JPH01149114A - Running control method for free-running robot - Google Patents

Abstract

PURPOSE:To easily prevent a robot from being not able to run by preliminarily calculating coordinates of a position, to which the robot should be returned by U-turn or the like, in accordance with position coordinates of the robot itself and position data of an obstacle and returning the robot to this position when the robot cannot run to an uncleaned area. CONSTITUTION:With respect to the running control of the self-running robot, obstacle position data detected by an obstacle sensor using an ultrasonic wave is stored in a storage device and position coordinates of the robot itself are calculated in a central processing unit which controls and commands the robot, in accordance with data of an encoder which measures the rotational frequency of wheels and a gyro which measures the change of the direction of the robot. If the free-running robot cannot U-turn neither move by judgement of the running method and commands, a position where the robot can turn before entering into a recessed part is operated in accordance with said calculated position coordinates of the robot itself. As the result, the robot returns to this position, where the robot can turn before entering into the recessed part or the like, by simple running like backward or forward straight running, thus controlling the running.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a self-propelled robot, and is intended to prevent the robot from stopping in its running due to an error in recognition of objects such as walls in a room or obstacles. This is related to a control method for a self-propelled robot that can prevent such problems ◇ [Conventional technology] What about a mobile robot that cleans the entire floor while detecting obstacles using conventional ultrasonic waves? The ffl control device uses ultrasonic V as described in Japanese Patent Laid-Open No. 60-95522.
An obstacle sensor detects walls, pillars, etc. by multiplying The driving side that returned (was doing 111 tl).

[Problem that the invention seeks to solve]

In the above-mentioned conventional technology, when the obstacle has a simple shape such as a square or a rectangle, as shown in FIG.
Slightly rough obstacle 9 For example, in the case of a square obstacle shown in Fig. 8 of the present invention, which has 100 and 101 protrusions, the obstacle sensor detects the obstacle in front. The robot runs in a straight line until it reaches the square obstacle 102, but the robot cannot make a U-turn at the 9th ledge (#1o1) and stops running. It enters 104 and becomes stuck at 105. Therefore, in the conventional technology, consideration is not given to how to deal with obstacles such as a U-turn caused by a protruding part of an obstacle, or how to escape when a robot enters a recessed part. In the ninth part of the class, there was a problem where the mobile robot became unable to move.

SUMMARY OF THE INVENTION An object of the present invention is to smoothly escape from the situation where a mobile robot cannot move due to a protrusion or recess of the obstacle, so that the mobile robot can continue running.

[Elaborate means to solve problems]

The above purpose is to detect an obstacle using an ultrasonic sensor and store the obstacle position data in a storage device. If a corner turn cannot be made, the central processing unit that issues control commands to the robot will guide the robot to avoid the protrusion based on the situation of the obstacle protrusion or recess in the storage device and the robot's own position coordinates. This is achieved by adopting a travel control method that allows the vehicle to turn back by simply traveling in a straight line, such as backwards or forwards, up to a point 11 where it can turn, or to a position where it can turn before entering a recess.

[Effect]

The traveling control method for a self-propelled robot according to the present invention is based on a central y! which stores obstacle position data detected by an obstacle sensor based on Choon Saka in a storage device and issues control commands to the robot. ,
31 device calculates the robot's self-position coordinates from data from an encoder that measures the number of rotations of the wheels and a gyro that measures changes in the robot's orientation. When the self-propelled robot determines and gives instructions on how to travel in the central processing unit, the robot can make a U-turn or other turn based on the information on the protrusion or depression of the obstacle in the storage device.
If the robot becomes unable to move, the robot'fr is determined based on the situation of the obstacle protrusion or recess in the storage device and the robot's self-position coordinates calculated above.
: The central processing unit calculates the positional coordinates at which the vehicle can turn while avoiding the protrusion of the obstacle or the position 'gA' at which the vehicle can turn before entering the recess. However, driving to a position where it is possible to turn is limited to simple driving methods such as straight-line driving backwards or forwards.
To prevent escaping from a protrusion or recess of an obstacle from becoming complicated. And in the central processing unit,
Calculate it yourself and set the position above to the seated position where you can turn while avoiding fc obstacles or the position coordinates where you can turn before entering a dent. The driving condition is eat-.

