A waterless brush sweeping system and method suitable for desert environments

Abstract

The application provides a kind of anhydrous brush cleaning system and method suitable for desert environment, it is related to photovoltaic intelligent operation and maintenance technical field, aiming at the extreme environmental characteristics of high sand dust, strong ultraviolet, large temperature difference and water resource scarcity of desert photovoltaic power station, the application realizes frame positioning through laser scanning and visual fusion, calculates sand dust risk index and dynamically plans return path, adjusts high pressure airflow and secondary brush parameters according to sand dust characteristics, completes zero water consumption cleaning and risk control of photovoltaic module.The system includes walking mechanism, sensing detection unit, central controller, brush assembly and negative pressure dust collection unit, adopts double rail positioning walking wheel system, intelligent thermal management unit and sand dust early warning and path planning module.The application solves the problem that the prior art cannot simultaneously consider efficient cleaning, reliable operation and safety risk avoidance, and is suitable for automatic cleaning of photovoltaic modules in extreme desert environment.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic intelligent operation and maintenance technology, and in particular to a waterless brush cleaning system and method suitable for desert environments. Background Technology

[0002] Desert photovoltaic power stations are located in environments characterized by frequent sandstorms, strong ultraviolet radiation, large diurnal temperature variations, and extreme water scarcity, posing severe challenges to their daily operation and maintenance. Dust accumulation on the surface of photovoltaic modules is a core factor affecting the power generation efficiency of photovoltaic power stations. In desert environments, this severely restricts the power generation revenue and operational efficiency of photovoltaic power stations. Therefore, routine cleaning of photovoltaic modules is a core necessity for the operation and maintenance of desert photovoltaic power stations.

[0003] To address the need for waterless cleaning of photovoltaic modules in deserts, existing technologies have been used to develop and apply related dry cleaning robots. However, their core shortcomings are as follows: existing dry cleaning methods mostly employ single brush cleaning or single airflow blowing techniques. Single brush cleaning is difficult to effectively remove fine, adhering sand and dust in desert environments and can easily cause scratches and microcracks on the glass surface of photovoltaic modules. Meanwhile, single airflow blowing solutions are ineffective at removing adhering sand and dust and can easily cause secondary pollution from sand and dust. After cleaning, sand and dust can easily be redeposited on adjacent modules. Regarding sand-proof sealing, existing equipment mostly uses a single static structure seal, which cannot prevent the intrusion of ultrafine desert sand and dust, easily causing failures in electrical and transmission components. The protection level generally cannot meet the requirements for long-term stable operation. Regarding the walking mechanism, existing frame-walking equipment mostly relies on grippers or negative pressure adsorption for fixation. Grippers are prone to deformation and damage to the photovoltaic frame, and negative pressure adsorption systems are prone to cavity blockage and adsorption failure in sandy environments, and have extremely high energy consumption. At the same time, the walking wheels are not optimized for gripping force under sandy conditions, and are prone to slipping, deviation, or even falling. Regarding environmental tolerance, existing equipment has not been specifically optimized for the strong ultraviolet radiation and extreme high temperature environment of the desert. Polymer components are prone to rapid aging and embrittlement, and electronic components are prone to overheating and failure under high temperature conditions in summer. The equipment has a short fault-free operation time and high maintenance costs, which cannot meet the needs of long-term large-scale operation and maintenance of desert photovoltaic power stations.

[0004] Therefore, the industry urgently needs a photovoltaic cleaning technology solution that can achieve zero water consumption, adapt to the extreme desert environment, and simultaneously ensure cleaning efficiency, equipment reliability, and operational safety. Summary of the Invention

[0005] The main objective of this invention is to provide a waterless brush cleaning system and method suitable for desert environments, which solves the problem that existing photovoltaic cleaning technologies cannot simultaneously achieve efficient cleaning, reliable operation, and safety avoidance. Through the dynamic adaptation and collaborative linkage of the system's core modules, it achieves automatic cleaning of desert photovoltaic modules with zero water consumption, high reliability, and long lifespan.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a waterless brush cleaning method suitable for desert environments, comprising: S1. Border positioning is achieved through laser scanning and visual fusion; S2. Calculate the dust storm risk index and dynamically plan the return path; S3. Based on the characteristics of sand and dust, the high-pressure airflow and secondary brush parameters are adaptively adjusted to complete the zero-water-consumption cleaning and hazard avoidance control of the photovoltaic module.

