Outdoor rainfall simulation control method and system for automatic driving automobile test

Abstract

The invention relates to the technical field of automatic driving vehicle testing, in particular to an outdoor rainfall simulation control method and system for automatic driving vehicle testing. The method comprises the following steps: acquiring target parameters of outdoor rainfall simulation, wherein the target parameters at least comprise target rainfall intensity, target duration and a target rainfall angle; a fuzzy controller is adopted, and according to the difference between the target rainfall intensity and the current rainfall intensity and the variable quantity of the difference between the target rainfall intensity and the current rainfall intensity, the adjusting quantity of the water yield of each spraying device in the multiple spraying devices is obtained; adjusting the water outlet amount of the spraying device according to the adjusting amount, adjusting the starting time of the spraying device according to the target duration time, and adjusting the water outlet angle of the spraying device according to the target rainfall angle. According to the invention, high-precision simulation and dynamic regulation and control of the outdoor rainfall environment are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving vehicle testing, in particular to an outdoor rainfall simulation control method and system for automatic driving vehicle testing. BACKGROUND

[0002] With the rapid development of automatic driving technology, its perception and decision-making ability under complex weather conditions has become one of the key factors affecting the safety and reliability of the system. Among various weather disturbances, rainfall, as a common but complex natural weather phenomenon, has a significant impact on the environmental perception module (such as camera, laser radar, millimeter wave radar, etc.) of the automatic driving system. When raindrops are suspended in the air or attached to the surface of the sensor, it will cause image blurring, increased point cloud noise, distorted reflection signal, etc., which in turn leads to target recognition failure, path planning deviation, and even safety hazards such as emergency braking malfunction.

[0003] Therefore, in the development and testing process of the automatic driving system, how to build a simulation platform that can truly restore the natural rainfall environment has become an important issue to improve the system's robustness and extreme weather adaptability. At present, in the functional verification stage of the automatic driving vehicle, the commonly used rainfall simulation means mainly relies on laboratory rain rooms or simple spraying devices. Among them, although the laboratory rain room can control the parameters such as rainfall intensity and duration to a certain extent, due to its space closure, high cost, poor deployment flexibility and other shortcomings, it is difficult to meet the testing needs of large-scale, dynamic, and scene diversification.

[0004] In view of this, the present application is proposed. SUMMARY

[0005] The purpose of the present application is to provide an outdoor rainfall simulation control method and system for automatic driving vehicle testing to realize high-precision simulation and dynamic regulation and control of outdoor rainfall environment.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions: In a first aspect, the present application provides an outdoor rainfall simulation control method for automatic driving vehicle testing, comprising: obtaining target parameters of outdoor rainfall simulation, the target parameters at least including target rainfall intensity, target duration and target rainfall angle; using a fuzzy controller to obtain the adjustment amount of the water output of each spraying device in the plurality of spraying devices according to the difference between the target rainfall intensity and the current rainfall intensity, and the change amount of the difference between the target rainfall intensity and the current rainfall intensity; The spraying controller adjusts the water output of the spraying device according to the adjustment amount, adjusts the opening time of the spraying device according to the target duration, and adjusts the water output angle of the spraying device according to the target rainfall angle.

[0007] In a second aspect, the present application provides an outdoor rainfall simulation control system for automatic driving vehicle testing, comprising: A man-machine interaction module is configured to obtain target parameters of outdoor rainfall simulation, wherein the target parameters at least include target rainfall intensity, target duration and target rainfall angle. A processor is configured to adopt a fuzzy controller to obtain an adjustment amount of water output of each spraying device in a plurality of spraying devices according to a difference between the target rainfall intensity and a current rainfall intensity and a change amount of the difference between the target rainfall intensity and the current rainfall intensity. The spraying controller adjusts the water output of the spraying device according to the adjustment amount, adjusts the opening time of the spraying device according to the target duration, and adjusts the water output angle of the spraying device according to the target rainfall angle.

[0008] Compared with the prior art, the present application has the following beneficial effects: (1) The rainfall process in an outdoor environment is simulated with high precision.