As a result, even if the self-propelled robot is unable to turn due to the protrusion of an obstacle and enters the recess, the self-propelled robot can smoothly escape from the MjFfr and continue running.
I can continue.

C English Example] Hereinafter, one embodiment of the present invention will be described with reference to the drawings using an example of a self-propelled cleaning robot for the purpose of cleaning.

Figure 3 is a perspective view showing the configuration of a self-propelled cleaning robot.This robot approaches the wall in the manner of a crab walking sideways (hereinafter referred to as sideways walking) in order to eliminate the residue left behind when cleaning near the wall. The next step is to create a self-propelled robot. In Fig. 3, 51 is the left wheel, 52 is the left wheel drive motor, and 53 is the left wheel drive motor.
is a right wheel, 54 is a right wheel drive motor, 55 is a wheel turning motor for turning the left and right wheels in the t-900 direction, 56
is ultrasonic [sending and receiving! 57 is a rotating disk, 58 is an ultrasonic parabolic antenna, 59 is an ultrasonic transmitter/receiver 56 and a parabolic antenna 58 are fully installed, and the ultrasonic wave generator rotates the rotating disk 57 counterclockwise in the direction of the arrow 57α. radar rotation motor,
6o is an ultra-f wave radar encoder that measures the number of rotations of the rotating disk 57, 61 is a gyro that measures the angular change in the direction of the self-propelled cleaning robot, 62 is a vacuum cleaner, 63 is a vacuum cleaner motor, 64 65 is a cleaning dust suction port, 65 is a measuring circuit unit, and 66 is a unit that calculates the robot's own position coordinates, calculates the position coordinates of obstacles detected by the ultrasonic radar of the ultrasonic transmitter/receiver and the parabolic antenna, and calculates the obstacle information. 67 is a control power source; 68 is a driving X source; 69 is the robot body foot; 70 is a body frame 69'ff; :1
Casters 71 that hold the wheel horizontally are robot bodies.

FIG. 4 is a system block diagram showing mH of the travel control system in FIG. 3. In FIG. 4, numerals 52 to 67 correspond to those in FIG. 6 described above.

76 is a central processing unit (CPLI) which is a component of the travel control unit 66; 74 is a central processing unit 76 that displays calculation results such as the robot's self-position, the position data of the camp book detected by the ultrasonic radar, the cleaned range, etc. a storage device that stores
75.76 will be 1t5i later in Figure 6! These are a left wheel encoder and a right wheel encoder that measure the rotational speed of the left and right wheels. 77 has the wheels oriented 9 for sideways running.
A turning angle switch 78 detects whether the vehicle has turned 90 degrees when the vehicle is turned by 0'', and a wheel return switch 78 is used to produce the angle when the wheels return to their original orientation after traveling sideways.

Figures 5 and 6 show a wheel drive mechanism in which only the wheels turn 90 degrees and return to the original direction without changing the direction of the robot 10 body in order to eliminate the residue left behind when cleaning near walls. In FIG. 5, 51 is the left wheel drive motor for the left wheel 52, 56 is the right wheel, 54 is the right wheel drive motor, 55 is a one-wheel turning motor for turning the left and right wheels by a full 90', and 80 is a motor for driving the left wheel. a bevel gear attached to the wheel turning motor 55;
81 and 82 are worms, 83 is a worm shaft, and 84 is a worm Idl18! 85 is a left wheel drive bevel gear, 86 is a right wheel drive bevel gear, 77 is the 900 turning angle switch, 78 is the wheel return switch, 87 is a switch that operates switches 78 and 78. The cam 69 is a KtI robot body frame to which these parts are fixed.

