Test method, system, electronic device and storage medium for vehicle cleaning device

Abstract

The application provides a test method, system, electronic device and storage medium of a vehicle cleaning device, the method comprising: simulating a dirty condition and an environmental parameter for a perception component according to a test requirement corresponding to the cleaning device to be tested, and driving the cleaning device to perform a cleaning operation on the perception component in the dirty state under the environmental parameter; during and after the cleaning operation performed by the cleaning device, collecting performance evaluation parameters from the perception component and the cleaning device; evaluating the performance evaluation parameters based on at least one evaluation dimension to obtain evaluation results of the performance evaluation parameters, and fusing the evaluation results to generate a test result of the cleaning device, and the evaluation dimension comprises at least one of a cleaning performance dimension, an operation performance dimension and an extreme performance dimension. The application can solve the problems of uncontrollable test conditions, low evaluation efficiency and single evaluation dimension of the current vehicle cleaning device, and realize standardized and efficient performance testing.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a testing method, system, electronic device, and storage medium for a vehicle cleaning device. Background Technology

[0002] With the rapid development of Advanced Driver Assistance Systems (ADAS) and autonomous driving technologies, onboard intelligent driving sensors (cameras, LiDAR, millimeter-wave radar, etc.) have become core components of environmental perception systems. Their cleanliness directly determines key functions such as image recognition accuracy and ranging accuracy, thus affecting the safety and reliability of the intelligent driving system. In actual driving, dust, mud, insects, and frost can easily form stubborn contaminants on sensor surfaces, leading to blurred images, ranging deviations, and even functional failures. Therefore, the decontamination performance of cleaning devices for these sensors has become an important indicator of the all-weather availability of intelligent driving systems.

[0003] However, current methods for evaluating the performance of cleaning devices generally suffer from problems such as uncontrollable testing conditions, low evaluation efficiency, and limited evaluation dimensions, making it difficult to objectively and efficiently verify the actual cleaning capabilities of the devices. Summary of the Invention

[0004] One of the purposes of this application is to provide a testing method for vehicle cleaning devices to solve the problems of uncontrollable testing conditions, low evaluation efficiency, and single evaluation dimensions in current vehicle cleaning device testing. This method aims to achieve standardized and efficient performance testing, comprehensively measure the cleaning performance of various vehicle cleaning devices, and thus ensure the availability of various sensors in the vehicle and maintain the effectiveness of the vehicle's intelligent driving system.

[0005] To achieve the above objectives, the technical solution adopted in this application is as follows: The first aspect of this application provides a testing method for a vehicle cleaning device, the method comprising: Based on the test requirements of the cleaning device to be tested, simulate dirt conditions and environmental parameters for the sensing component, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. During and after the cleaning operation is performed by the cleaning device, performance evaluation parameters are collected from the sensing components and the cleaning device. The performance evaluation parameters are evaluated based on at least one evaluation dimension to obtain the evaluation results of each performance evaluation parameter. The evaluation results are then fused to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0006] Optionally, the evaluation of the performance evaluation parameters based on at least one evaluation dimension to obtain evaluation results for each performance evaluation parameter, and the fusion of the evaluation results to generate test results for the cleaning device, includes: The basic weight of each performance evaluation parameter is determined based on the evaluation dimension corresponding to each performance evaluation parameter. Based on the test scenario for the cleaning device and the type of the sensing component, the correction coefficients corresponding to each of the basic weights are determined, and the correction coefficients are used to correct the corresponding basic weights to obtain each corrected weight. After preprocessing each of the performance evaluation parameters, a single score for each of the performance evaluation parameters is obtained based on the preprocessed performance evaluation parameters and their corresponding preset baselines. Based on the corrected weights, a weighted fusion is performed on each of the individual scores to generate a comprehensive score for the cleaning device, thereby obtaining test results including the comprehensive score.

[0007] Optionally, the method further includes: A multivariate evaluation chart is generated based on the individual scores of each of the aforementioned performance evaluation parameters. The multivariate evaluation chart is used to indicate the performance status of the cleaning device on each performance evaluation parameter. The test results also include the multivariate evaluation graph.

[0008] Optionally, the method further includes: Calculate the degree of deviation between each of the performance evaluation parameters and the corresponding preset baseline; Based on the performance evaluation parameters whose deviation exceeds a preset threshold, the corresponding performance indicators to be optimized are determined; Based on the performance indicators to be optimized, optimization guidance information is generated for the cleaning device.

[0009] Optionally, simulating dirt conditions and environmental parameters for the sensing component according to the testing requirements corresponding to the cleaning device under test includes: Obtain the test requirements corresponding to the cleaning device, including the test scenario and the sensing components to be cleaned; The sensing component is simulated with dirty conditions and environmental parameters based on the test scenario.

[0010] Optionally, simulating dirt conditions for the sensing component based on the test scenario includes: Obtain the preset dirt and physical release conditions corresponding to the test scenario; The preset dirt is applied to the sensing component according to the physical application conditions.

[0011] Optionally, after applying the preset dirt to the sensing component according to the physical application conditions, the method further includes: When the preset dirt meets the morphology adjustment conditions, the ambient temperature is adjusted to a preset temperature range and maintained for a preset duration, so that the preset dirt adheres to the surface of the sensing component in a specific form.

[0012] Optionally, obtaining the test requirements corresponding to the cleaning device includes: In response to a user command, the sensor components that need to be cleaned and the test scenario corresponding to the cleaning of the sensor components are determined, the test scenario including weather and driving speed range; The weather and the driving speed range are used to simulate dirt conditions and environmental parameters for the sensing component.

[0013] Optionally, the collection of performance evaluation parameters from the cleaning device includes: Obtain the device type of the cleaning device; Based on the device type of the cleaning device, the corresponding performance evaluation parameters are determined, and the determined performance evaluation parameters are collected from the cleaning device.

[0014] A second aspect of this application provides a testing system for a vehicle cleaning device, the testing system comprising: The simulation drive component is used to simulate dirt conditions and environmental parameters for the sensing component according to the test requirements of the cleaning device to be tested, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. An evaluation acquisition component is used to acquire performance evaluation parameters from the sensing component and the cleaning device during and after the cleaning operation is performed by the cleaning device. The result generation component is used to evaluate the performance evaluation parameters based on at least one evaluation dimension, obtain the evaluation results of each performance evaluation parameter, and fuse the evaluation results to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0015] A third aspect of this application provides an electronic device including a processor, a memory, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the steps of the test method for a vehicle cleaning device as described in the first aspect of this application.

[0016] A fourth aspect of this application provides a readable storage medium storing a computer program that, when executed by a processor, implements the steps of the test method for the vehicle cleaning device as described in the first aspect of this application.

[0017] The testing method provided in this application first obtains the user's testing requirements for the cleaning device. Different testing requirements correspond to different dirt conditions and environmental parameters under which the cleaning device operates. Then, based on the testing requirements, dirt conditions and environmental parameters are simulated for the sensing component that needs cleaning. This standardizes the dirt conditions that the sensing component may experience, avoiding inconsistencies in testing conditions caused by human intervention. Next, the cleaning device under test is driven to clean the sensing component in a dirty state. During and after the cleaning process, corresponding performance evaluation parameters are collected from the sensing component and the cleaning device. Test results for the cleaning device are generated based on these performance evaluation parameters. The collected performance evaluation parameters correspond to at least one evaluation dimension. By associating the performance evaluation parameters with the corresponding dimensions, the cleaning device is evaluated on at least one dimension, generating more comprehensive and objective test results and enhancing their reference value.

[0018] By adopting the testing scheme of this application, no manual intervention is required to apply dirt or collect / analyze test results, thus maintaining the uniformity of testing conditions and avoiding errors in test results caused by fluctuations in testing conditions. Furthermore, the automated processing method improves the efficiency of test result generation. Simultaneously, by conducting a multi-dimensional comprehensive evaluation of the cleaning device, the test results become more comprehensive. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of the steps of a testing method provided in an embodiment of this application; Figure 2 This is a structural block diagram of a testing system provided in an embodiment of this application; Figure 3 This is a schematic diagram of the architecture of a testing system provided in an embodiment of this application; Figure 4 This is a schematic diagram illustrating the implementation steps of a testing system provided in an embodiment of this application. Figure 5 This is a spider web diagram provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] The embodiments of this application will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be understood that the preferred embodiments are only for illustrating this application and are not intended to limit the scope of protection of this application.

