Swimming pool wall and water line autonomous cleaning method and swimming pool cleaning robot

Abstract

This invention discloses an autonomous cleaning method for pool walls and waterlines, comprising: a pool cleaning robot first climbs upwards along the pool wall, adjusting its height based on water outlet sensor signals to keep the cleaning components positioned at the waterline; then, it moves laterally along the waterline to clean, simultaneously collecting lateral point cloud data. When the number of point clouds with a distance less than a set threshold reaches a certain threshold, the robot descends from the wall to the pool bottom. The robot then retreats a set distance towards the wall in front of it at the pool bottom and re-detects the lateral point cloud. If the point cloud conditions are met, it performs a first-angle turn to align with the adjacent pool wall; otherwise, it tilts towards that adjacent pool wall with a second-angle turn. The robot then moves forward and corrects its angle with the corresponding pool wall. Once aligned with the pool wall, it climbs upwards along the pool wall again. This invention also discloses a pool cleaning robot capable of implementing the aforementioned cleaning method. This invention enables highly efficient automated cleaning of pool walls and waterline areas.

Description

Technical Field

[0001] This invention relates to a swimming pool cleaning method and a swimming pool cleaning robot, and more particularly to a method and robot for autonomous cleaning of swimming pool walls and waterlines. Background Technology

[0002] Pool cleaning robots are intelligent devices used to automatically clean the bottom, walls, and waterline of swimming pools. Most existing pool cleaning robots operate by random collisions or following simple pre-set paths, resulting in low cleaning coverage and inefficiency.

[0003] To improve cleaning efficiency, some advanced robots integrate inertial measurement units (IMUs). By acquiring their own attitude angle information and using closed-loop control algorithms to adjust their movement mechanisms, they can achieve stable vertical climbing and descending along the pool wall. However, these robots only solve the problem of attitude stability. Due to a lack of precise perception of the surrounding environment, such as the inability to accurately determine the distance to the pool wall, identify obstacles ahead, or accurately perceive the position of the waterline, they still have many shortcomings in practical applications. Specifically, the robots cannot effectively clean waterline areas or are prone to getting stuck when encountering complex structures such as pool corners or ladders. In addition, although existing technologies have sporadically appeared solutions using single sensors such as LiDAR or water immersion sensors, they generally lack a systematic solution that effectively integrates information from multiple sensors and applies it to a complete closed-loop operation process, from pool bottom wall finding, stable wall climbing, waterline identification, cleaning along the waterline, intelligent wall replacement, to safe descent. Summary of the Invention

[0004] To address the shortcomings of the existing technology, this invention provides a method for autonomous cleaning of swimming pool walls and waterlines, achieving highly efficient automated cleaning of these areas. Another objective of this invention is to provide a swimming pool cleaning robot to achieve the same autonomous cleaning method for swimming pool walls and waterlines.

[0005] The technical solution of this invention is as follows: A method for self-cleaning of swimming pool walls and waterlines, comprising:

[0006] The pool cleaning robot climbs up the pool wall;

[0007] Based on the water outlet sensor signal, the height of the pool cleaning robot on the pool wall is controlled so that the cleaning components of the pool cleaning robot are kept at the waterline position.

[0008] The horizontally driven pool cleaning robot moves along the waterline to clean and detects the lateral point cloud data of the pool cleaning robot corresponding to the direction of movement. When the number of points in the lateral point cloud data that are less than a set distance threshold reaches a certain threshold, the pool cleaning robot goes down the wall and reaches the bottom of the pool.

[0009] The pool cleaning robot moves backward a set distance relative to the front wall at the bottom of the pool and detects lateral point cloud data. When the number of points in the lateral point cloud data that are less than a set distance threshold reaches a certain threshold, it performs a first angle turn to face the adjacent pool wall of the current front pool wall; otherwise, it performs a second angle turn to tilt towards the current front pool wall.

[0010] The pool cleaning robot moves forward and corrects its angle with the pool wall it is facing, then climbs up the pool wall after it is directly facing the wall.

[0011] Furthermore, based on the water outlet sensor signal, the height of the pool cleaning robot on the pool wall is controlled to keep the cleaning components of the pool cleaning robot at the waterline position. Specifically, when the pool cleaning robot climbs up the pool wall, the adsorption component works with maximum suction. After the water outlet sensor detects that the pool cleaning robot has exited the water, the suction of the adsorption component is reduced, causing the height of the pool cleaning robot on the pool wall to decrease until the water outlet sensor detects that the pool cleaning robot has re-entered the water, at which point the suction of the adsorption component is stopped.

