Sensor cleaning control method of unmanned vehicle and related device thereof

Abstract

The invention relates to a sensor cleaning control method of an unmanned vehicle and a related device thereof, a sensor of the unmanned vehicle is provided with a pollution monitoring unit, and the method comprises the following steps: obtaining multi-element pollution data collected by the pollution monitoring unit; wherein the pollution detection unit at least comprises an image acquisition module and a laser detection module; the multi-element pollution data comprises image data acquired by the image acquisition module and laser data acquired by the laser detection module; the image data and the laser data are analyzed to obtain a pollution result of the sensor, and the pollution result is used for indicating the pollution degree of the sensor; when the pollution result indicates that the sensor is in a pollution state, determining a cleaning strategy based on the multi-element pollution data; and controlling a cleaning system of the unmanned vehicle to clean the sensor according to the cleaning strategy. According to the scheme provided by the invention, the pollution type and degree of the sensor can be accurately identified, a targeted cleaning strategy is adopted, and the cleaning efficiency and the cleaning capability are improved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a sensor cleaning control method and related apparatus for unmanned vehicles. Background Technology

[0002] Autonomous vehicles, such as those used in mining, agriculture, and logistics, rely heavily on various sensors (such as LiDAR, cameras, and millimeter-wave radar) to perceive their surroundings and achieve autonomous navigation and safe operation. However, these vehicles often operate in harsh environments, such as dust storms in mines, mud splashes in farmland, and inevitably encountering rain, snow, and fog during all-weather operations.

[0003] In related technologies, sensors are often covered with thick mud in special environments such as mines. Traditional regular cleaning methods lack specificity and cannot adjust cleaning strategies according to the actual pollution situation, resulting in waste of cleaning resources or poor cleaning effect. Moreover, there are no good targeted cleaning solutions for different types of pollutants, such as dust and liquid. Summary of the Invention

[0004] To address or partially address the problems existing in related technologies, this application provides a sensor cleaning control method and related apparatus for unmanned vehicles, which can accurately identify the type and degree of sensor contamination and adopt targeted cleaning strategies to improve cleaning efficiency and cleaning capacity.

[0005] This application provides a sensor cleaning control method for an unmanned vehicle. The unmanned vehicle's sensors are equipped with a pollution monitoring unit. The method includes: acquiring multivariate pollution data collected by the pollution detection unit; wherein the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; analyzing the image data and the laser data to obtain a pollution result for the sensor, the pollution result indicating the degree of pollution of the sensor; when the pollution result indicates that the sensor is in a polluted state, determining a cleaning strategy based on the multivariate pollution data; and controlling the unmanned vehicle's cleaning system to clean the sensor according to the cleaning strategy.

[0006] In conjunction with the first aspect, in one possible implementation of the first aspect, the laser data includes the laser angle and laser energy of the reflected laser; the step of analyzing the image data and the laser data to obtain the contamination result of the sensor includes: performing grayscale processing on the image data to obtain grayscale image data; identifying contaminated pixels with grayscale values ​​exceeding a preset grayscale value from the grayscale image data, and generating a contamination mask based on the contaminated pixels; calculating the average grayscale value of the contamination mask based on the grayscale values ​​of the contaminated pixels in the contamination mask; and obtaining the contamination result of the sensor based on the average grayscale value of the contamination mask, the laser angle, and the laser energy.

[0007] In conjunction with the first aspect, in one possible implementation of the first aspect, before determining the cleaning strategy based on the multivariate pollution data when the pollution result indicates that the sensor is in a polluted state, the method further includes: inputting the multivariate pollution data into a pollutant classification model to determine the category of pollutants on the sensor; wherein the category of pollutants includes at least dust or liquid; and determining the cleaning strategy based on the category of pollutants.

[0008] In conjunction with the first aspect, in one possible implementation of the first aspect, controlling the cleaning system of the unmanned vehicle to clean the sensor according to the cleaning strategy includes: when the type of pollutant is dust, determining the cleaning strategy as a first cleaning strategy; determining a first cleaning parameter in the first cleaning strategy based on the pollution result; wherein the first cleaning parameter includes at least the water spray volume, water spray intensity, and nozzle position of the water spray module in the cleaning system; and controlling the cleaning system of the unmanned vehicle to clean the pollutant based on the first cleaning strategy.