FIG. 6 is a cross section taken along line AA in FIG. 5, and although this is the left wheel, it has the same structure as the right wheel. In FIG. 6, 51 is the left wheel;
52 is the left wheel drive motor, 69 is the robot body frame, 81 and 83 are the worms and the worm shaft, 85 is the bevel gear, 75 is the left wheel encoder, 87 is the switch cam, and 89 is fixed to the body frame 89. Encoder holder, 88 is the pivot axis, 88α is the precision of the pivot axis, 90
91 and 92 are bevel gears fixed above and below the wheel drive shaft 90, 95 is a foil gear meshing with the worm 81, and 94 rotatably supports a rotation axis 88α fixed to the main body frame. The bearing 95 is a bevel gear on the left wheel shaft that meshes with the bevel gear 92.

Next, the operation will be explained.

The self-propelled cleaning robot in this example is a robot that thoroughly cleans the room by moving straight and making U-turns while avoiding walls and obstacles in the room. In addition, this self-propelled field removal robot is designed to eliminate unfinished cleaning of areas near walls.
The wheels are turned 90 degrees to approach the top of the robot. The robot travels by driving the left wheel 51 and the stone wheel 53 by interlocking the left wheel drive motor 52 and the right wheel drive motor. The upper part of the robot is provided with an ultrasonic transmitter/receiver 56 and a parabolic antenna 58 as shown in FIG.
9, it is rotating in the counterclockwise direction 57α.

The parabolic antenna 58 receives the ultrasonic waves tIf1. ``lr: Converts into a highly directional ultrasonic wave (dashed line 72) and emits it around j+7 ultrasonic waves in the counterclockwise direction mentioned above.The emitted ultrasonic waves are reflected when they hit the walls or obstacles in the room. The reflected ultrasonic waves return to the parabolic antenna 58 and are transmitted to the ultrasonic transceiver 56.
received at The directions of ultrasonic emission and reception from the parabolic antenna 58 are measured by an ultrasonic radar encoder 60. These ultrasonic transceiver 56 and parabolic antenna 5
8 and the ultrasonic radar Ecarda 60, the positions of walls and obstacles in the room can be measured from the time from when the ultrasonic waves are emitted to when they are received, and from the directions of the ultrasonic waves being emitted and received. Ru. A garbage suction port 64 is connected to the vacuum cleaner 62,
As the robot moves, Jl) Absorbs dust from the width of the robot body. Indicated by CPLI in FIG. 4, the central processing unit 7
3 is the work of calculating the self-position coordinates of the self-propelled cleaning robot, calculating the positions of walls and obstacles in surrounding rooms, and storing the position data of these walls and obstacles in the storage device 74;
The coordinate range of the cleaning garbage inlet 64 is calculated in the storage device, the judgment of the driving method to avoid surrounding walls and obstacles, and the drive control of the left and right wheels are performed. I do.

The self-position coordinates are based on wheel rotation speed data from left and right wheel encoders 75, 76, wheel travel distance, and gyro 6.
Data on the angle change amount of the robot's orientation in step 1 is Q, the travel start position is the coordinate origin, and the robot's orientation at the start position is f:X
It is calculated by accumulating one after another as the Y-axis direction of the Y coordinate. The position coordinates of the wall or obstacle are determined by the distance and direction data to the wall or obstacle measured by the ultrasonic transmitter/receiver 56, parabolic antenna 58, and radar encoder 60, and the self-position image of the robot. Calculate by adding up. The position coordinates of walls and obstacles are stored in r11@74 as described in 5 above.

The traveling control method performed by the central processing unit 73 in the self-propelled cleaning robot of the embodiment is as shown in FIG. 10, as shown in FIG. If this is not possible, the wheels can be turned 90 degrees as shown in Figures 5 and 6 to approach the wall sideways.