[0012] Furthermore, the reduction of the adsorption force of the adsorption component is achieved by gradually reducing the adsorption force of the adsorption component in a stepwise manner.

[0013] Furthermore, the first angle is 90°, and the second angle is no more than 45°.

[0014] Furthermore, the process of the pool cleaning robot advancing and correcting its angle relative to the pool wall specifically involves: after turning at the bottom of the pool, the robot calculates the angle between its current orientation and the pool wall's normal direction based on the forward laser radar point cloud computing data. By adjusting the robot's yaw angle, the robot is positioned directly facing the pool wall.

[0015] Furthermore, when climbing upwards along the pool wall, the roll angle obtained by the inertial measurement unit of the pool cleaning robot is used to calculate the differential speed control amount of the left and right movement mechanism of the pool cleaning robot through a proportional-derivative controller or a proportional-integral-derivative controller, so as to correct the left and right tilt of the robot body.

[0016] Furthermore, the differential control quantity The formula is:

[0017]

[0018] in, The error is the roll angle. The angular velocity of the roll angle. For proportional gain, This is the differential gain.

[0019] Furthermore, when climbing upwards along the pool wall, the roll angle obtained by the inertial measurement unit of the pool cleaning robot is used to calculate the difference in adsorption force between the left and right adsorption components of the pool cleaning robot through a proportional-derivative controller or a proportional-integral-derivative controller, thereby adjusting the left and right adsorption force of the pool cleaning robot to correct the left and right tilt of the robot body.

[0020] Another objective of this invention is to provide a swimming pool cleaning robot, comprising a robot body, wherein the robot body is equipped with a two-dimensional lidar, a water outlet sensor, an inertial measurement unit, a control module, a moving mechanism, and an adsorption component, the control module being electrically connected to the two-dimensional lidar, the water outlet sensor, the inertial measurement unit, the moving mechanism, and the adsorption component, and the control module comprising:

[0021] The memory is configured to store instructions;

[0022] The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement the aforementioned autonomous cleaning method for the pool walls and waterline.

[0023] Furthermore, the scanning field of view of the two-dimensional lidar is 220° to 270°.

[0024] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows:

[0025] By integrating two-dimensional lidar, water vent sensors, and inertial measurement units, the robot can accurately perceive all-around environmental information from the underwater pool wall outline to the water surface boundary, solving the problem of existing technologies having limited perception capabilities and being unable to accurately identify water lines and wall environments, thus achieving high-precision panoramic perception.

[0026] Through a closed-loop control process involving pool wall raising, water outlet detection, water line cleaning, adjacent wall detection, wall lowering, pool bottom reversal, and pool wall raising, perception, decision-making, and execution are integrated, enabling the robot to autonomously complete pool cleaning. This achieves truly unmanned, full-area cleaning, significantly improving the level of automation and cleaning coverage.

[0027] After the pool wall rises, it can adaptively determine the optimal adsorption force according to the actual pool conditions, ensuring that the cleaning roller brush acts precisely on the waterline area, solving the problem of poor cleaning effect due to different pool conditions, and improving adaptability to different pool environments. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the structure of the pool cleaning robot in an embodiment.

[0029] Figure 2 This is a schematic diagram of the various modules of the pool cleaning robot in this embodiment.

[0030] Figure 3 This is a schematic diagram of the process for the self-cleaning method of the pool wall and waterline in an embodiment.

[0031] Figure 4 This is a schematic diagram illustrating the principle of controlling the upward posture of the pool wall during the self-cleaning process of the pool wall and waterline.

[0032] Figure 5 This is a schematic diagram illustrating the pump speed calibration process during the self-cleaning method for pool walls and waterlines. Detailed Implementation

[0033] The present invention will be further described below with reference to embodiments. It should be understood that these embodiments are only for illustrating the present invention and are not intended to limit the scope of the present invention. After reading this description, any modifications of this description in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0034] Please combine Figure 1 , Figure 2 As shown, the swimming pool cleaning robot involved in this embodiment of the invention includes a robot body 1. The robot body 1 integrates a multi-sensor sensing unit, a moving mechanism 50 for performing motion, an adsorption component 60 for providing wall adhesion, and a control module 10. A cleaning roller brush 70 is provided at the front end of the robot body 1. The multi-sensor sensing unit is composed of various sensors, specifically including: a two-dimensional lidar 20 for detecting the outline of the environment in front and above, a water outlet sensor 30 for accurately determining the water surface boundary, and an inertial measurement unit 40 for real-time monitoring of the robot's own attitude.

