KR20060081131A - Cleaning method of robot for cleaner - Google Patents

Abstract

The present invention relates to a cleaning robot and a cleaning method using the same, the cleaning pattern on the top, the cleaning time is operated by the operation button 111, and remotely by the remote control unit 113 to operate the operation state and operation information display unit 112 And a bumper 120 having a shock absorbing port 121 for shock absorbing the front surface and an actuating opening 122 for transmitting a shock, and a driving motor in the lower center. The inlet 143 in which the driving wheel 131 is formed by the driving of the driving wheel 131 and the driving unit 130 is driven, and the suction motor 144 is formed while the brush 141 is rotated by the lower brush motor 142. A dust collecting part 140 for collecting dust and collecting dust, and a driving cleaner 100 having a power supply part 150 for supplying power to the rear, and an operation port of the bumper 120 of the driving cleaner 100. The sensing plate 210 is energized in a thin film form so as to be interlocked with the 122. It is formed so as to detect the obstacle contact to the front, and install the infrared sensor 220 in the bumper 120 radially, the infrared sensor 220 is installed densely installed the installation intervals of both sides than the installation interval of the center of the front To install a plurality of floors facing the ground on the lower front to form a floor detection sensor 230 for detecting the obstacle terrain, the encoder for measuring the speed of the driving wheel 131 rotated by the driving motor 132 in the driving unit ( 240 and the sensor unit 200 and the wheel lifting sensor 250 for detecting the lifting of the wheel, and calculated according to the optimal driving pattern according to the information detected by the sensor unit 200 to the controller 310 It characterized in that it comprises a central processing unit 300 for cleaning while generating a signal to control the running of the traveling cleaning unit (100).

Cleaning Robot, Random Bounce Pattern, Region Pattern, Wall Riding Pattern, Composite Pattern

Description

Cleaning robot and cleaning method using same {Cleaning Method of Robot for Cleaner}

1 is a perspective view showing a robot for cleaning according to the present invention.

Figure 2 is a block diagram showing the control structure of the cleaning robot according to the present invention.

3 is a process chart showing a cleaning method of a robot for cleaning according to the present invention.

Figure 4 is a block diagram showing the infrared sensor position of the cleaning robot according to the present invention.

Figure 5 is a block diagram showing the position of the floor sensor of the cleaning robot according to the present invention.

Figure 6 is a block diagram showing the position where the bumper of the cleaning robot according to the invention is detected.

7 is a conceptual view showing a cleaning pattern in a cleaning method of a general cleaning robot.

8 is a conceptual diagram showing a random bounce pattern in the cleaning method of the robot for cleaning according to the present invention.

Figure 9 is a conceptual diagram showing the principle of the random bounce pattern in the cleaning method of the robot for cleaning according to the present invention.

10 is a conceptual diagram showing a zigzag pattern in the cleaning method of the robot for cleaning according to the present invention.

11 is a conceptual diagram showing the principle of the zigzag pattern in the cleaning method of the cleaning robot according to the present invention.

12 is a conceptual view showing a wave pattern in the cleaning method of the robot for cleaning according to the present invention.

13 is a conceptual diagram showing a avoidance pattern of corners and bottlenecks in a wave pattern in the cleaning method of the robot for cleaning according to the present invention.

14 is a conceptual view showing a whirlwind pattern in a cleaning method of a robot for cleaning according to the present invention;

15 is a conceptual diagram showing a square rounding pattern in the cleaning method of the robot for cleaning according to the present invention.

16 is a conceptual diagram showing an obstacle avoidance pattern of the square rounding pattern in the cleaning method of the robot for cleaning according to the present invention.

Figure 17 is a conceptual diagram showing the principle of the star-shaped pattern in the cleaning method of the robot for cleaning in accordance with the present invention.

18 is a conceptual view showing a star pattern in the cleaning method of the robot for cleaning according to the present invention.

19 is a conceptual diagram showing the principle of a circular rounding pattern in the cleaning method of the robot for cleaning according to the present invention.

20 is a conceptual diagram showing a circular rounding pattern in the cleaning method of the robot for cleaning according to the present invention.

21 is a conceptual diagram showing the obstacle avoidance pattern of the circular rounding pattern in the cleaning method of the robot for cleaning according to the present invention.

22 is a conceptual view showing a wall riding pattern in the cleaning method of the robot for cleaning according to the present invention.

Figure 23 is a block diagram showing the operation sequence according to the condition of the random bounce pattern in the cleaning method of the robot for cleaning according to the present invention.

24 is a block diagram showing the operation sequence according to the conditions of the wave pattern in the cleaning method of the robot for cleaning according to the present invention.

25 is a block diagram showing the operation sequence according to the conditions of the wall riding pattern in the cleaning method of the cleaning robot according to the present invention.

* Description of the symbols for the main parts of the drawings *

1: cleaning robot 100: driving cleaning unit

110: operation unit 111: operation button

112: display unit 113: remote control

120: bumper 121: buffer

122: operating port 130: running part

131: driving wheel 132: driving motor

140: dust collector 141: brush

142: brush motor 143: suction port

144: suction motor 150: power supply

200: sensor unit 210: detection plate

220: infrared sensor 230: floor detection sensor

240: encoder 250: wheel lift sensor

300: central processing unit 310: control unit

The present invention relates to a cleaning robot and a cleaning method using the same, and more particularly, by operating a control button and a remote controller in a predetermined pattern or a combination of patterns and patterns set in advance or at a long distance, by efficiently driving the robot while cleaning the designated area. Cleans and detects obstacles with the contact sensor formed on the bumper and the infrared sensor formed in the front, and detects obstacles and obstacles on the ground driven by the floor sensor, encoder and wheel lift sensor to avoid switching to the optimum pattern. Therefore, the present invention relates to a cleaning robot and a cleaning method using the same, which improves the cleaning efficiency by enlarging the cleaning area while minimizing the occurrence of errors due to local minimization.

In general, the robot (Robot) is also called artificial human (人造 人間). The machine was assembled inside the doll, which was originally a human figure, and the hands and feet and other parts pointed to an automatic doll that operates like the original man. The term robot means `` robota '' in the Czech language.

Such robots are expensive equipments that implement complex algorithms, and their scope of application is very limited and theoretically feasible. However, robots are in a state that is not yet technically feasible. Although this trend is in full swing, it is still a long time to use the technology of robots easily at home, not in industrial form.

Accordingly, cleaning robots that operate with various algorithms by combining a sensing technology and a vacuum cleaner that can be implemented at present are developed and marketed.

In the prior art, the cleaning robot driving is based on the premise of recognizing the whole environment in a manner of having a driving plan by prioritizing global position discrimination, and only a low level of environment close to the robot using local environmental factors. It can be divided into the case of recognition.

When the entire environment is recognized and the cleaning path is generated in advance, the entire area is not only required for expensive equipment such as a camera and a beacon, but also needs a high-performance processor to process information obtained from them in real time. The system configuration is complicated and costs are increased because it requires separate equipment or algorithms to correct disturbances. It is slow and frequently occurs as too much data is sent and received for the simple task of cleaning. There was this.

In the case of using the local environmental factors, a simple system can be considered if the global environment is not recognized and preliminary path generation can be considered, resulting in a relatively low manufacturing cost, and a failure rate is reduced due to an uncomplicated structure, thereby ensuring product reliability. However, since there is no global information, the cleaning pattern is randomly generated and operated, so the reliability is low.When it reaches the special terrain by repeating the cleaning pattern, it is recognized as the local minimization and the cleaning robot stops as the cleaning robot stops. There was a problem that the efficiency of the cleaning area was reduced due to the reduction of the cleaning area.

In this manner, the tornado pattern, which is a pattern for cleaning in consideration of environmental factors and global factors, moves while changing the direction of the cleaning robot along the spiral running trajectory extending from the starting point to continuously suck dust remaining in the cleaning robot. When the cleaning robot rotates, the speed of the outer wheel is determined and the distance between the origin at the initial rotation end point and the center of the cleaning robot is determined, and then the driving wheel speed is adjusted according to the angle at which the direction of the cleaning robot is changed. To run along the toroidal spiral trajectory.

However, in the tornado pattern as described above, when the cleaning robot comes in contact with an obstacle, the cleaning area is gradually deflected, the area where the cleaning is not increased, and the cleaning area is greatly increased according to the encoder measurement that measures the speed of rotation by measuring the speed of the wheel. By changing, the angle of the whirlwind is greatly displaced according to the small error generated in the encoder, and there is a problem in that the area of the cleaning is greatly different from the setting.

In addition, the conventional cleaning robot tried to solve the problem by obtaining the global magnetic position and generating the optimum trajectory by using various sensor technology, but it is difficult to realize the technology, and the low-cost cleaning robot simply moves randomly because it is realized at high cost. To be. Cleaning robots having such random motions have a problem in that empty areas which are not cleaned are extended or cleaning time is lengthened.

In order to solve the problem, the main object of the present invention is to detect the position of the obstacle to detect the position of the obstacle when the obstacle is in contact with the bumper formed to cushion the shock formed in the front of the center, Infrared sensor is formed so as to be installed on the side side to detect obstacles in the front and side, and bottom detection sensor to detect the ground at the bottom, encoder to measure the speed of the running wheel and wheel lifter to detect the lifting of the wheel When the sensor unit that forms the sensor transmits information about the driving area to the central processing unit, the random bounce pattern, the zigzag pattern, the wave pattern, the tornado pattern, the square round pattern, the star pattern, the round round pattern, and the wall ride are adapted to each situation. By cleaning one or more combinations of patterns, obstacles and obstacles can be efficiently recovered. And by expanding the clean region and minimizes overlapping of the cleaning region it is to augment the effectiveness of the cleaning.

In addition, another object of the present invention is to reduce the production cost by effectively reducing the number of sensors installed to recognize that there is an obstacle between the two when the infrared sensor installed radially on the front is detected in the two areas according to each other.

Further, another object of the present invention is to use only the recognition of the obstacles and obstacles close to the sensor unit, so that the system configuration can be simplified and the cost can be reduced, and various cleaning methods can be combined to increase the cleaning efficiency as well as various Adaptation to the feature is consequently maximizing the cleaning efficiency.

In addition, another object of the present invention is to achieve the ultimate goal of cleaning the given cleaning area given as a cleaning robot as quickly as possible and without missing any part in the cleaning area, by performing agglomeration of cleaning in a certain area so as not to overlap specific areas. By defining two measures: coherence, which is a property of dense movement to a specific pattern, and roaming property, which is a property that a robot performing cleaning at a proper speed while moving a pattern quickly according to a situation does not stay in one area but moves to other areas. By presenting the transformation of various patterns of the cleaning robot and the direction of running, all the cleaning areas are cleaned evenly without the empty area at the minimum time to maximize the cleaning efficiency.

