A 2D radar-based transition and planning method for a stair cleaning robot between stairs and rest platforms

Abstract

The present invention relates to the field of robots and intelligent control technology, and in particular to a transition and planning method between stairs and rest platforms of a stair cleaning robot based on 2D radar, comprising: step S1: transition of the robot from stairs to rest platforms; step S2: analysis of landing points, determination of starting points and robot turning at the starting points; step S3: traversal of rest platforms; step S4: determination of target points; step S5: transition of the robot from the rest platform to the next level of stairs. The present invention applies 2D laser radar to the stair cleaning robot, so that the stair cleaning robot can realize detection and identification from stairs to rest platforms; efficient cleaning on the rest platform; and transition from the rest platform to the next level of stairs, thereby ensuring that the robot can better serve the cleaning work of high-rise buildings.

Description

Technical Field

[0001] The present invention relates to the field of robots and intelligent control technologies, and in particular to a transition and planning method between stairs and a rest platform of a stair cleaning robot based on a 2D radar. Background Art

[0002] In recent years, my country's urbanization has achieved remarkable achievements. Urban buildings have become the living and working space for most people. The cleaning of stairs in high-rise buildings is very complicated, and the flat-ground sweeping robots on the market cannot achieve the cleaning of stairs. Therefore, the research of safe, reliable, and efficient mobile stair cleaning robots is helpful to improve the efficiency of floor and stair cleaning, improve people's living and working environment, and improve the sanitation of the city; it is also of great significance to reduce the work pressure of sanitation workers, accelerate the formation of intelligent building systems, and accelerate the pace of building an intelligent society.

[0003] Lu Guiping and others designed a stair-adaptive cleaning robot that uses a planetary wheel mechanism and a Mecanum wheel to achieve flat ground walking and stair climbing. It uses infrared obstacle avoidance sensors and ultrasonic sensors to detect surrounding environment information, achieves certain path planning, and can avoid obstacles and prevent falls. However, there are problems such as instability when going downstairs and poor adaptive ability.

[0004] Tu Futao and others designed an intelligent stair cleaning robot that uses a ball screw structure to achieve stair climbing, and uses ultrasonic sensors, tentacle sensors, grayscale sensors and Arduino controllers to collect and process data. The robot can achieve climbing, obstacle avoidance and cleaning functions, but has poor climbing stability and cannot achieve autonomous navigation.

[0005] Although the rest platform is a flat ground, it is different from ordinary flat ground. Although the general flat ground path planning algorithm can realize traversal, the traversal end point is random, which is not conducive to the robot's search for the stairs down; secondly, even if it is found, the entire traversal and search process is bound to be time-consuming, resulting in a large waste of energy, or the robot needs to be equipped with high-end sensors, which increases the cost and is not conducive to the promotion of the robot. For this reason, the stair sweeping robot uses a low-cost and highly reliable laser radar as a sensor, which improves the robot's ability to collect data on the rest platform, provides a basis for further development of functions, and is conducive to further promoting the robot to the market. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings existing in the prior art, and proposes a transition and planning method between stairs and rest platforms for a stair cleaning robot based on 2D radar. By applying 2D lidar to the stair cleaning robot, the stair cleaning robot can realize detection and identification from stairs to rest platforms; efficient cleaning on the rest platform; and transition from the rest platform to the next level of stairs, thereby ensuring that the robot can better serve the cleaning work of high-rise buildings.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A 2D radar-based transition and planning method for a stair cleaning robot between stairs and a rest platform comprises the following steps:

[0009] Step S1: Transition of the robot from the stairs to the rest platform;

[0010] The robot cleans by going downstairs, so the robot detects whether it is a staircase or a rest platform downwards, and determines it through the first sounding sensor and the second sounding sensor; if the robot moves forward a certain distance, the first sounding sensor and the second sounding sensor both sense that there is no obstruction below, which means that the robot currently needs to determine whether the front is a rest platform or a staircase; after the robot finishes cleaning the current step, the 2D laser radar tilts a certain angle, and the 2D laser radar uses triangulation technology to output the distance value and angle, and can obtain the angle and distance data of the obstacle to achieve obstacle identification;

[0011] Step S2: Analyze the landing point, determine the starting point and the robot turns around at the starting point;

[0012] The robot reaches the rest platform from the stairs. There are two landing points, including the first position and the second position. The robot uses the radar to sense the distance to the wall to the left. If the distance is small, it is judged to be at the first position; if the distance is large, it is judged to be at the second position. Taking the first position as the starting point, the position of the robot at the second landing point needs to be offset by the length of the stairs to the first position.

