KR101638679B1 - The outdoor guide robot with automatic location based guidance and navigation - Google Patents

Abstract

BACKGROUND OF THE INVENTION Field of the Invention [0002] The present invention relates to an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance. More particularly, the present invention relates to a method and apparatus for processing a plurality of position sensor information, judging a current position on its own during a movement, autonomously navigating a guide in a database, It is expected that it will be possible to obtain a good effect that the reliability of the outdoor location recognition is high, the autonomous driving ability is maximized, and the versatility is maximized.

Description

[0001] The present invention relates to an outdoor-purpose general-purpose guide robot apparatus capable of autonomous navigation and location-

BACKGROUND OF THE INVENTION Field of the Invention [0002] The present invention relates to an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance. More particularly, the present invention relates to a method and apparatus for processing a plurality of position sensor information, judging a current position on its own during a movement, autonomously navigating a guide in a database, To a general purpose guidance robot apparatus for outdoor use.

2. Description of the Related Art In recent years, unlike conventional industrial robots, there have been many developments of robots used in people's lives such as office robots and pet robots. As for the work of the robots, guidance service to the destination in the office and a conversation partner of the elderly are assumed. For example, an office robot has developed a robot that asks a guest for a destination, hears a destination from a guest, and the robot moves toward the destination and guides the robot by walking the guest.

In addition, large-scale department stores and outdoor exhibition sites have a complicated structure and a wide range of movement. For convenience of users, information centers are installed at specific locations, and kiosks are installed to provide guidance services to users . Since the guide service of the prior art is performed at a specific location, the user must go to a place where the guide is located or a place where the kiosk is installed in order to receive the guide service. That is, in the case of the first-time visitor, there is a problem that it becomes an obstacle to the use of the guide place from finding the installation place of the information desk or the kiosk. That is, the conventional guidance system has a problem in that it can not provide an active guidance service for the user.

Such a problem consumes a lot of time for the guidance work for the users, which leads to a waste of labor costs, an inefficient waste of the working time, and a problem of lowering the mobility of the user.

Korean Patent Publication No. 10-1012288 discloses a portable type mobile communication service which provides services such as location guidance, local information provision, local advertisement provision, mobile coupon issuance and ticket sales to a user by performing an autonomous movement within a specific place and interacting with a user A robot, and a system thereof; A driving unit; A sensor unit; An input unit; A storage unit (DB); A communication unit; A central control unit; An output section; And the central control unit includes a central management server that manages the guide place through a communication network with the mobile coupon issuance information when the user issues a mobile coupon, a personal terminal of the mall, and a mobile coupon To the mobile communication terminal of the user who has issued the information.

In Korean Patent Laid-Open Publication No. 10-2013-0024159, a distance detecting sensor provided on a body of a robot constituted by a guide including a distance detecting sensor and a driving unit is rotated, and a front / rear / left / right (Measurement plane) positioned in the up / down (lower) direction as well as the distance between the obstacle (measurement plane) positioned in the direction of the obstacle and the obstacle Respectively.

Japanese Patent Publication No. 8-108381 relates to a traveling robot which can precisely follow a path set by a guide rail and which can improve the degree of following and which makes it difficult for the bent portion to pass through or to be easily transmitted A pair of right and left running wheels supported by an axle orthogonal to the pivot shaft and a pair of left and right running wheels supported by an axle orthogonal to the pivot shaft, And the guide roller is composed of one child roller which rolls along one side of the guide rail and two pressing rollers which are arranged so as to be able to roll along the other side of the guide rail, And is arranged symmetrically with respect to a vertical plane connecting the contact point and the pivot shaft. Specifically, And the roller and the pressing roller to be installed are disposed on the opposite side of the guide rail.

Japanese Patent Publication No. 2010-188429 discloses a route guidance robot including a time determining means, a sentence firing means, a time storing means, a gesture means, a shortened information storing means, a stopping means, and a shortening means .

