An automated cleaning robot for glass curtain walls

Abstract

This utility model relates to the field of curtain wall cleaning equipment and discloses an automated glass curtain wall cleaning robot, including a main body; an adsorption module disposed at the bottom of the main body for providing adsorption force when the robot approaches the curtain wall; an auxiliary propulsion system disposed on both sides of the main body for generating thrust to adjust the relative distance between the robot and the curtain wall; and a dual winch scheduling system including two independently controlled winch devices, with the traction ropes of the two winch devices respectively connected to both sides of the main body, and the main body is moved by coordinating the release and retraction of the two traction ropes. This utility model ensures the stability and controllability of the movement process, enabling the robot to remain stable even at high speeds, thereby improving the overall efficiency and safety of curtain wall cleaning.

Description

Technical Field

[0001] This utility model relates to the field of curtain wall cleaning equipment, and more specifically, to an automated cleaning robot for glass curtain walls. Background Technology

[0002] With the rapid development of modern building technology, glass curtain walls have been widely used as building envelopes in high-rise and super high-rise buildings. Their appearance adds a modern touch to the city, but the problem of cleaning the curtain wall surface has become increasingly prominent. Traditional cleaning methods mainly rely on manual labor, which is inefficient and yields inconsistent cleaning results. To overcome the drawbacks of manual cleaning, various automated or semi-automated cleaning equipment has emerged in the industry.

[0003] A patent application with publication number CN115886657A was found, disclosing a negative pressure adsorption self-climbing curtain wall cleaning robot. Addressing existing problems, the following solution is proposed: It includes a main frame, an upper cleaning module, a lower cleaning module, a rotor section, a water system, a moving winch, and a negative pressure chamber. A safety mechanism is installed on the main frame. Both the upper and lower cleaning modules are mounted on the main frame, and the rotor section is hinged to the front of the main frame. It has advantages such as high safety factor, multiple ropes for operation, and long service life.

[0004] Similar to the solutions described above, existing technologies also have the following drawbacks: First, installing a track or single hoisting system on the top of the curtain wall allows the cleaning equipment to move along a preset path. However, such equipment is often cumbersome to move, has a limited scheduling range, and is difficult to effectively handle curtain wall structures with complex protrusions, concavities, or large spans, lacking obstacle-crossing flexibility. Second, in existing vertical rope solutions, the high load characteristics of the robot body (the robot needs to carry heavy batteries, high-power cleaning mechanisms, or water bags, etc.) result in huge inertia. During horizontal acceleration / deceleration, the generated inertial force can create a disruptive torque, severely altering the force distribution of each adsorption unit, posing a risk of seal failure at the front or rear end of the adsorption chamber. This inherent defect forces the system to limit its movement speed and maintain high-power adsorption throughout, resulting in low scheduling efficiency and energy economy. Third, when encountering unstable airflows such as strong winds, the robot is at risk of swaying and spinning, resulting in insufficient reliability. Utility Model Content

[0005] This utility model aims to overcome at least one of the defects of the prior art and provides an automated cleaning robot for glass curtain walls. It solves the technical problem that the existing vertical rope solution generates a tilting torque due to the large load and high inertia of the robot, which leads to the robot having to limit its movement speed and adsorption with high power consumption throughout the process, resulting in low scheduling efficiency and poor energy economy.

[0006] The technical solution adopted by this utility model is an automated cleaning robot for glass curtain walls, including a main body; an adsorption module, set at the bottom of the main body, used to provide adsorption force after the robot approaches the curtain wall; an auxiliary propulsion system, set on both sides of the main body, used to generate thrust to adjust the relative distance between the robot and the curtain wall; and a dual winch scheduling system, including two independently controlled winch devices, with the traction ropes of the two winch devices respectively connected to both sides of the main body, and the main body is moved by coordinating the release and retraction of the two traction ropes.

