Vehicle-mounted HUD environment illumination self-adaptive brightness adjusting system

Abstract

The invention relates to the technical field of vehicle-mounted display, and discloses a vehicle-mounted HUD environment illumination adaptive brightness adjusting system, which comprises an illumination acquisition module, a parameter analysis module, a model construction module, a brightness calculation module, an error evaluation module, a calibration correction module, a threshold setting module, a result output module, a feedback optimization module, a system regulation module and the like. The illumination acquisition module analyzes the illumination parameter stability and interference difference and generates an illumination perception state value; the parameter analysis module screens an optimal brightness and light intensity combination; the model construction module is used for training samples to establish a brightness and light intensity relation model; the brightness calculation module generates a brightness threshold by combining real-time illumination data mapping; the error evaluation and calibration correction module optimizes a calculation rule; and the feedback optimization module operates feedback adjustment parameters to finally generate an environment illumination adaptive brightness adjustment scheme. The system realizes accurate and dynamic adjustment of the brightness of the vehicle-mounted HUD, effectively adapts to the environment illumination change, improves the display effect and the driving safety, and has the characteristics of self-learning and self-optimization.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle display, in particular to a brightness adjustment system for vehicle HUD ambient light adaptation. BACKGROUND

[0002] In the context of rapid development of automotive electronics technology, the vehicle head-up display (HUD) system has become an important configuration for improving driving safety and convenience. However, the existing vehicle HUD brightness adjustment technology has many shortcomings in dealing with complex ambient light. On the one hand, the traditional brightness adjustment system lacks precision in collecting ambient light, making it difficult to effectively distinguish between normal ambient light and interfering light (such as direct strong light, glare, etc.), resulting in poor stability of light parameters and inability to accurately reflect the true ambient light conditions. For example, when the vehicle exits a tunnel or encounters sudden strong light, the existing system often fails to capture the sudden change in light, resulting in mismatch between HUD display brightness and ambient light, causing driver visual glare or blurred display content.

[0003] The existing technology lacks dynamic optimization mechanism in parameter analysis and model construction. Most systems use fixed parameter models that cannot adaptively adjust based on real-time light data and display effects, resulting in low matching accuracy of brightness and light intensity. For example, under different weather conditions (sunny, cloudy, rainy) or different time periods (day, dusk, night), the fixed model cannot balance the comfort of brightness and the clarity of display content, and is prone to problems such as brightness adjustment lag or over-adjustment.

[0004] The existing brightness calculation and error evaluation methods have obvious defects. Traditional systems usually only calculate brightness based on simple light intensity values, without fully considering light feature change trends and error distribution, resulting in large deviations between calculated thresholds and actual needs. At the same time, there is a lack of effective calibration correction and feedback optimization mechanism, which cannot analyze and adjust errors in real time during the brightness adjustment process, and errors can accumulate over time, further reducing the adjustment accuracy. These problems not only affect the visual experience of drivers, but also increase the risk of driving safety due to unclear display information, especially in complex light environments, the shortcomings of existing technology are more prominent. Therefore, a vehicle HUD brightness adjustment system that can accurately perceive ambient light, dynamically optimize adjustment parameters, and real-time calibrate errors is needed to meet the increasingly complex driving environment requirements. SUMMARY

[0005] The present application relates to the technical field of vehicle display, in particular to a brightness adjustment system for vehicle HUD ambient light adaptation.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solution: a brightness adjustment system for vehicle HUD ambient light adaptation, the system comprising: The light collection module collects the range of the intensity of the ambient light of the vehicle-mounted HUD and the value of the interference light intensity, analyzes the stability of the light parameters and the difference of the interference, calculates the difference of the light perception between the environments, integrates the light perception parameter set, and obtains the light perception state value; The parameter analysis module extracts the feature combination of the HUD display brightness and the ambient light based on the light perception state value, screens the optimal brightness and light intensity combination, and obtains the parameter analysis parameter set. The model construction module extracts the current model training sample variation based on the parameter analysis parameter set, analyzes the relationship between the brightness and the light intensity, matches the sample features and the generated combination, and obtains the model construction parameter set. The brightness calculation module extracts the current HUD display brightness feature value based on the model construction parameter set, combines the real-time light data, performs mapping of the features and the brightness, sets the calculation rules and applies the rules to the matching of the light features, and obtains the brightness calculation threshold. The error evaluation module captures the brightness sampling point feature data based on the brightness calculation threshold, combines the feature changes of the sampling points caused by the ambient light, infers the feature changes, analyzes the feature change trend corresponding to the brightness calculation, classifies and organizes the feature change trend according to the inference result, adjusts and analyzes the brightness features of the classified data according to the feature change information, and obtains the error distribution evaluation value. The calibration correction module analyzes the error value of the features and the brightness based on the error distribution evaluation value through the real-time features and the calculation data, adjusts the calculation rules according to the error value, and obtains the brightness calibration correction scheme. The threshold setting module extracts the calibrated brightness parameters based on the brightness calibration correction scheme, analyzes the parameter fluctuation range, sets the brightness threshold interval and applies it to the error distribution evaluation, and obtains the brightness threshold parameter set. The result output module integrates the brightness calculation results based on the brightness threshold parameter set, formats the data according to the HUD display mode, and generates standardized output data. The feedback optimization module collects system operation feedback information based on the standardized output data, analyzes the source of the brightness error, adjusts the model training parameters and the parameter analysis rules, and obtains the system optimization parameter set. The system control module generates an adaptive HUD brightness adjustment and control scheme for the ambient light based on the system optimization parameter set.

[0007] Preferably, the light perception state value includes a disturbance light intensity parameter set, a delay response parameter set, a perception difference parameter set, the parameter analysis parameter set includes a brightness feature parameter, a light intensity feature parameter, a combination screening parameter, the model construction parameter set includes a sample change parameter, a feature matching parameter, a training adjustment parameter, the brightness calculation threshold includes a feature change parameter, a light matching parameter, a rule setting parameter, the error distribution evaluation value includes a feature trend analysis parameter, a brightness feature relationship parameter, the brightness calibration correction scheme includes an error analysis parameter, a rule adjustment parameter, the brightness threshold parameter set includes a parameter fluctuation parameter, an interval setting parameter, the standardized output data includes a mode integration parameter, a format processing parameter, the system optimization parameter set includes a feedback analysis parameter, a model adjustment parameter, and the adaptive HUD brightness adjustment and regulation scheme of the ambient light includes a configuration synchronization parameter and a process coordination parameter.

[0008] Preferably, the light collection module comprises: A data collection submodule collects the stability value and the interference value in the region based on the vehicle-mounted HUD ambient light intensity collection range and the interference light intensity value, locates invalid light data, removes abnormal interference data, arranges the extracted stability value and interference value in the order of the region, and generates a light parameter interference data set; An intensity analysis submodule analyzes the stability and interference between environments based on the light parameter interference data set, calculates the light parameter change ratio, sorts the difference values between environments by weight, marks the region with too large fluctuation difference, and obtains light perception difference data; A distribution integration submodule calls the light perception difference value based on the light perception difference data, performs multi-dimensional summary, screens light perception value difference, classifies by value size, and sequentially arranges the light perception value to generate a light perception state value.

