A cleaning robot and a method of positioning a cleaning robot

Abstract

The application discloses a cleaning robot and a cleaning robot positioning method, wherein the cleaning robot comprises a body, a driving mechanism, a driving wheel, a windproof hook, a sensor, a memory and a processor; the sensor is arranged on the windproof hook and faces a photovoltaic module; the sensor is used for sensing a frame of the photovoltaic module; the processor comprises a real-time positioning module which is configured to determine a real-time position of the cleaning robot in a photovoltaic array based on frame sensing signal counts collected by the sensor and real-time rotation information of the driving wheel; the processor further comprises a path planning module which is configured to plan a moving path of the cleaning robot according to the real-time position and a current running state of the cleaning robot. Through the application, accurate position positioning of the cleaning robot can be realized, and the reliability and safety of parking in a strong wind environment are significantly improved.

Description

Technical Field

[0001] This application relates to the field of photovoltaic cleaning technology, and more specifically, to a cleaning robot and a cleaning robot positioning method. Background Technology

[0002] In photovoltaic (PV) power generation systems, PV modules, as the core power generation units, are typically arranged in arrays and electrically connected between adjacent strings via flexible cable trays. To prevent dust accumulation on the module surfaces from affecting power generation efficiency, PV cleaning robots are commonly used for automated cleaning operations. However, if the robot encounters sudden severe weather such as strong winds while performing cleaning or charging tasks, it must immediately stop operation and shut down to prevent equipment damage. Because the robot's emergency stop location is uncertain, if it happens to stop on a flexible cable tray with limited structural strength, there is still a high risk of overturning or falling under the continued influence of strong winds, threatening equipment safety and the stable operation of the power station. Summary of the Invention

[0003] To address the aforementioned technical problems, this application discloses a cleaning robot and a method for positioning the cleaning robot. This application enables precise positioning of the cleaning robot, significantly improving its reliability and safety during unforeseen circumstances. Specifically, the technical solution of this application is as follows: In a first aspect, this application discloses a cleaning robot, comprising: ontology; The drive mechanism is built into the main body; A drive wheel is connected to the drive mechanism, which drives the drive wheel to rotate. The windproof hook includes a first connecting part and a second connecting part. The first connecting part is connected to one side of the body, and the second connecting part is located at the end of the first connecting part away from the body. The second connecting part is used to extend to the bottom of the photovoltaic module. A sensor is disposed on the windproof hook and faces the photovoltaic module; the sensor is used to sense the frame of the photovoltaic module. Memory and processor; The processor includes a real-time positioning module, which is configured to determine the real-time position of the cleaning robot in the photovoltaic array based on the count of the frame sensing signals collected by the sensor and the real-time rotation information of the drive wheel. The processor also includes a path planning module, configured to plan the movement path of the cleaning robot based on the real-time location and the current operating status of the cleaning robot.

[0004] In some embodiments, the sensor is disposed at the second connection portion, the sensor is located below the photovoltaic module, or the sensor is disposed at the first connection portion, the sensor is located on one side of the photovoltaic module; The real-time positioning module includes: A sensor counter is used to trigger a counting signal and record the current cumulative number of sensor counts whenever the sensor detects the frame of the photovoltaic module during the movement of the cleaning robot. A rotary wheel rangefinder is used to calculate the real-time walking distance of the cleaning robot based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated. A position corrector is used to logically compare the real-time walking distance with the standard distance between the center lines of adjacent photovoltaic modules, and determine and correct the real-time position of the cleaning robot based on the comparison result.

[0005] In some implementations, the position corrector is configured to perform the following logical steps: The judgment is based on the ratio of the real-time walking distance to the standard distance: If the ratio is within the first threshold range, the positioning status is determined to be normal, and the real-time position is calculated using the current sensor count, the real-time walking distance, and the standard distance. If the ratio is within the second threshold range, a sensor miss is determined, and the current sensor count is incremented by one. The corrected sensor count and the corrected real-time walking distance are calculated, and the real-time position is calculated using the corrected sensor count, the corrected real-time walking distance, and the standard distance. If the ratio exceeds the upper limit of the second threshold range, the drive wheel is determined to be slipping and spinning freely. In this case, the real-time position remains unchanged, and an abnormal status warning is triggered. The upper limit of the first threshold range is less than or equal to the lower limit of the second threshold range.

[0006] In some embodiments, the inductive counter is further configured to: The counting signals triggered within the first time interval are merged using a software filtering algorithm and identified as single valid counts.

[0007] In some implementations, the path planning module includes: A docking detector is used to acquire real-time environmental status data and determine whether the cleaning robot needs to dock based on the environmental status data. The strategy planner is used to determine the current operating status of the cleaning robot by analyzing the changing trend of the real-time location, and to plan a docking strategy by combining the current operating status with the real-time location. The drive controller, according to the planned docking strategy, controls the cleaning robot to move to the docking station and complete the docking.

[0008] In some implementations, the strategy planner is configured to perform the following logical steps: Based on the temporal analysis and change pattern recognition of the real-time location, the current operating state of the cleaning robot is analyzed, and the current operating state includes at least one of yaw, drive wheel slippage, and obstacle crossing. Based on a pre-built location coding database, the relative distance between the real-time position of the cleaning robot and each docking location code is calculated; Based on the relative distance and the current operating state, the docking strategy is dynamically generated. The docking strategy includes at least one of the following: the cleaning robot's deviation correction control strategy, the autonomous back-back and escape strategy when a blockage occurs, and the intelligent obstacle-crossing decision-making for path obstacles.