Japanese Patent Application Laid-Open No. 2004-216552 discloses a mobile robot for controlling a landmark installed in an up space to control its own position and autonomously moving, and an autonomous traveling system and method thereof, A communication module for transmitting a light source control signal for controlling blinking and a light source controlled to be blinked based on the light source control signal are detected from a video signal photographed by a camera and an image processing module for reading image coordinates of the light source, A position calculating module for calculating the position coordinates of the mobile robot using the coordinates and the world coordinates of the light source stored in advance and the position coordinates of the calculated mobile robot in correspondence with the space coordinates of the previously stored work space, A motion control module for setting and driving a path, And a main control module for performing overall operation control of the robot.

However, no technology has been proposed for a robot that carries out autonomous navigation and guidance based on positional information that can reliably determine its position in the open air.

Korean Patent Publication No. 10-1012288 Korean Patent Publication No. 10-2013-0024159 Japanese Patent Publication No. 8-108381 Japanese Patent Publication No. 2010-188429 Japanese Patent Publication No. 2004-216552

An object of the present invention is to provide an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance. More particularly, the present invention relates to a method and apparatus for processing a plurality of position sensor information, judging a current position on its own during a movement, autonomously navigating a guide in a database, And it is an object of the present invention to provide a general purpose guidance robot apparatus for outdoor use which can guide the robot.

In order to accomplish the above object, the present invention provides an outdoor general purpose guidance robot apparatus capable of autonomous navigation and position-based autonomous guidance,

Guiding input / output means for inputting guidance request information of the subject and performing display of the guidance information;

Position estimating means for processing at least one position display sensor information into an extended Kalman filter according to the inputted guidance request information;

Route selection means for receiving the position information by the position evaluation means at the main controller and moving the predetermined route while storing the position information according to the predetermined correction position calculated through the position evaluation means;

An autonomous running means for reading the predetermined route data from the main controller and moving along a route point stored from an initial position to a final position;

And an autonomous guide means for reading the video and audio data from the guide database through the guide input / output means by guiding the video and audio data at the position point located in the predetermined route section while proceeding the autonomous travel, The present invention provides an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance.

The at least one position indicating sensor may be a Dead-reckoning position indicating sensor, a DGPS position indicating sensor, and an electronic compass position indicating sensor.

The position estimating means calculates the moving distance and the moving direction of the robot apparatus from the encoder rotation information of the robot apparatus wheel, which is the estimated navigation position indicating sensor, and calculates the first new position and the first position of the robot apparatus using the moving distance and the moving direction information. Estimates the position error covariance,

Receiving the position information from the DGPS position indicating sensor and receiving direction information from the electronic compass position indicating sensor to estimate a second new position and a second position error covariance and determining the first new position and the second new position and the first position The reliability of each positional information is calculated using the difference between the error covariance and the second positional error covariance, the position of the robot is updated by updating the robot position in proportion to the reliability of each information, and the covariance of the corrected robot position is updated The Kalman filter may be a position evaluating means to which a Kalman expansion filter is applied.

The pre-path selection means may store the current point information at predetermined intervals by calculating the distance from the pre-stored storage location to the current location by storing the location information.

The predetermined interval may be 1 m. Preferably 0.5 m. More preferably 0.2 m. If the interval is larger than 1 m, the accuracy of the position information is lowered. If the interval is narrower than 0.2 m, the efficiency of storing the position information is lowered.

Wherein the autonomous running means sets position information from the DGPS position display sensor when the guidance request information of the target person is input and searches an objective point closest to the current position of the robot apparatus from the predetermined path data to select an initial target point And sets each predetermined route point as a target point in order and corrects the error from the position information of the position evaluation means at the current position to determine the current position and start autonomous travel, If the distance error is less than a predetermined value, the robot determines that the robot has reached the current target point. Otherwise, it repeats the process of maintaining the current target point. If the robot reaches the final target point, can do.

