An adaptive brightness adjustment method for LED car lights based on artificial intelligence

By installing multiple sensors on the vehicle and using CNN and RNN models for deep learning analysis, real-time monitoring and adjustment of the brightness of the headlights, the shortcomings of the existing intelligent headlight adjustment system in comprehensive consideration of multiple environmental factors and abnormal handling are solved, and the problem of the existing intelligent headlight adjustment system is achieved. Efficient and intelligent headlight adjustment improves driving safety and driving experience.

CN118829030BActive Publication Date: 2024-12-03ZHUHAI CITY ZHENGYUAN OPTOELECTRONICS TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411303596.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-19
Publication Date
2024-12-03
Estimated Expiration
2044-09-19

AI Technical Summary

Technical Problem

The existing intelligent car light adjustment system has shortcomings in the comprehensive consideration of various environmental factors and handling abnormal situations, resulting in poor accuracy and timeliness of car light adjustment, affecting driving safety.

Method used

Adaptive adjustment method of LED headlight brightness based on artificial intelligence, data is collected and deep learning analysis is carried out through multiple sensors installed on the vehicle, combining convolutional neural network (CNN) and recurrent neural network (RNN) models for real-time monitoring. and adjust the brightness of the headlights, and have abnormal detection and adaptive learning functions.

Benefits of technology

It realizes efficient and intelligent lighting brightness adjustment in complex driving environments, improves driving safety and driving experience, ensures that the brightness of the lights always meets the best lighting effect, and reduces the need for manual intervention.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118829030B_ABST
    Figure CN118829030B_ABST
Patent Text Reader

Abstract

The present invention discloses an adaptive adjustment method for LED headlight brightness based on artificial intelligence, which relates to the technical field of intelligent adjustment of LED headlight brightness, and comprises the following steps: Step 1: collecting data of the vehicle's surrounding environment and driving status through sensors installed on the vehicle, and performing normalization processing to convert the data into standardized data; Step 2: using a CNN model to analyze the processed standardized data, generate analysis results, and judge the characteristics of the current driving environment. The adaptive adjustment method for LED headlight brightness based on artificial intelligence can detect abnormal conditions in time and issue warnings, further improving the reliability and safety of the system. This dual analysis mechanism ensures the accuracy of headlight brightness adjustment and the stability of system operation. When an abnormality is detected, the system will immediately issue a warning signal to remind the driver or trigger an automatic correction mechanism to ensure that the system is always in the best operating state.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent adjustment of LED vehicle light brightness, and in particular to an adaptive adjustment method of LED vehicle light brightness based on artificial intelligence. Background Art

[0002] With the rapid development of the automobile industry and intelligent technology, the degree of intelligence of vehicle lighting systems is getting higher and higher. Traditional vehicle lighting systems are usually manually controlled by the driver to switch on and off and adjust the brightness of the lights. This method not only increases the burden on the driver, but may also cause insufficient or excessive lighting due to improper human operation, thereby affecting driving safety. For example, when driving at night or entering a tunnel, if the brightness of the lights cannot be adjusted in time, the driver's field of vision may be affected, increasing the risk of accidents. In recent years, with the development of artificial intelligence technology, especially the advancement of deep learning and sensor technology, intelligent headlight adjustment systems based on environmental perception have gradually become a hot spot for research and development. Through various sensors installed on the vehicle, such as light intensity sensors, vehicle speed sensors, and angle sensors, the vehicle can perceive changes in the surrounding environment in real time. These sensor data can be analyzed through deep learning algorithms to achieve adaptive adjustment of the brightness of the lights. For example, when the vehicle enters a tunnel, the system can automatically increase the brightness of the lights, and in the daytime with sufficient light, the system will automatically reduce the brightness of the lights to save energy.

[0003] However, some intelligent headlight adjustment systems on the market still have many shortcomings. For example, some systems can only perform simple brightness adjustment based on the data of a single sensor, and cannot comprehensively consider the influence of multiple environmental factors. The accuracy and timeliness of the adjustment are poor. In addition, some systems lack effective response mechanisms under abnormal conditions, such as sensor failure and sudden environmental changes, which may lead to errors in headlight adjustment and affect driving safety. Therefore, there is an urgent need for an artificial intelligence-based headlight brightness adaptive adjustment method. By introducing multiple sensors and combining deep learning algorithms, the environmental data can be analyzed and processed in real time to automatically adjust the headlight brightness. At the same time, the system should have anomaly detection and adaptive learning functions, and can realize efficient and intelligent headlight brightness adjustment in various complex driving environments to ensure driving safety and driving comfort. Summary of the invention

[0004] To solve the above technical problems, the present invention is implemented through the following technical solutions: an artificial intelligence-based LED headlight brightness adaptive adjustment method, comprising the following steps: step one: collecting data on the vehicle's surrounding environment and driving status through sensors installed on the vehicle, and performing normalization processing to convert the data into standardized data; step two: using the CNN model to analyze the processed standardized data, generate analysis results, judge the characteristics of the current driving environment, and output corresponding headlight brightness adjustment signals; step three: using the RNN model to monitor historical operating data, detect abnormal conditions of the headlight system and issue a warning signal; step four: according to the analysis results of the CNN model, through the brightness adjustment calculation module, using the weight coefficient of the standardized data, dynamically calculate and adjust the brightness of the headlights to adapt to different driving environments; step five: using the RNN model to monitor the operating data of the headlight brightness adjustment system in real time, issue a warning signal when an abnormality is found, and record the abnormality for further analysis.

[0005] Preferably, it includes a vehicle light brightness adjustment system, which includes a sensor data acquisition module, a brightness adjustment calculation module, a deep learning analysis module, a remote update module and a security management module;

[0006] Among them, the sensor data acquisition module: collects the vehicle's surrounding environment data as raw data through sensors installed on the vehicle, and transmits it to the headlight brightness adjustment system; the sensor data acquisition module includes a light intensity sensor, an angle sensor, a yaw angular velocity sensor, a vehicle speed sensor, an acceleration sensor and a headlight angle sensor.

