Intelligent vehicle lamp system with adjustable brightness

Abstract

The invention discloses an intelligent vehicle lamp system with adjustable brightness, and particularly relates to the technical field of intelligent vehicle lamps, and the intelligent vehicle lamp system is characterized in that vehicle lamp adjustment response information, driver vehicle lamp adjustment intervention feedback information, lowest recognizable brightness rising information and standby mode switching failure information of an existing intelligent vehicle lamp system are acquired; an intelligent vehicle lamp system performance evaluation model is constructed, a vehicle lamp brightness adjustment hidden danger index is generated, existing operation hidden dangers of the intelligent vehicle lamp system are sensed in time before the intelligent vehicle lamp system has obvious faults, and unstable factors of the system in actual operation are found in time. The stability condition of vehicle lamp brightness adjustment is discriminated and classified during actual operation of an existing intelligent vehicle lamp system, and when a stability reduction signal is generated, the difference value between a vehicle lamp brightness adjustment hidden danger index and a vehicle lamp brightness adjustment hidden danger index threshold value is calculated; and constructing a machine learning vehicle lamp brightness adjustment model by combining the real-time environment brightness data and the driving data of the opposite vehicle, and carrying out automatic adjustment on the vehicle lamp.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicle lamps, and more specifically, to an intelligent vehicle lamp system with adjustable brightness. Background Art

[0002] With the rapid development of automobile technology, adaptive lighting systems have gradually become an important part of modern automobiles. Traditional headlight systems are often unable to effectively cope with the ever-changing ambient light and traffic conditions when driving at night or in bad weather conditions, resulting in increased driving safety risks. Existing smart headlight technology usually relies on light sensors and cameras to detect the light source and environmental conditions ahead. However, due to long-term use or the irregularity of vehicle maintenance cycles, these systems cannot promptly perceive the operating risks before obvious failures occur. As a result, in the face of complex traffic environments, especially when the lights of oncoming vehicles are low or the light intensity is insufficient, misjudgments occur and the brightness of the high beam cannot be accurately adjusted, thus affecting driving safety. Therefore, improving the accuracy of light source detection and the system response speed has become the key to achieving efficient headlight adjustment. Summary of the invention

[0003] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a smart vehicle light system with adjustable brightness to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A brightness-adjustable intelligent vehicle light system, comprising an information acquisition module, a performance evaluation module, a difference identification module, and a vehicle light adjustment optimization module;

[0006] An information acquisition module, used to acquire headlight adjustment response information of an existing intelligent headlight system, driver headlight adjustment intervention feedback information, minimum recognizable brightness increase information, and standby mode switching failure information;

[0007] Performance evaluation module, which builds a performance evaluation model for the intelligent headlight system, generates a risk index for headlight brightness adjustment, and evaluates the stability of the existing intelligent headlight system in adjusting headlight brightness during actual operation;

[0008] A difference identification module compares the risk index of headlight brightness adjustment with a preset risk index threshold of headlight brightness adjustment, and identifies and classifies the stability of headlight brightness adjustment of the existing intelligent headlight system during actual operation;

[0009] The headlight adjustment optimization module builds a machine learning headlight brightness adjustment model based on real-time ambient brightness data and driving data of oncoming vehicles to automatically adjust the headlights when the stability of the existing intelligent headlight system in adjusting the headlight brightness decreases during actual operation.

[0010] In a preferred embodiment, the headlight adjustment response information includes a headlight adjustment response delay coefficient, the driver headlight adjustment intervention feedback information includes a driver headlight adjustment intervention feedback coefficient, the minimum recognizable brightness rise information includes a minimum recognizable brightness rise abnormality coefficient, and the standby mode switching failure information includes a standby mode switching failure ratio.

[0011] In a preferred embodiment, the logic for obtaining the vehicle light adjustment response delay coefficient is as follows:

[0012] During the actual driving process of the vehicle, the time point sj1 of each change event of the ambient light source and the time point sj2 of the start of headlight adjustment when the change event occurs are recorded, and the response time xy from the change of the ambient light source to the start of headlight adjustment is calculated, and the expression is as follows xy=|sj2-sj1|;

[0013] Calculate the average response time xyp, the expression is as follows where xy i represents the response time calculated for the i-th ambient light source change event, i = {1, 2, ..., I}, I is a positive integer;

[0014] Compare the response time calculated for each ambient light source change event with the average response time, mark the response time calculated for the ambient light source change event that is greater than the average response time as the delayed response time xyc, and calculate the headlight adjustment response delay coefficient Xcdx, which is expressed as follows where xyc n Indicates the response time that is marked as delayed response time for the nth time, n={1,2,...,N}, where N is a positive integer.