[0007] Preferably, in step S1, the border positioning is achieved by fusing laser scanning and vision, specifically including: The three-dimensional contour of the photovoltaic panel frame is obtained by laser dot matrix scanning, and visual images of the frame are acquired at the same time. The edge point set of the border in the laser dot matrix is ​​extracted using an edge detection algorithm; By using Hough to detect border lines in the visual image, the laser data is fused with the visual image to calculate the border width, and the wheel system posture is automatically corrected based on the identified border position.

[0008] Preferably, in step S1, the frame positioning also includes applying a preset test force to detect the initial gripping effect of the wheel system, measuring the slippage of the wheel system, determining that the gripping force is insufficient when the slippage exceeds a preset threshold, and starting the frame contact surface cleaning program; at the same time, the stiffness of the suspension system is adaptively adjusted according to the detected frame flatness.

[0009] Preferably, in step S2, the calculation of the dust risk index specifically includes: collecting the current dust concentration, wind speed, and temperature difference in the environment, normalizing them respectively, and then introducing a time correction factor based on the day and night time period and a seasonal correction factor based on the season. The normalized dust concentration, wind speed, and temperature difference are then weighted and summed to obtain the dust risk index.

[0010] Preferably, in step S2, the dynamic programming return path executes the corresponding operational strategy based on the calculated dust storm risk index level, including: When the risk index is below the first threshold, normal operation is maintained; when the risk index is between the first and second thresholds, the single operation time is shortened and the return inspection frequency is increased; when the risk index is above the second threshold, operation is suspended and the safety return procedure is initiated. The safe return procedure uses a path planning algorithm to plan the shortest return path along the border and dynamically controls the return speed based on visibility, wind speed, current battery level, and distance from the base station.

[0011] It also includes real-time calculation of the energy required for safe return and reserve of power for safe return. When the power is insufficient to meet the required energy, it immediately switches to energy-saving mode to reduce power and sends out a support request. When the risk index is higher than the third threshold, it shuts down in an emergency and activates the end emergency anchoring device.

[0012] Preferably, in step S3, the high-pressure airflow is adaptively adjusted according to the characteristics of the sand and dust. Specifically, this includes: estimating the sand and dust thickness and identifying the sand and dust type based on the visually detected sand and dust coverage or airflow resistance; determining the base pressure and base pulse duty cycle of the high-pressure airflow based on the sand and dust thickness; and then introducing a temperature compensation factor determined based on the current temperature and a humidity compensation factor determined based on the current relative humidity to correct the base pressure and base duty cycle, thereby generating dynamic high-pressure airflow parameters.

[0013] Preferably, in step S3, the secondary brush parameters are adaptively adjusted according to the characteristics of the sand and dust, specifically including: The cleaning process is carried out sequentially using primary and secondary brushes. The contact pressure and rotation speed of each brush are dynamically adjusted based on the estimated thickness of the sand and dust and the identified type of sand and dust. After the primary brush performs initial cleaning, the cleanliness level is checked. When the cleanliness level is lower than the preset threshold, the secondary brush starts fine cleaning.

[0014] Preferably, step S3 further includes: monitoring the drive load of the brush during the waterless cleaning process; when the load exceeds the preset clogging threshold, determining that the brush is clogged with sand and dust, interrupting the cleaning and executing the self-cleaning logic; controlling the brush drive motor to rotate in the opposite direction, and simultaneously starting the high-frequency vibration motor to make the brush generate axial high-frequency vibration, and adjusting the negative pressure dust collection unit to the maximum power to extract the shaken sand and dust, and automatically resuming cleaning after the load returns to normal.