[0007] By placing two winches on either side of the building's top or on two buildings with a gap between them, the traction ropes of the two winches and the main body form a spatial inverted V-shaped traction configuration. This allows the robot to remain suspended throughout the operation, enabling it to flexibly traverse complex protrusions and concave structures on the curtain wall surface, greatly enhancing the robot's three-dimensional scheduling capabilities and terrain adaptability. Under this architecture, the robot can completely shut down the adsorption module and auxiliary propulsion system during traction movement. Only after being pulled to the target position does the auxiliary propulsion system provide thrust to bring it against the curtain wall, subsequently activating the adsorption module for fixation. This achieves "on-demand start and stop" of the adsorption function, significantly reducing system energy consumption and improving the overall energy economy of the machine. Furthermore, the inverted V-shaped structure formed by the dual winch traction ropes naturally generates a horizontal component force, effectively counteracting the robot's swaying tendency due to inertia during horizontal movement. This ensures the stability and controllability of the movement process, allowing the robot to remain stable even at high speeds, thereby improving the overall efficiency and safety of curtain wall cleaning.

[0008] Furthermore, the adsorption module includes multiple independent sealed chambers, each equipped with a negative pressure adsorption mechanism. By setting up multiple independent sealed chambers, a redundant adsorption system is formed. Even if one or more chambers fail to seal due to uneven glass or encountering obstacles, the remaining chambers can still provide sufficient adsorption force, greatly improving the safety and reliability of the robot.

[0009] Furthermore, the negative pressure adsorption mechanism includes a fan, a pressure sensor, and a controller. The fan is mounted on the main body, with its air inlet connected to a corresponding sealed chamber. The pressure sensor is mounted on the main body and is used to detect the robot's posture data. The controller is integrated within the main body and adjusts the fan speed according to the parameters of the pressure sensor. By sensing changes in the robot's posture in real time through the pressure sensor and dynamically adjusting the fan speed by the controller to change the adsorption force, adaptive and precise control of the adsorption force is achieved. This ensures strong adsorption while avoiding excessive energy consumption, thus improving the system's energy economy.

[0010] Furthermore, the auxiliary propulsion system includes two rotor modules mounted on both sides of the main fuselage for adjusting the attitude of the main fuselage. The use of independently controlled rotor modules on both sides allows for flexible and rapid adjustment of the robot's yaw attitude in the air through differential thrust, enabling it to accurately align with and conform to the curtain wall, effectively enhancing the robot's obstacle-crossing ability; it also improves the robot's wind resistance.

[0011] Furthermore, it also includes a multi-sensor perception system mounted on the main body for environmental perception, path planning, and safety monitoring. This multi-sensor perception system includes a lidar sensor, an IMU module, and ultrasonic sensors. The lidar sensor, mounted at the top center of the main body, scans the surrounding environment of the glass curtain wall, constructs a local map, and plans the path. The IMU module, also mounted at the top center of the main body and integrated with the lidar sensor, perceives the robot's posture and trajectory in real time, providing data for navigation and path correction. The ultrasonic sensors, mounted at the front and rear ends of the main body, detect the distance between the robot and the glass curtain wall, as well as obstacles. By integrating this multi-sensor perception system, comprehensive and real-time perception of the working environment (such as building outlines and obstacles) and the robot's own state (such as posture and distance) is achieved, providing core data support for the robot's intelligent path planning, precise navigation, and safe obstacle avoidance.

[0012] Furthermore, the multi-sensor perception system also includes a thermal imaging camera, installed at the bottom center of the main body, for detecting the distribution of stains on the glass curtain wall surface and the surrounding environment. Through the coordinated configuration of lidar sensors, IMU, thermal imaging camera, and ultrasonic sensors, a multi-layered perception network combining long-range, mid-range, and short-range sensing, as well as visual, inertial, and distance perception, is constructed, ensuring that the robot can achieve accurate and reliable navigation and operation under various lighting and weather conditions.