[0015] In a preferred embodiment, the logic for obtaining the driver's headlight adjustment intervention feedback coefficient is as follows:

[0016] The duration, brightness difference, and angle difference of the driver's active headlight adjustment are obtained during vehicle driving, and the duration, brightness difference, and angle difference of each active headlight adjustment by the driver are integrated into a data set, and the duration, brightness difference, and angle difference in the data set are preprocessed, including data cleaning and data standardization;

[0017] Obtain the maximum and minimum values ​​of duration, the maximum and minimum values ​​of brightness difference, and the maximum and minimum values ​​of angle difference from the acquired data set;

[0018] The maximum and minimum values ​​of the duration are used as the upper and lower limits, respectively, to construct a data fluctuation interval, divide the data fluctuation interval into A duration sub-intervals, and calculate the probability distribution of each duration sub-interval Where P a represents the probability distribution of the a-th duration subinterval, a={1,2,...,A}, slc a represents the number of duration data points in the data set that fall into the ath duration subinterval, and ZN represents the total amount of data in the data set; the probability distribution P of each brightness difference subinterval is calculated according to the calculation method of the probability distribution of each duration subinterval b And the probability distribution P of each angle difference subinterval c , where P b represents the probability distribution of the b-th brightness difference subinterval, b={1,2,...,B}, P c represents the probability distribution of the c-th angle difference subinterval, c = {1, 2, ..., C};

[0019] Calculate the duration information entropy CXS, the expression is as follows

[0020] Calculate the brightness difference information entropy LDS, the expression is as follows

[0021] Calculate the angle difference information entropy JDS, the expression is as follows

[0022] Calculate the driver's headlight adjustment intervention feedback coefficient Jxyt, the expression is as follows Where xssj represents the vehicle driving time.

[0023] In a preferred embodiment, the logic for obtaining the lowest recognizable brightness rise abnormal coefficient is as follows:

[0024] Get the actual brightness value of the ambient light source that the intelligent headlight system fails to recognize each time, and record the time of the unrecognized event, establish a time series of the actual brightness values ​​of the unrecognized ambient light source in chronological order, compare the actual brightness values ​​of the ambient light sources of two adjacent unrecognized events, if the actual brightness value of the ambient light source of the latter unrecognized event is greater than the actual brightness value of the ambient light source of the previous unrecognized event, then add the actual brightness value of the ambient light source of the latter unrecognized event to the actual brightness value data set of the ambient light source of the unrecognized event, and calculate the minimum recognizable brightness rise abnormality coefficient Xdsb according to the actual brightness and time span in the actual brightness value data set. The expression is as follows where ldz j+1Indicates the j+1th actual brightness value in the actual brightness value data set, ldz j represents the jth actual brightness value in the actual brightness value data set, sjk j+1 represents the unidentified event time of the j+1th actual brightness value in the actual brightness value data set, sjk j Represents the unidentified event time of the jth actual brightness value in the actual brightness value data set, j = {1, 2, ..., J}, and J is a positive integer.

[0025] In a preferred implementation, the logic for obtaining the standby mode switching failure ratio is as follows:

[0026] In the historical driving records, the number of times the intelligent headlight system can successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF1, and the number of times the intelligent headlight system cannot successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF2, and the backup mode switching failure ratio Bymx is calculated. The expression is as follows

[0027] In a preferred embodiment, an intelligent headlight system performance evaluation model is constructed based on the headlight adjustment response delay coefficient, the driver's headlight adjustment intervention feedback coefficient, the lowest recognizable brightness rise abnormality coefficient, and the standby mode switching failure ratio to generate a headlight brightness adjustment hidden danger index CDTJ. The model is based on the following formula: CDTJ=4 α*Xcdx+β*Jxyt+γ*Xdsb+δ*Bymx , where α, β, γ, and δ represent the headlight adjustment response delay coefficient, the driver's headlight adjustment intervention feedback coefficient, the minimum recognizable brightness increase abnormal coefficient, and the preset proportional coefficient of the standby mode switching failure ratio, respectively, and α, β, γ, and δ are all greater than 0.

[0028] In a preferred embodiment, if the headlight brightness adjustment hidden danger index is greater than the headlight brightness adjustment hidden danger index threshold, a stability degradation signal is generated and the stability degradation signal is transmitted to the headlight adjustment optimization module; if the headlight brightness adjustment hidden danger index is less than or equal to the headlight brightness adjustment hidden danger index threshold, a normal stability signal is generated.