[0013] Step S3: traverse the rest platform;

[0014] After the robot reaches the rest platform, it will traverse the rest platform, and use the improved plow path planning method through 2D laser radar to achieve the traversal of the rest platform;

[0015] The plough path planning means that the robot starts from a certain point and moves in a straight line in a specific direction. When it encounters a boundary, the robot stops moving and rotates 90 degrees at the current position, then moves forward one body position, then remains still and rotates 90 degrees again, and finally continues to move forward. This method is repeated until the entire area is traversed.

[0016] Among them, the rest platform area calculation: A = l*w, A is the rest platform area, l is the length of the rest platform, w is the width of the rest platform;

[0017] Calculation of rest platform distance: D = (P*N) / S, where D is the rest platform distance, P is the number of rows of rest platforms, N is the length of each row, and S is the robot running speed;

[0018] Estimation of rest platform operation time: T = A / (R*E), where T is the tillage time, A is the tillage area, R is the rate of the robot cleaning operation, and E is the efficiency coefficient;

[0019] Step S4: determine the target point;

[0020] After continuous cycles, the stair cleaning robot can complete the cleaning traversal of the floor rest platform. When the distance in the 0° direction measured by the radar is less than the length of the robot, it stops moving and cleaning.

[0021] Step S5: the robot transitions from the rest platform to the next staircase;

[0022] When the robot reaches the target point, if the radar above the robot tilted at an angle of μ detects H1 = h*m, and the sounding sensor detects nothing downward, it means that the robot has reached the intersection of the rest platform and the next staircase, and there is a downward staircase in front.

[0023] On the contrary, when H2 detected by the radar is less than h*m, it indicates that the robot has not yet reached the intersection of the rest platform and the next staircase, and needs to move forward for a certain distance until the distance detected by the radar is H2 before it can go downstairs, thereby completing the stair cleaning and the transition between the stairs and the rest platform.

[0024] Preferably, in step S1, the specific method includes:

[0025] The radar can directly measure the distances m1, m2, and m3, and obtain θ1, θ2, and θ3 on the established coordinate system. If the radar can detect two or more line segments with different distances, L1=m1*sinθ1, L2=m2*sinθ2, and L1≠L2, then it is judged that there are still stairs in front of the robot; on the contrary, when the radar always detects a line segment with a distance of L3, L3=m3*sinθ3, then it is judged that there is a rest platform below, and the robot should complete the going downstairs action and then enter the rest platform planning control state.

[0026] Preferably, in step S2, the specific method includes:

[0027] Step S21: the robot moves from a position D1 away from the stair riser to a position D2 away from the stair riser in a direction perpendicular to the stair riser at the first position;

[0028] Step S22: After ensuring that D2 = D1 + half of the fuselage length, rotate counterclockwise;

[0029] Step S23: Move the robot to a position D1 away from the stair riser in a direction perpendicular to the stair riser to ensure that the corner can be effectively cleaned.

[0030] Preferably, in step S3, the specific method for the robot to traverse the route on the rest platform includes:

[0031] Step S31: Assuming that the stair cleaning robot transitions to the floor rest platform after completing the cleaning of the stair section, the robot body first rotates 90° clockwise;

[0032] Step S32: After the robot rotates 90°, the distances in the 0° direction and the 270° direction are measured by the radar, which are recorded as Length and Width. By setting the expected left and right movement speed of the robot, the time t = Width / speed for each left and right movement can be obtained according to the measured distance Width. The length of each forward movement of the robot is equal to the width of the robot, which is also achieved by controlling the speed and running time in the program.