[0009] Preferably, the parameter analysis module comprises: A parameter extraction submodule identifies the display features of the HUD in each region based on the light perception state value, records the light intensity and adjustment data of the brightness, normalizes the recorded data, classifies the normalized data by brightness and light intensity, and generates a feature parameter data set; A feature optimization submodule analyzes the brightness and light intensity values in the data set according to the feature parameter data set, screens parameter combinations with high matching degree with the perception state, records the matching results through feature matching, adjusts the parameter combinations, and generates a parameter combination optimization result; A parameter selection submodule retrieves the parameter combination optimization result, determines the brightness and light intensity combination with the optimal matching degree, adjusts the parameter analysis parameters, inputs the generated configuration, verifies the stability of the parameter set, and generates a parameter analysis parameter set.

[0010] Preferably, the model construction module comprises: a sample analysis submodule that analyzes a parameter set based on the parameters, collects sample data from a training library, including brightness deviation, light intensity distortion, and stray light effects, performs time series analysis on the data, eliminates abnormal samples, partitions the remaining data, and obtains sample analysis data; a parameter matching submodule that analyzes the influence of sample variables on brightness extraction and light intensity extraction based on the sample analysis data, calculates the influence degree of the change of each sample feature on parameter adjustment, determines the optimal matching parameter setting according to the influence score, adjusts the parameters in a loop to capture the optimal combination, and obtains a parameter docking result; a training integration submodule that selects a brightness and light intensity combination that matches the current sample condition from the parameter docking result, performs parameter training experiments, optimizes the parameter setting through multiple training and verification, determines and solidifies the parameters as a model standard, and generates a model construction parameter set.

[0011] Preferably, the brightness calculation module comprises: a feature extraction submodule that locates a brightness feature monitoring point based on the model construction parameter set, extracts feature change values in the monitoring area, continuously records feature increase and decrease rates, extracts a plurality of key change nodes corresponding to the change rates, sorts the node values in order, and obtains a current feature change feature value; a light matching submodule that analyzes node change values and real-time light data based on the current feature change feature value, calibrates according to a predetermined matching criterion, calls the matching criterion to perform feature redistribution within the light interval, and obtains a feature light matching structure; a rule allocation submodule that uses a brightness rule dynamic adjustment method to measure the distribution of features between change nodes based on the feature light matching structure, sets upper and lower limits for node rules, applies the rules to regional feature values, and allocates them to obtain a brightness calculation threshold.

[0012] Preferably, the brightness rule dynamic adjustment method comprises obtaining an allocation rule value of a matching feature change node, setting a node rule lower limit through feature weight, light weight, and dynamic adjustment weight to control the minimum matching.

[0013] Preferably, the error evaluation module comprises: a feature data capture submodule that applies an ambient light algorithm based on the brightness calculation threshold, captures feature data of sampling points, eliminates abnormal values and corrects errors, stores them in layers according to intervals, performs feature processing, and generates a feature data dataset; a brightness analysis submodule that divides intervals according to brightness categories based on the feature data dataset, extracts change trends and fluctuation features, and generates a brightness category and feature change feature set; The classification inference submodule adjusts feature parameters and calibrates trend data based on the brightness category and the feature change feature set, extracts a classification interval, and performs numerical prediction to obtain an error distribution evaluation value.

[0014] Preferably, the ambient light algorithm comprises calculating feature feature values, generating feature data sets, and comprehensively calculating feature values after feature through data point weights, sampling point original feature values, light values and category coefficients.

[0015] Preferably, the feedback optimization module comprises: The data analysis submodule extracts real-time features and calculates data based on the standardized output data, analyzes real-time brightness values and calculated values, corresponds brightness difference values to current feature information, and generates feature brightness error values; The parameter adjustment submodule sets training adjustment parameters of the model based on feature brightness error values, sets adjustment amplitudes for brightness categories with large errors, fine-tunes low error categories, filters and integrates matching parameter sets through comparison of adjustment effects, and generates a model adjustment parameter set. The optimization control submodule applies adjustment parameters at each training link based on the model adjustment parameter set, implements model optimization item by item, synchronously monitors features and brightness, gradually adjusts the optimization order of each brightness category model, and generates a system optimization parameter set.

[0016] Compared with the prior art, the beneficial effects of the present application are: The vehicle-mounted HUD ambient light self-adaptive brightness adjustment system realizes accurate perception and dynamic brightness adjustment of ambient light through the cooperation of multiple modules, significantly improves the HUD display effect and driving safety. The light collection module can deeply analyze the stability of light parameters and interference differences based on specific light intensity collection range and interference light intensity value, accurately calculate the light perception difference between environments, and generate a reliable basis for subsequent adjustment. This fine processing of light data effectively solves the problem of inaccurate perception of traditional systems in complex light environments, enabling the system to quickly identify interference such as strong light and backlight, laying a foundation for subsequent brightness adjustment.

[0017] The parameter analysis module extracts the optimal brightness and light intensity combination from the light perception state value, through feature extraction, optimization and selection steps, ensuring that the parameter combination is highly matched with the actual environment. Compared with the traditional fixed parameter mode, this module can dynamically filter the best parameters according to different light scenes, avoiding the problem of brightness adjustment lag or over-adjustment caused by fixed parameters, enabling the HUD display brightness to adapt to environmental changes in real time, and improving the visual comfort of the driver.

[0018] The model construction module constructs a dynamically adjustable brightness adjustment model based on the parameter analysis results through sample analysis, parameter matching, and training integration, etc. The model can analyze the relationship between brightness and light intensity according to the real-time sample data variation, effectively solving the problem of lack of adaptability of traditional models. For example, under different weather or time period light conditions, the model can automatically adjust the parameters to ensure the accuracy and timeliness of brightness adjustment.

[0019] The brightness calculation module combines model parameters and real-time light data to realize accurate setting of brightness calculation threshold through feature extraction, light matching, and rule assignment, etc. Especially, the application of the brightness rule dynamic adjustment method effectively controls the minimum matching degree by setting the lower limit of the node rule with multiple weights, so that the brightness calculation can more flexibly adapt to the changes of different light characteristics, further improving the accuracy of brightness adjustment.

[0020] The error evaluation and calibration correction module can real-time find and correct the deviation in the brightness calculation process by capturing the feature data of the brightness sampling points, analyzing the feature change trend, and adjusting the error. This closed-loop feedback mechanism ensures that the error does not accumulate in the long-term operation of the system, maintaining the stability and reliability of the brightness adjustment. The threshold setting module and the result output module further standardize the brightness threshold interval and the output format, making the system output more standardized.

[0021] The feedback optimization module and the system control module continuously adjust the model training parameters and analysis rules by collecting running feedback information, realizing the self-optimization of the system. This dynamic optimization mechanism enables the system to continuously adapt to new light environments and use scenarios in long-term use, avoiding the problem of declining adjustment effect caused by environmental changes in traditional systems, fundamentally improving the environmental adaptability and reliability of the vehicle-mounted HUD brightness adjustment system, and providing drivers with a safer and more comfortable driving experience. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 The working principle diagram of the vehicle-mounted HUD environmental light self-adaptive brightness adjustment system described in the present application; Figure 2 The design diagram of the light collection module; Figure 3 The design diagram of the parameter analysis module; Figure 4 The design diagram of the model construction module. DETAILED DESCRIPTION

[0023] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0024] Please refer to Figures 1-4 The present application relates to a vehicle-mounted HUD ambient light adaptive brightness adjustment system, and the specific implementation is as follows: The light collection module collects the range of the intensity of the ambient light of the vehicle-mounted HUD and the intensity of the interference light, analyzes the stability of the light parameters and the difference in interference, calculates the difference in light perception between environments, integrates into a set of light perception parameters, and obtains a light perception state value.