In addition, another object of the present invention is to input a variety of patterns in advance to the central processing unit to implement the appropriate pattern through the various proximity information sensed by the sensor unit and to increase the cleaning efficiency because the pattern is transferred in real time to effectively avoid obstacles. have.

In order to achieve the above object, the cleaning robot of the present invention and the cleaning method using the same have a cleaning pattern and a cleaning time on the upper part, and the remote controller is operated at a remote location and the operation state and operation information are displayed on the display unit. And a bumper provided with a shock absorber buffering a shock to the front and an actuating tool for transmitting a shock to the front surface, and a driving part rotating while the driving wheel is rotated by a driving motor in the lower center, The brush is rotated by a brush motor, and the suction part is formed by a suction motor, and a dust collecting part for collecting dust is formed. Formed to detect the obstacle contact to the front by forming a sensing plate that is energized in the form of And install the infrared sensor on the bumper radially, but the infrared sensor is installed on both sides more tightly than the installation distance of the center of the front, and installed on the lower front to face the ground in a plurality of floor detection to detect the obstacles The sensor unit forms a sensor, an encoder for measuring the speed of the driving wheel that rotates with the driving motor, a sensor unit including a wheel lifting sensor for detecting the lifting of the wheel, and an optimum driving pattern according to the information detected by the sensor unit. And a central processing unit which generates a signal to the control unit to perform the cleaning to control the driving of the driving cleaning unit and performs cleaning.

In addition, the infrared sensor of the sensor unit is installed in a radial direction, when detected from the front according to the area detected by each other requires a large avoidance angle, when detected from the side requires a relatively small avoidance angle toward the side When the two are detected together, it is characterized in that the installation and configuration to recognize that there is an obstacle between the two infrared sensors.

In addition, the cleaning setting step of setting the cleaning environment while watching the display unit displaying the status from the upper part with the operation button at a short distance and the remote control at a distance from the cleaning time, and the cleaning time and cleaning pattern set at the cleaning setting step. Follow the cleaning start step to transfer to the starting point to start the cleaning, and start cleaning with the transfer to the starting point of the cleaning in the cleaning start step random bounce pattern, zigzag pattern, wave pattern, whirlwind pattern, square rounding The pattern cleaning step of cleaning according to the pattern set in the operation unit by combining one or various patterns, star pattern, circular rounding pattern, and wall riding pattern, and cleaning in the pattern specified in the pattern cleaning step. When the obstacle is detected, the central processing unit depends on the position and shape of the obstacle. The obstacle avoidance step that runs while avoiding obstacles according to the signal of the controller in the optimum pattern that is processed and avoided, and the floor detection of the sensor part when the obstacles such as depression or stairs are reached while driving while avoiding obstacles in the obstacle avoidance step Disturbance terrain avoidance step of avoiding obstacles by converting from the central processing unit into a pattern that can detect and avoid the ground condition by the sensor, and avoiding obstacles in the optimal pattern, and cleaning obstacles while avoiding obstacles in the obstacles avoiding step When the area of the place is identified, it is combined with each other to compensate for the shortcomings of the random bounce pattern, zigzag pattern, wave pattern, tornado pattern, square round pattern, star pattern, circular round pattern and wall riding pattern. The composite pattern cleaning step and the optimum pattern cleaning step Avoid the danger space through each sensor unit while cleaning, or if the problem occurs inside and outside the abnormal terrain and the robot quickly avoids the risk according to the situation judgment, if the problem that can not avoid the avoidance pattern step Characterized in that comprises a.

In addition, in the pattern cleaning step, the random bounce pattern is basically a straight line, but according to the correlation between the radius of the cleaning robot, the maximum sensing distance of the sensor unit, and the stopping distance after detecting the obstacle according to the angle entering the obstacle. Cleaning is carried out by adjusting the cohesiveness and roamability by traveling through various paths by determining the movement route while rotating at the set angle using the set angle set in the below and the minimum avoidance angle in which the obstacle set in Equation 2 below is not detected again. Characterized in that.

(Equation 1)

, 청소용 로봇(1)의 반지름이

cm, 전면 적외선센서(220)의 최대 감지거리가

cm, 장애물을 감지한 후에 멈출 때까지 직진 이동한 거리가

cm 이다.Where the set angle is

, The radius of the cleaning robot (1)

cm, the maximum sensing distance of the front infrared sensor 220

cm, the distance traveled straight until it stopped after detecting the obstacle

cm.

(Equation 2)

, 랜덤각

이다.Here, the minimum evacuation angle is ∠avoidance, and the set angle is

, Random angle

to be.

In the pattern cleaning step, the zigzag pattern starts to rotate to the opposite side where the obstacle is sensed by the sensor unit, and when the side infrared sensor is detected during the rotation, the rotation is stopped, the angle rotated to the opposite side of the obstacle is calculated, and the distance is straight along the obstacle. After driving, the circumference is divided by the angle divided by the circumference and the rotated angle.The difference between the previous entry angle and the avoidance angle is small, so that the trajectory is generated with less deflection. When the corner is detected by the sensor unit, it is configured to avoid rotation by the opposite side of the corner.

In addition, in the pattern cleaning step, the wave pattern travels in a straight line at a predetermined distance, rotates in a circle in a moving direction, and turns in a parallel opposite direction. In the bottleneck where continuous obstacles are detected while avoiding obstacles, avoiding the first obstacle, and then avoiding other obstacles in succession, reducing the straight distance to 1/2 to move a narrow area, When it reaches, the straight line distance is shortened and this distance is smaller than the predetermined turning distance so as to change the traveling direction and avoid obstacles to perform cleaning.

In the pattern cleaning step, the tornado pattern is rotated while calculating the speed of the rotating wheel by setting the initial radius of rotation of the first spiral shape from the center of the set cleaning area, and rotating in a spiral trajectory while rotating 180 degrees. Increase the radius and apply the increase in the radius of rotation to the rotation speed of the driving wheel to increase the rotation speed of the wheel according to the increase of the angle every 180 degrees without calculating the approximate value to reduce the cumulative angle of the error range according to the approximation. It is characterized in that it is configured to travel while.

Further, in the pattern blue step, the square rounding pattern repeats the driving by increasing the radius by the increment of the rotated radius in the state in which the rectilinear distance is set in the initial radius and goes straight for a predetermined distance and is rotated 90 degrees in the direction of rotation. When it reaches a space with obstacles and reaches a space with obstacles, it avoids running in a square shape, creating overlapping areas regardless of the shape of obstacles, and after returning to the square rounding pattern after showing a pattern following the shape of obstacles It is characterized by driving to avoid obstacles.

Then, in the pattern cleaning step, the star-shaped pattern goes straight into a star shape that repeats the same operation n times while going straight for a certain distance and then rotated by an angle minus the circumferential ratio from the angle distributed according to the number of repetitions To move the cleaning robot from the center point of the star shape to the starting point and from the end of the pattern to the center point of the pattern, return to the center of the star once every time the pattern is finished, rotate it by a certain angle, and then perform the pattern again. Using the angle a calculated by Equation (1) is characterized in that it is configured to run after rotating to the angle calculated by the equation (2).

(Equation 1)

Here, the calculation angle is a, the number of rotations is n.

(Equation 2)

Here, the calculation angle is a.

In addition, in the pattern cleaning step, the circular rounding pattern is cleaned in a circular pattern having a diameter of half of the radius of the cleaning area in a state in which the circular cleaning area is set, and returns to the original position by Equation 1 below. Repeating the circular cleaning with the set number of times with half the radius of the cleaning area as the set number of times while rotating at the angle distributed according to the set pattern number of times, and rotating it 180 degrees when encountering obstacles, backwards the previous pattern and moves it to the starting point. It is characterized by traveling while avoiding obstacles by performing the normal next pattern in the state.

(Equation 1)

Here, the distributed angle is β and the number of executions is n.

In the pattern cleaning step, the wall riding pattern sets the maximum forward speed of the cleaning robot according to Equation 1 below and performs linear motion, and then determines whether there is an obstacle in each control cycle according to Equation 2 below. By controlling as shown in Equation 1, the forward speed is decreased while increasing the amount of rotation to the opposite side of the detected obstacle, and if there is no obstacle, the following pattern is repeated while following the detected wall, while following the outline of the obstacle to drive In the process of identifying and applying obstacles, the vehicle moves outside while changing its forward or rotational speed according to the position of the obstacle, and uses coherent motion that follows the outline to secure roaming, but returns to its place when c = 0. Since c = 0 and the moving distance s is greater than or equal to the threshold Tw, the process is terminated.

(Equation 1)

(Equation 2)

,

,

,

,

,

,

, 회전속도는

, 회전속도와 전진속도를 위한 게인행렬은

, 측면, 사선, 정면에 대한 회전속도와 전진속도를 위한 게인은

,

,

,

,

,

, 에러행렬은

이고, 측면

, 사선

, 정면

는 정면의 센서들이 연속해서 장애물을 감지하거나 감지하지 못한 횟수이며, 최대전진속도는

이다Here, the running speed of the equations (1) and (2) is

, Rotation speed

The gain matrix for the rotational speed and the forward speed

The gain for the rotational speed and the forward speed for the side, diagonal, and front

,

,

,

,

,

, The error matrix is

Side

, Diagonal

, face

Is the number of times the sensors in front of the vehicle are continuously detecting or not detecting obstacles, and the maximum forward speed is

to be

In addition, in the obstacle terrain avoidance step, the floor detection sensor of the sensor unit travels by setting a predetermined distance between the floor surface and the lower part of the cleaning robot as a target distance so as to detect the state of the terrain on which the cleaning robot is traveling, and the topographic map, the threshold which is impossible to travel. When it reaches an irregular abnormal terrain such as carpet and carpet, it detects the abnormal terrain and performs the avoidance pattern, and if the ground sensor does not detect the floor at a very short moment, it detects the terrain state. By ignoring and decelerating the maximum speed of the cleaning robot, if it does not detect more than the set time, it does not detect it.

In order to maximize the roamability in the composite pattern cleaning step, the vehicle travels over the minimum driving time by moving from the vertex to the vertex located diagonally from the vertex to move the farthest from the starting point, and moves a certain distance in the presence of an obstacle. Zigzag, wave, tornado, square round, star shape and circular rounding are combined to execute the regional pattern. When passing through a narrow place due to the detection of wealth, it switches to the wall riding pattern, and when it detects an obstacle or finishes the cleaning of a predetermined area, it moves to the wall riding pattern. The local pattern is maintained while being avoided even if it exists. For cleaning, set the wall cleaning time from the total cleaning time and drive it, but decrease it according to the cleaning time until the minimum driving time depending on the area where the cleaning overlaps, and if it exceeds the maximum driving time of the wall riding pattern, transfer to the next pattern for cleaning. As the process proceeds, the area of cleaning is divided according to the specifications of the cleaning robot, the time set for each pattern is distributed, the cleaning pattern is set according to the expected position according to the allocated time, and the shape, shape and size of the obstacle. To accurately clean according to the sensing frequency of the sensor part and suggest proper harmonization of the two patterns of cohesiveness and roaming, and then to perform cleaning by combining them in the order of traveling through wall pattern through local pattern through random bounce pattern. It features.

