Automobile radar performance optimization system and method in extreme weather

Abstract

The invention discloses an automobile radar performance optimization system and method in extreme weather, and the system comprises an environment sensing module which is used for collecting multi-dimensional environment data, and recognizing a weather type based on a multi-modal fusion interference classification model; the multiband radar array comprises a millimeter wave radar and a terahertz radar, and supports the dynamic adjustment of the proportion of millimeter wave and terahertz frequency bands according to the weather type and the driving scene; the intelligent signal processing unit is used for carrying out preprocessing, interference classification and signal compensation on the data; the main and auxiliary cooperative control unit is used for dynamically allocating roles of a main radar and an auxiliary radar in the multiband radar array according to the driving scene, and the auxiliary radar calibrates the main radar; the self-adaptive cleaning system is used for cleaning, antifogging and deicing the radar sensing surface; and the central control module generates a control instruction based on the weather type and the driving scene to realize system cooperative control. The environmental adaptability, the detection reliability and the system stability of the automobile radar under complex and severe weather conditions can be improved.

Description

Technical Field

[0001] This application relates to the field of intelligent radar systems, and in particular to a system and method for optimizing the performance of automotive radar under extreme weather conditions. Background Technology

[0002] With the rapid development of autonomous driving and advanced driver assistance systems (ADAS) technologies, automotive environmental perception systems, especially radar systems, have become core components for ensuring driving safety. However, in extreme weather conditions such as rain, snow, fog, and sandstorms, the detection performance of existing automotive radars will significantly decrease, severely restricting their reliable application in all-weather scenarios.

[0003] Currently, at the environmental perception level, common solutions rely on single rain sensors or simple visual algorithms for weather assessment. These methods struggle to accurately distinguish between weather types with vastly different physical characteristics. For example, they cannot effectively identify the different interference patterns caused by heavy rain, dense fog, and sandstorms on sensor signals, resulting in a lack of targeted response strategies for subsequent signal processing and decision-making systems.

[0004] In radar signal processing, the mainstream solutions are still based on the frequency-modulated continuous wave (FMCW) signal system, combined with conventional filtering algorithms (such as mean filtering and Kalman filtering) to suppress noise. These methods are effective for ideal Gaussian noise distribution, but their suppression capability is limited for complex meteorological noise with non-Gaussian distribution generated in extreme weather (such as random scattering of raindrops and Mie scattering of fog particles), which can easily lead to a sharp drop in signal-to-noise ratio and an increase in false alarm and missed detection rates.

[0005] In terms of sensor configuration and collaboration, existing technologies mostly adopt a fixed primary and secondary radar division of labor or a simple data overlay strategy. This static collaboration method cannot adapt to dynamically changing driving scenarios (such as vehicle steering and lane changes), and is prone to instantaneous detection blind spots when switching scenarios. In addition, multi-sensor fusion algorithms mostly rely on preset fixed weights and lack a mechanism for dynamic adjustment based on the reliability of the real-time environment, making it difficult to achieve performance complementarity between different sensors under extreme weather conditions.

[0006] Furthermore, cleaning and maintaining radar sensors is a significant technical challenge. Existing cleaning devices typically rely on dedicated cleaning fluids and pump / valve systems, which not only consume additional onboard resources but also pose a risk of the cleaning fluid freezing and causing system failure in low-temperature environments. Simultaneously, the lack of an effective integrated anti-fogging and de-icing design means that the sensor's optical or radio frequency window surfaces are prone to performance degradation due to contaminant buildup or frost condensation, further exacerbating the decline in perception capabilities under extreme weather conditions. Summary of the Invention

[0007] The purpose of this invention is to provide a system and method for optimizing the performance of automotive radar under extreme weather conditions, which can improve the environmental adaptability, detection reliability and system stability of automotive radar under complex and severe weather conditions.

[0008] The embodiments of the present invention are implemented as follows: In a first aspect, the vehicle radar performance optimization system under extreme weather conditions in this embodiment includes: The environmental perception module is used to collect multi-dimensional environmental data and identify weather types based on a multi-modal fusion interference classification model. The interference classification model is a Transformer-based model. Its input layer integrates data from the rain sensor, humidity sensor, camera, and lidar, learns cross-modal associations through a multi-head attention mechanism, and the output layer uses the Softmax function to identify weather types.

