Multifunctional cross-row operation photovoltaic cleaning robot, cleaning system and control method thereof

Abstract

The invention discloses a multifunctional cross-row operation photovoltaic cleaning robot, a cleaning system and a control method thereof.The cleaning system comprises a robot and a transfer trolley, and the robot comprises a robot body, a sensing system, a moving mechanism, a suspension module, an adsorption system, a sweeping system and a control unit; scenes of irregular arrangement and abnormal height difference between photovoltaic panels can be intelligently identified through multi-sensor fusion, and mechanisms such as lifting of a suspension wheel set, opening and closing of vacuum adsorption and pressing and supporting of a roller brush are correspondingly controlled to cooperatively complete stable crossing. The matched transfer trolley is provided with a track driving mechanism, an RTK positioning system, a scissor-fork type lifting platform, an inclination angle adjusting mechanism and a sensing obstacle avoidance system, autonomous transfer and accurate butt joint of the robot among different photovoltaic array rows can be achieved, the problems that an existing robot cannot adapt to complex working conditions, and row crossing operation is difficult are effectively solved, and the working efficiency of the robot is improved. And full-automatic, high-efficiency and high-reliability photovoltaic power station cleaning operation and maintenance are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic panel cleaning, in particular to a multifunctional cross-row operation photovoltaic cleaning robot, a cleaning system and a control method thereof. BACKGROUND

[0002] With the wide application of photovoltaic power generation technology, the cleaning and maintenance demand of large-scale photovoltaic power stations is increasingly prominent. The dust and snow accumulation on the surface of photovoltaic panels can significantly reduce the power generation efficiency, and therefore regular cleaning is required. At present, the manual cleaning method has problems such as low efficiency, high cost and poor safety, and is difficult to implement in large-scale and complex terrain power stations.

[0003] Most of the existing cleaning robots are designed for continuous and flat photovoltaic panel arrays, and face two major technical bottlenecks in actual application. First, due to installation errors, thermal expansion and contraction or foundation settlement, misalignment, gaps or irregular gaps often exist between photovoltaic panel components in the same row, which leads to robot tracking failure, jamming or falling. Second, in order to improve land utilization, photovoltaic arrays are often arranged according to the terrain, and there are significant differences in height and inclination between different array rows, which makes it difficult for the robot to autonomously and safely transfer and operate across rows.

[0004] The robots in the prior art lack precise perception and adaptive crossing ability for such complex abnormal scenarios, and their moving mechanisms, obstacle crossing strategies and overall control systems are often single-function, which cannot achieve efficient cleaning while considering stable crossing and fully autonomous operation under complex working conditions. SUMMARY

[0005] The purpose of the present application is to provide a multifunctional cross-row operation photovoltaic cleaning robot, a cleaning system and a control method thereof, which integrates intelligent perception, accurate decision-making and multi-mechanism collaborative control, and realizes fully autonomous and efficient cleaning operation of the robot in complex photovoltaic arrays.

[0006] To achieve the above purpose, the present application realizes the following technical scheme: a multifunctional cross-row operation photovoltaic cleaning robot, the robot comprising a robot body, the robot body being provided with: a perception system installed on the robot body, for identifying the height difference of photovoltaic panel components and determining the pollution type; a moving mechanism installed on both sides of the robot body, for driving the robot body to travel on the surface of the photovoltaic panel; a suspension module installed on one side of the robot body, for being suspended on the edge of the photovoltaic panel to provide support for stable operation of the robot body; an adsorption system provided at the bottom of the robot body, for providing stable support when crossing abnormal scenarios; The cleaning system is installed on the side of the robot body opposite to the suspension module, and at both the front and rear ends of the robot body; The control unit is located inside the robot body and is electrically connected to the sensing system, the moving mechanism, the suspension module, the adsorption system, and the cleaning system. The robot body is also equipped with a nine-axis inertial measurement unit for real-time monitoring of the robot body's tilt angle and acceleration.

[0007] A further technical solution is that the sensing system includes: LiDAR is used to detect the layout of photovoltaic panels, and to detect the edge position deviation and height difference of photovoltaic panels; A binocular vision camera is used to acquire images of the components in front and to verify the recognition results of the LiDAR using an image matching algorithm; An ultrasonic sensor is used for detection and collision avoidance, assisting the lidar and the binocular vision camera in verifying the results.

[0008] A further technical solution is provided, wherein the cleaning system includes: The front roller brush is connected to the robot body via an electric drive mechanism and is used to control the lifting and lowering of the front roller brush to achieve cleaning and obstacle crossing operations. The rear roller brush is connected to the robot body via an electric drive mechanism and is used to control the rear roller brush to lift and lower, so as to achieve cleaning and obstacle crossing operations. Pressure sensors are installed on both the front roller brush and the rear roller brush; The lifting roller brush achieves longitudinal lifting and lowering along the tilt direction of the photovoltaic panel through an electric winch, cable, and guide pulley assembly. The electric winch has a self-locking function to ensure that the position is maintained during the lifting and lowering process. It is driven to rotate by a drive motor, which is mounted on the mounting frame of the lifting roller brush. The electric winch is mounted on the robot body, and the cable is wound around the electric winch, with its free end fixed to the mounting frame of the lifting roller brush.