[0022] With the evolution of intelligent driving technology, vehicles are increasingly reliant on multiple sensing components. The detection accuracy of sensing components such as cameras and lidar directly determines the reliability of intelligent driving system decisions. However, in complex environments such as rain, snow, mud, and dust, the surfaces of sensing components are susceptible to foreign object adhesion, leading to field-of-view obstruction, increased point cloud noise, and inaccurate target recognition, severely weakening the robustness and safety of the intelligent driving system. Therefore, reliable sensing component cleaning devices have become a key technical requirement for ensuring the stable operation of intelligent driving functions.

[0023] However, current testing methods for evaluating the performance of cleaning devices generally suffer from the following problems: the application of contaminants lacks standardization, resulting in poor comparability of results between different test batches; the evaluation indicators only focus on whether the dirt is removed, without considering other dimensions; and the entire evaluation process relies on manual spraying, visual observation, and manual recording and analysis, which takes too long and is inefficient.

[0024] In response, this application provides a testing method for a vehicle cleaning device to solve one or more of the above-mentioned problems.

[0025] Reference Figure 1 The flowchart illustrates the steps of a testing method for a vehicle cleaning device provided in an embodiment of this application. The method may include: Step 101: Based on the test requirements of the cleaning device to be tested, simulate dirt conditions and environmental parameters for the sensing component, and drive the cleaning device to perform a cleaning operation on the sensing component in a dirty state under the environmental parameters. Step 102: During and after the cleaning operation performed by the cleaning device, performance evaluation parameters are collected from the sensing component and the cleaning device. Step 103: Evaluate the performance evaluation parameters based on at least one evaluation dimension to obtain the evaluation results of each performance evaluation parameter, and fuse the evaluation results to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0026] The testing method for a vehicle cleaning device provided in this application targets a cleaning device that cleans vehicle sensing components. During the testing phase, the cleaning performance of the device is evaluated using a multi-dimensional, standardized approach to quantify its cleaning capability. Furthermore, by testing the cleaning performance of the device, a qualified cleaning device can effectively and promptly remove dirt from the surface of the sensing components, thereby maintaining the availability of the vehicle sensing components. In other words, the testing method provided in this application can indirectly maintain the all-weather availability of the vehicle sensing components and protect the vehicle's intelligent driving functions. In some specific implementations, the testing method proposed in this application places the vehicle with the cleaning device deployed inside a functional compartment or at a vertical device location during testing. The functional compartment or vertical device provides the necessary environment for testing the cleaning device.

[0027] For step 101, according to the test requirements corresponding to the cleaning device to be tested, simulate dirt conditions and environmental parameters for the sensing component, and drive the cleaning device to perform a cleaning operation on the sensing component in a dirty state under the environmental parameters.

[0028] In this embodiment, the system supports receiving test requirements for the cleaning device to be tested through channels such as a human-computer interaction interface. Subsequently, relevant information from the obtained test requirements (such as the test scenario and the sensing components to be cleaned) is transmitted to subsequent execution steps via Ethernet communication or other means.

[0029] It is understood that the cleaning device is, for example, a portable cleaning device, which is installed directly near the sensing component to clean the sensing component in real time.

[0030] Furthermore, the sensing components in this application are parts installed in the vehicle to collect environmental information to maintain the vehicle's intelligent driving functions, such as cameras and lidar. That is, the cleaning device needs to perform cleaning on the target.

[0031] Among them, environmental parameters refer to the environmental parameters in the working environment where the cleaning device actually performs cleaning work, such as temperature, humidity, wind speed, and precipitation.

[0032] When the sensing component has been simulated for dirt conditions and environmental parameters, it means that the sensing component is currently in a dirty state and meets the actual working environment of the cleaning device. At this time, the cleaning device can perform cleaning work to evaluate its cleaning performance.

[0033] In a practical implementation, the entity running the current test method may have a communication connection with the cleaning device, thereby driving the cleaning device to perform the cleaning operation. Alternatively, it may display feedback indicating that the cleaning device is ready to perform the cleaning operation, prompting the user to turn on the cleaning device, etc.

[0034] For step 102, performance evaluation parameters are collected from the sensing component and the cleaning device during and after the cleaning operation is performed by the cleaning device.

[0035] For example, in this embodiment of the application, a comprehensive evaluation of the cleaning device will be conducted, including the cleaning effect and the operating status of the cleaning device (such as energy consumption, noise, etc.). Therefore, performance evaluation parameters will be collected from both the sensing component and the cleaning device to conduct a multi-dimensional comprehensive evaluation. At this time, the performance evaluation parameters collected from the sensing component are used to reflect the cleaning effect of the cleaning device.

[0036] Optionally, the performance evaluation parameters collected from the sensing component can be tailored to the type of the sensing component. For example, when the sensing component is a camera, its light transmittance can be collected as a performance evaluation parameter, while when the sensing component is a radar, its ranging accuracy can be collected as a performance evaluation parameter. The type of sensing component can be determined based on the testing requirements in step 101.

[0037] Subsequently, step 103 is used to evaluate the performance evaluation parameters based on at least one evaluation dimension, obtain the evaluation results of each performance evaluation parameter, and perform weighted fusion of each evaluation result to generate the test results of the cleaning device. The evaluation dimension includes at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0038] The cleaning performance evaluation dimension specifically evaluates the cleaning effect of the cleaning device, such as residual dirt after cleaning. These parameters can be performance evaluation parameters collected after the cleaning operation, such as those collected by the sensing components or calculated based on data collected from the sensing components and the cleaning device. The operational performance dimension specifically evaluates the user experience of the cleaning device, such as noise and energy consumption during use. These parameters can be performance evaluation parameters collected from the cleaning device during the cleaning operation. The extreme performance dimension specifically evaluates the long-term stability of the cleaning device, including performance degradation after thousands of uses and working time in extreme environments. These parameters can be performance evaluation parameters collected after a single cleaning operation, combined with historical parameters for a comprehensive judgment. These parameters are collected from the sensing components and are used to evaluate the durability and extreme response capability of the cleaning device. For example, the working time of the cleaning device in extreme environments can be determined by collecting the time taken for the sensing components to recover from contamination to normal output in extreme scenarios.

[0039] After obtaining the evaluation results of the performance evaluation parameters, the results will be merged to generate the final test results. The evaluation results are obtained by measuring the performance evaluation parameters against preset standards. In practice, the merging of evaluation results can be achieved through methods including, but not limited to, weighted merging, arithmetic average, extreme value merging, and ordinal value merging. The weights for weighted merging of the evaluation results can be determined by referring to the evaluation dimension to which the performance evaluation parameters belong.

[0040] It is understood that this application involves multiple evaluation dimensions, and the performance evaluation parameters collected for the cleaning operations performed by the cleaning device belong to at least one of these dimensions. In most cases, performance evaluation parameters from two or more evaluation dimensions will be collected to obtain the test results for the cleaning device. Therefore, the cleaning device can be evaluated from multiple perspectives, resulting in more comprehensive test results.

[0041] The testing method for a vehicle cleaning device provided in this application first obtains the user's testing requirements for the cleaning device. Different testing requirements correspond to different dirt conditions and environmental parameters under which the cleaning device operates. Then, based on the testing requirements, dirt conditions and environmental parameters are simulated for the sensing component that needs cleaning. This standardizes the dirt conditions that the sensing component may experience, avoiding inconsistencies in testing conditions caused by human intervention. Next, the cleaning device under test is driven to clean the dirty sensing component. During and after the cleaning process, corresponding performance evaluation parameters are collected from the sensing component and the cleaning device. Test results for the cleaning device are generated based on these performance evaluation parameters. The collected performance evaluation parameters correspond to at least one evaluation dimension. By associating the performance evaluation parameters with the corresponding dimensions, the cleaning device is evaluated on at least one dimension, generating more comprehensive and objective test results and enhancing the reference value of the test results.

[0042] By adopting the testing scheme of this application, no manual intervention is required to apply dirt or collect / analyze test results, thus maintaining the uniformity of testing conditions and avoiding errors in test results caused by fluctuations in testing conditions. Furthermore, the automated processing method improves the efficiency of test result generation. Simultaneously, by conducting a multi-dimensional comprehensive evaluation of the cleaning device, the test results become more comprehensive.

[0043] In one embodiment of this application, step 101, which simulates dirt conditions and environmental parameters for the sensing component according to the testing requirements of the cleaning device to be tested, may include: S01, Obtain the test requirements corresponding to the cleaning device, the test requirements including the test scenario and the sensing components to be cleaned; S02, simulate dirt conditions and environmental parameters for the sensing component according to the test scenario.