[0029] In a preferred embodiment, when the stability reduction signal is generated, the corresponding vehicle headlight brightness adjustment hidden danger index is obtained, and the difference between the vehicle headlight brightness adjustment hidden danger index and the vehicle headlight brightness adjustment hidden danger index threshold is calculated;

[0030] Collect ambient brightness data and driving data of oncoming vehicles;

[0031] The environmental brightness data includes but is not limited to the light source intensity and the rate of change of light illumination, and the driving data of the oncoming vehicle includes but is not limited to the vehicle speed and the distance between the oncoming vehicle and the own vehicle;

[0032] Preprocessing the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, illumination change rate, vehicle speed, and the distance between the oncoming vehicle and the own vehicle;

[0033] A machine learning headlight brightness adjustment model is constructed based on the bp neural network. The bp neural network includes an input layer, a hidden layer, and an output layer. The input layer contains five neurons, which correspond to the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, light change rate, vehicle speed, and the distance between the oncoming vehicle and the vehicle itself. The number of neurons in the hidden layer is twice the number of neurons in the input layer plus one. The number of neurons in the output layer is three, which correspond to the three outputs of the machine learning headlight brightness adjustment model: headlight brightness, illumination range, and illumination quantity.

[0034] The training process of the machine learning headlight brightness adjustment model includes forward propagation, loss calculation, and back propagation. The machine learning headlight brightness adjustment model is continuously optimized through multiple iterative processes until the loss function reaches the expected minimum value. The iterative process is stopped to complete the construction of the machine learning headlight brightness adjustment model.

[0035] Automatically adjust the headlights based on the machine learning headlight brightness adjustment model.

[0036] Technical effects and advantages of the present invention:

[0037] 1. The present invention acquires the headlight adjustment response information of the existing intelligent headlight system, the driver's headlight adjustment intervention feedback information, the minimum recognizable brightness increase information and the standby mode switching failure information, constructs an intelligent headlight system performance evaluation model, generates a headlight brightness adjustment hidden danger index, evaluates the stability of the headlight brightness adjustment of the existing intelligent headlight system during actual operation, timely perceives the operation hidden dangers of the intelligent headlight system before an obvious fault occurs, timely discovers the unstable factors of the system in actual operation, improves the stability of the intelligent headlight system, enhances driving safety, compares the headlight brightness adjustment hidden danger index with a preset headlight brightness adjustment hidden danger index threshold, identifies and classifies the stability of the headlight brightness adjustment of the existing intelligent headlight system during actual operation, acquires the corresponding generated headlight brightness adjustment hidden danger index when a stability reduction signal is generated, and calculates the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, constructs a machine learning headlight brightness adjustment model in combination with real-time environmental brightness data and driving data of oncoming vehicles, automatically adjusts the headlights, provides accurate headlight adjustment, and improves the safety and intelligence of automobile lighting. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0039] Figure 1 Schematic diagram of the structure of the system of the embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Embodiment: The present invention provides as Figure 1 The intelligent vehicle lighting system with adjustable brightness shown includes an information acquisition module, a performance evaluation module, a difference identification module, and a vehicle lighting adjustment optimization module;

[0042] An information acquisition module, used to acquire headlight adjustment response information of an existing intelligent headlight system, driver headlight adjustment intervention feedback information, minimum recognizable brightness increase information, and standby mode switching failure information;

[0043] Performance evaluation module, which builds a performance evaluation model for the intelligent headlight system, generates a risk index for headlight brightness adjustment, and evaluates the stability of the existing intelligent headlight system in adjusting headlight brightness during actual operation;

[0044] A difference identification module compares the risk index of headlight brightness adjustment with a preset risk index threshold of headlight brightness adjustment, and identifies and classifies the stability of headlight brightness adjustment of the existing intelligent headlight system during actual operation;

[0045] The headlight adjustment optimization module, when the stability of the headlight brightness adjustment of the existing intelligent headlight system decreases during actual operation, combines the real-time ambient brightness data and the driving data of oncoming vehicles to build a machine learning headlight brightness adjustment model to automatically adjust the headlights;

[0046] An information acquisition module, used to acquire headlight adjustment response information of an existing intelligent headlight system, driver headlight adjustment intervention feedback information, minimum recognizable brightness increase information, and standby mode switching failure information;

[0047] The headlight adjustment response information includes a headlight adjustment response delay coefficient, the driver headlight adjustment intervention feedback information includes a driver headlight adjustment intervention feedback coefficient, the minimum recognizable brightness increase information includes a minimum recognizable brightness increase abnormality coefficient, and the standby mode switching failure information includes a standby mode switching failure ratio;

[0048] The light adjustment response delay coefficient is used to measure the response delay and degree of the existing intelligent light system to the light adjustment during actual operation. When the existing intelligent light system has a response delay to the light adjustment during actual operation, it may cause the following effects:

[0049] Increased safety risks: At night or in low-light conditions, delays may cause the lights to fail to adapt to road and traffic conditions in a timely manner, which can affect the driver's field of vision and increase the risk of accidents;

