calculation processing and provide results to the user. For example, the control unit according to the present disclosure may include both a computer and a server, or may take any form.

Here, the computer may include, for example, a laptop equipped with a web browser, a desktop, a laptop, a tablet PC, a slate PC, a single board computer, etc.

A server device is a server that processes information by communicating with external devices and may include an application server, computing server, database server, file server, mail server, proxy server, and web server.

Portable terminals are, for example, wireless communication devices that ensure portability and mobility, such as PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), and PDA ( Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminal, smart phone All types of handheld wireless communication devices, such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted-device (HMD), etc. It can be included.

The unmanned vehicle according to the present disclosure detects whether an abnormal reboot of the unmanned vehicle has occurred, and if an abnormal reboot occurs based on the detected result, determines whether the charging signal from the docking station has been received, and if the charging signal is not received, , perform a kidnap recovery mode to find the position in space, and run in a preset zigzag pattern from a preset first point to a second point so that the unmanned vehicle finds the position in space, using image data and LIDAR. Collect data, input the collected video data into the image extraction model to extract spatial data within the video, and input the spatial data and lidar data within the video into the location tracking model to output expected position data in space of the unmanned moving object. can be provided.

These unmanned vehicles can quickly and accurately find their current location in space.

In order for an unmanned vehicle to move to a destination, it must know where it currently is before it can move to the destination. MCL, a LiDAR-based location estimation technique, requires the starting point on a grid map to estimate where the unmanned vehicle is when it has moved. In other words, the unmanned vehicle must know the starting point so that it can easily estimate its position in space from then on. For this reason, the unmanned vehicle can set the docking station as the starting point.

However, there may be cases where the unmanned mobile device is turned off due to lack of power in a place other than the docking station, or is rebooted due to a system error, etc. At this time, it is possible to check whether the unmanned mobile object is in front of the docking station using the docking station's charging signal. If a charging signal is received, it can be seen that the unmanned vehicle exists in front of the docking station and that the unmanned vehicle is currently at the starting point. Conversely, if there is no charging signal, the unmanned vehicle is not in front of the docking station and may determine that it is currently in a different location from the starting point.

Additionally, there may be cases where the robot is moved by someone while the unmanned vehicle is successfully estimating its location. Since the unmanned vehicle does not know that its position has changed, it has no choice but to continuously estimate its position from the position before it was moved. This problem is called the kidnapped robot problem. The reason this problem is difficult is because it is not easy to know that an unmanned vehicle has failed to recognize its own location. Therefore, a location recognition performance measure (quality measure) is needed that can determine whether location recognition of an unmanned vehicle is being successfully performed or has been kidnapped.

Below, we will look at the unmanned mobile device in detail.

1 is a diagram showing the configuration of an unmanned mobile vehicle according to the present disclosure. FIG. 2 is a diagram illustrating map information on a space stored in the memory of FIG. 1 as an example.

Referring to FIG. 1, the unmanned mobile vehicle 100 may be a mobile vehicle that recognizes the external environment, judges the situation, and performs a mission on its own without human assistance. For example, the unmanned mobile object 100 may be a mobile robot. The unmanned mobile device 100 may include a communication unit 110, a control unit 120, and a driving unit 130.

The communication unit 110 may be electrically connected to the control unit 120 and may communicate with the docking station 10 for charging. The communication unit 110 may receive a charging signal from the docking station 10. At this time, the communication unit 110 includes global system for mobile communication (GSM), code division multiple access (CDMA), wideband code division multiple access (WCDMA), and universal wireless communication (UMTS) modules, in addition to the Wi-Fi module and the wireless broadband module. It may include a wireless communication module that supports various wireless communication methods, such as mobile telecommunications system), Time Division Multiple Access (TDMA), Long Term Evolution (LTE), 4G, 5G, and 6G.

The control unit 120 performs the above-described operations using a memory 121 that stores data for an algorithm for controlling the operation of components in the device or a program that reproduces the algorithm, and the data stored in the memory 121. It may be implemented with at least one processor 122. Here, the memory 121 and the processor 122 may each be implemented as separate chips. Additionally, the memory 121 and processor 122 may be implemented as a single chip.