[0009] In some embodiments, there are two windproof hooks, which are respectively connected to both sides of the main body; each windproof hook on both sides has a sensor, and the counting signals of the two sensors are independent of each other; So that when the body of the cleaning robot tilts, the strategy planner can identify the yaw direction of the cleaning robot by comparing the difference between the counting signals on both sides.

[0010] In some implementations, the strategy planner is further configured to perform the following logical steps: Based on the relative distance, feasible paths are prioritized, and when the distances are similar, the path direction that is consistent with the current movement trend is selected first. When the highest priority feasible path becomes infeasible due to obstruction, it adaptively switches to the second-best backup path.

[0011] Secondly, this application also discloses a cleaning robot positioning method, executed by the cleaning robot described in any of the above embodiments, comprising the following steps: Based on the frame sensing signal count collected by the sensor and the real-time rotation information of the drive wheel, the real-time position of the cleaning robot in the photovoltaic array is determined. Based on the real-time location and the current operating status of the cleaning robot, the movement path of the cleaning robot is planned.

[0012] In some embodiments, determining the real-time position of the cleaning robot within the photovoltaic array based on the count of edge sensing signals collected by the sensors and the real-time rotation information of the drive wheels includes: During the movement of the cleaning robot, each time the sensor detects the frame of the photovoltaic module, a counting signal is triggered, and the current cumulative number of sensor counts is recorded. The real-time walking distance of the cleaning robot is calculated based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated. The real-time walking distance is logically compared with the standard distance between the center lines of adjacent photovoltaic modules, and the real-time position of the cleaning robot is determined and corrected based on the comparison result.

[0013] Compared with the prior art, this application has at least one of the following beneficial effects: 1. The cleaning robot of this application achieves accurate positioning through the real-time positioning module of the processor, and plans the current path for the cleaning robot through the path planning module. When the cleaning robot is in an abnormal state, it can quickly adjust the path and move to a safe docking station.

[0014] 2. The cleaning robot of this application achieves precise positioning through a dual positioning algorithm that fuses sensor data with data on the number of wheel revolutions. This technology provides accurate location data for wind protection, enabling the robot to quickly locate and move to a safe docking station. Simultaneously, during docking, it can diagnose yaw, slippage, and other conditions in real time, dynamically planning a safe path that includes correction and obstacle avoidance strategies, significantly improving the reliability and safety of docking in strong winds.

[0015] 3. This application employs a low-cost sensor combined with a simple software filtering and verification algorithm, replacing the expensive LiDAR and vision system. By logically comparing sensor pulse counts with wheel travel distance, and using a threshold-based state discrimination mechanism, it significantly reduces hardware costs and algorithm complexity while ensuring positioning accuracy. This also improves the system's robustness under low-light conditions such as nighttime, demonstrating significant practical and economic advantages. Attached Figure Description

[0016] The preferred embodiments will now be described in a clear and easy-to-understand manner, in conjunction with the accompanying drawings, to further explain the above-mentioned characteristics, technical features, advantages, and implementation methods of this application.

[0017] Figure 1 This is a schematic structural block diagram of one embodiment of a cleaning robot according to this application; Figure 2This is a schematic diagram showing the location of the sensor installed on one side of the cleaning robot body in an embodiment of the present application; Figure 3 This is a schematic diagram showing the positions of sensors installed on both sides of the cleaning robot body in an embodiment of this application; Figure 4 This is a schematic diagram of the processor structure in an embodiment of a cleaning robot according to this application; Figure 5 This is a schematic diagram of a photovoltaic module structure in one embodiment of this application; Figure 6 This is a flowchart illustrating the steps of one embodiment of a positioning method for a cleaning robot according to this application. Detailed Implementation

[0018] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0019] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or sets.

[0020] To keep the drawings concise, each figure only schematically shows the parts relevant to the invention, and these do not represent the actual structure of the product. Furthermore, to facilitate understanding, in some figures, only one of components with the same structure or function is schematically depicted, or only one is labeled. In this document, "one" not only means "only one," but can also mean "more than one."

[0021] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0022] In this document, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0023] Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the specific implementation methods of this application will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings and other implementation methods can be obtained based on these drawings without creative effort.

[0025] Currently, the positioning technology for photovoltaic cleaning robots faces numerous challenges, including poor environmental adaptability and high costs. In existing technologies, to minimize interference with the photovoltaic power generation process, the operation of photovoltaic cleaning robots is typically scheduled for nighttime. However, the mainstream positioning scheme relying on the fusion of LiDAR and visual cameras performs poorly under these conditions. While LiDAR can provide accurate point cloud data, its ability to identify large, flat surfaces of photovoltaic modules without significant features is limited. Visual cameras, on the other hand, heavily rely on ambient light; their imaging quality deteriorates significantly under low-light conditions at night, making feature extraction difficult and leading to a substantial decrease in positioning accuracy or even failure. Furthermore, the high-precision LiDAR and industrial cameras used in this solution are expensive, resulting in a high hardware investment for the entire system and hindering its widespread application in large-scale photovoltaic power plants.

[0026] On the other hand, some lower-cost systems employ odometry positioning based on drive wheel encoders, which calculates the walking distance and position by accumulating the number of wheel revolutions. This method can maintain basic positioning under ideal conditions of flat, dry conditions and stable robot load, but in actual operation, especially in special weather conditions such as strong winds, rain, and snow, its inherent technical defects become fully apparent under complex conditions.