[0009] In conjunction with the first aspect, in one possible implementation of the first aspect, controlling the cleaning system of the unmanned vehicle to clean the sensor according to the cleaning strategy includes: when the type of the contaminant is liquid, determining the cleaning strategy as a second cleaning strategy; determining a second cleaning parameter in the second cleaning strategy based on the contamination result; wherein the second cleaning parameter includes at least the heating temperature of the heating module and the water spray volume of the spray module in the cleaning system; and controlling the cleaning system of the unmanned vehicle to clean the contaminant based on the second cleaning strategy.

[0010] In conjunction with the first aspect, in one possible implementation of the first aspect, after controlling the cleaning system of the unmanned vehicle to clean the sensor, the method further includes: controlling the image acquisition module to acquire target image data of the sensor; inputting the target image data into an image recognition model to obtain a recognition result; the recognition result is used to indicate whether the sensor has been successfully cleaned; if the recognition result indicates that the sensor has not been successfully cleaned, then adjusting the cleaning parameters in the cleaning strategy, and controlling the cleaning system of the unmanned vehicle to clean the sensor according to the adjusted cleaning parameters.

[0011] A second aspect of this application provides a sensor cleaning control device for an unmanned vehicle, comprising: an acquisition module for acquiring multivariate pollution data collected by a pollution detection unit; wherein the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; an analysis module for analyzing the image data and the laser data to obtain a pollution result of the sensor, the pollution result indicating the degree of pollution of the sensor; a determination module for determining a cleaning strategy based on the multivariate pollution data when the pollution result indicates that the sensor is in a polluted state; and a cleaning module for controlling the cleaning system of the unmanned vehicle to clean the sensor according to the cleaning strategy.

[0012] A third aspect of this application provides an electronic device, comprising: Processor; and A memory that stores executable code, which, when executed by the processor, causes the processor to perform the method described above.

[0013] A fourth aspect of this application provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method described above.

[0014] The fifth aspect of this application provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the method described above.

[0015] The technical solution provided in this application may include the following beneficial effects: This application discloses a sensor cleaning control method and related apparatus for an unmanned vehicle. The unmanned vehicle's sensors are equipped with a contamination monitoring unit. The method includes: acquiring multivariate contamination data collected by the contamination detection unit; wherein the contamination detection unit includes at least an image acquisition module and a laser detection module; the multivariate contamination data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; analyzing the image data and laser data to obtain a contamination result for the sensor, the contamination result indicating the degree of contamination of the sensor; when the contamination result indicates that the sensor is in a contaminated state, determining a cleaning strategy based on the multivariate contamination data; and controlling the unmanned vehicle's cleaning system to clean the sensor according to the cleaning strategy, which can accurately identify the type and degree of sensor contamination and adopt targeted cleaning strategies to improve cleaning efficiency and cleaning capability.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0017] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments thereof.

[0018] Figure 1 This is a schematic flowchart illustrating the sensor cleaning control method for an unmanned vehicle according to an embodiment of this application; Figure 2 This is a schematic diagram of the sensor cleaning control device for an unmanned vehicle shown in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation

[0019] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0020] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a” and “the” as used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0021] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0022] In harsh environments such as mines, sensors on autonomous vehicles face serious contamination problems. Mine environments typically contain large amounts of dust, extreme temperature differences, rain, snow, splashes of water, and mud, among other contaminants. These contaminants adhere to the sensor surfaces, severely affecting their sensitivity and normal operation. This is particularly true for optical and laser sensors on autonomous vehicles; surface contamination can lead to distorted data and even complete sensor failure.

[0023] In related technologies, sensor cleaning solutions mainly suffer from the following shortcomings: First, traditional periodic cleaning methods lack specificity and cannot adjust cleaning strategies according to actual pollution levels, leading to wasted cleaning resources or poor cleaning results. Second, existing pollution detection methods are limited, often relying solely on visual or simple physical detection, making it difficult to accurately determine the degree and type of pollution. Furthermore, there is a lack of differentiated treatment solutions for different types of pollutants, such as dust and liquids, resulting in low cleaning efficiency. Especially in special environments such as mines, sensors often face thick layers of mud. In such cases, conventional cleaning methods are often ineffective in removing pollutants and may even cause the cleaning system itself to become clogged and malfunction, affecting not only the normal operation of autonomous vehicles but also potentially posing safety hazards.