[0009] A multi-band radar array, comprising at least one set of millimeter-wave radar and one set of terahertz radar, wherein the multi-band radar array supports dynamic adjustment of the ratio of millimeter-wave to terahertz frequency bands according to the weather type and driving scenario; The intelligent signal processing unit is used to preprocess, classify interference, and compensate the data from the environmental perception module and the multi-band radar array. The main and auxiliary coordination control unit is used to dynamically allocate the roles of the main radar and auxiliary radar in the multi-band radar array according to the driving scenario, and the auxiliary radar calibrates the main radar. An adaptive cleaning system is used to clean, defog, and de-ice radar sensor surfaces; The central control module is connected to the environmental perception module, the intelligent signal processing unit, the main and auxiliary collaborative control unit, and the adaptive cleaning system, respectively, and is used to generate control commands based on weather type and driving scenario to achieve coordinated control of the entire system.

[0010] In a possible implementation, the multi-dimensional environmental data collected by the environmental perception module includes one or more combinations of the following four types of data: Rainfall intensity, raindrop size, and frequency collected by a rain sensor; relative humidity and dew point temperature collected by a humidity sensor; environmental visibility, atmospheric scattering coefficient, and weather image features collected by a camera, including snowflake morphology and fog distribution; and point cloud noise density, particulate matter trajectory, and echo intensity attenuation rate collected by a lidar.

[0011] In a possible implementation, the interference classification model is a Transformer-based model whose input layer integrates data from the rain sensor, humidity sensor, camera, and lidar, and identifies weather types through a multi-head attention mechanism.

[0012] In a possible implementation, the master-slave collaborative control unit is configured as follows: In scenarios where vehicles are traveling straight, the medium-sized radar is the primary radar. In a scenario where the vehicle is turning left, switch the left radar to become the primary radar. In a scenario where the vehicle is turning right, switch the right radar to become the primary radar. The auxiliary radar performs real-time calibration of the main radar's detection results at a period of no more than 100ms.

[0013] In a possible implementation, the adaptive cleaning system includes: Rainwater harvesting unit for collecting and filtering rainwater; Thermoelectric cooling elements are used to prevent fogging on radar sensor surfaces; Ultrasonic vibration devices are used to remove ice and stubborn stains from radar sensor surfaces.

[0014] In a possible implementation, the central control module is configured to construct a "weather-scene-parameter" mapping matrix and dynamically configure at least one of the following parameters based on this matrix: Radar's transmit power, scan period, and frequency band percentage; CFAR detection threshold in signal processing; Trigger thresholds and operating modes of adaptive cleaning systems.

[0015] In a possible implementation, the system also includes a predictive maintenance module for analyzing historical sensor data to predict performance degradation trends and triggering the calibration process in advance.

[0016] Secondly, this application also discloses a method for optimizing the performance of automotive radar under extreme weather conditions, including the following steps: Collect multi-dimensional environmental data through multiple sensors; The environmental data is processed using a multimodal fusion-based interference classification model to identify the current weather type; Based on the weather type and real-time driving scenario, the operating parameters of the multi-band radar array are dynamically adjusted, including dynamically adjusting the ratio of millimeter wave and terahertz frequency bands. Based on the driving scenario, the roles of the main radar and auxiliary radar in the multi-band radar array are dynamically switched, and the auxiliary radar data is used to calibrate the main radar data. When the contamination level on the radar sensor surface exceeds the threshold, a multimodal cleaning process is initiated.

[0017] In a possible implementation, the dynamic adjustment of the operating parameters of the multi-band radar array further includes: Increase the transmission power of millimeter-wave radar during heavy rain; In extreme weather conditions, the radar scanning cycle can be shortened. The frequency band ratio of millimeter-wave radar and terahertz radar is adjusted according to vehicle speed.

[0018] In a possible implementation, initiating the multimodal cleaning process includes: Activate high-pressure jet cleaning to utilize collected rainwater for cleaning; Simultaneously activate the thermoelectric cooling element to prevent fogging; Simultaneously activate the ultrasonic vibration device to remove ice.

[0019] The embodiments of this application have the following beneficial effects: (1) Through multi-dimensional and fine-grained environmental perception and Transformer-based intelligent classification algorithm, the accurate identification and classification of extreme weather can be achieved, and the weather identification accuracy rate can be increased to more than 95%, which provides a reliable premise for subsequent optimization and significantly improves environmental adaptability.