[0013] Furthermore, the winch device is equipped with a tension sensor and an emergency braking mechanism. The tension sensor detects the tension of the traction rope, and the emergency braking mechanism brakes the traction rope immediately when the tension sensor detects abnormal data. By linking the tension monitoring of the winch device with the adsorption pressure monitoring of the robot body, a system-level safety redundancy protection mechanism is constructed. This design means that dangers caused by either traction system malfunctions (such as cable jamming or entanglement) or adsorption system failures (such as insufficient negative pressure) can be quickly identified and triggered by emergency braking. This provides comprehensive safety protection for the entire process of robot high-altitude operations, addressing multiple risk sources and significantly reducing the risk of falls due to single system failures.

[0014] Furthermore, it also includes two sets of cleaning rollers and auxiliary wheels installed at the front and rear ends of the main body, respectively. The length of the cleaning rollers is adapted to the width of the body. Each set of cleaning rollers has an auxiliary wheel on each side of its front end, and the auxiliary wheels are made of polyurethane material. The cleaning rollers, together with the polyurethane auxiliary wheels, can stably maintain the optimal working distance between the bristles and the glass, ensuring both uniform cleaning effect and preventing excessive wear of the glass by the bristles.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: By placing two winch devices on either side of the building's top or on two buildings with a gap, the traction ropes of the two winches and the main body form a spatial inverted V-shaped traction configuration. Under this architecture, the robot can completely shut down the adsorption module and auxiliary propulsion system during hoisting and movement. Only after being pulled to the target position does the auxiliary propulsion system provide thrust to make it adhere to the curtain wall, and then the adsorption module is activated to complete the fixation. This achieves "on-demand start and stop" of the adsorption function, significantly reducing system energy consumption and improving the overall energy economy of the machine. In addition, the inverted V-shaped structure formed by the double winch traction ropes naturally generates a horizontal component force, which can effectively counteract the swaying tendency of the robot due to inertia during horizontal movement, ensuring the stability and controllability of the movement process. This allows the robot to remain stable even at high speeds, thereby improving the overall efficiency and operational safety of curtain wall cleaning. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the working scenario of the automated cleaning robot for glass curtain walls.

[0017] Figure 2 This is a schematic diagram of the top structure of the automated cleaning robot for this glass curtain wall.

[0018] Figure 3 This is a schematic diagram of the bottom structure of the automated cleaning robot for glass curtain walls.

[0019] Figure 4 This is a schematic diagram showing the location of the ultrasonic sensor at the bottom of the automated cleaning robot for this glass curtain wall.

[0020] In the diagram: 1. Main body; 2. Adsorption module; 3. Cleaning roller brush; 4. Ultrasonic sensor; 5. Auxiliary wheel; 6. Auxiliary propulsion system; 7. LiDAR sensor; 8. IMU module; 9. Pressure sensor; 10. Thermal imaging camera; 11. Traction rope; 12. Winching device; 13. Traction ring. Detailed Implementation

[0021] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this invention. To better illustrate the following embodiments, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0022] This solution discloses an automated cleaning robot for glass curtain walls, including a main body 1, a dual-winch scheduling system, an adsorption module 2, an auxiliary propulsion system 6, cleaning execution components, a multi-sensor perception system, a safety protection system, and an intelligent control system. These modules work together to form a complete operational system, with the specific structure as follows: Main Body 1: The main body 1 is the core load-bearing structure of the robot, made of lightweight, high-strength aluminum alloy, with a weight controlled between 30-35kg, balancing structural strength and flexibility. The body integrates an intelligent control system and a power module (powered by a lithium battery pack, providing approximately 8 hours of runtime). Externally, it has pre-installed interfaces for various sensors and cleaning execution components, ensuring stable assembly of each module and enabling data interaction.