[0050] Glare risk: Delayed adjustment may cause the high beam to fail to switch to low beam in time, affecting the sight of oncoming vehicles and pedestrians, and increasing the safety risks caused by glare;

[0051] Degraded driving experience: Drivers’ trust in lighting adjustment decreases, which may lead to more frequent manual intervention, affecting driving smoothness and comfort;

[0052] Increased energy consumption: Failure to adjust the light brightness in time may lead to unnecessary energy waste and increase the overall energy consumption of the car;

[0053] When the light adjustment response delay coefficient is larger, it indicates that the stability of the existing intelligent light system in actual operation is worse, and the probability of operation hazards is greater. By calculating the light adjustment response delay coefficient, the response speed of the intelligent light system can be quantified, which helps to evaluate its performance stability in various environments, timely discover and identify the response delay problem of the intelligent light system, and help to carry out targeted maintenance and optimization to reduce safety hazards.

[0054] The logic for obtaining the light adjustment response delay coefficient is as follows:

[0055] During the actual driving process of the vehicle, the time point sj1 of each change event of the ambient light source and the time point sj2 of the start of headlight adjustment when the change event occurs are recorded, and the response time xy from the change of the ambient light source to the start of headlight adjustment is calculated, and the expression is as follows xy=|sj2-sj1|;

[0056] Calculate the average response time xyp, the expression is as follows where xy i represents the response time calculated for the i-th ambient light source change event, i = {1, 2, ..., I}, I is a positive integer;

[0057] Compare the response time calculated for each ambient light source change event with the average response time, mark the response time calculated for the ambient light source change event that is greater than the average response time as the delayed response time xyc, and calculate the headlight adjustment response delay coefficient Xcdx, which is expressed as follows where xycn represents the response time marked as delayed response time for the nth time, n = {1, 2, ..., N}, N is a positive integer;

[0058] The driver's headlight adjustment intervention feedback coefficient is used to measure the frequency and degree of the driver's manual intervention in adjusting the headlights during the actual operation of the existing intelligent headlight system. The larger the driver's headlight adjustment intervention feedback coefficient is, the more it indicates that the existing intelligent headlight system has deficiencies in light source detection or adjustment accuracy, resulting in the system being unable to effectively adapt to environmental changes. The driver's trust in the intelligent headlight system may gradually decrease, and may tend to actively intervene in the adjustment rather than rely on the system's automatic adjustment. By calculating the driver's headlight adjustment intervention feedback coefficient, an important reference basis can be provided for monitoring the operating hazards of the existing intelligent headlight system, which is helpful for optimizing and adjusting the existing intelligent headlight system.

[0059] The logic for obtaining the driver's headlight adjustment intervention feedback coefficient is as follows:

[0060] The duration, brightness difference, and angle difference of the driver's active headlight adjustment are obtained during vehicle driving, and the duration, brightness difference, and angle difference of each active headlight adjustment by the driver are integrated into a data set, and the duration, brightness difference, and angle difference in the data set are preprocessed, including data cleaning and data standardization;

[0061] The data cleaning includes missing value processing and outlier detection. Common methods include filling with mean, median, mode or front and back values, or deleting missing value records, IQR outlier detection, and data standardization. Common methods include Min-Max normalization and Z-score standardization.

[0062] Obtain the maximum and minimum values ​​of duration, the maximum and minimum values ​​of brightness difference, and the maximum and minimum values ​​of angle difference from the acquired data set;

[0063] The maximum and minimum values ​​of the duration are used as the upper and lower limits, respectively, to construct a data fluctuation interval, divide the data fluctuation interval into A duration sub-intervals, and calculate the probability distribution of each duration sub-interval Where P a represents the probability distribution of the a-th duration subinterval, a={1,2,...,A}, slc a represents the number of duration data points in the data set that fall into the ath duration subinterval, and ZN represents the total amount of data in the data set; the probability distribution P of each brightness difference subinterval is calculated according to the calculation method of the probability distribution of each duration subinterval b And the probability distribution P of each angle difference subinterval c , where P brepresents the probability distribution of the b-th brightness difference subinterval, b={1,2,...,B}, P c represents the probability distribution of the c-th angle difference subinterval, c = {1, 2, ..., C};

[0064] Calculate the duration information entropy CXS, the expression is as follows

[0065] Calculate the brightness difference information entropy LDS, the expression is as follows

[0066] Calculate the angle difference information entropy JDS, the expression is as follows

[0067] Calculate the driver's headlight adjustment intervention feedback coefficient Jxyt, the expression is as follows Where xssj represents the vehicle driving time;

[0068] It should be noted that before calculating the driver's headlight adjustment intervention feedback coefficient, it is necessary to ensure that the duration information entropy, brightness difference information entropy, angle difference information entropy, and vehicle driving time are all normalized. Commonly used normalization methods include Min-Max normalization, Z-Score normalization, etc.