[0009] The present application can realize dynamic regulation and control of rainfall intensity, direction and distribution in an outdoor complex environment by constructing a complete rainfall simulation system and introducing a closed-loop feedback mechanism based on fuzzy control, thereby significantly improving the authenticity and stability of rainfall simulation and providing more natural weather testing conditions for automatic driving vehicles.

[0010] (2) Differentiated rainfall control in multiple areas is supported to meet diversified testing requirements.

[0011] The system has spatial partition control capability and can set differentiated rainfall strategies for different road types such as curves, intersections and slopes, such as enhancing lateral rainfall, simulating local water accumulation and adjusting the spraying angle, so as to comprehensively verify the perception and decision-making performance of the automatic driving system in various rainy weather scenarios.

[0012] (3) Efficient linkage with an automatic driving testing platform to improve the automation level of testing.

[0013] The present application realizes data intercommunication with the automatic driving testing platform through a standard interface, supports automatic triggering of specific rainfall modes according to test cases, synchronously records rainfall events and vehicle response data, greatly improves the intelligentization and automation degree of the testing process, and provides reliable data support for system evaluation and algorithm optimization. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the specific embodiments or the prior art of the present application, the following will briefly introduce the drawings needed to be used in the specific embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 is a flow chart of an outdoor rainfall simulation control method for autonomous vehicle testing provided by an embodiment of the present application. DETAILED DESCRIPTION

[0016] The exemplary embodiments of the present application are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present application to help understanding. They should be considered only as exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Also, for the sake of clarity and conciseness, the description below omits the description of well-known functions and structures.

[0017] The present application is further described in detail below in conjunction with the embodiments.

[0018] Figure 1 is a flow chart of an outdoor rainfall simulation control method for autonomous vehicle testing provided by an embodiment of the present application. The method can be executed by an outdoor rainfall simulation control system for autonomous vehicle testing. As shown in Figure 1 The present embodiment provides an outdoor rainfall simulation control method for autonomous vehicle testing, including the following steps: S110, obtaining target parameters of outdoor rainfall simulation, the target parameters at least including: target rainfall intensity, target duration and target rainfall angle.

[0019] The user can directly input the target parameters to the system through the human-computer interaction module, or input a test case to the system. The system loads and parses the test case to obtain the target parameters of outdoor rainfall simulation.

[0020] The current rainfall intensity is the rainfall per unit time, with the unit of mm / h. The current rainfall angle represents the inclination of rainfall, with the unit of angle, simulating the scene of rain plus wind. The target duration is the duration of rainfall, for example, 10 minutes.

[0021] In a specific embodiment, a specific rainfall mode is automatically triggered according to a test case to meet the test requirements under different weather scenarios. For example, the "heavy rain mode" can rapidly increase the target rainfall intensity from 1 mm / h to 25 mm / h in a short time to simulate a sudden rainstorm environment; and the "shower mode" can simulate an intermittent rainfall process by periodically switching the rainfall intensity between 5 mm / h and 10 mm / h to comprehensively verify the perception and decision-making capabilities of the autonomous driving system under dynamic rainfall conditions. It can be seen that the target parameter can be a fixed value or a set of time-varying values.

[0022] In S120, a fuzzy controller is used to obtain the adjustment amount of the water output of each of the plurality of spraying devices according to the difference between the target rainfall intensity and the current rainfall intensity and the change amount of the difference between the target rainfall intensity and the current rainfall intensity.

[0023] To achieve high-precision dynamic adjustment and stable control of the rainfall intensity in the outdoor rainfall simulation process, the embodiment introduces a closed-loop feedback control strategy based on fuzzy control theory. This method can effectively cope with interference factors caused by complex and variable outdoor environments (such as changes in wind speed, fluctuations in temperature and humidity, etc.), ensuring that the actual rainfall intensity always approximates the set target value.

[0024] Optionally, a fuzzy controller (which integrates a fuzzy control algorithm) is used to obtain the adjustment amount of the water output of each of the plurality of spraying devices according to the difference between the target rainfall intensity and the current rainfall intensity and the change amount of the difference between the target rainfall intensity and the current rainfall intensity. The fuzzy control algorithm is described in detail below.