FIG. 10 is an example of a travel trajectory when there is an obstacle 106 in a vertically long rectangular room. To set the coordinates, as described above, the traveling start point of the robot is set as the coordinate origin, and the direction of the robot at the traveling start point is set in the positive direction of Yltl.

The self-propelled cleaning robot moves from the starting point 1 in Figure 10.
From 07 to 108, I drove straight and U-turn repeatedly, and at 108, I couldn't make a U-turn because of the wall on the left side of the room, so from 108 to 109, I ran sideways (the first horizontal run) and hit the wall. bring them closer.

Here we will explain how to approach a wall by running sideways.

As shown in FIG. 7, for lateral travel, without changing the direction of the robot body 1α, the left wheel 51 (L and right wheel 52a) are rotated by 90 degrees to form 51b and 52b, and the left wheel 51 (L) and right wheel 52a are turned 51b and 52b from the robot body 71α to the left wall. According to the distance 111LaK to 97, the travel distance Lb is the same, and the left and right i is indicated by the arrows 51e and 5.
Drive the vehicle in the direction of 2e, and then move the wheel 5 toward the wall.
1C and wheel 52c are turned 90 degrees to the left to return 5jd and 52d to their original positions.

The operation of changing the direction of the wheels by 90 degrees is performed by driving the wheel turning motor 55 shown in FIG. When the wheel turning motor 55 rotates, the worm l1
The II 83 rotates, and the left and right worms 81 and 82 rotate at the same time. When the worm 81 rotates, the wheel gear 93 meshing with the worm 81 shown in FIG. Axis center CCC narrow axis rotation. Whether the wheel has turned 90 degrees is detected by the rotation angle switch 77 shown in FIG. 5 based on the rotation of a switch cam 87 fixed to the rotation axis 88α. Conversely, whether the wheel has returned to its original state is similarly detected by the wheel return switch 78 based on the reverse rotation of the switch cam 87. By the above operation, only the direction of the wheel is turned by 90 degrees and the operation is made to return to the original direction.

Returning to the explanation of the running trajectory in FIG. 10, after approaching the wall by running sideways, the robot retreats along the left wall from 109 to 110 and cleans near the wall. From 110 to 111, the robot is brought close to the wall during the second horizontal run. This self-propelled cleaning robot performs cleaning twice in order to overlap the cleaning width of the garbage suction port and eliminate cleaning residue during parallel and backward travel. From 111 to 112, go straight ahead again along the left wall. Now that the cleaning near the wall has been completed, the next step is to move to the uncleaned area 117, which has not yet been cleaned.

In order to move to an area that has not been cleaned, the direction of the robot must be changed. In the example of FIG. 10, the robot body 115 indicating the outer shape of the robot body is too close to the left wall of the room at 112 after cleaning near the wall, so it cannot turn. Therefore, in the self-propelled cleaning robot of the embodiment, from 112 to 114, the lateral movement is used to move the robot away from the wall (hereinafter referred to as escape lateral movement). Next, at 114, robots) 1
Turn the car 90 degrees to the left and point it in the direction of 115, then drive straight until it reaches 115. Further 115 to 116,1
The vehicle is caused to travel to the unreturned area 117 on route No. 18.

However, an error of 10 to 20 degrees cannot be avoided in the position data of the obstacle detected by the ultrasonic transmitter/receiver 56, the laboratory antenna 58, and the ultrasonic radar encoder 60 shown in FIG.

Therefore, the dots in Fig. 11 are obstacle position data detected by the ultrasonic transmitter/receiver and parabolic antenna while the robot was traveling straight and repeatedly making U-turns from t-107 to t-112, and traveling sideways to the wall. 119 and 12 in the diagram
An error occurs in the obstacle position data as shown by the zero point.

This obstacle position data 120 is in the same situation as if there was a protrusion of the obstacle at that position. Therefore, the first
Escape and lateral travel from 112 to 114 in Figure 0 may not be possible.