[0035] The control module 10 is electrically connected to the 2D LiDAR 20, the water outlet sensor 30, the inertial measurement unit 40, the moving mechanism 50, and the adsorption component 60. The 2D LiDAR 20, the water outlet sensor 30, and the inertial measurement unit 40 serve as input devices, transmitting the collected distance information, water outlet status signals, and attitude angle data to the control module 10 in real time. The control module 10 includes a memory and a processor. The memory is configured to store instructions; the processor is configured to retrieve instructions from the memory and, when executing instructions, analyze and make decisions based on the input information, generating corresponding control instructions, which are then output to the moving mechanism 50 and the adsorption component 60 to control the movement, turning, climbing, descending, and adsorption force adjustment of the pool cleaning robot, achieving an autonomous cleaning method for the pool walls and waterline.

[0036] The two-dimensional lidar 20 is mounted at the upper left or upper right corner of the front of the robot body 1, with a scanning field of view of 220° to 270°. This position allows its scanning plane to cover the area in front of and above the robot's direction of travel, enabling it to detect the pool wall in front when at the bottom of the pool and the water surface or obstacles above when climbing walls. To protect the delicate and relatively fragile two-dimensional lidar 20 from damage by collisions or large suspended objects in the complex underwater environment, it is equipped with a protective grid cover. This protective grid cover consists of a grid structure composed of horizontal and vertical grids, which effectively blocks foreign objects while maximizing the passage of the laser beam.

[0037] The water discharge sensor 30 is installed in the drainage area on the top of the robot body 1. Figure 1 (Not shown in the image), its function is to determine whether the robot has climbed to the waterline position. To ensure the accuracy of the detection, the water outlet sensor 30 is placed inside a light-shielding structure. This light-shielding structure satisfies both light-shielding and water-permeable conditions: firstly, it blocks light to prevent ambient light (especially strong sunlight) from interfering with the sensor and causing misjudgments; secondly, it allows water to flow freely, ensuring that the water environment around the sensor changes rapidly when the top of the robot touches or leaves the water surface, thus enabling the water outlet sensor to promptly detect changes in the medium's state and avoid delays or misjudgments caused by stagnant water or trapped air. For example, the light-shielding structure can be a dark-colored cover with water vents. In this embodiment, the water outlet sensor 30 can be an optical sensor, which determines the water outlet state by detecting the difference in the refractive index or reflectivity of the medium (water or air) around the sensor probe.

[0038] The inertial measurement unit 40, which typically includes a three-axis accelerometer and a three-axis gyroscope, is integrated on a circuit board inside the robot body 1. It is used to measure the robot's three-dimensional attitude angles in real time, including the pitch angle and roll angle, which are crucial for the climbing attitude.

[0039] Please combine Figures 3 to 5 As shown, the autonomous cleaning method for pool walls and waterlines mainly includes the following six steps: S100 pool bottom reversal, S200 pool wall rising and attitude adjustment, S300 water outlet detection, S400 waterline cleaning, S500 adjacent wall detection, and S600 wall descent. After descent, the pool cleaning robot can re-enter the pool bottom reversal area in S100 and begin cleaning the next wall until all walls are cleaned or the preset working time is reached.

[0040] The following is a detailed explanation of the work process for each step:

[0041] Step S100: Change direction at the bottom of the pool.