[0024] To address the aforementioned issues, this application provides a sensor cleaning control method for unmanned vehicles, which can accurately identify the type and degree of sensor contamination and adopt targeted cleaning strategies to improve cleaning efficiency and capabilities.

[0025] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0026] Figure 1 This is a schematic flowchart illustrating the sensor cleaning control method for an unmanned vehicle as shown in an embodiment of this application.

[0027] See Figure 1 A sensor cleaning control method for an unmanned vehicle, wherein the sensors of the unmanned vehicle are equipped with a pollution monitoring unit, including: S110: Acquire multivariate pollution data collected by the pollution detection unit; wherein the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module.

[0028] Specifically, a pollution monitoring unit refers to a detection device installed on a sensor for real-time monitoring of pollutant status. A pollution monitoring unit includes at least an image acquisition module and a laser detection module. Multivariate pollution data refers to a set of multidimensional information that can jointly characterize the pollution status of the sensor surface, which is collected synchronously or asynchronously by these two modules. For example, it can be obtained by using two-dimensional visual data output by the image acquisition module and three-dimensional reflection data acquired by the laser detection module. It can improve the accuracy of pollution detection by combining spatial coverage and reflection characteristics differences.

[0029] For example, the image acquisition module can be a camera, such as a CMOS camera with a 2-megapixel resolution, whose lens angle can completely cover the monitored sensor surface. To cope with different lighting conditions, the module can be equipped with a supplementary lighting unit, such as an infrared LED, to ensure clear image data can be acquired even at night or in low-light environments such as mines. The laser detection module can be a device containing a laser emitter (such as a 905 nm laser diode) and a photodetector. It can be a single-point ranging module or a miniature scanning galvanometer, used to emit a laser beam to one or more predetermined points on the sensor surface and receive the reflected signal. The image data can be one or more frames of digital images captured by the camera (e.g., a 1920x1080 pixel RGB or grayscale image), while the laser data are the physical property parameters of the reflected laser measured by the laser detection module.

[0030] S120: Analyzes image data and laser data to obtain sensor contamination results, which are used to indicate the degree of sensor contamination.

[0031] Specifically, the contamination results can be used to assess the contamination status of the sensor, and based on the contamination results, it can be determined whether to trigger a cleaning process for the sensor.

[0032] In one possible implementation, the laser data includes the laser angle and laser energy of the reflected laser. The step of analyzing the image data and laser data to obtain the contamination result of the sensor includes: performing grayscale processing on the image data to obtain grayscale image data; identifying contaminated pixels with grayscale values ​​exceeding a preset grayscale value from the grayscale image data, and generating a contamination mask based on the contaminated pixels; calculating the average grayscale value of the contamination mask based on the grayscale values ​​of the contaminated pixels in the contamination mask; and obtaining the contamination result of the sensor based on the average grayscale value of the contamination mask, the laser angle, and the laser energy.

[0033] Specifically, grayscale processing refers to the process of converting a color image into a grayscale image. For example, in image data, the pixel values ​​of the RGB channels can be converted into a single grayscale value to obtain grayscale image data. A preset grayscale value can be set in advance, for example, to 200. Contaminated pixels refer to pixels whose grayscale values ​​exceed a preset threshold. A contaminated mask is generated based on these contaminated pixels, and the contaminated mask can be generated using a binarization algorithm. The contamination result can be calculated based on the contaminated mask, laser angle, and laser energy. For example, the formula for calculating the contamination result is: PollutionScore=w1*AreaCoverage+w2*AvgGrayValue+w3*f(LaserAngleDeviation)+w4*g(1 / LaserEnergy) In this formula, w1, w2, w3, and w4 are weighting coefficients calibrated based on experimental data and can be preset; AreaCoverage is the contaminated area coverage rate; AvgGrayValue is the average gray value; f(LaserAngleDeviation) and g(1 / LaserEnergy) are functions that normalize the laser angle deviation and the reciprocal of the laser energy, respectively; and PollutionScore is the pollution result. This formula integrates multi-dimensional sensor information into a numerical value that intuitively reflects the degree of pollution. For example, the scoring range can be set to 0-100, where 0-10 points are defined as "clean," 11-50 points as "moderately polluted," and 51-100 points as "severely polluted."