[0020] (2) By adopting multi-band radar dynamic coordination and adaptive signal processing technology, the effective detection distance in rainy weather is extended from 80 meters to more than 120 meters, and the signal-to-noise ratio is improved by 40%; the target recognition accuracy is maintained at more than 90% in foggy weather, and the false alarm rate is reduced to less than 0.3 times / thousand kilometers.

[0021] (3) Through dynamic switching of main and auxiliary radars and intelligent parameter configuration, the detection reliability in complex scenarios such as turning and lane changing is improved by 23%, the system response time is shortened to 0.2 seconds, the detection blind zone is effectively eliminated, and the system has good scene adaptability.

[0022] (4) The adaptive cleaning system utilizes rainwater resources, reducing the consumption of special cleaning fluid by up to 60%; the predictive maintenance mechanism reduces the sensor failure rate by more than 30%, extending the system's service life.

[0023] (5) It is compatible with existing millimeter-wave radar and lidar hardware platforms. The performance can be greatly improved by upgrading the software algorithm, which facilitates the technical iteration and OTA upgrade of mass-produced models. The system has strong compatibility and iterability. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and therefore should not be considered as a limitation on the scope of protection of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart of a method for optimizing the performance of automotive radar under extreme weather conditions is shown. Detailed Implementation

[0026] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0027] The components of the embodiments of this application described and illustrated in the accompanying drawings can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0028] In the following, the terms “comprising,” “having,” and their cognates, which may be used in various embodiments of this application, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as excluding, firstly, the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more features, numbers, steps, operations, elements, components, or combinations thereof.

[0029] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0030] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of this application pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be construed as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of this application.

[0031] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0032] This application provides a system and method for optimizing automotive radar performance under extreme weather conditions based on multimodal fusion. Addressing the radar performance degradation problem in extreme weather, a four-level linkage optimization mechanism of "environmental perception - dynamic adjustment - multi-source fusion - intelligent cleaning" is constructed. Through deep collaboration and feedback among the modules, the global optimal system performance can be achieved. The core connections and collaborative working mechanisms among the modules are as follows: The environmental perception module outputs parameter information (weather type and level) to the central control module. Based on this information, the central control module issues control commands to the multi-band radar array, intelligent signal processing unit, and adaptive cleaning system, forming a primary closed loop of "perception-decision-execution", which can realize the closed loop of perception and decision-making.

[0033] The central control module also acquires real-time vehicle CAN bus information (such as vehicle speed and steering angle). This upgrades system decision-making from a purely "weather-driven" approach to a "weather-scenario dual-driven" approach. For example, steering signals trigger the main and auxiliary collaborative control units to switch roles; vehicle speed signals affect radar parameter configuration and the working sequence of the cleaning system, achieving dynamic adaptation between scenarios and execution.

[0034] The intelligent signal processing unit not only processes the raw radar signal, but its output interference classification results (from the improved Transformer model) and compensated high-quality data, in turn, provide the environmental perception module with more accurate weather classification and provide reliable input for the data calibration of the master-slave collaborative control unit, forming a fine-tuning closed loop within the system.

[0035] As an independent execution unit, the adaptive cleaning system is triggered and controlled by both the environmental perception module (such as the degree of contamination detected by the camera) and the central control module (which determines the cleaning timing based on vehicle speed, acceleration, etc.), ensuring the cleanliness and reliability of the entire system's perception foundation.

[0036] Through the above four interconnected closed loops, this invention achieves comprehensive and adaptive optimization from macroscopic environment to microscopic signal, and from hardware parameters to software algorithms.

[0037] Specifically, this embodiment provides a vehicle radar performance optimization system under extreme weather conditions. The system includes an environmental perception module, a multi-band radar array, an intelligent signal processing unit, a main / auxiliary cooperative control unit, an adaptive cleaning system, and a central control module.

[0038] The environmental perception module consists of a rain sensor, a humidity sensor, a camera, and a lidar. The rain sensor, mounted on the wiper mechanism, has a sampling frequency of at least 10Hz and is used to collect precipitation intensity, raindrop size, and frequency. The camera is equipped with a polarized filter lens to acquire environmental visibility, atmospheric scattering coefficient, and weather image features (such as snowflake morphology and fog distribution). The lidar is used to acquire point cloud noise density, particle trajectory, and echo intensity attenuation rate. These data, after preprocessing, are input to the intelligent signal processing unit. The intelligent signal processing unit includes a preprocessing module, an interference classification model, and a compensation algorithm. The preprocessing module uses adaptive median filtering (dynamically adjusting the window size from 3×3 to 7×7) combined with wavelet threshold denoising to process the raw data, increasing the effective point cloud retention rate of the lidar to over 90%.