[0042] This step involves determining whether the robot can remount after cleaning the pool bottom or descending from the pool wall, or whether it needs to continue moving along the pool bottom to change walls, once it is close to the pool wall in front of it. After the robot starts, the 2D LiDAR 20 is activated, scanning at a certain frequency (e.g., 10 Hz) to acquire point cloud data within a 270-degree range in front of the robot. This point cloud data is analyzed in real time to calculate the distances to the nearest obstacles directly in front of and to the sides of the robot. When the distance ahead is determined to be less than a preset proximity threshold... When the distance is 0.2 meters (for example), it is considered to be close to the pool wall. Accordingly, the robot executes the preset wall-facing procedure: first, the control mechanism 50 moves the robot backward a small safe distance. (For example, 0.3 meters) to ensure the robot has sufficient turning space. Then, according to the preset global cleaning strategy (e.g., always cleaning the pool clockwise), the robot is controlled to turn to the right by a fixed angle (e.g., 90 degrees). This is to prepare for subsequent movement parallel to the wall and ultimately achieve a face-to-face position. Next, the robot continues to move forward, continuously fine-tuning its attitude using LiDAR data until the distance between its front and the pool wall reaches a preset wall-climbing preparation distance (e.g., 0.1 meters), and the angle between its heading angle and the normal direction of the pool wall is less than a preset alignment angle threshold (e.g., 5 degrees). At this point, the robot has completed its attitude adjustment to face the pool wall and is ready to begin climbing.

[0043] Specifically, a two-dimensional lidar continuously detects the distance in front. When detected At that point, it is determined that the robot has approached the pool wall, and the robot is instructed to retreat in the opposite direction until the distance in front meets the requirement. This is to ensure that the robot has enough room to turn.

[0044] After completing the back-off distance, proceed according to the current cleaning direction. Select the corresponding lateral radar scanning area (clockwise or counterclockwise). When the cleaning direction is clockwise, select the right scanning area; when the cleaning direction is counterclockwise, select the left scanning area. Within the selected scanning area, the statistical distance is less than the lateral threshold. Effective LiDAR ranging point cloud quantity When detected When the set threshold is reached, it indicates that continuous sidewalls can be detected in the current cleaning direction, indicating that the bottom wall replacement operation can be performed, and the predetermined rotation angle for bottom wall replacement is set to a larger angle. Generally, it is 90°; when detected When the set threshold is reached, it indicates insufficient information from the side pool walls, making it more suitable to continue moving along the pool bottom. Therefore, the predetermined rotation angle should be set to a smaller angle. ,in ,For example It is 30°.

[0045] The robot then rotates a predetermined angle along the current cleaning direction and moves straight forward in the current direction after the rotation is complete. During this straight-line movement, the forward distance measurement value is continuously monitored; when detected... When a set threshold is reached, the robot is deemed to be approaching the pool wall again, entering the feasibility assessment stage for climbing the wall. In this stage, the robot's laser radar scan area in front of it is counted for objects within a distance less than the pool wall's threshold. Effective point cloud quantity .when When the set threshold is reached, it indicates that the continuity of the pool wall ahead is insufficient or the local structure is unsuitable for performing the wall-climbing action, and it is determined to be in a state where wall-climbing is not possible. The process then returns to the pool bottom reversal procedure to continue performing pool bottom movement or wall-changing operations. When a set threshold is reached, the robot is deemed capable of climbing onto the pool wall. If the conditions for climbing are met, the robot calculates the normal direction of the pool wall based on the points detected by the forward LiDAR, and then calculates the angle between the robot's current orientation and the pool wall's normal direction. By fine-tuning the robot's yaw angle, the robot can be made to face the pool wall.

[0046] Step S200: Pool wall rise and attitude adjustment.

[0047] During the ascent of the pool wall, the robot's attitude angles, including pitch angle, are output in real time via the inertial measurement unit (IMU). Roll angle and yaw angle Among them, roll angle This is used to characterize the robot's left and right tilt relative to the pool wall in the normal direction. To keep the robot vertical and avoid incorrect orientation or slippage caused by tilting, the control system... Construct differential control parameters and drive the left and right tracks to run at different speeds.

[0048] Specifically, after the robot faces and approaches the pool wall, the control module 10 sends a command to the adsorption assembly 60, causing its internal water pump to operate at higher power to generate sufficient negative pressure, thereby allowing the robot to firmly adhere to the pool wall. Subsequently, the moving mechanism 50 (in this embodiment, the left and right tracks) climbs at a baseline speed. The ascent begins. Maintaining attitude stability is crucial throughout the ascent. Control module 10 continuously acquires real-time roll angle data from inertial measurement unit 40. Ideally, the robot should maintain a vertical climb, i.e., a roll angle of approximately 0.5°. It should be 0. Any value deviating from 0 indicates that the robot has tilted left or right. See also Figure 4 It shows a schematic diagram of the attitude control principle. When the roll angle is detected... If the value is not zero, immediately perform closed-loop correction. First, calculate the roll angle error. The target roll angle is 0. Simultaneously, to predict the tilting trend, the angular velocity of the roll angle also needs to be calculated. This angular velocity can be obtained by considering the roll angle at the current moment. Roll angle at the time of the previous control cycle We approximate the calculation by using differences, that is:

[0049]

[0050] in, This is the time interval of the control cycle. Next, a proportional-derivative controller (or PID controller) is used to calculate the differential control quantity. :

[0051]

[0052] in, This is the proportional gain coefficient. This represents the differential gain coefficient. It should be noted that these two parameters need to be tuned based on the robot's specific mechanical characteristics and dynamic model. Proportional term. Its function is to apply a restoring force based on the current tilt angle; the greater the tilt, the greater the restoring force. Differential term Its function is to suppress the tilting angular velocity to prevent the robot from oscillating due to excessive adjustments. The differential control value is calculated. Then, it is assigned to the left and right tracks. The target speed for the left track... and the target speed of the right track The calculations are as follows:

[0053]

[0054]

[0055] For example, when the robot tilts to the right, If positive, the error The calculated differential control value is negative. It also tends to be negative. This will result in the left track speed. Increase, while right track speed The left-faster-right-slower motion generates a leftward corrective torque M, which restores the robot to a vertical posture with a roll angle of 0. By repeating this process in each control cycle (e.g., every 10 milliseconds), the robot can resist disturbances such as water flow and steadily climb vertically upwards along the pool wall.

[0056] Step S300: Water effluent detection.

[0057] During the robot's ascent, the signal from the water-out sensor 30 located at its top is monitored. When most of the robot's top surface is above the water, the medium surrounding the water-out sensor 30 changes from water to air, causing a clear jump in its output signal. After this water-out signal is detected continuously for a preset time (e.g., 0.5 seconds), it is determined that the robot has reached the waterline. Specifically, the water-out sensor 30 measures the received light intensity I, when... Time output If within the time window Detected N times consecutively The water discharge is deemed complete.

[0058] At this point, in order to clean the water line in the most energy-efficient and effective manner subsequently, this embodiment provides a preferred method for calibrating the water pump speed to adjust the adsorption force generated by the adsorption component 60. See also Figure 5 The diagram illustrates the pump speed calibration process. After the initial determination that the water level has been reached, a one-time pump speed calibration procedure is performed. Specifically, the speed of the adsorption component 60 (i.e., the pump) is gradually and stepwise reduced. For example, decreasing by 50 revolutions per second. As the adhesive force decreases, the robot will begin to slowly slide downwards due to gravity, as shown in its height curve. It also descends accordingly. During this process, the signal from the water outlet sensor 30 is continuously monitored. When the robot slides down to the point where the water outlet sensor 30 at the top is submerged again, the sensor signal flips to the "in-water" state, and the water pump speed at this moment is recorded and defined as the calibration speed. The calibration process occurs at the time of calibration completion. The end. In other words, this... Rotational speed is the critical minimum rotational speed required to keep the robot hovering just above the waterline under the specific water level and pool wall conditions of the pool.

[0059] The specific process is as follows:

[0060] (a) Complete water discharge determination stage

[0061] As the robot rises along the pool wall, when the water outlet sensor 30 continuously detects water outlet signals, the speed of the adsorption pump is set to its maximum speed. This allows the water pump to drain water at full power, generating maximum suction force. In this state, the robot can stably maintain an upright posture and continue to climb up the pool wall until the entire robot is completely out of the water. At this point, it is determined that the robot has completed the water exit process.

[0062] (b) Identification of waterline height deviation

[0063] After the robot has completely exited the water, the cleaning roller brush 70 may move away from the waterline area due to excessive suction, causing the brush to dry-brush on the pool wall surface, affecting the cleaning effect and accelerating mechanical wear. To avoid the above problems, after confirming that the water has completely exited the water, the waterline cleaning mode is not entered immediately; instead, the water pump speed calibration process is started.

[0064] (c) Pump speed reduction calibration stage

[0065] During the pump calibration phase, while maintaining the robot's vertical attachment to the pool wall, the pump speed is gradually reduced according to a preset step size. As the pump speed gradually decreases, the robot's adhesion to the pool wall decreases, and its attachment position slowly moves downward along the pool wall, with the cleaning roller 70 gradually approaching the waterline area.