[0027] For example, in windy weather, strong winds may directly push the robot body to a different position, causing the actual displacement to differ from the wheel rotation. Slippery or dusty roads can easily cause the drive wheels to slip or spin freely, resulting in serious distortion of mileage information. These factors can all cause cumulative errors in pure odometer positioning, which cannot be detected and corrected by the robot itself. Ultimately, this leads to drift in the robot's position information, which not only affects the integrity of cleaning coverage but also makes it difficult to reliably navigate to a safe area in situations requiring emergency stopping, such as in strong winds, due to the inability to know the accurate location. This poses a risk of the equipment falling or being damaged.

[0028] To address the aforementioned technical deficiencies, this application provides a cleaning robot and a cleaning robot positioning method. Based on dual-verification positioning using low-cost sensors and drive wheel mileage data fusion, and combined with real-time operation status monitoring and dynamic path planning, the photovoltaic cleaning robot achieves highly reliable and accurate positioning and autonomous and safe docking in complex environments such as nighttime and strong winds.

[0029] This application provides an embodiment of a cleaning robot, with reference to the attached specification. Figure 1 As shown, the cleaning robot includes: a body, a drive mechanism, drive wheels, a windproof hook, sensors, a memory, and a processor.

[0030] The drive mechanism is built into the main body. The drive wheel is connected to the drive mechanism. The drive mechanism is used to drive the drive wheel to rotate.

[0031] The windproof hook includes a first connecting part and a second connecting part. The first connecting part is connected to one side of the main body, and the second connecting part is located at the end of the first connecting part away from the main body. The second connecting part is used to extend to the bottom of the photovoltaic module.

[0032] Specifically, the windproof hook is designed in a hook shape, with its first and second connecting parts arranged perpendicularly to each other. The first connecting part is used to securely connect to one side of the robot body, serving a fixing function. The second connecting part extends from the end of the first connecting part and is designed to reach under the photovoltaic module, thereby achieving reliable hooking and restraint of the photovoltaic module's edge. The windproof hook is mainly used to enhance the cleaning robot's tensile and displacement resistance on the photovoltaic module, preventing it from falling off or shifting due to airflow or vibration during robot movement or cleaning.

[0033] A sensor, mounted on a windproof hook and facing the photovoltaic module, is used to sense the frame of the photovoltaic module.

[0034] For details, please refer to the attached instruction manual. Figure 2As shown, the sensor is used to sense the position information of the photovoltaic module frame. During operation, the robot initiates a fusion localization algorithm to determine its exact position within the photovoltaic array. The fusion localization algorithm in this application relies on a combination of software and hardware. Optionally, the sensor is an M30 inductive sensor.

[0035] In some embodiments, the cleaning robot includes two windproof hooks, each connected to one side of the main body. Each windproof hook on both sides has a sensor, and the counting signals from the two sensors are independent of each other. A sensor is also mounted on each of the windproof hooks on the lower sides of the robot, as shown in the attached instruction manual. Figure 3 As shown. The sensor is wired to the electrical control box and connected to the robot's main control board, transmitting signals to the main control board. Both the sensor and the electrical control box selected in this application are dustproof and waterproof (IP65).

[0036] The sensor is located at the second connection part, below the photovoltaic module, or at the first connection part, on one side of the photovoltaic module; the processor includes a real-time positioning module, configured to determine the real-time position of the cleaning robot in the photovoltaic array based on the count of the frame sensing signals collected by the sensor and the real-time rotation information of the drive wheel.

[0037] Specifically, as the robot walks on the modules, sensors detect and pass over the metal frame of each photovoltaic module, generating a pulse signal each time. Simultaneously, the encoder on the drive motor continuously records the real-time number of rotations of the drive wheels. The system integrates a positioning algorithm, comparing and coupling the number of sensor triggers with the actual walking distance calculated from the wheel rotation count and wheel diameter. Software calculations and judgments are then used to filter out any potential false signals.

[0038] Through the fusion and cross-verification of this dual information source, the system can calculate a relatively reliable position under various operating conditions, including normal operation, sensor signal loss, and drive wheel slippage, thereby achieving accurate positioning.

[0039] The processor also includes a path planning module, configured to plan the movement path of the cleaning robot based on the real-time location and the current operating status of the cleaning robot.

[0040] Specifically, after obtaining the real-time position code, the system further analyzes the trend of this code's changes to determine the robot's current operating status. For example, by comparing whether the counts of the left and right sensors are synchronized, it can be determined whether the robot has yawed. By observing whether the increase in the position code matches the wheel rotation, it can be determined whether an obstacle has been encountered, causing a stall. Combining these status judgments with the accurate real-time position code, the system begins to plan specific docking strategies.

[0041] The cleaning robot, based on a planned docking strategy, sends specific movement commands to its motion control module via the control system. During control, it continuously monitors positioning information to ensure the robot moves along the predetermined path towards the target docking station. Once the robot has reached the location code corresponding to the target docking station, confirmed by both sensor counts and wheel lap counts, the control system commands the robot to decelerate and eventually stop, completing the docking.

[0042] Based on the above embodiments, this application provides another embodiment of a cleaning robot, as detailed in the appendix to the specification. Figure 4 As shown, the real-time positioning module includes: The sensor counter is used to trigger a counting signal and record the current cumulative number of sensor counts whenever the sensor detects the frame of the photovoltaic module during the movement of the cleaning robot.

[0043] Specifically, as the robot walks along the long side of the photovoltaic array module, whenever the sensor effectively detects the edge of the metal frame of the photovoltaic module, the internal circuit state changes abruptly, generating an electrical pulse signal. This signal is transmitted to the robot's main control board in real time, and the counter module in the main control board responds to this pulse by incrementing the cumulative sensing count of the corresponding sensor channel by 1.