[0044] It is understood that the test scenario in this application embodiment refers to the environment in which the sensing component to be cleaned is located, that is, the environment in which the current vehicle (the vehicle on which the sensing component is mounted) is located. The current test scenario is also the working environment of the cleaning device. For example, the test scenario includes environmental parameters such as temperature, humidity, wind speed, and precipitation.

[0045] In practice, the system can receive a simple combination of conditions selected by the user as a test requirement, and retrieve the environmental parameters such as temperature, humidity, and wind speed corresponding to the combination of conditions from the local database (or online database), thereby reducing the complexity of the test requirements provided by the user and eliminating the need for accurate input from the user.

[0046] Subsequently, based on the test scenario, the sensing component was subjected to simulated dirt conditions and environmental parameters; In this embodiment, the test scenario includes not only environmental parameters, such as the aforementioned temperature, humidity, wind speed, and precipitation, but also the existing dirt conditions within the current environmental parameters. The dirt conditions determine the specific dirt adhering to the surface of the sensing component, while the environmental parameters affect the adhesion pattern of the dirt on the sensing component surface and the working environment of the cleaning device. Both factors together determine the cleaning difficulty faced by the cleaning device in actual scenarios.

[0047] For example, in sunny driving scenarios, the dirt conditions encountered by the sensing components, such as dust and insect residue, tend to dry and solidify under high temperatures. In rainy driving scenarios, the dirt conditions encountered by the sensing components, such as rainwater and mud, tend to exhibit uneven, patchy adhesion at high speeds. In different driving scenarios, each sensing component corresponds to a different level of dirt, meaning the cleaning device needs to handle different types of dirt. To address this, the testing method provided in this application sets relevant data on dirt conditions for environmental parameters representing different driving scenarios. This allows the testing of the cleaning device to simulate not only its operating environment but also the dirt conditions of the sensing components it needs to handle in that environment.

[0048] Therefore, the relevant steps in this application do not require manual intervention (such as manually smearing or splashing the corresponding dirt onto the sensing components), supporting the provision of complete and standardized dirt conditions and environmental parameters for the cleaning device. Since no manual intervention is required, the automated simulation process can not only provide the working scenario and target of the cleaning device more quickly; the standardized dirt conditions and environmental parameters can also avoid deviations caused by inconsistencies in initial states between multiple rounds of testing, improving the reliability of the test results.

[0049] Furthermore, since the sensing component has already been selected in the test requirements in step S01, in the subsequent steps of step 101, the cleaning device simply performs a cleaning operation on the selected sensing component. Simultaneously, it supports identifying the type of the current sensing component in the instruction for selecting the sensing component, so that corresponding evaluation parameters can be collected for that sensing component type later; alternatively, different sensing components carry their corresponding type identifiers, and after selecting a sensing component, the corresponding type identifier will be obtained to confirm the type of the sensing component.

[0050] In this example, the test scenario and the sensing components to be cleaned are obtained from the test requirements. The test scenario includes the dirt conditions that the cleaning device needs to handle and the environmental parameters during the cleaning device's operation. The test scenario comprehensively considers the working conditions of the cleaning device, including the object being cleaned (dirt conditions) and the working environment (represented by environmental parameters). The parameters in the test scenario support a standardized and comprehensive simulation of the cleaning device's working state, which can improve the accuracy of the test results for the cleaning device.

[0051] In one embodiment of this application, the test requirements corresponding to the cleaning device may include: In response to a user command, the sensor components that need to be cleaned and the test scenario corresponding to the cleaning of the sensor components are determined, the test scenario including weather and driving speed range; The weather and the driving speed range are used to simulate dirt conditions and environmental parameters for the sensing component.

[0052] That is, the current sensing component that needs to be cleaned and its corresponding test scenario can be obtained by receiving a combination of conditions contained in the user command, or by directly receiving the user command. The test scenarios in this embodiment include weather (such as sunny high temperature, sunny low temperature; light rain, moderate rain, heavy rain, etc.) and vehicle driving speed range (such as static, low speed dynamic, medium speed dynamic, and high speed dynamic, etc.), which are used to jointly confirm the simulated dirt conditions and environmental parameters that need to be simulated.

[0053] For example, the correspondence between the test scenario (weather and driving speed range) and the dirt conditions and environmental parameters is shown in Table 1: Table 1:

[0054] In this context, RH stands for Relative Humidity, which is the practical unit used in the embodiments of this application. The dynamic full-scene refers to low-speed dynamic, medium-speed dynamic, and high-speed dynamic.

[0055] It is understandable that different driving speed ranges will affect the wind speed in the environmental parameters. For example, in a sunny, normal-temperature driving scenario, low-speed dynamic, medium-speed dynamic, and high-speed dynamic can respectively correspond to adjustable wind speeds of 0-10 m / s, 10-20 m / s, and 20-30 m / s in the environmental parameters, and so on. Furthermore, the correspondence between different driving speed ranges and wind speed, as well as the wind speed itself, can be adjusted according to user needs. To avoid redundancy, this is not shown in detail in Table 1.

[0056] In some specific implementations, users can click to select the weather and driving speed range (i.e., select the test scenario) in the visualization interface. Then, according to the correspondence shown in Table 1, the corresponding environmental parameters and dirt conditions will be retrieved to simulate dirt conditions and environmental parameters for the sensing component based on the test scenario (weather and driving speed range).

[0057] In addition, the sensing components that need cleaning can also be directly carried in user commands, or confirmed by the user through the selection of commands on a visual interactive device.

[0058] This embodiment provides a method for confirming test scenarios. Users can obtain detailed scenarios in the current test method by providing only simple and intuitive weather and driving speed ranges, reducing user operation complexity and improving user experience.

[0059] In one embodiment of this application, the evaluation of the performance evaluation parameters based on at least one evaluation dimension to obtain evaluation results for each performance evaluation parameter, and the fusion of the evaluation results to generate the test results of the cleaning device, may specifically include: S11, Determine the basic weight of each performance evaluation parameter according to the evaluation dimension corresponding to each performance evaluation parameter; S12, based on the test scenario for the cleaning device and the type of the sensing component, determine the correction coefficient corresponding to each of the basic weights and use the correction coefficient to correct the corresponding basic weights to obtain each corrected weight; S13, after preprocessing each of the performance evaluation parameters, obtain the individual score of each of the performance evaluation parameters based on the preprocessed performance evaluation parameters and their corresponding preset baselines; S14, perform weighted fusion on each of the individual scores based on the corrected weights to generate a comprehensive score for the cleaning device, so as to obtain test results including the comprehensive score.

[0060] That is, based on the performance requirements of the cleaning device in the intelligent driving scenario, different weights are applied to the evaluation results of the collected performance evaluation parameters, and the results are weighted and fused to obtain the final comprehensive score, making the comprehensive score more consistent with the working state of the cleaning device in the intelligent driving scenario. In other embodiments, the arithmetic mean of each evaluation result can be directly calculated to obtain the final comprehensive score, or the evaluation results can be sorted according to their numerical values, and then each evaluation result can be assigned a score or weight according to the sorting, and the final comprehensive score can be obtained based on the sum of the assigned scores or the weighted sum of the assigned weights.

[0061] Specifically, taking weighted fusion as an example, the basic weight of each performance evaluation parameter can be determined first based on the evaluation dimension to which the performance evaluation parameter belongs. For example, the basic weight of the cleaning performance dimension > the basic weight of the operational performance dimension > the basic weight of the extreme performance dimension. That is, different evaluation dimensions have different impacts on the comprehensive evaluation, and even in multi-dimensional evaluation, the cleaning performance dimension is still the most important evaluation dimension.

[0062] Optionally, the base weight for the cleaning performance dimension is 0.60, the base weight for the extreme performance dimension is 0.25, and the base weight for the operational performance dimension is 0.15.

[0063] Furthermore, based on different test scenarios and the types of perception components, correction coefficients are obtained for the basic weights corresponding to the performance evaluation parameters. That is, the specific operating environments of the cleaning device and perception components may present different intelligent driving requirements. Therefore, adjustments are made based on the test scenario and the type of perception component. For example, in scenarios where the vehicle travels at slow speeds (e.g., when the perception component is a camera), the weights of the cleaning performance and operational performance dimensions need to be appropriately increased. In this case, intelligent driving focuses on panoramic views such as parking, requiring assurance of cleaning effectiveness and the user's experience inside the vehicle; it does not prioritize characteristics related to extreme performance dimensions, thus requiring a decrease in the corresponding weights of extreme performance dimensions.