[0015] This invention provides a waterless brush cleaning system suitable for desert environments, including a walking mechanism, a sensing and detection unit, a central controller, a brush assembly, and a negative pressure dust collection unit; The walking mechanism is installed at the bottom of the robot and uses a tracked or wheeled chassis; The sensing and detection unit is integrated and installed at the front end of the robot and near the brush assembly, including a vision sensor, a torque detection module and a positioning module; The central controller uses a microcontroller or embedded controller, which incorporates algorithms for judging the degree of contamination and self-diagnostic algorithms for brush clogging. The brush assembly includes a flexible, wear-resistant brush roller, a drive motor, and an eccentric vibration motor; The negative pressure dust collection unit includes a centrifugal fan, a filter screen, and a removable dust collection box. The suction port is located next to the brush assembly and behind the brush in the direction of rotation.

[0016] Beneficial effects: 1. This invention employs a composite cleaning mechanism that combines high-pressure airflow with two-stage brushes. Through the initial peeling by the primary brush and the fine cleaning by the secondary brush, combined with the real-time extraction by the negative pressure dust collection unit, it effectively solves the problems of incomplete removal of fine-particle adhesive dust and easy scratching of the photovoltaic module surface by single brush cleaning. At the same time, it overcomes the shortcomings of single airflow blowing in removing adhesive dust and easily generating secondary dust pollution, achieving zero-water-consumption and high-efficiency cleaning of the photovoltaic module surface.

[0017] 2. This invention achieves real-time perception and graded response to extreme weather conditions in the desert by using a multi-factor dynamic assessment mechanism of the dust risk index, which integrates dust concentration, wind speed, temperature difference and time / seasonal correction factors. Through dynamic planning of the shortest path along the edge and multi-factor speed control, it ensures that the equipment can safely return to the base station under adverse conditions such as reduced visibility and increased wind speed, effectively improving the operational safety and risk avoidance reliability of the equipment in the extreme desert environment.

[0018] 3. This invention estimates the thickness and type of sand and dust through visual detection and airflow resistance, and introduces temperature compensation factors and humidity compensation factors to dynamically correct the high-pressure airflow parameters and brush parameters. This allows the cleaning intensity to match the characteristics of sand and dust, temperature and humidity conditions in the desert environment in real time, avoiding the problem of surface damage to components caused by over-cleaning or substandard cleanliness caused by insufficient cleaning, and achieving the best balance between cleaning efficiency and surface protection.

[0019] 4. This invention achieves automatic correction of wheel system posture through frame positioning technology that combines laser scanning and vision; by applying a test force to detect wheel system slippage and initiating a frame contact surface cleaning program, combined with adaptive adjustment of suspension system stiffness, it effectively solves the problem of slippage, deviation, or even fall of existing equipment under sand and dust cover conditions, and significantly improves the gripping stability and operational reliability of the walking mechanism on the photovoltaic frame. Attached Figure Description

[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of a waterless brush cleaning method suitable for desert environments according to the present invention; Figure 2 This is a structural diagram of a waterless brush cleaning system suitable for desert environments according to the present invention. Detailed Implementation

[0021] Example 1 like Figure 1 As shown, a waterless brush cleaning method suitable for desert environments includes the following steps: S1. Collect meteorological data, assess the condition of the frame, assess the surface contamination of the components, and predict the risk of sandstorm activity in the next 24 hours; S2. Obtain the three-dimensional contour of the frame through laser dot matrix scanning, automatically correct the attitude and position of the wheel system, apply a 10N force to test the initial gripping effect, and adjust the suspension stiffness according to the flatness of the frame. S3. Calculate the sandstorm risk index, implement the corresponding strategy according to the risk level, return along the original route along the edge, dynamically adjust the speed according to the visibility, and reserve 30% of the power for emergency return. S4. Assess the thickness and type of sand and dust based on visual and airflow resistance, dynamically adjust high-pressure airflow parameters, brush pressure and rotation speed, complete the secondary cleaning verification and optimize sand and dust collection. S5. Assess sealing performance, check wheel condition, monitor brush wear, count dust collection volume, and perform predictive maintenance.