Furthermore, as actually shown in FIG. 12, if there is a protrusion 121 on the front wall of the room, the obstacle position data shown at 122 in FIG. 13 is detected, and because of the obstacle position data 122, Only the horizontal escape runs from 112 to 114 shown in FIG. 10 are possible.

The self-propelled robot traveling control method of the present invention is thus:
This is a control method that allows the robot to smoothly escape from the situation when it becomes unable to move due to data on the protrusion of an obstacle, and in this example, to continue traveling to the uncleaned area. be.

The traveling control method will be explained with reference to FIGS. 1, 2, and 14. FIG. 1 is a flowchart of control performed when the robot cannot move due to a protrusion or depression of an obstacle. FIG. 2 shows a method of controlling the self-propelled cleaning robot according to the embodiment to move it close to a wall by going straight, making a U-turn, and moving sideways, and the first method.
2 is a flowchart showing the necessary timing and relationship of the control shown in FIG.

FIG. 14 is a plan view showing 71 cll harmful object position data detected by ultrasonic waves and a specific method of escaping from the position where the robot is stuck. In FIG. 14, 112 is W shown in FIG. 10, and the robot's self-position coordinates (x, y) after the last cleaning is completed. 113 is the outer shape of the robot body at the robot position 112, 122 is the obstacle position data of the protrusion 121 in FIGS. Wall position data 125 is the position data of the left wall of the room. The self-propelled cleaning robot is in the position shown in Figure 14! 112, and it becomes impossible to escape and run sideways in the right direction at the tth point in the obstacle position data 122,
That is, when the judgment 3 in FIG. 1 that there is an l1lI harmful substance on the escape side is YES, first, in processes 4 to 14, the XY coordinates of the point at which the vehicle should turn back until it is possible to escape sideways are calculated.

The method is 4 in Figure 1 to determine the backward travel distance! Set to 5 subdivisions in total. The self-position coordinates of the nine robots that have been moved back over a distance are the coordinates t-(X, Y
), then in process 8, the Y coordinate becomes YBteY+ノ, and the X coordinate becomes (X, YB) of X.

The reason for proceeding to process 8 is that the direction of movement of the robot at the self-position 112 is the negative direction of the Y axis (arrow 126).
The process proceeds to friend processing 8 in which the determination as to whether the traveling direction of No. 5 is the Y-axis is ①S, and the determination as to whether No. 6 is the positive direction of the Y-axis is NO. Next, judgment 12
Then, it is determined whether there is any obstacle data on the escape side of the robot position coordinates (XYB).

The judgment is made by retracting the distance by 1 and the above coordinates (X
, YB), the robot body moves a distance Lm to the right by which it can turn by 90 degrees, and it is determined whether there is obstacle data at the outer shape 119 (shaded area αbcct) of the robot body at nine points 118.

Since there is obstacle data 122 within the robot outline 119, the determination in step 12 is NO, and the process proceeds to step 14.

In process 14, 5crn is added to the retreat distance 1, and the robot position coordinates (X, YB) t- are calculated again with Y coordinate f, YB=Y-)4, and again in process 12, the position (XY
Determine whether there is an obstacle on the escape side of B). Similarly below,
Repeat this process by adding 5 cm to the backward travel distance @1 until there are no obstacles in the robot body on the escape side. Then, based on the judgment of 12, obstacle data 12 was placed on the escape side.
The robot position where 2 disappears is 127 in FIG. The robot body 128 moves to the right from 127 by a distance that allows it to turn by 90 degrees, and the position after the escape horizontal run is 129, and the external shape of the robot body at 129 is 130 (efgh surrounded by a broken line).

The calculation of the XY sitting image of the point where Takumi turns back until he can escape and run sideways from Process 4 above is when the robot's advancing direction is the negative direction of the Y axis, the above XY coordinate (XYB) t - the above backward traveling distance J' k S cm was added in increments and calculated in step 8, YB=Y-)4, but the direction of movement of the robot was Y
If the axis is in the positive direction, XYJi[0YJ
l[fYBt-1 processing 7 tv YB=Y-14te calculation.