[0033] Step S33: After the first left movement, the robot is ready to move forward, and a variable Count is set to record the number of forward movements, and the initial value of Count is 0;

[0034] Step S34: After the first forward movement, the robot is ready to move right, at which time Count+1, Count=1;

[0035] Step S35: After the first rightward movement, the robot is ready to move forward;

[0036] Step S36: After the second forward movement, Count is increased by 1, and the value of Count is 2 at this time.

[0037] Preferably, in step S3, the specific method for handling obstacles encountered by the robot on the rest platform includes:

[0038] When the stair cleaning robot is working on the floor rest platform: Assuming that the robot cleans according to the plowing path planning algorithm, when the robot moves to the left, it detects points within a range greater than 240° and less than 300° through the 2D laser radar, and counts points with a distance less than 250mm to set the obstacle avoidance distance. If the total number of counted points is less than 5, there is no need to avoid obstacles; otherwise, it is judged that the obstacle is on the left side of the robot; when the robot moves to the right, it detects points within a range of 60° to 120° through the 2D laser radar, and counts points with a distance less than 250mm to set the obstacle avoidance distance. If the total number of counted points is less than 5, there is no need to avoid obstacles, otherwise, it is judged that the obstacle is on the right side of the robot;

[0039] On the basis of realizing the position discrimination of obstacles on the floor rest platform, a new obstacle avoidance strategy is proposed: on the floor rest platform, assuming that an obstacle appears in the left direction of the robot, the robot will slow down when it detects the obstacle, then move forward 250mm until no obstacle is detected on the left, then move to the left, finally bypass the obstacle and continue cleaning; assuming that an obstacle appears in the right direction of the robot, the robot will slow down when it detects the obstacle, then move forward 250mm until no obstacle is detected on the right, then move to the right, finally bypass the obstacle and continue cleaning.

[0040] Preferably, in step S4, the specific method includes:

[0041] When the robot uses the improved plow path planning to clean the sides and corners of the rest platform, the radar can detect that the distance between the robot and the wall is D, and establish a plane coordinate system on the radar, defining the initial position of 0°. The angles θ1, θ2, and θ3 of the laser beam can be refracted. According to the geometric theorem, d1=D*tanθ1, d1+d2=D*tanθ2, and d1+d2+d3=D*tanθ3 are obtained, and the coordinates of point 1, point 2, point 3, and point n are obtained. If it is found that the difference in the x-coordinate is less than or equal to 5mm, it means that these points are on a straight line, and the y-coordinate is an arithmetic progression. Through fitting, it can be obtained that the laser beam refracted by the radar at this time is a uniformly distributed point, which proves that it is a wall.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The present invention applies 2D laser radar to the stair cleaning robot, so that the stair cleaning robot can realize detection and identification from the stairs to the rest platform; efficient cleaning on the rest platform; and transition from the rest platform to the next level of stairs, thereby ensuring that the robot can better serve the cleaning work of high-rise buildings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Schematic diagram of the structure of the stair cleaning robot in the present invention;

[0045] Figure 2 It is a schematic diagram of the radar plane coordinate system in the present invention;

[0046] Figure 3 A schematic diagram of the landing point of the robot on the rest platform in the present invention;

[0047] Figure 4 A schematic diagram of the plow path planning in the present invention;

[0048] Figure 5 A schematic diagram of the floor rest platform robot in the present invention reaching the target point position;

[0049] Figure 6 A schematic diagram of the robot in the present invention transitioning from a rest platform to a staircase;

[0050] Figure 7 The robot radar in the present invention determines whether it has reached the rest platform;

[0051] Figure 8 A schematic diagram of the rotation details of the robot at the starting point in the present invention;

[0052] Fig. 9 A schematic diagram of the cleaning path of the robot on the floor rest platform in the present invention;

[0053] Fig.10 A schematic diagram of obstacle avoidance path planning of the robot on a floor rest platform in the present invention;