In addition, when the danger space is recognized in the avoidance pattern step, when the bottom sensor is not detected, the other sensor unit detects an obstacle and rotates 180 degrees to prevent reentry by reversing a predetermined distance from the random bounce or wall riding pattern. Carry out the cleaning by performing the avoidance pattern, and if the sensor detects that the obstacle continuously exceeds the set maximum detection time of 6 seconds ~ 9 seconds, it stops the operation by sending an error. When the wheel lifting condition is detected, the floor sensor does not detect the abnormal terrain and the wheel is lifted and it is caused by the user. Therefore, it stops the operation completely and sends an error when the set maximum detection time exceeds 6 seconds to 9 seconds. Characterized in that.

In the avoidance pattern step, it is determined that there is no obstacle in the state where the sensor cannot detect the wheel due to the terrain, carpet, and obstacle protruding from the bottom of the cleaning robot, which is determined to be an abnormal terrain, and the area of cleaning in the random bounce pattern. In case of regional patterns, each pattern is completed by ignoring the cleaning area, and in case of wall riding pattern, it is not detected continuously and rotates in one direction. Set the maximum driving time, which is the time that can be checked, and execute the avoidance pattern when the set time is exceeded.If the bumper of the cleaning robot is continuously pressed, it is judged as an obstacle inability to recognize the bumper due to bumper failure, user intention, or infrared sensor failure. Drive backwards a certain distance to avoid obstacles, In order to prevent falling by moving more than 1/2 of the diameter of the cleaning robot in the abnormal topless surface, the evacuation pattern is carried out by continuously moving more than 1/2 of the diameter. Characterized in that configured to stop.

In addition, the overload and failure of the brush motor caused by the foreign matter caught in the brush of the cleaning robot in the avoidance pattern step and the user's intention, the overload and failure of the driving motor generated when the robot is severely pressed in the ground direction Detects the overload and malfunction of the suction motor caused by the closing and sends out an error and stops operation if it is judged to be the failure of the robot, and if it is determined by the topographical characteristics, it rotates 150 degrees at the maximum speed and cleans it at the maximum speed. It reverses as much as 1/2 of the diameter of the robot and escapes from the terrain while avoiding the emergency, and acquires the overload signal input from the circuit that drives each motor. Stop and send an error.

As described above, a preferred embodiment of the cleaning robot and a cleaning method using the same according to the present invention will be described with reference to the accompanying drawings.

1 is a perspective view showing a cleaning robot according to the present invention, Figure 2 is a block diagram showing a control structure of the cleaning robot according to the present invention, Figure 3 is a process diagram showing a cleaning method of the cleaning robot according to the present invention, Figure 4 is a block diagram showing the position of the infrared sensor of the cleaning robot according to the invention, Figure 5 is a block diagram showing the position of the bottom sensor sensor of the cleaning robot according to the present invention, Figure 6 is a bumper of the cleaning robot according to the invention 7 is a conceptual diagram illustrating a cleaning pattern in a cleaning method of a general cleaning robot, FIG. 8 is a conceptual diagram illustrating a random bounce pattern in a cleaning method of a cleaning robot according to the present invention, and FIG. Conceptual view showing the principle of the random bounce pattern in the cleaning method of the robot for cleaning according to the present invention, Figure 10 is A conceptual diagram illustrating a zigzag pattern in a cleaning method of a utility robot, FIG. 11 is a conceptual diagram illustrating a zigzag pattern in a cleaning method of a cleaning robot according to the present invention, and FIG. 12 is a wave pattern in a cleaning method of a cleaning robot according to the present invention. 13 is a conceptual diagram illustrating a avoidance pattern of corners and bottlenecks as a wave pattern in the cleaning method of the robot for cleaning according to the present invention, and FIG. 14 illustrates a tornado pattern in the cleaning method of the robot for cleaning according to the present invention. 15 is a conceptual diagram illustrating a square rounding pattern in the cleaning method of the robot for cleaning according to the present invention, FIG. 16 is a conceptual diagram illustrating an obstacle avoidance pattern of the square rounding pattern in the cleaning method of the robot for cleaning according to the present invention; 17 is a circle of the star pattern in the cleaning method of the cleaning robot according to the present invention 18 is a conceptual diagram showing a star-shaped pattern in the cleaning method of the cleaning robot according to the present invention, Figure 19 is a conceptual diagram showing the principle of a circular rounding pattern in the cleaning method of the cleaning robot according to the present invention, 20 is a conceptual diagram showing a circular rounding pattern in the cleaning method of the cleaning robot according to the present invention, Figure 21 is a conceptual diagram showing an obstacle avoidance pattern of the circular rounding pattern in the cleaning method of the cleaning robot according to the present invention, Figure 22 is the present invention Fig. 23 is a conceptual diagram showing a wall riding pattern in a cleaning method of a cleaning robot according to the present invention. FIG. 23 is a block diagram showing an operation sequence according to a condition of a random bounce pattern in the cleaning method of a cleaning robot according to the present invention. Is a block diagram showing the operation sequence according to the conditions of the wave pattern in the cleaning method of the cleaning robot according to 25 is a block diagram showing the operation sequence according to the condition of the wall riding pattern in the cleaning method of the robot for cleaning according to the present invention.

As shown in Figures 1 to 25, the cleaning robot of the present invention to form a control unit 110 for manipulating the setting of the cleaning on the top and to form a bumper 120 to cushion the shock on the front, the driving wheel at the bottom The driving unit 130 is formed to run to 131 and the dust collection unit 140 to suck the garbage by the rotation of the brush 141 to the lower center is formed, and the power supply unit 150 for supplying power to the rear side is formed Sensor unit for installing the cleaning unit 100, a plurality of sensors in the front, bottom and radially of the driving cleaning unit 100 and detecting the driving speed and the lifting of the wheel of the driving unit 130 to detect the state of the location of the obstacle. And a central processing unit 300 for controlling the driving cleaning unit 100 in an optimal driving pattern by combining the information detected by the sensor unit 200.

The driving and cleaning unit 100 operates a cleaning pattern and a cleaning time on the upper part with the operation button 111, and operates the remote control unit 113 at a long distance and displays the operation state and operation information on the display unit 112. ) And a bumper 120 having a shock absorbing hole 121 for cushioning the shock to the front and an actuating hole 122 for transmitting a shock, and a driving wheel driven by the driving motor 132 at the lower center. The 131 forms a traveling part 130 that travels while rotating, and sucks the dust through the suction port 143 in which the suction motor 144 is formed while rotating the brush 141 with the lower brush motor 142. The dust collecting unit 140 is formed, and is configured to form a power supply unit 150 for supplying power to the rear portion.

The sensor unit 200 forms a sensing plate 210 that is energized in the form of a thin film so as to be interlocked with the operation hole 122 of the bumper 120 of the driving and cleaning unit 100, and is formed to detect an obstacle contact to the front, and a bumper. Install the infrared sensor 220 radially on the 120, the infrared sensor 220 is installed more densely installed spacing of both sides than the installation interval of the center of the front, and installed so as to face the ground in plurality on the lower front Form a floor sensor 230 for detecting the obstacle terrain, the encoder 240 for measuring the speed of the driving wheel 131 rotated by the driving motor 132 and the wheel lifting sensor for detecting the lifting wheel ( 250).

Here, the infrared sensor 220 of the sensor unit 200 is installed radially and requires a large avoidance angle when detected from the front according to the area detected by each other, and requires a relatively small avoidance angle when detected from the side. As it is formed closer to the side, and when the two are detected together, it is installed and configured to recognize that there is an obstacle between the two infrared sensors 220.

Acquisition of data due to the detection of the sensor unit 200 as described above assumes that most obstacles encountered in the environment in which the cleaning robot 1 runs are larger than the cleaning robot 1 and the curvature is not large. Compared to the case detected from the side and the case detected from the side, the former requires a large avoidance angle, while the latter requires a relatively small avoidance angle. On the other hand, the wall riding motion is advantageous to smoothly follow the outer surface of the obstacle when the obstacle to the side is more accurate than the front. Therefore, the sensor of the cleaning robot is concentrated on the side so that the limited sensor can be used efficiently. To meet this need even more, a series of sensors arranged in series can detect obstacles between them.

For example, as shown in FIG. 4, when the sensors 0 and 1 detect an obstacle, the obstacle is at 80.5 degrees, and when 0, 1, and 2 are detected, the obstacle is at 71 degrees.

The angle of the obstacle for the robot according to Equation 1 is as follows.

는

번째 감지된 센서의 각도이고,

은 감지된 센서의 수이다.From here,

Is

Angle of the first detected sensor,

Is the number of sensors detected.

By applying the resolution expansion method used to the bumper sensor, obstacle avoidance after collision can be performed more smoothly.

The central processing unit 300 generates a signal to the control unit 310 by calculating to fit the optimal driving pattern according to the information detected by the sensor unit 200 to perform cleaning while controlling the driving of the driving cleaning unit 100. Configure.

Thus, looking at the cleaning method of the cleaning robot of the present invention configured as follows.

First, the cleaning setting step (S1) of setting the cleaning environment while watching the display unit 112 displaying the status from the upper portion with the operation button 111 at a short distance and the remote control 113 at a distance based on the cleaning time and the cleaning pattern. After that, cleaning is started in the cleaning start step (S2) to start the cleaning by transferring to the starting point started according to the cleaning time and cleaning pattern set.

In this way, when the cleaning is started while being transported to the starting point of cleaning in the cleaning start step S2, one random bounce pattern, a zigzag pattern, a wave pattern, a square round pattern, a star shape pattern, a circular round pattern and a wall riding pattern are one. Alternatively, the cleaning is performed while driving the trajectory corresponding to the pattern in the pattern cleaning step (S3) for cleaning according to the pattern set in the operation unit 110 by various combinations.

If the cleaning efficiency is defined as how many areas of cleaning have been performed in a minimum amount of time, in order to clean the area, the cleaning robot overlaps more areas in order to increase the cleaning efficiency. You must move without.

For this movement of the cleaning robot 1, two measures of roaming and cohesion are defined. Roamability is a property that the robot moves to other areas without staying in one area, and cohesiveness is a property that densely moves in a specific pattern without overlapping specific areas.