[0025] First, the configuration of the spraying device is initialized according to the target rainfall intensity, the target rainfall angle, the three-dimensional terrain data of the rainfall area, and the surface material type. See the following formula: ; Wherein, is the surface height of the rainfall area at coordinate point . represents the quantized data of the surface material type of the rainfall area at coordinate point , such as asphalt, cement, green belt, etc., which affects the rainwater infiltration and water accumulation distribution. is the target rainfall intensity, is the target rainfall angle. is the initial configuration of the i-th spraying device, including the water output and the water output angle. The function is a preset quadratic function for mapping the relationship between the terrain, material, and target parameter to the initial configuration. The embodiment does not limit the specific form of the quadratic function, which is a function obtained in advance according to experiments / experience.

[0026] Next, the following control variables are defined as input and output variables of the fuzzy controller: Input variable one: the difference between the target rainfall intensity and the current rainfall intensity (error) ; The calculation formula is as follows: ; Wherein, is the target rainfall intensity (unit: mm / h), is the actual rainfall intensity collected by the rain sensor at the t time (unit: mm / h).

[0027] Input variable two: the change of the difference between the target rainfall intensity and the current rainfall intensity ; Reflect the trend of error change over time, used to predict future rainfall state, the calculation formula is as follows: ; Wherein, is the error of the previous time.

[0028] Output variable: the adjustment amount u of the water output of the i-th sprinkler at the current t time i , so as to affect the overall rainfall intensity: u i (t)∈[-1,+1]; In order to enable the fuzzy controller to handle the non-linear and highly uncertain rainfall simulation problem, the system adopts the language variable method to fuzzy process the input and output variables, and defines seven basic fuzzy sets, as follows:

[0029] These fuzzy sets correspond to different language descriptions of error, error change rate and adjustment amount, for example, when the error is "positive large", it means that the current rainfall intensity is much lower than the target value, and the water output needs to be increased significantly.

[0030] The core of fuzzy control lies in the design of fuzzy rule base. Based on expert experience and historical data training, the system constructs the following fuzzy control rules to form a complete control logic system. Each rule adopts the form of "IF... AND... THEN...", which is expressed as: IF <e is A> AND <Δe is B> THEN ; Where A and B are fuzzy sets of input variables e and Δe, respectively, and C is the fuzzy set of output variable u. Some typical rules are as follows: IF e is PB AND Δe is PB THEN u is PB; IF e is PM AND Δe is ZO THEN u is PM; IF e is ZO AND Δe is ZO THEN u is ZO; IF e is NS AND Δe is PS THEN u is ZO; The above rules indicate what adjustment strategies the fuzzy controller should adopt to approximate the target rainfall intensity under different deviations and their changing trends.

[0031] After completing the fuzzy rule reasoning, the system uses the centroid method (also known as the area center method) for defuzzification to obtain the specific control output value (i.e., the adjustment amount of the water output). Its mathematical expression is as follows: ; in, This indicates the number of activated fuzzy rules. Indicates the first The weights of each rule are calculated using a weighted average method based on rule matching degree. Specifically, only the three fuzzy rules with the highest matching degree to the current input variable are selected, ignoring the influence of other rules. The weight of each rule is determined by the similarity between its preconditions and the input variable, quantified using discrete values ​​of "complete match = 1, partial match = 0.5, no match = 0". The final control output is obtained by weighted averaging of the output values ​​of each selected rule, thereby achieving a reasonable approximation of the system behavior and control decision. Indicates the first The output value corresponding to each rule (i.e., the representative value of a certain fuzzy set). Through this method, the system can convert the fuzzy inference results into precise numerical signals for the spraying device to execute.

[0032] Finally, the fuzzy controller determines the output value u based on the defuzzified value. i (t) The water output of each spray device is adjusted in real time, and the update formula is as follows: ; in: Indicates the first The water output (L / min) of each nozzle at time t. This is the gain coefficient, typically ranging from [0.05, 0.2]. For the first a regulation amount of water output of the spray head at the current time t ([-1, 1]). If > 0, it indicates that the current rainfall intensity is low, and the water output needs to be increased; otherwise, the water output is reduced, so as to realize dynamic closed-loop regulation of the rainfall intensity.