[0064] For example, in some specific implementations, the corresponding relationship of the correction coefficients is confirmed according to the test scenario and the type of perception component, as shown in Table 2: Table 2:

[0065] It is understandable that the "dynamic" and "static" in Table 2 refer to the dynamic full-scenario and static scenarios within the driving speed range, and the first three examples do not involve weather conditions. A correction factor is set. This avoids significant adjustments to the basic weights.

[0066] After obtaining the correction coefficients corresponding to the current performance evaluation parameters, the corresponding basic weights are adjusted using the correction coefficients to obtain the final corrected weights used for comprehensive calculation. For example, if the collected performance evaluation parameters M1, M2, and M3 belong to the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension, respectively; and the current test scenario and the type of sensing component are static + camera, then the original basic weights of performance evaluation parameters M1, M2, and M3 (0.60, 0.15, and 0.25) will be adjusted by coefficients of 1.1, 1.1, and 0.8. Therefore, the corrected weight for M1 is 0.6 × 1.1 = 0.66, the corrected weight for M2 is 0.15 × 1.1 = 0.165, and the corrected weight for M3 is 0.25 × 0.8 = 0.2.

[0067] The performance evaluation parameters can be processed rather than directly collected. For example, if the required performance evaluation parameter is the dirt residue rate under the cleaning performance dimension, directly collecting the area of ​​the dirty area cannot directly participate in the calculation of the comprehensive score. It needs to be converted into a single performance evaluation parameter of dirt residue rate first, then evaluated, and the evaluation result will participate in the final weighted fusion to obtain the comprehensive score.

[0068] For example, preprocessing of each performance evaluation parameter, such as denoising and / or averaging of multiple detection results, can eliminate the influence of detection errors. For instance, the 3σ principle can be used to remove outliers, thus completing denoising.

[0069] The preset baselines are standard indicators for different performance evaluation parameters. For example, the preset baseline for dirt residue rate is ≤5%. The preset baselines for each performance evaluation parameter are shown in Table 3. Table 3:

[0070] It should be noted that the different performance indicators in Table 3 belong to different evaluation dimensions. For example, the cleaning performance dimension includes cleanliness, cleaning efficiency, and cleaning effect; the operational performance dimension includes energy consumption and noise; and the extreme performance dimension includes extreme effectiveness and durability.

[0071] It is understandable that, as shown in the quantification methods in Table 3, the performance evaluation parameters in this application can be obtained either directly by data collection or through simple calculation. For example, cleaning time can be directly collected, and its corresponding quantification method does not require any further calculation; while dirt residue rate needs to be obtained through simple area calculation, and its corresponding quantification method includes performing calculations according to the corresponding calculation formulas. Therefore, referring to the quantification methods in Table 3, after obtaining all performance evaluation parameters, preprocessing can be performed.

[0072] Subsequently, each performance evaluation parameter is compared with its corresponding preset baseline, and a score is determined based on the comparison results, resulting in an individual score for each performance evaluation parameter. This individual score is one form of representation of the evaluation results.

[0073] For example, in one implementation, a preset baseline is the optimal ideal value. The smaller the deviation of the performance evaluation parameter from the optimal ideal value, the higher the score; the larger the deviation, the lower the score. Furthermore, for negative parameters such as dirt residue rate, the smaller the performance evaluation parameter, the higher its corresponding score; for positive parameters such as light transmittance, the larger the performance evaluation parameter, the higher its corresponding score.

[0074] In another implementation, several continuous numerical ranges can be pre-set for each performance evaluation parameter, with each range corresponding to a rating level or rating value. The preset baseline is the boundary value between adjacent ranges, and the range into which the parameter falls corresponds to its rating level or rating value.

[0075] Furthermore, if the current individual scores are all scores corresponding to each performance evaluation parameter calculated in a single instance, then under the same conditions, multiple tests can be performed, and the average of each individual score can be calculated to balance the error from multiple measurements.

[0076] Finally, after obtaining the individual scores, the individual scores can be weighted and fused together with their corresponding correction weights to generate the comprehensive score of the cleaning device.

[0077] Understandably, the sum of the correction weights should typically be 1. Therefore, if the sum of the correction coefficients obtained after adjusting the base weights is not 1, the correction coefficients need to be normalized. For example, in the previous example, the sum of the correction weights corresponding to M1-M3 is 1.025; then all correction coefficients can be uniformly divided by 1.025 before being used in the calculation. That is, the final correction weights used in the calculation change from 0.66, 0.165, and 0.2 to 0.66 / 1.025, 0.165 / 1.025, and 0.2 / 1.025.

[0078] For example, taking the weighted summation form in weighted fusion as an example, the formula for calculating the comprehensive score is as follows:

[0079] Among them, S 总 The calculated overall score, To apply normalized adjustments to the weights, For each performance evaluation parameter after multiple measurements and averaging, i represents the individual score of each performance evaluation parameter. , , The total number of evaluation dimensions corresponding to the three types—cleaning performance, operational performance, and extreme performance—is used for mean calculation. A higher score indicates better performance.

[0080] In this embodiment, the collected performance evaluation parameters are weighted and fused, for example, by weighted summation. Specifically, based on the intelligent driving scenario, different environments, and the requirements of the perception components for different performance evaluation parameters, the weights of the collected different performance evaluation parameters in the calculation are adjusted so that the final output comprehensive score is more in line with the current intelligent driving scenario, thus obtaining the test results of the cleaning device under the current intelligent driving scenario.

[0081] Optionally, a performance rating can also be provided after generating a comprehensive score.

[0082] For example, performance ratings are divided into four levels: Excellent, Good, Average, and Poor. The core judgment principle is that cleaning performance is a mandatory compliance item; if any performance evaluation parameter in this category fails to meet the standard, it is directly judged as Poor. The correspondence between each performance rating and the overall score is shown in Table 4. Table 4:

[0083] In other words, in this embodiment, in addition to providing a comprehensive score, a performance rating can also be provided as a test result. This allows for the provision of multiple forms of comprehensive evaluation.

[0084] In this embodiment of the application, the method may further include: A multivariate evaluation chart is generated based on the individual scores of each of the aforementioned performance evaluation parameters. The multivariate evaluation chart is used to indicate the performance status of the cleaning device on each performance evaluation parameter. The test results also include the multivariate evaluation graph.

[0085] It is understood that the multivariate evaluation chart is used to visually represent the different performance evaluation parameters of the cleaning device. The multivariate evaluation chart can be, for example, a spider web chart, where different axes are assigned values ​​by at least one of the performance evaluation parameters. Alternatively, it can be a bar chart, pie chart, or other schematic diagram that visually represents the numerical values.

[0086] For example, when the multivariate evaluation chart is a spider web diagram, each axis of the spider web diagram can be the performance index in the first column of Table 3. Based on this, when the performance index is cleanliness, this axis is only assigned a value by the dirt residue rate; while when the performance index is cleaning efficiency, this axis is jointly assigned a value by cleaning time and cleaning liquid / gas consumption per unit area, for example, by averaging the corresponding individual scores.

[0087] In this embodiment, a comprehensive score and a multivariate evaluation chart (such as a spider web chart) can be provided simultaneously as test results, so that users can not only obtain an overall performance evaluation of the cleaning device through the comprehensive score, but also intuitively perceive the individual performance of the cleaning device through the multivariate evaluation chart.

[0088] In one embodiment of this application, the method may further include the following steps: S21, calculate the degree of deviation between each of the performance evaluation parameters and the corresponding preset baseline; S22, determine the corresponding performance index to be optimized based on the performance evaluation parameters whose deviation exceeds a preset threshold; S23, Based on the performance indicators to be optimized, generate optimization guidance information for the cleaning device.

[0089] That is, in addition to comprehensive scores and multivariate evaluation charts, the test results in this application embodiment can also provide the parts of the cleaning device that need improvement based on the deviation of the performance evaluation parameters from the corresponding preset baseline.

[0090] For example, calculate the deviation of each performance evaluation parameter from the preset baseline. When the preset baseline does not directly correspond to the measured value of each performance parameter (such as the dirt residue rate in Table 3, where the preset baseline is not an area value), the corresponding quantitative calculation can be performed first with reference to the quantification method in Table 3, and then compared with the preset baseline.

[0091] If a performance evaluation parameter deviates significantly from its corresponding preset baseline, meaning the performance evaluation indicator is unqualified, then either the performance indicator to which the parameter belongs needs optimization, or the performance evaluation parameter itself needs optimization. Subsequently, corresponding optimization guidance information can be generated based on the performance indicator to be optimized, or based on the performance evaluation parameter.