[0009] In a further technical solution, the electric drive mechanism is mounted on the robot body and includes two electric push rods, which are used to control the movement of the front roller brush and the rear roller brush, respectively. One end of each of the two electric push rods is hinged to the robot body, and the other end is hinged to the front roller brush and the rear roller brush, respectively.

[0010] In a further technical solution, the moving mechanism is a tracked moving mechanism, including a left drive motor, a right drive motor, a left track structure, and a right track structure, which are driven independently.

[0011] The adsorption system includes a negative pressure fan and multiple adsorption chambers.

[0012] In a further technical solution, the suspension module includes: The system includes a suspension arm and a wheel assembly, wherein the suspension arm is fixed to the robot body and the wheel assembly is hinged to the suspension arm, and can be raised or lowered under the command of the control unit. It also includes a power source that is poweredly connected to the wheel assembly.

[0013] A multi-functional photovoltaic cleaning system for cross-row operations includes the aforementioned robot and a transfer vehicle. The transfer vehicle is used to transfer the robot and includes: A mobile chassis, wherein a track drive mechanism is provided at the bottom of the mobile chassis for omnidirectional autonomous movement; A lifting platform is mounted on a mobile chassis. The lifting platform includes a scissor lift structure and a support frame, which is used to support the robot body. The RTK positioning system, integrated into a mobile chassis, is used to achieve centimeter-level precise positioning. The tilt adjustment mechanism is connected between the scissor lift structure and the support frame, and is used to adjust the tilt angle of the support frame to match the tilt angle of the photovoltaic panel; The perception and obstacle avoidance system, mounted on a mobile chassis, includes a multi-line LiDAR and a wide-angle vision camera, used to build environmental maps and avoid obstacles in real time; The tilt adjustment mechanism includes a hydraulic rod for changing the angle between the support frame and the top of the scissor lift structure.

[0014] A control method for a multifunctional cross-row photovoltaic cleaning system, the control method being based on the aforementioned cleaning system, comprising the following steps: Perception step: Acquire environmental data collected by the robot's perception system; Decision-making steps: Based on the aforementioned environmental data, identify any abnormal scenarios that need to be crossed between photovoltaic panel modules; Control steps: In response to the detection of the abnormal scene, generate and issue a corresponding set of control instructions to adjust the posture of the robot's actuator, plan the travel path and control the mobile mechanism to perform traversing actions; Recovery steps: After confirming the completion of the crossing based on the aforementioned environmental data, generate and issue instructions to resume the robot's regular cleaning operation.

[0015] A further technical solution is that the abnormal scenario is that the photovoltaic panel modules in the same row are not neatly arranged; The decision-making steps include: generating a first recognition signal when the deviation of the component edge position exceeds a first threshold based on the LiDAR scanning data, and confirming the scenario after verification by a feature point matching algorithm based on visual image data; The control steps include: generating a first set of control commands to control the suspension mechanism to lift the wheel assembly and start the multi-chamber vacuum adsorption module; subsequently, planning an oblique travel path based on the fusion data of the inertial measurement unit and the lidar, and performing the crossing by controlling the left drive motor and the right drive motor through differential speed control.

[0016] A further technical solution is that the abnormal scenario is that there is a height difference between photovoltaic panels in the same row; The decision-making steps include: generating a second recognition signal when the height difference of the components in front is detected to be within a threshold range based on the LiDAR scanning data, and confirming the scene after verification by a stereo matching algorithm based on the binocular vision three-dimensional image data; The control steps include: generating a second set of control commands to control the front fixed roller brush to descend and press against the surface of the adjacent high component; subsequently, based on feedback data from the pressure sensor and the inertial measurement unit, controlling the front roller brush motor to output support torque and coordinate with the moving mechanism to perform a crossing.