The memory 121 can store data supporting various functions of the device and a program for the operation of the control unit, can store input/output data, and can store a plurality of application programs (application programs or applications) running on the device. (application)), data and commands for operation of the device can be stored. At least some of these applications may be downloaded from an external server via wireless communication.

The memory 121 may be a flash memory type, a hard disk type, a solid state disk type, an SDD type (Silicon Disk Drive type), or a multimedia card micro type. micro type), card type memory (e.g. SD or XD memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), EEPROM (electrically erasable) It may include at least one type of storage medium among programmable read-only memory (PROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, and optical disk. Additionally, the memory 211 is separate from the device, but may be a database connected by wire or wirelessly.

The memory 121 may store global routes for each destination. The global route for each destination may be a route for the unmanned mobile device 100 to arrive at a specific destination in a space such as a hospital, airport, mall, etc.

The memory 121 may store map information (A). At this time, as shown in FIG. 2, the map information (A) may be a grid map. The grid map can be created in a form for 2D or 3D autonomous driving through sensing of the unmanned vehicle 100 or by using various SLAM (Simutaneous Localization And Mapping) techniques in advance. For example, a grid map can be created as a 2D map in the form of image files such as png or pgm, or as a 3D map through point cloud data containing 3-axis spatial information. At this time, the first point (B) of the map information (A) may be a starting point for finding a location in space, and the second point (C) may be an end point for finding a location in space.

The processor 122 may control the operation of finding a location in space.

The processor 122 may detect whether an abnormal reboot has occurred. If an abnormal reboot occurs based on the detected result, the processor 122 may determine whether a charging signal from the docking station 10 has been received through the communication unit 150.

If the charging signal is not received, the processor 122 may perform a kidney recovery mode to find a location in space. In order to find a location in space, the processor 122 may collect map information-based image data and LIDAR data to create an inference model by driving in a preset zigzag pattern from a preset first point to a second point. At this time, the processor 122 may rotate 360 degrees at each of the plurality of points within the first point and the second point to further collect a unit number of image data and LIDAR data for each rotation angle.

The processor 122 may extract spatial data within the image by inputting the collected image data into an image extraction model. The processor 122 may input spatial data and LIDAR data in the image into a position tracking model and output expected position data in space of the unmanned moving object. Here, spatial data may be feature data. At this time, the processor 122 clusters the expected location data between clusters, selects a cluster among the clustered clusters in which the number of expected location data is greater than a preset number, and if there are the same number of clusters among the clustered clusters, Clusters with low variance can be selected. Additionally, if the expected position data deviates from a preset error value, the processor 122 may input the LiDAR data into an Iterative Closet Point (ICP) algorithm and further output corrected expected position data.

The driving unit 130 drives an unmanned moving object (unmanned moving object) to travel from a first point to a second point in order to find a position in space based on driving information including zigzag pattern driving, 360-degree rotation driving, etc. output through the processor 122. 100) can be operated. For example, the driving unit 130 includes a battery, an electronic component that receives power from the battery, a wheel or a drive belt that receives power from the battery and drives the unmanned vehicle 100 in order to provide power necessary for driving the unmanned vehicle 100. It may include mechanical driving parts such as.

3 to 5 are flowcharts showing a method for finding the location of an unmanned moving object in space according to the present disclosure. FIG. 6 is a diagram illustrating an example of a process in which an unmanned mobile device finds a position in space through the processor of FIG. 1.

FIG. 7 is a diagram showing an example of a process for storing expected position data and corrected expected position data in the memory of FIG. 1 . FIG. 8 is a diagram illustrating an example of the process of clustering expected location data between clusters and selecting the optimal cluster using the processor of FIG. 1.

Referring to FIGS. 3 to 8, the method for finding the location of an unmanned moving object in space includes a detection step (S310), a decision step (S320), a mode execution step (S330), a collection step (S340), and an extraction step (S350). , may include an output step (S360).