[0044] The rotary wheel rangefinder is used to calculate the real-time walking distance of the cleaning robot based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated.

[0045] Specifically, the wheel rangefinder calculates the robot's actual displacement by driving the physical movement of the wheels. The real-time walking distance is calculated as the distance the robot travels on a single photovoltaic module; therefore, the real-time walking distance is reset to zero after each counting signal trigger and recalculated. Based on the sensor counter and wheel rangefinder, the accumulated sensor counts record the total number of photovoltaic modules the robot has traversed, while the real-time walking distance calculates the distance the robot has traveled on the current photovoltaic module. Using the standard distance L between the centerlines of adjacent photovoltaic modules and the accumulated sensor counts, the total length of photovoltaic modules traversed by the robot can be calculated. Adding this to the distance traveled on the current photovoltaic module gives the robot's total walking distance.

[0046] Optionally, a high-precision rotary encoder is installed on the output shaft of the drive motor. This encoder records the number of rotations of the motor output shaft or the drive wheel itself in real time, or accurately records the time taken for the drive wheel to rotate using a timer. Combined with the known outer diameter D of the drive wheel (in millimeters), the system uses circular kinematics formulas for calculation. If the number of rotations n (in revolutions) is known, then the real-time travel distance Y0 (in millimeters) is: Y0 = n * π * D.

[0047] In other implementations, the travel distance can also be obtained if the rotation time t (in seconds) is known and converted by the motor speed.

[0048] A position corrector is used to logically compare the real-time walking distance with the standard distance between the center lines of adjacent photovoltaic modules, and determine and correct the real-time position of the cleaning robot based on the comparison result.

[0049] Specifically, the system executes the core fusion positioning and data correction logic. The algorithm first obtains the standard distance L between the center lines of adjacent photovoltaic modules in the photovoltaic array, which is a preset known constant. The real-time walking distance Y0 calculated in step S220 is logically correlated and compared with the number of sensor counts i.

[0050] It should be noted that the standard distance L mentioned above is the sum of the width of a single photovoltaic module and the gap distance between adjacent photovoltaic modules.

[0051] Based on the above embodiments, this application provides another embodiment of a cleaning robot, wherein the induction counter is further configured to: perform merging processing on the counting signals triggered within a first time interval using a software filtering algorithm, and identify them as single valid counts.

[0052] Specifically, to ensure counting accuracy, the system implements a filtering algorithm at the software level. The software filtering algorithm merges the counting signals triggered within the first time interval, identifying them as single valid counts. Because the gap between the frames of adjacent photovoltaic modules is very small, the sensor may trigger two signals in a very short time when passing quickly. The software detects the time interval between pulses; if the interval is less than a preset first time interval, it determines that this is a jitter signal generated by passing through the gap between adjacent photovoltaic modules and filters it out, thus ensuring that the count increases only once for each complete photovoltaic module passed. Since the gap between adjacent photovoltaic modules is small, and there are photovoltaic module frames on both sides of the gap, when the cleaning robot crosses the gap, the sensor senses the frames on both sides and triggers two signals. The time interval between the two signals is approximately equal to the time it takes for the sensor to move from one frame to the other; that is, the width of the gap between the two photovoltaic modules determines the size of this time interval. In this case, the time interval between the two signals can be determined by measuring the gap width and the cleaning robot's moving speed; this time interval is approximately equal to the first time interval. By monitoring the time difference between two adjacent signals, it can be determined whether the cleaning robot has passed through the gap between two adjacent photovoltaic modules, thereby filtering out redundant signals. In some embodiments, the first time interval ranges from 0.1 seconds to 1 second, for example, the first time interval is 0.1 seconds, 0.2 seconds, 0.3 seconds, 0.4 seconds, 0.5 seconds, 0.6 seconds, 0.7 seconds, 0.8 seconds, 0.9 seconds, or 1 second.

[0053] Based on the above embodiments, this application provides another embodiment of a cleaning robot, wherein the position corrector is configured to perform the following logical steps: The judgment is based on the ratio of the real-time walking distance to the standard distance: If the ratio is within the first threshold range, the positioning status is determined to be normal, and the real-time position is calculated based on the current number of sensor counts, the real-time walking distance, and the standard distance.

[0054] If the ratio is within the second threshold range, it is determined that a sensor missed detection has occurred, and the current sensor count is incremented by one. The corrected sensor count and the corrected real-time walking distance are calculated. The real-time position is calculated using the corrected sensor count, the corrected real-time walking distance, and the standard distance.

[0055] Among them, the corrected sensor count is equal to the current sensor count plus one, and the corrected real-time walking distance is equal to the current real-time walking distance minus the standard distance.

[0056] If the ratio exceeds the upper limit of the second threshold range, it is determined that the drive wheel is slipping and spinning freely. At this time, the real-time position remains unchanged and an abnormal status warning is triggered.

[0057] Wherein, the upper limit of the first threshold range is less than or equal to the lower limit of the second threshold range.

[0058] In practice, the algorithm continuously calculates the ratio of the walking distance to the standard distance between adjacent components. If this ratio is approximately within a first threshold range, the positioning is considered normal. Under normal circumstances, the width of individual components on the array is consistent, and for every sensor trigger count, the walking distance is the width of the individual component. The robot's real-time position code can be obtained by accumulating the number of sensor triggers.