[0007] The light intensity sensor is used to collect light intensity values ​​so as to adjust the brightness of the headlights according to the brightness of the surrounding environment to ensure the best lighting effect; the angle sensor is used to measure the angle value of the car so that the headlights can adjust the lighting direction according to the turning angle to ensure the integrity and accuracy of the road lighting when turning; the yaw rate sensor is used to measure the yaw rate value of the car to help provide better lighting when the vehicle turns and avoid lighting blind spots caused by high-speed turns or sharp turns; the speed sensor is used to measure the current speed value of the car so that the headlights can adjust the lighting mode according to the speed change: the vehicle control system dynamically adjusts the lighting mode of the headlights according to the vehicle speed, such as providing a longer lighting range when driving at high speeds, and reducing the lighting intensity to reduce energy consumption when driving at low speeds in the city; the acceleration sensor is used to obtain the acceleration value of the vehicle to adjust the illumination angle to adapt to different terrains: by analyzing the acceleration and deceleration data of the vehicle, the illumination angle of the headlights is adjusted to adapt to different terrain changes and ensure lighting stability; the headlight angle sensor is used to obtain the brightness value of the headlights to ensure that the lighting angle of the headlights can accurately match the steering action of the vehicle, optimize the road lighting effect, and achieve more precise lighting control;

[0008] That is to say, the raw data includes: light intensity value, turning angle value, yaw angular velocity value, vehicle speed value, acceleration value and headlight brightness value.

[0009] Through the data collection of these sensors, the vehicle control system can fully perceive the vehicle's current driving status and changes in the surrounding environment, and transmit these raw data to the subsequent deep learning analysis module and brightness adjustment calculation module, laying the foundation for realizing intelligent headlight brightness adjustment.

[0010] Preferably, the LED headlight brightness adaptive adjustment method further includes a data analysis step, which is performed by a deep learning analysis module. The deep learning analysis module uses a convolutional neural network (CNN) and a recurrent neural network (RNN) in a deep learning algorithm to analyze and monitor the raw data collected by the sensor data acquisition module, wherein the convolutional neural network (CNN) is marked as a CNN model, and the recurrent neural network (RNN) is marked as an RNN model:

[0011] Data preprocessing: First, the collected raw data is preprocessed by normalization to convert the raw data into standardized data suitable for processing by the deep learning analysis module.

[0012] Among them, normalization is used to preprocess the original data. First, the original data is aggregated into: ;

[0013] Where I represents the light intensity value, Indicates the angle value, represents the yaw rate value, V represents the vehicle speed value, a represents the acceleration value, L represents the vehicle light brightness value, and D represents the original data set. The original data value includes: I, , ,V,a,L;

[0014] The normalization processing steps include: normalizing the light intensity value, the turning angle value, the yaw angular velocity value, the vehicle speed value, the acceleration value and the headlight brightness value, and the formula is: ;

[0015] in, Represents the original data value; and Respectively represent the minimum and maximum values ​​in the original data set; f norm It represents the standardized data after the original data value is normalized, that is, the impact corresponding to the original data value; the normalized original data ensures that data of different dimensions are processed within the same range.

[0016] Data analysis: The preprocessed standardized data is input into the pre-trained CNN model. The CNN model learns from the labeled data to understand the relationship between the driving environment and the headlight brightness adjustment, which is used to accurately judge the characteristics of the current driving environment and output the corresponding headlight brightness adjustment signal.

[0017] At the same time, the RNN model monitors the historical operation data of the headlight brightness adjustment system to process time series data, promptly detect abnormal conditions in the operation of the headlight brightness adjustment system, such as abnormal brightness adjustment, sensor failure, etc., and issue early warning signals.

[0018] Preferably, the headlight brightness adjustment is implemented by a brightness adjustment calculation module, which uses the following formula to calculate the final brightness, and the final brightness is recorded as L a :

[0019] Car light brightness adjustment formula: ;

[0020] Among them, L base is the base brightness value; They are the weight coefficients of light intensity, turning angle, yaw rate, vehicle speed, acceleration and headlight brightness respectively; Indicates the influence of light intensity value; Indicates the influence of the angle value; Indicates the influence of the yaw rate value; Indicates the influence of vehicle speed value; Indicates the influence of acceleration value, Indicates the influence of the brightness value of the car light.

[0021] The brightness L calculated by the above formula a , can dynamically adapt to the needs of different driving environments, provide optimal lighting effects in complex environments such as night, tunnels, and rainy days, and reduce brightness to save energy under good lighting conditions.

[0022] Preferably, the vehicle light brightness adjustment system further includes an adaptive learning module, which further optimizes the accuracy and adaptability of the vehicle light brightness adjustment through the following steps:

[0023] Adaptive adjustment parameter learning: The headlight brightness adjustment system dynamically adjusts the weight coefficient of the headlight brightness adjustment based on the real-time collected raw data, historical headlight brightness adjustment records and driving environment feedback .

[0024] The adaptive learning module is used to perform regression analysis on the headlight brightness adjustment effect under different driving environments, extract important parameters in the original data that affect the headlight brightness adjustment, and update the weight coefficients of important parameters in real time.

[0025] Closed-loop feedback mechanism: After the headlight brightness is adjusted, the headlight brightness adjustment system obtains real-time feedback data through the sensor data acquisition module.

[0026] The adaptive learning module adjusts important parameters based on feedback data to ensure that the headlight brightness adjustment always meets the requirements of the best lighting effect; the closed-loop feedback mechanism is used to continuously optimize the performance of the headlight brightness adjustment system during the adjustment process to cope with dynamically changing driving environments and vehicle conditions.

[0027] Weight coefficient update: Through continuous learning and optimization, the headlight brightness adjustment system automatically updates the weight coefficient in the headlight brightness adjustment formula.

[0028] The adaptive learning module calculates the difference between the current weight coefficient and the feedback data, and uses the optimization algorithm to gradually adjust each weight coefficient to ensure that the influence of each parameter on the brightness adjustment of the headlight is in the optimal state;

[0029] Long-term optimization and personalized adjustment: The adaptive learning module not only optimizes the single driving environment, but also combines the long-term operation data of the vehicle to provide personalized lighting adjustment solutions for different drivers;

[0030] For example, by analyzing the driver's habits and common routes, the headlight brightness adjustment system can pre-adjust the weight coefficient to better meet specific driving needs and improve the driver's comfort and safety;

[0031] Among them, the adaptive learning module has an anomaly detection and warning unit, which monitors and detects anomalies in the historical operating data of the headlight brightness adjustment system through the RNN model.

[0032] Among them, the composition of the adaptive learning module is as follows:

[0033] Data collection and preprocessing: continuously collect raw data and system adjustment history records, and perform normalization to ensure data quality;

[0034] Regression analysis: Perform regression analysis on the brightness adjustment effect under different driving environments, generate analysis results, identify important parameters affecting brightness adjustment, and update the weight coefficients of important parameters based on feedback;

[0035] Weight update mechanism: Dynamically adjust the weight coefficient in the adjustment formula according to the regression analysis results , ensuring that brightness adjustment always meets the best lighting effect in various environments;

[0036] Long-term learning and optimization: Combined with the vehicle's long-term operating data, personalized adjustment plans are provided for different drivers. By analyzing the driver's habits and frequently used routes, the weight coefficients are adjusted in advance to improve personalization and comfort.