[0054] Fig.11 This is a schematic diagram of the robot radar identifying a wall in the present invention. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings, so that those skilled in the art can better understand the advantages and features of the present invention, thereby making a clearer definition of the protection scope of the present invention. The embodiments described in the present invention are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0056] Reference Figure 1 The present invention adds a laser 2D radar to the structure of the stair cleaning robot. Figure 1 The three-view schematic diagram shows that the robot body is equipped with four Mecanum wheels symmetrical to the left and right middle lines of the body, which can move in all directions. When turning on the spot, the Mecanum wheels can greatly reduce lateral slippage. A laser 2D radar with adjustable tilt angle is installed directly above the robot body to provide road information for the robot to transition from the rest platform to the stairs. The first and second sounding sensors are installed at the front of the robot body. The first and second sounding sensors are used to detect whether the front is suspended and whether there are stairs leading down, so as to complete the transition from the stairs to the rest platform and from the rest platform to the stairs.

[0057] A 2D radar-based transition and planning method for a stair cleaning robot between stairs and a rest platform comprises the following steps:

[0058] Step S1: Transition of the robot from the stairs to the rest platform;

[0059] The robot goes downstairs to clean, so the robot detects whether it is a staircase or a rest platform. In order to make the robot smoothly transition from the stairs to the rest platform, the first sounding sensor and the second sounding sensor are used to distinguish. If the robot moves forward a certain distance, the first sounding sensor and the second sounding sensor both sense that there is no obstruction below, which means that the robot currently needs to determine whether the front is a rest platform or a staircase. After the robot finishes cleaning the current step, the 2D laser radar is tilted at a certain angle, and the distance measurement characteristics such as Figure 2 2D LiDAR uses triangulation technology to output distance values ​​(the actual distance between the rotation center and the sampling point) and angles (the angle between the current sampling point and the radar's own orientation). Obstacles can be identified by obtaining the angle and distance data of obstacles.

[0060] Step S2: Analyze the landing point, determine the starting point and the robot turns around at the starting point;

[0061] The robot reaches the rest platform from the stairs. There are two landing points, such as Figure 3 As shown, it includes the first position 1 and the second position 2. Among them, the rotary arms of the robot going downstairs of the present invention are on the left and right sides of the robot, so the posture of going down the stairs from the upper floor can only be along the direction of the stair riser, that is, Figure 3As shown in the first position 1 or the second position 2. Therefore, it is necessary to first determine whether it is in the first position 1 or the second position 2. The radar can be used to sense the distance to the wall to the left. If the distance is small, it is determined to be in the first position 1; if the distance is large, it is determined to be in the second position 2; taking the first position 1 as the starting point, the position of the robot at the landing point second position 2 needs to be offset by the length of the stairs to the first position 1;

[0062] Step S3: traverse the rest platform;

[0063] After the robot reaches the rest platform, it will traverse the rest platform, and use the improved plow path planning method through 2D laser radar to achieve the traversal of the rest platform;

[0064] The plow path planning means that the robot starts from a certain point and moves in a straight line in a specific direction. When it encounters a boundary, the robot stops moving and rotates 90 degrees at the current position, then moves forward one body position, then remains still and rotates 90 degrees again, and finally continues to move forward. This method is repeated until the entire area is traversed. Figure 4 shown.

[0065] The calculation of the rest platform area is: A = l*w, A is the area of ​​the rest platform (usually in square meters), l is the length of the rest platform (meters), and w is the width of the rest platform (meters);

[0066] Calculation of rest platform distance: D = (P*N) / S, where D is the rest platform distance (meters), P is the number of rows of rest platforms, N is the length of each row, and S is the robot running speed (meters / second);

[0067] Estimation of the rest platform operating time: T = A / (R*E), T is the tillage time (seconds or minutes), A is the tillage area, R is the rate of the robot's cleaning operation (usually the area that can move per second or per minute), and E is the efficiency coefficient; factors such as turning and empty travel are taken into account.