The distance error may be 0.1 m. Preferably 0.05 m. More preferably 0.01 m. The distance error value should be set to have a value as small as possible.

delete

delete

The video and audio data of the autonomous guide means are classified into three types as main guide data, auxiliary guide data, and other guide data. The main guide data is a place where the main guide should be performed in the entire guide space, And the auxiliary guidance data performs the guidance while the robot apparatus is moving and needs to explain the surrounding situation while the other guidance data is a description of the surrounding situation during the movement of the robot apparatus And the guidance information is related to the guide position so that the target person to be guided when passing through the unnecessary area is not tedious. The video and audio data are transmitted to the robot device And if there is no corresponding guiding data, the other guiding data is retrieved from the guiding database And can be carried out the instructions and other data descriptions main guide data or the auxiliary guide data during the performing performs the other guide data, guide data to perform primary or secondary guide data stops when the current position corresponds to the robot moves.

According to the present invention, when using only GPS, the position error can not be recognized if the position error is large and the information is not received from the satellite. However, the present technology uses the digital compass and the encoder information of the robot wheel, Not only the GPS alone is improved but also the recognition of the location is possible even when the GPS information is not received for a short period of time, so that the reliability of the outdoor location recognition is high.

Secondly, if the route to be driven in advance is manually operated, the robot itself memorizes the route and recognizes the route as a guide route (or inputs guide route information in advance instead of manual travel) Since the robot runs autonomously by itself, there is no need for the user to control the robot, and the robot itself has an autonomous running ability capable of guiding the passengers and guiding them.

Thirdly, it is not a method of programming the guidance procedure in advance, but the robot searches the guidance database for guidance information corresponding to the position of the robot on the move, and performs the guidance, so that it is not necessary to re- If the database is built only, it can be applied as it is without modification, regardless of the location. In other words, in case of changing the guide route in the same place, there is no need to modify any program, and there is a general possibility that the robot operating in the place A can be applied as it is in the place B without any modification.

1 is a conceptual diagram of a DR localization system for an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance.
FIG. 2 is a conceptual diagram of moving state of a robot according to time for EKF localization of an outdoor general purpose guide robot apparatus capable of autonomous navigation and location-based autonomous guidance.
FIG. 3 is an experimental place where a reference position of an outdoor general purpose guide robot apparatus capable of autonomous navigation and location-based autonomous guidance is set.
6 is a graphical representation of error characteristics of an outdoor general purpose guidance robot apparatus capable of autonomous navigation and position-based autonomous guidance.
7 is a Kalman filter-based position estimation method for an outdoor general purpose guidance robot apparatus capable of autonomous navigation and position-based autonomous guidance.
11 is an autonomous guidance performing algorithm of an outdoor general purpose guidance robot apparatus capable of autonomous navigation and location-based autonomous guidance.
12 is an autonomic guidance experiment environment photograph of an outdoor general purpose guide robot apparatus capable of autonomous navigation and location-based autonomous guidance.
13 is a photograph of an autonomous guidance experiment result of an outdoor general purpose guide robot apparatus capable of autonomous navigation and position-based autonomous guidance.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. It should be noted that the same reference numerals are used to denote the same or similar components in different drawings, even if they are shown in different drawings.

Also, the terms and words used in the present specification and claims should not be construed as limited to ordinary or dictionary terms, and the inventor should properly define the concept of the term to describe its invention in the best way The present invention should be construed in accordance with the meaning and concept consistent with the technical idea of the present invention. Therefore, the constitutions described in the embodiments described herein are merely the most preferred embodiments of the present invention, and are not intended to represent all of the technical ideas of the present invention. Therefore, various equivalents which can be substituted at the time of application It should be understood that variations can be made.

The DR (Dead-Reckoning) localization principle is a method to estimate the current position by calculating the trajectory of the robot by detecting the moving distance and direction of the robot using only the installed sensor after setting the initial position of the robot. DR does not require external support facilities and can continuously detect the position of the robot. This localization is a navigation system that calculates the new position of the robot from the moving distance and rotation angle information and the information of the encoder mounted on both wheels when moving from the known position to the next step.

DR is a navigation system that calculates the new position of the robot from information of movement distance and rotation angle information and encoders mounted on both wheels when moving from the known position to the next step, But provides relatively precise information for a short period of time. As shown in Fig. 1, when the moving distances of the left and right wheels of the robot are D _L and D _R , respectively, and the distance between the two wheels is L,

. In this equation, R and α are unknown. From this, R and α are obtained as follows.