[0017] In summary, this invention offers the following advantages: It provides a multifunctional photovoltaic cleaning system for cross-row operations, comprising a robot and a transport vehicle. This system boasts advantages of full automation, high efficiency, and high reliability. The robot, through a perception system integrating LiDAR and binocular vision cameras, can intelligently identify abnormal scenarios such as uneven arrangement and height differences on the photovoltaic panel surface and autonomously initiate corresponding crossing strategies. Its tracked mobile mechanism, lifting suspension module, and vacuum adsorption system work synergistically, combined with a controllable roller brush with pressure feedback, enabling the robot to smoothly cross gaps and steps, effectively solving the technical challenges of jamming and falling under complex working conditions. Simultaneously, the accompanying transport vehicle possesses RTK positioning, scissor lift, and hydraulic tilt adjustment functions, enabling precise docking and autonomous transport of the robot between different photovoltaic arrays. This compensates for situations where the robot may be unable to autonomously cross rows, with the transport vehicle handling the transport and crossing. The entire system forms a complete closed-loop operation where "the robot focuses on cleaning, and the transport vehicle is responsible for scheduling," significantly improving the automation level and cleaning efficiency of photovoltaic power plant operation and maintenance. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a first three-dimensional schematic diagram of the robot body of this application; Figure 2 This is a second three-dimensional schematic diagram of the robot body of this application; Figure 3 This is a third-dimensional schematic diagram of the robot body of this application; Figure 4 This is a first three-dimensional schematic diagram of the transfer vehicle used in this application; Figure 5 This is a second perspective view of the transfer vehicle in this application; Figure 6 This is a schematic diagram of the robot's cross-row operation in this application; Figure 7 This is a schematic diagram illustrating a scenario where the photovoltaic panels are not neatly arranged, as described in Embodiment 1 of this application. Figure 8 This is a schematic diagram illustrating the crossing of the photovoltaic panel height difference scenario in Embodiment 2 of this application; Figure 9 This is a schematic diagram of the control method of this application.

[0019] In the diagram: 101. Main body; 102. Rear roller brush; 103. Lifting roller brush; 104. Electric winch; 105. Suspension module; 106. Casters; 107. Adsorption system; 108. Cable; 109. Moving mechanism; 110. Front roller brush; 111. Second drive source; 112. Rotary wheel; 113. Sensing system; 114. Suspension arm; 115. Wheel assembly; 116. Power source; 201. Mobile chassis; 202. Lifting platform; 203. Multi-line lidar; 204. Track drive mechanism; 205. Scissor lift structure; 206. RTK positioning system; 207. Support frame; 208. Hydraulic rod; 209. DC motor. Detailed Implementation

[0020] To more clearly illustrate the overall concept of this application, a detailed explanation is provided below with reference to the accompanying drawings.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below.

[0022] Furthermore, it should be understood in the description of this application that the terms "top", "bottom", "inner", "outer", "axial", "radial", "circumferential", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0023] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a communication connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] In this application, unless otherwise expressly specified and limited, the "above" or "below" of the second feature can mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. In the description of this specification, references to terms such as "an embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples.

[0025] like Figures 1-8 As shown, a multi-functional photovoltaic cleaning robot for cross-row operations is disclosed. The robot includes a robot body, on which are provided: The sensing system 113 is installed on the robot body and is used to identify the height difference of the photovoltaic panel components and determine the type of contamination on the surface of the photovoltaic panel. The moving mechanism 109 is used to drive the robot body to move on the surface of the photovoltaic panel; Suspension module 105 is used to suspend the edge of the photovoltaic panel to provide support for the stable operation of the robot body; Adsorption system 107 is used to provide stable support when crossing abnormal scenarios; The cleaning system is installed on the side of the robot body opposite to the suspension module 105, as well as at the front and rear ends of the robot body. It can not only clean the dust on the photovoltaic panel, but also clean the snow on the photovoltaic panel, realizing the function of one machine for multiple purposes. The control unit is located inside the robot body and is electrically connected to the sensing system 113, the moving mechanism 109, the suspension module 105, the adsorption system 107, and the cleaning system.

[0026] Specifically, the robot body includes a main body 101, on which a sensing system 113, a moving mechanism 109, a suspension module 105, an adsorption system 107, and a cleaning system are installed.

[0027] The sensing system 113 includes: a lidar for detecting the layout of photovoltaic panels and detecting edge position deviations and height differences of the photovoltaic panels; A binocular vision camera is used to acquire images of the components in front, and to determine the type of contamination through an image matching algorithm, as well as to verify the recognition results of the lidar; An ultrasonic sensor is used for detection and collision avoidance, assisting the lidar and the binocular vision camera in verifying the results.

[0028] The mobile mechanism 109 is a tracked structure, including a left drive motor, a right drive motor, a left track structure, and a right track structure, which are driven independently. The left drive motor and the right drive motor are used to drive the left track structure and the right track structure, respectively. Differential speed control enables the robot to turn and move diagonally. The track is a wide-body rubber track, which has the advantages of low ground pressure, good adhesion, and strong obstacle crossing ability compared to a wheeled structure. Especially on photovoltaic panels with snow, slippery surfaces, or dust, it ensures the accuracy of the cleaning path.

[0029] The suspension module 105 includes a suspension arm 114 and a wheel assembly 115. The suspension arm 114 is fixed to the main body 101 of the robot body. The wheel assembly 115 is hinged to the suspension arm 114 and can be raised or lowered under the command of the control unit. The wheel assembly 115 can be composed of a wheel 112 and an L-shaped rod. The wheel surface is made of rubber and has vertical stripes to enhance guidance and prevent deviation or slippage on the aluminum alloy frame. During operation, the wheel assembly 115 rolls close to the top edge of the photovoltaic panel under the command of the control unit, forming a reliable "top-bottom" double constraint with the moving mechanism 109, greatly reducing the risk of the equipment tipping over or falling. It also includes a power source 116, which is an electric push rod assembly. There are two wheel assemblies 115, and the electric push rod assembly is poweredly connected to the wheel assembly 115. Specifically, in one embodiment, the electric push rod assembly is installed between the two wheel assemblies 115, with one end of the electric push rod assembly hinged to one of the wheel assemblies 115 and the other end hinged to the other wheel assembly 115. In another embodiment, there may be two electric push rod assemblies, with one end of each electric push rod assembly hinged to the suspension arm 114 and the other end hinged to the wheel assembly 115 through a connector. The operation of the electric push rod assembly drives the wheel assembly 115 to be raised or lowered.

