JP2004326692A - Autonomous travelling robot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To attain a reliable work in a short time in a work area without storing a map of the work area in advance. <P>SOLUTION: A running motor is controlled and driven to turn a cleaning robot so that the front side of the robot faces the advancing direction. Then, the robot advances straight forward, and cleans a floor surface at the same time. When an obstacle sensor detects that there is an obstacle such as a wall ahead, the robot moves to the vicinity of the obstacle and stops advancing. The cleaning robot always estimates its own location and detects whether there is any obstacle, and records a cleaned area and the position of the obstacle in an attribute map as areas where the advance is disabled. When the advance is detected and stopped, the nearest uncleaned area is retrieved by using the attribute map. When retrieval is completed, the robot makes the nearest uncleaned area a target point, derives the shortest course to the target point, and moves along the obtained shortest course. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an autonomous traveling robot that travels autonomously in a work target area and performs work while creating a map.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, an autonomous mobile robot stores a map of a work area created in advance and performs work while traveling in the work area while referring to the map based on a work plan. 2. Description of the Related Art There is known a work area in which a map of a work area is created and stored, and the work is performed while traveling in the work area while referring to the map based on a work plan (for example, see Patent Document 1).
[0003]
[Patent Document 1]
JP-A-05-46239 (paragraphs "0012" to "0024" etc.)
[0004]
[Problems to be solved by the invention]
However, in the case where a map is created by making a single round around the work area in advance, it takes time and effort to create the map. Particularly, when a map of a work area having a complicated shape is created in advance, a large amount of work is required. There is a problem that it takes time, and it takes a long time to complete the actual work.
The present invention provides an autonomous mobile robot capable of performing a reliable work in a short time in a work area without storing a map of the work area in advance.
[0005]
[Means for Solving the Problems]
The present invention provides an autonomous traveling robot that includes a traveling unit and a working unit, and performs an operation with the working unit while autonomously traveling with the traveling unit.In the autonomous traveling robot, an obstacle detecting unit that detects an obstacle existing in a work target area, Map creation that divides the work target area into cells and records the obstacles detected by the obstacle detection means and the positions where work has been completed by the work means in this work target area in cell units at the time of work and creates a map Means, a recognition means for recognizing the closest unworked cell in the map from the current position at the time of work, and traveling to the unworked cell recognized by the recognition means when the obstacle detection means detects an obstacle ahead in the traveling direction. Control means for controlling travel of the means.
[0006]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In this embodiment, an embodiment in which the present invention is applied to a self-propelled cleaning robot will be described.
[0007]
1 and 2 show the configuration of a self-propelled cleaning robot. An operation in which various buttons, indicators, and the like are provided on the front upper portion of a housing 1 having a substantially circular lower portion and a substantially hemispherical upper portion. A panel 2 is arranged, a bumper 3 is provided from the front side to a lower side of the housing 1, and a plurality of obstacle sensors 4 including ultrasonic sensors are arranged on the bumper 3 as obstacle detecting means. . The obstacle sensors 4 are arranged, for example, three at a predetermined interval at a position visible from the front and two at a predetermined interval on the left and right side surfaces. The obstacle sensor 4 detects an obstacle in front and on the left and right.
[0008]
In the housing 1, a cleaner motor 5, a fan 6 rotated by the motor 5, and a dust collecting chamber 8 for sucking and collecting dust from a suction port 7 provided at the bottom by the rotation of the fan 6 are housed as working means. Have been.
[0009]
Also, a left driving wheel 9a and a right driving wheel 9b are respectively attached to the left and right of substantially the center of the bottom of the housing 1, and these driving wheels 9a and 9b are driven to rotate by a left traveling motor 10a and a right traveling motor 10b, respectively. ing. The driving wheels 9a and 9b and the traveling motors 10a and 10b constitute traveling means. The rotation of the driving wheels 9a and 9b is detected by rotary encoders 11a and 11b, respectively.
[0010]
At the center of the bottom rear end of the housing 1 is mounted a swivel wheel 12 which is rotatable and freely turns left and right. The housing 1 houses a circuit board 13 incorporating control circuit components such as a CPU, a ROM, and a RAM, and a battery 14 for supplying power to each unit.