The high roaming pattern has an advantage that the ability to reduce the cleaning time is improved because the transition to other cleaning areas is fast, and the high cohesive pattern performs compact cleaning, thereby minimizing empty areas.

Cleaning efficiency is defined as cleaning the maximum area in the shortest time, and can be expressed as a combination of two measures of cohesion and roaming.

In the cleaning method according to the present invention, after the cleaning pattern for the cleaning robot 1 is prepared and the roaming and cohesiveness of each pattern is analyzed, the weights of the two scales are changed so that the cleaning efficiency can be maximized. Factors for changing the weight of the two scales are the cleaning time set by the user and the remaining cleaning time, local terrain information obtained from recently detected obstacles, the distance the cleaning robot has moved, and the rotation angle. Depending on these factors, the weight of the roaming and cohesiveness of each cleaning pattern is changed to ultimately optimize the cleaning efficiency.

에서

로 이동하는 동안 총 이동거리는

이고

는

이다. Looking at the definition of the general pattern cleaning as described above, as shown in Figure 7, the cleaning robot for a unit time

in

The total distance traveled while

ego

Is

to be.

In this case, the coverage may be defined in Equation 2.

c: Coverage, S: Total travel distance, L: Travel distance for unit time

The value obtained in Equation 2 helps to determine the characteristic of the cleaning robot moved from the starting point to the last point. In the first case, s = l ≠ 0, which is effectively away from the initial position, and c = 1. In the second case, s> l moves to 0 <c <l. Lastly, s> l = 0, which is a continuous forward and reverse operation, but the actual distance is but c = 0 because the starting and ending points are in the same position.

The goal of the cleaning robot is to clean a large area so that there is no empty area in a short time.

It is desirable to follow a pre-planned zigzag trajectory for the most optimal cleaning, but it is not suitable for robots without sensors because it requires perfect magnetic positioning.

Therefore, the cleaning robot needs not only good transfer ability to other cleaning areas but also the ability to travel without leaving a specific area. From the above, the following figure of merit is defined.

Wandering: When c is close to 1, the transition to other cleaning areas is rapid

Cohesion: Because c is close to zero, it lacks the ability to transfer to other areas of cleansing but minimizes the empty area because it is less likely to escape within a certain area.

라 하고

이어야 한다는 부가 조건이 필요하다. 물론 사용자에 의한 특정영역에 대한 집중적인 청소는 예외이다.However, it is important to minimize the overlapping areas so

And

An additional condition is to be required. The exception is intensive cleaning of specific areas by the user.

The movement pattern which causes the cleaning robot 1 to have the highest roamability is a linear driving pattern as in the first case. In view of the entire area, the direction of the cleaning robot 1 may return to its place due to obstacle avoidance, but it is a pattern having the highest roaming force locally.

However, if complex features exist, the roaming can be drastically reduced during the avoidance process. At this time, since the wall riding pattern moves along the outline of the feature, it can be used as an auxiliary pattern to help move to other cleaning areas.

Afterwards, we propose a method of generating and managing movement patterns that maximize cohesion and roaming, respectively, and propose appropriate distribution methods for them.

In this pattern cleaning step (S3), the random bounce pattern is basically a straight line, but the radius of the cleaning robot 1, the maximum sensing distance of the sensor unit 200, the stopping distance after the obstacle detection, depending on the angle entering the obstacle. Cleaning is carried out by adjusting the roaming ability by traveling through various paths by determining the movement route while rotating at the set angle using the set angle set by the correlation and the minimum avoidance angle at which obstacles are not detected again.

As described above, the random bounce pattern moves straight until it meets the obstacle, and when the obstacle is detected, the random bounce pattern rotates at a random angle to avoid the obstacle.

This pattern is probabilistic, and has a high merit that it is possible to quickly move the clearing area. However, the following problems occur in the process of avoiding obstacles, so the roaming can be drastically lowered.

이므로 배회성이 급격히 낮아진다. Since the evacuation angle is randomly selected when rotating at a random angle after detecting the obstacle, the evacuation path overlaps the entry path severely,

Because of this, the roaming rate is drastically lowered.

To solve this problem, the following algorithm is used.

의 각도로 회전하면 회피할 수 있으며, 측면으로 진입 하면 45도+

각도, 직각으로 진입하면 0도+

각도로 회전하면 회피할 수 있다. As shown in FIG. 8, when the cleaning robot 1 runs along the arrow and enters the front, 90 ° +

Can be avoided by rotating at an angle of 45 ° +

Angle, right angles, 0 degrees +

Rotation at an angle can be avoided.

는 각각 90, 45, 0도 회전하더라도 측면이 지속적으로 장애물을 감지하는 것을 방지하기 위해서 추가되는 설정각도이다.From here

Is a set angle added to prevent the side from continuously detecting obstacles even when rotated 90, 45, 0 degrees respectively.

cm, 전면 적외선 센서의 최대 감지거리가

cm, 장애물을 감지한 후에 멈출 때까지 직진 이동한 거리가

cm 일 때 설정각도는 아래의 수학식 3의 상관관계식에 의해 구해진다. 이 때 물론

이어야 한다 This, the set angle is the radius of the cleaning robot 1 as shown in FIG.

cm, the maximum sensing distance of the front infrared sensor

cm, the distance traveled straight until it stopped after detecting the obstacle

The set angle at cm is obtained by the correlation of Equation 3 below. Of course at this time

Should be

When the obstacle avoidance angle of the cleaning robot 1 is obtained in the same manner as in Equation 4 using the set angle obtained according to the above method, it is as follows.

는 0보다 크고 일정 범위를 갖는 난수로서 두 가지 역할을 한다. From here,

Is a random number that is greater than zero and has a range.

를 포함시켜서 청소용 로봇의 이동경로에 무작위성을 부여하여 확률적으로는 다양한 경로를 선택할 수 있도록 하는 것과 난수의 범위를 설정함으로써 배회 성을 조절하는 역할을 수행한다. The angle except this is the minimum evacuation angle at which the cleaning robot does not detect the obstacle.

By including the randomness to the moving path of the cleaning robot to randomly select a variety of paths and plays a role of controlling the roamability by setting a range of random numbers.

In addition, in the pattern cleaning step (S3), the zigzag pattern starts to rotate to the opposite side where the obstacle is detected by the sensor unit 200, and when the side infrared sensor is detected while rotating, the rotation is stopped and the angle rotated to the opposite side of the obstacle is calculated. After driving for a certain distance along the obstacle, if you drive by turning the circumference divided by the bisected angle and the rotated angle, the difference between the previous entering angle and the avoiding angle is small, which creates a trajectory with less deflection. Cleaning is performed while following, and when the corner is detected by the sensor unit is configured to avoid by rotating to the opposite side of the corner.

The zigzag pattern of the related art is a highly coherent pattern, and the following process is repeated as shown in FIG. 10.

를 청소용 로봇의 직경보다 짧게 선택하므로 빈 영역이 없이 세밀하게 청소할 수 있다는 장점을 가지고 있으며, 이상적으로는 주어진 청소영역을 최단 시간에 커버할 수 있는 것으로 알려져 있다. The pattern is generally rotated 90 degrees in the zigzag pattern's advancing direction with a certain distance straight, and is rotated 90 degrees in the previous direction of rotation with the rotation straight for a certain distance.

Since it is shorter than the diameter of the robot for cleaning, it has the advantage that it can be cleaned finely without empty areas, and ideally it is known that it can cover a given cleaning area in the shortest time.

However, this pattern generates the following problems during the generation process.

As the traveling direction is gradually deflected due to the positive displacement superimposed on the cleaning robot in a conventional zigzag run, it becomes unsuitable as a cohesive pattern.

In addition, according to the obstacle avoidance method, an area that is not regularly cleaned along the wall surface is generated, and thus cleaning near the obstacle becomes impossible, and the obstacle avoiding process is complicated because it is divided into several steps.

In order to solve some of these problems, the zigzag pattern of the present invention is used. The zigzag pattern of the present invention uses the trajectory as shown in FIG. 11, and the reaction taken when an obstacle is detected is very characteristic as follows, and solves the problem by this method.

Such a zigzag pattern of the present invention stops the rotation when the side infrared sensor is detected while rotating by starting the opposite side of the obstacle detected by the sensor unit 200, and calculates an angle rotated to the opposite side of the obstacle and is fixed along the obstacle. After going straight, the car is rotated by an angle obtained by summing the circumferential ratio in half and the rotated angle.

In practical applications, most of the parallel wall obstacles occur, and the error of setting the direction of the cleaning robot 1 is largely generated when it rotates, rather than going straight. If you get the angle of entry for obstacles, the distance from the previous entry angle is used by using it in the exiting process. In other words, it becomes possible to generate a trajectory with less deflection.

In addition, if the side infrared sensor is detected during the rotation, the rotation stops, calculates the angle rotated to the opposite side of the obstacle, and if it goes straight along the obstacle, the wall adapts to the wall, so the cleaning on the wall can generate a smooth trajectory. .

This pattern can be continually facing the corner, so it is desirable to avoid it in roaming motion at close overlaps.

In addition, since the angle between the infrared sensor detected when falling into a corner or a trap similar to the corner is about 90 degrees, this condition can be used to quickly and easily detect corners. After detecting the corner in this way, the traveling direction of the cleaning robot 1 rotates to the opposite side of the corner to travel.

In addition, it is advantageous when the robot follows the wall pattern while avoiding to the center of the other side of the corner, because the frequency of obstacles around it is likely to be low, so that the area pattern can be maintained for a long time, which is advantageous to increase the cleaning area. Do.

This method has the additional benefit of ensuring the cleaning of the corners.

In the pattern cleaning step (S3), the wave pattern travels in a straight line at a predetermined distance, rotates in a circle in a traveling direction, and rotates in a direction opposite to parallel to move in a smooth and quick manner. After securing (1) the turning radius, drive around while avoiding obstacles by turning to avoid obstacles. If the bottleneck is detected, even if the obstacles are detected or if it is detected while going straight, the obstacles do not go straight for the desired distance. After the second avoidance motion, set the straight distance to 1/2 of the previous straight distance to move to the narrow area.When the corner terrain is reached, the straight line is shorter and if this distance is smaller than the predetermined turn distance, the direction of travel is avoided. While cleaning.

Such a wave pattern has a similar shape as zigzag as shown in FIG. 12, but does not require two rotations in the process of changing the direction of the linear motion, so that the wave pattern can be connected more quickly. Satisfaction with watching a smoother motion can also occur.