[0066] (d) Water ingress threshold detection and calibration completed

[0067] During the decrease in rotational speed, the status change of the water outlet sensor 30 is continuously monitored. When the water outlet sensor 30 detects a switch from water outlet to water inlet, it is determined that the cleaning roller brush 70 has descended to the water surface boundary height. At this time, the corresponding water pump speed is recorded as the water pump calibration speed during the waterline process. .

[0068] Step S400: Water line cleaning.

[0069] During the waterline cleaning process, the robot needs to maintain a vertical posture against the pool wall and move horizontally along the waterline. The IMU is used to acquire the attitude angle in real time, and PID control is employed to differentially adjust the drive wheels (or tracks) to stabilize the robot's posture in a vertical position.

[0070] After completing the water pump speed calibration, set the speed of the adsorption component 60 to the speed just calibrated. (or slightly higher to increase safety margin) ,For example This ensures that the robot's cleaning brush operates precisely on the waterline area where water and air meet, achieving optimal cleaning results while avoiding energy waste and component wear caused by excessive suction. Subsequently, the movement mechanism 50 is controlled to move the robot horizontally along the pool wall (e.g., moving to the right in a clockwise direction). Attitude control is equally important during horizontal movement. Here, not only the roll angle of the inertial measurement unit 40 is utilized... To keep the robot from tilting left or right (target roll angle is 0 degrees), pitch angle must also be used to ensure the robot's body remains perpendicular to the wall (target pitch and roll angles are: , Its control algorithm is similar to that of step S200, both using a PID controller to correct posture errors in real time to ensure that the robot maintains a stable posture during translation.

[0071] Step S500: Detect adjacent walls.

[0072] As the robot moves horizontally along the waterline, its 2D LiDAR 20 remains operational. The point cloud data returned by the LiDAR from the right side (or the side facing the cleaning direction) is analyzed. When the number of points in the lateral point cloud data whose distance is less than a set distance threshold reaches a certain threshold, it indicates that the robot is about to reach a corner.

[0073] Step S600: Lower the wall.

[0074] Once the robot has completed cleaning the waterline on the pool wall, or reached the preset total working time, it executes step S600 to descend from the wall. The robot gradually and smoothly reduces the suction force of the adsorption component 60, while utilizing the attitude closed-loop control algorithm in step S200 to ensure that the robot maintains a vertical attitude under the influence of gravity and descends stably along the pool wall. When the inertial measurement unit 40 detects that the robot's pitch angle changes from nearly 90 degrees (vertical) to nearly 0 degrees (horizontal), or when the bottom collision sensor (if any) is triggered, it is determined that the robot has safely reached the bottom of the pool. At this point, the adsorption component 60 completely stops working, and a complete cycle of cleaning the pool wall and waterline, namely "pool bottom → wall up → water out → waterline → wall replacement → wall down → pool bottom," is completed.

[0075] Example 2

[0076] As an optional implementation, this embodiment further proposes an optimized attitude control scheme based on Embodiment 1, namely, coordinated control of attitude control and suction adjustment. This scheme aims to provide a more powerful and faster attitude correction capability to cope with more complex working conditions, such as uneven pool wall surfaces, strong water flow impacts, or slight slippage of one track.

[0077] The system hardware structure of this embodiment is basically the same as that of Embodiment 1, including a robot body 1, a control module 10, a two-dimensional lidar 20, a water outlet sensor 30, an inertial measurement unit 40, and a moving mechanism 50. The key difference from Embodiment 1 is that the adsorption component 60 is designed to allow for zoned or differential adjustment of suction force. For example, the adsorption component 60 may include two or more independently controlled water pumps or valves, corresponding to the suction cup areas on the left and right sides of the robot, respectively.

[0078] In this embodiment, during step S200, when the robot is climbing along the pool wall or moving laterally along the waterline, the roll angle is monitored in real time by the inertial measurement unit 40. When a tilt is detected in the robot, such as a tilt to the right (i.e., roll angle), If the value is positive, it will not only perform differential control as described in Example 1 (i.e., instruct the left track to accelerate and the right track to decelerate), but will also send a coordinated control command to the adsorption component 60 at the same time.