[0059] If the ratio is significantly within the second threshold range, it may be determined that the sensor signal is lost. The system will automatically perform data compensation, adding 1 to the current count to correct the current count, i.e., adding 2 to the cumulative number of accurate border detections relative to the previous count. Optionally, the minimum value of the second threshold range is greater than or equal to the maximum value of the first threshold range.

[0060] For example, the first threshold range is 0.5-1.5. The system determines whether the walking distance (time) / distance to adjacent components (time) is greater than 0.5 and less than or equal to 1.5. If so, the system considers the walking distance (time) / distance to adjacent components (time) to be approximately equal to 1, and records this code in the background.

[0061] In some implementations, the second threshold ranges from 1.5 to 2.5. The system determines that the walking distance (time) / distance to adjacent components (time) is greater than 1.5 and less than or equal to 2.5. If so, it is assumed that the walking distance (time) / distance to adjacent components (time) is approximately equal to 2, indicating that the robot has traversed two photovoltaic modules, and the position sensor signal for the first photovoltaic module has been lost. This is reported to the backend with an alert, the sensor count is incremented by 1, and the backend data is corrected and a larger code number is recorded.

[0062] In other implementations, if the ratio is abnormally high, for example, greater than the upper limit of the second threshold range, it is determined that the drive wheel is slipping and spinning freely. At this time, the positioning accuracy decreases, and the system can only roughly determine that the robot is located on one of the three consecutive components.

[0063] For example, the upper limit of the second threshold range is 2.5. If the distance traveled (time) / distance to adjacent components (time) is greater than 2.5, it is determined that the robot's wheels are slipping on that component and spinning idly. Feedback is sent to the backend and an alert is issued, and the backend records the code number. The backend uses the code number to locate which component the robot is on and issues an alert, requiring timely troubleshooting.

[0064] It should be noted that since the real-time walking distance is cleared after each counting signal is triggered and the real-time walking distance is recalculated, under normal circumstances, the maximum value of the real-time walking distance is the distance traveled during the time from the previous clearing to this clearing. Therefore, an additional highest value before being cleared needs to be stored before the real-time walking distance is cleared, and the judgment of which threshold range is made based on this highest value.

[0065] However, in a special case, the wheels of the sweeping robot slip on the photovoltaic module and rotate idly in place. That is to say, the real-time walking distance will continuously increase over time and cannot move forward to the next photovoltaic module to trigger the counting signal. In this case, once the walking distance (time) / distance between adjacent modules (time) exceeds the upper limit of the second threshold range, it is determined that the wheels of the sweeping robot slip on this module and rotate idly, and there is no need to wait until the next counting signal is triggered to make a judgment.

[0066] In some other embodiments, the switch is in the front-back direction of the robot body, the origin is at the parking position, and the switch anchors the origin. Refer to the attached Figure 5 As shown in the figure, the sensor in the left-right direction of the robot body counts 1 after passing a photovoltaic panel, and the number is +1 or -1. The number of the parking position is 0, and the number of the return position is n. During the process of moving from the parking position to the return position, the number is +1. During the process of returning from the return position to the parking position, the number is -1.

[0067] Refer to Figure 5 As shown in the figure, the real-time position of the robot on the module: Y0 = n * π * D (mm).

[0068] If the sensor is triggered normally and the wheels of the robot do not rotate idly, for example, 0 < Y0 ≤ L, then the real-time position of the robot on the array: i a+1 = i a +1, , the initial position: i0 = 0, Y = 0, a = 0.

[0069] If the sensor is triggered abnormally and the wheels of the robot do not rotate idly, for example, L < Y0 ≤ 2L, then the real-time position of the robot on the array: i a+1 = i a +2, .

[0070] If the wheels of the robot rotate idly, for example, Y0 > 3L, then the real-time position of the robot on the array: a+1 = i a , In this case, the system cannot perform precise positioning and can only infer based on the last valid code i a that the robot is still within the range of at most three modules starting from this code and immediately trigger a stall alarm.

[0071] By employing the algorithm described in this application, which cross-validates and logically compares physical walking distance with sensor event counts, the system can dynamically identify anomalies during the positioning process (such as signal loss or wheel slippage) and adaptively correct the position encoding. This allows the system to maintain a relatively reliable and accurate determination of the robot's real-time position even under various non-ideal working conditions.

[0072] Based on the above embodiments, this application provides another embodiment of a cleaning robot, as detailed in the appendix to the specification. Figure 4 As shown, the path planning module includes: The docking determination unit is used to acquire the current environmental status data in real time and determine whether the cleaning robot needs to dock based on the environmental status data.

[0073] Specifically, during the docking process, the system first acquires environmental status data through environmental sensors. Based on this environmental status data, it determines whether the cleaning robot is in severe weather, such as strong winds, heavy rain, or heavy snow.

[0074] This embodiment uses windy weather as an example for illustration. Optionally, the core environmental status data is wind speed data. Wind speed data is collected through a dedicated wind speed sensor installed in the robot's communication box, photovoltaic module, or the robot itself. This sensor continuously monitors the instantaneous wind speed in the robot's working area and transmits it in real-time as a digital signal to the robot's main control system or backend monitoring platform, such as a Supervisory Control and Data Acquisition (SCADA) system. If the current environmental status data meets the docking criteria, the robot is determined to dock.

[0075] In some optional embodiments, the docking determiner is used to perform the following steps: Acquire current environmental status data, including wind speed data. Determine if the wind speed has reached or exceeded a preset first wind speed threshold. If the wind speed reaches or exceeds the first wind speed threshold, determine if the cleaning robot needs to perform a docking action. Issue a docking command to the cleaning robot.