[0022] According to an embodiment of the present invention, in step S1, meteorological data is collected synchronously at a frequency of 10Hz using a local sensor array, including temperature, humidity, wind speed, and dust concentration. In this embodiment, the dust concentration sensor employs the principle of laser scattering, detecting particle sizes ranging from 0.3 to 10 μm and measuring from 0 to 1000 mg / m³. Then, four sets of laser ranging sensors are used to scan the three-dimensional contour of the frame, combined with a 1.3-megapixel CMOS vision sensor for frame edge detection. The Canny edge detection algorithm is used to extract frame features, and Hough transform is used to detect straight line boundaries and identify the frame type. Simultaneously, a multispectral sensor measures light reflectance to quantify the pollution level, a miniature laser scatterer analyzes the dust particle size distribution, and a miniature force sensor measures dust adhesion. Based on five years of historical dust activity data, an LSTM neural network model is used to predict the dust risk for the next 24 hours. Input features include: current dust concentration, wind speed, temperature, humidity, seasonal factors, and time factors; the model output is the hourly dust probability distribution.

[0023] In this embodiment, the LSTM neural network model adopts a 2-layer LSTM structure, with 128 hidden units in each layer. The input layer receives a 6-dimensional feature vector [dust concentration, wind speed, temperature, humidity, seasonal factor, time factor]. The output layer consists of 24 nodes corresponding to the predictions for the next 24 hours, using the ReLU activation function and the Adam optimizer. The training data consists of 5 years of historical data, divided into training and test sets in an 8:2 ratio, with a batch size of 32 and a learning rate of 0.001. Train for 200 rounds until the loss converges.

[0024] According to an embodiment of the present invention, in step S2, the border positioning and alignment includes: border detection by fusing laser data and visual images. The laser data processing employs an edge detection algorithm, sets a threshold of 2mm to extract border edge point sets, and calculates the border width. After preprocessing the visual image, including brightness adjustment and contrast enhancement, Canny edge detection and Hough transform are used to detect border lines and calculate the visual width. The data fusion formula is shown in equation (1) below: (1); in, To score the reliability of visual recognition, This is the width of the merged border, used for subsequent alignment control.

[0025] Based on the identified border position, attitude adjustment is achieved through differential gear control. The offset of the border centerline is defined as... The control quantity is generated using the PID control algorithm. As shown in equation (2): (2); In this embodiment, the scaling factor Integral coefficient Differential coefficients The control objective is to keep the offset within ±2mm within 0.5 seconds.

[0026] A 10N test force is applied to measure the wheel system slippage. If the slippage exceeds 1mm, insufficient grip is determined, and the side contact surface cleaning procedure is initiated. Based on the side flatness test results, the suspension system stiffness is adaptively adjusted. As shown in equation (3): (3); in, N / mm is the base stiffness. For adjustment coefficients, This represents the standard deviation of the border height.

[0027] The risk index is calculated by combining dust concentration, wind speed, temperature difference, time factor, and seasonal factor. As shown in equation (4): (4); in, To normalize the dust concentration, among which . For normalized wind speed, among which . For normalized temperature difference, where . The time correction factor is 1.3 for nighttime (19:00-07:00) and 1.0 for daytime. The seasonal adjustment factor is set at 1.5 for spring (March-May), 1.2 for winter (November-February), and 1.0 for other seasons. The risk index is limited to the range [0,1].

[0028] Warning level classification and response strategy: Low risk ( <0.3): Normal operation, maintaining standard operating parameters; Medium risk (0.3≤R<0.6): Reduce single operation time to 60% and increase return inspection frequency to once every 30 minutes; High risk (R≥0.6): Suspend operations and immediately initiate the safety return procedure; Extreme risk (R≥0.9): Emergency shutdown, all systems enter protection mode, and the end emergency anchoring device is activated.

[0029] The A* algorithm is used to plan the shortest return path along the border, avoiding crossing component gaps. The dynamic speed control function is shown in equation (5) below: (5); in, Based on m / s velocity The visibility factor is 1.0 when visibility is >100m, 0.5 when visibility is 50-100m, and 0.2 when visibility is <50m. For wind speed factor, where . This is the power factor, which is 0.5 when the power is less than 30% and 1.0 otherwise. It is a distance factor; it is 0.6 when the distance to the base station is less than 10m, and 1.0 otherwise.

[0030] Real-time calculation of the energy required for return Ensure current battery level satisfy ,in This is the total battery capacity. If the battery is low, immediately switch to power saving mode to reduce power by 50% and request support.