In the detection step, whether an abnormal reboot of the unmanned mobile device 100 has occurred can be detected through the processor 122 (S310). For example, an abnormal reboot may include at least one of a state in which the device turns itself off due to insufficient power or a state in which the device turns on by itself due to a system error. That is, the processor 122 may determine that the reboot is abnormal if the reboot command is not based on the user's power input signal.

In the determination step, if an abnormal reboot occurs based on the detected result, the processor 122 may determine whether a charging signal from the docking station 10 has been received (S320). At this time, the processor 122 can check whether it is in front of the docking station 10 using the charging signal from the docking station 10. That is, when the processor 122 receives a charging signal, it can determine that it exists in front of the docking station 10 and recognize that it is currently at the starting point. If the processor 122 does not receive a charging signal, it may determine that it does not exist in front of the docking station 10 and recognize it as being in a different position from the current starting point.

In the mode performance step, if the charging signal is not received through the processor 122 (S320), a kidney recovery mode to find a location in space may be performed (S330). Here, the kidnapping state is a state in which the unmanned vehicle has been moved to another location where it is not known which location it was moved to, a state in which the unmanned vehicle has been moved by someone, or the direction or location of the unmanned vehicle due to a collision with an unexpected object, etc. It may include at least one of a changed state and a state in which the unmanned moving object cannot determine the pre-designated location. Accordingly, in the mode execution step, the kidnap recovery mode can be performed so that the unmanned vehicle can find its location in space even when it is in the kidnap state.

As shown in FIG. 6, the collection step is performed through the processor 122 by performing a preset zigzag movement from a preset first point (B) to a second point (C) so that the unmanned mobile object 100 finds its position in space. By driving the pattern (P), image data and lidar data can be collected to create an inference model (S340). The processor 122 may collect image data and LIDAR data obtained from the camera 140 and the sensor unit 150. The processor 122 may collect image data and LIDAR data by driving a zigzag pattern (P) at each of a plurality of points (P1 to P11,...) from the first point (B) to the second point (C). At this time, the first point (B) may be a starting point for finding a location in space, and the second point (C) may be an end point for finding a location in space. Here, the image data may include the position coordinates of the unmanned moving object, and the LIDAR data may include distance values of surrounding objects of the unmanned moving object.

In addition, in the collection step, through the processor 122, each of the plurality of points (P1 to P11, ...) within the first point (B) and the second point (C) is rotated 360 degrees to collect a unit number of image data for each rotation angle. You can collect more LIDAR data (S340). For example, the processor 122 may collect 36 pieces of image data and LIDAR data per 10 degree rotation for each of the plurality of points (P1 to P11,...). The processor 122 collects image data and LIDAR data by rotating 360 degrees and driving in a zigzag pattern (P) at each of the plurality of points (P1 to P11, ...) to the first point (B) and the second point (C). can do. For example, the processor 122 collects 36 image data and LIDAR data per 10-degree rotation at point P1 until the unmanned vehicle completes a 360-degree rotation (first process), and the unmanned vehicle rotates 360 degrees at point P1. Once the 360-degree rotation is completed, video data and lidar data can be collected by driving in a preset zigzag pattern (P) to point P2 (second process). This unmanned vehicle travels in a zigzag pattern (P) at each of a plurality of points (P1 to P11,...) from the first point (B) to the second point (C) and rotates 360 degrees to collect image data and lidar data. It can be collected.

Here, the camera 140 may be provided as a video camera using an image sensor. The camera 140 can process image frames such as still images or videos acquired through an image sensor. Meanwhile, when there are a plurality of cameras 140, they may be arranged to form a matrix structure. A plurality of image information having various angles or focuses can be input through the video cameras forming this matrix structure. Additionally, video cameras may be arranged in a stereo structure to acquire left and right images to implement a three-dimensional stereoscopic image. Additionally, the camera 140 may be provided as an AI camera. The AI camera can finely adjust images recognized through a wide-angle sensor that mimics the human retina using a brain neural network algorithm. AI cameras can adjust shutter speed, light exposure, saturation, color depth, dynamic range, contrast, etc. Additionally, AI cameras can output captured images as high-quality images. The sensor unit 150 may include at least one of an ultrasonic sensor, a radar sensor, and a lidar sensor. The sensor unit 150 can acquire LiDAR data through a LiDAR sensor.