[0023] Dual-winch dispatching system: As a core dispatching component, it serves as the "power center" for large-scale spatial dispatching. It comprises two independent winch units 12 mounted on the rooftop, symmetrically arranged at both ends of the rooftop work area. Each winch unit 12 is equipped with a servo motor, precision reducer, and tension sensor, allowing independent control of the length and speed of the traction rope 11. Traction rings 13 are symmetrically positioned on both sides of the top of the main body 1, with the centerline of the main body 1 as the axis of symmetry. These rings are made of high-strength alloy (e.g., CrNiMo alloy steel (such as 42CrMo) or 4140 alloy steel), with a load-bearing capacity ≥500kg, reducing the risk of falls. The traction ropes 11 of the two winch units 12 are connected to the traction rings 13 on both sides via universal joints, ensuring balanced force on the body during rope retraction and detour without twisting.

[0024] Adsorption module 2: Located at the bottom of the main body 1 (attached) Figure 3The system comprises multiple independent sealed chambers, each approximately 180mm × 160mm × 35mm in size. Each chamber is equipped with a fan, a pressure sensor 9, and a controller. The fan inlet connects to a pre-drilled hole in the inner wall of the sealed chamber, primarily providing negative pressure. The pressure sensor 9 is installed at the bottom of the main body to monitor the adsorption pressure in real time. When the pressure is abnormal, adjustments can be made promptly to ensure the reliability of adsorption. Specifically, the controller is integrated into the main body 1 and electrically connected to the fan and pressure sensor 9. It controls the fan speed based on the electrical signal transmitted by the pressure sensor 9 to adjust the adsorption force. Each chamber has its own controller. Even if some chambers experience leakage, the remaining chambers can still maintain adsorption force, greatly ensuring the stability of the robot's adsorption and preventing overall adsorption failure due to a single chamber malfunction, such as partial suspension.

[0025] The sealed chamber comprises a sealing suction frame, silicone foam strips, and nylon cloth. The sealing suction frame uses a material with moderate hardness, such as rubber, which provides both support and adjustable suction force. For example, while enhancing support strength, it can adaptively conform to the slight deformation of the glass curtain wall. When the suction pressure is adjusted, the elastic rubber can compress or rebound synchronously, precisely matching the sealing gap under different negative pressure requirements. The ends of the sealing suction frame are fixed with foam strips wrapped in nylon cloth, forming a sealing contact interface with the glass curtain wall surface. This ensures resilience, wear resistance, durability, and protection against scratches on the glass panel. After repeated testing, this design fully meets practical functions. Compared to the existing metal rigid frame and sealing strip structure (or complex pneumatic sealing components), it also solves the cost problems of complex structure and high cost in existing technologies.

[0026] The auxiliary propulsion system helps the robot move more flexibly on the glass curtain wall and adapt to different cleaning paths. It includes two rotor modules located on both sides of the main body 1. Each rotor module comprises a motor, blades, and arms, which are technologies well-known to those skilled in the art; their components and working principles will not be detailed here. The blades are made of carbon fiber (lightweight and high-strength, single-wing weight <1kg), with a wingspan of 300-400mm. The motors are brushless motors, and the thrust is adjusted by regulating the motor speed. After the dual winches transport the robot to a certain distance from the curtain wall (usually tens of centimeters), the control system controls the servo motors to drive the blades at high speed, generating thrust, like an "aerial pusher," quickly pushing the suspended robot to the curtain wall surface. Different blade speeds on both sides result in different thrusts, allowing adjustment of the deflection angle of the main body 1. During this process, the lidar sensor 7 and ultrasonic sensor 4 in the multi-sensor perception system continuously calibrate the distance and angle to ensure the robot's stable and vertical contact with the curtain wall, preparing for vacuum adsorption. The adsorption module 2 and auxiliary propulsion system can effectively enhance wind resistance, with an adsorption failure rate of <1% in strong wind (level 5 and below) environments.