[0025] The parameter analysis module extracts the feature combination of the HUD display brightness and the ambient light based on the light perception state value, selects the optimal brightness and light intensity combination, and obtains a set of parameter analysis parameters.

[0026] The model construction module extracts the current model training sample variation based on the set of parameter analysis parameters, analyzes the relationship between brightness and light intensity, matches the sample features and generated combinations, and obtains a set of model construction parameters.

[0027] The brightness calculation module extracts the current HUD display brightness feature value based on the set of model construction parameters, combines real-time light data, performs mapping of features and brightness, sets calculation rules and applies the rules to the matching of light features, and obtains a brightness calculation threshold.

[0028] The error evaluation module captures the feature data of the brightness sampling points based on the brightness calculation threshold, combines the feature changes of the sampling points caused by the ambient light, analyzes the feature change trend corresponding to the brightness calculation, classifies and organizes the feature change trend according to the inference result, adjusts and analyzes the brightness features of the classified data based on the feature change information, and obtains an error distribution evaluation value.

[0029] The calibration correction module analyzes the error value of the features and brightness based on the error distribution evaluation value through real-time features and calculation data, adjusts the calculation rules based on the error value, and obtains a brightness calibration correction scheme.

[0030] The threshold setting module extracts the calibrated brightness parameters based on the brightness calibration correction scheme, analyzes the parameter fluctuation range, sets the brightness threshold interval and applies it to the error distribution evaluation, and obtains a set of brightness threshold parameters.

[0031] The result output module integrates the brightness calculation results based on the set of brightness threshold parameters, formats the data according to the HUD display mode, and generates standardized output data.

[0032] The feedback optimization module collects system operation feedback information based on standardized output data, analyzes the sources of brightness error, adjusts model training parameters and parameter analysis rules, and obtains the system optimization parameter set.

[0033] The system control module generates an adaptive HUD brightness adjustment scheme based on the system optimization parameter set.

[0034] Example 1: The set of disturbance light intensity parameters included in the illumination perception state value is a series of parameters used to characterize the intensity changes of interfering light in the environment. In practical applications, these parameters are obtained by collecting and quantifying the intensity of interfering light incident at different time periods and angles. For example, during vehicle operation, the intensity changes of interfering light from different sources such as direct sunlight, road reflections, and other vehicle lights with time and vehicle driving status are recorded in this parameter set. The delay response parameter set reflects the system's response delay characteristics to changes in illumination. It covers parameters related to the time interval between a change in ambient illumination and the system starting to react and adjust the HUD brightness, as well as the time delay parameters at each stage of the adjustment process. Obtaining these parameters requires real-time monitoring and recording of the system's response time under different illumination change scenarios. The perception difference parameter set reflects the degree of difference in illumination perception under different environments. For example, in different environments such as sunny days, cloudy days, rainy days, and inside tunnels, the difference parameters between the data collected by the same illumination intensity sensor and the actual illumination intensity perceived by the human eye are determined by comparing subjective human evaluation and objective sensor data.

[0035] The brightness characteristic parameters in the parameter analysis set describe various characteristics of the HUD display brightness, including maximum, minimum, and average brightness values, variation range, and rate of change. These parameters are obtained through real-time monitoring and analysis of the HUD's brightness output in different display modes. Luminous intensity characteristic parameters characterize the intensity features of ambient light. Besides basic luminous intensity values, these include luminous intensity distribution characteristics and frequency of change, such as the difference in intensity distribution between the horizontal and vertical directions, and the periodic variation of luminous intensity over time. These parameters are obtained through multi-dimensional acquisition and analysis by an ambient light sensor. The combination screening parameters play a role in selecting the optimal combination of brightness and luminous intensity. These include various weighting parameters and threshold parameters used to evaluate the matching degree of different combinations. For example, when evaluating brightness and luminous intensity combinations, different weights are assigned to factors such as eye comfort and display clarity; these weighting parameters are part of the combination screening parameters.

[0036] The sample change parameter in the model construction parameter set records the change amount of the model training sample, including the related parameters of sample data addition, deletion, modification, etc., and the change amplitude parameter of the sample data characteristic value, etc. In the model training process, with the continuous input of new sample data, these parameters are updated in real time to reflect the change of the sample. The feature matching parameter is used for matching the sample feature and the generated combination, which includes the calculation parameter of feature similarity, the matching threshold parameter, etc. For example, when matching the sample feature and the generated combination, a feature similarity threshold is set, and when the similarity of the two reaches or exceeds the threshold, it is considered that the matching is successful. This threshold is one of the feature matching parameters. The training adjustment parameter can adjust the model training process, including the learning rate, the number of iterations, the regularization parameter, etc. The setting of these parameters will directly affect the training effect and convergence speed of the model, which needs to be adjusted according to the actual situation during the training process.

[0037] The feature change parameter in the brightness calculation threshold represents the change of the brightness feature, including the change amplitude, change direction, change frequency, etc. of the brightness feature, which is obtained by real-time monitoring and analysis of the HUD display brightness feature. The light matching parameter is used for matching the feature with the real-time light data, including the sampling frequency of the light data, the related parameters of the matching algorithm, etc. For example, when matching the brightness feature with the real-time light data, a certain matching algorithm is used, and the parameters in the algorithm are the light matching parameters. The rule setting parameter is used when setting the calculation rule, including the condition parameter, weight parameter, etc. of various calculation rules. For example, when setting the brightness calculation rule, it is specified that when the ambient light intensity reaches a certain threshold, the HUD brightness is adjusted by a certain proportion. The threshold and proportion parameters here are rule setting parameters.

[0038] The feature trend analysis parameter in the error distribution evaluation value is used to analyze the trend of the brightness calculation corresponding feature change, including the direction, slope, stability, etc. of the trend, which is obtained by statistical analysis of the feature change data in the brightness calculation process. The brightness feature relationship parameter reflects the relationship between the brightness features, such as the correlation parameter between different brightness features, the degree of mutual influence parameter, etc., which is determined by joint analysis of multiple brightness features.

[0039] The error analysis parameter in the brightness calibration correction scheme is used for error value analysis of the feature and the brightness, including the error calculation method parameter, error range parameter, etc. For example, when calculating the error of the feature and the brightness, methods such as mean square error or absolute error are used. The related parameters of these methods are error analysis parameters. The rule adjustment parameter can adjust the calculation rule, including the modification method, adjustment amplitude parameter, etc. When it is found that the calculation rule has an error, the rule is adjusted according to the error analysis result, and the adjustment amplitude and method are determined by the rule adjustment parameter.