[0034] Specifically, when contaminants are present on the sensor surface, the image data captured by the camera is processed into grayscale. High-grayscale contaminant pixels are identified and a contamination mask is generated. The area covered by the contamination mask and the average grayscale value together reflect the degree and density of contaminant coverage. Simultaneously, the laser beam emitted by the laser sensor experiences a reflection angle shift and energy attenuation when encountering contaminants. Changes in the reflection angle can determine the surface morphology of the contaminant; for example, liquid contaminants may cause specular reflection, while dust contaminants may cause diffuse reflection. The degree of energy attenuation is related to the light transmittance or absorptive properties of the contaminant. By fusing and analyzing the average grayscale value in the image data with the laser angle and energy data, the degree of contaminant coverage and physical characteristics can be comprehensively assessed, thereby accurately evaluating the sensor's contamination status. This allows for multi-dimensional analysis of the sensor's contamination status based on image and laser data, improving the accuracy of contamination detection.

[0035] For example, the acquired image data is first converted to grayscale, transforming the color image into a grayscale image. Then, the grayscale value of each pixel in the grayscale image is judged according to a preset grayscale threshold. If the grayscale value of a pixel exceeds 200, the pixel is marked as a contaminated pixel. A binarized contaminated mask image is generated based on all marked contaminated pixels. The average grayscale value of the contaminated mask is calculated. For example, if there are 1000 contaminated pixels in the contaminated mask, and the sum of their grayscale values ​​is 220000, the average grayscale value is 220. Based on the average grayscale value of 220, the laser angle deviation of 5°, and the laser energy attenuation of 20%, the contamination result of the sensor can be obtained. For example, when the average grayscale value is greater than 200, the laser angle deviation is greater than 3°, and the laser energy attenuation is greater than 15%, the sensor is determined to be heavily contaminated.

[0036] S130: When the contamination result indicates that the sensor is in a contaminated state, a cleaning strategy is determined based on multivariate contamination data.

[0037] Specifically, a cleaning strategy refers to a combination of cleaning parameters generated based on the type and degree of contamination. For example, parameters such as water volume and nozzle position can be adjusted according to the cleaning strategy to carry out differentiated treatment for different contaminants.

[0038] In one possible implementation, before determining a cleaning strategy based on multivariate pollution data when the pollution result indicates that the sensor is in a contaminated state, the method further includes: inputting the multivariate pollution data into a pollutant classification model to determine the category of pollutants on the sensor; wherein the category of pollutants includes at least dust or liquid; and determining a cleaning strategy based on the category of pollutants.

[0039] Specifically, the pollutant classification model can be a pre-trained machine learning model. Inputting diverse pollution data into the model allows it to identify the differences in the physical properties of different pollutants. For example, dust exhibits a particulate distribution, while liquids show flow marks or reflective properties, thus classifying the pollutants. This model can be a Support Vector Machine (SVM), decision tree, random forest, or a more complex deep neural network. Pollutant categories can be determined based on their morphological characteristics, such as by analyzing texture distribution, edge features, and differences in laser reflection patterns in image data. Dust pollution can manifest as discrete particle aggregation, while liquid pollution can exhibit continuous coverage or specular reflection characteristics. This enables accurate differentiation between dust and liquid pollutants, allowing cleaning strategies to be dynamically adjusted based on pollutant characteristics. For instance, heating liquid pollution can effectively avoid the problem of residual water stains on the surface caused by water spraying, while directional water spraying can quickly remove particulate matter from dust pollution.

[0040] For example, after acquiring multivariate pollution data collected by the pollution detection unit, the multivariate pollution data is input into a pollutant classification model. This model can be a deep learning-based image recognition model, trained on a large amount of labeled data. The image data is first preprocessed, including normalization and cropping, and then input into a convolutional neural network to extract features. The laser data, after noise reduction, is fused with the image features. The fused feature vector is input into a fully connected layer, ultimately outputting the probability distribution of pollutant categories. Pollutant categories include dust and liquids. Dust is further subdivided into fine dust, coarse dust, etc.; liquids are subdivided into water, oil, etc. The pollutant classification model outputs the probability of each category, for example, [P(dust), P(liquid), P(mud), P(ice / snow)]. The category with the highest probability is the identification result. Based on the identified pollutant category, a corresponding cleaning strategy is determined. For example, for dust pollution, a strategy of high-pressure airflow followed by water spraying is used; for liquid pollution, a strategy of heating and evaporation followed by wiping is used. The cleaning strategy includes specific parameter settings, such as air pressure, water volume, and heating temperature.