[0027] Cleaning execution components: including front and rear sets of cleaning roller brushes 3 and auxiliary wheels 5 (attached) Figure 2 The length of the cleaning roller brush 3 is adapted to the width of the machine body (1200-1500mm). Each set of cleaning roller brushes 3 has a polyurethane auxiliary wheel 5 (50mm in diameter) on each side of the front end. The wheel surface is smooth to reduce friction with the glass curtain wall. During cleaning, the cleaning roller brush 3 is driven by a servo motor (speed adjustable, 50-150r / min) mounted on the main body 1, rotating clockwise / counterclockwise alternately to achieve efficient removal of dust and stains from the glass surface. The auxiliary wheels 5 roll on the curtain wall surface, providing support and ensuring that the cleaning roller brush 3 maintains a constant distance (3-5mm) from the glass surface, preventing the bristles from being excessively squeezed or detached from the glass, thus ensuring the optimal cleaning distance.

[0028] Multi-sensor perception system: includes four main categories of sensors, namely LiDAR sensor 7, IMU module 8, ultrasonic sensor 4, and thermal imaging vision camera 10, as detailed below: The lidar sensor 7 is installed at the top center of the main body 1. It is preferably a 2D lidar sensor 7 (with a detection range of 0.1-50m and an angle of 360°) to scan the surrounding environment of the glass curtain wall, build a local map, identify the building outline, curtain wall corners, etc., and provide a basis for path planning.

[0029] IMU module 8, integrated with LiDAR sensor 7, includes a six-axis IMU (using an accelerometer with an accuracy of ±0.01g and a gyroscope with an accuracy of ±0.001° / s). It is used to sense the robot's attitude (tilt angle, pitch angle) and motion trajectory (speed, displacement), preventing the robot from deviating from the preset path. LiDAR sensor 7 and IMU module 8 work together. LiDAR sensor 7 scans and acquires environmental information, combining it with a preset 3D model of the glass curtain wall. IMU module 8 provides real-time feedback on the robot's attitude and motion data. Through algorithm processing, the robot achieves precise navigation on the glass curtain wall, ensuring that the robot performs cleaning operations according to the predetermined path.

[0030] Four ultrasonic sensors (4 in total) are installed: two at the front and rear ends of the bottom of the main body 1 (approximately 50mm above the ground when the robot is placed on the ground), and the other two at the front ends of the two cleaning roller brushes 3 (at least 30mm from the sides of the cleaning roller brushes 3). Sensors with a detection range of 0.02-4m are selected. These sensors are used to detect the distance between the robot and the glass curtain wall (ensuring a gap of <1mm during adsorption) and to determine the distance to obstacles (such as glass seams and metal frames) to avoid collisions. A lidar sensor 7 (0.1-50m detection range) constructs a building outline map for long-distance path planning. The four ultrasonic sensors 4 (0.02-5m detection range) calibrate the distance to the curtain wall in real time (accuracy ±1mm) and identify nearby obstacles (such as protruding decorations). The data from both sensors are fused using a Kalman filter (50Hz frequency) to solve the problem of "blind spot perception in complex curtain wall environments."

[0031] The thermal imaging vision camera 10 is installed at the bottom center of the main body 1. It is preferred to have a resolution of at least 640×512 and a frame rate of at least 30fps. In low light conditions or complex weather conditions such as night, rain, and fog, the thermal imaging vision camera 10 can clearly capture the distribution of stains on the glass curtain wall surface (identifying stain areas through temperature differences) and the robot's surrounding environment to assist in judging the cleaning effect and environmental safety, and can effectively improve the cleaning effect in harsh environments.

[0032] The safety protection system includes a dual protection system of "active early warning + passive backup". Active early warning: When the pressure sensor 9 detects that the adsorption pressure is lower than -60kPa, or the tension sensor of the winch device 12 detects abnormal rope tension (>800N or <200N), an early warning is immediately triggered and the operation is suspended. Passive backup: The traction rope 11 of the double winch (i.e., the winch device 12) is made of cut-resistant steel wire rope (8mm in diameter, breaking strength ≥10kN). The winch device 12 integrates an emergency braking device, which can quickly lock the rope to lock the machine position when the adsorption fails.