[0037] Preferably, the abnormality detection and early warning unit monitors and detects abnormalities on the historical operation data of the vehicle light brightness adjustment system through an RNN model, and the process includes the following steps:

[0038] Data collection and preprocessing: Collect historical operating data of the headlight brightness adjustment system, including data from various sensors at different time points: such as light intensity value, turning angle value, yaw angular velocity value, vehicle speed value, acceleration value and corresponding headlight brightness value, and normalize these historical operating data so that their value range is between 0 and 1, so that the RNN model can be used to effectively process these data and obtain a historical data sequence.

[0039] Construction of the RNN model: Determine the structure of the input layer, hidden layer and output layer of the RNN model. The input layer of the RNN model receives the normalized historical data sequence, the hidden layer extracts and memorizes the features of the historical data sequence through recurrent neurons, and the output layer generates a warning signal for detecting abnormalities in the headlight brightness adjustment system.

[0040] Anomaly detection and warning: The error between the actual headlight brightness value and the brightness value predicted by the RNN model is calculated using the following formula: ;

[0041] in, is the normalized predicted brightness value predicted by the RNN model, is the actual normalized headlight brightness value, and the detection threshold is set to T;

[0042]

[0043] When the error >T, there is an abnormality and a warning signal S is generated;

[0044] When S=1, the driver or the headlight brightness adjustment system is notified through indicator lights, sounds or communications.

[0045] Preferably, the LED headlight brightness adaptive adjustment method further includes a matrix beam control module and a pedestrian detection system; the pedestrian detection system cooperates with the matrix beam control module, and when a pedestrian is detected, the pedestrian detection system generates a corresponding pedestrian detection result, and the headlight brightness adjustment system illuminates the pedestrian and reminds the driver and the pedestrian to improve safety;

[0046] How to illuminate pedestrians: The headlight brightness adjustment system is used in conjunction with the matrix beam control module, which controls the switching of LED monomers to accurately control the lighting area, such as shielding the beam of oncoming vehicles when meeting to avoid glare:

[0047] Beam adjustment: When vehicles meet, the headlight brightness adjustment system will block the beam of the oncoming vehicle to avoid dazzling other vehicles;

[0048] The matrix beam control module is used to dynamically adjust the lighting area based on the vehicle's surrounding environment and the input of the pedestrian detection system, ensuring that the beam is focused on the road area that needs to be illuminated and improving safety.

[0049] Combined with real-time monitoring of the RNN model: When controlling the light beam, the headlight brightness adjustment system is used to timely detect the effect of the light beam adjustment and adjust the important parameters of the light beam control according to the feedback data to ensure the lighting effect while saving energy.

[0050] Preferably, the headlight brightness adjustment method of the headlight brightness adjustment system further includes a dynamic light adjustment mechanism to improve nighttime driving safety:

[0051] Dynamic turn signal adjustment: When the vehicle is turning, the headlight brightness adjustment system adjusts the direction and brightness of the headlights through the input of the steering angle sensor and the yaw angular velocity sensor. The dynamic turn signal adjusts the flashing frequency and brightness to imitate the effect of flowing lights and improve the recognition of night driving.

[0052] Puddle lamp adjustment: When the car door is open at night, the headlight brightness adjustment system automatically illuminates the road surface around the car through puddle lamp adjustment.

[0053] Preferably, the adaptive learning module adaptively adjusts the detection threshold and the sensitivity of early warning signal generation by learning historical operation data while detecting anomalies:

[0054] Adaptive adjustment: The RNN model not only analyzes the current data, but also combines the characteristics of historical operating data to dynamically adjust the detection threshold and the triggering conditions of the warning signal; by comparing the changing trends of current data and historical operating data, the headlight brightness adjustment system is used to identify potential anomalies and adjust the sensitivity of the headlight brightness adjustment system in a timely manner to avoid false alarms or missed alarms.

[0055] Learning feedback mechanism: After the headlight brightness adjustment system detects an anomaly and issues a warning signal, it will save all the data of this process in the database for subsequent analysis and RNN model optimization. During this period, the adaptive learning module regularly uses the feedback data to retrain the RNN model to further improve the accuracy of anomaly detection and the stability of the headlight brightness adjustment system.

[0056] Through this mechanism, the headlight brightness adjustment system can continuously improve its intelligence level in a constantly changing environment, ensuring that the best decisions can always be made in various complex driving scenarios.

[0057] Preferably, the headlight brightness adjustment system continuously optimizes the CNN model and the RNN model through the remote update module and the safety management module to ensure the adaptability and reliability of the headlight brightness adjustment system in various driving environments:

[0058] Data upload and analysis: The vehicle uploads the original operating data, headlight brightness adjustment results, and anomaly detection information to the cloud server via the wireless network. The cloud server processes the data using big data analysis and reinforcement learning algorithms, and generates new training data sets to discover the deficiencies of the headlight brightness adjustment system in actual applications and generate new training data sets.

[0059] Model optimization and remote upgrade: The cloud server optimizes the CNN model and RNN model based on the latest training data set, including adjusting the parameters of the two models, improving the network structure, and introducing new feature extraction methods to improve the model's adaptability and prediction accuracy;

[0060] Safety verification and continuous feedback: After the CNN model and RNN model were upgraded, the headlight brightness adjustment system was re-verified for comprehensive safety to ensure the stability and effectiveness of the new CNN model and RNN model under various driving conditions;

[0061] During the verification process, if an abnormality is found, the headlight brightness adjustment system will automatically roll back to the previous version of the model and record relevant information for further analysis and improvement. If no abnormality is found, the headlight brightness adjustment system will keep the current version of the model and record relevant information for storage as historical operation data.

[0062] During the operation of the headlight brightness adjustment system, the vehicle continuously collects historical operation data and regularly feeds it back to the cloud server to form a closed loop of continuous improvement, ensuring that the model is continuously optimized over time to adapt to changing environmental needs;

[0063] Real-time exception handling and correction: During real-time monitoring, when the RNN model detects an abnormality, the headlight brightness adjustment system not only sends a warning signal, but also activates an automatic correction mechanism. The correction mechanism takes appropriate countermeasures based on the type and severity of the abnormality, such as adjusting the headlight brightness adjustment strategy, restarting the sensor, or switching to safety mode to ensure the stable operation of the headlight brightness adjustment system.