[0040] The parameter fluctuation parameter in the luminance threshold parameter set reflects the fluctuation of the calibrated luminance parameter, including the range, frequency, and amplitude of the fluctuation, which is obtained by real-time monitoring and statistics of the calibrated luminance parameter. The interval setting parameter is used to set the luminance threshold interval, including the upper and lower limit parameters of the interval, the standard parameter for interval division, etc. For example, according to different vehicle driving environments, the luminance threshold interval is divided into multiple sub-intervals, and the upper and lower limit parameters of each sub-interval are the interval setting parameters.

[0041] The mode integration parameter in the standardized output data is used to integrate different HUD display modes, including priority parameters and switching condition parameters of different modes. For example, when integrating navigation mode, multimedia mode and other display modes, the priority of each mode and the switching condition parameters are set, which are the mode integration parameters. The format processing parameter formats the data, including the encoding method, storage format, and transmission protocol parameters, to ensure that the output data can be correctly received and processed by the HUD display device.

[0042] The feedback analysis parameter in the system optimization parameter set is used to analyze system running feedback information, including feedback information collection frequency and analysis method parameters. By collecting and analyzing the feedback information generated during system operation, the source of luminance error is found. The model adjustment parameter can adjust the model training parameter, including model structure adjustment parameter and training algorithm parameter. According to the feedback analysis result, the model training parameter is adjusted to improve the accuracy and adaptability of the model.

[0043] The configuration synchronization parameter in the environmental light adaptive HUD luminance adjustment and control scheme ensures the synchronization of system configuration, including configuration information transmission frequency and consistency verification parameters. For example, when updating the configuration of each module in the system, the configuration synchronization parameter is used to ensure the consistency of the configuration information of each module. The process coordination parameter is used to coordinate the operation of each process in the system, including the execution order and time interval parameters of each process, to ensure that each process in the system can run in order and efficiently, and to realize adaptive adjustment of HUD luminance. These parameters are interrelated and interact with each other to form the parameter system of the system, so that the system can accurately adjust the display brightness of the HUD according to the change of environmental light, and improve the visual comfort and driving safety of the driver. In practical application, these parameters need to be reasonably set and optimized according to the specific vehicle model, HUD device characteristics, and use environment, etc., to ensure that the performance of the system reaches the best state.

[0044] Example 2: The module includes a data acquisition submodule, an intensity analysis submodule, and a distribution integration submodule, each submodule realizes accurate perception of ambient light through hierarchical data processing logic.

[0045] The data acquisition submodule collects the stability value and the interference value in the region based on the vehicle-mounted HUD ambient light intensity collection range and the interference light intensity value. In actual operation, first, according to the installation position and the field of view of the vehicle-mounted HUD, the collection area is divided into multiple sub-areas, such as the area above the front windshield and the area around the instrument panel, each sub-area corresponds to different light collection weights. Through multiple light sensors arranged inside and outside the vehicle, the light intensity data of each sub-area is collected in real time, including natural light, road reflection light, other vehicle lights, etc. In the collection process, the system will locate the invalid light data in real time, for example, when the sensor data is abnormal due to obstruction or failure, the invalid data is identified through a pre-set threshold judgment mechanism. The process of removing abnormal interference data combines time series analysis and spatial correlation analysis, for example, if the data of a certain sensor fluctuates beyond the normal range within a short period of time, and the data of adjacent sensors does not change similarly, it is determined that the data is abnormal and is removed. Then, arrange the stability value and the interference value extracted according to the region order to generate a light parameter interference data set, which contains the original value of the light intensity of each region, the interference light intensity component, the collection timestamp, etc.

[0046] The intensity analysis submodule analyzes the stability and interference between environments based on the light parameter interference data set. First, statistical analysis is performed on the stability value in the data set to calculate the mean, variance, and other parameters of the light intensity in each region to represent the stability of the ambient light. For the interference value, different frequency interference sources are identified through frequency domain analysis method, for example, high frequency interference may come from the flicker of vehicle lights, and low frequency interference may come from the influence of cloud movement on natural light. When calculating the light parameter change ratio, the difference and mean of light intensity of adjacent time points or adjacent regions are used to reflect the degree of light change. In the process of ordering the environmental difference value by weight, the setting of weight considers the importance of region and the degree of light influence, for example, the weight of the area directly in front of the front windshield is higher than that of the side area of the instrument panel. By setting a difference threshold, mark the area with too large fluctuation difference, for example, when the light change ratio of a certain area exceeds the pre-set threshold, it is determined that there is significant light interference in this area, thereby obtaining light perception difference data, which contains the position of the difference area, the difference amplitude, the difference occurrence time, etc.

[0047] The distribution integration module calls the light perception difference value for multi-dimensional summarization based on the light perception difference data. The multi-dimensional includes time dimension, space dimension and frequency dimension. The duration and trend of light difference are analyzed in the time dimension. The distribution range of difference area and the influence of adjacent area are analyzed in the space dimension. The distribution of different frequency interference is analyzed in the frequency dimension. When screening the light perception numerical difference, a clustering analysis method is used to classify similar difference values into a class. For example, high light difference values caused by direct sunlight are classified into a class, and difference values caused by strong light changes at the entrance of the tunnel are classified into another class. After classification according to the numerical value, the light perception value is arranged in order to generate a light perception state value. The state value exists in the form of a multi-dimensional vector. Each dimension corresponds to a different light feature, such as the change rate of the time dimension, the area difference coefficient of the space dimension, and the interference frequency component of the frequency dimension.

[0048] In the data collection process, the division of the collection area needs to be adjusted according to the specific model of the vehicle and the installation position of the HUD. For example, the angle of the front windshield of an SUV is different from that of a car, which will cause differences in the division of the light collection area. The positioning algorithm of invalid light data needs to be combined with the fault diagnosis information of the sensor and the historical data rules. For example, a sensor often has abnormal data in rainy weather, so the verification threshold of the sensor data is increased in rainy weather. The removal rule of abnormal interference data needs to consider dynamic environmental factors. For example, when the vehicle enters a tunnel, light data fluctuations for a short time are allowed to avoid misjudging normal light mutations as interference.

[0049] The setting of the weight in the intensity analysis needs to be determined through a large number of actual road tests. For example, in different weather and road conditions, the influence of light in each area on the driver's vision is tested to adjust the weight parameters accordingly. When marking areas with large fluctuation differences, the threshold setting needs to consider both accuracy and robustness. If the threshold is set too low, too many normal areas will be marked. If it is set too high, the real interference area may be missed.

[0050] The dimension selection of multi-dimensional summarization in distribution integration needs to be based on the actual needs of HUD brightness adjustment. For example, for night driving scenarios, more attention is paid to the influence of high-frequency interference, so the analysis weight in the frequency dimension is increased. The classification and sorting rules need to ensure that the light perception state value can accurately reflect the driver's visual perception. For example, light difference values that have a greater impact on vision are placed in the front row for subsequent modules to prioritize processing.

[0051] Embodiment 3: The parameter analysis module is composed of a parameter extraction submodule, a feature optimization submodule and a parameter selection submodule. Through the ordered connection of the data processing flow, the light perception state value and the HUD brightness feature are deeply analyzed.