의 로봇 중심에 대한 다음 위치

는 다음과 같이 구해진다. 먼저 x_i 는,Previous location

Next location for robot center of

Is obtained as follows. First x _i is,

In the same way, y _i is as follows.

Next, the moving direction of the robot with respect to the reference coordinates is calculated as follows.

Also, the position of the robot relative to the reference coordinates is calculated as follows.

here,

to be.

However, the DR increases infinitely as the distance traveled by the robot increases. The reason is due to the unevenness of the wheel, the slip between the wheel and the floor, the unevenness of the floor, and the error of the encoder. The dual wheel imbalance is a systematic error and has the property of infinitely extending according to the moving distance. Other error factors, on the other hand, have random characteristics and can be processed statistically.

When considering DGPS Localization, GPS (Global Positioning System) is used for military purposes by receiving radio waves emitted from satellites and obtaining positioning information. However, it has been expanded to private sector in 1983, It is available 24 hours a day, anywhere in the world, and is resistant to interference and disturbance outside weather conditions and uses the World Geodetic System (WGS-84) worldwide. GPS (Global Positioning System) is divided into space part, control part and user part. The space section has four satellites assigned to each of the six orbital planes, three of which are reserved satellites. Since the altitude of the satellite is about 20,183 km, the orbital period is about 12 hours, and the inclination angle of the orbits is at orbit around the same orbital plane, I can see it. The control part tracks the satellites, determines the trajectory of each satellite, maintains the precise time, transmits the information to the satellites, and receives the operating status. The user part is composed of a GPS receiver, an antenna and software for data processing And receives the radio wave transmitted from the satellite to measure the coordinates of the receiving point.

DGPS (Differential GPS) is a GPS receiver that is installed at a reference point that already knows the location, and it receives the GPS satellite signal with its own precise measured position value and compares the position calculated by the receiver, Is high. The DGPS of Korea is composed of one central office of Daejeon of the Central Government, 11 base stations and 8 monitoring stations. The central office real-time remote monitoring and control of the DGPS operation status of the reference station and the monitoring station using the PSDN network, and simultaneously monitoring the positioning information, monitoring the satellite status integrity, checking the availability status of the GPS through the reference station and monitoring station positioning error analysis Information such as positioning data (Post Processing Data: PPS), Reciprocal Independent Exchance data (RINEX) and SSF (Standard Storage Format) are provided free of charge on the web (http://www.ndgps.go.kr) .

The reference station installed a GPS antenna at the correct reference point, received each GPS satellite signal, compares the measured distance with a known distance, and then corrected the satellite error value. The RTCM SC-104 (Radio Technical Commission for Maritime Services, 104) using a medium frequency (283.5-325 KHz) according to the format format. The monitoring station uses a GPS antenna at a distance of about 100 NM (180 km) from the reference station. When the satellite error correction signal exceeds the threshold and when the satellite signal is abnormal, the alarm message is transmitted to the central office. .

However, since the DGPS location information also includes various error factors, there is a position error. Major location error factors include random errors, location error, ionospheric error, communication error, and human error. Among them, the artificial error is deliberately interfering with the error using the error function in the US satellite control center to limit the use to some extent except the military purpose. For several years, the artificial error intervention was stopped to provide accurate information to the private sector. In addition, GPS localization requires reliable reception of information from at least four satellites.

When considering the localization based on the extended Kalman filter (EKF), the necessity of the sensor fusion is continuously accumulated as DR is used to evaluate the position, and the position error gradually increases with time. In addition, when using only DGPS, the error does not accumulate with time, but an error of 5m or more may occur. In addition, according to the satellite state, the position information can not be received from the four satellites. Therefore, the DR navigation method can be effective for the local short-haul driving, and the localization for the long-time movement is better by the DGPS. Therefore, it is possible to perform optimal localization by combining only the advantages of both methods.

The localization that combines DR and DGPS has already been reported, but the error is large in the case of straight running, such as corner running, where the position is evaluated effectively. This is because DGPS only provides location information but does not provide direction information. In order to compensate for the drawbacks of such curved lines, this task develops a localization method that uses the digital compass to further use the angle information. A digital compass is a sensor that provides absolute azimuth information from earth magnetic field information. Such DGPS or digital compass is advantageous in that there is no error accumulation due to a bias error, which is a disadvantage of an inertial sensor such as a gyro sensor, as an inertial sensor.