[0064] Security and privacy protection: During data transmission and remote upgrades, the headlight brightness adjustment system uses advanced encryption technology and access control measures to ensure data security and user privacy protection. All data processing and model updates are performed in a secure computing environment to avoid any unauthorized access and data leakage.

[0065] The present invention provides an artificial intelligence-based LED headlight brightness adaptive adjustment method, which has the following beneficial effects:

[0066] 1. The AI-based LED headlight brightness adaptive adjustment method, through the above-mentioned multiple sensors installed on the vehicle, the system can collect relevant data about the vehicle's surrounding environment in real time. These data provide comprehensive environmental perception information for subsequent brightness adjustment, ensuring that the headlight brightness can be adaptively adjusted according to the current driving environment, improving driving safety and driving experience. The application of this technical solution reduces the need for manual intervention and realizes intelligent and automated headlight control.

[0067] 2. This AI-based LED headlight brightness adaptive adjustment method, by introducing convolutional neural networks and recursive neural networks, the system can conduct in-depth analysis of the collected data and accurately judge the characteristics of the current driving environment. The CNN model has mastered the relationship between different driving environments and headlight brightness adjustment by learning a large amount of labeled data, ensuring that the headlight brightness can be automatically adjusted in various environments. The RNN model can detect abnormal situations in time and issue early warnings by monitoring the system's historical operation data, further improving the reliability and safety of the system. This dual analysis mechanism ensures the accuracy of headlight brightness adjustment and the stability of system operation.

[0068] 3. This AI-based LED headlight brightness adaptive adjustment method uses the RNN model to monitor the system operation data and detect anomalies. The system can promptly identify abnormal conditions in brightness adjustment, such as brightness abnormalities, sensor failures, etc. The RNN model can capture minor anomalies in system operation by processing time series data, and quantify the anomalies through error calculation formulas. When an anomaly is detected, the system will immediately issue a warning signal to remind the driver or trigger an automatic correction mechanism to ensure that the system is always in the best operating state and avoid safety hazards caused by brightness adjustment errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flow chart of a method for adaptively adjusting brightness of LED vehicle lights based on artificial intelligence according to the present invention;

[0070] Figure 2 This is a flow chart of the vehicle light brightness adjustment system. DETAILED DESCRIPTION

[0071] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

[0072] like Figure 1 and Figure 2 As shown, the present invention provides a technical solution: an artificial intelligence-based LED headlight brightness adaptive adjustment method, comprising the following steps: step one: collecting data on the vehicle's surrounding environment and driving status through sensors installed on the vehicle, and performing normalization processing to convert the data into standardized data; step two: using the CNN model to analyze the processed standardized data, generate analysis results, judge the characteristics of the current driving environment, and output a corresponding headlight brightness adjustment signal; step three: using the RNN model to monitor historical operating data, detect abnormal conditions of the headlight system and issue a warning signal; step four: according to the analysis results of the CNN model, through the brightness adjustment calculation module, using the weight coefficient of the standardized data, dynamically calculate and adjust the brightness of the headlight to adapt to different driving environments; step five: using the RNN model to monitor the operating data of the headlight brightness adjustment system in real time, issue a warning signal when an abnormality is found, and record the abnormality for further analysis.

[0073] The LED headlight brightness adaptive adjustment method based on artificial intelligence includes a headlight brightness adjustment system, which includes a sensor data acquisition module, a brightness adjustment calculation module, a deep learning analysis module, a remote update module and a safety management module;

[0074] Among them, the sensor data acquisition module: collects the vehicle's surrounding environment data as raw data through sensors installed on the vehicle, and transmits it to the headlight brightness adjustment system; the sensor data acquisition module includes a light intensity sensor, an angle sensor, a yaw angular velocity sensor, a vehicle speed sensor, an acceleration sensor and a headlight angle sensor.

[0075] The light intensity sensor is used to collect light intensity values ​​so as to adjust the brightness of the headlights according to the brightness of the surrounding environment to ensure the best lighting effect;

[0076] The turning angle sensor is used to measure the turning angle of the car so that the headlights can adjust the lighting direction according to the turning angle to ensure the integrity and accuracy of the road lighting when turning;

[0077] The yaw rate sensor is used to measure the yaw rate value of the vehicle, helping to provide better lighting when the vehicle is turning and avoiding lighting blind spots caused by high-speed turns or sharp turns;

[0078] The vehicle speed sensor is used to measure the current vehicle speed so that the headlights can adjust the lighting mode according to the vehicle speed: the vehicle control system dynamically adjusts the lighting mode of the headlights according to the vehicle speed, such as providing a longer lighting range when driving at high speeds, and reducing the lighting intensity to reduce energy consumption when driving at low speeds in the city;

[0079] The acceleration sensor is used to obtain the acceleration value of the vehicle to adjust the illumination angle to adapt to different terrains: by analyzing the acceleration and deceleration data of the vehicle, the illumination angle of the headlights is adjusted to adapt to different terrain changes and ensure lighting stability;

[0080] The headlight angle sensor is used to obtain the headlight brightness value, ensure that the headlight lighting angle can accurately match the vehicle's steering action, optimize the road lighting effect, and achieve more precise lighting control.

[0081] Through the above-mentioned multiple sensors installed on the vehicle, the system can collect relevant data about the vehicle's surrounding environment in real time. These data provide comprehensive environmental perception information for subsequent brightness adjustment, ensuring that the brightness of the headlights can be adaptively adjusted according to the current driving environment, thereby improving driving safety and driving experience. The application of this technical solution reduces the need for manual intervention and realizes intelligent and automated headlight control.

[0082] The LED headlight brightness adaptive adjustment method also includes a data analysis step, which is completed by a deep learning analysis module. The deep learning analysis module uses a convolutional neural network (CNN) and a recurrent neural network (RNN) in a deep learning algorithm to analyze and monitor the raw data collected by the sensor data acquisition module, wherein the convolutional neural network (CNN) is marked as a CNN model, and the recurrent neural network (RNN) is marked as an RNN model:

[0083] Data preprocessing: First, the collected raw data is preprocessed by normalization to convert the raw data into standardized data suitable for processing by the deep learning analysis module.