[0031] According to an embodiment of the present invention, the waterless cleaning strategy in step S4 includes: S41. Pollution Assessment and Parameter Adaptation: Estimate dust thickness h (range 0-3 mm) based on visually detected dust coverage and airflow resistance; identify dust type (fine sand / coarse sand / sticky dust / mixed type) through spectral analysis. High-pressure airflow pressure. and pulse duty cycle The adaptive adjustment formula is shown below: (6); in, MPa is the base pressure Based on duty cycle. For temperature compensation factor, where . This is a temperature compensation factor; when the temperature is >45℃ ,otherwise . For humidity compensation factor, where .

[0032] S42. Brush control algorithm: Dynamically adjusts brush pressure based on dust thickness, type, and temperature. and rotational speed As shown in the following formula: (7); in, N / cm 2 Based on pressure, rpm is the base rotational speed. Here, is the thickness compensation factor, where . The type compensation factor has specific values ​​for different types, including 0.9 for fine sand, 1.3 for coarse sand, 1.5 for sticky dust, and 1.1 for mixed types. For temperature compensation factor, where .

[0033] S43. Secondary Cleaning Verification: After the primary brush completes the initial cleaning, the cleanliness is detected by infrared reflectivity. If the cleanliness is <90%, the secondary brush is activated for fine cleaning. The primary brush uses a carbon fiber-silicone composite material with a diameter of 0.8mm, a density of 150 fibers / cm², and a pressure of 0.8-1.5N / cm². The secondary brush uses a microfiber-TPU composite material with a diameter of 0.3mm, a density of 300 fibers / cm², and a pressure of 0.3-0.8N / cm².

[0034] S44. Dust Collection Optimization: Based on dust volume and homework time Adjust the working mode of the dust collection system as shown in equation (8): (8); electrostatic deposition power The value is adjusted according to humidity: 100% when humidity is <30%, 70% when humidity is between 30% and 60%, and 30% when humidity is >60%.

[0035] According to an embodiment of the present invention, in step S5, system status monitoring and maintenance includes: S501. Sealing performance evaluation: Real-time monitoring of internal and external pressure difference using a differential pressure sensor. ,like If the pressure remains at 300 Pa for 5 seconds, a seal failure is determined, and emergency mode is activated. At this point, the airflow velocity increases to 3.0 m / s, and the pressure increases to 300 Pa. An infrared thermal imager detects abnormal temperatures in the sealing area to identify the leak location.

[0036] S502, Wheel Condition Check: Perform an adhesion test after every 100m of travel. Monitor wheel system current. ,like If the wheel surface is found to be clogged with sand and dust after 10 seconds, cleaning will be temporarily suspended and the self-cleaning program will be started.

[0037] S503, Brush Wear Monitoring: The condition of the brush is assessed through motor current feedback and cleaning performance, and a wear index is defined. As shown in equation (9): (9); in, For the new brush current, To discard the brush current, This is the current. When The system alarms when Forced replacement at any time.

[0038] S504. Predictive Maintenance: Predicting the remaining life of critical components based on the Arrhenius model, as shown in the following formula: (10); in, Time already used kJ / mol is the activation energy. J / (mol·K), This refers to the actual operating temperature. The system provides a 30-day advance warning of maintenance needs based on the rated temperature.

[0039] Example 2 like Figure 2 As shown, this embodiment provides a waterless brush cleaning system for a photovoltaic cleaning robot suitable for desert environments. The system includes: a walking mechanism, a sensing and detection unit, a central controller, a brush assembly, and a negative pressure dust collection unit.

[0040] Photovoltaic panel surface (attached with sand and dust): As an object to be cleaned, its surface is covered with dust, fine sand or a small amount of scale deposited in the desert environment.

[0041] The locomotion mechanism, mounted on the bottom of the robot, uses a tracked or wheeled chassis to support the robot body and move across the photovoltaic panel array surface. The locomotion mechanism is electrically connected to the central controller and receives path commands from the central controller.

[0042] Sensing and detection unit: Integrated and installed at the front end of the robot and near the brush assembly, including at least: A visual sensor uses a grayscale camera to capture images of the photovoltaic panel surface to determine the degree of contamination. A torque detection module is used to monitor the drive load of the brush assembly in real time. The positioning module uses RTK-GPS to record the robot's real-time location and the cleaned area.