[0039] The interference classification model is based on an improved Transformer architecture. The input layer integrates all the aforementioned environmental perception data; the model employs an 8-layer Transformer encoder with an 8-head attention mechanism (64 dimensions per head) to learn cross-modal associations; the output layer uses a Softmax function for classification. This model was trained using over 100,000 extreme weather samples, achieving a classification accuracy of over 95% for rain, snow, fog, and dust storms.

[0040] The compensation algorithm dynamically adjusts the CFAR detection threshold based on the classification results (reducing it by 10%-15% in rain / snow and increasing it by 5%-10% in fog); for millimeter-wave radar, the Doppler frequency shift compensation algorithm eliminates the velocity error caused by raindrop scattering, with a compensation accuracy of ≤0.5km / h; for lidar, the distance measurement value is corrected based on the atmospheric attenuation model to ensure that the distance compensation error in fog is ≤1m.

[0041] The multi-band radar array includes millimeter-wave radar and terahertz radar. The central control module dynamically adjusts the frequency ratio of these two types of radar based on weather type output from the intelligent signal processing unit and vehicle speed information acquired via the CAN bus, using beamforming technology in 5% increments. For example, when the vehicle speed is below 30 km / h, the terahertz radar ratio is increased to 60% to enhance near-range detection capabilities; when the vehicle speed is above 80 km / h, the millimeter-wave radar power is increased by 20%, extending the detection range to 200 meters.

[0042] The central control module determines the switching between the main and auxiliary radars based on the steering angle signal: when driving straight, the central radar is the primary radar; when turning left, the left radar becomes the primary radar, with the central and right radars as secondary radars; when turning right, the right radar becomes the primary radar, with the central and left radars as secondary radars. The auxiliary radar performs real-time calibration of the main radar's detection results with a period of 100ms (shortened to 50ms when the vehicle speed is >60km / h), achieving a calibration accuracy of 0.1m, effectively eliminating blind spots during steering.

[0043] The adaptive cleaning system includes a rainwater harvesting unit, a thermoelectric cooling coil, and an ultrasonic vibration device (frequency 20-40kHz). The rainwater harvesting unit collects rainwater flowing down the windshield, filters it, and stores it. When the camera detects that the radar surface is contaminated with more than 5% of the obstructed area, the system initiates high-pressure jet cleaning. Simultaneously, the thermoelectric cooling coil operates to prevent fogging, and the ultrasonic vibration device activates to remove ice and stubborn stains. The central control module intervenes based on vehicle speed information; for example, it delays cleaning when the vehicle speed is >100km / h to prevent fogging, and pauses cleaning during rapid acceleration or deceleration to prevent spray deviation.

[0044] The central control module acquires vehicle status information via the CAN bus and constructs a "weather-scene-parameter" mapping matrix, enabling coordinated control of all the aforementioned modules. Furthermore, the module incorporates predictive maintenance functionality, analyzing historical sensor data to anticipate performance degradation trends and triggering calibration processes in advance to prevent problems before they occur.

[0045] The working principle of this invention is as follows: After the system starts, the environmental perception module continuously collects data, and the intelligent signal processing unit classifies the weather and compensates for the signal in real time. Based on the weather conditions and vehicle status (vehicle speed, steering), the central control module, on the one hand, instructs the multi-band radar array and the main-auxiliary collaborative control unit to adjust their operating modes for optimal detection; on the other hand, it monitors the cleanliness of the sensors and intelligently activates the cleaning system when necessary. The entire process forms a dynamic, adaptive, optimized closed loop, ensuring that the radar system maintains optimal performance under any extreme weather and driving scenario.

[0046] This application also provides a method for optimizing vehicle radar performance under extreme weather conditions, referencing... Figure 1 This includes the following steps: Step S101: Collect multi-dimensional environmental data, that is, synchronously collect raw environmental data through multiple sensors deployed on the vehicle. Specifically, this includes: Activate the rain sensor and collect precipitation intensity, raindrop size, and frequency at a sampling frequency of no less than 10Hz.

[0047] Activate the humidity sensor to collect relative humidity and dew point temperature.

[0048] The camera equipped with a polarizing filter lens is activated to acquire environmental visibility, atmospheric scattering coefficient, and weather image features including snowflake patterns and fog particle distribution.