[0068] Step S4: determine the target point;

[0069] After the above continuous cycle, the stair cleaning robot can complete the cleaning traversal of the floor rest platform. When the distance measured by the radar in the 0° direction is less than the length of the robot, it stops moving and cleaning. The robot's stop position at this time may be Figure 5 (A) If the value of the robot's forward movement count is an odd number, the robot can directly rotate 90° clockwise to reach Figure 5 (C) The robot’s stop position may also be Figure 5(B) If the robot moves forward and the Count value is an even number, the robot will move right again and reach Figure 5 (A) and then rotate 90° clockwise to reach Figure 5 (C) The position shown makes the front of the robot face down one staircase.

[0070] Step S5: the robot transitions from the rest platform to the next staircase;

[0071] A standard stair height h (the height of the vertical part of the step) is usually between 150 mm and 180 mm, and the standard height between floors (i.e., floor height) is usually between 2.5 m and 3.0 m. Taking a common floor height of 2.7 meters as an example, if the standard step height of 160 mm is used, the total rise distance of the stairs is 2.7 meters, and the number of steps required is approximately: the number of steps required is m = 16.875, which requires approximately 17 steps. A short stairwell may only require 8 steps.

[0072] When the robot reaches the vicinity of the target point, if the radar at an angle of μ above the robot detects H1=h*m, considering that some stairwells may be relatively short and the number of stairs is not large, the actual minimum value is lowered to 150mm*8=1200mm (generally the height should be 180mm*17=3060mm), and the threshold value of h*m is: 1.2m~3.06m. In actual work, the robot can set a suitable threshold value according to the actual stairs. At the same time, if the sounding sensor detects empty downwards, it means that the robot has reached the intersection of the rest platform and the next floor of stairs, and there are stairs leading down ahead; if Figure 6 (a) shown.

[0073] On the contrary, if Figure 6 As shown in (b), when H2 detected by the radar is less than h*m, it indicates that the robot has not yet reached the intersection of the rest platform and the next staircase. It needs to move forward for a certain distance until the distance detected by the radar is H2 before it can go downstairs, thereby completing the stair cleaning and the transition between the staircase and the rest platform.

[0074] Specifically, in step S1, the specific method includes:

[0075] like Figure 3 As shown, the radar can directly measure the distances m1, m2, and m3, and obtain θ1, θ2, and θ3 on the established coordinate system. If the radar can detect two or more line segments with different distances, such as Figure 7(a) shows the line segment with distances L1 and L2. L1=m1*sinθ1, L2=m2*sinθ2, and L1≠L2, so it is judged that there is still a staircase in front of the robot. Considering that it is difficult to be absolutely equal in practice, the actual operation can be judged by the large difference between the two. On the contrary, if Figure 7 As shown in (b), when the radar always detects a line segment with a distance of L3, L3 = m3*sinθ3, it is judged that the rest platform is below, and the robot should complete the downstairs action and then enter the rest platform planning control state. Considering that it is difficult to be absolutely equal in practice, the difference between the two can be judged in operation. The specific value depends on the actual situation.

[0076] In order to realize the plow-type planning proposed below, the robot needs to rotate 90° at the first position 1. When the robot rotates 90° at the first position 1, factors such as the size of the robot body, the length of the side brush, and the turning radius need to be considered. In step S2, the specific method includes:

[0077] Step S21: Figure 8 As shown, the robot moves from a position D1 away from the stair riser to a position D2 away from the stair riser in a direction perpendicular to the stair riser at a first position 1;

[0078] Step S22: After ensuring that D2 = D1 + half of the fuselage length, rotate counterclockwise (turn around);

[0079] Step S23: Move the robot to a position D1 away from the stair riser in a direction perpendicular to the stair riser to ensure that the corner can be effectively cleaned.

[0080] Specifically, in step S3, the specific method for the robot to traverse the route on the rest platform includes:

[0081] Step S31: Assume that the stair cleaning robot transitions to the floor rest platform after completing the cleaning of the stair section, and the starting position is as follows: Fig. 9 As shown in (a), the robot body must first rotate 90° clockwise; this is because the rotary arms are located on the left and right sides of the robot, which will block the cleaning area at the edge of the wall;

[0082] Step S32: After the robot rotates 90°, the position of the robot is as follows: Fig. 9 As shown in (b), the distances in the 0° direction and the 270° direction are measured by the radar, which are recorded as Length and Width. By setting the expected left and right movement speed of the robot, the time t = Width / speed for each left and right movement can be obtained according to the measured distance Width. The length of each forward movement of the robot is equal to the width of the robot, which is also achieved by controlling the speed and running time in the program.