[0052] The parameter extraction submodule identifies the display characteristics of each region of the HUD based on the illumination perception state value. Specifically, the system first divides the screen into different functional regions such as the main display area, auxiliary information area, etc. according to the display region division rules of the vehicle-mounted HUD, and each region corresponds to an independent brightness adjustment strategy. When identifying the display characteristics, the brightness output value, pixel grayscale distribution, dynamic contrast ratio, etc. of each region are obtained by reading the register data of the driving chip of the HUD or real-time acquisition of the voltage and current signals of the display panel. When recording the light intensity and adjustment data of the brightness, the real-time light intensity data of the ambient light sensor is synchronously collected, including multi-dimensional information such as spectral distribution, polarization characteristics, etc., and is associated with the current brightness adjustment action of the HUD, such as adjustment amplitude, adjustment direction and adjustment time, etc. When normalizing the recorded data, linear transformation or nonlinear mapping methods are used to map parameters of different dimensions to the [0, 1] interval, for example, the brightness value is mapped from [0, 1000] cd / m 2 to [0, 1], and the light intensity value is mapped from [0, 50000] lux to [0, 1]. The normalized data is sorted according to brightness and light intensity to generate a feature parameter data set, which contains fields such as timestamp, region identifier, normalized brightness value, normalized light intensity value, adjustment state, etc., forming a structured data record.

[0053] The feature optimization submodule analyzes the brightness and light intensity values in the feature parameter data set. In the analysis process, the Pearson correlation coefficient of brightness and light intensity is calculated using correlation analysis method to identify the linear relationship strength between the two; at the same time, the data points are divided into different illumination scene clusters such as sunny strong light, overcast weak light, tunnel transition, etc. by clustering algorithm. When screening parameter combinations with high matching degree to the perception state, a matching degree evaluation model is constructed, which considers factors such as human eye visual characteristic curve (such as CIE V(λ) curve), HUD display color gamut coverage range, ambient light interference coefficient, etc. to calculate the matching degree score for each parameter combination. For example, in sunny scenes, the matching degree model will focus on the dynamic range matching of brightness and light intensity; in night scenes, it will focus on the brightness uniformity matching under low light intensity. When matching features and recording matching results, sliding window technology is used to perform time series matching on historical data, and the window size is dynamically adjusted according to the illumination change rate, such as reducing the window size to improve real-time performance when there is a sudden change in strong light, and expanding the window size to accumulate more data when there is a slow change in weak light. When adjusting the parameter combination, the matching degree score is used as the optimization objective based on genetic algorithm or particle swarm optimization algorithm to iteratively generate a better parameter combination, and the parameter combination optimization result is generated, which contains several sets of optimized brightness-light intensity parameter pairs and corresponding scene labels.

[0054] The parameter selection sub-module retrieves the parameter combination optimization result to determine the optimal brightness and light intensity combination with the best matching degree. The retrieval process adopts a multi-condition filtering mechanism. First, the scene category is determined according to the current light perception state value, and then the parameter combinations are arranged in descending order of matching degree score under the scene category. When adjusting the parameters, the weight coefficients of the matching degree evaluation model, the threshold parameters of scene classification, and the like are analyzed. For example, the weight coefficient of the light intensity change rate is increased in the tunnel scene. After the input generation configuration is generated, the parameter set stability is verified. The verification methods include backtracking test and real-time stress test. The backtracking test uses historical light data playback to check the adjustment effect of the parameter combination in the historical scene. The real-time stress test monitors the response speed and adjustment accuracy of the parameter combination under simulated extreme light environments. When generating the parameter analysis parameter set, the parameter set contains core parameters such as the scene mapping table, the matching degree threshold, and the weight coefficient matrix, which are stored in a structured file format for subsequent module calls.

[0055] In the parameter extraction process, the division of the display area needs to be combined with the specific model of the HUD. For example, the display area division rules of AR-HUD and ordinary HUD are different. The former needs to consider the depth of field distribution of virtual images. The selection of the normalization processing method needs to consider the calculation efficiency and information retention. For the HUD display panel with nonlinear response, a nonlinear normalization method such as gamma correction needs to be used.

[0056] The construction of the matching degree evaluation model in feature optimization needs to be based on a large number of subjective evaluation experiments. Drivers of different ages and different vision levels are invited to score the HUD display effect. The scoring results are associated with objective parameters to determine the weights of each evaluation factor. The selection of the clustering algorithm needs to consider the data distribution characteristics. For non-convex distributed data, the DBSCAN density clustering algorithm is better than the K-means algorithm.

[0057] The division granularity of the scene category in parameter selection needs to be adjusted according to actual application requirements. Too fine division will increase the calculation complexity, and too coarse division will affect the adjustment accuracy. The test cases for stability verification need to cover various extreme light scenes, such as noon direct sunlight, rainy night headlight glare, and continuous entry and exit of tunnel groups, to ensure the reliability of the parameter set in complex environments.

[0058] Embodiment 4: The model construction module includes a sample analysis sub-module, a parameter matching sub-module, and a training integration sub-module. Each sub-module realizes the construction of the brightness adjustment model through data processing and algorithm iteration. Taking the application of a certain type of vehicle-mounted HUD in urban road environments as an example, the specific implementation of this module is as follows: The sample analysis submodule collects sample data from the training library based on the parameter analysis parameter set. The training library stores running data of the vehicle in different scenarios, for example, when the vehicle drives to an open road at noon on a sunny day, the ambient light intensity suddenly increases, and the HUD brightness needs to be correspondingly increased; for example, in the evening, the vehicle enters a tunnel, and the ambient light intensity suddenly decreases, and the HUD brightness needs to be quickly reduced to adapt to the driver's visual needs. The sample data includes brightness deviation, light intensity distortion, and stray light effects, etc. The brightness deviation is the difference between the actual display brightness and the expected brightness of the HUD, the light intensity distortion reflects the difference between the data collected by the ambient light sensor and the true light intensity, and the stray light effect includes the influence of interfering light from oncoming vehicle headlights, road reflection, etc.

[0059] When performing time series analysis on the data, the system will sort out the change trend of the sample data in chronological order. Taking the tunnel scenario as an example, from 10 seconds before the vehicle enters the tunnel to 20 seconds after the vehicle completely enters the tunnel, the light intensity data and the HUD brightness data in this period of time will be continuously recorded. By analyzing these data, it can be found that the light intensity will have a significant downward trend at the moment of entering the tunnel, and the HUD brightness will also be adjusted accordingly. When removing abnormal samples, if the light intensity data in a sample has an unreasonable and violent fluctuation in a short period of time, and the difference with the adjacent sample data is too large, the sample is determined to be abnormal and is removed. For example, under normal weather conditions, the light intensity data of a sample suddenly shows an extremely low value, while the light intensity data of the adjacent sample does not have this phenomenon, and the vehicle is not in a special scenario such as entering a tunnel at this time, the sample is considered to be abnormal. When partitioning the remaining data, the data can be divided into strong light area, medium light area and weak light area according to the size of the ambient light intensity. Taking 5000 lux as the boundary, 5000 lux and above is the strong light area, 1000-5000 lux is the medium light area, and 1000 lux and below is the weak light area. The data in each area is processed separately to obtain sample analysis data.