The Kalman filter is the most efficient way to evaluate optimal values by combining two or more pieces of information. The Extended Kalman Filter (EKF) is an algorithm widely used for information fusion that evaluates the optimal measurement value with minimum error from multiple measurements containing noises.

로 정의하고 는 이동거리, 그리고

는 이동방향으로 정의한다.Fig. 2 shows a state in which the robot moves to a position at a time k and a position at a time k + 1. A state variable indicating the position of the robot

And distance traveled, and

Is defined as the movement direction.

The system model for the position of the robot is shown in Fig.

는,Here,

Quot;

는 다음과 같이 가정한다.And error

Is assumed as follows.

는 평균이 0이고 공분산이 Q(k)인 가우시안이 된다. (9)식에서 d(k)와

는 제어입력으로서 상수로 가정한다. 또한 Q(k)는 각 상태변수의 오차 표준편차로서 대각 행렬이 된다.In other words,

Becomes a Gaussian with an average of 0 and a covariance of Q (k). (9), d (k) and

Is assumed to be a constant as a control input. In addition, Q (k) is the diagonal matrix of the error standard deviation of each state variable.

     Next, the measurement model for measuring the position and orientation information from DGPS and digital compass is as follows.

는 평균이 0이고 공분산이 R(k)인 가우시안 잡음이며, R(k)는 각 측정변수의 오차의 표준편차로 구성되는 대각 행렬이다. DGPS 및 디지털 컴퍼스는 정확한 위치에 단지 랜덤 잡음(오차)만 개입된 것으로 가정하여 Z(X(k), S_t)를 다음과 같이 정의한다.here

Is a Gaussian noise with an average of 0 and a covariance of R (k), and R (k) is a diagonal matrix consisting of the standard deviation of the error of each measurement variable. The DGPS and the digital compass define Z (X (k), S _t ) as follows, assuming that only the random noise (error) is involved in the correct position.

The system model and the measurement model defined above are applied to the EKF algorithm to evaluate the position of the robot as follows. First, the position of the robot at time k + 1 from the system model and the control input u (k) is predicted as follows.

는 다음과 같이 계산된다.The covariance matrix associated with this prediction

Is calculated as follows.

는 상태 천이함수

의 자코비안으로서 아래 식과 같이 구해진다.here,

State transition function

The following equation is obtained.

Next, the predicted measurement model from the system model and the control input u (k) is obtained as follows.

Also, actual measured sensor actual model from DGPS and digital compass is shown as follows.

는 DGPS로부터 측정된 위치정보이고

는 디지털 컴퍼스의 방위정보이다. 이 실측 위치와 시스템 모델로부터 예측된 위치의 차로 구성되는 innovation 행렬

는 다음과 같다.here

Is the position information measured from the DGPS

Is the orientation information of the digital compass. An innovation matrix consisting of the difference between the actual position and the predicted position from the system model

Is as follows.

에 수반되는 공분산 행렬

은 다음과 같다.

The covariance matrix

Is as follows.

는 측정모델의 자코비안으로서 다음과 같이 단위행렬로 주어지는데 그 이유는 DGPS 및 디지털 컴퍼스의 정보는 시스템 모델과는 무관하게 독립적으로 정보를 주기 때문이다.here

Is given as a unit matrix as a Jacobian of the measurement model because the information of DGPS and digital compass gives independent information irrespective of the system model.

The EKF evaluates the correspondence between the measured value and the predicted value and corrects the position using an innovation only when there is a certain degree of relevance. The evaluation of the association of two values uses the following validation gate.

는 설계 파라메타이다. Validation gate의 의미는 측정값과 예측 값의 차와 측정오차의 분산과의 비를 나타내는데, 측정오차가 오차의 분산보다 일정한 값 이하일 때만 위치보정이 유효하게 된다. here

Is the design parameter. The meaning of the validation gate is the ratio of the difference between the measured value and the predicted value to the variance of the measurement error. Only when the measurement error is less than the variance of the error, the position correction becomes effective.