The above method has a problem of repeating the same trajectory in a corner or bottleneck topography as shown in FIG. In the bottleneck type, even if the obstacle is detected during the rotation or the straight line, the obstacle cannot go straight as much as the desired distance due to successive obstacles.After the second evasion, the straight distance is set to 1/2 of the previous straight distance to move to the narrow area. To help. In the case of corner terrain, if the straight line is shortened for the same reason as the bottleneck, and this distance is smaller than the predetermined turning distance, the moving direction is changed so that the cleaning robot can easily escape from the corner.

In the pattern cleaning step (S3), the whirlwind pattern is rotated while calculating the speed of the rotating wheel 131 by setting the initial radius of rotating the first spiral shape from the center of the set cleaning area, and rotating the spiral by 180 degrees. By applying an increase to the rotational speed of the driving wheel 131 increased by the rotation radius increases by increasing the increase of the rotational speed of the driving wheel 131 according to the increase of the angle every 180 degrees without approximation while minimizing the load of the calculation It is configured to travel while minimizing the deflection of the trajectory by reducing the cumulative angle of the error range according to the approximation.

In general, as shown in FIG. 14, the conventional tornado pattern has a very coherent pattern similar to zigzag or wave. This pattern moves along the spiral trajectory extending from the starting point, but also changes the direction of the cleaning robot, which also has the advantage of continuously sweeping away the dust left in the cleaning robot 1. The process of creating this pattern can be summarized as follows.

In this conventional whirlwind pattern, the distance from the center point to the cleaning robot is increased by an appropriate increase in preparation for the cumulative angle of the cleaning robot 1 with respect to the center point while the initial distance of the cleaning robot is set from the center point. The speed of the wheel 131 is obtained and applied until the cleaning robot 1 reaches the predetermined maximum distance.

Such a conventional tornado pattern has the following problems in the production method and operation process.

Obtaining and applying the speed of the driving wheel 131 while increasing the distance according to the increase in preparation for the cumulative angle with respect to the center point, the speed of the driving wheel 131 should be obtained and applied whenever the proper time or angle changes. Accompany the load.

In addition, it is difficult to obtain a speed that accurately converges to all the results (distance, angle) according to the speed change of the driving wheel, so that an estimated value is generated according to an approximation, and when they accumulate, the entire trajectory is also modified.

In addition, if the distance from the center point is infinitely increased, the speed difference between the two wheels is also infinitely small, so that the resolution of the encoder must be infinitely large in order to perform this.

The tornado pattern of the present invention, which solves the problem of the conventional tornado pattern, updates the speed every 180 degrees, so that no operation load occurs. If the conventional traveling speed is calculated to update the speed every 1 degree, the present invention, which is updated once every 180 degrees, generates about 180 times the efficiency.

In addition, even if the cumulative angle generated by the approximate calculation occurs every 180 degrees, severe deflection due to their accumulation is reduced.

In addition, in the pattern cleaning step (S3), the rectangular rounding pattern is repeated with a predetermined distance and the radius is increased by the increment of the radius rotated in a straight state by a certain distance while going straight for a predetermined distance and rotating 90 degrees in the rotation direction. If you reach a space with obstacles while expanding and driving in a wide rectangular shape, do not move while moving in a rectangular shape while making overlapping areas, excluding movements that comply with the obstacles, or after a short time of movement. Drive back to avoid obstacles while returning to the square round pattern.

Since the square rounding pattern is not shown in FIG. 15, since the arc trajectory is not used, if the distance from the center point of the tornado pattern is infinitely increased, the speed difference between the two wheels is also infinitely small, so that the resolution of the encoder must be infinitely large. It is a pattern that solves the burden that the initial radius is set to go straight for a certain distance and rotates 90 degrees in the direction of rotation while driving for a certain distance. Drive while expanding.

Like the tornado pattern, the square rounding pattern increases cohesion when there are no obstacles, but when traveling in an obstacle space due to the needs of the user, the coarse rounding pattern can be expanded while maintaining cohesion. Can be. First, the left one has the advantage of being able to perform more accurate motion, although the overlapping area is continuously generated without the movement of the obstacle. The figure on the right is a straight forward acclimation movement along the perimeter of the obstacle in order to maintain the movement of this pattern.

In the pattern cleaning step (S3), the star-shaped pattern goes straight after a predetermined distance, and then rotates by the angle subtracted from the circumference at an angle distributed according to the number of repetitions. After moving from the center point of the star shape to the starting point and from the end point of the pattern to the center point of the star, return to the center at the end of each pattern, rotate it by a certain angle, and then repeat the pattern. Configure to run.

이 되도록 하였으나 본 발명의 별형상패턴은 특정 영역을 집중적으로 청소할 수 있게

,

는 가능한 크도록 만들어진 패턴이다. The regional patterns we have introduced so far for maximum cohesion

However, the star-shaped pattern of the present invention can clean a specific area intensively

,

Is a pattern made to be as large as possible.

Of course, it is possible to do this by adjusting the patterns introduced so far, but the star-shaped pattern has the advantage that it is relatively easy to produce and can be removed through large rotation even if a large material remains at the bottom of the cleaning robot.

In addition, since there are various types of dust, rubbish, carpets, and so on, a particular area may be approached at various angles, and thus, the thread or hair twisted in a specific direction may be efficiently removed and cleaned.

만큼 직진한 후에 하기의 수학식 5에서 산출된 각도 a를 이용하여 수학식 6에서 산출된 각도로 회전하는 과정을

회 반복한다.Such a star-shaped pattern, as shown in Figure 17,

After going straight forward by using the angle a calculated in Equation 5 below to rotate to the angle calculated in Equation 6

Repeat times.

When the above process is performed, the star pattern shown in FIG. 18 is completed.

Depending on the number of repetitions and the straight distance, the blank area becomes more prominent or the user wants intensive cleaning according to the user's request, but in order to avoid possible movement of the center point, the process moves from the center point to the start point and from the end point to the center point. After the movement of the pattern, return to the center at the end of each pattern, rotate by a certain angle, and then apply the pattern again.

In addition, in the pattern cleaning step (S3), the circular rounding pattern is cleaned with a circular pattern having a diameter of half the radius of the cleaning area in the state of setting the circular cleaning area, and returns to the original position according to the set number of times of pattern execution. Repeat the circular cleaning with the set number of times by rotating half the radius of the cleaning area by the assigned angle, and if it meets the obstacle, rotate it 180 degrees to reverse the previous pattern and move it to the starting point. Drive while avoiding obstacles.

As shown in FIG. 19, the circular rounding pattern is a pattern that serves as a star shape pattern and is generated through the following process.

라고 할 때 지름을

로 하는 원을 따라서 주행하다가, 제자리에 돌아오면, 패턴 수행회수를

이라고 할 때, 수학식 7의 각도로 회전한다.First, the radius of the cleaning area

The diameter

Drive along the circle, and when it comes back

Is rotated by the angle of the equation (7).

Repeating the above process to complete the pattern.

If the circular rounding pattern is repeated 12 times, it can be seen that the driving trajectory is uniformly crossed as shown in FIG. 20.

Such a circular rounding pattern is difficult to cope with when a star-shaped pattern meets an obstacle, while the circular rounding pattern of the present invention moves to a central point to have an advantage that the response to the obstacle is simple and secures a constant cohesion. Rotate

In response to the obstacle, as illustrated in FIG. 21, when the obstacle is encountered, the controller 180 rotates 180 degrees, and the previous pattern is moved backwards to the starting point, thereby performing the normal next pattern.

In addition, in the pattern cleaning step (S3), the wall riding pattern sets the initial speed of the cleaning robot 1 and performs a straight motion, and if there is an obstacle in every control period, the forward speed is increased while increasing the amount of rotation to the opposite side of the detected obstacle. If there is a decrease and there is no obstacle, the driving pattern is repeated while following the detected wall and increasing the forward speed, and following the outline of the obstacle.In the process of identifying and applying obstacles, the acceleration or rotation speed is changed depending on the position of the obstacle. While driving along the wall outline.

In complex terrain, since the straight line distance of the random bounce pattern is shortened, the roaming property is not drastically reduced, and even when using a regional pattern, the roaming property is very low due to a large number of obstacles. The riding pattern follows the outline of the obstacle, making it easy to escape from complex terrain.

On the other hand, the wall riding pattern is a basic driving method of most mobile robots and is a representative cleaning pattern used by most cleaning robots. In the case of the cleaning robot, a sensor for this pattern is usually set separately. The wall riding pattern of the present invention proposes a wall riding method using seven infrared sensors 220 installed on the front as a main sensor.

Conventional wall riding pattern is equipped with one infrared sensor for wall riding on the right to determine whether they are detected by adapting to the wall, or distance sensor looking at two parallel places as a sensor for wall riding 1 A jaw was adopted and adapted to the wall using the distance difference detected from each.

However, the cleaning robot 1 of the present invention proposes a wall riding pattern using seven infrared sensors 220 mounted on the front surface to detect obstacles, and as shown in FIG. After setting initial speed and performing straight motion, if there is an obstacle, the rotation speed is increased to the left to decrease the forward speed. If there is no obstacle, the rotation speed is increased to the right to increase the forward speed. Drive while following the outline.

The characteristic of the wall riding pattern that runs as described above is that the acceleration or deceleration of the forward or rotational speed varies depending on the position of the obstacle with respect to the cleaning robot during the obstacle identification and application process. Through this wall riding pattern, it is more flexible than other cleaning robots and can adapt to the wall, precisely, the outline of obstacles.

The sensor information continuously obtained from the sensor unit 200 can be induced at the speed of the cleaning robot 1 in Equations 8 and 9 below.

, , ,

,,,

는 주행속도이고

는 회전속도이다. 또한

는 각각 회전속도와 전진속도를 위한 게인행렬이고

,

,

,

,

, 그리고

는 각각 측면, 사선, 정면에 대한 회전속도와 전진속도를 위한 게인이다.

는 에러 행렬로

는 측면,

는 사선,

는 정면의 센서들이 연속해서 장애물을 감지하거나 감지하지 못한 횟수이다. 또한,

는 최대 전진속도이다.ego

Is driving speed

Is the speed of rotation. Also

Are gain matrices for rotational speed and forward speed, respectively

,

,

,

,

, And

Are the gains for the rotational speed and the forward speed for the side, diagonal and front, respectively.

Is an error matrix

Side,

Diagonal,

Is the number of times the sensors in front of the car have continuously detected or failed to detect an obstacle. Also,

Is the maximum forward speed.

In the wall riding pattern running at a speed like this, the wall is mounted on the wall to the right and if it is detected, the sign is + and the sign is-. It should be noted here that the sensor unit 200 of the oblique line or the front has a value of 0 while not detecting the wall so as not to excessively react to the wall. In addition, the infrared sensors 220 on the left side are all regarded as the front side to lower the complexity and to adapt to various surfaces. In the case of gain matrix, it is desirable to adjust the gain on the front side so that the gain on the front side is in the order of the largest, diagonal, and lateral order so that it is sensitive to the collision at the front, that is, it does not collide.