[0083] Step S33: After the first left shift, arrive at Fig. 9 (c) The robot is ready to move forward (the robot does not need to turn around to switch from left movement to forward movement. It can directly use the characteristics of the Mecanum wheel to switch the wheels on the same side from rotating in the same direction to rotating in opposite directions). A variable Count is set to record the number of forward movements. The initial value of Count is 0.

[0084] Step S34: After the first forward movement, the Fig. 9 (d) The robot is ready to move right, and Count+1, Count=1;

[0085] Step S35: After the first right shift, arrive at Fig. 9 (e) The robot is then ready to move forward (there is no need to turn around to switch from right movement to forward movement, as the characteristics of the Mecanum wheel are directly utilized);

[0086] Step S36: After the second forward movement, the robot reaches Fig. 9 At the position shown in (f), Count is increased by 1, and the value of Count is 2.

[0087] Specifically, in step S3, the specific method for handling obstacles encountered by the robot on the rest platform includes:

[0088] When the stair cleaning robot is working on the floor rest platform: Assuming that the robot cleans according to the plowing path planning algorithm, the robot detects points within the range greater than 240° and less than 300° through the 2D laser radar when moving left, and counts the points with a distance less than 250mm set obstacle avoidance distance. If the total number of counted points is less than 5, obstacle avoidance is not required, which can remove some noise points; otherwise, it is judged that the obstacle is on the left side of the robot, such as Fig.10 As shown in (a); when the robot moves to the right, it detects points within the range of 60° to 120° through the 2D laser radar, and counts the points whose distance is less than the obstacle avoidance distance set at 250mm. If the total number of counted points is less than 5, obstacle avoidance is not required, which can remove some noise points. Otherwise, it is judged that the obstacle is on the right side of the robot, as shown in Fig.10 (b) as shown.

[0089] Based on the recognition of the obstacle position on the floor rest platform, a new obstacle avoidance strategy is proposed: on the floor rest platform, assuming that the obstacle appears in the left direction of the robot, when the robot detects the obstacle, it slows down and then moves forward 250mm until no obstacle is detected on the left, then moves to the left, and finally bypasses the obstacle and continues cleaning. The whole process is as follows: Fig.10As shown in (a); Assuming that an obstacle appears in the right direction of the robot, when the robot detects the obstacle, it slows down and then moves forward 250mm until no obstacle is detected on the right side, then moves to the right, and finally bypasses the obstacle and continues cleaning. The whole process is shown in Fig.10 (b) as shown.

[0090] Specifically, in step S4, the specific method includes:

[0091] When the robot uses the improved plow path planning to clean the sides and corners of the rest platform, in order to ensure the cleaning coverage, the robot needs to move parallel to the wall. The problem of radar recognition of the wall needs to be solved, such as Fig.11 As shown in the figure, the radar can detect that the distance between the robot and the wall is D. A plane coordinate system is established on the radar, and the initial position of 0° is defined. The angles of the laser beam θ1, θ2, and θ3 can be refracted. By using the geometric theorem, d1=D*tanθ1, d1+d2=D*tanθ2, and d1+d2+d3=D*tanθ3 are obtained, and the coordinates of point 1, point 2, point 3, and point n are obtained. If the x coordinate difference is less than or equal to 5mm, it means that these points are on a straight line, and the y coordinate is an arithmetic progression. By fitting, it can be obtained that the laser beam refracted by the radar at this time is a uniformly distributed point, which proves that it is a wall. In actual application, in order to ensure that the robot's cleaning range is as large as possible, the robot should be as close to the wall as possible and move parallel to the wall. Since the robot's posture is not actually ideally parallel to the wall, it is necessary to add an IMU sensor to obtain the robot's posture and obtain posture compensation when the movement is offset to ensure that the robot can move in a straight line or parallel to the wall.