[0033] Working principle: Taking the cleaning of a high-rise building's curtain wall as an example: Before operation, two independent winches 12 are installed at both ends of the rooftop. The robot is positioned at the edge of the rooftop, and the steel wire ropes of the winches 12 for traction are connected to the traction rings 13 via universal joints. The control system is activated, the multi-sensor perception system begins to work, the lidar sensor 7 scans the building outline, and performs global path planning in conjunction with the preset 3D curtain wall model. The IMU module 8 is initialized to determine the robot's initial posture.

[0034] Using a dual-winch device 12 mounted on the rooftop, the robot is propelled to move in three-dimensional space at high altitude by extending and retracting the traction rope 11 connected to it. Based on a pre-planned path and real-time obstacle locations, the winch device 12 precisely adjusts the rope length and extension / retraction speed, allowing the robot to traverse obstacle areas or move to the vicinity of the target work area, completing "coarse positioning." When the robot is lowered to a predetermined airspace tens of centimeters above the curtain wall surface, the auxiliary propulsion system is activated. The blades on both sides rotate at high speed according to commands from the control system, generating thrust, acting like an "aerial pusher" to smoothly and quickly push the suspended robot towards the curtain wall. Simultaneously, ultrasonic sensors 4 monitor the gap between the robot and the curtain wall in real time, ensuring the robot approaches precisely in a vertical posture during propulsion, preparing for adsorption.

[0035] When the robot approaches the curtain wall to a gap of only millimeters, the adsorption module 2 immediately activates. Fans in each chamber start, creating negative pressure within the sealed chamber, allowing the robot to firmly adhere to the glass surface. Pressure sensor 9 monitors the adsorption status of each chamber in real time, and the controller dynamically adjusts the fan speed accordingly to maintain a stable adsorption force. Even if a leak occurs in one chamber, the remaining chambers can still operate independently, ensuring overall safety.

[0036] After adsorption stabilizes, the robot enters the cleaning operation mode. At this time, the ropes of the dual winch device 12 are in a slack and ready state. During subsequent cleaning operations, if small adjustments to the position are required (such as avoiding minor obstacles), the dual winch device 12 can assist in fine-tuning, or work in conjunction with the auxiliary propulsion system and adsorption module to achieve more flexible operation coverage. The IMU module 8 continuously provides feedback on the robot's posture, and the lidar sensor 7 and ultrasonic sensor 4 work together to detect the surrounding environment and obstacles (such as frames and seams) to ensure accurate movement paths and avoid collisions. At the same time, the cleaning execution components begin to work. The front and rear cleaning rollers 3, driven by servo motors, rotate alternately in both directions at adjustable speeds to clean the glass surface. Other cleaning equipment can also be equipped to achieve scrubbing cleaning. The polyurethane auxiliary wheel 5 at the front end of the cleaning roller 3 rolls on the glass, maintaining a contact distance of 3-5mm between the bristles and the glass surface to maintain optimal cleaning conditions, ensuring cleaning efficiency while protecting the glass from scratches. In poor lighting or complex weather conditions, the thermal imaging vision camera 10 can assist in identifying stained areas and optimizing the cleaning path.

[0037] Throughout the entire operation, the safety protection system provides continuous monitoring. If the pressure sensor 9 detects an abnormal drop in adsorption pressure (typically below -60 kPa) or the tension sensor of the winch 12 detects that the rope tension exceeds the safe range (e.g., >800 N or <200 N), the system will immediately trigger an alarm and suspend the operation. In the event of extreme situations such as adsorption failure, the emergency braking device within the winch 12 will instantly lock the rope, securely suspending the robot in its original position, achieving a "passive safety net."

[0038] When sensing elements such as the lidar sensor 7 and ultrasonic sensor 4 detect obstacles such as protrusions on the curtain wall surface or floating objects in the air, the control system quickly instructs the adsorption module 2 to stop adsorption and simultaneously shuts down the rotor power. This step is to prevent the robot from forcibly crossing obstacles while adsorbing the curtain wall, which could damage the equipment and the curtain wall, and also to prevent the rotor from getting caught in debris and causing malfunctions.