[0076] In practice, the system internally presets a first wind speed threshold as the critical point to trigger safety protection. Optionally, the first wind speed threshold can range from 15 m / s to 20 m / s, for example, 15 m / s, 16 m / s, 17 m / s, 18 m / s, 19 m / s, or 20 m / s. When the real-time wind speed data continuously exceeds this threshold, it is determined that the current ambient wind force poses a safety risk to the outdoor cleaning robot, potentially causing it to tip over, derail, or damage the photovoltaic modules. Based on this risk assessment, the system logic switches from the normal "cleaning operation mode" to "high wind protection mode." To ensure equipment safety and stability, the cleaning task must be interrupted, and a docking command must be generated immediately.

[0077] The strategy planner is used to determine the current operating status of the cleaning robot by analyzing the changing trend of the real-time location, and to plan a docking strategy by combining the current operating status with the real-time location.

[0078] Specifically, the memory pre-stores the codes for all permitted docking stations along the cleaning path. The core of the robot processor's planning strategy is calculating the path distance between the robot's current position code and each preset docking position code, and selecting the one with the closest real-time distance as the target docking station. If the robot is close to two docking stations, it is controlled to move in the direction of the wind to dock as quickly as possible. This planning process considers both efficiency and safety, ensuring that the robot reaches the docking station via the shortest path.

[0079] The drive controller, according to the planned docking strategy, controls the cleaning robot to move to the docking station and complete the docking.

[0080] Specifically, in some optional embodiments, if the wind speed sensor detects a decrease in wind speed and it falls below a preset second wind speed threshold during the entire movement process, the system can automatically cancel the docking command and control the robot to resume cleaning operations on the spot. Otherwise, it will remain docked until the severe weather conditions are resolved.

[0081] In some alternative embodiments, the method further includes: after the cleaning robot stops at a preset position, continuously acquiring current environmental state data. It determines whether the current environmental state data meets the cleaning conditions, i.e., is lower than a preset second wind speed threshold. If so, it controls the cleaning robot to leave the stopping position and continue performing the cleaning task. In some embodiments, the second wind speed threshold ranges from 11 m / s to 13 m / s, for example, 11 m / s, 12 m / s, or 13 m / s.

[0082] This application enables a cleaning robot to achieve fully automated and highly reliable docking in harsh environments such as strong winds, encompassing risk assessment, autonomous positioning, status diagnosis, path planning, and safe execution. Compared to existing technologies (LiDAR + vision camera), it is more versatile and easier to implement in terms of logic algorithms and software / hardware interfaces. Furthermore, it has lower implementation costs and is easier to promote.

[0083] Based on the above embodiments, this application provides another embodiment of a cleaning robot, wherein the strategy planner is configured to execute the following logical steps: Based on the temporal analysis and change pattern recognition of the real-time location, the current operating state of the cleaning robot is analyzed, and the current operating state includes at least one of yaw, drive wheel slippage, and obstacle crossing.

[0084] Specifically, the robot continuously tracks and records the changes in its position encoding sequence over time, including its rate of increase or decrease, the consistency of the independent counting sequences of the left and right sensors, and the degree of matching between the position encoding update and the expected displacement calculated from the number of rotations of the drive wheels. For example, by continuously monitoring the temporal changes in the real-time position encoding, if the encoding sequence shows discontinuities or reverse jumps, the robot is determined to be yawing. If the encoding change stagnates while the drive wheel speed signal remains constant, the drive wheels are determined to be slipping. If the encoding remains unchanged for an extended period and the drive wheel torque exceeds the normal range, the robot is determined to have encountered an obstacle.

[0085] In some optional implementations, a sensor is mounted on each of the windproof hooks on both sides of the robot's body. The counting signals of the two sensors are independent of each other, so that the strategy planner can identify the robot's yaw direction by comparing the difference in the counting signals on both sides when the robot's body tilts. When the system detects a continuous and increasing difference in the cumulative counts of the left and right sensors within the same time period, it identifies a "yaw" state, indicating that the robot's travel axis has deviated from the reference direction of the photovoltaic array.

[0086] In some alternative embodiments, if the system detects through an algorithm that the theoretical travel distance Y0 calculated based on the high-precision number of rotations of the drive wheel is consistently and significantly greater than the standard component spacing L, and this difference exceeds the preset tolerance range used to determine normal sliding, the system determines that the drive wheel has "slipped" or "idled," which means that the wheel is slipping on the component surface and the physical displacement is much smaller than the theoretical rotation amount.

[0087] In some alternative embodiments, if the position code stops updating during travel, while the drive wheel encoder feedback shows that the motor is still continuously outputting torque in an attempt to rotate, but the calculated actual displacement is minimal or zero, the system combines this contradictory information to infer that a physical "obstacle" has likely been encountered ahead, causing the robot to "block." These state analyses are performed in real time and dynamically, and are applicable to dynamic conditions during docking.

[0088] Based on a pre-built location coding database, the relative distance between the real-time position of the cleaning robot and each docking location coding is calculated.

[0089] Specifically, in some embodiments, the method further includes pre-establishing a location coding database for photovoltaic modules, and marking the location and number of each photovoltaic module and docking station along the cleaning path. The docking location coding can be set at the cleaning robot's docking stations, the cleaning robot's return position, and photovoltaic modules near the support columns. Photovoltaic modules near the support columns are relatively stable on windy days due to the stable support of the columns, and the cleaning robot can select these photovoltaic modules as docking locations on windy days.

[0090] This location coding database records a digital map of the entire photovoltaic cleaning path, clearly marking the location code of each designated stop. When a stop is required, the system immediately reads the current real-time location code and compares it with the location codes of all stops in the database. The optimal target stop is then preliminarily selected.