로부터

즉, 시간

에서의 최적의 위치를 평가하고 거기에 수반된 공분산 행렬

을 갱신하는 것이다. 먼저 잘 알려진 칼만 게인

은 다음과 같이 정의된다.The final stage of the location assessment is the predicted location

from

That is,

And the covariance matrix associated with it

Lt; / RTI > First well known Kalman gain

Is defined as follows.

The meaning of this Kalman gain is the extent to which the position measured by the sensor is modified by weighting the estimated position of the hypothetical navigation.

Using the Kalman gain, correct the predicted position as follows.

Finally, the covariance matrix associated with this position correction is updated as follows.

In this way, optimal position estimation is performed by estimating the position correction and covariance matrix at each sampling time.

DGPS differs in accuracy depending on the method, and the higher the precision, the higher the price that costs millions of won. In this project, low-cost DGPS is used for practical application. Although the product specification has a position accuracy of 3 m, it has been observed that an error of up to 5 m occurs in actual test results. Table 1 shows the precision test results using DGPS. Although the specification shows a precision of 2.5 m, the actual test result shows that the maximum error is about 4.9 m as shown in Table 2.

The digital compass also uses a relatively inexpensive HMC model for commercialization, with a bearing accuracy of 3 degrees. Table 3 shows the specifications of the digital compass. The actual test result showed an average error of about 2.6 degrees.

In order to analyze the performance of the developed localization method, an actual driving experiment was performed. Experiments were carried out in such a way that the robot passed through 6 positions where the position information was already known, and the error between the actual position and the position evaluated by the robot was evaluated. The school site track as shown in Fig. 3 was selected as the experimental site. The playground is located long to the north and south, and six places marked with x in the figure are set as reference points for measuring the position error.

delete

As shown in FIG. 3, the reference position number is located at the lower right of the first position and the remaining positions are in the clockwise direction. The EKF localization method using the extended Kalman filter was chosen arbitrarily for the experiment because it is independent of the reference coordinate origin selection.

delete

The developed EKF localization method was performed several times on selected tracks. Table 5 shows the position error characteristics analyzed through experiments. In Table 5, the best error characteristic shows the best result of position evaluation during the experiment, and the worst error characteristic shows the opposite result. The reason why the error characteristics vary with each experiment is that the DGPS position error characteristic changes from experiment to experiment, and the EKF error characteristic also depends on the DGPS error characteristic.

For the best error characteristics, the DGPS maximum error was about 3.6m and the mean error was about 2.5m, while the EKF error was 1.7m maximum and 0.8m average. The maximum error of DGPS was 4.78m and the mean error was 3.2m, while the maximum error was 2.23m and the average was 1.9m. The error ratio is based on the mean error. The best EKF error is 32% of the DGPS error and the worst case is 57%. Overall error is about half of the error. Also, the maximum error is reduced to about half compared with DGPS, and the error characteristics are remarkably improved. The maximum error of DGPS localization is 4.78m, the mean error is 2.83m, and the standard deviation is 1.08m. In the case of EKF localization, the maximum error is 2.23m, the mean error is 1.33m, And a standard deviation of 0.73 m. In conclusion, the EKF localization has a maximum error of 2.23m and 95.4% reliability (twice the standard deviation), which can estimate the position of the robot with an error of 1.46m and the error is reduced by half compared with the DGPS position. . FIG. 6 is a graph showing error characteristics.