[0039] After cleaning the designated area, the winch system retrieves the rope, lifts the robot, and safely returns it to the rooftop, ending the operation.

[0040] In high-altitude planar curtain wall cleaning operations, the rooftop double winches are the "power center" for robots to achieve large-scale spatial scheduling. Their operation, in coordination with various modules, constructs a precise work process.

[0041] Obviously, the above embodiments of this utility model are merely examples for clearly illustrating the technical solution of this utility model, and are not intended to limit the specific implementation of this utility model. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of this utility model should be included within the protection scope of the claims of this utility model.

Claims

1. An automated cleaning robot for glass curtain walls, characterized in that: include Main fuselage (1); The adsorption module (2) is located at the bottom of the main body (1) and is used to provide adsorption force when the robot approaches the curtain wall; An auxiliary propulsion system is set on both sides of the main body (1) to generate thrust to adjust the relative distance between the robot and the curtain wall; The dual-winch dispatching system includes two independently controlled winch devices (12). The traction ropes (11) of the two winch devices (12) are respectively connected to both sides of the main body (1). The main body (1) is moved by coordinating the release and retraction of the two traction ropes (11).

2. The automated cleaning robot for glass curtain walls according to claim 1, characterized in that: The adsorption module (2) includes multiple independent sealed chambers, each of which is equipped with a negative pressure adsorption mechanism.

3. The automated cleaning robot for glass curtain walls according to claim 2, characterized in that: The sealed chamber includes a sealing adsorption frame, a silicone foam sponge strip, and nylon cloth; the port of the sealing adsorption frame is fixed with a silicone foam sponge strip wrapped in nylon cloth, forming a contact interface that seals with the surface of the glass curtain wall.

4. The automated cleaning robot for glass curtain walls according to claim 1, characterized in that: The auxiliary propulsion system includes two rotor modules installed on both sides of the main fuselage (1) for adjusting the attitude of the main fuselage (1).

5. The automated cleaning robot for glass curtain walls according to claim 1, characterized in that: It also includes a multi-sensor perception system installed on the main body (1) for environmental perception, path planning and safety monitoring.

6. The automated cleaning robot for glass curtain walls according to claim 5, characterized in that: The multi-sensor perception system includes a lidar sensor (7), an IMU module (8), and an ultrasonic sensor (4). The lidar sensor (7) is installed at the top center of the main body (1) to scan the surrounding environment of the glass curtain wall, build a local map and plan the path; The IMU module (8) is installed at the top center of the main body (1) and integrated with the lidar sensor (7) to perceive the robot's posture and motion trajectory in real time, providing data for navigation and path correction; An ultrasonic sensor (4) is installed at both ends of the main body (1) to detect the distance between the robot and the glass curtain wall and obstacles.

7. The automated cleaning robot for glass curtain walls according to claim 6, characterized in that: The multi-sensor perception system also includes a thermal imaging vision camera (10), which is installed at the bottom center of the main body (1) to detect the distribution of stains on the surface of the glass curtain wall and the surrounding environment.

8. The automated cleaning robot for glass curtain walls according to claim 1, characterized in that: It also includes a pressure sensor (9), which is installed at the bottom of the main body (1) to detect the pressure between the main body (1) and the glass curtain wall.

9. The automated cleaning robot for glass curtain walls according to claim 8, characterized in that: The winch (12) is equipped with a tension sensor and an emergency braking mechanism. The tension sensor is used to detect the tension of the traction rope (11), and the emergency braking mechanism is used to brake the traction rope (11) in an emergency when the tension sensor detects abnormal data.

10. An automated cleaning robot for glass curtain walls according to any one of claims 1-9, characterized in that: It also includes two sets of cleaning roller brushes (3) and auxiliary wheels (5) installed at the front and rear ends of the main body (1), respectively. The length of the cleaning roller brushes (3) is adapted to the width of the body. Each set of cleaning roller brushes (3) has an auxiliary wheel (5) on each side of the front end. The auxiliary wheels (5) are made of polyurethane material.

Images