As shown in FIG. 7, the extraction step inputs the collected image data (ID1) into the image extraction model (M1) through the processor 122 to extract feature data (OD1) in the image (S350). . At this time, the image extraction model (M1) may be a model that can robustly recognize and extract images.

In the output step, feature data (OD1) and lidar data (ID2) in the image are input into the position tracking model (M2) through the processor 122 to output expected position data (OD2) in space of the unmanned moving object. (S360). At this time, the expected position data may include the position coordinates (x, y) of the unmanned mobile object, and the rotation angle (θ) at which the unmanned mobile object looks at least one of front, side, and rear. For example, the position coordinates (x, y) of the unmanned mobile device may include 36 x and y coordinates, and the rotation angle (θ) may be the position of the unmanned mobile device in at least one of the 1 o'clock to 12 o'clock directions. It could be the viewing angle.

Here, the server 20, which is electrically connected to the control unit 120, learns the input values of feature data (OD1) and lidar data (ID2) in the image, which are metadata, based on the location tracking model (M2) to make recommendations. The result of the expected position data (OD2) within the space of the unmanned moving object can be output. At this time, feature data OD1 in the image may include the spatial coordinates of the unmanned moving object, and LiDAR data ID2 may include distance values of surrounding objects of the unmanned moving object. Additionally, the location tracking model (M2) can be built to learn through correlation between feature data (OD1) and lidar data (ID2) in the image included in the input data. The location tracking model (M2) can be constructed and reinforced as a learning data set using the feature data (OD1) and lidar data (ID2) in the image using the DNN algorithm.

The server 20 may transmit the expected position data (OD2) in space of the unmanned moving object determined based on the characteristic data (OD1) in the image and the lidar data (ID2) to the memory 121 of the control unit 120. The memory 121 may store the received expected position data OD2 in space of the unmanned moving object. At this time, the memory 211 may store the position coordinates (x, y) of the unmanned moving object having 36 x and y coordinates in the expected position data (OD2) in space, and the unmanned moving object may be positioned between 1 o'clock and 12 o'clock. The rotation angle (θ) viewed in at least one of the directions or 36 directions may be stored.

When image data and LIDAR data corresponding to the current position in space acquired from the camera 140 and the sensor unit 150 are received, the processor 122 generates the expected position data in space of the unmanned moving object based on artificial intelligence. It can be extracted.

As shown in FIG. 8, the output step may cluster the expected location data between clusters (A, B, C) through the processor 122 (S371). In the output step, if there are the same number of clusters among the clustered clusters (A, B, C) through the processor 122 (S372), a cluster with low variance can be selected (S373). In the output step, if there is no same number of clusters among the clustered clusters (A, B, C) through the processor 122 (S372), a cluster in which the number of expected location data is greater than the preset number can be selected (S374). .

Here, the point clouds of the cluster (A, B, C) correspond to the expected position data and may include the position coordinates (x, y) of the unmanned moving object with 36 x and y coordinates, and the unmanned moving object may include a rotation angle (θ) viewed in at least one of the 1 o'clock direction to the 12 o'clock direction, and the rotation angle may be calculated. For example, if there are the same number of A clusters and B clusters, the processor 122 may select B clusters with low variance. For another example, if there are no clusters with the same number, the processor 122 may select cluster C in which the number of expected location data is greater than the preset number. In order to increase the accuracy of the expected location data, the processor 122 clusters the expected location data between clusters (A, B, C), and clusters or clusters with low variance in the point clouds of the clusters (A, B, C). You can select a cluster with a large number of expected location data.