[0033] Through the above-mentioned closed-loop rainfall regulation mechanism based on fuzzy control, the system can realize high-precision and self-adaptive control of the rainfall intensity in a complex outdoor environment, and significantly improve the authenticity and reliability of the automatic driving vehicle test.

[0034] S130, adjusting the water output of the spray device according to the regulation amount, adjusting the opening time of the spray device according to the target duration, and adjusting the water output angle of the spray device according to the target rainfall angle.

[0035] The water output, the target duration, and the target rainfall angle obtained in S120 are sent to the programmable control unit of each spray device to control the spray device to meet the aforementioned water output, the opening time to meet the target duration, and the water output angle to meet the target rainfall angle.

[0036] Optionally, the target parameters further include a ground water film thickness threshold, which is used to simulate local waterlogging phenomenon, and the unit is mm.

[0037] After obtaining the ground water film thickness threshold, the current thickness of the ground water film is determined through the image recognition device (such as a camera) deployed outdoors. According to the gap between the current thickness of the ground water film and the ground water film thickness threshold, the water output of the spray device and the drainage speed of the drainage device are controlled. For example, if the current thickness of the ground water film is greater than the ground water film thickness threshold, the drainage speed is increased and the water output is reduced. It should be noted that if the target rainfall intensity and the ground water film thickness threshold are set at the same time, the spray device needs to adjust the water output to meet the target rainfall intensity first; at this time, if there is a certain gap between the current thickness of the ground water film and the ground water film thickness threshold, the drainage speed of the drainage device is adjusted to meet the ground water film thickness threshold.

[0038] Optionally, the target parameters further include a rainfall simulation strategy for different regions. In order to more realistically restore the natural rainfall environment and simulate the diversified rainfall conditions faced by different regions in a complex road scene, this embodiment proposes a multi-region differentiated rainfall control strategy based on functional region division. This strategy allows the system to set independent target parameters for different road sections (such as straight roads, curved roads, intersections, and slopes) in the same test region, and realize dynamic regulation of the rainfall intensity and distribution of each sub-region.

[0039] ​First, according to the geographical information and road structure of the test area, the entire test area is divided into several sub-areas with specific functions. Each sub-area corresponds to a typical traffic scene and sets the corresponding rainfall simulation target. For example, the following table:

[0040] Through such fine division, the system can provide more challenging and realistic test environment for autonomous vehicles.

[0041] For example, if the target rainfall intensity of each sub-area is to be controlled separately, an independent fuzzy controller needs to be deployed for each sub-area. Each fuzzy controller obtains the adjustment amount of the water output of the spraying device in each sub-area according to the difference between the target rainfall intensity and the current rainfall intensity of each sub-area, and the change amount of the difference between the target rainfall intensity and the current rainfall intensity of each sub-area. For details, please refer to the description of the above embodiments, which will not be repeated here.

[0042] For example, the rainfall simulation strategy for the curve area includes the rainfall simulation strategy for the curve area. First, considering factors such as rainwater easily splashing horizontally and tire grip decreasing in the curve area, the system adopts an enhanced rainfall strategy in such areas. Based on this, during the simulation of rainfall, if the rainfall area is a curve area, the target rainfall intensity and the first correction coefficient are obtained to obtain the target curve rainfall intensity, as shown in the following formula: wherein, represents the target curve rainfall intensity, represents the first correction coefficient, the range is [0.1~0.3], and represents the proportion of the increase in the target rainfall intensity. The more curved the curve area is, the larger the first correction coefficient is. The system controls the spraying device in the curve area to spray water according to the corrected rainfall intensity in the curve area. Specifically, through the fuzzy controller of the curve area, the adjustment amount of the water output of each spraying device in the curve area is obtained according to the difference between the target curve rainfall intensity and the current rainfall intensity, and the change amount of the difference between the target curve rainfall intensity. At the same time, the system controls the spraying device to increase the water output outside the curve area to form a biased rainfall effect, so as to simulate the extreme situation that the vehicle may encounter in the wet and slippery curve.