[0011]
FIG. 3 is a block diagram showing the configuration of the control unit. Reference numeral 21 denotes a CPU constituting the control unit main body, reference numeral 22 denotes a ROM in which a program for controlling each unit by the CPU 21 is stored, and reference numeral 23 denotes a memory for storing various data. RAM. Reference numeral 24 denotes a motor control unit 25 for controlling rotation of the operation panel 2, the obstacle sensor 4, the cleaner motor 5, a motor control unit 26 for controlling rotation of the left and right traveling motors 10a and 10b, and the left and right encoders 11a and 11b. This is an I / O port that controls input and output of signals. The CPU 21, the ROM 22, the RAM 23, and the I / O port 24 are electrically connected via a bus line 28.
[0012]
FIG. 4 is a functional block diagram functionally showing the configuration of the control unit. This cleaning robot is functionally composed of the RAM 23, and the storage unit 31, which is backed up by the battery 14 for power supply, the CPU 21, The control unit 32 includes a complex of the ROM 22 and the I / O port 24.
[0013]
The storage unit 31 is provided with a map storage unit 311 for storing a map created for the work target area.
The control unit 33 calculates a moving amount and a moving direction from outputs of the left and right encoders 11a and 11b, and calculates a moving amount and a moving direction from the outputs of the left and right encoders 11a and 11b. Position estimating unit 322 for estimating the position, a cleaner control unit 323 for controlling the motor control unit 25, a cleaning target area is divided into cells, and an obstacle detected by the obstacle sensor 4 is detected in the cleaning target area during a cleaning operation. A map creating unit 324 is provided for creating a map by recording an object and a cleaned position in cell units.
[0014]
When cleaning starts, the cleaning robot first initializes a map in the map storage unit 311. Two maps are prepared: an attribute map MA1 shown in FIG. 5 and a potential map MA2 shown in FIG. These maps MA1 and MA2 divide the work target area into cells.
[0015]
The attribute map MA1 is a map for managing the work target area with attributes of an uncleaned area, a cleaned area, and a non-travelable area, and the corresponding cells are referred to as an unworked cell, a completed cell, and a non-travelable cell, respectively. The potential map MA2 is a map expressing a path length from a certain position to an arbitrary position in the work target area as a potential.
[0016]
When the cleaning robot initializes the map, the cleaning robot performs the traveling control shown in FIG. First, in S1, the traveling direction is adjusted. In this case, the left and right traveling motors 10a and 10b are driven and controlled so that the front turns in the traveling direction. Subsequently, in S2, the right and left traveling motors 10a and 10b are drive-controlled to travel straight. At this time, the cleaner motor 5 is simultaneously driven to clean the floor. Then, in S3, when the obstacle sensor 4 detects that there is an obstacle such as a wall in front, the obstacle sensor 4 moves to a position near the obstacle and then stops traveling.
[0017]
In this operation, the cleaning robot always estimates the self-position by the self-position estimating unit 322, detects the presence or absence of the front and left obstacles by the obstacle sensor 4, and displays the cleaned area and the obstacle in the attribute map MA1. The position of the object is recorded as a non-traveling area. Note that all of the attribute maps MA1 are uncleaned areas when initialized.
[0018]
When an obstacle is detected and stopped, subsequently, in S4, the closest uncleaned area is searched from the attribute map MA1. The search for the uncleaned area uses the potential map MA2. That is, as shown in FIG. 6, a potential map centering on the current position of the cleaning robot RC is created. Here, the potential of the current position of the cleaning robot RC is set to −1, and the potential is sequentially increased from “1” as the distance increases.
[0019]
In addition, for example, infinity 与 え る is given as a travel prohibition potential to a cell of the potential map MA2 corresponding to a position recorded as a travel impossible area due to an obstacle in the attribute map MA1. In addition, a margin is set so that the cleaning robot RC does not approach the travel impossible area too much. That is, the traveling prohibition potential is also applied to cells around the cell to which the traveling prohibition potential is applied.
[0020]
Therefore, it is assumed that the cleaned area recorded in the attribute map MA1 is an area M1 indicated by oblique lines in the figure, an unrunnable area is an area M2 indicated by a cross in the figure, and an uncleaned area is an area M3 indicated by white in the figure. , It is determined that an uncleaned area exists, and the closest uncleaned cell is searched for in the uncleaned area M3. The closest uncleaned cell is the cell having the smallest potential in the uncleaned area in the potential map MA2. In FIG. 6, the cell has a potential of “8”.