이 되므로,

이고 이동거리(

)가 문턱치

이상이면 종료하도록 해야 한다. The wall riding pattern ultimately uses coherent motion that follows the outline to ensure roaming, but when it follows obstacles of unusual structure such as columns,

This becomes

And travel distance (

) Threshold

If it is abnormal, it should be terminated.

As described above, when the cleaning is performed in the pattern specified in the pattern cleaning step S3 and the obstacle is detected by the sensor unit 200, the controller is processed to the optimal pattern which is processed and avoided by the central processing unit 300 according to the position and shape of the obstacle. After the obstacle avoidance step (S4) to run while avoiding the obstacle in accordance with the signal of 310, the floor detection sensor 230 of the sensor unit 200 when the obstacle is reached, such as sinking or stairs while avoiding the obstacle In the obstacle pattern avoidance step (S5) of avoiding obstacles by traveling in the optimal pattern by converting the central processing unit 300 into a pattern that can be avoided by grasping the state of the ground in the driving pattern.

Here, the floor detection sensor 230 of the sensor unit 200 in the obstacle terrain avoiding step (S5) detects the presence or absence of the floor, so as to determine whether the cleaning robot 1 can be driven according to the target distance, the floor surface and If a certain distance to the bottom of the cleaning robot (1) is set as the target distance, the distance between the floor sensor 230 and the ground momentarily exceeds the target distance in an irregular abnormal terrain such as a general map, a road map that is impossible to drive, and a threshold and a carpet. The longer the floor sensor 230 ignores not detecting the floor in a very short time and lowers the maximum speed of the robot 1 for cleaning, and if it does not detect for a specific time, it does not detect it by performing a avoiding motion for a certain distance. It is configured to avoid rotating while reversing.

As described above, looking at the position of the cleaning robot (1) to perform the equivalent acceleration movement while detecting the floor with the bottom sensor 230 formed as shown in Figure 5 as follows.

, 초기속도

, 그리고 가속도

로 한다.Initial position

, Initial speed

, And acceleration

Shall be.

이라고 하면 약 0.16초 내에 약 1cm 정도 초과하여 멈추는 것이 가능하므로 안전하다. 참고로 정상속도를 유지하면 약 4.5cm 초과하게 되다. Lower the maximum speed of the cleaning robot from the basic speed of 30cm / s to 15cm / s,

In this case, it is safe to stop by more than about 1cm within about 0.16 seconds. For reference, if the normal speed is maintained, it exceeds 4.5cm.

The above method can be used in the infrared sensor 220 having the same specifications in terms of productivity, there is an advantage that can also ensure user safety.

In this way, when the obstacle area avoidance step (S5) to avoid the obstacle shape, while driving and grasping the area of the cleaning place is random bounce pattern, zigzag pattern, wave pattern, star pattern, circular rounding pattern, square rounding pattern and wall riding pattern In the composite pattern cleaning step (S6) in which the complex cleaning is performed by the optimal cleaning pattern by combining to compensate for the shortcomings, the cleaning is performed by combining the complex patterns according to the position and size of the cleaning area and the obstacle.

The needs of the user can be divided into the area cleaning where the cleaning area is relatively small and the concentration is large according to the cleaning area and the concentration level, and the whole cleaning that does not require the big concentration level.

In the case of local cleaning, circular rounding, star shape, wave, and whirlwind patterns with high cohesion are advantageous. In the case of whole cleaning, it is necessary to transfer and concentrate the cleaning area. The compound pattern cleaning step (S6) is performed.

Depending on the cleaning condition over time, the remaining space without cleaning gradually decreases, but it does not decrease after a certain amount of time. This trend is particularly noticeable when using a roaming pattern because small uncleaned areas remain everywhere and are known to take four to five times longer than cleaning the same area with an ideal zigzag trajectory. .

In order to solve this problem, as described above, the distribution of the roaming motion and the cohesive motion is properly distributed, and the cohesiveness is increased toward the end of the cleaning to minimize the empty area and reduce the cleaning time. In addition, since the user is likely to drive the cleaning robot 1 in a place where there are few obstacles, such as the center point of the cleaning area, initially, the coherence pattern is first executed to minimize the empty area. This method has excellent performance in reducing the empty area to less than about 10% even in the presence of many obstacles.

 In the composite pattern cleaning step (S6), the combined patterns are sequentially run through the random bounce pattern through the regional pattern incorporating zigzag, wave, tornado, square rounding, star shape and circular rounding. By combining, it runs more than the minimum driving time by moving from the vertex to the vertex at the diagonal to move the farthest from the starting point to maximize the roaming, and moves the distance pattern in the presence of obstacles to execute the regional pattern. In order to ensure fair distribution and coherence, the local pattern is immediately executed when exceeding about three times the minimum driving time, and when it passes through a narrow place by the detection of the sensor unit 200, it switches to a wall riding pattern and changes the local pattern. After detecting obstacles or cleaning of the pre-determined area, the vehicle will run in the wall riding pattern. In the later stage, local patterns are maintained while avoiding obstacles in order to maximize coherence, and the wall riding pattern runs over the set time considering the wall surface of the total cleaning time for the cleaning of all walls, but the area where the cleaning overlaps. According to the cleaning time, the minimum driving time is reduced according to the cleaning time, and if the maximum driving time of the wall riding pattern is exceeded, the next pattern is transferred to the next pattern. Distribute the set time to set the cleaning pattern suitable for the expected position according to the allocated time, and accurately determine the shape, shape and size of the obstacle according to the frequency of detection of the sensor unit 200 to identify two patterns of coherence and roaming Properly harmonized directionality, through random bounce patterns, through local patterns to wall ride patterns It is configured to perform cleaning in combination in the order of running.

In order to mimic the behavior of humans moving after the near cleaning by sequentially processing the roaming pattern and the cohesive pattern in the compound pattern cleaning step (S6), the cleaning area is transferred to the roaming pattern. Fine cleaning using cohesive patterns is the basic sequence. In addition, since the escape is necessary in a complicated environment, the wall ride pattern for escape must be connected last. On the other hand, it is difficult to maintain optimal cohesiveness in an obstacle-based environment, so transition to a wall riding pattern is preferable.

In the complex pattern cleaning step (S6), the degree of transition of the cleaning area to control the complex pattern is dependent on the size of the area to be cleaned, so it is necessary to grasp the geometric characteristics of the cleaning area from the user's time input. .

이고, 청소용 로봇의 직경을

라고 할 때, 청소용 로봇이 단위시간당 청소하는 단위청소영역을 수학식 11과 같이 정의한다.The average speed of the cleaning robot 1

The diameter of the cleaning robot

In this case, the unit cleaning area that the cleaning robot cleans per unit time is defined as in Equation (11).

, 총 청소시간을

라고 하면 수학식 12과 같은 관계를 유도할 수 있다. And the area of the given cleaning area

Total cleaning time

In this case, a relationship as shown in Equation 12 can be derived.

는

에 따른 청소효율로서 주어진 영역을 겹치지 않고 완벽하게 청소하며

이나 일반적인 경우에는

이다. 장애물에 대한 회피각이 임의로 선택되면

이다. 사용자가 선택하는 것은 시간이므로 위의 수학식 12를 수학식 13과 같이 변형한다. From here,

Is

Cleaning efficiency according to the

Or in common cases

to be. If the avoidance angle for the obstacle is randomly selected

to be. Since the user selects time, Equation 12 is transformed to Equation 13.

In this case, assuming that the cleaning area is a quadrangle, the maximum outline length of the cleaning area is expressed by Equation 14.

를 선택하고, 장애물이 존재하므로

,

,

라고 하면 청소영역과 외곽선의 길이는 수학식 15와 같다. The user

Select it, and there are obstacles

,

,

In this case, the length of the cleaning area and the outline is expressed by Equation 15.

The methodology for distributing each pattern using the above information and the characteristics of each pattern is as follows.

First, in the compound pattern cleaning step S6, as shown in FIG. 23, the random bounce pattern should be able to move to the farthest point from the starting point in order to maximize the roamability. In this case, since the longest distance is a movement from one vertex to a vertex located diagonally, the minimum driving time is as follows.

In order to secure the stable cleaning area of the area, which is the next pattern, the obstacle must not exist around the minimum driving time. Therefore, it is preferable to start the area pattern when moving to at least the minimum travel distance after the obstacle avoidance.

However, even if this condition is not satisfied, local cleaning should be performed when the cleaning time exceeds three times the minimum driving time to ensure fair roaming and cohesion.

In addition, if the cleaning robot 1 moves in a very narrow area, it must be smooth to escape. If the time of linear driving during the recent four obstacle detection is less than the maximum obstacle detecting linear driving time, the wall is easy to avoid. Run the ride pattern.

의 범위를 점차 늘려가도록 한다. As mentioned above, the coherence needs to be increased in the late stage of cleaning, so that the random bounce cleaning time is gradually reduced and random stamping is performed in the process of calculating the avoidance angle.

Gradually increase the range of.

In the complex pattern cleaning step S6, as shown in FIG. 24, the area cleaning may be broken due to cohesiveness. Therefore, when the obstacle is detected, it is preferable to escape through the wall riding pattern, and each pattern sets a maximum area cleaning time. When the cleaning of the area over the predetermined size is completed, the wall riding pattern is performed immediately.

In order to maximize the coherence at the end of cleaning, increase the running time on the obstacles and overcome the obstacles through obstacle avoidance and maintain the cleaning even if the obstacle is detected.

In the composite pattern cleaning step (S6), as shown in FIG. 25, the wall riding pattern needs to be cleaned for all the wall surfaces during the cleaning time, and thus, the wall riding pattern should be driven over a predetermined time in consideration of the total driving time.

However, as time passes, the overlapping area becomes larger, and according to the cleaning time, it is reduced to the minimum driving time necessary for escape of the abnormal terrain. This method naturally serves to increase the late cohesion of the cleaning.

The wall riding pattern is basically transferred to the next pattern when the maximum travel time is exceeded. In this case, when the cleaning robot to be considered follows an obstacle whose total length of the outline is very short, it is preferable to perform the following pattern since it is no longer necessary to follow the outline after one rotation. In order to do this, the rotation angle of the cleaning robot is integrated at every moment to obtain the current angle of the cleaning robot, and when it is rotated +/- 360 degrees from the time of performing this pattern, the next motion is performed.