[0043] For example, the rainfall simulation strategy for the slope area includes the rainfall simulation strategy for the slope area. In the slope area, in order to make the rainfall direction closer to the natural state, the system automatically adjusts the spraying angle according to the terrain slope. Based on this, during the simulation of rainfall, if the sub-area is a slope area, the water output angle of each spraying device in the slope area is determined according to the slope angle of the slope area and the second correction coefficient, as shown in the following formula: ;​ wherein, is the water outlet angle of each spray device in the ramp area, is the slope angle of the ramp area, is the second correction coefficient, and the value range is [0.1~0.3]. Through this method, the spray device sprays water according to the adjusted water outlet angle, and the spray flow direction can be adaptively adjusted according to the terrain change, thereby improving the authenticity of the rainfall simulation.

[0044] The rainfall simulation strategy of different sub-regions includes the rainfall simulation strategy of the intersection area. Due to poor drainage and other reasons, local water accumulation often occurs in the intersection area. Therefore, the system sets the minimum water film thickness requirement in this type of area to simulate the local water accumulation effect. Based on this, during the simulation of rainfall, the specific ground water film thickness threshold of the intersection area is determined according to the ground water film thickness threshold and the third correction coefficient. For example, the third correction coefficient is determined according to the degree of depression of the intersection area, and the specific ground water film thickness threshold is obtained by adding the third correction coefficient to the ground water film thickness threshold. The current thickness of the ground water film is determined by the image recognition device deployed outdoors, and the water outlet amount of the spray device and the drainage speed of the drainage device in the intersection area are controlled according to the difference between the current thickness of the ground water film and the specific ground water film thickness threshold, so as to meet the following formula: ; wherein, is the current thickness of the ground water film, is the difference between the specific ground water film thickness threshold, and the value range is set to is 2mm. The system realizes this effect by increasing the water outlet amount and appropriately slowing down the drainage speed.

[0045] Optionally, although the target parameters of each sub-region are independently set, the system still needs to ensure the coordination and consistency of the overall rainfall simulation process. For this purpose, each fuzzy controller exchanges real-time rainfall information of the sub-region, such as rainfall amount and rainfall angle, through the communication module. Each fuzzy controller shares global meteorological data, such as wind speed and humidity. In terms of control strategy, after obtaining the target parameters of outdoor rainfall simulation, each fuzzy controller does not directly control the spray device based on the target parameters of the corresponding sub-region, but considers the rainfall intensity of the adjacent region to realize smooth transition between regions and avoid abrupt switching. Specifically: each fuzzy controller smoothes the rainfall simulation strategy of the sub-region according to the real-time rainfall information of the adjacent sub-region. For example, the target rainfall intensity of the sub-region is 100, and the real-time rainfall intensity of the adjacent region is 80, so the target rainfall intensity is set to a linearly increasing sequence from 80 to 100, and the target rainfall intensity is increased from 80 to 100 in 10 minutes.

[0046] Optionally, the system also supports a priority scheduling mechanism. If it is determined according to the rainfall simulation strategy of different sub-regions that changes in target rainfall intensity occur in more than two target sub-regions at the same time, a high-risk area, such as a curve area, is selected from the target sub-regions according to driving risk. The water output of the sprinkler device in the high-risk area is controlled to reach the target rainfall intensity.

[0047] The above multi-region differentiated rainfall control strategy is used to achieve high-precision simulation of various rainfall scenarios in complex road environments, and significantly improve the test efficiency and safety evaluation level of the perception, decision-making and execution capabilities of the autonomous vehicle in rainy conditions.

[0048] The outdoor rainfall simulation control system for autonomous vehicle testing provided in the embodiment comprises: A human-computer interaction module is configured to obtain target parameters of outdoor rainfall simulation, wherein the target parameters at least include target rainfall intensity, target duration and target rainfall angle.