[0060] The parameter matching submodule analyzes the influence of sample variables on brightness extraction and light intensity extraction through sample analysis data. Taking the data in the strong light area as an example, the sample variables include the change rate of light intensity and the adjustment amplitude of HUD brightness, etc. When calculating the influence degree of the change of each sample feature on parameter adjustment, if the light intensity change rate is fast and the HUD brightness adjustment amplitude cannot keep up in time, it will lead to driver visual discomfort, and in this case, the influence degree of the change of this sample feature on parameter adjustment is larger. The optimal matching parameter setting is determined according to the influence score, and the parameters corresponding to the features with high influence scores will be adjusted first. For example, in the strong light area, the influence score of the light intensity change rate is higher, so when adjusting the parameters, the parameter settings related to the light intensity change rate will be mainly considered. When adjusting the parameters in a loop to capture the optimal combination, the system will continuously try different parameter combinations to find the combination that can make the HUD brightness adjustment effect optimal, thereby obtaining the parameter matching result.

[0061] The training integrated sub-module selects the brightness and light intensity combination matching the current sample condition from the parameter matching result for parameter training test. Taking the sample condition in the medium light area as an example, a set of brightness and light intensity combination is selected for brightness adjustment training of the HUD in the simulated medium light environment. Through multiple training and verification optimization of the parameter setting, the system will evaluate the brightness adjustment effect of the HUD after each training, and if the effect does not meet the expectation, the parameter will be adjusted and training will be performed again. For example, it is found that the brightness of the HUD is adjusted too much in the medium light environment after the first training, resulting in over-bright display, so the amplitude parameter of the brightness adjustment is reduced and training is performed again. After multiple training and adjustment, the parameters are determined and solidified as the model standard, and the model construction parameter set is generated.

[0062] In the sample analysis process, the construction of the training library needs to accumulate data of vehicles in various different scenarios for a long time. Taking a certain automobile manufacturer as an example, it can collect data by a large number of road test vehicles in different regions, different seasons and different weather conditions to ensure that the data of the training library is comprehensive and representative. The method of time series analysis needs to be selected according to the characteristics of the data. For data with obvious periodicity, periodic analysis method can be used; for non-periodic data, trend analysis method is used. The exclusion standard of abnormal samples needs to be adjusted according to the actual situation. For example, in severe weather conditions, the fluctuation range of light intensity data may increase, and the standard for excluding abnormal samples should also be relaxed accordingly.

[0063] In the parameter matching process, the calculation method of the influence degree needs to consider multiple factors, such as the visual perception of the driver, the display performance of the HUD, etc. The evaluation results of the driver on the HUD display effect under different parameter combinations can be combined with the objective data to calculate the influence degree of each sample feature. The evaluation standard needs to be clear, for example, the influence degree is divided into high, medium and low three levels, and different levels correspond to different scoring ranges. The cycle strategy of parameter adjustment needs to be reasonable to avoid falling into local optimal solution, and the combination of random adjustment and directional adjustment can be used.

[0064] In the training integration process, the number of training tests needs to be determined according to the complexity of the model and the difficulty of parameter adjustment. For a simple model, a few training may achieve the desired effect; for a complex model, more training times are needed. The verification method can include subjective evaluation and objective test, subjective evaluation is to invite the driver to score the display effect of the HUD, and objective test is to measure the brightness, contrast and other parameters of the HUD through professional equipment. The standard of parameter solidification needs to ensure that the model can work stably and reliably in various scenarios, for example, the brightness adjustment error of the HUD should be controlled within a certain range in different light intensity environments.

[0065] Example 5: The brightness calculation module includes a feature extraction submodule, a light matching submodule, and a rule allocation submodule. Each submodule realizes accurate calculation of the HUD display brightness through data processing and logical operation. Taking a new energy vehicle equipped with AR-HUD driving on a suburban road as an example, the specific implementation of this module is as follows: The feature extraction submodule locates the brightness feature monitoring points based on the model construction parameter set. In AR-HUD, the display area needs to correspond to the depth of field distribution of the virtual image, so the monitoring points are divided according to the positions of different display elements such as road signs and vehicle virtual models. For example, the virtual turning arrow area at 20 meters in front and the speed limit sign area at 50 meters are taken as independent monitoring points. When extracting the feature change value in the monitoring area, the image sensor is used to collect the pixel gray scale value of the HUD display picture in real time, and the change amount of the pixel gray scale per unit time is calculated. The feature increase and decrease rate is recorded continuously. Taking the turning arrow area as an example, when the vehicle is about to turn, the brightness of this area will increase in a short time, and the system will record the increase and decrease rate of the brightness per second during this process. A plurality of key change nodes corresponding to the change rate are extracted, for example, when the brightness increase and decrease rate exceeds a preset threshold (such as a change of 50 cd / m 2 ) per second, it is determined as a key change node. The node values are sorted in order to obtain the current feature change characteristic value, which includes the timestamp, brightness change amount, change rate, etc. of each node.

[0066] The light matching submodule analyzes the node change value and real-time light data based on the current feature change characteristic value. The real-time light data is collected by the ambient light sensor arranged on the roof, which includes light intensity, spectral distribution, polarization degree, etc. Taking the case of the vehicle entering the shadow area of the viaduct as an example, the ambient light sensor will detect that the light intensity decreases in a short time, and the feature extraction submodule will capture the change nodes of the HUD display brightness. Calibration is performed according to the predetermined matching criteria, which predefines the reference model of brightness change under different light scenes, for example, when the light intensity in the shadow area decreases, the HUD brightness should increase by 15%-20% within 0.5 seconds. The matching criteria are called to perform feature re-allocation in the light interval, match the light change characteristics of the shadow area with the pre-set model, adjust the brightness allocation proportion of each monitoring point, and obtain the feature light matching structure, which clearly defines the brightness adjustment priority and amplitude of each monitoring point under the current light condition.

[0067] The rule distribution sub-module adopts a brightness rule dynamic adjustment method based on a feature light matching structure to measure the distribution of features between change nodes. Taking a speed limit sign area as an example, during the process of a vehicle entering a shadow area, the distribution of the brightness feature of the area at each time node is measured from the time period when the light intensity starts to drop to the time period when the vehicle completely enters the shadow area. The upper and lower limits of the node rule are set, the lower limit of the node rule is set through the feature weight, the light weight and the dynamic adjustment weight, for example, the importance feature weight of the speed limit sign is 0.4, the light weight of the current light change is 0.3, and the dynamic adjustment weight is 0.3, and the minimum amplitude of the brightness adjustment of the area is determined to be 10% after weighted calculation of the three, so as to control the minimum matching. The rule is applied to the area feature value and distributed, for example, the brightness of the speed limit sign area is increased by 12% within 0.3 seconds, the turning arrow area is increased by 8%, and the other auxiliary information area is increased by 5%, so as to obtain the brightness calculation threshold, which includes the brightness target value, the adjustment time window, the priority coefficient and other parameters of each area.