[0041] S140: According to the cleaning strategy, control the cleaning system of the unmanned vehicle to clean the sensors.

[0042] Specifically, the cleaning system can be a cleaning device integrating mechanical actuators, such as a multi-angle nozzle array, a high-pressure water pump, and an electric heating element. It can remove contaminants through a dual mechanism of physical rinsing and thermal action. It can accurately quantify the degree of contamination through the collaborative analysis of image and laser data, and generate adaptive cleaning parameters based on the contaminant type identification results. This effectively solves the problem of sensor failure caused by combined contamination such as dust and liquid in the mining environment, and improves the accuracy and environmental adaptability of cleaning control.

[0043] For example, the system can first continuously capture image data of the sensor surface using a camera, while the laser module emits a probe laser to the sensor surface and records the reflected signal. The contaminated area is identified based on the image data, and the degree of contamination is determined by combining the image data with a preset threshold. When the comprehensive evaluation result exceeds the set threshold, the system matches the corresponding cleaning parameters based on the current contamination characteristics.

[0044] In one possible implementation, the cleaning system of the unmanned vehicle is controlled to clean the sensors according to a cleaning strategy, including: when the type of pollutant is dust, determining the cleaning strategy as a first cleaning strategy; determining a first cleaning parameter in the first cleaning strategy based on the pollution result; wherein the first cleaning parameter includes at least the water spray volume, water spray intensity and nozzle position of the water spray module in the cleaning system; and controlling the cleaning system of the unmanned vehicle to clean the pollutants based on the first cleaning strategy.

[0045] Specifically, the first cleaning strategy refers to a cleaning plan designed for dust-type pollutants. The water spray volume can be calculated by the ratio of the average gray value of the contaminated mask to a preset threshold. For example, when the average gray value exceeds the threshold by 50%, the water spray volume is set to 0.5 ml per square centimeter. The water spray intensity is divided into three levels: low, medium, and high. The corresponding level is selected according to the degree of laser energy attenuation. When the laser energy drops by more than 60%, the high-intensity mode is activated. The nozzle position is determined by the spatial distribution of the contaminated mask. If the contaminated area is concentrated in the center area of ​​the sensor, the nozzle is adjusted to spray vertically downward at a 15-degree angle. The cleaning parameters automatically matched according to the degree of contamination can prevent equipment damage caused by over-cleaning, while ensuring the thorough removal of stubborn dust.

[0046] Specifically, in dust contamination scenarios, the dust coverage level is determined based on the average grayscale value of the contaminated mask. When the grayscale value exceeds a threshold, the cleaning parameter generation module maps the contamination level to a corresponding water spray volume level; for example, light contamination corresponds to a low water volume mode, while heavy contamination triggers a high-pressure, high-volume mode. The nozzle position control module drives the nozzles to move above the concentrated contamination area based on the spatial distribution characteristics of the contaminated mask. During the execution phase, the water pump outputs water flow of corresponding intensity according to a preset pressure curve, while the nozzles, driven by the positioning mechanism, perform reciprocating scanning motion to ensure that the water flow covers the physical area corresponding to all contaminated pixels. During the cleaning process, the water spray intensity is positively correlated with the dust adhesion intensity; for example, high-frequency pulsed water flow is used to peel off tightly adhered dust.

[0047] For example, cleaning parameters can be determined based on the PollutionScore value. For instance, if the PollutionScore is between 11 and 30 (light dust), the following procedure is performed: the central nozzle sprays water at 2 bar pressure for 0.5 seconds. If the PollutionScore is between 31 and 70 (moderate dust), the following procedure is performed: the movable nozzle sweeps water back and forth at 4 bar pressure for 1.5 seconds to ensure full coverage. If the PollutionScore is greater than 70 (heavy dust or hardened soil), a "pre-wetting-powerful rinsing" combination procedure is performed: first, soften the soil with low-pressure water spray for 0.5 seconds, wait 2 seconds, and then peel it off with water spray at the maximum pressure of 6 bar for 2.5 seconds.

[0048] In one possible implementation, the cleaning system of the unmanned vehicle is controlled to clean the sensors according to a cleaning strategy, including: when the type of contaminant is liquid, determining the cleaning strategy as a second cleaning strategy; determining a second cleaning parameter in the second cleaning strategy based on the contamination result; wherein the second cleaning parameter includes at least the heating temperature of the heating module and the water spray volume of the water spray module in the cleaning system; and controlling the cleaning system of the unmanned vehicle to clean the contaminants based on the second cleaning strategy.