을 추정하고 이 추정에 연관된 위치오차 공분산

을 계산한다. 또한 위성으로부터 수신된 DGPS 위치정보 및 전자나침판(Digital compass)로부터의 방향정보를 이용하여 로봇의 현재 위치 및 방향

를 추정하고 이 추정에 연관된 오차공분산

를 계산한다. 다음으로

과

및

과

의 차이를 이용하여 각 위치정보의 신뢰성을 계산하고 각 정보의 신뢰성에 비례하여 로봇의 수정된 위치

를 평가하여 출력하며 아울러 수정위치에 대한 공분산을 갱신한다. 이와 같은 과정을 매 싸이클마다 수행함으로써 확장칼만필터 기반 위치평가를 수행한다.7 is a flowchart summarizing a Kalman filter-based position evaluation method. First, the robot calculates the moving distance d and direction θ from the information of each encoder attached to the wheel. Using this information, the position of the new robot

And the position error covariance associated with this estimate

. Also, using the DGPS position information received from the satellite and the direction information from the electronic compass, the current position and direction of the robot

And the error covariance associated with this estimate

. to the next

and

And

and

The reliability of each positional information is calculated using the difference between the position of the robot and the corrected position of the robot in proportion to the reliability of each information

And updates the covariance for the correction position. This process is performed every cycle to perform an extended Kalman filter based position estimation.

In the preliminary path recognition method, the pilot route for autonomous navigation of the guide robot travels the robot through the manual operation of the route to be guided by the robot to acquire the route data. In the manual mode, the main controller moves the position information of the satellite DGPS, direction information of Digital Comapss, and the position information of the robot's own encoder in real time while moving, and calculates the correction position calculated by the extended Kalman filter every 0.5m interval .

delete

delete

The autonomous running of the robot is performed by reading the path data stored in advance by the main program during autonomous driving and moving along the stored path from the initial position to the final position. The path travel control is performed by calculating the direction from the current position of the robot to the target point and controlling it to follow the direction continuously.

delete

delete

delete

delete

delete

delete

The guide image and voice data are stored based on the position (target point of the guide) where the data is to be executed to establish the guide database. That is, if any guide image and voice data are related to the building A, the position at which the robot reaches the building A and performs the guidance is the file name of the guide data. The guide information is classified into three types as main guide (A type), auxiliary guide (B type) and other guide (C type). The main guide is a place where major guidance should be performed in the whole guide space. In order to carry out the guidance. The auxiliary guidance data indicates a case in which the robot needs to explain its surroundings supplementarily while moving. The assistant guidance performs necessary guidance while the robot is not moving during the movement. Since the main guidance and the auxiliary guidance are guidance related to the position, the robot itself searches for guidance data corresponding to the position according to the position of the robot and performs guidance. On the other hand, the other guidance is a position-independent guide in order to avoid bored people being guided when the robot passes through an area that does not need to explain the surrounding situation while moving.
Further, when the guide database is constructed, it is possible to distinguish whether the guide data is primary, secondary, or other, so that the robot can determine whether to stop when performing the guidance or to perform guidance while driving. For example, if the guide information file name is '135_32_A', the reference position to be guided is the longitude 135 and the latitude 32, and the guide form indicates that the guide is the main guide of the stop guide.
The position-based autonomous guidance is performed when the robot estimates an optimum position by the position estimation method during movement and recognizes the current position and then searches whether or not there is the guidance data corresponding to the current position from the guidance database Method. This method does not need to arrange the guide information according to the guide order in advance, but also can guide the guide world without any program modification in the world when the global guide database is built in the existing Internet network.
The location based autonomous guidance execution algorithm is shown in the flowchart of FIG. When the robot estimates its optimal position by using the extended Kalman filter while moving, the robot always searches from the database whether or not there is guidance data corresponding to the position. If there is no guidance data corresponding to the current robot position, the robot carries out the autonomous running along the guiding route, and performs the guidance by randomly calling the other guidance from the database. When performing other guidance, the robot does not stop and continues the guiding route. If the guidance information corresponding to the guidance information is present in the guide database, it is determined whether it is the main guide form or the auxiliary guide form, and if it is the main guide form, the robot is stopped and guidance is performed. When the guidance is completed, the robot performs autonomous travel along the guide route again. In the auxiliary guiding mode, the robot performs guidance while traveling along the guiding path without stopping.
If another type of guidance data is searched while the assistant guidance is being performed while the robot is moving, guidance with a higher priority is performed according to the guidance type priority. The guide priority is set to be the highest (A type), followed by the assistant guidance (B type), and the other guidance (C type) is the lowest ranking. Accordingly, if the main guidance data is detected while the robot is moving while performing the auxiliary guidance, the auxiliary guidance is immediately stopped and the main guidance is performed. If the main guidance or auxiliary guidance data is detected during the other guidance, Perform secondary guidance.
By performing the guidance in this way, not only the necessary guidance on the guidance route is effectively performed but also the guidance is always performed even in the absence of the main guide or auxiliary guide, so that the guideee can always concentrate on guidance without feeling bored.