On the other hand, when the direction of movement of the robot is the X-axis, the coordinates (XB, Y) of the point of return are as follows: Calculate by XB=X-4, and if it is in the negative direction, calculate by XB=X-14 in process 11.

In processes 5 to 14, if the traveling direction in FIG. 14 is the negative direction of Y@, the XX coordinates of the point at which you should return until you can escape sideways are calculated using the coordinates (X, YB) of point 127. Next, in process 15 in FIG. 1, the central processing unit rIL73 (
CPLI) C, command the wheel drive 11h motor II 52.54 to retreat for escape. 112 to 1 in Figure 14
During escape and retreat up to 27, the decision in step 1 in FIG. 1 is YES, processes 2 to 15 are skipped, and the next process for connector (2) is performed. If there is no obstacle behind the robot, the X coordinate of the robot's self-position coordinates (x, y) is determined by judgment 17.18 in FIG.
Move backward until it is within IIMt of the Y coordinate YB of B).

If the direction of movement of the robot is Move back until you are within 1 cm.

In judgment 17 in FIG. 1, the robot's self-position coordinate (X).

When the X coordinate of Y) is within one position of the X coordinate of the coordinate 127 (X, YB) and reaches the point 127 in FIG. , issues a stop command to the wheel driver. Next, in step 21, a flag indicating that escape and retreat has ended is reset. Accordingly, the judgment of 1 in Figure 1 whether the escape is in retreat is NO (J and N, and the judgment of 3 whether there is obstacle data on the escape sideways side is NO at point 127) In the next process 22 of connector (2), an escape lateral run command is issued. In response to the command for escape lateral movement in process 22, the robot performs lateral movement in which only the wheels turn 90 degrees (Fig. 7) by driving the wheel turning motors shown in Figs. Figure 1
From 27 to 129, perform an escape lateral run away from the wall.

Subsequently, if the escape lateral run away from the wall is completed as determined at 23 in FIG. 1, the escape run flag is reset at step 24, and the wall run end flag is set at step 25.

The traveling control in FIG. 1 is the details of the process 46 in FIG. 2, and 25.24.25 in FIG. 1 is 47°4 in FIG.
It corresponds to 8 and 49.

Since the wall running end flag is set at 25 in FIG. 1, the determination at 31 in FIG. 2 is YES. Next, in the subroutine 45, a search for an unremoved area and a search for a running route are performed. Subsequently, in the second (9) step 37, the vehicle travels to the uncleaned area.

In FIG. 14, when traveling to an uncleaned area, the direction of the robot is changed from arrow 131 to 13 at point 129 after the escape horizontal travel.
2, turn 90 degrees to the left, go straight on the travel route to the uncleaned area 17 in FIG. 10 to 115, and then travel as 11t5 and 11B.

According to this embodiment, as shown in FIG. 11, due to an error in the obstacle position data detected by the ultrasonic transceiver and the parabolic antenna, the situation is the same as if there was a protrusion of the obstacle at 120. As a result, it becomes impossible to escape sideways to drive to an uncleaned area and then turn.
If the robot becomes unable to move, or if the protrusion 121 actually enters the recess as shown in Figure 12, the robot will no longer be able to escape and run sideways to reach the area that has not been cleaned. In this case, the central processing unit calculates in advance the XX coordinates of the point at which it turned back until it can escape sideways, and the robot returns to that point by reversing, so the robot can use data on protrusions and depressions of obstacles. It has the effect of allowing you to smoothly escape from a stuck situation and continue running to an area that has not been cleaned.

In addition, in the method of calculating the XX coordinates of the point where the robot turned back until it was able to escape sideways in this example, as explained in steps 5 to 11 in FIG. L (In the example in Figure 14, the robot's direction of movement is parallel to Ylll, so the coordinates of the point of return are:
The X coordinate remains the same as the robot's self-position coordinates (x, y) before turning back in 112, and only the X coordinate is changed to ■), and the situation is such that it cannot move even if it runs simply backwards. This has the effect of preventing the robot from colliding with the walls or obstacles in the room.