[0092] For example, if the average cleaning noise deviates significantly from its corresponding preset baseline, such as excessive average cleaning noise, optimization guidance information can be provided for the average cleaning noise (performance evaluation parameter) or noise (performance indicator): "Optimize the sealing structure of the cleaning device to reduce noise generated by airflow / vibration, and adjust the cleaning frequency to balance cleaning effect and noise control." If the function recovery time deviates significantly from its corresponding preset baseline, i.e., the function recovery time is too long, optimization guidance information can be provided for the function recovery time (performance evaluation parameter) or extreme effectiveness (performance indicator): "Optimize the drive power of the cleaning device and adjust the pressure parameters of the gas-liquid mixture injection." If the energy consumption of a single cleaning operation deviates significantly from its corresponding preset baseline, i.e., the energy consumption exceeds the standard, it can be suggested to "Optimize the power control logic to reduce real-time energy consumption during the cleaning process, and adjust the liquid / gas injection volume to improve cleaning efficiency per unit area."

[0093] It is understandable that the performance evaluation parameters in step S21 can be directly collected or obtained after simple quantization through the quantization process shown in Table 3, or they can be the preprocessed performance evaluation parameters in step S13.

[0094] Optionally, if the individual scores of each performance evaluation parameter have already been calculated based on a preset baseline, the performance index to be optimized can be determined based on the individual scores. For example, if the smaller the deviation of a performance evaluation parameter from its optimal ideal value, the higher its individual score, then the performance evaluation parameter with the lowest individual score can be used as the performance index to be optimized.

[0095] Optionally, if all performance evaluation parameters meet the preset baseline, the weakest performance evaluation parameter can be selected. The "weakest point" can be the minimum value among all current individual scores, or a value under any preset filtering conditions, and can be adjusted according to user needs.

[0096] This application not only provides various forms of test results, but also provides optimization guidance information for poor performance evaluation parameters, thereby improving the user experience.

[0097] In one embodiment of this application, simulating dirty conditions for the sensing component according to the test scenario may include the following steps: S31, Obtain the preset dirt and physical release conditions corresponding to the test scenario; S32, apply the preset dirt to the sensing component according to the physical application conditions.

[0098] As shown in Table 1, different test scenarios (weather and driving speed range) correspond to different dirt conditions, that is, there are different preset dirt conditions.

[0099] Optionally, different types of dirt can be stored in different areas to avoid cross-contamination and ensure the stability of dirt characteristics.

[0100] In this embodiment, different types of contaminants also have corresponding physical release conditions. For example, dust requires diffuse spraying, while rainwater requires large-area coverage. To better suit actual usage environments, different types of contaminants can be released according to their corresponding physical release conditions.

[0101] For example, an adjustable nozzle (adjustable spray angle / force / density / distance) can be configured to adapt to different physical release conditions of contaminants (such as large-area water mist adhesion, uniform dust dispersion, atomized spray condensation of frost, etc.). After confirming the physical release conditions, the corresponding nozzle is selected, or the nozzle is adaptively adjusted to complete the release of the contaminant.

[0102] In addition, electromagnetic control valve assemblies with a response time of ≤50ms can be provided for controlling the switching or distribution of different types of contaminants.

[0103] In this embodiment, different types of contaminants are released according to their corresponding physical release conditions, making the simulated contaminant conditions more consistent with the actual working scenario and improving the accuracy of the test.

[0104] In one embodiment of this application, after step S32, which involves applying the preset dirt to the sensing component according to the physical application conditions, the method may further include: When the preset dirt meets the morphology adjustment conditions, the ambient temperature is adjusted to a preset temperature range and maintained for a preset duration, so that the preset dirt adheres to the surface of the sensing component in a specific form.

[0105] In this embodiment of the application, considering that dirt may exist in different forms in different scenarios, a temperature condition step is also set to make the dirt more consistent with its corresponding form.

[0106] As shown in Table 1, the mud needs to be in a heavily dried state under extreme high-temperature conditions, while it is impossible to achieve this state at normal temperatures. Even with simulated environmental parameters, the short-term environmental parameter values ​​are insufficient to make the mud meet the requirements for a heavily dried state. Therefore, it is necessary to adjust the ambient temperature to a preset temperature range and maintain it for a preset duration to ensure that the mud's morphology meets the requirements for a heavily dried state, thus allowing the mud to adhere to the surface of the sensing component in a heavily dried state.

[0107] For example, in order to achieve the drying, solidification or condensation of dirt and frost, and to ensure that the form of dirt and frost matches the actual scene, temperature control can be performed, with an adjustment range of -40℃ to 85℃. The low temperature mode (-40℃ to 0℃) is used for condensation of dirt and frost, and the high temperature mode (40℃ to 85℃) is used for drying dirt and frost (such as the drying of dust and mud).

[0108] It is understandable that the morphology adjustment conditions refer to the pre-defined forms of dirt that require special adjustment, including but not limited to heavily dried mud or frost. These are all parameters that can be set according to user needs, and are not limited to the examples above. Pre-set temperature ranges and preset durations are also parameters that can be adjusted by the user according to their needs.

[0109] In this embodiment, considering that dirt may have different morphological requirements in different scenarios, a temperature condition step is set to make the dirt more consistent with its corresponding morphology, so that the environment simulated by the cleaning device is more consistent with the actual working environment, thereby improving the accuracy of the cleaning device test.

[0110] In one embodiment of this application, collecting performance evaluation parameters from the cleaning device may specifically include the following steps: S51, Obtain the device type of the cleaning device; S52, determine the corresponding performance evaluation parameters according to the device type of the cleaning device, and collect the determined performance evaluation parameters from the cleaning device.

[0111] In this embodiment of the application, as shown in Table 3, different cleaning devices have different data collection parameters. When the type of cleaning device is uncertain, the parameters required by all types of cleaning devices can be collected to ensure the smooth execution of the current testing method.

[0112] In this embodiment, the type of the cleaning device can also be obtained. Only the parameters corresponding to the current type of cleaning device are collected, which can reduce redundant steps in data collection and improve detection efficiency.

[0113] For example, cleaning devices include liquid, gas, and gas-liquid hybrid types. For the performance evaluation parameter of liquid / gas consumption per unit area, when the type of cleaning device is uncertain, both total liquid consumption and total gas consumption will be collected simultaneously. If the cleaning device under test is a gas-type cleaning device, collecting the total liquid consumption is a redundant step; if executed, the resulting parameter will only be 0.

[0114] In this case, we propose first determining the type of cleaning device, and then only collecting parameters corresponding to that type of cleaning device. That is, in this situation, the total liquid consumption is not collected at all, which improves detection efficiency.

[0115] Optionally, while receiving user instructions and confirming test requirements, the type of cleaning device input or selected by the user can be obtained.

[0116] Reference Figure 2 The diagram illustrates a structural block diagram of a test system for a vehicle cleaning device according to an embodiment of this application. The test system may include: The simulation drive component 201 is used to simulate dirt conditions and environmental parameters for the sensing component according to the test requirements corresponding to the cleaning device to be tested, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. The evaluation acquisition component 202 is used to acquire performance evaluation parameters from the sensing component and the cleaning device during and after the cleaning operation is performed by the cleaning device. The result generation component 203 evaluates the performance evaluation parameters based on at least one evaluation dimension, obtains the evaluation results of each performance evaluation parameter, and merges the evaluation results to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0117] Optionally, the analog drive component 201 includes: The requirement parsing component is used to obtain the test requirements corresponding to the cleaning device, which include the test scenario and the sensing components to be cleaned. An environment simulation component is used to simulate dirt conditions and environmental parameters for the sensing component based on the test scenario.

[0118] Optionally, the result generation component 203 includes: The basic weight determination component is used to determine the basic weight of each performance evaluation parameter based on the evaluation dimension corresponding to each performance evaluation parameter. The corrected weight acquisition component is used to determine the correction coefficient corresponding to each of the basic weights based on the test scenario for the cleaning device and the type of the sensing component, and to correct the corresponding basic weights using the correction coefficients to obtain each corrected weight. A single-item score generation component is used to preprocess each of the performance evaluation parameters and obtain a single-item score for each of the performance evaluation parameters based on the preprocessed performance evaluation parameters and their corresponding preset baselines. A comprehensive score generation component is used to perform weighted fusion on each of the individual scores based on the corrected weights to generate a comprehensive score for the cleaning device, so as to obtain test results including the comprehensive score.