[0049] Specifically, when liquid contamination is detected on the sensor surface in a mining environment, the heating module is first activated to directionally heat the contaminated area. For example, when mud-like liquid contamination is detected, the heating temperature is set to a critical value that allows water to evaporate, and the heating time is automatically matched according to the contaminated area. After the initial evaporation treatment is completed, the water spray module rinses the sensor surface with a specific amount of water, which is dynamically adjusted according to the adhesion strength of the residual contaminants.

[0050] For example, when the category is "ice and snow," the contamination score is 40 points: the system will initiate a second cleaning strategy. First, the heating module is activated, raising the sensor surface temperature to 15°C and maintaining it for 45 seconds to melt the ice and snow into water; then, water is sprayed briefly at a low pressure of 1 bar for 0.5 seconds to blow away the melted water droplets and prevent residual water stains. If the category is "mud," the contamination score is 80 points: the mud can be briefly heated at 30°C to reduce its viscosity, and then rinsed using a high-pressure water spray program. If the category is "rainwater," the contamination score is 20 points: the rainwater can be removed by heating with the heating module. Heating promotes the evaporation of liquid contaminants, reduces their adhesion, and improves the subsequent washing effect. By adjusting the heating temperature and water spray volume, different levels of liquid contamination can be accommodated, effectively removing liquid contaminants while reducing water waste and improving cleaning efficiency.

[0051] In one possible implementation, after the cleaning system of the unmanned vehicle cleans the sensor, the method further includes: controlling the image acquisition module to acquire target image data of the sensor; inputting the target image data into the image recognition model to obtain the recognition result; using the recognition result to indicate whether the sensor has been cleaned successfully; if the recognition result indicates that the sensor has not been cleaned successfully, adjusting the cleaning parameters in the cleaning strategy, and controlling the cleaning system of the unmanned vehicle to clean the sensor according to the adjusted cleaning parameters.

[0052] Specifically, the target image data refers to the sensor surface image data re-captured by the image acquisition module after the cleaning operation, reflecting the actual cleanliness status after cleaning. The image recognition model can be a pre-trained convolutional neural network model. After the initial cleaning operation, the image acquisition module immediately acquires images of the sensor surface. The acquired target image data is input into the pre-trained image recognition model, which analyzes the texture features and color distribution in the image to determine whether there are residual contaminants. If the model output indicates that the cleaning is not up to standard, the system will automatically adjust the cleaning parameters according to the current degree of residual contamination, such as increasing the water spray volume by 20% or extending the heating time by 30 seconds, and then start the secondary cleaning process. This process can be repeated until the image recognition model confirms that the cleanliness meets the standard.

[0053] This application discloses a sensor cleaning control method for an unmanned vehicle. The unmanned vehicle's sensors are equipped with a contamination monitoring unit. The method includes: acquiring multivariate contamination data collected by the contamination detection unit; wherein the contamination detection unit includes at least an image acquisition module and a laser detection module; the multivariate contamination data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; analyzing the image data and laser data to obtain a contamination result for the sensor, the contamination result indicating the degree of contamination of the sensor; when the contamination result indicates that the sensor is in a contaminated state, determining a cleaning strategy based on the multivariate contamination data; and controlling the unmanned vehicle's cleaning system to clean the sensor according to the cleaning strategy, which can accurately identify the type and degree of sensor contamination and adopt targeted cleaning strategies to improve cleaning efficiency and cleaning capability.

[0054] Corresponding to the aforementioned application function implementation method embodiments, this application also provides a sensor cleaning control device, electronic device, and corresponding embodiments for unmanned vehicles.

[0055] Figure 2 This is a schematic diagram of the sensor cleaning control device for an unmanned vehicle shown in an embodiment of this application.

[0056] See Figure 2 A sensor cleaning control device for an unmanned vehicle, comprising: The acquisition module 210 is used to acquire multivariate pollution data collected by the pollution detection unit; wherein the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module.

[0057] The analysis module 220 is used to analyze image data and laser data to obtain the contamination results of the sensor, which are used to indicate the degree of contamination of the sensor.

[0058] The determination module 230 is used to determine a cleaning strategy based on multivariate pollution data when the pollution result indicates that the sensor is in a polluted state.