[0049] Start the lidar to obtain point cloud noise density, particle motion trajectory and echo intensity attenuation rate.

[0050] Step S102: Weather Recognition Based on Multimodal Fusion. All multi-dimensional environmental data collected in Step S101 are input into a pre-trained interference classification model based on the Transformer architecture. The input layer of this model fuses multi-source heterogeneous data. Through its 8-layer encoder's multi-head attention mechanism (8 heads, 64 dimensions), the model automatically learns cross-modal associations between data from different sensors (e.g., associating high humidity with a white haze feature in an image to identify fog). Finally, the model's output layer uses the Softmax function to perform fine-grained classification of weather types (such as heavy rain, light snow, dense fog, dust storms, etc.), achieving an accuracy of over 95% and outputting weather levels from H0 to H4.

[0051] Step S103: Dynamically adjust radar parameters based on weather and scenario. The central control module receives the weather recognition result from step S102 and simultaneously acquires driving scenario information such as vehicle speed and steering angle in real time via the vehicle's CAN bus. Based on the built-in "weather-scenario-parameter" mapping matrix, control commands are generated, specifically including: Dynamic frequency band allocation adjustment: Based on weather type and vehicle speed, the frequency band energy distribution between millimeter-wave radar and terahertz radar is dynamically adjusted. For example, when the vehicle speed is below 30 km / h and in foggy weather, the multi-band radar array is instructed to increase the terahertz radar's allocation to 60% to enhance close-range imaging accuracy; when the vehicle speed is above 80 km / h and in heavy rain, the transmission power of the millimeter-wave radar is increased by 20% to extend the maximum detection range to 200 meters. The adjustment step size is accurate to 5%.

[0052] Adjust signal processing parameters: Dynamically adjust the threshold of the CFAR detection algorithm in the intelligent signal processing unit according to the weather type. For example, lower the threshold by 10%-15% in rainy or snowy weather to capture weak targets, and raise the threshold by 5%-10% in foggy weather to suppress clutter.

[0053] Step S104: Dynamically switch and calibrate the main and auxiliary radars according to the driving scenario. The main-auxiliary coordination control unit executes a dynamic switching strategy for the main and auxiliary radar roles based on the real-time steering angle signal, specifically including: When the steering angle signal indicates that the vehicle is going straight, the central radar is set as the main radar, and the left and right radars are auxiliary radars.

[0054] When the steering angle signal indicates that the vehicle is turning left, immediately switch the left radar to the main radar and the middle and right radars to the auxiliary radars.

[0055] When the steering angle signal indicates that the vehicle is turning right, immediately switch the right radar to the main radar and the center and left radars to the auxiliary radars.

[0056] During this process, the auxiliary radar performs real-time calibration of the main radar's detection results at a period of no more than 100ms (reduced to 50ms when driving at high speeds), and uses algorithms such as RANSAC to remove abnormal point clouds, ensuring that the calibration accuracy reaches 0.1m and effectively eliminating steering blind spots.

[0057] Step S105: Intelligently trigger and execute multimodal cleaning. The adaptive cleaning system continuously monitors the condition of the radar-sensing surface: When the radar surface contamination level is determined to exceed a threshold (e.g., 5% obstruction area) through camera image analysis, the system is immediately triggered and initiates a multimodal cleaning process: High-pressure jetting: Start the high-pressure water pump and use the filtered rainwater stored in the rainwater collection unit to spray and rinse the sensor surface.

[0058] Synchronous anti-fog: The thermoelectric cooling element is activated immediately to actively control the temperature of the sensor window and prevent fogging on the surface due to temperature differences.

[0059] Simultaneous de-icing: Simultaneously activate the ultrasonic vibration device (frequency 20-40kHz) to break up and remove thin ice or stubborn stains on the surface through high-frequency mechanical vibration.

[0060] The timing of the entire cleaning process is intelligently managed by the central control module. For example, it delays execution when the vehicle speed is >100km / h to prevent obstruction of vision, and pauses execution during rapid acceleration and deceleration to prevent spray deviation.

[0061] This application also provides a computer device, exemplary of which includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to enable the computer device to perform the functions of the methods described above or the various modules in the systems described above.

[0062] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a Central Processing Unit (CPU), Graphics Processing Unit (GPU), Network Processor (NP), Digital Signal Processor (DSP), Application-Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0063] The memory can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory is used to store computer programs, and the processor can execute the computer programs accordingly after receiving execution instructions.