[0030] The adsorption system 107 is located at the bottom of the robot body and includes a negative pressure fan and multiple adsorption chambers. The negative pressure fan is fixedly installed inside the main body 101 of the robot body. Vacuum suction cups are installed in the adsorption chambers and are connected to the negative pressure fan. The vacuum suction cups are located on the bottom surface of the robot body. Each adsorption chamber can generate negative pressure independently or synchronously, so that the robot body is adsorbed onto the surface of the photovoltaic panel. When the system determines that it has entered the large tilt angle operation mode, the vacuum generator is automatically started, so that negative pressure adsorption force is quickly generated in each vacuum suction cup. This adsorption force, together with the constraint force of the suspension module 105 and the driving force of the track in the moving mechanism 109, constitutes a stable three-dimensional force system, ensuring that the robot can maintain safety and stability when cleaning, turning or even crossing obstacles at extreme tilt angles.

[0031] The cleaning system includes a front roller brush 110, which is connected to the robot body via an electric drive mechanism to control the lifting and lowering of the front roller brush 110. The front roller brush 110 includes a first brush body, a first drive source, a first mounting frame, and a first electric drive mechanism. The first mounting frame is hinged to the main body 101, and the first brush body is rotatably mounted on the first mounting frame. The first drive source is fixed on the first mounting frame and poweredly connected to the first brush body to drive the first brush body to rotate. The first electric drive mechanism is installed between the main body 101 and the first mounting frame to drive the lifting and lowering of the first mounting frame. The rear roller brush 102 is connected to the robot body via an electric drive mechanism to control the lifting and lowering actions of the rear roller brush 102. The rear roller brush 102 includes a second brush body, a second drive source 111, a second mounting frame, and a second electric drive mechanism. The second mounting frame is hinged to the main body 101, and the second brush body is rotatably mounted on the second mounting frame. The second drive source 111 is fixed on the second mounting frame and is poweredly connected to the second brush body. The second electric drive mechanism is installed between the main body 101 and the second mounting frame to drive the lifting and lowering of the second mounting frame. Pressure sensors are installed on the shafts of both the front roller brush 110 and the rear roller brush 102 to provide real-time feedback of pressure values. The lifting roller brush 103 achieves longitudinal lifting along the tilt direction of the photovoltaic panel through the electric winch 104, cable 108 and guide pulley group. The lifting stroke can be designed from 0 to 500 mm as needed. The electric winch 104 has a self-locking function to ensure that the position is maintained during the lifting process. It is driven by a drive motor, which is mounted on the mounting bracket of the lifting roller brush 103. The electric winch 104 is mounted on the main vehicle body 101.

[0032] Specifically, the lifting roller brush 103 includes a third brush body, a third drive source, and a third mounting frame. The third brush body is rotatably mounted on the third mounting frame, and the third drive source is fixed on the third mounting frame and poweredly connected to the third brush body. Two universal wheels 106 are provided on the third mounting frame to facilitate the movement of the lifting roller brush 103 on the photovoltaic panel.

[0033] The electric drive mechanism is mounted on the robot body and includes two electric push rods, which respectively constitute the first electric drive mechanism and the second electric drive mechanism. They are used to control the movement of the front roller brush 110 and the rear roller brush 102. One end of each of the two electric push rods is hinged to the main body 101 of the robot body, and the other end is hinged to the first mounting frame of the front roller brush 110 and the second mounting frame of the rear roller brush 102, respectively, and is used to control the lifting and lowering of the first mounting frame and the second mounting frame.

[0034] Specifically, optical shafts are fixed on both the first and second mounting brackets. One end of the electric push rod of the first electric drive mechanism is hinged to the main body 101, and the other end is hinged to the optical shaft on the first mounting bracket, which is used to drive the front roller brush 110 to lift and fall. One end of the electric push rod of the second electric drive mechanism is hinged to the main body 101, and the other end is hinged to the optical shaft on the second mounting bracket, which is used to drive the rear roller brush 102 to lift and fall.

[0035] The main vehicle body 101 is also equipped with a nine-axis inertial measurement unit for real-time monitoring of the robot's tilt angle and acceleration.