[0079] Specifically, this cooperative control command can momentarily and briefly enhance the suction force of the robot's right-side suction cup area and / or weaken the suction force of the left-side suction cup area. This suction difference generates an additional restoring torque at the bottom of the robot to resist tilting. This restoring torque, along with the corrective torque M generated by the differential speed of the tracks (e.g., ... Figure 4 (As shown) are superimposed and act together on the robot body 1, so that the robot's posture can be pulled back to the vertical state more quickly and powerfully.

[0080] This collaborative control strategy introduces a new dimension of control. Track differential speed primarily generates torque through friction; however, its effectiveness is affected when the pool walls are slippery or the track adhesion is insufficient. In contrast, the restoring torque generated by differential suction acts directly on the robot body, unaffected by track slippage, resulting in a faster response and more robust performance. In this way, the robot's posture stability is significantly improved when facing sudden disturbances, enabling it to better maintain its predetermined trajectory and enhancing cleaning quality and operational safety.

[0081] Example 3

[0082] This embodiment discloses a simplified working mode for a specific application scenario, namely the "waterline-specific cleaning mode". This mode aims to meet users' specific needs for quickly and efficiently cleaning pool waterline sludge, thereby simplifying user operation and improving product usability.

[0083] The hardware system of this embodiment is exactly the same as that of Embodiment 1. Its feature lies in that the pool cleaning robot provides at least two working mode options, which can be selected, for example, through a physical button on the robot body or a companion mobile application: one is the "fully automatic mode" that performs the complete cleaning process described in Embodiment 1; the other is the "waterline-specific mode" of this embodiment.

[0084] After the user selects the "Waterline Dedicated Mode", the cleaning process of the control module 10 will be simplified. After the user starts the "Waterline Dedicated Mode", the robot will directly start from "Step S200: Pool Wall Rise and Attitude Adjustment".

[0085] By offering this mode, users can directly focus on maintaining the most heavily contaminated waterline areas without waiting for the robot to complete the time-consuming process of cleaning the pool bottom and finding the walls. This greatly improves cleaning efficiency and product usability, making it especially suitable for daily quick cleaning tasks.

[0086] Example 4

[0087] This embodiment, based on embodiment 1, further deepens the application of 2D LiDAR 20 data. By introducing an intelligent obstacle avoidance algorithm based on point cloud data pattern recognition, it enhances the robot's robustness and autonomous operation success rate in irregularly shaped swimming pools or environments with complex obstacles.

[0088] The hardware system in this embodiment is the same as that in Embodiment 1. Its core lies in the software algorithm of the control module 10, especially in "Step S100: Pool Bottom Reversal" and "Step S400: Waterline Cleaning", which embeds advanced obstacle recognition and escape subroutines.

[0089] The control module 10 is configured to analyze the fan-shaped scan point cloud data returned by the two-dimensional LiDAR 20 in real time and at high frequency. The core function of this subroutine is to identify specific point cloud distribution patterns, which are usually associated with typical scenarios that can cause the robot to get stuck.

[0090] For example, when a robot moves along the bottom of a pool or laterally along the waterline, if the point cloud data in front of it exhibits a sharply concave "U" or "V" shaped distribution—meaning the detection distance on the sides is relatively short while the detection distance directly in front suddenly increases—this usually indicates that the robot is entering a 90-degree inner corner. Without intervention, the robot is likely to get stuck in the corner due to insufficient turning space. Similarly, when the point cloud data exhibits a complex, discontinuous, stepped distribution, this may indicate that the robot has encountered an entry step or stainless steel ladder in the pool. These structures are also common causes of traditional robots getting stuck.

[0091] In this embodiment, when the identification algorithm of the control module 10 detects the predefined point cloud data features related to potential jamming risks (for example, detecting that the point cloud data forms a U-shaped indentation distribution feature within a predetermined distance in front of the robot body), the system will immediately interrupt the current straight-line or regular turning logic and instead trigger and execute a set of preset escape procedures.

[0092] As one implementation, the escape procedure may include the following sequence of actions:

[0093] 1. Stop moving forward.

[0094] 2. Control the moving mechanism 50 to precisely retreat a safe distance, for example, 50 centimeters, along the original path used to enter the area. It should be noted that this precise retreat can be achieved using odometer or inertial navigation data.

[0095] 3. After moving backward into position, rotate the robot in place by a specific angle, such as 45 degrees or -45 degrees, to change the robot's orientation.