[0069] The minimum recognizable brightness rise abnormal coefficient is used to measure the intelligent headlight system's ability to identify the brightness of the ambient light source. The larger the minimum recognizable brightness rise abnormal coefficient, the worse the intelligent headlight system's ability to identify the ambient light source. It may have a trend of continuous decline. A higher minimum recognizable brightness rise abnormal coefficient may cause the system to be unable to effectively identify the light source or obstacle ahead at night or in bad weather conditions, thereby increasing the risk of accidents. By calculating the minimum recognizable brightness rise abnormal coefficient, potential problems of insufficient light source recognition can be effectively identified, ensuring the driver's visibility in complex environments and timely discovering potential operational hazards of the intelligent headlight system.

[0070] The logic for obtaining the lowest recognizable brightness increase abnormal coefficient is as follows:

[0071] Get the actual brightness value of the ambient light source that the intelligent headlight system fails to recognize each time, and record the time of the unrecognized event, establish a time series of the actual brightness values ​​of the unrecognized ambient light source in chronological order, compare the actual brightness values ​​of the ambient light sources of two adjacent unrecognized events, if the actual brightness value of the ambient light source of the latter unrecognized event is greater than the actual brightness value of the ambient light source of the previous unrecognized event, then add the actual brightness value of the ambient light source of the latter unrecognized event to the actual brightness value data set of the ambient light source of the unrecognized event, and calculate the minimum recognizable brightness rise abnormality coefficient Xdsb according to the actual brightness and time span in the actual brightness value data set. The expression is as follows where ldz j+1 Indicates the j+1th actual brightness value in the actual brightness value data set, ldz j represents the jth actual brightness value in the actual brightness value data set, sjk j+1 represents the unidentified event time of the j+1th actual brightness value in the actual brightness value data set, sjk j represents the unidentified event time of the jth actual brightness value in the actual brightness value data set, j = {1, 2, ..., J}, J is a positive integer;

[0072] The standby mode switching failure ratio is used to measure whether the intelligent headlight system can successfully switch to the standby mode in a very short time when a fault occurs. The larger the standby mode switching failure ratio is, the greater the proportion of the intelligent headlight system that fails to successfully switch to the standby mode in a very short time when a fault occurs, which indicates that the existing intelligent headlight system may have certain operational risks. Calculating the standby mode switching failure ratio can help evaluate the emergency response capability of the existing intelligent headlight system when a fault occurs and understand its stability in emergency situations. A high standby mode switching failure ratio indicates that the intelligent headlight system fails to quickly switch to the standby mode when a fault occurs, which may lead to driving safety risks.

[0073] The logic for obtaining the failure ratio of standby mode switching is as follows:

[0074] In the historical driving records, the number of times the intelligent headlight system can successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF1, and the number of times the intelligent headlight system cannot successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF2, and the backup mode switching failure ratio Bymx is calculated. The expression is as follows

[0075] It should be noted that the time period [t1, t2] can be determined by technicians in the field based on historical driving records and reference to the best practices and standards for emergency response of smart vehicle lighting systems in the industry to ensure that the preset time period meets safety requirements;

[0076] Performance evaluation module, which builds a performance evaluation model for the intelligent headlight system, generates a risk index for headlight brightness adjustment, and evaluates the stability of the existing intelligent headlight system in adjusting headlight brightness during actual operation;

[0077] According to the light adjustment response delay coefficient, the driver's light adjustment intervention feedback coefficient, the lowest recognizable brightness increase abnormal coefficient, and the standby mode switching failure ratio, an intelligent light system performance evaluation model is constructed to generate the light brightness adjustment hidden danger index CDTJ. The model is based on the following formula: CDTJ = 4 α*Xcdx+β*Jxyt+γ*Xdsb+δ*Bymx, where α, β, γ, and δ represent the light adjustment response delay coefficient, the driver light adjustment intervention feedback coefficient, the lowest recognizable brightness rise abnormal coefficient, and the preset proportional coefficient of the standby mode switching failure ratio, respectively, and α, β, γ, and δ are all greater than 0;

[0078] It should be noted that before building the intelligent headlight system performance evaluation model, it is necessary to ensure that the headlight adjustment response delay coefficient, the driver's headlight adjustment intervention feedback coefficient, the minimum recognizable brightness increase abnormal coefficient, and the standby mode switching failure ratio are all normalized. Common normalization methods include Min-Max normalization and Z-Score standardization. α, β, γ, and δ are set according to actual conditions. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of various indicators through professional opinion surveys and comprehensive evaluations;