As described above, while cleaning in the optimum pattern cleaning step (S6) and the danger space is recognized through each sensor unit, or if a problem occurs in the abnormal terrain and the robot for cleaning, the situation is quickly avoided according to the situation judgment If a problem that cannot be avoided occurs, the problem generated in the avoidance pattern step S7 for generating an error is handled.

On the other hand, the working space of the cleaning robot (1) not only has a dangerous space that is harmful to the user and the robot, there may be an abnormal terrain that the cleaning robot exhibits inoperable or abnormal operation, due to problems inside and outside the cleaning robot Problems may also occur at each site. The problem is summarized as follows.

When the danger space is recognized in the avoidance pattern step (S7), if the floor sensor 230 does not detect it, it is regarded as an obstacle when it occurs in the area, and it can be safely handled only by processing the obstacle, so that it can be random bounce or wall ride. In case of a pattern, it moves back and performs a dodging motion that rotates and continuously executes each pattern, and the amount of rotation is 180 degrees so as not to enter again, and the secondary problem is that each sensor detects the maximum time that is continuously detected. If the time exceeds 6 seconds to 9 seconds to notify the error and stop all the motion, if the wheel lifting sensor 250 displays the wheel lifted state, the floor sensor 230 did not detect the abnormal terrain or the user Because it is caused by, the operation stops completely and an error is notified when the maximum detection time exceeds 6 seconds ~ 9 seconds.

The reason for limiting the maximum detection time is that this range is the optimal condition to send an error by experimental value.

Since the cleaning robot 1 is almost incapable of detecting the sensor unit 200 when the wheel is idle due to the terrain, carpet or obstacle protruding from the bottom of the cleaning robot determined to be the abnormal terrain in the avoidance pattern step (S7). There is no way to solve this problem. To solve this problem, set the maximum driving time without obstacles and dodge the pattern if exceeded. The maximum driving time is 7 ~ 10m in the normal home, so the random bounce pattern In case of local pattern, it is effective to ensure that each movement is completed by ignoring this value, and in case of wall riding pattern, it is not detected continuously and rotates in one direction, and if it makes one turn, random bounce Because the transition to the pattern, it is possible to check naturally, the bumper of the cleaning robot (1) is continuously pressed In this case, it may be caused by an inability to recognize an obstacle due to a bumper failure, a user's intention, or an infrared sensor failure.In order to avoid it, it is preferable to perform a backward motion to escape completely to an obstacle. If you move more than / 2, you can fall on terrain without a floor, so if you move more than 1/2 of the diameter continuously, you can avoid the pattern, and if it occurs continuously, stop all the movements and inform you of the error. In the case where the infrared sensor 220 of (1) continuously detects an obstacle, it may be a failure of the infrared sensor, driving on a very narrow feature, and the cleaning robot is continuously rotated, depending on the nature of each pattern. Or driving on unusual terrain can also cause this problem. If the amount exceeds perform avoid pattern continues to occur, configured to notify an error to stop all motion.

In the case of the overload or failure of the brush motor 142 which is determined to have a problem inside and outside the cleaning robot 1 in the avoidance pattern step S7, a long carpet or hair may be caught, and the driving motor may be stuck. 132) in the case of overload and failure of the user may occur when the robot is intentionally pressed hard to the ground direction, in the case of overload and failure of the suction motor 144 may also occur in the closing of the flow path, cleaning robot ( If there is no solution to the cause of 1) self-defect, there is no solution, so it is determined that the error is caused by the topographical characteristics, and it rotates 150 degrees at the maximum speed, and reverses by 1/2 of the diameter of the cleaning robot at the maximum speed to avoid the emergency. It takes out of the terrain and acquires the overload signal input from the circuit that drives the motor, and the number of signals generated per unit time is continuously maintained for 3 to 5 seconds. If occurs, configured to notify an error to stop all motion.

The major cause of such an overload signal is topographical characteristics. In other words, the terrain is difficult to turn the brush or the wheel is difficult to rotate. In this case, it is necessary to get out of the special terrain, so if there is a problem, avoid the motion. If the problem continues to occur, notify the error and stop all operation. At this time, the condition for executing the avoidance motion is satisfied if the number of signals generated per unit time is obtained by acquiring an overload signal input from a circuit for driving each motor. Threshold values can vary because the circuit design and motor specifications can vary, but 3 to 5 seconds is appropriate because the sampling time is too large and the hardware can be damaged.

As described above, the cleaning robot and the cleaning method using the same according to the present invention are configured to detect the position of the obstacle by detecting the contacted position by the sensor plate when the obstacle is in contact with the bumper formed to cushion the shock, and at the center of the front surface. Infrared sensor is formed so as to be installed on the side side to detect obstacles in the front and side, and bottom detection sensor to detect the ground at the bottom, encoder to measure the speed of the running wheel and wheel lifter to detect the lifting of the wheel When the sensor unit that has formed the sensor transmits information about the driving area to the central processing unit, the random bounce pattern, the zigzag pattern, the wave pattern, the square round pattern, the star shape pattern, the round round pattern, and the wall riding pattern are set to match each situation. Or various combinations of cleaning to effectively avoid obstacles and obstacles. Expanding the cleaned region and to minimize duplication of cleaning zones, and provides the effect of increasing the efficiency of cleaning.

In addition, in the cleaning method of the cleaning robot of the present invention, the infrared sensor installed radially on the front side is recognized as an obstacle between the two when detected together according to the area detected by each other effectively reduce the amount of sensors installed to reduce the manufacturing cost It provides an effect.

In addition, since the cleaning method of the cleaning robot of the present invention uses only the recognition of obstacles and obstacles in proximity to the sensor unit, the system configuration can be simplified, the cost can be reduced, and the cleaning efficiency can be increased by combining various cleaning methods. In addition, it can adapt to various features, resulting in the effect of maximizing the cleaning efficiency.

In addition, the cleaning method of the cleaning robot of the present invention, in order to achieve the ultimate goal of cleaning all the areas inside the cleaning area as quickly as possible, without missing any part of the cleaning robot, the cohesiveness is carried out by agglomeration in a certain area. Two measures: coherence, which is a property of dense movement of a specific pattern without overlapping specific areas, and roaming property, which is a property that a robot which performs cleaning at a proper speed while transferring a pattern quickly according to a situation does not stay in one area but moves to other areas. By defining the directionality of the transformation and the driving of various patterns of the cleaning robot, all the cleaning areas are evenly cleaned without empty areas in the minimum time provides the effect of maximizing the cleaning efficiency.

In addition, the cleaning method of the robot for cleaning according to the present invention inputs a variety of patterns in advance to the central processing unit, implements a suitable pattern through various proximity information sensed by the sensor unit, and the pattern is transferred in real time, thereby effectively avoiding obstacles, thereby improving cleaning efficiency. It gives rise to an effect.

Claims (16)

In the upper part, the cleaning pattern and the cleaning time are operated by the operation button 111, and the remote control unit 113 is operated at a long distance, and the operation unit 110 which displays the operation state and operation information on the display unit 112 is formed, A bumper 120 having a shock absorber 121 for cushioning shock and an actuating hole 122 for transmitting a shock is formed, and the driving wheel 131 rotates while the driving wheel 131 is driven by the driving motor 132 in the lower center. And forming a dust collecting part 140 for collecting dust by sucking the suction part 143 formed with the suction motor 144 while rotating the brush 141 with the lower brush motor 142. And a driving cleaning unit 100 including a power supply unit 150 for supplying power to the rear portion, Forming a sensing plate 210 that is energized in the form of a thin film so as to be interlocked with the operation hole 122 of the bumper 120 of the driving and cleaning unit 100 and formed to detect the obstacle contact to the front, and radially on the bumper 120 The infrared sensor 220 is installed, but the infrared sensor 220 is installed on both sides of the installation intervals more densely than the center of the front of the front, and installed on the lower front to face the ground in a plurality of floor detection to detect the obstacles A sensor 230 is formed, and an encoder 240 for measuring the speed of the driving wheel 131 rotating with the driving motor 132 and the wheel lifting sensor 250 for detecting the lift of the wheel are formed on the driving unit 130. One sensor unit 200, Based on the information detected by the sensor unit 200 to calculate the optimal driving pattern to generate a signal to the control unit 310 to control the running of the driving cleaning unit 100 to clean the central processing unit 300 Cleaning robot comprising a. The method of claim 1, The infrared sensor 220 of the sensor unit 200 is installed radially and needs a large avoidance angle when it is detected from the front according to the area detected by each other, and requires a relatively small avoidance angle when detected from the side. The more compactly formed toward the side, and when the two are detected together, the cleaning robot, characterized in that the installation and configuration to recognize that there is an obstacle between the two infrared sensors 220. The cleaning setting step (S1) for setting the cleaning environment while watching the display unit 112 displaying the status at the upper portion of the cleaning time, the cleaning pattern with the operation button 111 at a short distance, and the remote control 113 at a distance, and A cleaning start step (S2) for starting cleaning by transferring to a starting point started according to the cleaning time and cleaning pattern set in the cleaning setting step (S1); In the cleaning start step (S2), if the cleaning starts to be transported to the starting point, the random bounce pattern, zigzag pattern, wave pattern, whirlwind pattern, square rounding pattern, star shape pattern, circular rounding pattern and wall riding pattern A pattern cleaning step (S3) for cleaning according to the pattern set in the operation unit 110 by combining one or various, When cleaning the pattern specified in the pattern cleaning step (S3) and the sensor unit 200 detects an obstacle, the control unit 310 is processed by the central processing unit 300 according to the position and shape of the obstacle to the optimal pattern avoided. Obstacle avoiding step (S4) to run while avoiding obstacles in accordance with the signal of), While avoiding obstacles in the obstacle avoidance step (S4), when reaching a obstacle such as a depression or a stair, the bottom detection sensor 230 of the sensor unit 200 detects the state of the ground and avoids the center in a pattern. Disability terrain avoidance step (S5) for converting in the processing unit 300 to run in an optimal pattern to avoid obstacle terrain, In the obstacle terrain avoidance step (S5), if the area of the cleaning place is identified while avoiding and driving the obstacle terrain, a random bounce pattern, a zigzag pattern, a wave pattern, a whirlwind pattern, a square rounding pattern, a star shape pattern, a circular rounding pattern and a wall ride Compound pattern cleaning step (S6) to perform a complex cleaning with the optimal cleaning pattern by combining to compensate for the shortcomings of the pattern, While cleaning with the optimal pattern in the compound pattern cleaning step (S6), while perceiving dangerous space through each sensor unit, or if a problem occurs inside and outside the terrain and robot, the risk is quickly avoided and avoided according to the situation judgment. Cleaning method of a robot for cleaning, characterized in that it comprises a avoidance pattern step (S7) for generating an error if a problem does not occur. The method of claim 3, wherein In the pattern cleaning step (S3), the random bounce pattern is basically a straight line, and the radius of the cleaning robot 1, the maximum sensing distance of the sensor unit 200, and the stopping distance after sensing the obstacle according to the angle of entering the obstacle. Based on the correlation, the set angle set in Equation 1 below and the minimum avoidance angle in which obstacles set in Equation 2 below are not detected are determined. A cleaning method for a robot for cleaning, characterized in that cleaning is carried out while adjusting the over roamability. (Equation 1)