[0059] The cleaning module 240 is used to control the cleaning system of the unmanned vehicle to clean the sensors according to the cleaning strategy.

[0060] In one possible implementation, the analysis module 220 is further configured to: 1) convert the laser data, including the laser angle and laser energy of the reflected laser, to grayscale to obtain grayscale image data; 2) identify contaminated pixels in the grayscale image data whose grayscale values ​​exceed a preset grayscale value, and generate a contaminated mask based on the contaminated pixels; 3) calculate the average grayscale value of the contaminated mask based on the grayscale values ​​of the contaminated pixels in the contaminated mask; and 4) obtain the contamination result of the sensor based on the average grayscale value of the contaminated mask, the laser angle, and the laser energy.

[0061] In one possible implementation, the analysis module 220 is further configured to input multivariate pollution data into a pollutant classification model to determine the category of pollutants on the sensor; wherein the category of pollutants includes at least dust or liquid. Based on the category of pollutants, a cleaning strategy is determined.

[0062] In one possible implementation, the cleaning module 240 is further configured to determine the cleaning strategy as a first cleaning strategy when the type of pollutant is dust; determine the first cleaning parameters in the first cleaning strategy based on the pollution result; wherein the first cleaning parameters include at least the water spray volume, water spray intensity and nozzle position of the water spray module in the cleaning system; and control the cleaning system of the unmanned vehicle to clean the pollutants based on the first cleaning strategy.

[0063] In one possible implementation, the cleaning module 240 is further configured to determine a second cleaning strategy when the type of contaminant is liquid; determine a second cleaning parameter in the second cleaning strategy based on the contamination result; wherein the second cleaning parameter includes at least the heating temperature of the heating module and the water spray volume of the spray module in the cleaning system; and control the cleaning system of the unmanned vehicle to clean the contaminant based on the second cleaning strategy.

[0064] In one possible implementation, the analysis module 220 is further configured to control the image acquisition module to acquire target image data from the sensor; input the target image data into the image recognition model to obtain a recognition result; and use the recognition result to indicate whether the sensor has been successfully cleaned. If the recognition result indicates that the sensor has not been successfully cleaned, the cleaning parameters in the cleaning strategy are adjusted, and the cleaning system of the unmanned vehicle is controlled to clean the sensor according to the adjusted cleaning parameters.

[0065] This application discloses a sensor cleaning control device for an unmanned vehicle. The unmanned vehicle's sensors are equipped with a contamination monitoring unit, comprising: an acquisition module for acquiring multivariate contamination data collected by the contamination detection unit; wherein the contamination detection unit includes at least an image acquisition module and a laser detection module; the multivariate contamination data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; an analysis module for analyzing the image data and laser data to obtain the sensor's contamination result, which indicates the degree of contamination of the sensor; a determination module for determining a cleaning strategy based on the multivariate contamination data when the contamination result indicates that the sensor is in a contaminated state; and a cleaning module for controlling the unmanned vehicle's cleaning system to clean the sensors according to the cleaning strategy, accurately identifying the type and degree of sensor contamination and adopting targeted cleaning strategies to improve cleaning efficiency and cleaning capability.

[0066] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0067] This application also provides an electronic device. Figure 3 This is a schematic diagram of the hardware structure of an embodiment of the electronic device of this application. The electronic device includes a memory 320 and at least one processor 310. The memory 320 is electrically connected to the at least one processor 310. The memory 320 stores instructions. The at least one processor 310 calls the instructions in the memory 320 to cause the electronic device to execute the sensor cleaning control method for an unmanned vehicle according to any of the foregoing embodiments of this application.

[0068] Specifically, the processor 310 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0069] Memory 320 may include a mass storage device for data or instructions. For example, and not limitingly, memory 320 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 320 may include removable or non-removable (or fixed) media. Where appropriate, memory 320 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 320 is non-volatile solid-state memory. In a particular embodiment, memory 320 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0070] In one example, the control device may also include a communication interface 330 and a bus 340. The processor 310, memory 320, and communication interface 330 are connected via the bus 340 and communicate with each other.

[0071] The communication interface 330 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0072] Bus 340 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 340 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0073] Furthermore, in conjunction with the sensor cleaning control method for unmanned vehicles in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. This computer-readable storage medium stores executable code, which, when executed by a processor, implements any of the sensor cleaning control methods for unmanned vehicles in the above embodiments.