[0091] Based on the relative distance and the current operating state, the docking strategy is dynamically generated. The docking strategy includes at least one of the following: the cleaning robot's deviation correction control strategy, the autonomous back-back and escape strategy when a blockage occurs, and the intelligent obstacle-crossing decision-making for path obstacles.

[0092] Specifically, the dynamically generated comprehensive docking strategy is not simply a "move in a straight line from point A to point B" instruction, but rather an intelligent control sequence that integrates state-based responses. The strategy generator combines relative distance information and the current operating state. For example, if the robot is yawed, a closed-loop correction control instruction based on the difference in counts from the left and right sensors is generated. The robot then plans its movement path to the target docking station after returning to the centerline of the path. If the robot is slipping, a stall disengagement segment containing intermittent backtracking and forward movement instructions is inserted into the movement path. If an obstacle is detected, a corresponding obstacle-crossing action instruction is generated based on the obstacle's orientation assessment. Finally, all action instructions are integrated temporally and spatially based on the absolute positioning information provided by real-time position encoding to form a dynamic docking path adaptable to real-time operating conditions.

[0093] In some optional embodiments, if the state analysis indicates that the robot is "yawing," a "yawing correction control strategy" will be embedded in the generated strategy. Optionally, during the journey to the target docking station, the system will continuously compare the counts of the left and right sensors and dynamically adjust the speed difference between the left and right drive wheels through a proportional control algorithm, so that the robot gradually returns to the correct path parallel to the component frame.

[0094] In some alternative embodiments, if a "stuck" or persistent "slippage" is diagnosed, the strategy will include an "autonomous retreat and escape strategy." Optionally, the robot is controlled to first stop moving forward, and then commanded to move in the opposite direction a preset short distance, which may be calculated based on the number of wheel laps, to attempt to escape the obstacle or slippage area, before attempting to move forward again or assessing whether to change the path.

[0095] In some alternative embodiments, if the system infers the presence of an "obstacle," it triggers an "intelligent obstacle-crossing decision." Optionally, it issues a command to the cleaning robot to raise the cleaning brush or adjust the chassis posture, or controls the robot to perform a small obstacle-crossing maneuver. This ensures that the robot can efficiently and safely approach the nearest docking station while dealing with complex working conditions.

[0096] Based on the above embodiments, this application provides another embodiment of a cleaning robot, wherein the strategy planner is further configured to perform the following logical steps: Based on the relative distance, feasible paths are prioritized, and in the case of similar distances, the path direction that is consistent with the current movement trend is selected first.

[0097] When the highest priority feasible path becomes infeasible due to obstruction, it adaptively switches to the second-best backup path.

[0098] Specifically, the system possesses the ability to adaptively switch paths in dynamic environments to handle sudden obstacles. Once the path generated based on the highest priority is initiated, the system continuously monitors the robot's operational status for real-time feasibility verification. If the system continuously detects a "stuck" state or a fixed obstacle while the robot is traveling along this path, it determines that the highest priority path is currently "blocked and infeasible." At this point, the backup path switching logic is activated. The path is re-planned based on the new target location encoding. The entire switching process is autonomously decided by the algorithm without human intervention, ensuring that even in complex situations where some paths are unexpectedly blocked, the robot can still find a passable alternative route and ultimately safely complete its docking task.

[0099] Based on the same concept, this application also discloses a cleaning robot positioning method, executed by the cleaning robot described in any of the above embodiments, as shown in the appendix to the specification. Figure 6 As shown, the positioning method for the cleaning robot in this embodiment includes the following steps: S100, based on the frame sensing signal count collected by the sensor and the real-time rotation information of the drive wheel, the real-time position of the cleaning robot in the photovoltaic array is determined.

[0100] S200, based on the real-time location and the current operating status of the cleaning robot, plan the movement path of the cleaning robot.

[0101] In some embodiments, step S100, which determines the real-time position of the cleaning robot in the photovoltaic array based on the frame sensing signal count collected by the sensor and the real-time rotation information of the drive wheel, includes: S110, during the movement of the cleaning robot, whenever the sensor detects the frame of the photovoltaic module, a counting signal is triggered, and the current cumulative number of sensor counts is recorded.

[0102] S120, calculate the real-time walking distance of the cleaning robot based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated.

[0103] S130, the real-time walking distance is logically compared with the standard distance between the center lines of adjacent photovoltaic modules, and the real-time position of the cleaning robot is determined and corrected based on the comparison result.

[0104] In some implementations, step S200, which involves planning the movement path of the cleaning robot based on the real-time location and the current operating status of the cleaning robot, specifically includes: S210, acquire current environmental status data in real time, and determine whether the cleaning robot needs to dock based on the environmental status data.

[0105] S220: By analyzing the changing trend of the real-time position, the current operating status of the cleaning robot is determined, and a docking strategy is planned by combining the current operating status with the real-time position.

[0106] S230, according to the planned docking strategy, control the cleaning robot to move to the docking station to complete the docking.

[0107] The cleaning robot and the cleaning robot positioning method of this application have the same technical concept, and the technical details of the embodiments of the two are applicable to each other. In order to reduce repetition, they will not be repeated here.

[0108] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of program modules is merely an example. In practical applications, the above functions can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program units or modules to complete all or part of the functions described above. The program modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software program unit. Furthermore, the specific names of the program modules are only for easy differentiation and are not intended to limit the scope of protection of this application.