[0084] Among them, normalization is used to preprocess the original data. First, the original data is aggregated into:

[0085] ;

[0086] Where I represents the light intensity value, Indicates the angle value, represents the yaw rate value, V represents the vehicle speed value, a represents the acceleration value, L represents the vehicle light brightness value, and D represents the original data set. The original data value includes: I, , ,V,a,L;

[0087] The normalization processing steps include: normalizing the light intensity value, the turning angle value, the yaw angular velocity value, the vehicle speed value, the acceleration value and the headlight brightness value, and the formula is:

[0088] ;

[0089] in, Represents the original data value; and Respectively represent the minimum and maximum values ​​in the original data set; f norm It represents the standardized data after the original data value is normalized, that is, the impact corresponding to the original data value.

[0090] Data analysis: The preprocessed standardized data is input into the pre-trained CNN model. The CNN model learns from the labeled data to understand the relationship between the driving environment and the headlight brightness adjustment, which is used to accurately judge the characteristics of the current driving environment and output the corresponding headlight brightness adjustment signal.

[0091] At the same time, the RNN model monitors the historical operation data of the headlight brightness adjustment system to process time series data, promptly detect abnormal conditions in the operation of the headlight brightness adjustment system, such as abnormal brightness adjustment, sensor failure, etc., and issue early warning signals;

[0092] Building a CNN model: determine the input layer, convolution layer, pooling layer, fully connected layer, and output layer;

[0093] Receiving preprocessed raw data through an input layer, wherein the raw data includes data such as light intensity, angle, yaw angular velocity, vehicle speed, acceleration, and headlight angle; and making the dimension of the input layer consistent with the number of features of the input raw data;

[0094] The convolutional layer extracts spatial features from the raw data, and the filter performs convolution operations on the raw data to extract relevant high-level features, such as lighting patterns and vehicle dynamic characteristics in different driving environments.

[0095] Then, downsampling is completed through the pooling layer through maximum pooling and average pooling to reduce the data dimension and retain important features to avoid model overfitting;

[0096] Then, the spatial features, high-level features, and important features extracted by the convolution layer and the pooling layer are connected to the output layer through the fully connected layer for further processing and classification. The output of the fully connected layer represents the classification results of different driving environments, where the classification results include: spatial features, high-level features, and important features.

[0097] Finally, the output layer processes the classification results through the activation function Softmax to obtain the output results, and converts the output results into specific brightness adjustment signals. The brightness adjustment signals are used to indicate the brightness that the LED lights should be adjusted in the current environment.

[0098] During the application process, assuming that a vehicle enters a tunnel from an environment with sufficient daylight, the sensor detects that the light intensity is significantly reduced. The preprocessed data is input into the CNN model. The CNN model extracts the current environmental features through the convolution layer and classifies it as a tunnel environment. The output layer generates an exclusive brightness adjustment signal with a tunnel environment logo, indicating that the brightness of the headlights should be significantly increased. The system responds immediately and increases the brightness of the LED headlights to cope with the dark environment of the tunnel, ensuring the driver's clear vision.

[0099] After being trained with a large amount of labeled data, the RNN model can learn the relationship between different driving environments and the adjustment of headlight brightness, so as to accurately judge the characteristics of the current driving environment, such as daytime, nighttime, tunnels, rainy days, etc.; it can automatically adjust the brightness of LED headlights according to the analysis results of the CNN model to adapt to different driving environments. For example, at night or in tunnels, the headlight brightness is increased to enhance the lighting effect; during the day or in good lighting conditions, the headlight brightness is reduced to save energy.

[0100] By introducing convolutional neural networks and recursive neural networks, the system can conduct in-depth analysis of the collected data and accurately judge the characteristics of the current driving environment. The CNN model has mastered the relationship between different driving environments and the adjustment of headlight brightness by learning a large amount of labeled data, ensuring that the headlight brightness can be automatically adjusted in various environments. The RNN model can detect abnormal situations in a timely manner and issue early warnings by monitoring the system's historical operation data, further improving the reliability and safety of the system. This dual analysis mechanism ensures the accuracy of headlight brightness adjustment and the stability of system operation.

[0101] The headlight brightness is adjusted by the brightness adjustment calculation module, which uses the following formula to calculate the final brightness L:

[0102] Car light brightness adjustment formula:

[0103] ;

[0104] Among them, L baseis the base brightness value; They are the weight coefficients of light intensity, turning angle, yaw rate, vehicle speed, acceleration and headlight brightness respectively; Indicates the influence of light intensity value; Indicates the influence of the angle value; Indicates the influence of the yaw rate value; Indicates the influence of vehicle speed value; Indicates the influence of acceleration value, Indicates the influence of the brightness value of the car light.

[0105] By introducing the brightness adjustment formula, the system can dynamically calculate and adjust the brightness of the headlights to adapt to different driving environment requirements. The formula takes multiple influencing factors into consideration and realizes multi-dimensional brightness adjustment. This technical solution ensures the best lighting effect in complex environments such as night, tunnels, and rainy days. At the same time, it reduces the brightness under good lighting conditions to save energy, thereby improving the system's energy efficiency and user experience.

[0106] The headlight brightness adjustment system also includes an adaptive learning module, which further optimizes the accuracy and adaptability of the headlight brightness adjustment by the following steps:

[0107] Adaptive adjustment parameter learning: The headlight brightness adjustment system dynamically adjusts the weight coefficient of the headlight brightness adjustment based on real-time sensor data, historical headlight brightness adjustment records and driving environment feedback .

[0108] The adaptive learning module is used to perform regression analysis on the headlight brightness adjustment effect under different driving conditions, extract important parameters that affect the headlight brightness adjustment, and update the weight coefficients of important parameters in real time.

[0109] Closed-loop feedback mechanism: After the headlight brightness is adjusted, the headlight brightness adjustment system obtains real-time feedback data through the headlight angle sensor, light intensity sensor and other environmental sensors.

[0110] The adaptive learning module adjusts the adjustment parameters according to the feedback data to ensure that the headlight brightness adjustment always meets the requirements of the best lighting effect; the closed-loop feedback mechanism is used for the headlight brightness adjustment system to continuously optimize its own performance during the adjustment process to cope with the dynamically changing driving environment and vehicle status.

[0111] Weight coefficient update: Through continuous learning and optimization, the headlight brightness adjustment system automatically updates the weight coefficient in the headlight brightness adjustment formula.

[0112] The adaptive learning module calculates the difference between the current weight coefficient and the feedback data, and uses the optimization algorithm to gradually adjust each weight coefficient to ensure that the influence of each parameter on the brightness adjustment of the headlight is in the optimal state;

[0113] For example, when light intensity changes significantly The weight of the sensor is automatically increased to respond more accurately to changes in ambient light; at high speeds, The weight of the light will be increased to better meet the lighting needs at high speeds.