[0043] The signal output terminal of the sensing and detection unit is connected to the signal input terminal of the central controller.

[0044] Central controller: Employs a microcontroller unit (MCU) or embedded controller, incorporating algorithms for determining contamination levels and for self-diagnosing brush clogging. The control signal outputs of the central controller are connected to the walking mechanism, brush assembly, and negative pressure dust collection unit, respectively.

[0045] Brush Assembly: Mounted on the bottom or front of the robot, it includes a flexible, wear-resistant brush roller, a drive motor, and an eccentric vibration motor. Under the control of the central controller, the brush assembly performs the following actions: rotating the brush in either forward or reverse direction; and axial high-frequency vibration to shake off sand and dust adhering to the bristles. The working surface of the brush assembly maintains elastic contact with the photovoltaic panel surface.

[0046] Negative pressure dust collection unit: Includes a centrifugal fan, filter, and removable dust collection box. The suction port of the negative pressure dust collection unit is located adjacent to the brush assembly, behind it in the direction of brush rotation. This unit is connected to the working area of ​​the brush assembly via air ducts. Under the command of the central controller, it can generate negative pressure of different levels to extract the sand and dust blown off by the brush and collect it into the dust collection box.

[0047] The complete working process of this system performing waterless cleaning in a desert environment is as follows: (1) Before starting the assignment: The central controller receives image data of the photovoltaic panel surface from the sensing unit and determines the pollution level of the current area using a built-in algorithm. If the pollution level is below the first threshold T1, it enters the "waterless dry mode"; if it is above T1, it enters the "enhanced mode" and simultaneously activates brush rotation and negative pressure adsorption.

[0048] (2) During cleaning operations: The central controller sends path commands to the walking mechanism and simultaneously sends a forward rotation command to the drive motor of the brush assembly, causing the flexible brush rollers to rotate and scrape the surface of the photovoltaic panel, removing the attached sand and dust. At the same time, the central controller activates the negative pressure dust collection unit, generating negative pressure airflow to extract the sand and dust scraped off by the brushes in real time through the suction port. After being intercepted by a filter, the dust falls into the dust collection box. This achieves waterless dry cleaning by "sweeping and vacuuming simultaneously."

[0049] (3) Self-cleaning process: During the cleaning process, the torque detection module in the sensing unit continuously monitors the current or torque value of the drive motor of the brush assembly. When this value exceeds the second threshold T2, it indicates that excessive sand and dust have become entangled or clogged on the brush rollers. At this time, the central controller immediately interrupts normal cleaning and executes the self-cleaning logic: Control the brush drive motor to rotate in the opposite direction, causing the brush bristles to swing in the opposite direction; At the same time, the high-frequency vibration motor is started to make the brush roller generate axial high-frequency vibration, with a frequency range of 50Hz to 200Hz, to further shake off the sand and dust embedded in the root of the bristles. Set the fan of the negative pressure dust collection unit to maximum power to forcefully suck up the shaken-off sand and dust.

[0050] After the self-cleaning process continues for a preset time (e.g., 5-10 seconds), the central controller determines that the torque value has returned to normal and then automatically resumes the cleaning operation.

[0051] (4) After completing the assignment: Once the positioning module in the sensing and detection unit reports that the robot has traversed the entire planned path, and the vision sensor confirms that the grayscale value of the panel surface has reached the cleanliness standard, the central controller generates a cleaning completion signal, stops all actuators, and uploads the operation log to the remote server. Collected sand and dust are stored in the dust collection box for centralized cleaning during maintenance.

[0052] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A waterless brush cleaning method suitable for desert environments, characterized by: include: S1. Border positioning is achieved through laser scanning and visual fusion; S2. Calculate the dust storm risk index and dynamically plan the return path; S3. Based on the characteristics of sand and dust, the high-pressure airflow and secondary brush parameters are adaptively adjusted to complete the zero-water-consumption cleaning and hazard avoidance control of the photovoltaic module.

2. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S1, the border is located by fusing laser scanning and vision, specifically including: The three-dimensional contour of the photovoltaic panel frame is obtained by laser dot matrix scanning, and visual images of the frame are acquired at the same time. The edge point set of the border in the laser dot matrix is ​​extracted using an edge detection algorithm; By using Hough to detect border lines in the visual image, the laser data is fused with the visual image to calculate the border width, and the wheel system posture is automatically corrected based on the identified border position.

3. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S1, the frame positioning also includes applying a preset test force to detect the initial gripping effect of the wheel system, measuring the slippage of the wheel system, determining that the gripping force is insufficient when the slippage exceeds a preset threshold, and starting the frame contact surface cleaning program; at the same time, the stiffness of the suspension system is adaptively adjusted according to the detected frame flatness.

4. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S2, the dust risk index is calculated, which specifically includes: collecting the dust concentration, wind speed and temperature difference of the current environment, normalizing them respectively, and then introducing a time correction factor based on the day and night time period and a seasonal correction factor based on the season. The normalized dust concentration, wind speed and temperature difference are then weighted and summed to obtain the dust risk index.

5. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S2, the dynamic programming return path executes the corresponding operational strategy based on the calculated dust storm risk index level, including: When the risk index is below the first threshold, normal operation is maintained; when the risk index is between the first and second thresholds, the single operation time is shortened and the return inspection frequency is increased; when the risk index is above the second threshold, operation is suspended and the safety return procedure is initiated. The safe return procedure uses a path planning algorithm to plan the shortest return path along the border and dynamically controls the return speed based on visibility, wind speed, current battery level, and distance from the base station.

6. The waterless brush cleaning method suitable for desert environments according to claim 5, characterized in that: It also includes real-time calculation of the energy required for safe return and reserve of power for safe return. When the power is insufficient to meet the required energy, it immediately switches to energy-saving mode to reduce power and sends out a support request. When the risk index is higher than the third threshold, it shuts down in an emergency and activates the end emergency anchoring device.

7. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S3, the high-pressure airflow is adaptively adjusted according to the characteristics of the sand and dust. Specifically, this includes: estimating the sand and dust thickness and identifying the sand and dust type based on the visually detected sand and dust coverage or airflow resistance; determining the base pressure and base pulse duty cycle of the high-pressure airflow based on the sand and dust thickness; and then introducing a temperature compensation factor determined based on the current temperature and a humidity compensation factor determined based on the current relative humidity to correct the base pressure and base duty cycle, thereby generating dynamic high-pressure airflow parameters.

8. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: In step S3, the parameters of the secondary brush are adaptively adjusted according to the characteristics of the sand and dust, specifically including: The cleaning process is carried out sequentially using primary and secondary brushes. The contact pressure and rotation speed of each brush are dynamically adjusted based on the estimated thickness of the sand and dust and the identified type of sand and dust. After the primary brush performs initial cleaning, the cleanliness level is checked. When the cleanliness level is lower than the preset threshold, the secondary brush starts fine cleaning.

9. The waterless brush cleaning method suitable for desert environments according to claim 1, characterized in that: Step S3 also includes: monitoring the drive load of the brush during the waterless cleaning process; when the load exceeds the preset clogging threshold, determining that the brush is clogged with sand and dust, interrupting the cleaning and executing the self-cleaning logic; controlling the brush drive motor to rotate in the opposite direction, and simultaneously starting the high-frequency vibration motor to make the brush generate axial high-frequency vibration, and adjusting the negative pressure dust collection unit to the maximum power to extract the shaken sand and dust, and automatically resuming cleaning after the load returns to normal.

10. A waterless brush cleaning system suitable for desert environments, characterized in that: It includes a walking mechanism, a sensor detection unit, a central controller, a brush assembly, and a negative pressure dust collection unit; The walking mechanism is installed at the bottom of the robot and uses a tracked or wheeled chassis; The sensing and detection unit is integrated and installed at the front end of the robot and near the brush assembly, including a vision sensor, a torque detection module and a positioning module; The central controller uses a microcontroller or embedded controller, which incorporates algorithms for judging the degree of contamination and self-diagnostic algorithms for brush clogging. The brush assembly includes a flexible, wear-resistant brush roller, a drive motor, and an eccentric vibration motor; The negative pressure dust collection unit includes a centrifugal fan, a filter screen, and a removable dust collection box. The suction port is located next to the brush assembly and behind the brush in the direction of rotation.

Images