[0092] In summary, the present invention applies 2D lidar to the stair cleaning robot, so that the stair cleaning robot can realize detection and identification from stairs to rest platforms; efficient cleaning on the rest platforms; and transition from the rest platforms to the next level of stairs, thereby ensuring that the robot can better serve the cleaning work of high-rise buildings.

[0093] The description and practice disclosed in the present invention are easy to think and understand for ordinary technicians in the technical field, and several improvements and modifications can be made without departing from the principles of the present invention. Therefore, modifications or improvements made without departing from the spirit of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A transition and planning method between stairs and rest platforms for a stair cleaning robot based on 2D radar, characterized in that: The steps include: Step S1: Transition of the robot from the stairs to the rest platform; The robot cleans by going downstairs, so the robot detects whether it is a staircase or a rest platform downwards, and determines it through the first sounding sensor and the second sounding sensor; if the robot moves forward a certain distance, the first sounding sensor and the second sounding sensor both sense that there is no obstruction below, which means that the robot currently needs to determine whether the front is a rest platform or a staircase; after the robot finishes cleaning the current step, the 2D laser radar tilts a certain angle, and the 2D laser radar uses triangulation technology to output the distance value and angle, and can obtain the angle and distance data of the obstacle to achieve obstacle identification; Step S2: Analyze the landing point, determine the starting point and the robot turns around at the starting point; The robot reaches the rest platform from the stairs. There are two landing points, including the first position and the second position. The robot uses the radar to sense the distance to the wall to the left. If the distance is small, it is judged to be at the first position; if the distance is large, it is judged to be at the second position. Taking the first position as the starting point, the position of the robot at the second landing point needs to be offset by the length of the stairs to the first position. Step S3: traverse the rest platform; After the robot reaches the rest platform, it will traverse the rest platform, and use the improved plow path planning method through 2D laser radar to achieve the traversal of the rest platform; The plough path planning means that the robot starts from a certain point and moves in a straight line in a specific direction. When it encounters a boundary, the robot stops moving and rotates 90 degrees at the current position, then moves forward one body position, then remains still and rotates 90 degrees again, and finally continues to move forward. This method is repeated until the entire area is traversed. Among them, the rest platform area calculation: A= l*w， A is the area of ​​the rest platform, l is the length of the rest platform, w is the width of the rest platform; Calculation of rest platform distance: D = (P*N) / S, where D is the rest platform distance, P is the number of rows of rest platforms, N is the length of each row, and S is the robot running speed; Estimation of rest platform operation time: T=A / (R*E), T is the tillage time, A is the tillage area, R is the rate of the robot cleaning operation, and E is the efficiency coefficient; Step S4: determine the target point; After continuous cycles, the stair cleaning robot can complete the cleaning traversal of the floor rest platform. When the distance in the 0° direction measured by the radar is less than the length of the robot, it stops moving and cleaning. Step S5: the robot transitions from the rest platform to the next staircase; When the robot reaches the vicinity of the target point, if the radar above the robot tilted at an angle of μ detects H1=h*m, and the sounding sensor detects nothing downward, it means that the robot has reached the intersection of the rest platform and the next staircase, and there is a staircase leading down ahead, where h is the height of the vertical part of the step, and m is the number of steps; On the contrary, when the H2 detected by the radar is less than h*m, it indicates that the robot has not yet reached the intersection of the rest platform and the next staircase, and needs to move forward for a distance until the distance detected by the radar is H2 before it can go downstairs, thus completing the stair cleaning and the transition between the staircase and the rest platform. In step S3, the specific method for handling obstacles encountered by the robot on the rest platform includes: When the stair cleaning robot is working on the floor rest platform: Assuming that the robot cleans according to the plowing path planning algorithm, when the robot moves to the left, it detects points within a range greater than 240° and less than 300° through the 2D laser radar, and counts points with a distance less than 250mm to set the obstacle avoidance distance. If the total number of counted points is less than 5, there is no need to avoid obstacles; otherwise, it is judged that the obstacle is on the left side of the robot; when the robot moves to the right, it detects points within a range of 60° to 120° through the 2D laser radar, and counts points with a distance less than 250mm to set the obstacle avoidance distance. If the total number of counted points is less than 5, there is no need to avoid obstacles, otherwise, it is judged that the obstacle is on the right side of the robot; On the basis of realizing the position discrimination of obstacles on the floor rest platform, a new obstacle avoidance strategy is proposed: on the floor rest platform, assuming that an obstacle appears in the left direction of the robot, the robot will slow down when it detects the obstacle, then move forward 250mm until no obstacle is detected on the left, then move to the left, finally bypass the obstacle and continue cleaning; assuming that an obstacle appears in the right direction of the robot, the robot will slow down when it detects the obstacle, then move forward 250mm until no obstacle is detected on the right, then move to the right, finally bypass the obstacle and continue cleaning.