[0114] Long-term optimization and personalized adjustment: The adaptive learning module not only optimizes the single driving environment, but also combines the long-term operation data of the vehicle to provide personalized lighting adjustment solutions for different drivers;

[0115] Based on the above technical solution, for a vehicle that frequently switches between urban and rural roads, the headlight brightness adjustment system discovers through an adaptive learning module that the driver prefers low-brightness mode in urban environments and high-brightness mode on rural roads; after regression analysis, the headlight brightness adjustment system adaptively adjusts the brightness adjustment parameters to ensure energy saving on urban roads and provide stronger lighting effects on rural roads, thereby improving the driver's comfort and safety.

[0116] Among them, the adaptive learning module has an anomaly detection and warning unit, which monitors and detects anomalies in the historical operating data of the headlight brightness adjustment system through the RNN model.

[0117] The anomaly detection and early warning unit monitors and detects anomalies in the historical operating data of the headlight brightness adjustment system through the RNN model. The process includes the following steps:

[0118] Data collection and preprocessing: Collect historical operating data of the headlight brightness adjustment system, including data from various sensors at different time points: such as light intensity value, turning angle value, yaw angular velocity value, vehicle speed value, acceleration value and corresponding headlight brightness value, and normalize these historical operating data so that their value range is between 0 and 1, so that the RNN model can be used to effectively process these data and obtain a historical data sequence.

[0119] Construction of RNN model: Determine the structure of the input layer, hidden layer and output layer of the RNN model;

[0120] The input layer receives historical data in the form of time series, which includes light intensity values, turning angle values, yaw angular velocity values, vehicle speed values, acceleration values, headlight brightness values, and system operating status, etc. These historical data are then input into the input layer of the RNN model in chronological order; that is, the input layer of the RNN model receives a normalized historical data sequence in order to capture the temporal dependency in the data.

[0121] The hidden layer processes the input time series data through recurrent neurons, and generates the current state of the hidden layer by combining the previous hidden state of the hidden layer, which is recorded as the current state, and transmits the current state to the output layer;

[0122] The output layer generates a warning signal for abnormal status of the headlight brightness adjustment system according to the current state of the hidden layer. That is to say, when abnormal headlight brightness adjustment or sensor failure is detected, the RNN model will generate a warning signal.

[0123] It also includes error calculation, that is, anomaly detection and early warning. In other words, the RNN model determines the normal operation status of the system by calculating the error between the actual headlight brightness value and the model-predicted brightness value:

[0124] Anomaly detection and warning: The error between the actual headlight brightness value and the brightness value predicted by the RNN model is calculated using the following formula:

[0125] ;

[0126] in, is the normalized predicted brightness value predicted by the RNN model, is the actual normalized headlight brightness value, and the detection threshold is set to T;

[0127] When the error >T, there is an abnormality and a warning signal S is generated;

[0128] .

[0129] When S=1, the driver or the headlight brightness adjustment system is notified through indicator lights, sounds or communications;

[0130] When S=0, there is no abnormality, no warning signal will be generated, and the headlight brightness adjustment system is in normal operation.

[0131] That is to say, according to the above technical solution: when the vehicle is driving at night for a long time, the system continuously collects and records the data of the headlight brightness and related sensors. The RNN model analyzes these time series data and finds that the headlight brightness in a certain period of time does not match the ambient light intensity. When the error exceeds the set threshold, the RNN immediately sends a warning signal to notify the driver or automatically adjust the headlight brightness to prevent driving safety problems caused by insufficient or excessive brightness.

[0132] The LED headlight brightness adaptive adjustment method also includes a matrix beam control module and a pedestrian detection system;

[0133] The pedestrian detection system works with the matrix beam control module. When a pedestrian is detected, the headlight brightness adjustment system will illuminate the pedestrian and alert the driver and pedestrian to improve safety.

[0134] The headlight brightness adjustment system is used in conjunction with the matrix beam control module to control the switch of the LED monomers for precise control of the lighting area, such as shielding the beam of the oncoming vehicle when meeting the vehicle to avoid glare:

[0135] Beam adjustment: When vehicles meet, the headlight brightness adjustment system will block the beam of the oncoming vehicle to avoid dazzling other vehicles;

[0136] The matrix beam control module is used to dynamically adjust the lighting area based on the vehicle's surrounding environment and the input of the pedestrian detection system, ensuring that the beam is focused on the road area that needs to be illuminated and improving safety.

[0137] Combined with real-time monitoring of the RNN model: When the headlight brightness adjustment system controls the light beam, it is used to timely detect the effect of the light beam adjustment, and adjust the adjustment parameters of the light beam control according to the feedback data to ensure the lighting effect while saving energy.

[0138] The headlight brightness adjustment method of the headlight brightness adjustment system further includes a dynamic light adjustment mechanism to improve nighttime driving safety:

[0139] Dynamic turn signal adjustment: When the vehicle is turning, the headlight brightness adjustment system adjusts the direction and brightness of the headlights through the input of the steering angle sensor and the yaw angular velocity sensor. The dynamic turn signal imitates the effect of flowing lights by adjusting the flashing frequency and brightness, thereby improving the recognition of night driving.

[0140] Puddle light adjustment: When the car door is open at night, the headlight brightness adjustment system automatically illuminates the road around the car through puddle light adjustment to prevent the driver and passengers from encountering puddles or other obstacles in the dark; the brightness of the puddle light will also be automatically adjusted according to the surrounding light intensity to ensure safety while reducing energy consumption.

[0141] Through the combination of dynamic turn signals and puddle light technology, the system further improves the safety of night driving. The dynamic turn signal technology adjusts the direction and brightness of the lights according to the vehicle's steering information, improving the recognition of night driving and ensuring complete lighting coverage when turning. The puddle light technology automatically illuminates the road surface around the vehicle when the door is opened at night, preventing the driver and passengers from encountering puddles or other obstacles in the dark. The integration of these technologies effectively improves the safety of night driving and avoids potential dangerous situations.

[0142] While detecting anomalies, the adaptive learning module adaptively adjusts the detection threshold and the sensitivity of warning signal generation by learning from historical operation data:

[0143] Adaptive adjustment: The RNN model not only analyzes the current data, but also combines the characteristics of historical operating data to dynamically adjust the detection threshold and the triggering conditions of the warning signal; by comparing the changing trends of current data and historical operating data, the headlight brightness adjustment system is used to identify potential anomalies and adjust the sensitivity of the headlight brightness adjustment system in a timely manner to avoid false alarms or missed alarms.