Where the set angle is

, The radius of the cleaning robot (1)

cm, the maximum sensing distance of the front infrared sensor 220

cm, the distance traveled straight until it stopped after detecting the obstacle

cm. (Equation 2)

Here, the minimum evacuation angle is ∠avoidance, and the set angle is

, Random angle

to be. The method of claim 3, wherein In the pattern cleaning step (S3), the zigzag pattern starts to rotate to the opposite side where the obstacle is detected by the sensor unit 200, and when the side infrared sensor is detected while rotating, the rotation is stopped and the angle rotated to the opposite side of the obstacle is calculated. After going straight along a certain distance, if you run by rotating the circumference divided by the angle divided by the angle and the rotation angle, the difference between the previous entry angle and the avoiding angle is small, so it generates a low deflection trajectory, and follows the wall surface at a certain distance straight ahead. While cleaning is performed, the cleaning method for a cleaning robot, characterized in that configured to avoid the rotation by the opposite side of the corner when the corner is detected by the sensor unit. The method of claim 3, wherein In the pattern cleaning step (S3), the wave pattern travels in a straight line at a predetermined distance, rotates in a circle in a moving direction, and turns in a parallel opposite direction. When an obstacle is detected, the cleaning robot 1 is reversed to secure a rotation radius. After driving the car while avoiding obstacles, in the bottleneck where continuous obstacles are detected, the first obstacle is avoided, and after avoiding other continuous obstacles, the straight distance is reduced to 1/2 to move the narrow area. The method of cleaning a robot for cleaning characterized in that when reaching the edge-shaped terrain, the straight line is shortened and this distance is smaller than the predetermined turning distance, thereby avoiding obstacles while changing the traveling direction. The method of claim 3, wherein In the pattern cleaning step S3, the tornado pattern is rotated while calculating the speed of the rotating wheel 131 by setting the initial radius of rotating the first spiral shape from the center of the set cleaning area, and rotating it by 180 degrees. By increasing the radius of rotation by the spiral trajectory, by applying the increase of the increase in the radius of rotation to the rotational speed of the driving wheel 131 increases the rotational speed of the running wheel 131 according to the increase of the angle every 180 degrees without calculating an approximation Cleaning method of the robot, characterized in that configured to drive while reducing the cumulative angle of the error range according to the approximation. The method of claim 3, wherein In the pattern cleaning step (S3), the square rounding pattern rotates by increasing the radius by an increase of the radius rotated in a state in which the rectilinear distance is set in the initial radius and goes straight at a predetermined distance while being rotated 90 degrees in the rotational direction. Repeatedly expanding to a wide rectangular shape and reaching an obstacled space, avoiding it while traveling in a rectangular shape, creating an overlapping area regardless of the shape of the obstacle, and returning to the square rounding pattern after showing a pattern that follows the shape of the obstacle. While cleaning so as to avoid obstacles cleaning robot cleaning method characterized in that the driving. The method of claim 3, wherein In the pattern cleaning step (S3), the star-shaped pattern goes straight to a certain distance and then rotates by the angle subtracted from the circumferential ratio from the angle distributed according to the number of repetitions, and travels in the star shape which repeats the same operation n times while going straight for a certain distance. Move the cleaning robot (1) from the center point of the star shape to the start point and from the end point of the pattern to the center point of the star shape. Cleaning method for a cleaning robot, characterized in that configured to run after rotating to the angle calculated by the formula (2) using the angle a calculated by the following equation (1) to perform. (Equation 1)

Here, the calculation angle is a, the number of rotations is n. (Equation 2)

Here, the calculation angle is a. The method of claim 3, wherein In the pattern cleaning step S3, the circular rounding pattern is cleaned in a circular pattern having a diameter of half the radius of the cleaning area in a state in which the circular cleaning area is set, and returns to the original position as shown in Equation 1 below. Repeat the circular cleaning with the set number of times using the half of the radius of the cleaning area as the diameter while rotating at the angle distributed according to the pattern number of times set by, and rotate 180 degrees when encountering the obstacle. A cleaning method of a cleaning robot, characterized by traveling while avoiding obstacles by performing a normal next pattern while moving. (Equation 1)

Here, the distributed angle is β and the number of executions is n. The method of claim 3, wherein In the pattern cleaning step (S3), the wall riding pattern is set by Equation 1 below to set the maximum forward speed of the cleaning robot 1, and performs linear motion. By determining and controlling as shown in Equation 1 below, the forward speed is decreased while increasing the amount of rotation to the opposite side of the detected obstacle, and if there is no obstacle, the pattern of the obstacle is repeated while following the detected wall and increasing the forward speed. Drive along the road, drive the outside while changing the forward or rotational speed according to the position of the obstacle in the process of identifying and applying the obstacle, and use the coherent motion following the outside line to secure the roaming, but when c = 0 The cleaning furnace, characterized in that to return to the place, so that c = 0 and the moving distance (s) is greater than the threshold Tw to end. How to clean a bot. (Equation 1)

(Equation 2)

,

,

,

Here, the running speed of the equations (1) and (2) is

, Rotation speed

The gain matrix for the rotational speed and the forward speed

The gain for the rotational speed and the forward speed for the side, diagonal, and front

,

,

,

,

,

, The error matrix is

Side

, Diagonal

, face

Is the number of times the sensors in front of the vehicle are continuously detecting or not detecting obstacles, and the maximum forward speed is

to be. The method of claim 3, wherein In the obstacle-avoidance step (S5), the floor detection sensor 230 of the sensor unit 200 has a predetermined distance between the floor surface and the lower part of the cleaning robot 1 so as to detect the state of the terrain on which the cleaning robot 1 runs. Set the target distance and drive, and if you reach the irregular abnormal topography such as the impossible topographic map, the threshold and the carpet, it detects the abnormal topography and performs the avoidance pattern, and the distance between the ground momentarily becomes longer than the target distance. If the sensor 230 does not detect the floor at a very short moment, it ignores the terrain and decelerates the maximum speed of the cleaning robot 1 so that it does not detect if it does not detect for more than the set time. Cleaning method of a robot for cleaning, characterized in that configured to avoid while rotating. The method of claim 3, wherein In the composite pattern cleaning step (S6) to travel the longest distance from the vertex to move to the farthest point from the starting point to maximize the roamability to the maximum running time by moving to a vertex located diagonally, moving a certain distance in the presence of obstacles Zigzag, wave, tornado, square round, star shape and circular rounding are integrated, and the local pattern is executed immediately after exceeding 3 times the minimum driving time to ensure fair distribution and coherence. When passing through a narrow place due to the detection of the unit 200, it switches to a wall riding pattern, and while performing an area pattern, detects an obstacle or runs a wall riding pattern when cleaning a predetermined area, and maximizes cohesiveness at the end of cleaning. The local pattern is maintained while avoiding obstacles, and the wall riding pattern is used to clean all walls. For this purpose, set the wall cleaning time at the total cleaning time and drive it, but decrease it according to the cleaning time until the minimum driving time according to the area where the cleaning overlaps, and if it exceeds the maximum driving time of the wall riding pattern, transfer to the next pattern for cleaning. As the process proceeds, the area of cleaning is divided according to the specification of the cleaning robot 1, the time set for each pattern is distributed, the cleaning pattern is set to the expected position according to the distributed time, and the shape of the obstacle and Accurately grasp the shape and size according to the sensing frequency of the sensor unit 200, and present two patterns of coherence and roaming properly, and combine them in the order of traveling through wall patterns through local patterns through random bounce patterns. A cleaning method for a cleaning robot, characterized in that configured to perform cleaning. The method of claim 3, wherein When the danger space is recognized in the avoidance pattern step S7, when another sensor unit 200 detects the floor sensor 230 in a state where the floor sensor 230 is not detected, it is regarded as an obstacle and is a predetermined distance from the random bounce or wall riding pattern. Carry out the cleaning by performing the avoidance pattern rotated 180 degrees so as not to re-enter the re-entry, and sends an error if the sensor unit 200 continuously detects that the obstacle exceeds the set maximum detection time 6 ~ 9 seconds When the wheel lifting state is detected by the wheel lifting sensor 250, the bottom detection sensor 230 does not detect the abnormal terrain, and the wheel is lifted by the user and stops the operation completely. Cleaning method for a cleaning robot, characterized in that configured to send an error if the time exceeding 6 ~ 9 seconds. The method of claim 3, wherein In the avoidance pattern step (S7), the bottom of the cleaning robot (1), which is determined to be an abnormal topography, the terrain, the carpet, the obstacle protrudes due to the wheel, the sensor unit 200 determines that there is no obstacle in the state that can not be detected. Therefore, the random bounce pattern is set according to the area of cleaning, and in the case of local patterns, the area is ignored by the cleaning area, and in the case of the wall riding pattern, it is not continuously detected and rotates in one direction and makes one turn. If it exceeds the set time, the evacuation pattern is executed. If the bumper of the cleaning robot 1 is continuously pressed, the bumper failure, user intention, It is determined that the obstacle cannot be recognized due to the failure of the infrared sensor, and the vehicle is driven backward by a certain distance to avoid the obstacle. In order to avoid falling by moving more than 1/2 of the diameter of the cleaning robot 1 in the abnormal terrain without the bottom surface, the evacuation pattern is performed by continuously moving 1/2 or more of the diameter. Cleaning method for a cleaning robot, characterized in that configured to send and stop operation. The method of claim 3, wherein Overload and failure of the brush motor 142 caused by the foreign matter caught in the brush 141 of the cleaning robot 1 in the avoidance pattern step (S7) and the user deliberately generated when the robot is pressed hard to the ground direction Detects the overload and failure of the suction motor 144 caused by the overload and failure of the driving motor 132 and the closing of the flow path, and if it is determined that the cleaning robot 1 has its own failure, it sends an error and stops the operation. If it is determined by the topographical characteristics, it rotates 150 degrees at the maximum speed, reverses by 1/2 of the diameter of the cleaning robot (1) at the maximum speed, and escapes the terrain while avoiding the emergency. Each motor (132, 142, 144) It is configured to stop all the operation and send an error when the number of signals generated per unit time is continuously acquired by acquiring the overload signal input from the driving circuit. Cleaning the cleaning robot.

Images