[0119] Optionally, the testing system further includes: An evaluation chart generation component is used to generate a multivariate evaluation chart based on the individual scores of each of the performance evaluation parameters. The multivariate evaluation chart is used to indicate the performance status of the cleaning device on each performance evaluation parameter. The test results also include the multivariate evaluation chart.

[0120] Optionally, the testing system further includes: A deviation calculation component is used to calculate the degree of deviation between each of the performance evaluation parameters and the corresponding preset baseline; An optimization confirmation component is used to determine the corresponding performance index to be optimized based on the performance evaluation parameters whose deviation exceeds a preset threshold. A guidance generation component is used to generate optimization guidance information for the cleaning device based on the performance indicators to be optimized.

[0121] Optionally, the environment simulation component includes: The release condition acquisition component is used to acquire the preset dirt and physical release conditions corresponding to the test scenario; An application component is used to apply the preset dirt to the sensing component according to the physical application conditions.

[0122] Optionally, the testing system further includes: A shape adjustment component is used to adjust the ambient temperature to a preset temperature range and maintain it for a preset duration when the preset dirt meets the shape adjustment conditions, so that the preset dirt adheres to the surface of the sensing component in a specific shape.

[0123] Optionally, the requirement parsing component includes: A scene acquisition component is used to respond to user commands to determine the sensing component that needs to be cleaned and the test scene corresponding to the cleaning of the sensing component. The test scene includes weather and driving speed range. The weather and driving speed range are used to simulate dirt conditions and environmental parameters for the sensing component.

[0124] Optionally, the evaluation acquisition component 202 includes: A type acquisition component is used to acquire the device type of the cleaning device; The evaluation parameter acquisition component is used to determine the corresponding performance evaluation parameters based on the device type of the cleaning device, and to acquire the determined performance evaluation parameters from the cleaning device.

[0125] The testing system provided in this application first obtains the user's testing requirements for the cleaning device. Different testing requirements correspond to different dirt conditions and environmental parameters under which the cleaning device operates. Then, based on the testing requirements, it simulates the dirt conditions and environmental parameters for the sensing component that needs cleaning, standardizing the dirt conditions the sensing component may experience and avoiding inconsistencies in testing conditions caused by human intervention. Subsequently, it drives the cleaning device under test to clean the sensing component in a dirty state, collecting corresponding performance evaluation parameters from the sensing component and the cleaning device during and after the cleaning process. Based on these performance evaluation parameters, it generates test results for the cleaning device. The collected performance evaluation parameters correspond to at least one evaluation dimension. By associating the performance evaluation parameters with the corresponding dimensions, the cleaning device can be evaluated on at least one dimension, generating more comprehensive and objective test results and improving the reference value of the test results.

[0126] By adopting the testing scheme of this application, no manual intervention is required to apply dirt or collect / analyze test results, thus maintaining the uniformity of testing conditions and avoiding errors in test results caused by fluctuations in testing conditions. Furthermore, the automated processing method improves the efficiency of test result generation. Simultaneously, by conducting a multi-dimensional comprehensive evaluation of the cleaning device, the test results become more comprehensive.

[0127] In some implementations, the aforementioned testing system adopts a three-layer architecture of "application layer - execution layer - evaluation layer," with each layer's modules working collaboratively to achieve a fully automated closed loop of "scenario creation - cleaning testing - multi-dimensional quantitative evaluation and summary." The execution layer is responsible for scenario replication and data collection, the evaluation layer for core analysis and intelligent evaluation, and the application layer for result output and interactive adaptation. Specifically, the execution layer includes the environment simulation function in the aforementioned simulation-driven component 201 and the performance evaluation parameter collection function in the evaluation collection component 202; the evaluation layer includes the evaluation function in the aforementioned result generation component 203; and the application layer includes the test requirement acquisition function in the aforementioned simulation-driven component 201 and displays the test results generated by the result generation component 203.

[0128] The specific structure of the three-tier architecture is as follows: Application Layer (Human-Computer Interaction and Data Management). As the core of human-computer interaction and data coordination, the application layer is responsible for parameter input, result display, and data traceability. It is composed of hardware and software working together and specifically includes: Hardware components: industrial-grade touch display screen, high-performance industrial host, Ethernet communication module; Software components: Human-computer interaction interface (supports manual input / preset recall of parameters such as sensor type (i.e., sensing component type) and operating condition parameters (i.e., environmental parameters), data management module (stores evaluation data and scenario manufacturing data (such as environmental parameters and / or dirt condition parameters), supports historical data query and export), and report generation module (automatically generates quantitative evaluation reports, including individual scores, comprehensive scores, optimization suggestions, etc.). The application layer receives test requirements input by the user (such as sensor type: 200W pixel front-view camera, cleaning device: gas-liquid mixing cleaning equipment, test scenario: scenario working condition, see Table 1 for reference), and transmits relevant parameters to the execution layer and evaluation layer through the Ethernet communication module. At the same time, it receives the results and optimization suggestions fed back by the evaluation layer.

[0129] Execution Layer (Scene Creation and Cleaning Testing). As the system's execution terminal, the execution layer is responsible for completing standardized scene creation and cleaning device linkage testing. It is the core hardware support for multi-dimensional quantitative evaluation and summarization, specifically including: Hardware components: mainly include a scene creation module (dirt creation module, environmental parameter simulation module), a cleaning linkage module (real-time collection of performance evaluation parameters and feedback to the evaluation layer), and a supporting communication interface (to realize two-way data interaction with the application layer and the evaluation layer). Software components: Scene creation module (issues control commands for dirt creation and environmental simulation), cleaning linkage module (controls cleaning devices to complete standardized testing operations and synchronously transmits scene operation status and cleaning test status to the evaluation layer). The execution layer receives evaluation parameters input from the application layer through a communication interface. The scene control submodule regulates the scene creation module to build a preset scene. After the scene parameters stabilize, the cleaning control submodule regulates the cleaning linkage module to complete the standardized cleaning test operation. The status feedback submodule collects scene parameters and cleaning test process parameters in real time and transmits them synchronously to the evaluation layer to provide basic data for multi-dimensional quantitative evaluation and summary.

[0130] Specifically, the scenario creation module is used to simulate various dirt adhesion scenarios and driving conditions in the actual use of intelligent driving sensors, providing a standardized and repeatable scenario basis for cleaning tests and multi-dimensional quantitative evaluation. It is composed of a dirt creation module and an environmental parameter simulation module.

[0131] The dirt manufacturing module is used to standardize and precisely attach five types of typical intelligent driving sensor dirt: rainwater, dust, mud, insect residue, and frost. Its core components include dirt storage, spray control, and high / low temperature environmental chambers. The specific hardware components and functions are as follows: Dirt storage unit: used to classify and store various dirt media to avoid cross-contamination and ensure stable dirt characteristics; Spray control unit: used to achieve precise metering and standardized spraying of various dirt media, ensuring controllable dirt attachment amount and uniformity; Dedicated spray unit: equipped with adjustable nozzles to adapt to different dirt attachment forms; also equipped with electromagnetic control valve groups to achieve rapid switching between dirt types and spray modes, linking with the scene control submodule at the execution layer to receive control commands and adjust spray parameters; High / low temperature environmental chamber: provides a suitable environment for dirt manufacturing, enabling the drying, solidification, or condensation of dirt or frost, ensuring the dirt form matches the actual scene.

[0132] In addition, it may include a sensor positioning unit: used to accurately fix and position the intelligent driving sensor and its matching controller, ensuring that the dirt spraying unit is aligned with the sensor window and avoiding deviation that affects the evaluation accuracy.

[0133] The environment simulation module within the scenario creation module simulates various environmental conditions encountered during actual driving. Working in conjunction with the dirt creation module, it achieves scenario-based simulation of "dirt + working environment." Its core components include hardware for wind speed control, rainfall control, and environmental control, specifically: Wind speed control unit: simulates airflow environments at different driving speeds to ensure dirt adhesion patterns closely resemble actual driving conditions; Rainfall control unit: simulates driving scenarios with different rainfall intensities, working in conjunction with rainwater dirt creation to simulate rainwater adhesion, including light, moderate, and heavy rainfall intensities; Environmental control unit: simulates environmental parameters such as temperature, humidity, and light during actual driving to ensure the realism of the scenario simulation, working in conjunction with the dirt creation module to enhance the realism of dirt adhesion. Linkage control unit: links with the scenario control submodule, receives working condition parameters from the application layer, coordinates the operation of the wind speed control, rainfall control, and environmental control units, and achieves rapid switching and stable operation in different environments.