[0144] Learning feedback mechanism: After the headlight brightness adjustment system detects an anomaly and issues a warning signal, it will save all the data of this process in the database for subsequent analysis and RNN model optimization. During this period, the adaptive learning module regularly uses the feedback data to retrain the RNN model to further improve the accuracy of anomaly detection and the stability of the headlight brightness adjustment system.

[0145] Through this mechanism, the headlight brightness adjustment system can continuously improve its intelligence level in a constantly changing environment, ensuring that the best decisions can always be made in various complex driving scenarios.

[0146] The RNN model is used to monitor the system operation data and detect anomalies. The system can promptly identify abnormal situations in brightness adjustment, such as brightness anomalies, sensor failures, etc. The RNN model can capture minor anomalies in system operation by processing time series data, and quantify the anomalies through error calculation formulas. When an anomaly is detected, the system will immediately issue a warning signal to remind the driver or trigger an automatic correction mechanism to ensure that the system is always in the best operating state and avoid safety hazards caused by brightness adjustment errors.

[0147] The headlight brightness adjustment system continuously optimizes the CNN model and RNN model through the remote update module and the safety management module to ensure the adaptability and reliability of the headlight brightness adjustment system in various driving environments:

[0148] Data upload and analysis: The vehicle uploads the operating sensor data, headlight brightness adjustment results and anomaly detection information to the cloud server via the wireless network. The cloud server processes the data using big data analysis and reinforcement learning algorithms to identify deficiencies in the actual application of the headlight brightness adjustment system and generate new training data sets.

[0149] Model optimization and remote upgrade: The cloud server optimizes the CNN model and RNN model based on the latest training data set, including adjusting the parameters of the two models, improving the network structure, and introducing new feature extraction methods to improve the model's adaptability and prediction accuracy;

[0150] The optimized model is remotely deployed to the headlight brightness adjustment system through a secure communication channel to ensure that the model can respond quickly and adapt to various complex driving environments in actual applications. After receiving the new model, the headlight brightness adjustment system will update the model in a non-operating state and use the latest model for headlight brightness adjustment and anomaly detection at the next startup.

[0151] Safety verification and continuous feedback: After the CNN model and RNN model were upgraded, the headlight brightness adjustment system was re-verified for comprehensive safety to ensure the stability and effectiveness of the new CNN model and RNN model under various driving conditions;

[0152] During the verification process, if an anomaly is found, the headlight brightness adjustment system will automatically roll back to the previous version of the model and record relevant information for further analysis and improvement;

[0153] During the operation of the headlight brightness adjustment system, the vehicle continuously collects historical operating data and regularly feeds it back to the cloud server to form a closed loop of continuous improvement, ensuring that the model is continuously optimized over time and adapts to changing environmental requirements.

[0154] The present invention can not only provide precise headlight brightness adjustment in complex driving environments, but also ensure the intelligence and adaptability of the headlight brightness adjustment system through continuous learning and optimization. Through data feedback and model optimization, the performance and safety standards of the headlight brightness adjustment system are further improved, further improving the user's driving experience and road safety.

[0155] The artificial intelligence-based LED headlight brightness adaptive adjustment method of the present invention can provide accurate and real-time lighting adjustment in various complex driving environments. At the same time, the intelligence level and safety of the headlight brightness adjustment system are improved through continuous learning and remote updating of modules, providing the driver with the best driving experience and safety guarantee.

[0156] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. A method for adaptively adjusting the brightness of LED lights based on artificial intelligence, characterized in that: The steps include: Step 1: Collect data about the vehicle's surrounding environment and driving status through sensors installed on the vehicle, and perform normalization processing to convert the data into standardized data; Step 2: Use the CNN model to analyze the processed standardized data, generate analysis results, determine the characteristics of the current driving environment, and output the corresponding headlight brightness adjustment signal; Step 3: Use the RNN model to monitor historical operation data, detect abnormal conditions of the lighting system and issue warning signals; Step 4: Based on the analysis results of the CNN model, the brightness adjustment calculation module uses the weight coefficient of the standardized data to dynamically calculate and adjust the brightness of the headlights to adapt to different driving environments; Step 5: Use the RNN model to monitor the operating data of the headlight brightness adjustment system in real time, issue a warning signal when an abnormality is found, and record the abnormality for further analysis; It includes a vehicle light brightness adjustment system, which includes a sensor data acquisition module, a brightness adjustment calculation module, a deep learning analysis module, a remote update module and a safety management module; Among them, the sensor data acquisition module: collects vehicle surrounding environment data as raw data through sensors installed on the vehicle, and transmits it to the headlight brightness adjustment system; the sensor data acquisition module includes a light intensity sensor, an angle sensor, a yaw angular velocity sensor, a vehicle speed sensor, an acceleration sensor and a headlight angle sensor; The light intensity sensor is used to collect light intensity values, the angle sensor is used to measure the angle value of the car, the yaw angular velocity sensor is used to measure the yaw angular velocity value of the car, the vehicle speed sensor is used to measure the current vehicle speed value, the acceleration sensor is used to obtain the acceleration value of the vehicle, and the headlight angle sensor is used to obtain the headlight brightness value; The LED headlight brightness adaptive adjustment method further includes a data analysis step, which is completed by a deep learning analysis module. The deep learning analysis module uses a convolutional neural network (CNN) and a recurrent neural network (RNN) in a deep learning algorithm to analyze and monitor the raw data collected by the sensor data acquisition module, wherein the convolutional neural network (CNN) is marked as a CNN model and the recurrent neural network (RNN) is marked as an RNN model. The data analysis steps are as follows: First, the collected raw data is preprocessed and normalized to convert the raw data into standardized data suitable for processing by the deep learning analysis module; Before the normalization process is used to preprocess the original data, the original data is firstly aggregated into: ; Where I represents the light intensity value, Indicates the angle value, represents the yaw rate value, V represents the vehicle speed value, a represents the acceleration value, L represents the vehicle light brightness value, and D represents the original data set. The original data value includes: I, , ,V,a,L; Then the light intensity value, turning angle value, yaw angular velocity value, vehicle speed value, acceleration value and headlight brightness value are normalized, and the formula is: ; in, Represents the original data value; and Respectively represent the minimum and maximum values ​​in the original data set; f norm It represents the standardized data after the original data value is normalized, that is, the impact corresponding to the original data value; The preprocessed standardized data is input into the pre-trained CNN model. The CNN model learns the labeled data to understand the relationship between the driving environment and the headlight brightness adjustment, so as to accurately judge the characteristics of the current driving environment and output the corresponding headlight brightness adjustment signal. At the same time, the RNN model monitors the historical operation data of the headlight brightness adjustment system and is used to process the historical data sequence; The brightness adjustment calculation module is used for adjusting the brightness of the headlights, and is specifically used for calculating the final brightness L of the headlights. a The brightness adjustment calculation module calculates the final brightness through the following formula, and the final brightness is recorded as L a : Car light brightness adjustment formula: ; in, is the base brightness value; They are the weight coefficients of light intensity, turning angle, yaw rate, vehicle speed, acceleration and headlight brightness respectively; Indicates the influence of light intensity value; Indicates the influence of the angle value; Indicates the influence of the yaw rate value; Indicates the influence of vehicle speed value; Indicates the influence of acceleration value, Indicates the influence of the brightness value of the headlight; The vehicle light brightness adjustment system further includes an adaptive learning module: The headlight brightness adjustment system dynamically adjusts the weight coefficient of the headlight brightness adjustment based on the real-time collected raw data, historical headlight brightness adjustment records and driving environment feedback. ; The adaptive learning module is used to perform regression analysis on the headlight brightness adjustment effect under different driving environments, extract important parameters from the original data that affect the headlight brightness adjustment, and update the weight coefficients of important parameters in real time; After the headlight brightness is adjusted, the headlight brightness adjustment system obtains real-time feedback data through the sensor data acquisition module; The adaptive learning module adjusts important parameters according to feedback data. The closed-loop feedback mechanism is used for the headlight brightness adjustment system to continuously optimize its own performance during the adjustment process. Through continuous learning and optimization, the headlight brightness adjustment system automatically updates the weight coefficients in the headlight brightness adjustment formula. The adaptive learning module calculates the difference between the current weight coefficient and the feedback data, and uses the optimization algorithm to gradually adjust each weight coefficient. The adaptive learning module not only optimizes the single driving environment, but also combines the long-term operation data of the vehicle to provide lighting adjustment solutions for different drivers. The adaptive learning module has an abnormality detection and warning unit, which monitors and detects abnormalities in the historical operation data of the headlight brightness adjustment system through the RNN model. The abnormality detection and early warning unit monitors and detects abnormalities on the historical operation data of the headlight brightness adjustment system through the RNN model. The process includes the following steps: Collect historical operation data of the headlight brightness adjustment system, including historical operation data of each sensor at different time points, and perform normalization processing so that its value range is between 0 and 1, and obtain a historical data sequence; Determine the structure of the input layer, hidden layer and output layer of the RNN model. The input layer of the RNN model receives the normalized historical data sequence, the hidden layer extracts and memorizes the features of the historical data sequence through recurrent neurons, and the output layer generates a warning signal for detecting abnormalities in the headlight brightness adjustment system. The error between the actual headlight brightness value and the brightness value predicted by the RNN model is calculated using the following formula: ; in, is the normalized predicted brightness value predicted by the RNN model, is the actual normalized headlight brightness value, and the detection threshold is set to T; When the error >T, there is an abnormality and a warning signal S is generated; .