[0074] This application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0075] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0076] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0077] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A sensor cleaning control method for an unmanned vehicle, characterized in that, The unmanned vehicle is equipped with a pollution monitoring unit on its sensors, and the method includes: The pollution detection unit acquires multivariate pollution data collected by the pollution detection unit; wherein the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module. The image data and the laser data are analyzed to obtain the contamination result of the sensor, which is used to indicate the degree of contamination of the sensor; When the pollution result indicates that the sensor is in a polluted state, a cleaning strategy is determined based on the multivariate pollution data; According to the cleaning strategy, the cleaning system of the unmanned vehicle is controlled to clean the sensors.

2. The method according to claim 1, characterized in that, The laser data includes the laser angle and laser energy of the reflected laser; the step of analyzing the image data and the laser data to obtain the contamination result of the sensor includes: The image data is converted to grayscale to obtain grayscale image data; The contaminated pixels with grayscale values ​​exceeding a preset grayscale value are identified from the grayscale image data, and a contaminated mask is generated based on the contaminated pixels; The average gray value of the contaminated mask is calculated based on the gray values ​​of the contaminated pixels in the contaminated mask. The contamination result of the sensor is obtained based on the average gray value of the contamination mask, the laser angle, and the laser energy.

3. The method according to claim 2, characterized in that, Before determining a cleaning strategy based on the multivariate contamination data when the contamination result indicates that the sensor is in a contaminated state, the method further includes: The multivariate pollution data is input into a pollutant classification model to determine the category of pollutants on the sensor; wherein the category of pollutants includes at least dust or liquid; The cleaning strategy is determined based on the type of contaminant.

4. The method according to claim 3, characterized in that, The step of controlling the cleaning system of the unmanned vehicle to clean the sensors according to the cleaning strategy includes: When the type of pollutant is dust, the cleaning strategy is determined to be the first cleaning strategy; Based on the pollution results, a first cleaning parameter is determined in the first cleaning strategy; wherein the first cleaning parameter includes at least the water spray volume, water spray intensity, and nozzle position of the water spray module in the cleaning system; Based on the first cleaning strategy, the cleaning system of the unmanned vehicle is controlled to clean the pollutants.

5. The method according to claim 3, characterized in that, The step of controlling the cleaning system of the unmanned vehicle to clean the sensors according to the cleaning strategy includes: When the type of contaminant is liquid, the cleaning strategy is determined to be the second cleaning strategy; Based on the pollution results, a second cleaning parameter is determined in the second cleaning strategy; wherein the second cleaning parameter includes at least the heating temperature of the heating module in the cleaning system and the water spray volume of the water spray module; Based on the second cleaning strategy, the cleaning system of the unmanned vehicle is controlled to clean the pollutants.

6. The method according to claim 4 or 5, characterized in that, After the cleaning system of the unmanned vehicle cleans the sensors, the process further includes: The image acquisition module is controlled to acquire target image data from the sensor; The target image data is input into an image recognition model to obtain a recognition result; the recognition result is used to indicate whether the sensor has successfully cleaned the image. If the identification result indicates that the sensor has not been cleaned successfully, the cleaning parameters in the cleaning strategy are adjusted, and the cleaning system of the unmanned vehicle is controlled to clean the sensor according to the adjusted cleaning parameters.

7. A sensor cleaning control device for an unmanned vehicle, characterized in that, include: An acquisition module is used to acquire multivariate pollution data collected by the pollution detection unit; wherein, the pollution detection unit includes at least an image acquisition module and a laser detection module; the multivariate pollution data includes image data acquired by the image acquisition module and laser data acquired by the laser detection module; An analysis module is used to analyze the image data and the laser data to obtain the contamination result of the sensor, and the contamination result is used to indicate the degree of contamination of the sensor; The determination module is used to determine a cleaning strategy based on the multivariate pollution data when the pollution result indicates that the sensor is in a polluted state; The cleaning module is used to control the cleaning system of the unmanned vehicle to clean the sensors according to the cleaning strategy.

8. A computer program product comprising a computer program / instructions, characterized in that, The When a computer program / instruction is executed by a processor, it implements the method of any one of claims 1-6.

9. An electronic device, characterized in that, include: processor; as well as A memory having executable code stored thereon, which, when executed by the processor, causes the processor to perform the method as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, It stores executable code that, when executed by a processor of an electronic device, causes the processor to perform the method as described in any one of claims 1-6.

Images