[0036] A multi-functional photovoltaic cleaning system for cross-row operations includes a cleaning robot and a transfer vehicle. The transfer vehicle works in conjunction with the robot to move the robot to other photovoltaic panels. The transfer vehicle includes: Mobile chassis 201, with a track drive mechanism 204 at the bottom of the mobile chassis 201. The track drive mechanism 204 is driven by a DC motor 209 mounted on the mobile chassis 201 and is used for omnidirectional autonomous movement. A lifting platform 202 is mounted on a mobile chassis 201. The lifting platform 202 includes a scissor lift structure 205 and a support frame 207. The support frame 207 is used to support the robot body. The RTK positioning system 206 is fixedly integrated on the mobile chassis 201 and is used to achieve centimeter-level precise positioning. The tilt adjustment mechanism is connected between the scissor lift structure 205 and the support frame 207, and is used to adjust the tilt angle of the support frame 207 to match the tilt angle of the photovoltaic panel; The perception and obstacle avoidance system, mounted on the mobile chassis 201, includes a multi-line lidar 203 and a wide-angle vision camera for real-time environmental mapping and obstacle avoidance.

[0037] The tilt adjustment mechanism includes a hydraulic rod 208, used to change the angle between the support frame 207 and the top of the scissor lift structure 205.

[0038] like Figure 9 As shown, a control method for a multifunctional cross-row photovoltaic cleaning system includes the following steps: S10 Perception Step: Acquire environmental data collected by the robot's perception system. S20 Decision Step: Based on the aforementioned environmental data, identify abnormal scenarios that need to be crossed between photovoltaic panel modules; S30 control steps: In response to the identification of the abnormal scene, generate and issue a corresponding set of control instructions to adjust the posture of the robot's actuator, plan the travel path and control the mobile mechanism to perform traversing actions; S40 Recovery Step: After confirming the completion of the crossing based on the aforementioned environmental data, generate and issue instructions to resume the robot's regular cleaning operation.

[0039] When the abnormal scenario is that the photovoltaic panel modules in the same row are not arranged neatly; The S20 decision-making step includes: generating a first recognition signal when the deviation of the component edge position exceeds a first threshold based on the LiDAR scanning data, and confirming the scene after verification by a feature point matching algorithm based on visual image data; The S30 control step includes: generating a first control instruction set to control the suspension mechanism to lift the wheel assembly 115 and start the adsorption system 107; then, planning an oblique travel path based on the fusion data of the inertial measurement unit and the lidar in the perception system, and performing the crossing by controlling the left and right drive motors through differential speed control, the left and right drive motors being the left drive motor and the right drive motor.

[0040] When the abnormal scenario is that there is a height difference between photovoltaic panel modules in the same row; The decision-making steps include: generating a second recognition signal when the height difference of the components in front is detected to be within a traversable threshold range based on the LiDAR scanning data, and confirming the scene after verification by a stereo matching algorithm based on the three-dimensional image data of the binocular vision camera; The control steps include: generating a second set of control commands to control the front fixed roller brush to descend and press against the surface of the adjacent high component; subsequently, based on feedback data from the pressure sensor and the inertial measurement unit, controlling the front roller brush motor to output support torque and coordinate with the moving mechanism to perform a crossing.

[0041] Specifically: Example 1: Crossing control in scenarios with irregular arrangement: When the robot performs a horizontal cleaning operation on the photovoltaic panel, the control unit continuously processes the data collected by the sensing system 113. The specific control process is as follows: 1. Scene recognition stage: The lidar scans the photovoltaic panel layout within a 10m range in front at a frequency of 20Hz. When it detects that the deviation of the edge position of the component from the preset path exceeds 5cm, the first threshold is set to 5cm, and the first identification signal is immediately generated.

[0042] After receiving the first recognition signal, the control unit starts the feature point matching process of the vision camera: the feature points of the currently acquired image are extracted by the SIFT algorithm and matched with the preset standard template. When the matching degree is less than 85%, it is confirmed as the scene of "the photovoltaic panel components in the same row are not neatly arranged".

[0043] 2. Skipping the execution phase: The control unit responds within 50ms and executes the following control sequence: First, generate the first control instruction set: The power source 116 of the control suspension mechanism is activated to fully lift the wheel assembly 115, so that it is detached from the current edge surface of the photovoltaic panel, thus avoiding being stuck by the misaligned edge of the assembly during the crossing process.

[0044] The vacuum adsorption module is activated simultaneously, and each adsorption chamber establishes a negative pressure of no less than -60 kPa within 200 ms, so that the robot body is firmly adsorbed on the current photovoltaic panel surface, providing a stable foundation for subsequent crossings.

[0045] Next, the control unit integrates real-time tilt data from the nine-axis IMU and position data updated by the LiDAR scan, and calculates an optimal diagonal travel path from the edge of the current component to the adjacent component using the A* algorithm. This path planning ensures that the deviation between the actual travel trajectory and the theoretical path is no more than 2 cm.

[0046] Based on the planned path, the control unit sends differential control commands to the left and right drive motors of the moving mechanism. By precisely adjusting the rotation speed of the two tracks (left track structure and right track structure) (steering angle adjustment accuracy is 0.5° / step), the robot is driven to move stably along the diagonal path and successfully cross the component gap with a width of ≤30cm.