Meanwhile, in the output stage, if the expected position data deviates from the preset error value (S375) through the processor 122, the lidar data (ID2) is input to the ICP (Iterative Closet Point) algorithm (M3), and the corrected Expected position data (OD3) can be further output. The processor 122 provides the position coordinates (x, y) of the unmanned moving object having 36 x-coordinates and y-coordinates corresponding to the expected position data, and the position coordinates (x, y) of the unmanned moving object looking in at least one direction from 1 o'clock to 12 o'clock. It can be determined whether the rotation angle θ is outside a preset error value. The ICP algorithm (M3) can analyze expected location data more precisely by using an algorithm that registers two different point clouds. The ICP algorithm (M3) can find correlations using the closest points of expected location data, and move and rotate the current data based on the found correlations to add it to the existing dataset. Since the processor 122 using the ICP algorithm (M3) takes 0.1 seconds to process one image, it takes 3.6 seconds to process 36 images, and the Kidnap recovery mode can be performed within about 4 seconds. there is.

When the unmanned mobile device 100 finds its location in space based on the above-described expected location data or corrected expected location data, it can travel along an accurate path to a preset destination or a docking station 10 for charging.

100: Unmanned mobile device 110: Communication department
120: Control unit 121: Memory
122: processor 130: driving unit
140: Camera 150: Sensor unit

Claims (10)

In a method of outputting expected position data in space of a mobile robot using a position tracking model,
Detecting whether an abnormal reboot of the mobile robot has occurred;
If the abnormal reboot occurs based on the detected result, determining whether a charging signal from the docking station has been received;
If the charging signal is not received, performing a kidnap recovery mode to find a location in space;
Collecting image data and LIDAR data to create an inference model by driving in a preset zigzag pattern from a preset first point to a second point so that the mobile robot can find its position in the space;
extracting spatial data within the image by inputting the collected image data into an image extraction model; and
Inputting the spatial data in the image and the LIDAR data into a position tracking model to output expected position data in space of the mobile robot; Including,
The expected position data includes position coordinates of the mobile robot and rotation angles at which the mobile robot looks forward, sideways, and backward. According to paragraph 1,
The collection step is,
A method characterized in that each of the plurality of points within the first point and the second point is rotated 360 degrees to further collect a unit number of image data and lidar data for each rotation angle. According to paragraph 2,
The output step is,
Clustering the predicted location data between clusters,
A method characterized in that, among the clustered clusters, a cluster in which the number of expected location data is greater than a preset number is selected. According to paragraph 3,
The output step is,
A method characterized in that, if there are the same number of clusters among the clustered clusters, a cluster with low variance is selected. According to paragraph 4,
The output step is,
If the expected location data exceeds the preset error value,
A method characterized by inputting the LIDAR data into an Iterative Closet Point (ICP) algorithm and further outputting corrected expected position data. a communication unit communicating with the docking station; and
a processor that controls the operation of finding a position in space;
The processor,
Detects whether an abnormal reboot occurs,
If the abnormal reboot occurs based on the detected result, determine whether a charging signal from the docking station has been received through the communication unit,
If the charging signal is not received, a kidney recovery mode is performed to find the location in the space,
In order to find the location in the space, video data and lidar data are collected to create an inference model by driving in a preset zigzag pattern from a preset first point to a second point,
Input the collected image data into an image extraction model to extract spatial data within the image,
Input spatial data in the image and the lidar data into a position tracking model to output expected position data in the space,
The expected position data includes position coordinates, rotation angles looking forward, sideways, and backward, a mobile robot. According to clause 6,
The processor,
A mobile robot, characterized in that it rotates 360 degrees at each of the plurality of points within the first point and the second point and further collects a unit number of image data and lidar data for each rotation angle. In clause 7,
The processor,
Clustering the predicted location data between clusters,
A mobile robot, characterized in that among the clustered groups, a group in which the number of expected location data is greater than a preset number is selected. In article 8,
The processor,
A mobile robot, characterized in that if there are the same number of clusters among the clustered clusters, a cluster with low variance is selected. According to clause 9,
The processor,
If the expected location data exceeds the preset error value,
A mobile robot, characterized in that the lidar data is input into an ICP (Iterative Closet Point) algorithm to further output corrected expected position data.

Images