[0134] The cleaning linkage module is used to coordinate with the scene creation module. When the dirt and environment are ready, it triggers the cleaning device to be tested to perform cleaning and collects relevant cleaning test data in real time, which is then fed back to the evaluation layer. The specific components are as follows: Drive control unit: responsible for linking the scene creation module, receiving the working condition ready signal in real time (dirt adhesion meets the standard, environmental parameters are stable), and driving the cleaning device under test to start, run and stop; supports manual and automatic triggering modes, and presets the cleaning start time according to the working condition requirements to ensure that the cleaning process is synchronized with the test scene.

[0135] Cleaning Test Unit: Responsible for collecting various key data in real time during the cleaning process, including visual data, noise data, energy consumption data, time consumption data, and intelligent driving signal correlation data. Specifically, it consists of a visual inspection device, a sound pressure meter, a pressure gauge, a liquid level gauge, a timer, and an energy consumption meter (with a built-in current / voltage monitoring device).

[0136] After receiving scene parameters (dirt type, operating condition type, etc.) from the application layer, the scene control submodule synchronously sends control commands to the dirt generation module and the environment simulation module. The environment simulation module first sets up preset environmental operating conditions (such as high speed 80km / h, moderate rain, temperature 25℃). After the environmental operating condition parameters stabilize, the dirt generation module completes the precise positioning of the sensors through the sensor surface positioning device, adjusts the high and low temperature environment chamber to the appropriate temperature, and the dirt storage unit and the spray control unit work together to complete standardized dirt spraying / condensation, ensuring that the dirt is uniformly attached to the sensor window. At the same time, various detection sensors of the dirt generation module and the environment simulation module collect the current environmental scene parameters and cleaning test data in real time, and feed them back to the evaluation layer to form a closed-loop control of scene generation, ensuring the standardization, authenticity and repeatability of scene simulation, and providing an accurate scene basis for cleaning test and evaluation summary.

[0137] The evaluation layer (quantitative evaluation and summary recommendations) serves as the system's data analysis and evaluation summary unit. It is responsible for receiving scenario and cleaning test data collected by the execution layer, and using a dynamic weight adaptive model and multi-dimensional quantification algorithms to complete multi-dimensional quantitative evaluation and summary. Simultaneously, it generates optimization suggestions, linking the application layer and execution layer to form a closed loop. Its specific components and quantitative evaluation are as follows: Software components: data classification and processing module, multi-dimensional quantitative evaluation module (combining hardware-collected data to complete the score calculation and comprehensive summary for each dimension); Collaborative Logic: The evaluation layer receives scene parameters and cleaning test data transmitted from the execution layer. Through dynamic weight model and quantitative evaluation algorithm, it completes multi-dimensional quantitative evaluation and comprehensive summary, generates evaluation summary report and optimization suggestions, and transmits them to the application layer for display, forming an autonomous closed loop of "application control - scene creation - cleaning test - quantitative evaluation".

[0138] Quantitative evaluation: Based on the classification principles of cleaning performance dimensions (cleanliness, cleaning efficiency, cleaning effect), operational performance dimensions (energy consumption, noise), and extreme performance dimensions (extreme effectiveness, durability), combined with sensor type, driving speed range (static / dynamic), etc., a quantitative evaluation model is formed. The quantitative evaluation process can be divided into four steps: sensor signal processing, quantitative calculation, dynamic evaluation, and evaluation summary.

[0139] Step S1: Sensor signal processing. The raw data collected by the execution layer is classified, statistically analyzed, and baseline-calibrated based on the test scenario, sensor type, and usage status, providing a standardized dataset for subsequent quantization calculations. Step S2: Quantitative Calculation. Based on the quantification method in Table 3, quantitative calculations are performed on the standardized data processed in Step S1. Step S3: Dynamic Evaluation. Based on the test requirements received by the application layer (sensor type, cleaning device type, test scenario requirements), and using a dynamic weight adaptive model of "base weight + correction coefficient", the weights are dynamically adjusted and the comprehensive score is calculated.

[0140] Step S4: Evaluation and Summary. The summary and recommendations, combining the rating results, performance across various dimensions, test scenarios, sensor types, and other core elements, are provided in three parts: "Summary Overview," "Targeted Recommendations," and "Scenario Adaptation Recommendations." Summary and Overview: Briefly summarize the overall rating of the cleaning device, its shortcomings in core strengths, clarify its overall adaptability, and highlight the outstanding performance in the dimensions of cleaning performance, operational performance, and extreme performance (displayed in a multivariate evaluation chart). At the same time, it relates the test scenarios and sensor types to explain the current scope of adaptability.

[0141] Targeted recommendations: Provide specific and feasible optimization directions for the weakest indicators, and coordinate the adjustment of the structure and parameters of the cleaning equipment.

[0142] Scene adaptation suggestions: Based on the weather and driving speed range, identify the suitable and unsuitable scenarios for the cleaning device and provide scenario-based adjustment suggestions. For example, "This cleaning device performs well in sunny and normal weather and light rain in static / low-speed dynamic scenarios, and can be adapted to camera-type sensors; it performs poorly in extreme low temperature frost, heavy rain, strong airflow, and high-speed dynamic scenarios, and is not suitable for high-speed driving scenarios with lidar. It is recommended to add an auxiliary cleaning module or adjust the cleaning start timing in extreme scenarios."

[0143] In one specific implementation, a 200W pixel automotive front-view camera is used as the sensing component to be cleaned, with a total viewing window area of ​​1.2cm. 2 The baseline transmittance for dirt-free surfaces is 99.8%, and it is compatible with gas-liquid mixing cleaning devices (spray pressure 3 Bar, basic spray volume 2 mL / time). The test scenario is a dynamic full-scene under all weather conditions, and the corresponding correction coefficients in Table 2 are: cleaning performance dimension K1=1.2, operational performance dimension K2=0.8, and extreme performance dimension K3=1.1.

[0144] according to Figure 3 The three-tier architecture shown completes parameter deployment. Among them, Figure 3 This includes the software components of the aforementioned layers.

[0145] Application layer: Input test requirements (camera + gas-liquid mixing cleaning device + all environmental conditions), and generate a report including quantitative scores, spider web diagrams, and optimization suggestions; Execution layer: The scene creation module simulates the core rainy environment (rainfall 20mm / h, temperature 15℃, wind speed 15m / s), and standardizes the creation of rainwater + light mud and dirt; the cleaning linkage module collects cleaning test data through a visual inspection device, energy consumption meter, and sound pressure meter, with a collection frequency of 10Hz. Evaluation layer: Preset performance evaluation parameters in Table 3 and quantitative rating standards in Tables 2 and 4.

[0146] like Figure 4 As shown, the execution layer transmits the collected raw cleaning test data (dirt images, cleaning time, liquid consumption, light transmittance, energy consumption, noise, etc.) to the evaluation layer. The data classification and processing module completes data processing according to the cleaning performance dimension, operational performance dimension, and extreme performance dimension. The specific process is as follows: Based on the quantification method in Table 3, the standardized data obtained from S1 were quantitatively calculated using a single index. The calculation results were then labeled with "camera + dynamic + full operating condition". The specific measured results (all meeting the baseline reference requirements) are as follows: Cleanliness: Dirt residue rate = 3.5% (≤5%); Cleaning efficiency: Cleaning time = 0.7s (≤1s), single liquid consumption 1.8mL (≤2mL, gas-liquid mixture baseline). Cleaning effect: Light transmittance = 98.9% (≥98%); Extreme efficacy: Functional recovery time = 2.3s (≤3s); Energy consumption: 13Wh per cleaning cycle (≤20Wh); Noise: Average cleaning noise level 52dB (≤55dB); Durability: Performance degradation rate = 5.8% (≤10%).

[0147] Based on the base weights and correction coefficients, the weight adjustment and comprehensive performance score calculation are completed. The specific process is as follows: Basic weight allocation: Cleaning performance dimension (0.60), extreme performance dimension (0.25), and operational performance dimension (0.15), with equal weighting for single indicators within the class; Correction coefficient superposition: Based on the combination of dynamic and camera, take K1=1.2, K2=0.8, K3=1.1, and finally get: dynamic weight of cleaning performance dimension 0.65, extreme performance dimension 0.26, and operation performance dimension 0.09; Overall score calculation: The overall performance score of this embodiment is 92.5 points.

[0148] Based on this, a quantitative rating is completed by combining the achievement of individual indicators and the overall score, and then... Figure 5 A spider web diagram of multi-dimensional cleaning performance is used to visualize the results, and ultimately, scenario-based optimization suggestions are output.