[0068] The error evaluation module includes a feature data capture sub-module, a brightness analysis sub-module and a classification inference sub-module. The feature data capture sub-module applies an ambient light algorithm based on the brightness calculation threshold to capture feature data of sampling points. Taking a shadow area of a viaduct as an example, the brightness of each monitoring point is sampled every 20 ms during the HUD brightness adjustment process, and the collected data includes actual brightness value, ambient light intensity value, sampling time, etc. Abnormal values are removed and errors are corrected. If the difference between the brightness value of a sampling point and that of an adjacent sampling point exceeds 20% and the ambient light intensity does not change significantly, the value is determined to be abnormal and is removed, and the error is corrected by linear interpolation method. The feature data in each interval is stored separately, the sampling time is divided into multiple intervals, such as 0-0.5 seconds, 0.5-1 second, etc., the feature data in each interval is stored separately, and the feature data set is generated by feature processing, which includes the brightness mean value, the change variance, the light change rate and other feature parameters of each interval.

[0069] The brightness analysis sub-module divides the intervals according to the brightness categories based on the feature data set. The brightness categories are divided into high-brightness area (>800 cd / m 2 ), medium-brightness area (400-800 cd / m 2 ) and low-brightness area (<400 cd / m 2 ) according to the display requirements of the HUD. The change trend and fluctuation feature in each interval are extracted, for example, in the 0-0.5 second interval, the brightness of the speed limit sign area is increased from the medium-brightness area to the high-brightness area, the change trend is upward, and the fluctuation feature is the standard deviation 0.05. The brightness category and the feature change feature set are generated, which includes the time interval, the change amplitude, the fluctuation parameter and other information corresponding to each brightness category.

[0070] The classification inference submodule adjusts feature parameters and calibrates trend data based on the luminance category and the feature change feature set. Taking the speed limit sign area as an example, if the actual luminance increase amplitude does not meet the requirement of the calculated threshold, the weight coefficient in the feature parameter is adjusted, such as increasing the light weight to 0.4, and the trend data is recalibrated. The classification interval is extracted, and numerical prediction is performed. According to the adjusted parameters, the luminance change value in the next time interval is predicted, and the error distribution evaluation value is obtained. The evaluation value includes the luminance prediction error range, error occurrence probability and other information of each area.

[0071] The feedback optimization module includes a data analysis submodule, a parameter adjustment submodule, and an optimization control submodule. The data analysis submodule extracts real-time features and calculates data based on standardized output data. Taking the shadow area of the viaduct as an example, the actual luminance value (750 cd / m 2 ) and the calculated value (800 cd / m 2 ) of the speed limit sign area are extracted, and the difference between the two is -50 cd / m 2 . The luminance difference value is corresponded with the current feature information (such as light intensity drop rate, monitoring point importance) to generate a feature luminance error value.

[0072] The parameter adjustment submodule sets the training adjustment parameters of the model based on the feature luminance error value. For large error luminance categories (such as high-light area error exceeding 5%), a larger adjustment amplitude is set, for example, the learning rate of this category is increased by 0.1; for low error categories (such as medium-light area error <2%), fine-tuning is performed, and the learning rate is increased by 0.05. By comparing the adjustment effect, the matching parameter set is selected and integrated, for example, the learning rate is first adjusted, the error change is observed, if the error is reduced, the adjustment is retained, otherwise the original parameter is restored and other parameters are adjusted, and the model adjustment parameter set is generated.

[0073] The optimization control submodule applies the adjustment parameters in each training link based on the model adjustment parameter set. Taking the model training of the shadow area as an example, the new learning rate is applied in the parameter update link, and the model optimization is implemented item by item. The features and luminance are monitored synchronously, for example, every 50 training rounds, the luminance adjustment error of the shadow area is retested. The optimization order of each luminance category is adjusted gradually, the high-light area with larger error is optimized first, then the medium-light area and the low-light area are optimized, and the system optimization parameter set is generated. The parameter set includes the optimization priority of each luminance category, the adjustment parameter combination, the verification threshold and other information.

[0074] In the feature extraction process, the positioning of the monitoring points needs to be combined with the specific display content of the HUD, for example, the monitoring point division of the navigation information area is different from that of the multimedia information area, and the former needs a higher monitoring frequency for driving safety. The extraction of the feature change value needs to consider the human eye visual persistence characteristics, and the luminance features with rapid changes are smoothed to avoid misjudgment caused by high-frequency noise.

[0075] When matching the light, the predetermined matching criteria need to be based on a large number of real vehicle test data, such as the light change scene of entering the shadow area at different vehicle speeds, recording the best brightness adjustment parameters to form the criteria library. The strategy of feature redistribution needs to consider the importance level of display elements, such as the priority of brightness adjustment of navigation arrows is higher than that of mileage information.

[0076] In the rule allocation, the weight setting of the brightness rule dynamic adjustment method needs to be determined through driver subjective evaluation experiment, such as inviting 20 drivers to score the HUD display effect under different weights, and selecting the weight combination with the highest score. The upper and lower limit setting of node rule needs to consider the display effect and hardware limitation, avoiding brightness adjustment beyond the physical output range of HUD.

[0077] In the ambient light algorithm of error evaluation, the setting of data point weight needs to consider the nearness of sampling time, and the weight of recent sampling points is higher than that of historical sampling points. In the feature processing, principal component analysis method is used to reduce data dimension and improve calculation efficiency.

[0078] In the parameter adjustment submodule of feedback optimization, the setting of adjustment amplitude needs to follow the gradient descent principle to avoid model oscillation caused by too large adjustment amplitude. The verification threshold of optimization control submodule needs to be set according to the actual demand of vehicle-mounted system, such as the brightness error threshold of civilian vehicles can be set to ±5%, while special vehicles may require higher accuracy of ±2%.

[0079] Through the coordinated operation of these submodules, the brightness calculation module can combine real-time light data and model parameters to accurately calculate the display brightness of each region of HUD, and the error evaluation and feedback optimization module can continuously optimize system parameters to ensure that the brightness adjustment of HUD can meet the visual comfort of drivers and ensure the clarity and visibility of driving information in complex light environments (such as day and night alternation, tunnel group running, rain glare, etc.). The processing delay of the whole module needs to be controlled within 30ms to adapt to the real-time demand of vehicle high-speed running, and incremental learning algorithm is used to avoid retraining all models every time optimization is performed, reducing the consumption of computing resources.

[0080] It should be noted that in this text, relationship terms such as first and second are only used to distinguish one entity or operation from another, and do not necessarily require or imply any actual relationship or order between the entities or operations. Moreover, the term "includes" "contains" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment.