[0064] This application also provides a computer-readable storage medium for storing the computer program used in the aforementioned computer device. For example, the computer-readable storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0065] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, in alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0066] In addition, the functional modules or units in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0067] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0068] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A vehicle radar performance optimization system under extreme weather conditions, characterized in that, include: The environmental perception module is used to collect multi-dimensional environmental data and identify weather types based on a multi-modal fusion interference classification model; A multi-band radar array, comprising at least one set of millimeter-wave radar and one set of terahertz radar, wherein the multi-band radar array supports dynamic adjustment of the ratio of millimeter-wave to terahertz frequency bands according to the weather type and driving scenario; The intelligent signal processing unit is used to preprocess, classify interference, and compensate the data from the environmental perception module and the multi-band radar array. The main and auxiliary coordination control unit is used to dynamically allocate the roles of the main radar and auxiliary radar in the multi-band radar array according to the driving scenario, and the auxiliary radar calibrates the main radar. An adaptive cleaning system is used to clean, defog, and de-ice radar sensor surfaces; The central control module is connected to the environmental perception module, the intelligent signal processing unit, the main and auxiliary collaborative control unit, and the adaptive cleaning system, respectively, and is used to generate control commands based on weather type and driving scenario to achieve system collaborative control.

2. The system according to claim 1, characterized in that, The multi-dimensional environmental data includes: Rainfall intensity, raindrop size, and frequency were collected by a rain sensor; Environmental visibility, atmospheric scattering coefficient, and weather image characteristics captured by cameras; Point cloud noise density, particle motion trajectory, and echo intensity attenuation rate collected by lidar.

3. The system according to claim 2, characterized in that, The interference classification model is a Transformer-based model. Its input layer integrates data from the rain sensor, humidity sensor, camera, and lidar, and uses a multi-head attention mechanism to identify weather types.

4. The system according to claim 1, characterized in that, The primary and secondary collaborative control unit is configured as follows: In scenarios where vehicles are traveling straight, the medium-sized radar is the primary radar. In a scenario where the vehicle is turning left, switch the left radar to become the primary radar. In a scenario where the vehicle is turning right, switch the right radar to become the primary radar. The auxiliary radar performs real-time calibration of the main radar's detection results at a period of no more than 100ms.

5. The system according to claim 1, characterized in that, The adaptive cleaning system includes: Rainwater harvesting unit for collecting and filtering rainwater; Thermoelectric cooling elements are used to prevent fogging on radar sensor surfaces; Ultrasonic vibration devices are used to remove ice and stubborn stains from radar sensor surfaces.

6. The system according to claim 1, characterized in that, The central control module is configured to construct a "weather-scene-parameter" mapping matrix and dynamically configure at least one of the following parameters based on this mapping matrix: Radar's transmit power, scan period, and frequency band percentage; CFAR detection threshold in signal processing; Trigger thresholds and operating modes of adaptive cleaning systems.

7. The system according to claim 1, characterized in that, The system also includes a predictive maintenance module, which analyzes historical sensor data to predict performance degradation trends and triggers the calibration process in advance.

8. A method for optimizing the performance of automotive radar under extreme weather conditions, characterized in that, Includes the following steps: Collect multi-dimensional environmental data through multiple sensors; The environmental data is processed using a multimodal fusion-based interference classification model to identify the current weather type; Based on the weather type and real-time driving scenario, the operating parameters of the multi-band radar array are dynamically adjusted, including dynamically adjusting the ratio of millimeter wave and terahertz frequency bands. Based on the driving scenario, the roles of the main radar and auxiliary radar in the multi-band radar array are dynamically switched, and the auxiliary radar data is used to calibrate the main radar data. When the contamination level on the radar sensor surface exceeds the threshold, a multimodal cleaning process is initiated.

9. The method according to claim 8, characterized in that, The dynamic adjustment of the operating parameters of the multi-band radar array also includes: Increase the transmission power of millimeter-wave radar during heavy rain; In extreme weather conditions, the radar scanning cycle can be shortened. The frequency band ratio of millimeter-wave radar and terahertz radar is adjusted according to vehicle speed.

10. The method according to claim 8, characterized in that, The initiation of the multimodal cleaning process includes: Activate high-pressure jet cleaning to utilize collected rainwater for cleaning; Simultaneously activate the thermoelectric cooling element to prevent fogging; Simultaneously activate the ultrasonic vibration device to remove ice.

Images