[0149] Quantitative rating: Based on the quantitative rating criteria in Table 4, the overall score is 92.5 points (≥90 points), and all cleaning performance indicators are better than the baseline reference, while the remaining indicators meet the baseline requirements. Therefore, it is rated as excellent. Visualization of multivariate evaluation charts: such as Figure 5 As shown, the spider diagram uses cleanliness, cleaning efficiency, cleaning effect, extreme effectiveness, energy consumption, noise, and durability as the axes. The axis values ​​are the normalized scores (0-1) of each dimension, or the average score of each item in each dimension. In this embodiment, the spider diagram shows that cleanliness and cleaning effect are the advantageous dimensions, liquid / gas consumption is basically up to standard, and functional recovery time in extreme scenarios is a relative weakness, but overall there is no significant disadvantage in the scores of each dimension.

[0150] Summary and recommendations: Summary: This gas-liquid hybrid cleaning device is suitable for cleaning vehicle front-view cameras in dynamic rainy or moderate rain scenarios. It has an excellent overall rating, good recovery of core perception functions, and meets the cleaning needs of intelligent driving. Targeted recommendations: To address the relatively low liquid consumption per unit area, it is recommended to optimize the nozzle atomization angle and reduce the single spray volume to 1.5-1.6mL to improve economy while ensuring cleaning effectiveness; to address the short functional recovery time, it is advisable to optimize the cleaning device's start-up response logic to shorten the response time from 80ms to within 50ms. Scene adaptation suggestions: This device can be directly applied to camera cleaning scenarios in sunny weather and light / moderate rain. If it is to be extended to extreme low temperature and frost scenarios, it is recommended to add a low temperature preheating module or antifreeze cleaning fluid to avoid the liquid freezing and affecting the cleaning effect.

[0151] Compared with existing technologies, this system has the following significant advantages: More comprehensive evaluation dimensions: Based on a multi-dimensional quantitative evaluation system, it breaks through the limitations of the existing single dimension, covers the core needs of the whole vehicle for intelligent driving sensor cleaning, and the quantitative standards of each dimension are clear. Compared with traditional qualitative evaluation or single-dimensional quantification, the comprehensiveness and accuracy of the evaluation are significantly improved.

[0152] Richer scenario adaptation: Based on "standardized pollution spraying + cross-scenario simulation adaptation", the impact of pollution and environmental differences is reduced. Combined with dynamic weight adaptive model, it adapts to multiple sensors and all working conditions, solving the pain points of poor experimental repeatability and insufficient scenario adaptability of existing technologies. More autonomous closed-loop system: Constructing an "application layer-execution layer-evaluation layer" system architecture: Integrating modules such as scenario simulation, cleaning linkage, and data processing to achieve fully automated evaluation. At the same time, based on the evaluation data feedback, it outputs summary evaluation suggestions for the cleaning device, forming a complete closed loop. Compared with the current situation where there is a lot of manual intervention and no optimization capability, it greatly improves the efficiency of R&D and testing.

[0153] Reference Figure 6 This application also provides an electronic device, such as... Figure 6 As shown, it includes a processor 601, a communication interface 602, a memory 603, and a communication bus 604, wherein the processor 601, the communication interface 602, and the memory 603 communicate with each other through the communication bus 604. Memory 603 is used to store computer programs; When processor 601 executes a program stored in memory 603, it performs the following steps: Based on the test requirements of the cleaning device to be tested, simulate dirt conditions and environmental parameters for the sensing component, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. During and after the cleaning operation is performed by the cleaning device, performance evaluation parameters are collected from the sensing components and the cleaning device. The performance evaluation parameters are evaluated based on at least one evaluation dimension to obtain the evaluation results of each performance evaluation parameter. The evaluation results are then fused to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

[0154] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0155] The communication interface is used for communication between the aforementioned terminal and other devices.

[0156] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0157] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0158] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores instructions that, when executed on a computer, cause the computer to perform the test method of any of the vehicle cleaning devices described in the above embodiments.

[0159] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium such as a solid-state disk (SSD).

[0160] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0161] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0162] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A test method for a vehicle cleaning device, characterized in that, The method includes: Based on the test requirements of the cleaning device to be tested, simulate dirt conditions and environmental parameters for the sensing component, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. During and after the cleaning operation is performed by the cleaning device, performance evaluation parameters are collected from the sensing components and the cleaning device. The performance evaluation parameters are evaluated based on at least one evaluation dimension to obtain the evaluation results of each performance evaluation parameter. The evaluation results are then fused to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, the operational performance dimension, and the extreme performance dimension.

2. The method according to claim 1, characterized in that, The evaluation of the performance evaluation parameters based on at least one evaluation dimension to obtain evaluation results for each performance evaluation parameter, and the fusion of the evaluation results to generate test results for the cleaning device, includes: The basic weight of each performance evaluation parameter is determined based on the evaluation dimension corresponding to each performance evaluation parameter. Based on the test scenario for the cleaning device and the type of the sensing component, the correction coefficients corresponding to each of the basic weights are determined, and the correction coefficients are used to correct the corresponding basic weights to obtain each corrected weight. After preprocessing each of the performance evaluation parameters, a single score for each of the performance evaluation parameters is obtained based on the preprocessed performance evaluation parameters and their corresponding preset baselines. Based on the corrected weights, a weighted fusion is performed on each of the individual scores to generate a comprehensive score for the cleaning device, thereby obtaining test results including the comprehensive score.

3. The method according to claim 2, characterized in that, The method further includes: A multivariate evaluation chart is generated based on the individual scores of each of the aforementioned performance evaluation parameters. The multivariate evaluation chart is used to indicate the performance status of the cleaning device on each performance evaluation parameter. The test results also include the multivariate evaluation graph.

4. The method according to claim 2, characterized in that, The method further includes: Calculate the degree of deviation between each of the performance evaluation parameters and the corresponding preset baseline; Based on the performance evaluation parameters whose deviation exceeds a preset threshold, the corresponding performance indicators to be optimized are determined; Based on the performance indicators to be optimized, optimization guidance information is generated for the cleaning device.

5. The method according to claim 1 or 2, characterized in that, The step of simulating dirt conditions and environmental parameters for the sensing component based on the testing requirements corresponding to the cleaning device to be tested includes: Obtain the test requirements corresponding to the cleaning device, including the test scenario and the sensing components to be cleaned; The sensing component is simulated with dirty conditions and environmental parameters based on the test scenario.

6. The method according to claim 5, characterized in that, The step of simulating dirty conditions for the sensing component according to the test scenario includes: Obtain the preset dirt and physical release conditions corresponding to the test scenario; The preset dirt is applied to the sensing component according to the physical application conditions.

7. The method according to claim 6, characterized in that, After applying the preset dirt to the sensing component according to the physical application conditions, the method further includes: When the preset dirt meets the morphology adjustment conditions, the ambient temperature is adjusted to a preset temperature range and maintained for a preset duration, so that the preset dirt adheres to the surface of the sensing component in a specific form.

8. The method according to claim 5, characterized in that, The process of obtaining the test requirements corresponding to the cleaning device includes: In response to a user command, the sensor components that need to be cleaned and the test scenario corresponding to the cleaning of the sensor components are determined, the test scenario including weather and driving speed range; The weather and the driving speed range are used to simulate dirt conditions and environmental parameters for the sensing component.

9. The method according to claim 1, characterized in that, The collection of performance evaluation parameters from the cleaning device includes: Obtain the device type of the cleaning device; Based on the device type of the cleaning device, the corresponding performance evaluation parameters are determined, and the determined performance evaluation parameters are collected from the cleaning device.

10. A testing system for a vehicle cleaning device, characterized in that, The testing system includes: The simulation drive component is used to simulate dirt conditions and environmental parameters for the sensing component according to the test requirements of the cleaning device to be tested, and drive the cleaning device to perform cleaning operations on the sensing component in a dirty state under the environmental parameters. An evaluation acquisition component is used to acquire performance evaluation parameters from the sensing component and the cleaning device during and after the cleaning operation is performed by the cleaning device. The result generation component evaluates the performance evaluation parameters based on at least one evaluation dimension, obtains the evaluation results of each performance evaluation parameter, and merges the evaluation results to generate the test results of the cleaning device. The evaluation dimensions include at least one of the cleaning performance dimension, operational performance dimension, and extreme performance dimension.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the test method as described in any one of claims 1 to 9.

12. A readable storage medium, characterized in that, The readable storage medium stores a computer program that, when executed by a processor, implements the test method as described in any one of claims 1 to 9.

Images