[0081] While embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and variations can be made to these embodiments without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A vehicle-mounted HUD ambient light adaptive brightness adjustment system, characterized in that, The system includes: The illumination acquisition module analyzes the stability of illumination parameters and the difference in interference based on the ambient illumination intensity acquisition range and interference light intensity value of the vehicle HUD, calculates the difference in illumination perception between environments, integrates them into a set of illumination perception parameters, and obtains the illumination perception state value. The parameter analysis module extracts the characteristic combination of HUD display brightness and ambient light based on the illumination perception state value, filters the optimal brightness and light intensity combination, and obtains the parameter analysis parameter set. The model building module, based on the parameter analysis parameter set, extracts the changes in the current model training samples, analyzes the relationship between brightness and light intensity, matches sample features with generated combinations, and obtains the model building parameter set. The brightness calculation module constructs a parameter set based on the model, extracts the current HUD display brightness feature value, combines it with real-time illumination data, maps the features to brightness, sets calculation rules, and applies the rules to the matching of illumination features to obtain the brightness calculation threshold. The error assessment module, based on the brightness calculation threshold, captures the feature data of the brightness sampling points, combines the ambient light to infer the feature changes of the sampling points, analyzes the feature change trend corresponding to the brightness calculation, classifies and organizes the feature change trend according to the inference results, and combines the feature change information to adjust and analyze the brightness features of the classified data to obtain the error distribution assessment value. The calibration correction module, based on the error distribution evaluation value, analyzes the error values ​​of features and brightness through real-time features and calculated data, and obtains a brightness calibration correction scheme by combining the error value adjustment calculation rules. The threshold setting module, based on the brightness calibration correction scheme, extracts the calibrated brightness parameters, analyzes the parameter fluctuation range, sets the brightness threshold interval and applies it to the error distribution evaluation to obtain the brightness threshold parameter set. The result output module integrates the brightness calculation results based on the brightness threshold parameter set, performs data formatting processing according to the HUD display mode, and generates standardized output data. The feedback optimization module, based on the standardized output data, collects system operation feedback information, analyzes the sources of brightness error, adjusts model training parameters and parameter analysis rules, and obtains a set of optimized system parameters. The system control module generates an ambient light adaptive HUD brightness adjustment scheme based on the system optimization parameter set.

2. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The illumination perception state value includes a perturbation light intensity parameter set, a delay response parameter set, and a perception difference parameter set. The parameter analysis parameter set includes brightness feature parameters, light intensity feature parameters, and combination screening parameters. The model construction parameter set includes sample change parameters, feature matching parameters, and training adjustment parameters. The brightness calculation threshold includes feature change parameters, illumination matching parameters, and rule setting parameters. The error distribution evaluation value includes feature trend analysis parameters and brightness feature relationship parameters. The brightness calibration correction scheme includes error analysis parameters and rule adjustment parameters. The brightness threshold parameter set includes parameter fluctuation parameters and interval setting parameters. The standardized output data includes mode integration parameters and format processing parameters. The system optimization parameter set includes feedback analysis parameters and model adjustment parameters. The ambient light adaptive HUD brightness adjustment and control scheme includes configuration synchronization parameters and process coordination parameters.

3. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The light acquisition module includes: The data acquisition submodule, based on the ambient light intensity acquisition range and interference light intensity value of the vehicle HUD, performs regional acquisition of stability value and interference value, locates invalid light data, removes abnormal interference data, and arranges the extracted stability value and interference value in regional order to generate a light parameter interference dataset. The intensity analysis submodule analyzes the stability and interference between environments based on the light parameter interference dataset, calculates the light parameter change ratio, sorts the differences between environments by weight, marks areas with excessive fluctuation differences, and obtains light perception difference data. The distribution integration submodule, based on the light perception difference data, calls the light perception difference values ​​for multi-dimensional aggregation, screens the differences in light perception values, classifies them according to the magnitude of the values, and arranges the light perception values ​​in an orderly manner to generate light perception status values.

4. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The parameter analysis module includes: The parameter extraction submodule identifies the display characteristics of each HUD area based on the illumination perception state value, records the light intensity and adjustment data of the brightness, normalizes the recorded data, sorts the normalized data according to brightness and light intensity, and generates a feature parameter dataset. The feature optimization submodule analyzes the brightness and light intensity values ​​in the feature parameter dataset, filters parameter combinations with high matching degree with the perception state, adjusts the parameter combinations and generates parameter combination optimization results by feature matching and recording the matching results. The parameter selection submodule retrieves the optimization results of the parameter combination, determines the optimal brightness and light intensity combination with matching degree, adjusts the parameter analysis parameters, inputs the generation configuration, verifies the stability of the parameter set, and generates the parameter analysis parameter set.

5. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The model building module includes: The sample analysis submodule, based on the parameter analysis parameter set, collects sample data through the training library, including brightness deviation, light intensity distortion and stray light effect, performs time series analysis on the data, removes abnormal samples, partitions the remaining data, and obtains sample analysis data. The parameter matching submodule analyzes the sample analysis data to analyze the impact of sample variables on brightness extraction and light intensity extraction, calculates the degree of influence of changes in each sample feature on parameter adjustment, determines the optimal matching parameter settings based on the impact score, iteratively adjusts the parameters to capture the optimal combination, and obtains the parameter matching results. The training integration submodule selects a combination of brightness and light intensity that matches the current sample conditions from the parameter docking results, conducts parameter training experiments, optimizes parameter settings through multiple training and verifications, determines and solidifies the parameters as model standards, and generates a model construction parameter set.

6. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The brightness calculation module includes: The feature extraction submodule constructs a parameter set based on the model, locates brightness feature monitoring points, extracts feature change values ​​in the monitoring area, continuously records feature increase / decrease rates, extracts multiple key change nodes corresponding to the change rate, sorts the node values ​​in order, and obtains the current feature change feature value. The illumination matching submodule analyzes the node change value and real-time illumination data based on the current feature change feature value, performs calibration according to a predetermined matching criterion, calls the matching criterion to perform feature redistribution within the illumination interval, and obtains the feature illumination matching structure. The rule allocation submodule, based on the feature illumination matching structure, adopts a brightness rule dynamic adjustment method to calculate the distribution of features among changing nodes, sets upper and lower limits for node rules, applies the rules to the regional feature values, allocates them, and obtains the brightness calculation threshold.

7. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 6, characterized in that: The brightness rule dynamic adjustment method includes obtaining the allocation rule value of matching feature change nodes, and setting the node rule lower limit through feature weight, illumination weight and dynamic adjustment weight to control minimum matching.

8. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The error assessment module includes: The feature data capture submodule, based on the brightness calculation threshold, applies an ambient lighting algorithm to capture feature data of sampling points, removes outliers and corrects errors, stores the data in a hierarchical manner according to intervals, performs feature processing, and generates a feature data dataset. The brightness analysis submodule, based on the characteristic data dataset, divides the intervals according to brightness categories, extracts the changing trends and fluctuation features, and generates a brightness category and feature change feature set; The classification inference submodule, based on the brightness category and feature change feature set, adjusts feature parameters and calibrates trend data, extracts classification intervals, and performs numerical prediction to obtain an error distribution evaluation value.

9. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 8, characterized in that: The ambient lighting algorithm includes calculating feature values, generating a feature dataset, and comprehensively calculating the feature values ​​after feature generation by combining data point weights, original feature values ​​of sampling points, lighting values, and category coefficients.

10. The vehicle-mounted HUD ambient light adaptive brightness adjustment system according to claim 1, characterized in that: The feedback optimization module includes: The data analysis submodule extracts real-time features and calculated data based on the standardized output data, analyzes the real-time brightness value and calculated value, and correlates the brightness difference with the current feature information to generate a feature brightness error value. The parameter adjustment submodule sets the training adjustment parameters of the model based on the feature brightness error value. It sets the adjustment range for brightness categories with large errors and fine-tunes the categories with low errors. By comparing the adjustment effects, it selects and integrates the matching parameter set to generate the model adjustment parameter set. The optimization control submodule adjusts the parameter set based on the model, applies the adjusted parameters in each training stage, implements model optimization item by item, monitors features and brightness synchronously, gradually adjusts the optimization order of each brightness category model, and generates a system optimization parameter set.

Images