[0047] 3. Recovery phase: Once both the LiDAR and the binocular vision camera confirm that the robot has completely reached the surface of the adjacent component and its pose is stable, the control unit generates a third set of control instructions: The power source 116 lowers the wheel assembly 115 so that it re-contacts the edge surface of the photovoltaic panel.

[0048] Shut down the adsorption system 107 to release the negative pressure.

[0049] Restore the robot's lateral cleaning path and continue performing routine cleaning operations.

[0050] The entire process from identification to recovery has a total response time of less than 100ms, ensuring the continuity and high efficiency of the cleaning operation.

[0051] Example 2: Crossing control in scenarios with height differences: When a photovoltaic panel module with a height difference appears in front of the robot, the control process is as follows: 1. Scene recognition stage: The lidar scans a 5m range in front of it. When it detects that the height difference between the component in front and the current component is within the range of 1-10cm (which can cross the threshold), it generates a second identification signal.

[0052] A binocular vision camera simultaneously acquires 3D images of the modules in front, accurately calculates the height difference using a stereo matching algorithm, and analyzes the continuity of the module edges. When the height difference is ≤10cm and the edge transition is continuous without abrupt changes, it is confirmed as a scenario where "there is a height difference between photovoltaic modules in the same row".

[0053] 2. Skipping the execution phase: The control unit responds within 30ms and executes the following control sequence: A second set of control commands is generated to control the electric drive mechanism of the front fixed roller brush, causing it to descend and securely adhere to the surface of the adjacent higher component. At this point, the front roller brush and the rear roller brush, which remains in contact, together form a front and rear support point spanning the height difference.

[0054] The drive motor of the front fixed roller brush outputs a specific torque, while the pressure sensor at the roller brush shaft provides real-time feedback of the pressure value (usually controlled within the range of 50-100N), providing precise support for the slight lifting of the front of the vehicle body.

[0055] The control unit synchronously acquires pitch angle and other pose data from the nine-axis IMU feedback, combines it with pressure sensor readings, and generates feedforward torque control commands through fusion calculations to dynamically adjust the support force of the front roller brush. Simultaneously, it controls the output driving force of the left and right drive motors of the movement mechanism. With the coordinated support of the roller brush and the drive of the movement mechanism, the robot smoothly traverses obstacles with a height difference ≤10cm. During this process, if a protrusion on the same horizontal line is ≤2cm, the suspension mechanism simultaneously provides auxiliary support and buffering protection.

[0056] 3. Recovery phase: Once the LiDAR and binocular vision camera confirm that the robot has fully traversed the adjacent higher component surface and its pose is stable (e.g., the pitch angle has recovered to within ±2°), the control unit generates the fourth set of control commands: The electric drive mechanism is controlled to raise the front fixed roller brush back to the normal cleaning height (usually 5-10mm from the board surface).

[0057] Restore the horizontal sweeping path and continue the sweeping operation.

[0058] Throughout the entire height difference crossing process, real-time monitoring by pressure sensors and IMU ensured that the support force and vehicle lifting angle remained within safe thresholds, effectively preventing damage to the photovoltaic modules.

[0059] The robot can determine the type of contamination on the photovoltaic panel and adjust the rotation speed of the front roller brush 110, rear roller brush 102, and lifting roller brush 103 according to the different types of contamination, as follows: A binocular vision camera acquires real-time images of the photovoltaic panel surface, and a deep learning image recognition algorithm is used to classify the types of contamination, including: 1. Dust: The image appears as a uniform gray tone with fine texture; 2. Dry snow: Overall bright white with blurred edges; 3. Wet snow: White but uneven, may appear in patches or streaks; 4. Ice formation: The surface has a high gloss reflection and the texture is hard and smooth.

[0060] Based on the identified type of contamination, the following cleaning strategy is adopted: 1. Regarding dust pollution: Set the roller brush speed to low speed mode, such as 30-40 RPM; use single-pass cleaning mode; maintain normal travel speed, such as 4-6 meters / minute.

[0061] 2. Regarding dry snow pollution: Set the roller brush speed to medium speed mode, such as 50-60 RPM; use single-pass cleaning mode; appropriately reduce the travel speed, such as 3-5 meters / minute.

[0062] 3. Regarding wet snow pollution: Set the roller brush speed to high speed mode, such as 70-80 RPM; adopt a multiple reciprocating cleaning mode, such as 2-3 times; significantly reduce the travel speed, such as 2-3 meters / minute.

[0063] 4. Regarding icing pollution: Set the roller brush speed to ultra-high speed mode, such as 90-100 RPM, and use multiple slow reciprocating cleaning modes, such as 3-4 times; extremely slow travel speed, such as 1-2 meters / minute.

[0064] If necessary, activate the auxiliary de-icing device. For example, a hot air mechanism can be installed on the main body 101 and the lifting roller brush 103 to help blow out hot air and melt the ice.