2. The method for adaptively adjusting the brightness of LED vehicle lights based on artificial intelligence according to claim 1, characterized in that: The LED headlight brightness adaptive adjustment method also includes a matrix beam control module and a pedestrian detection system; The pedestrian detection system works with the matrix beam control module. When a pedestrian is detected, the pedestrian detection system generates a corresponding pedestrian detection result, and the headlight brightness adjustment system illuminates the pedestrian. Ways to illuminate pedestrians: The headlight brightness adjustment system is used in combination with the matrix beam control module, which controls the switching of LED units to control the lighting area: when vehicles meet, the headlight brightness adjustment system will shield the light beam of the oncoming vehicle to avoid dazzling other vehicles; the matrix beam control module is used to dynamically adjust the lighting area based on the vehicle's surrounding environment and the input of the pedestrian detection system; when controlling the light beam, the headlight brightness adjustment system is used to timely detect the effect of the light beam adjustment and adjust the important parameters of the light beam control based on the feedback data.

3. The method for adaptively adjusting the brightness of LED vehicle lights based on artificial intelligence according to claim 2, characterized in that: The vehicle light brightness adjustment method of the vehicle light brightness adjustment system further includes a dynamic light adjustment mechanism: When the vehicle is turning, the headlight brightness adjustment system adjusts the direction and brightness of the headlights through the input of the steering angle sensor and the yaw rate sensor, and the dynamic turn signal adjusts the flashing frequency and brightness; When the car door is opened at night, the headlight brightness adjustment system automatically illuminates the road around the car by adjusting the puddle lights.

4. The method for adaptively adjusting the brightness of LED vehicle lights based on artificial intelligence according to claim 3, characterized in that: The adaptive learning module adaptively adjusts the detection threshold and the sensitivity of early warning signal generation by learning historical operation data while detecting anomalies: The RNN model not only analyzes the current data, but also combines the characteristics of historical operation data to dynamically adjust the detection threshold and the triggering conditions of the warning signal. By comparing the changing trends of the current data with the historical operation data, the headlight brightness adjustment system is used to identify potential anomalies and adjust the sensitivity of the headlight brightness adjustment system in a timely manner. After detecting an anomaly and issuing a warning signal, the headlight brightness adjustment system will save all data of this process in the database for subsequent analysis and RNN model optimization; during this period, the adaptive learning module will regularly use feedback data to retrain the RNN model.

5. The method for adaptively adjusting the brightness of LED vehicle lights based on artificial intelligence according to claim 4, characterized in that: The vehicle light brightness adjustment system continuously optimizes the CNN model and the RNN model through the remote update module and the safety management module: The vehicle uploads the original data, light brightness adjustment results and abnormal detection information during operation to the cloud server through the wireless network. The cloud server processes the data using big data analysis and reinforcement learning algorithms and generates a new training data set. The cloud server optimizes the CNN model and RNN model based on the latest training data set. After the CNN model and RNN model are upgraded, the headlight brightness adjustment system is re-verified for comprehensive safety. During the verification process, if an abnormality is found, the headlight brightness adjustment system will automatically roll back to the previous version of the model and record the relevant information; If no abnormality is found, the headlight brightness adjustment system will maintain the current version of the model and record relevant information; during the operation of the headlight brightness adjustment system, the vehicle continues to collect historical operation data and regularly feeds it back to the cloud server.

Citation Information

Patent Citations

  • LED light source intelligent control system and method based on data analysis

    CN118317485A

  • Evaluation method for vehicle lamp passing light brightness and darkness cut-off line

    CN118624177A