The experimental space as shown in FIG. 12 was set up to test the developed robot optimum position estimation method and the position based autonomous guidance performance method. The start is about 700m from the Central Building of Building 4 of the College of Engineering to the College of Social Science through Information Communication Center. The portion indicated by "o" in FIG. 12 represents the guide reference position where the main and auxiliary guide data should be performed. The guide reference position is located at the right side of the drawing from the fourth building of the College of Engineering, the information communication center, the parking lot, Science college order. Fig. 13 shows the autonomous guidance experiment result. In the figure, "X" is a position where the robot actually performed the guidance. The robot started from the No. 4 building of the engineering college and finished guiding at the center entrance of the Sociological Center. The maximum error of the guiding reference position and actual guiding position was measured to be 2.3m and 0.9m. The success rate of the robot guidance was 100% and the self-driving was also performed very well as shown in Fig.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit and scope of the invention will be.

Claims (7)

A universal guiding robot for outdoor use capable of autonomous navigation and position-based autonomous guidance,
Guiding input / output means for inputting guidance request information of the subject and performing display of the guidance information;
A dead-reckoning position display sensor, a DGPS position display sensor, and an electronic compass position display sensor information are processed by an extended Kalman filter according to the inputted guidance request information, Way;
Route selection means for receiving the position information by the position evaluation means at the main controller and moving the predetermined route while storing the position information according to the predetermined correction position calculated through the position evaluation means;
An autonomous traveling means for reading the predetermined route data from the main controller to set each point on the predetermined route from an initial position to a final position as a target point in order and moving along a stored route point; And
An autonomous guide means for autonomously guiding a guide image and voice data corresponding to a current position point of the robot to the robot in the guide database while autonomously running along the preset route and performing autonomous guidance through the guide input / ; / RTI >
The predetermined path data of the pre-path selection means moves the robot apparatus in the manual mode and displays the position information of the DGPS, the direction information of the digital compass and the dead-reckoning in real- The main controller receives the position information and stores the correction positions calculated through the extended Kalman filter at intervals of 0.2 m to 0.5 m,
The video and audio data of the autonomous guiding means are transmitted to the robot through a main guide for performing guidance in a stationary state at a place where the robot should perform the main guidance, auxiliary guidance for performing necessary guidance while moving, Wherein the guide route is selectively carried out during the movement of the guide with high priority among the other guides performing the guidance irrelevant to the position of the hazard.

delete The method according to claim 1,
The position estimating means calculates the moving distance and the moving direction of the robot apparatus from the encoder rotation information of the robot apparatus wheel, which is the estimated navigation position indicating sensor, and calculates the first new position and the first position of the robot apparatus using the moving distance and the moving direction information. Estimates the position error covariance,
Receiving position information from the DGPS position indicating sensor and receiving direction information from the electronic compass position indicating sensor to estimate a second new position and second position error covariance,
Calculating reliability of each positional information using the difference between the first new position, the second new position, the first positional error covariance and the second positional error covariance,
And a position estimating means for estimating a position of the robot by updating the robot position in proportion to the reliability of each information and applying a Kalman expansion filter for updating the covariance of the corrected robot position. General purpose guidance robot apparatus for outdoor use.
delete delete delete The method according to claim 1,
When the robot device evaluates its position while moving along the preset route, the robot device searches for the guidance data corresponding to the position, and if there is no corresponding guidance data, calls the other guidance data from the guidance database Wherein when the main guidance data or the auxiliary guidance data corresponds to the current robot position during the execution of the guidance and the other guidance data while performing the other guidance data, the execution of the other guidance data is stopped and the main guidance data or the auxiliary guidance data is executed. A general purpose guidance robot device for outdoor use capable of position - based autonomous guidance.

Images