[0049] A processor is configured to use a fuzzy controller to obtain an adjustment amount of water output of each of the plurality of sprinkler devices according to a difference between the target rainfall intensity and the current rainfall intensity and a change amount of the difference between the target rainfall intensity and the current rainfall intensity.

[0050] A sprinkler controller is configured to adjust the water output of the sprinkler device according to the adjustment amount, adjust the opening time of the sprinkler device according to the target duration, and adjust the water output angle of the sprinkler device according to the target rainfall angle.

[0051] An environment modeling module is configured to obtain three-dimensional terrain data and ground surface material types of a test area. The three-dimensional terrain data of the test area is obtained by means of unmanned aerial vehicle aerial photography, LiDAR scanning, etc. A road topology model is established to label functional areas such as curves, slopes and intersections. Ground material information (such as asphalt, cement and green belts) is recorded. A meteorological sensor comprises an anemometer, a temperature and humidity sensor, a rain gauge and other devices. Real-time collected meteorological data includes wind speed, wind direction angle, environmental temperature, relative humidity and actual rainfall intensity.

[0052] A visual detection module comprises high-definition cameras and infrared imaging devices to collect rainfall information such as road surface water film thickness and reflection intensity in real time.

[0053] The system provided by the embodiment can be linked with an automatic driving test platform, a data interaction channel between an outdoor rainfall simulation control system and the automatic driving test platform is constructed through a standard communication protocol. The communication interface adopts an API interface module, supports bidirectional transmission of multiple data formats (such as JSON and XML), and ensures that the system can be seamlessly connected with mainstream automatic driving simulation platforms (such as CARLA and Prescan) and real vehicle test platforms (such as Autoware and Apollo).

[0054] After the communication connection is established, a user can directly set target parameters of outdoor rainfall simulation in the automatic driving test platform, including target rainfall intensity, target duration and target rainfall angle and the like. The rainfall simulation control system receives control instructions from the test platform, and automatically completes initialization configuration and starting process of the spraying device, thereby significantly improving the automation degree and operation convenience of the test process.

[0055] The data recording and test module in the outdoor rainfall simulation control system records rainfall information in real time during the rainfall simulation process, and obtains vehicle response data based on a data synchronization mechanism with the automatic driving test platform; and generates an automatic driving test report based on the rainfall information and the vehicle response data. Specifically, time stamps, rainfall intensity change curves, spatial distribution characteristics and other environmental data of rainfall events are collected and stored in real time. At the same time, through the data synchronization mechanism with the automatic driving test platform, speed, steering angle, braking state, perception recognition results and other key response information of the vehicle are obtained. All data are uniformly labeled and stored in a local or cloud database, which is convenient for subsequent playback, comparison and analysis. After the test is completed, the system can automatically generate a structured test report, which records test time, rainfall mode type, rainfall parameter change curve, vehicle behavior performance and key event nodes and the like in detail. The report supports export in PDF, Excel or JSON format, which is convenient for test personnel to review and evaluate, and also provides strong data support for performance optimization and algorithm iteration of the automatic driving system.

[0056] The method provided by the present application will be described in detail through three specific embodiments as follows: Specific Embodiment 1: Basic Rainfall Simulation Control

[0057] In a closed automatic driving test field, a user wants to simulate a moderate rain (rainfall amount is about 5 mm / h) with a duration of 30 minutes. The test area is a flat road surface, mainly used to verify the stability of automatic driving vehicles in identifying road signs and keeping lanes under general rainy conditions.

[0058] The system first collects topographic data of the test area and imports it into the system for modeling; then, the user sets parameters such as target rainfall intensity, target rainfall angle, and target duration on the control terminal. The system automatically calculates the initial parameters of the sprinkler device according to the set values combined with site information and starts the equipment.

[0059] During the rainfall simulation process, the system collects real-time feedback data such as current rainfall intensity, water film thickness, and road surface reflection conditions through rain sensors and image recognition modules deployed in the site. Based on the fuzzy control algorithm, the water flow and water angle of each nozzle are continuously adjusted to ensure that the rainfall process always approximates the set target.