[0065] The transfer vehicle is an intelligent mobile platform that enables robots to autonomously cross rows of photovoltaic panels. It is key to improving the overall operation and maintenance efficiency of photovoltaic power plants. When a robot determines that it cannot independently cross between photovoltaic panels, the transfer vehicle is used to transport the robot across. The transfer vehicle includes the following structural components: Mobile chassis 201: It has omnidirectional autonomous movement capability. The bottom of the chassis is symmetrically arranged with track drive mechanisms 204, which enables it to easily cope with the common gravel roads, soft soil and sloping terrain in photovoltaic power stations. Its passability is far superior to that of wheeled chassis.

[0066] RTK Positioning System 206: Integrates a high-precision GNSS RTK positioning module, enabling real-time centimeter-level accurate positioning by receiving signals from multiple satellite constellations and accessing RTK reference station networks within or near the power station. The positioning data from this system forms the basis for global path planning, tracking, and precise docking.

[0067] Lifting platform 202: The scissor lift structure 205 is made of steel and is driven by a high-thrust hydraulic cylinder or a high-precision electric screw, which has excellent stability and load-bearing capacity.

[0068] Support frame 207: The dimensions are usually designed to be 1200mm×900mm, and the surface is covered with anti-slip material to ensure the safe parking of the robot body.

[0069] Lifting range: The lifting stroke range is 400-1600mm, which can cover the height difference between different array rows in most photovoltaic power plants.

[0070] Tilt adjustment: One of its core functions is its tilt adjustment capability. By using hydraulic rods 208 or electric push rod mechanisms, the angle between the support frame 207 and the top of the scissor lift structure 205 can be changed, thereby enabling its platform plane to align with the surface of photovoltaic panels with different installation tilt angles, forming a smooth and seamless transition ramp, which greatly facilitates the robot's entry and exit.

[0071] Perception and Obstacle Avoidance System: Multi-line LiDAR 203 and wide-angle vision cameras are installed at key locations at the front, rear, and sides of the transport vehicle. The multi-line LiDAR 203 scans at a frequency of 10-20Hz, constructing a real-time 3D point cloud map of the surrounding environment. Through built-in obstacle recognition and classification algorithms, it can effectively detect and avoid photovoltaic supports, cable troughs, maintenance passages, and other temporary obstacles. The vision camera assists in semantic information extraction and the identification of specific targets.

[0072] Task completion and status reporting: After the cleaning task of the current row is completed, the robot autonomously navigates to the preset docking point at the end of the row and sends a "job completed, requesting transfer" signal to the transfer vehicle and central control system via wireless communication link, along with its own positioning information.

[0073] Intelligent Response and Precise Delivery: Upon receiving instructions, the transport vehicle plans the optimal safe path based on RTK global positioning and a pre-stored high-precision electronic map, and autonomously drives to the robot's location. During the approach, it integrates multi-line LiDAR point cloud data and visual data for local path planning and dynamic obstacle avoidance. Finally, through "visual servo" control, it achieves millimeter-level precise alignment with the end of the photovoltaic panel. Subsequently, it automatically adjusts the height and tilt angle of the lifting platform 202 to perfectly connect with the photovoltaic panel surface, forming a safe ramp. The robot then smoothly moves onto the transport vehicle platform.

[0074] Cross-row transfer: The transfer vehicle carries the robot and, relying on RTK high-precision tracking and real-time obstacle avoidance, safely and efficiently travels to the next target work row.

[0075] For any parts not mentioned in this application, existing technologies may be used or referenced.

[0076] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0077] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A multi-functional photovoltaic cleaning robot for cross-row operations, comprising a robot body, characterized in that, The robot body is equipped with: A sensing system, installed on the robot body, is used to identify the height difference of the photovoltaic panel components and determine the type of pollution; A mobile mechanism, installed on both sides of the robot body, is used to drive the robot body to move on the surface of the photovoltaic panel; The suspension module, installed on one side of the robot body, is used to suspend the robot from the edge of the photovoltaic panel, providing support for the stable operation of the robot body; An adsorption system, located at the bottom of the robot body, is used to provide stable support when crossing abnormal scenarios; The cleaning system is installed on the side of the robot body opposite to the suspension module, and at both the front and rear ends of the robot body; The control unit is located inside the robot body and is electrically connected to the sensing system, the moving mechanism, the suspension module, the adsorption system, and the cleaning system. The robot body is also equipped with a nine-axis inertial measurement unit for real-time monitoring of the robot body's tilt angle and acceleration.

2. The multifunctional photovoltaic cleaning robot for cross-row operations according to claim 1, characterized in that, The sensing system includes: LiDAR is used to detect the layout of photovoltaic panels, and to detect the edge position deviation and height difference of photovoltaic panels; A binocular vision camera is used to acquire images of the components in front and to determine the type of contamination and verify the recognition results of the lidar through an image matching algorithm; An ultrasonic sensor is used for detection and collision avoidance, assisting the lidar and the binocular vision camera in verifying the results.