[0079] It can be seen from the above calculation expression that the larger the headlight adjustment response delay coefficient, the larger the driver headlight adjustment intervention feedback coefficient, the larger the minimum recognizable brightness rise abnormal coefficient, and the larger the standby mode switching failure ratio, the larger the headlight brightness adjustment hidden danger index, which may indicate that the stability of the existing intelligent headlight system in headlight brightness adjustment during actual operation is worse, and the probability of its operation hidden danger is greater. Conversely, the smaller the headlight adjustment response delay coefficient, the smaller the driver headlight adjustment intervention feedback coefficient, the smaller the minimum recognizable brightness rise abnormal coefficient, and the smaller the standby mode switching failure ratio, the smaller the headlight brightness adjustment hidden danger index, indicating that the stability of the existing intelligent headlight system in headlight brightness adjustment during actual operation is better, and the probability of its operation hidden danger is smaller;

[0080] A difference identification module compares the risk index of headlight brightness adjustment with a preset risk index threshold of headlight brightness adjustment, and identifies and classifies the stability of headlight brightness adjustment of the existing intelligent headlight system during actual operation;

[0081] If the hidden danger index of headlight brightness adjustment is greater than the hidden danger index threshold of headlight brightness adjustment, it indicates that the stability of the headlight brightness adjustment of the existing intelligent headlight system shows a downward trend during actual operation, a stability reduction signal is generated, and the stability reduction signal is transmitted to the headlight adjustment optimization module; if the hidden danger index of headlight brightness adjustment is less than or equal to the hidden danger index threshold of headlight brightness adjustment, it indicates that the stability of the headlight brightness adjustment of the existing intelligent headlight system shows no obvious downward trend during actual operation, and a stability normal signal is generated;

[0082] The headlight adjustment optimization module, when the stability of the headlight brightness adjustment of the existing intelligent headlight system decreases during actual operation, combines the real-time ambient brightness data and the driving data of oncoming vehicles to build a machine learning headlight brightness adjustment model to automatically adjust the headlights;

[0083] When a stability reduction signal is generated, a corresponding headlight brightness adjustment hidden danger index is obtained, and a difference between the headlight brightness adjustment hidden danger index and a headlight brightness adjustment hidden danger index threshold is calculated to measure the degree of stability reduction of the headlight brightness adjustment of the existing intelligent headlight system during actual operation;

[0084] Collect ambient brightness data and driving data of oncoming vehicles;

[0085] The environmental brightness data includes but is not limited to the light source intensity and the rate of change of light illumination, and the driving data of the oncoming vehicle includes but is not limited to the vehicle speed and the distance between the oncoming vehicle and the own vehicle;

[0086] It should be noted that the ambient brightness data and the driving data of the oncoming vehicle can be obtained through vehicle-mounted sensors such as light sensors, radar sensors, and visual sensors (such as cameras);

[0087] Preprocessing the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, illumination change rate, vehicle speed, and the distance between the oncoming vehicle and the own vehicle;

[0088] A machine learning headlight brightness adjustment model is constructed based on the bp neural network. The bp neural network includes an input layer, a hidden layer, and an output layer. The input layer contains five neurons, which correspond to the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, light change rate, vehicle speed, and the distance between the oncoming vehicle and the vehicle itself. The number of neurons in the hidden layer is twice the number of neurons in the input layer plus one, and the number of neurons in the output layer is three, which correspond to the three outputs of the machine learning headlight brightness adjustment model, namely, headlight brightness, illumination range, and illumination quantity;

[0089] The training process of the machine learning headlight brightness adjustment model includes forward propagation, loss calculation, and back propagation. The machine learning headlight brightness adjustment model is continuously optimized through multiple iterative processes until the loss function reaches the expected minimum value. The iterative process is stopped to complete the construction of the machine learning headlight brightness adjustment model.

[0090] Automatically adjust the lights based on the machine learning light brightness adjustment model;

[0091] The present invention acquires headlight adjustment response information of an existing intelligent headlight system, headlight adjustment intervention feedback information of a driver, minimum identifiable brightness increase information, and standby mode switching failure information, so as to construct an intelligent headlight system performance evaluation model, generate a headlight brightness adjustment hidden danger index, evaluate the stability of the headlight brightness adjustment of the existing intelligent headlight system during actual operation, timely perceive the operation hidden dangers of the intelligent headlight system before an obvious fault occurs in the intelligent headlight system, timely discover the unstable factors of the system during actual operation, improve the stability of the intelligent headlight system, enhance driving safety, compare the headlight brightness adjustment hidden danger index with a preset headlight brightness adjustment hidden danger index threshold, identify and classify the stability of the headlight brightness adjustment of the existing intelligent headlight system during actual operation, acquire the corresponding generated headlight brightness adjustment hidden danger index when a stability reduction signal is generated, calculate the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, construct a machine learning headlight brightness adjustment model in combination with real-time environmental brightness data and driving data of oncoming vehicles, automatically adjust the headlights, provide accurate headlight adjustment, and improve the safety and intelligence of automobile lighting.