[0109] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A cleaning robot, characterized in that, include: ontology; The drive mechanism is built into the main body; A drive wheel is connected to the drive mechanism, which drives the drive wheel to rotate. The windproof hook includes a first connecting part and a second connecting part. The first connecting part is connected to one side of the body, and the second connecting part is located at the end of the first connecting part away from the body. The second connecting part is used to extend to the bottom of the photovoltaic module. A sensor is disposed on the windproof hook and facing the photovoltaic module, the sensor being used to sense the frame of the photovoltaic module; Memory and processor; The processor includes a real-time positioning module, which is configured to determine the real-time position of the cleaning robot in the photovoltaic array based on the count of the frame sensing signals collected by the sensor and the real-time rotation information of the drive wheel. The processor also includes a path planning module, configured to plan the movement path of the cleaning robot based on the real-time location and the current operating status of the cleaning robot.

2. A cleaning robot as described in claim 1, characterized in that, The sensor is disposed at the second connection part, and the sensor is located below the photovoltaic module; or, the sensor is disposed at the first connection part, and the sensor is located on one side of the photovoltaic module. The real-time positioning module includes: A sensor counter is used to trigger a counting signal and record the current cumulative number of sensor counts whenever the sensor detects the frame of the photovoltaic module during the movement of the cleaning robot. A rotary wheel rangefinder is used to calculate the real-time walking distance of the cleaning robot based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated. A position corrector is used to logically compare the real-time walking distance with the standard distance between the center lines of adjacent photovoltaic modules, and determine and correct the real-time position of the cleaning robot based on the comparison result.

3. A cleaning robot as described in claim 2, characterized in that, The position corrector is configured to perform the following logical steps: The judgment is based on the ratio of the real-time walking distance to the standard distance: If the ratio is within the first threshold range, the positioning status is determined to be normal, and the real-time position is calculated using the current sensor count, the real-time walking distance, and the standard distance. If the ratio is within the second threshold range, a sensor miss is determined, and the current sensor count is incremented by one. The corrected sensor count and the corrected real-time walking distance are calculated, and the real-time position is calculated using the corrected sensor count, the corrected real-time walking distance, and the standard distance. If the ratio exceeds the upper limit of the second threshold range, the drive wheel is determined to be slipping and spinning freely. In this case, the real-time position remains unchanged, and an abnormal status warning is triggered. The upper limit of the first threshold range is less than or equal to the lower limit of the second threshold range.

4. A cleaning robot as described in claim 2, characterized in that, The inductive counter is also configured to: The counting signals triggered within the first time interval are merged using a software filtering algorithm and identified as single valid counts.

5. A cleaning robot as described in claim 1, characterized in that, The path planning module includes: A docking detector is used to acquire real-time environmental status data and determine whether the cleaning robot needs to dock based on the environmental status data. The strategy planner is used to determine the current operating status of the cleaning robot by analyzing the changing trend of the real-time location, and to plan a docking strategy by combining the current operating status with the real-time location. The drive controller, according to the planned docking strategy, controls the cleaning robot to move to the docking station and complete the docking.

6. A cleaning robot as described in claim 5, characterized in that, The strategy planner is configured to perform the following logical steps: Based on the temporal analysis and change pattern recognition of the real-time location, the current operating state of the cleaning robot is analyzed, and the current operating state includes at least one of yaw, drive wheel slippage, and obstacle crossing. Based on a pre-built location coding database, the relative distance between the real-time position of the cleaning robot and each docking location code is calculated; Based on the relative distance and the current operating state, the docking strategy is dynamically generated. The docking strategy includes at least one of the following: the cleaning robot's deviation correction control strategy, the autonomous back-back and escape strategy when a blockage occurs, and the intelligent obstacle-crossing decision-making for path obstacles.

7. A cleaning robot as described in claim 6, characterized in that, The cleaning robot includes two windproof hooks, which are respectively connected to both sides of the main body; each windproof hook on both sides has a sensor, and the counting signals of the two sensors are independent of each other; So that when the body of the cleaning robot tilts, the strategy planner can identify the yaw direction of the cleaning robot by comparing the difference between the counting signals on both sides.

8. A cleaning robot as described in claim 7, characterized in that, Also includes: The strategy planner is also configured to perform the following logical steps: Based on the relative distance, feasible paths are prioritized, and when the distances are similar, the path direction that is consistent with the current movement trend is selected first. When the highest priority feasible path becomes infeasible due to obstruction, it adaptively switches to the second-best backup path.

9. A positioning method for a cleaning robot, characterized in that, Performed by the cleaning robot according to any one of claims 1-8, the process includes the following steps: Based on the frame sensing signal count collected by the sensor and the real-time rotation information of the drive wheel, the real-time position of the cleaning robot in the photovoltaic array is determined. Based on the real-time location and the current operating status of the cleaning robot, the movement path of the cleaning robot is planned.

10. A positioning method for a cleaning robot as described in claim 9, characterized in that, The method of determining the real-time position of the cleaning robot in the photovoltaic array based on the frame sensing signal count and the real-time rotation information of the drive wheel collected by the sensor includes: During the movement of the cleaning robot, each time the sensor detects the frame of the photovoltaic module, a counting signal is triggered, and the current cumulative number of sensor counts is recorded. The real-time walking distance of the cleaning robot is calculated based on the number of rotations or rotation time of the drive wheel. The real-time walking distance is reset to zero after each counting signal is triggered, and the real-time walking distance is recalculated. The real-time walking distance is logically compared with the standard distance between the center lines of adjacent photovoltaic modules, and the real-time position of the cleaning robot is determined and corrected based on the comparison result.

Images