[0021]
After searching for the nearest uncleaned cell, subsequently, in S5, Point To Point travel control (hereinafter, referred to as PTP travel control) is performed. That is, the cleaning robot RC makes a plan to travel to the nearest cell in the uncleaned area M3 as the destination by the shortest route there. This plan utilizes the potential map MA2. That is, the shortest route can be obtained by calculating a route that reaches the current position by tracing the cell so that the potential decreases in order from the potential “8” at the destination point. Then, the vehicle travels in the reverse direction along the obtained route as shown by the arrow in FIG. Thereby, the cleaning robot RC can move to the nearest uncleaned cell in the uncleaned area M3 by the shortest path while bypassing the untravelable area M2.
[0022]
When the PTP running control in S5 ends, the process returns to S1 to adjust the traveling direction, and repeats straight running. Then, in S4, when it is determined that the uncleaned area M3 has disappeared, the series of cleaning control ends.
[0023]
For example, suppose that a rectangular room surrounded by a wall 41 and having a pair of obstacles 42 and 43 projecting from an opposing surface of the wall 41 is cleaned as shown in FIG. The X axis and the Y axis are defined for this room as shown in the figure.
[0024]
First, the cleaning robot RC adjusts the traveling direction to the positive direction of the Y axis at the position A at the lower left corner, travels straight to the position B, and performs cleaning. Eventually, an obstacle is detected ahead and stops at the position B.
[0025]
Then, the nearest uncleaned cell is searched. The closest uncleaned cell at this time is the cell on the right, so the cleaning robot RC turns 90 degrees to the right and moves to that cell. At this position, the traveling direction is adjusted in the negative direction of the Y axis, the vehicle travels straight toward the position C, and cleaning is performed. Eventually, an obstacle is detected ahead and stops at the C position.
[0026]
Then, the nearest uncleaned cell is searched. At this time, the closest uncleaned cell is the cell on the left, so the cleaning robot RC turns 90 degrees to the left and moves to that cell. At this position, the traveling direction is adjusted to the positive direction of the Y axis, the vehicle travels straight toward the position D, and cleaning is performed.
In this manner, in the work area in which the continuous uncleaned area continues, the cleaning robot RC performs cleaning while traveling in a general zigzag manner by the series of travel control, and also displays the cleaned area M1 and the disabled area in the attribute map MA1. The area M2 is recorded.
[0027]
When reaching the position D, the cleaning robot RC searches for the closest uncleaned cell, but there is no uncleaned cell adjacent to either the left or right at the position D. In this case, the position E is searched as the closest uncleaned cell. Then, the vehicle moves from the position D to the position E by the PTP traveling control.
[0028]
When reaching the position E, the cleaning robot RC adjusts its traveling direction and then goes straight ahead. When it detects an obstacle ahead, it searches for the nearest uncleaned cell and moves to that position by PTP traveling control. This series of operations is repeated to perform cleaning while running as indicated by the arrow in the figure. Then, when reaching the F position, the nearest uncleaned cell is searched for even at this position. Here, the uncleaned cells are present at equidistant positions on both the left and right sides. In order to cope with such a case, a default rule is determined in advance as to which of the X axis and the Y axis should be given priority. For example, when giving priority to the negative direction of the X-axis, the cleaning robot RC turns 90 degrees to the left and moves to the cell. At this position, the traveling direction is adjusted in the negative direction of the Y axis, the vehicle travels straight toward the position G, and cleaning is performed. Eventually, it detects an obstacle ahead and stops at the G position.
[0029]
Then, the nearest uncleaned cell is searched. At this time, the nearest uncleaned cell is located at the H position on the left side, so that the cleaning robot RC turns 90 degrees to the left and moves to the H position by the PTP traveling control. Align in the positive direction of the axis, go straight to position I and clean. When the vehicle reaches the position I in this way, all the work target areas are the cleaned area and the non-traveling area, so that there is no uncleaned area and the cleaning ends.
[0030]
FIG. 9 is a diagram showing a recording state of the attribute map MA1 when the room having the layout shown in FIG. 8 is cleaned. The whole cell-divided area is a work target area, and the entire area is initially an uncleaned area M3. . Then, the cleaning robot RC starts moving from the position A and records the cleaned area M1 while performing the cleaning, detects the obstacle with the obstacle sensor 4, and records it as the traveling impossible area M2.