[0096] 4. After rotating, use the 2D lidar 20 again to detect a new path ahead. If the new path is open, continue moving forward; if it is still an obstructed area, try retreating again and rotating at a different angle until a viable path is found.

[0097] Through the aforementioned closed-loop control logic of "perception-recognition-decision-escape," the robot no longer blindly executes predetermined actions but possesses the ability to assess risks and proactively avoid them based on the geometric characteristics of the real-time environment. This enables the pool robot provided in this application to better adapt to pools of various irregular shapes, as well as pools with complex internal structures such as ladders, steps, and curved corners, thereby significantly reducing the probability of needing human intervention and achieving a higher level of autonomous cleaning.

Claims

1. A method for self-cleaning swimming pool walls and waterlines, characterized in that, include: The pool cleaning robot climbs up the pool wall; Based on the water outlet sensor signal, the height of the pool cleaning robot on the pool wall is controlled so that the cleaning components of the pool cleaning robot are kept at the waterline position. The horizontally driven pool cleaning robot moves along the waterline to clean and detects the lateral point cloud data of the pool cleaning robot corresponding to the direction of movement. When the number of points in the lateral point cloud data that are less than a set distance threshold reaches a certain threshold, the pool cleaning robot goes down the wall and reaches the bottom of the pool. The pool cleaning robot moves backward a set distance relative to the front wall at the bottom of the pool and detects lateral point cloud data. When the number of points in the lateral point cloud data that are less than a set distance threshold reaches a certain threshold, it performs a first angle turn to face the adjacent pool wall of the current front pool wall; otherwise, it performs a second angle turn to tilt towards the current front pool wall. The pool cleaning robot moves forward and corrects its angle with the pool wall it is facing, then climbs up the pool wall after it is directly facing the wall.

2. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, Based on the water outlet sensor signal, the height of the pool cleaning robot on the pool wall is controlled to keep the cleaning components of the pool cleaning robot at the waterline position. Specifically, when the pool cleaning robot is climbing up the pool wall, the adsorption component works with maximum suction. After the water outlet sensor detects that the pool cleaning robot is out of the water, the suction of the adsorption component is reduced, causing the height of the pool cleaning robot on the pool wall to drop, until the water outlet sensor detects that the pool cleaning robot has re-entered the water, and the suction of the adsorption component is stopped.

3. The self-cleaning method for pool walls and waterlines according to claim 2, characterized in that, The adsorption force of the adsorption component is reduced gradually in a stepwise manner.

4. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, The first angle is 90°, and the second angle is no more than 45°.

5. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, The process of the pool cleaning robot moving forward and correcting its angle with the pool wall it is facing is as follows: After turning at the bottom of the pool, the pool cleaning robot calculates the angle between the robot's current orientation and the pool wall's normal direction based on the normal direction of the pool wall calculated by the forward laser radar point cloud computing. By adjusting the yaw angle of the pool cleaning robot, the robot is made to face the pool wall.

6. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, When climbing up the pool wall, the roll angle obtained by the inertial measurement unit of the pool cleaning robot is used to calculate the differential speed control amount of the left and right movement mechanism of the pool cleaning robot through a proportional-derivative controller or a proportional-integral-derivative controller, so as to correct the left and right tilt of the robot body.

7. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, The differential control quantity The formula is: in, The error is the roll angle. The angular velocity of the roll angle. For proportional gain, This is the differential gain.

8. The self-cleaning method for pool walls and waterlines according to claim 1, characterized in that, When the robot is climbing up the pool wall, the roll angle obtained by the inertial measurement unit of the pool cleaning robot is used to calculate the difference in adsorption force between the left and right adsorption components of the pool cleaning robot through a proportional-derivative controller or a proportional-integral-derivative controller, and adjust the left and right adsorption force of the pool cleaning robot to correct the left and right tilt of the robot body.

9. A swimming pool cleaning robot, characterized in that, The robot body includes a two-dimensional lidar, a water outlet sensor, an inertial measurement unit, a control module, a movement mechanism, and an adsorption assembly. The control module is electrically connected to the two-dimensional lidar, water outlet sensor, inertial measurement unit, movement mechanism, and adsorption assembly. The control module includes: The memory is configured to store instructions; The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement the aforementioned autonomous cleaning method for the pool walls and waterline.

10. The pool cleaning robot according to claim 1, characterized in that, The scanning field of view of the two-dimensional lidar is 220° to 270°.

Images