3. The multifunctional photovoltaic cleaning robot for cross-row operations according to claim 1, characterized in that, The cleaning system includes: The front roller brush is connected to the robot body via an electric drive mechanism and is used to control the lifting and lowering of the front roller brush to achieve cleaning and obstacle crossing operations. The rear roller brush is connected to the robot body via an electric drive mechanism and is used to control the rear roller brush to lift and lower to achieve sweeping and obstacle crossing operations. Pressure sensors are installed on both the front roller brush and the rear roller brush; The lifting roller brush achieves longitudinal lifting and lowering along the tilt direction of the photovoltaic panel through an electric winch, cable, and guide pulley assembly. The electric winch has a self-locking function to ensure that the position is maintained during the lifting and lowering process. It is driven to rotate by a drive motor, which is mounted on the mounting frame of the lifting roller brush. The electric winch is mounted on the robot body, and the cable is wound around the electric winch, with its free end fixed to the mounting frame of the lifting roller brush.

4. The multifunctional photovoltaic cleaning robot for cross-row operations according to claim 3, characterized in that, The electric drive mechanism is mounted on the robot body and includes two electric push rods, which are used to control the movement of the front roller brush and the rear roller brush, respectively. One end of each electric push rod is hinged to the robot body, and the other end is hinged to the front roller brush and the rear roller brush, respectively.

5. A multifunctional photovoltaic cleaning robot for cross-row operations according to claim 1, characterized in that, The moving mechanism is a tracked moving mechanism, including a left drive motor, a right drive motor, a left track structure, and a right track structure, which are driven independently. The adsorption system includes a negative pressure fan and multiple adsorption chambers.

6. The multifunctional photovoltaic cleaning robot for cross-row operations according to claim 1, characterized in that, The suspension module includes: The system includes a suspension arm and a wheel assembly, wherein the suspension arm is fixed to the robot body and the wheel assembly is hinged to the suspension arm, and can be raised or lowered under the command of the control unit. It also includes a power source that is poweredly connected to the wheel assembly.

7. A multifunctional photovoltaic cleaning system for cross-row operations, comprising a cleaning robot and a transfer vehicle as described in any one of claims 1-6, characterized in that, The transfer vehicle includes: A mobile chassis, wherein a track drive mechanism is provided at the bottom of the mobile chassis for omnidirectional autonomous movement; A lifting platform is mounted on a mobile chassis. The lifting platform includes a scissor lift structure and a support frame, which is used to support the robot body. The RTK positioning system, integrated into a mobile chassis, is used to achieve centimeter-level precise positioning. The tilt adjustment mechanism is connected between the scissor lift structure and the support frame, and is used to adjust the tilt angle of the support frame to match the tilt angle of the photovoltaic panel; The perception and obstacle avoidance system, mounted on a mobile chassis, includes a multi-line LiDAR and a wide-angle vision camera, used to build environmental maps and avoid obstacles in real time; The tilt adjustment mechanism includes a hydraulic rod for changing the angle between the support frame and the top of the scissor lift structure.

8. A control method for a multi-functional cross-row photovoltaic cleaning machine system, the control method being based on the cleaning robot described in any one of claims 1 to 6, characterized in that, Includes the following steps: Perception step: Acquire environmental data collected by the robot's perception system; Decision-making steps: Based on the aforementioned environmental data, identify any abnormal scenarios that need to be crossed between photovoltaic panel modules; Control steps: In response to the detection of the abnormal scene, generate and issue a corresponding set of control instructions to adjust the posture of the robot's actuator, plan the travel path and control the mobile mechanism to perform traversing actions; Recovery steps: After confirming the completion of the crossing based on the aforementioned environmental data, generate and issue instructions to resume the robot's regular cleaning operation.

9. The control method according to claim 8, characterized in that, The abnormal scenario is that the photovoltaic panel modules in the same row are not arranged neatly; The decision-making steps include: generating a first recognition signal when the deviation of the component edge position exceeds a first threshold based on the LiDAR scanning data, and confirming the scenario after verification by a feature point matching algorithm based on visual image data; The control steps include: generating a first set of control commands to control the suspension mechanism to lift the wheel assembly and start the multi-chamber vacuum adsorption module; subsequently, planning an oblique travel path based on the fusion data of the inertial measurement unit and the lidar, and performing the crossing by controlling the left drive motor and the right drive motor through differential speed control.

10. The control method according to claim 8, characterized in that, The abnormal scenario is that there is a height difference between photovoltaic panel modules in the same row; The decision-making steps include: generating a second recognition signal when the height difference of the components in front is detected to be within a threshold range based on the LiDAR scanning data, and confirming the scene after verification by a stereo matching algorithm based on the binocular vision three-dimensional image data; The control steps include: generating a second set of control commands to control the front fixed roller brush to descend and press against the surface of the adjacent high component; subsequently, based on feedback data from the pressure sensor and the inertial measurement unit, controlling the front roller brush motor to output support torque and coordinate with the moving mechanism to perform a crossing.

Images