[0031]
Therefore, when the cleaning robot RC reaches the H 'position before the I position, the recording state of the attribute map MA1 is as shown in FIG. At this time, since the front wall has not yet been detected by the obstacle sensor 4, all the areas shown in white are uncleaned areas M3. Thereafter, when the cleaning robot RC travels, the wall on the right side is detected as an obstacle, and when the wall in front of which the cleaning robot RC reaches the position I is detected as an obstacle, the periphery of the cleaned area M1 is detected. All are surrounded by the non-traveling area M2. At this point, the cleaning robot determines that it cannot move to the outside of the unrunnable area M2, and thus regards the uncleaned area M3 outside the unrunnable area M2 as the same as M2. As a result, it is determined that the uncleaned area has disappeared, and the cleaning is terminated.
[0032]
In this way, instead of creating a map of the work area in advance and then performing cleaning, a map is created while filling each cell of the work target area as a cleaned area and a non-travelable area while performing cleaning, and an uncleaned area is created. Since the cleaning is terminated when the battery runs out, the cleaning operation can be performed in a short time. In addition, since the cleaning is performed while always checking the uncleaned area, reliable cleaning can be performed without leaving the uncleaned area.
[0033]
Further, in order to move the cleaning robot RC to the nearest uncleaned cell, the potential map MA2 is used, and a path that reaches the position of the cleaning robot CR by tracing the cell so that the potential decreases in order from the potential at the destination point. Since the vehicle is calculated and travels in reverse of the calculated route, it is possible to move to the cell closest to the target uncleaned area by the shortest route, and thus to move in a short time.
[0034]
FIG. 10 shows a cleaning path when a U-shaped obstacle 44 is present in a square room surrounded by a wall 41. The X-axis and the Y-axis are also determined for this room as shown in the figure.
[0035]
First, the cleaning robot RC adjusts the traveling direction to the positive direction of the Y axis at the position A at the lower left corner, travels straight to the position B, and performs cleaning. Eventually, an obstacle is detected ahead and stops at the position B. Then, the cell next to the right is searched for as the closest uncleaned cell, and the cleaning robot RC turns 90 degrees to the right and moves to that cell. At this position, the traveling direction is adjusted in the negative direction of the Y axis, the vehicle travels straight toward the position C, and cleaning is performed. Eventually, an obstacle is detected ahead and stops at the C position.
[0036]
Then, the cell on the left is searched for as the closest uncleaned cell, and the cleaning robot RC turns 90 degrees to the left and moves to that cell. At this position, the traveling direction is adjusted to the positive direction of the Y axis, the vehicle travels straight, and cleaning is performed. Then, since the obstacle sensor 4 detects the obstacle 44 at the position D, the cleaning robot RC stops.
[0037]
Here, the cell on the right side is searched for as the closest uncleaned cell, and the cleaning robot RC turns 90 degrees to the right and moves to that cell. At this position, the traveling direction is adjusted to the negative direction of the Y axis, the vehicle travels straight, cleaning is performed, the wall 41 is detected ahead, and the vehicle stops at the E position.
In this way, when the cleaning robot RC detects an obstacle ahead, it searches for the nearest uncleaned cell and performs cleaning while traveling by the PTP travel control, and also displays the cleaned area M1 and the non-travelable area M2 in the attribute map MA1. Record
[0038]
When the cleaning robot RC eventually reaches the position F, it searches for the nearest uncleaned cell also at this position. In this case, the uncleaned cells exist at equidistant positions on both the left and right sides. For example, if priority is given to the negative direction of the X axis, the cleaning robot RC turns 90 degrees to the left and moves to the cell. At this position, the traveling direction is adjusted to the negative direction of the Y axis, the vehicle travels straight, cleaning is performed, the obstacle 44 is detected ahead, and the vehicle stops at the position G.
[0039]
Thereafter, the cleaning robot RC performs cleaning while traveling, and records the cleaned area M1 and the non-travelable area M2 in the attribute map MA1. Then, when reaching the H position, there are no uncleaned cells on both left and right sides thereof, and the cleaning robot RC searches the I position as the closest uncleaned cell, and moves to that position by the PTP traveling control.