[0092] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0093] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0094] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0095] If the functions are implemented in the form of software functional units 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 the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0096] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A brightness adjustable intelligent vehicle lighting system, characterized in that: It includes information acquisition module, performance evaluation module, difference identification module and headlight adjustment optimization module; An information acquisition module, used to acquire headlight adjustment response information of an existing intelligent headlight system, driver headlight adjustment intervention feedback information, minimum recognizable brightness increase information, and standby mode switching failure information; Performance evaluation module, which builds a performance evaluation model for the intelligent headlight system, generates a risk index for headlight brightness adjustment, and evaluates the stability of the existing intelligent headlight system in adjusting headlight brightness during actual operation; A difference identification module compares the risk index of headlight brightness adjustment with a preset risk index threshold of headlight brightness adjustment, and identifies and classifies the stability of headlight brightness adjustment of the existing intelligent headlight system during actual operation; The headlight adjustment optimization module builds a machine learning headlight brightness adjustment model based on real-time ambient brightness data and driving data of oncoming vehicles to automatically adjust the headlights when the stability of the existing intelligent headlight system in adjusting the headlight brightness decreases during actual operation.

2. The intelligent vehicle lighting system with adjustable brightness according to claim 1, characterized in that: The headlight adjustment response information includes the headlight adjustment response delay coefficient, the driver headlight adjustment intervention feedback information includes the driver headlight adjustment intervention feedback coefficient, the minimum recognizable brightness increase information includes the minimum recognizable brightness increase abnormality coefficient, and the standby mode switching failure information includes the standby mode switching failure ratio.

3. The intelligent vehicle lighting system with adjustable brightness according to claim 2, characterized in that: The logic for obtaining the light adjustment response delay coefficient is as follows: During the actual driving process of the vehicle, the time point sj1 of each change event of the ambient light source and the time point sj2 of the start of headlight adjustment when the change event occurs are recorded, and the response time xy from the change of the ambient light source to the start of headlight adjustment is calculated, and the expression is as follows xy=|sj2-sj1|; Calculate the average response time xyp, the expression is as follows where xy i represents the response time calculated for the i-th ambient light source change event, i = {1, 2, ..., I}, I is a positive integer; Compare the response time calculated for each ambient light source change event with the average response time, mark the response time calculated for the ambient light source change event that is greater than the average response time as the delayed response time xyc, and calculate the headlight adjustment response delay coefficient Xcdx, which is expressed as follows where xyc n Indicates the response time that is marked as delayed response time for the nth time, n={1,2,...,N}, where N is a positive integer.

4. The intelligent vehicle lighting system with adjustable brightness according to claim 2, characterized in that: The logic for obtaining the driver's headlight adjustment intervention feedback coefficient is as follows: The duration, brightness difference, and angle difference of the driver's active headlight adjustment are obtained during vehicle driving, and the duration, brightness difference, and angle difference of each active headlight adjustment by the driver are integrated into a data set, and the duration, brightness difference, and angle difference in the data set are preprocessed, including data cleaning and data standardization; Obtain the maximum and minimum values ​​of duration, the maximum and minimum values ​​of brightness difference, and the maximum and minimum values ​​of angle difference from the acquired data set; The maximum and minimum values ​​of the duration are used as the upper and lower limits, respectively, to construct a data fluctuation interval, divide the data fluctuation interval into A duration sub-intervals, and calculate the probability distribution of each duration sub-interval Where P a represents the probability distribution of the a-th duration subinterval, a={1,2,...,A}, slc a represents the number of duration data points in the data set that fall into the ath duration subinterval, and ZN represents the total amount of data in the data set; the probability distribution P of each brightness difference subinterval is calculated according to the calculation method of the probability distribution of each duration subinterval b And the probability distribution P of each angle difference subinterval c , where P b represents the probability distribution of the b-th brightness difference subinterval, b={1,2,...,B}, P c represents the probability distribution of the c-th angle difference subinterval, c = {1, 2, ..., C}; Calculate the duration information entropy CXS, the expression is as follows Calculate the brightness difference information entropy LDS, the expression is as follows Calculate the angle difference information entropy JDS, the expression is as follows Calculate the driver's headlight adjustment intervention feedback coefficient Jxyt, the expression is as follows Where xssj represents the vehicle driving time.