2. The method for transition and planning between stairs and rest platforms of a 2D radar-based stair cleaning robot according to claim 1, characterized in that: In step S1, the specific method includes: The radar can directly measure the distances m1, m2, and m3, and obtain θ1, θ2, and θ3 on the established coordinate system. If the radar can detect two or more line segments with different distances, L1=m1*sinθ1, L2=m2*sinθ2, and L1≠L2, then it is judged that there is still a staircase in front of the robot; on the contrary, when the radar always detects a line segment with a distance of L3, L3=m3*sinθ3, it is judged that there is a rest platform below, and the robot should complete the going downstairs action and then enter the rest platform planning control state; among them, m1, m2, and m3 are the actual distances between the rotation center and the sampling point, and θ1, θ2, and θ3 are the angles of the current sampling point relative to the radar's own direction.

3. The transition and planning method between stairs and rest platforms of a 2D radar-based stair cleaning robot according to claim 1 is characterized in that: In step S2, the specific method includes: Step S21: the robot moves from a position D1 away from the stair riser to a position D2 away from the stair riser in a direction perpendicular to the stair riser at the first position; Step S22: After ensuring that D2 = D1 + half of the fuselage length, rotate counterclockwise; Step S23: Move the robot to a position D1 away from the stair riser in a direction perpendicular to the stair riser to ensure that the corner can be effectively cleaned.

4. The method for transition and planning between stairs and rest platforms of a 2D radar-based stair cleaning robot according to claim 1, characterized in that: In step S3, the specific method of the robot traversing the route on the rest platform includes: Step S31: Assuming that the stair cleaning robot transitions to the floor rest platform after completing the cleaning of the stair section, the robot body first rotates 90° clockwise; Step S32: After the robot rotates 90°, the distances in the 0° direction and the 270° direction are measured by the radar, which are recorded as Length and Width. By setting the expected left and right movement speed of the robot, the time t=Width / speed for each left and right movement can be obtained according to the measured distance Width. The length of each forward movement of the robot is equal to the width of the robot, which is achieved by controlling the speed and running time in the program. Step S33: After the first leftward movement, the robot is ready to move forward, and a variable Count is set to record the number of forward movements, and the initial value of Count is 0; Step S34: After the first forward movement, the robot is ready to move right, at this time Count+1, Count=1; Step S35: After the first rightward movement, the robot is ready to move forward; Step S36: After the second forward movement, Count is increased by 1, and the value of Count is 2 at this time.

5. The method for transition and planning between stairs and rest platforms of a 2D radar-based stair cleaning robot according to claim 1, characterized in that: In step S4, the specific method includes: When the robot uses the improved plow path planning to clean the sides and corners of the rest platform, the radar can detect that the distance between the robot and the wall is D, and establish a plane coordinate system on the radar, defining the initial position of 0°. The angles θ1, θ2, and θ3 of the laser beam can be refracted. According to the geometric theorem, d1=D*tanθ1, d1+d2=D*tanθ2, and d1+d2+d3=D*tanθ3 are obtained, and the coordinates of point 1, point 2, point 3, and point n are obtained. If the difference in the x-coordinate is less than or equal to 5mm, it means that these points are on a straight line, and the y-coordinate is an arithmetic progression. Through fitting, it can be obtained that the laser beam refracted by the radar at this time is a uniformly distributed point, which proves that it is a wall.

Images