[0040]
The cleaning robot RC that has reached the position I travels straight in the negative direction of the Y axis, performs cleaning, detects the obstacle 44 ahead, and stops at the position J. The closest uncleaned cell at the J position is searched, but the K position is searched as the closest uncleaned cell that is not immediately nearby. The cleaning robot RC moves from the J position to the K position by PTP traveling control.
[0041]
When it reaches the K position, it turns 90 degrees to the right, adjusts the traveling direction to the negative direction of the Y axis, and then goes straight ahead to perform cleaning. Eventually, the wall 41 is detected ahead and stops at the L position. Then, the nearest uncleaned cell is searched, but the remaining uncleaned area is only in the vicinity of the position M. The cleaning robot RC reverses the direction and travels straight in the positive direction of the Y axis to perform cleaning. When the obstacle sensor 4 detects the wall 41 at the position M, the cleaning robot RC stops. When the robot reaches the M position, all the work target areas are the cleaned area and the non-traveling area, so that there is no uncleaned area and the cleaning ends.
[0042]
Even if the layout of the room to be cleaned is complicated in this way, reliable cleaning can be performed in a short time.
This embodiment has been described with reference to the case where the present invention is applied to a cleaning robot. However, the present invention is not limited to this, and can also be applied to a robot that performs other work than cleaning while traveling.
[0043]
【The invention's effect】
As described in detail above, according to the present invention, it is possible to provide an autonomous mobile robot capable of performing a reliable work in a short time in a work area without storing a map of the work area in advance.
[Brief description of the drawings]
FIG. 1 is a front view showing an external configuration of a cleaning robot according to an embodiment of the present invention.
FIG. 2 is a partially cutaway side view showing the internal configuration of the cleaning robot according to the embodiment.
FIG. 3 is a block diagram showing a hardware configuration of a control unit in the embodiment.
FIG. 4 is a functional block diagram functionally showing a configuration of a control unit in the embodiment.
FIG. 5 is a view showing an example of an attribute map formed in a map storage unit according to the embodiment;
FIG. 6 is a diagram showing an example of a potential map formed in a map storage unit according to the embodiment.
FIG. 7 is a flowchart showing traveling control of the cleaning robot in the embodiment.
FIG. 8 is a view showing an example of cleaning travel by the cleaning robot in the embodiment.
FIG. 9 is a diagram showing a recording state of an attribute map in the cleaning traveling example of FIG. 8;
FIG. 10 is a view showing another example of the cleaning traveling by the cleaning robot in the embodiment.
[Explanation of symbols]
4 Obstacle sensor, 5 Cleaner motor, 9a, 9b Driving wheel, 10a, 10b Running motor, 11a, 11b Rotary encoder, 21 CPU, 22 ROM, 23 RAM, 311 Map storage unit , 321: travel control unit, 322: self-position estimation unit, 323: cleaner control unit, 324: map creation unit.

Claims (4)

An autonomous traveling robot that includes a traveling unit and a working unit, and performs an operation with the working unit while autonomously traveling with the traveling unit.
Obstacle detecting means for detecting an obstacle present in the work target area, and the work target area is divided into cells, and at the time of work, the obstacle detected by the obstacle detect means in the work target area and the work means Map creating means for creating a map by recording the position where work has been completed in cell units, a recognizing means for recognizing the closest unworked cell in the map from a current position at the time of work, and the obstacle detecting means An autonomous mobile robot, comprising: a control means for controlling the travel of the traveling means to an unworked cell recognized by the recognition means when the obstacle is detected ahead of the traveling direction. The map creating means records the obstacle position in the work target area as an unworkable cell, records the worked position as a worked cell, and the rest as unworked cells. 2. The autonomous mobile robot according to claim 1, wherein the running means controls the running of the running means to a non-working cell so as to avoid a cell. The control means sets the worked cell and the unworked cell as runnable cells, and gives a higher potential to the runnable cell as the cell travels longer from the current position, and has the lowest potential among the unworked cells. 3. The autonomous mobile robot according to claim 2, wherein a detected cell is detected, and the running means is controlled to travel to the cell. 4. The control unit according to claim 3, wherein the control unit plans a route in which the potential is sequentially reduced to a cell having the lowest potential among the unworked cells, and performs traveling control so as to follow the route in reverse order. An autonomous mobile robot.

Images