5. The intelligent vehicle lighting system with adjustable brightness according to claim 2, characterized in that: The logic for obtaining the lowest recognizable brightness increase abnormal coefficient is as follows: Get the actual brightness value of the ambient light source that the intelligent headlight system fails to recognize each time, and record the time of the unrecognized event, establish a time series of the actual brightness values ​​of the unrecognized ambient light source in chronological order, compare the actual brightness values ​​of the ambient light sources of two adjacent unrecognized events, if the actual brightness value of the ambient light source of the latter unrecognized event is greater than the actual brightness value of the ambient light source of the previous unrecognized event, then add the actual brightness value of the ambient light source of the latter unrecognized event to the actual brightness value data set of the ambient light source of the unrecognized event, and calculate the minimum recognizable brightness rise abnormality coefficient Xdsb according to the actual brightness and time span in the actual brightness value data set. The expression is as follows where ldz j+1 Indicates the j+1th actual brightness value in the actual brightness value data set, ldz j represents the jth actual brightness value in the actual brightness value data set, sjk j+1 represents the unidentified event time of the j+1th actual brightness value in the actual brightness value data set, sjk j Represents the unidentified event time of the jth actual brightness value in the actual brightness value data set, j = {1, 2, ..., J}, and J is a positive integer.

6. The intelligent vehicle lighting system with adjustable brightness according to claim 2, characterized in that: The logic for obtaining the failure ratio of standby mode switching is as follows: In the historical driving records, the number of times the intelligent headlight system can successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF1, and the number of times the intelligent headlight system cannot successfully switch to the backup mode within the time period [t1, t2] when a fault occurs is counted CF2, and the backup mode switching failure ratio Bymx is calculated. The expression is as follows 7. The intelligent vehicle lighting system with adjustable brightness according to claim 2, characterized in that: According to the light adjustment response delay coefficient, the driver's light adjustment intervention feedback coefficient, the lowest recognizable brightness increase abnormal coefficient, and the standby mode switching failure ratio, an intelligent light system performance evaluation model is constructed to generate the light brightness adjustment hidden danger index CDTJ. The model is based on the following formula: CDTJ = 4 α*Xcdx+β*Jxyt+γ*Xdsb+δ*Bymx , where α, β, γ, and δ represent the headlight adjustment response delay coefficient, the driver's headlight adjustment intervention feedback coefficient, the minimum recognizable brightness increase abnormal coefficient, and the preset proportional coefficient of the standby mode switching failure ratio, respectively, and α, β, γ, and δ are all greater than 0.

8. The intelligent vehicle lighting system with adjustable brightness according to claim 7, characterized in that: If the headlight brightness adjustment hidden danger index is greater than the headlight brightness adjustment hidden danger index threshold, a stability reduction signal is generated, and the stability reduction signal is transmitted to the headlight adjustment optimization module; If the vehicle headlight brightness adjustment hidden danger index is less than or equal to the vehicle headlight brightness adjustment hidden danger index threshold, a stability normal signal is generated.

9. The intelligent vehicle lighting system with adjustable brightness according to claim 8, characterized in that: When a stability reduction signal is generated, a corresponding headlight brightness adjustment hidden danger index is obtained, and a difference between the headlight brightness adjustment hidden danger index and a headlight brightness adjustment hidden danger index threshold is calculated; Collect ambient brightness data and driving data of oncoming vehicles; The environmental brightness data includes but is not limited to the light source intensity and the rate of change of light illumination, and the driving data of the oncoming vehicle includes but is not limited to the vehicle speed and the distance between the oncoming vehicle and the own vehicle; Preprocessing the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, illumination change rate, vehicle speed, and the distance between the oncoming vehicle and the own vehicle; A machine learning headlight brightness adjustment model is constructed based on the bp neural network. The bp neural network includes an input layer, a hidden layer, and an output layer. The input layer contains five neurons, which correspond to the difference between the headlight brightness adjustment hidden danger index and the headlight brightness adjustment hidden danger index threshold, light source intensity, light change rate, vehicle speed, and the distance between the oncoming vehicle and the vehicle itself. The number of neurons in the hidden layer is twice the number of neurons in the input layer plus one. The number of neurons in the output layer is three, which correspond to the three outputs of the machine learning headlight brightness adjustment model: headlight brightness, illumination range, and illumination quantity. The training process of the machine learning headlight brightness adjustment model includes forward propagation, loss calculation, and back propagation. The machine learning headlight brightness adjustment model is continuously optimized through multiple iterative processes until the loss function reaches the expected minimum value. The iterative process is stopped to complete the construction of the machine learning headlight brightness adjustment model. Automatically adjust the headlights based on the machine learning headlight brightness adjustment model.

Images