[0060] After the rainfall ends, the system automatically generates a rainfall process log and turns off all the sprinkler devices. The entire test process achieves high-precision reproduction of the rainfall environment, providing reliable data support for the perception and decision-making capabilities of the autonomous driving system.

[0061] Specific embodiment 2: complex scene linkage test.

[0062] In another test case, the autonomous vehicle needs to pass through a section of curved area while encountering side wind accompanied by heavy rain. This scenario is used to evaluate the vehicle's path tracking ability and emergency braking response on a slippery curve.

[0063] The system loads the "heavy rain + side wind" combination mode according to the preset test case (with specific parameters bound, no need to set / update again). The sprinkler device cooperates with the wind simulation equipment to create a rainfall effect with a certain inclination angle, enhancing the feeling of rainwater splashing horizontally, simulating the real curve driving environment.

[0064] When the autonomous vehicle enters the test area, the system records key indicators such as braking distance, steering response, and lane deviation warning trigger frequency on the slippery road surface in real time. At the same time, the system continuously optimizes the sprinkler strategy through reinforcement learning mechanism, making the rainfall simulation more close to the set target and improving the credibility of the test results.

[0065] During the entire test process, the rainfall simulation system and the autonomous driving test platform maintain synchronous communication, which can dynamically adjust the rainfall intensity and direction according to the vehicle position, realize the true meaning of "on-demand rainfall", and effectively improve the authenticity and controllability of the test.

[0066] Specific embodiment 3: multi-region differentiated rainfall simulation.

[0067] In the city road comprehensive test field, the test personnel hope to simulate a composite rainfall environment containing multiple typical traffic scenes, including straight roads, curves, intersections, and slopes, to comprehensively verify the response ability of the autonomous vehicle in complex weather conditions.

[0068] The system divides the entire test area into several sub-areas, and sets independent rainfall parameters for each sub-area. For example, in the curve area, the lateral rainfall intensity is enhanced to simulate the risk of tire skidding; in the intersection area, the local water accumulation effect is simulated to increase the road reflection and visual interference; in the slope area, the spray angle is adjusted to be more consistent with the direction of natural rainfall.

[0069] Each sub-area spray device independently runs the fuzzy control strategy, and adjusts in a closed loop according to local meteorological sensor data and image feedback, to ensure uniform and stable rainfall intensity distribution. At the same time, the system coordinates the transition between regions through the communication transmission module, to avoid sudden changes or discontinuities in rainfall.

[0070] In addition, the system is also linked with the automatic driving test platform, and can automatically match the corresponding rainfall mode when the vehicle switches the driving area, to realize seamless multi-scene test experience. This embodiment significantly improves the adaptability of the automatic driving system in diversified road environments and the safety evaluation level.

[0071] It should be understood that the various forms of flow shown above can be reordered, added or deleted steps. For example, each step described in the present application can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions disclosed in the present application can be achieved, which is not limited herein.

[0072] The above specific embodiments do not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. An outdoor rainfall simulation control method for testing autonomous vehicles, characterized in that, include: Obtain target parameters for outdoor rainfall simulation, the target parameters including at least: target rainfall intensity, target duration, and target rainfall angle; A fuzzy controller is used to obtain the adjustment amount of the water output of each of the multiple sprinkler devices based on the difference between the target rainfall intensity and the current rainfall intensity, and the change in the difference between the target rainfall intensity and the current rainfall intensity. The water output of the spray device is adjusted according to the adjustment amount, the opening time of the spray device is adjusted according to the target duration, and the water outlet angle of the spray device is adjusted according to the target rainfall angle.

2. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 1, characterized in that, The target parameters also include: a threshold for the thickness of the ground water film; The method further includes: The current thickness of the water film on the ground is determined by image recognition devices deployed outdoors; Based on the difference between the current thickness of the ground water film and the ground water film thickness threshold, the water output of the spray device and the drainage speed of the drainage device are controlled.

3. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 1, characterized in that, The target parameters also include: rainfall simulation strategies for different sub-regions; the rainfall simulation strategies include target rainfall intensity; A fuzzy controller is used to determine the adjustment amount of the water output of each of the multiple sprinkler systems based on the difference between the target rainfall intensity and the current rainfall intensity, and the change in the difference between the target rainfall intensity and the current rainfall intensity. This adjustment includes: An independent fuzzy controller is configured for each sub-region. The adjustment amount of the water output of the sprinkler device in each sub-region is obtained based on the difference between the target rainfall intensity and the current rainfall intensity in each sub-region, and the change in the difference between the target rainfall intensity and the current rainfall intensity in each sub-region.

4. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 3, characterized in that, Rainfall simulation strategies for different sub-regions include rainfall simulation strategies for bend areas; The method employs a fuzzy controller to obtain the adjustment amount of the water output of each of the multiple sprinkler devices based on the difference between the target rainfall intensity and the current rainfall intensity, and the change in the difference between the target rainfall intensity and the current rainfall intensity. This includes: The target rainfall intensity at the bend is obtained based on the target rainfall intensity and the first correction factor; By using a fuzzy controller in the bend area, the adjustment amount of the water output of each sprinkler device in the bend area is obtained based on the difference between the target bend rainfall intensity and the current rainfall intensity, and the change in the target bend rainfall intensity difference.

5. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 3, characterized in that, Rainfall simulation strategies for different sub-regions include rainfall simulation strategies for slope areas; The method further includes: The water outlet angle of each sprinkler in the slope area is determined based on the slope angle and the second correction factor.

6. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 3, characterized in that, Rainfall simulation strategies for different sub-regions include rainfall simulation strategies for intersection areas; The method further includes: The specific ground water film thickness threshold for the intersection area is determined based on the ground water film thickness threshold and the third correction coefficient. The current thickness of the water film on the ground is determined by image recognition devices deployed outdoors; Based on the difference between the current thickness of the ground water film and a specific ground water film thickness threshold, the water output of the sprinkler system and the drainage speed of the drainage system in the intersection area are controlled.

7. The outdoor rainfall simulation control method for autonomous vehicle testing according to claim 3, characterized in that, Each fuzzy controller exchanges real-time rainfall information for its sub-region via a communication module. After obtaining the target parameters for outdoor rainfall simulation, the following steps are also included: Each fuzzy controller smooths the rainfall simulation strategy for its own sub-region based on real-time rainfall information from neighboring sub-regions; If, based on the rainfall simulation strategies for different sub-regions, it is determined that two or more target sub-regions simultaneously experience changes in target rainfall intensity; Based on the driving hazard, select high-risk areas from the target sub-regions; Prioritize controlling the water output of sprinkler systems in high-risk areas to achieve the target rainfall intensity.

8. The outdoor rainfall simulation control method for testing autonomous vehicles according to any one of claims 1-7, characterized in that, Obtain the target parameters for outdoor rainfall simulation, including: Load test cases; The test cases were analyzed to obtain the target parameters for outdoor rainfall simulation.

9. An outdoor rainfall simulation control system for autonomous vehicle testing, characterized in that, include: The human-computer interaction module is used to acquire target parameters for outdoor rainfall simulation, the target parameters including at least: target rainfall intensity, target duration and target rainfall angle; The processor is used to employ a fuzzy controller to obtain the adjustment amount of the water output of each of the multiple sprinkler devices based on the difference between the target rainfall intensity and the current rainfall intensity, and the change in the difference between the target rainfall intensity and the current rainfall intensity. The sprinkler controller adjusts the water output of the sprinkler device according to the adjustment amount, adjusts the start time of the sprinkler device according to the target duration, and adjusts the water output angle of the sprinkler device according to the target rainfall angle.

10. The outdoor rainfall simulation control system for autonomous vehicle testing according to claim 9, characterized in that, Also includes: The environment modeling module is used to acquire 3D terrain data and surface material types of the test area; Weather sensors are used to collect weather data in real time; A visual inspection module is used to collect rainfall information in real time; The data recording and testing module is used to record rainfall information in real time during rainfall simulation and to acquire vehicle response data based on the data synchronization mechanism with the autonomous driving test platform; and to generate an autonomous driving test report based on the rainfall information and vehicle response data.

Images