Describing the problems of the prior art, in the example shown in FIG. 8, if the self-propelled cleaning robot control method of the present invention is applied, the robot will move straight when the protrusion 101 hits the obstacle 100, as shown in FIG. When the robot reaches point 102 by repeating U-turns, the direction of movement of the robot is parallel to the Y axis, so
Retract downward parallel to the Y axis from 2 to 133, and 103
You can then make a U-turn to the left at 154 and continue driving from 134 onwards. Furthermore, as shown in FIG. 16, if an obstacle 103 has a dent [104] and the robot enters the dent 105, if the control method of the present invention is applied, the direction of movement of the robot can be changed to Y. Since it is an axis, 105 to 136
It is possible to retreat downward parallel to the Y axis until 138, make a U-turn at 138, and continue driving after 138. However, if you are unable to turn due to the protruding part of the obstacle in the prior art,
Even if the robot falls into a dent, it can smoothly escape from the spot and continue running.

〔Effect of the invention〕

According to the present invention, the height of the obstacle is detected using ultrasonic waves, and data on the protrusion or recess of the obstacle is used to enable the robot to make a U-turn and escape from the wall as described in this embodiment. Even if the robot is unable to run to an uncleaned area, the robot calculates in advance the coordinates of a return point where it can make a U-turn or run sideways from the robot's own position coordinates and the position data of obstacles, and then retreats to that point. By making the robot run back, the robot can smoothly escape from a stuck situation, which has the effect of allowing it to continue traveling to areas that have not been cleaned.

Further, according to the present invention, the X
Change the coordinate of Y seat w in the robot's direction of travel (Y coordinate if the robot's traveling direction is parallel to Y@, X seat 411 if parallel to Since the robot can escape from a stuck situation by running, it has the effect of preventing the robot from colliding with the walls of the MB store or obstacles.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing a method of controlling a self-propelled cleaning robot according to an embodiment of the present invention, and FIG.
A flowchart showing a control method for turning and moving close to a wall; FIG. 3 is a perspective view showing the overall configuration of the self-propelled cleaning robot of the embodiment;
Figure 4 is a system block diagram of the travel control system for a self-propelled 10-wheel drive vehicle, Figures 5 and 6, and Figure 7 are plane and cross-sectional views showing the wheel drive mechanism for lateral travel, and Figure 8. , FIG. 9 is an explanatory diagram of a driving example of the prior art, and FIGS. 10, 11, and 12. FIGS. 13, 14, 15, and 16 are plan views showing examples of how the self-propelled cleaning robot moves. 51...Left wheel, 52...Left wheel drive motor, 53
...Right wheel, 54...Right single corridor movement motor, 56...
・Ultrasonic transceiver, 5th...Parabolic antenna, 60th
... Ultrasonic radar encoder, 61 ... Gyro, 65 ... Measurement circuit section, 66 ... Travel control section, 73
...Central processing unit, 74...Storage device, 75...
Left wheel encoder, 76... Right wheel encoder.

Claims (1)

[Claims] 1. A wheel and a wheel drive device, a position measuring device that measures the robot's self-position, an obstacle sensor that uses ultrasonic waves to detect surrounding obstacles, calculation of the self-position, and position coordinates of the obstacle. A self-propelled robot is equipped with a central processing unit that determines the calculation and running method and instructs the single-wheel drive unit to run, and a memory unit that stores information about walls and obstacles in the room. If the robot is unable to move due to an obstacle protrusion or recess on the coordinates of the obstacle data obtained by the obstacle sensor, the robot can move only in the direction of movement at the position where it is unable to move, such as turning, etc. A travel control method for a self-propelled robot, characterized by performing travel control to return to a point where it can move.